www.elsevier.com/locate/ecoenv Ecotoxicology and Environmental Safety

Editor Kurunthachalam Kannan Wadsworth Center, New York State Department of Health and Department of Environmental Health Sciences, State University of New York at Albany, New York, USA

Associate Editors Richard D. Handy Paul Sibley Hyo-Bang Moon University of Plymouth School of Environmental Sciences Associate Professor Department of Biology University of Guelph Marine Environment Analytical Plymouth, Guelph, Ontario Laboratory (MEAL) Devon PL4 8AA N1G 2W1 Dept. of Marine Sci. and United Kingdom Canada Convergent Tech., Hanyang University, Korea Editorial Board

Shashi Bhushan Agrawal Taisen Iguchi Ann Miracle Irena Twardowska Fernando Barbosa Jr Hisato Iwata Haruhiko Nakata Paule Vasseur Swaran J. S. Flora D. Johnson Jae-Sung Rhee Wen-Xiong Wang C.A.M. van Gestel Rai S. Kookana Susan Shaw Po-Keung Wong Helmut Greim Yi-Fan Li Louis A. Tremblay Tao Zhang Ecotoxicology and Environmental Safety 146 (2017) 1–3

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Editorial ff Following the : What we know about the e ects of MARK oil on birds?

The April 20, 2010 explosion and subsequent sinking of the Deepwater Horizon (DWH) mobile drilling unit in Mississippi Canyon Block 252 (MC252) resulted in 11 fatalities and the release of an unprecedented volume of South Louisiana sweet crude oil (507 million liters) into the Gulf of Mexico (GOM). This oil covered an area of 112,100 square kilometers, including more than 10% of the GOM shoreline. The Natural Resource Damage Assessment (NRDA) Trustee Council for this spill, which included the US Department of Interior, US Environmental Protection Agency, Department of Commerce, Department of Agriculture and the five affected states, conducted injury assessment activities from 2010 to 2015. Included as one of these injury assessments was the development of an avian toxicity testing program developed by the US Fish and Wildlife Service (USFWS) as part of their avian injury assessment program (for the full funding application see: https://www.doi.gov/sites/doi.gov/files/migrated/deepwaterhorizon/upload/AvianToxicity.pdf. Unlike many previous oil spills, the DWH event occurred for an extended period in warm waters resulting in both bird deaths and large numbers of birds that were observed alive. The Live Oiled Bird Model (LOBM) was developed by USFWS as a means of estimating injury to those groups of birds. One of the primary requirements for this model was a determination of the “fate” of the oiled birds. That is, the likelihood a bird would suffer adverse effects from oiling or die. Unfortunately, while there is was an extensive body of literature available from which to draw conclusions on the effects of heavy oiling on birds, the literature available for assessment of the adverse outcomes for birds with light to moderate oil coverage was largely unquantified prior to DWH. As part of their injury assessment, USFWS initiated the “Blood Physiology Study” (https://www.doi.gov/sites/doi.gov/files/migrated/ deepwaterhorizon/upload/Final-DOI-NRDA569AP-1-July-2011-2.pdf) to determine if there was a link between lighter oiling categories and one of the better understood toxic effects of ingested oil, hemolytic anemia. In an important migratory stopover point such as the GOM, this particular sublethal endpoint could have broad adverse impacts on flight ability and migratory success. Although limited in scope, the “Blood Physiology Study” did suggest that birds collected from oiled areas were suffering from hemolytic anemia in the form of reduced packed cell volume, and increased incidence of reticulocytes and Heinz bodies, regardless of whether oil was visible on their feathers or not. This study could not provide a definitive link between exposure to MC252 oil and toxicity. As a result, the USFWS commissioned a series of expert panels to provide input on planning a series of experimental studies that would accurately characterize the toxic effects of MC252 oil on birds. Studies were planned that would attempt to fill some of the gaps in our understanding of oil toxicity to birds and the adverse outcomes to flight abilities (refer to https://www.doi.gov/sites/doi.gov/files/migrated/deepwaterhorizon/upload/AvianToxicity.pdf for full study plans). In brief these studies were divided into oil ingestion studies that would attempt to produce comprehensive dose-response relationships between oil ingestion rates and the development of sublethal adverse outcomes, and studies that would explore the nature of the effects of oil on flight and thermo- regulation. The initial goals of these studies were very broad due to the unknown experimental factors such as the exact nature of the toxicity of MC252 oil, the specific toxicological responses of the species involved and the more general issue of dose delivery of a noxious substance that requires ingestion. As these studies were part of the larger NRDA being undertaken by USFWS, they were conducted within the limitations of litigation sensitive work, including stringent quality assurance/quality control (QA/QC) and sample and data retention, which unfortunately somewhat limited the number of experiments that could be conducted within the time frame for the NRDA. However, the benefits of such work include access to samples and data for the larger scientific community. Nonetheless, these studies were able to build on previous oil toxicity studies to develop dosing methods that will be of great use to future studies of oil toxicity, and provide a more in-depth understanding of the adverse outcomes on avian physiology and flight behavior. The first set of studies suggested by the expert panels were pilot oral dosing studies to determine if methods described in the literature were applicable to the species available/chosen for study. Earlier oral ingestion studies were either conducted through gavage of juvenile or subadult birds for short periods of time or through feeding trials for longer duration studies. In general, accurate dosing was difficult due to issues with regur- gitation or rapid elimination. These same issues were experienced with the oral dosing studies designed for the DWH avian toxicity testing program. All four species tested in the pilot oral dosing study (Dean et al., 2017a) either regurgitated the oil, as in the homing pigeon, or displayed rapid defecation (western sandpiper, laughing gull and double-crested cormorant). Despite the challenge of accurate dosing during these studies, there were changes in blood chemistry and oxidative stress markers in the double-crested cormorant that indicated early stages of oil toxicity. De- termination of hemolytic anemia was somewhat hampered by a lack of agreement between slide readers on the presence and number of Heinz bodies present in the samples collected, resulting in a shift to the use of electron microscopy for definitive proof of Heinz body identification for each species. Further modifications were made to dosing methods for the western sandpiper, laughing gull and double-crested cormorant to determine if http://dx.doi.org/10.1016/j.ecoenv.2017.08.068

Available online 09 September 2017 0147-6513/ © 2017 Published by Elsevier Inc. Editorial Ecotoxicology and Environmental Safety 146 (2017) 1–3 better oral dose delivery could be achieved. Oral dose delivery method development was not continued because the field flight work showed that a single application of oil to feathers caused irreversible damage to feather structure in this species, making investigation of changes in flight performance without toxicity a priority for the NRDA, as some migratory birds would have experienced similar light oiling during stopover. Western sandpiper dosing methods could not be changed because the only additional option for dosing was to inject oil into meal worms, and this resulted in significant leakage of oil from the meal worms. Instead the method was altered to reduce daily dose amount and increase the period of exposure. Once again there were few changes in standard plasma biochemistry measurements, but due to the success of the pigeon flight work, a switch was made to focus on the effects of oil applied to feathers on flight and thermoregulation in western sandpipers. Further method development for direct oral dosing of laughing gulls was achieved through injection of oil into the body cavity of dead feeder fish that made up part of the captive diet (Horak et al., 2017). Laughing gulls, like many other species of gulls, can take advantage of variations in food sources, and as such they are adept at discerning the quality of food resources. In our case this meant that if oil was detected by the birds, through visual, olfactory or taste cues, the birds were unlikely to consume the treated fish. This presented a considerable challenge for the scientists and technical staff tasked with dosing, and while they were able to achieve dosing rates comparable to the nominal doses, it was not without intensive work on the part of the researchers, making the dosing method in its current state very difficult to replicate for this species. The study was successful in that the direction of the endpoint changes observed for laughing gulls was similar to that observed in the double-crested cormorants, but to a lesser extent. Further method development for oral dosing is required in this species to determine if results were confounded due to dose delivery, or if this particular species is not as sensitive to the effects of oil. In the case of the double-crested cormorants, fingerling catfish readily survived injection of oil into the body cavity under anaesthetic (Cunningham et al., 2017). These fish were consumed normally by the cormorants throughout the course of the day, achieving dose rates similar to nominal doses. Dosing at the higher rate of 10 ml/kg body weight/day resulted in anemia by day 7 and oil intoxication by day 14 of dosing, indicating that this method of oral dosing was suitable for this species. This is not to say that further method development is not required, and the authors of the studies are the first to admit that while this method was largely successful in terms of defining measurable endpoints of oil toxicity for this species that there were other factors that likely contributed to the substantial toxicity observed. First, these birds were housed with large water tanks that they often defecated into, resulting in oil being visible on feathers. This oil may have been preened, resulting in re-consumption, and was also likely to constitute a thermoregulatory stressor. Second, as dosing continued cormorants began to show a food aversion, and regardless of whether this was a taste aversion or caused by the toxicity of the oil, it resulted in considerable weight loss and exacerbated the loss of condition. Nonetheless, this method, used for dosing in this species, was very successful in determining the potential effects of MC252 oil on birds, and also in re-defining oil toxicity in birds. We were able to observe changes in liver and kidney function, measure the development of oxidative stress in the liver, expand our understanding of CYP upregulation, and accurately define hemolytic anemia through the use of electron microscopy (Alexander et al., 2017; Dean et al., 2017b; Harr et al., 2017a, 2017b, 2017c; Pritsos et al., 2017) While there were some threads of evidence that were outside of the purview of these NRDA-based experiments, such as expanding our knowledge of how oil affects adrenal and immune function, there were indications in the form of changes in adrenal histology, changes in gastrointestinal parasite burdens, and increases in white blood cell counts and gamma globulin concentrations that suggest further in depth study is warranted. Additionally, we were able to observe changes in function for both the laughing gulls and double-crested cormorants that had not yet been reported in birds. These included extending blood clotting time, “flaccid” appearance of cardiac muscle and damage to red blood cell organelles. Unfortunately constraints placed on the research by the NRDA process meant that there were no further direct oral dosing studies undertaken, so dose-response relationships could not be developed for MC252 oil. In the view of the editors, this is one of the unfortunate limitations of this work, that the original plans for these studies were not able to be followed through to completion. If the original study plans outlined by the expert panels were followed, this would have represented a significant leap forward in understanding the relationship between oil dose and toxicity in birds that could have been of great benefit to all other oil spills in the future. Data from such studies would aid in clarifying NRDA claims effectively and in a more timely fashion, reduce or target any research required for the specific type of oil involved in the spill, and provide a baseline suite of measures to assess the level of injury to an individual bird, and outline a course of action for appropriate treatment. Instead, an additional pilot study was undertaken on double-crested cormorants, this time using application of oil to feathers as an indirect oral dosing mechanism, using the assumption that all birds will preen to maintain feather integrity. This particular study incorporated investigations of the cardiac effects of oil as well as potential thermoregulatory effects; however, it was limited to only a single dose group. The study was exceedingly successful, showing changes in many of the standard clinical plasma measurements investigated, as well as the development of an echocardigram technique that was able to show birds likely suffer from a dilated cardiomyopathy, and that despite no changes in body weight or loss of ther- moregulatory ability, food consumption increased (Cunningham et al., 2017; Dean et al., 2017b; Harr et al., 2017). The impact of reduced cardiac function combined with hemolytic anemia and thermoregulatory stress is likely to have deleterious effects, particularly in populations living in oiled areas where food supplies dwindle, as birds will need more access to food, but potentially be unable to access it due to reduced capacity. These birds were also observed to have histological changes in organs such as liver and kidneys indicative of oil toxicity. Once again, the suite of endpoints investigated was somewhat limited by the NRDA process, so changes in parasite burden, clotting ability, and cardiac markers such as troponin could not be investigated to any great extent. Also, although the dosing was successful, in that oil was clearly preened from feathers, dose could not be accurately determined. The estimated application every three days to provide a “moderate” (20–40%) body coverage was not correct for the duration of the study, and by day 9, and by day 9 most of the feather surface was visibly covered in oil, indicating a “heavy” (> 40%) oiling (Cunningham et al., 2017). While this is unlikely to change the rate of oil consumption by the birds, it caused skin irritation resulting in feather plucking (Cunningham et al., 2017), and is likely to increase thermoregulatory stresses further. Although it was not possible within the confines of these studies to investigate the interaction between oil toxicity and flight ability, we obtained some excellent data on what might be happening to birds that experience a single oiling event. The work conducted on the western sandpipers and homing pigeons was coordinated between the two laboratories undertaking the work to ensure that there was as much overlap as possible. The “low” (< 20%) levels of oil application to the feathers on wings and tail were scaled for species, but oil was applied in the same way. Following training for either test, birds were oiled immediately prior to testing, with no time for oil ingestion to occur, and after the single oiling event, birds were flown again without additional oiling. One of the interesting differences observed between the two species that should be noted for those interested in conducting future studies was that in the two week period between application of oil on western sandpipers and subsequent flight trial, the birds were no longer visibly oiled; however, pigeons were still clearly visibly oiled. This may be a species-specificdifference in preening ability or type of preening involved. Western sandpipers had water readily available for preening, while the pigeons had access to dirt for dust bathing while in their loft, but could have had access to water or dust for bathing when stopped during homing flights.

2 Editorial Ecotoxicology and Environmental Safety 146 (2017) 1–3

While not published in this issue, a wind tunnel study representative of migratory flight (Maggini et al., 2017b) concluded that immediately after oil application birds with light to moderate (~ 30%) oiling on wings, tail and body would show reductions in flight speed, increases in drag, and 22% and 45% increase in flight costs respectively for oiled birds. Further, Maggini et al. (2017a) in this issue showed that takeoff of these birds was also impaired, taking off slower and at a shallower angle compared to the unoiled birds. The homing pigeon work provided further insight into con- sequences of this oiling and the adaptions that the birds are able to make. Perez et al. (2017a, 2017b, 2017c) were able to show that homing pigeons following a single oil application to feathers continued to show impairments in flight performance after multiple flights. These impairments included longer times to return home, more frequent stops, changes in flight path and inability to return body weight to normal within the same period of time as the control birds. There were some very interesting similarities between the homing pigeon and the western sandpiper flight work that warrant further consideration, including the impact of altered flight path and higher flight cost on migration. It should be noted that these impacts occurred without any obvious signs of oil toxicity having occurred by the end of the studies. All of these birds underwent blood sampling and necropsy at the conclusion of the studies, and there was no evidence of oil toxicity. Further study designs may seek to incorporate oil ingestion prior to flight testing to determine the full effects of oil on flight, particularly in light of the double-crested cormorant echocardiogram results, and anemia development, that would impose a physiological burden on flight ability. These studies also did not consider the impacts of oiled feathers or oil toxicity on diving ability in birds; however, the literature investigations of oil resulting in loss of buoyancy and thermoregulatory ability likely indicate that diving costs could increase as well. The studies presented in this issue could not have been undertaken without the foresight of Drs. Kim Trust and Mike Hooper who understood that there were large gaps in the literature that needed to be addressed in light of the DWH oil spill. They sought to put together a comprehensive test program that addressed the vast number of birds in the Gulf of Mexico that could have been affected by oil. Unfortunately, “the best laid plans” are often side tracked or slowed by the nature of research itself, particularly within the confines of NRDA processes, so we were unable to complete all of the work as planned. However, the research presented here did benefit from the extensive planning and input of all of the individual experts who contributed to the study planning from the outset. Without them it is unlikely that we would have been able to investigate the metabolic demands of flight imposed by oiled feathers, or discovered that the effects of oil on avian red blood cells goes beyond formation of Heinz bodies, that oil can cause cardiac damage to adult birds and that blood clotting deficiencies occur. It is our hope that the research presented in this issue will provide a stepping-stone for more comprehensive studies to elucidate how oil has these effects, and what can be done to prevent or mitigate those effects. Finally, we would like to thank all of the contributors to this special issue for their efforts in preparing these manuscripts, and their desire to see their work published in a cohesive manner that will benefit future research into the effects of oil on birds.

References

Alexander, C., Hooper, M.J., Cacela, D., Smelker, K.D., Calvin, C.S., Dean, K.M., Bursian, S.J., Cunningham, F.L., Hanson-Dorr, K.C., Horak, K.E., Isanhart, J.P., Link, J.E., Shriner, S.A., Godard-Codding, C.A., 2017. CYP1A protein expression and catalytic activity in double-crested cormorants experimentally exposed to Deepwater Horizon Mississippi Canyon 252 oil. Ecotoxicol. Environ. Saf. 142, 79–86. Cunningham, F.L., Dean, K.M., Hanson-Dorr, K.C., Harr, K.E., Healy, K.A., Horak, K.E., Shriner, S.A., Bursian, S.J., Dorr, B.S., 2017. Development of methods for avian oil toxicity studies using the double crested cormorant (Phalacrocorax auritus). Ecotoxicol. Environ. Saf. 141, 199–208. Dean, K.M., Bursian, S.J., Cacela, D., Carney, M.W., Cunningham, F.L., Dorr, B.S., Hanson-Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Lipton, I., McFadden, A.K., McKernan, M.A. and Harr, K.E. 2017a. Changes in White Cell Estimates and Plasma Chemistry Measurements following Oral Or External Dosing of Double-crested Cormorants, Phalacrocorax auritus, with Artificially Weathered MC252 Oil. Dean, K.M., Cacela, D., Carney, M.W., Cunningham, F.L., Ellis, C., Gerson, A.R., Guglielmo, C.G., Hanson-Dorr, K.C., Harr, K.E., Healy, K.A., Horak, K.E., Isanhart, J.P., Kennedy, L.V., Link, J.E., Lipton, I., McFadden, A.K., Moye, J.K., Perez, C.R., Pritsos,C.A., Pritsos, K.L., Muthumalage, T., Shriner, S.A. and Bursian, S.J. 2017b. Testing of an Oral Dosing Technique for Double-Crested Cormorant, Phalacrocorax auritus, Laughing Gull, Leucophaeus atricilla, Homing Pigeon, Columba livia, and Western Sandpiper, Calidris mauri, with Artificially Weather MC252 Oil. Harr, K.E., Cunningham, F.L., Pritsos, C.A., Pritsos, K.L., Muthumalage, T., Dorr, B.S., Horak, K.E., Hanson-Dorr, K.C., Dean, K.M., Cacela, D., Link, J.E., Healy, K., Tuttle, P., Bursian, S.J., 2017a. Weathered MC252 crude oil-induced anemia and abnormal erythroid morphology in double-crested cormorants (Phalacrocorax auritus) with light microscopic and ultra- structural description of Heinz bodies. Ecotoxicol. Environ. Saf (submitted for publication). Harr, K.E., Reavill, D.R., Bursian, S.J., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Shriner, S.A., Schmidt, R.E., 2017b. Organ weights and histopathology of double-crested cormorants (Phalacrocorax auritus) dosed orally or dermally with artificially weathered Mississippi Canyon 252 crude oil. Ecotoxicol. Environ. Saf (In press), (Citation to be finalized during editorial review). Harr, K.E., Rishniw, M., Rupp, T.L., Cacela, D., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C., Healy, K., Horak, K., Link, J.E., Reavill, D., Bursian, S.J., Cunningham, F.L., 2017c. Dermal exposure to weathered MC252 crude oil results in echocardigraphically identifiable systolic myocardial dysfunction in double crested cormorants (Phalacrocorax auritus). Ecotoxicol. Environ. Saf (In press). Horak, K.E., Bursian, S.J., Ellis, C., Dean, K., Link, J., Hanson-Dorr, K., Cunningham, F., Harr, K., Pritsos, C., Pritsos, K., Healy, K., Cacela, D., Shriner, S., 2017. Toxic Effects of Orally Ingested Oil from the Deepwater Horizon Spill on Laughing Gulls (Submitted for publication EES-16-1418). Maggini, I., Kennedy, L.V., Elliott, K.H., Dean, K.M., MacCurdy, R., Macmillan, A., Pritsos, C.A., Guglielmo, C.G., 2017a. Trouble on Takeoff: Crude Oil on Feathers Reduces Escape Performance of Shorebirds. Maggini, I., Kennedy, L.V., Macmillan, A., Elliot, K.H., Dean, K., Guglielmo, C.G., 2017b. Light oiling of feathers increase flight energy expenditure in a migratory shorebird. J. Exp. Biol. 220, 2372–2379. http://jeb.biologists.org/content/220/13/2372.full-text.pdf. Perez, C., Moye, J., Cacela, D., Dean, K., Pritsos, C.A. 2017a. Body Mass Change in Flying Homing Pigeons Externally Exposed to Deepwater Horizon Crude Oil. Perez, C.R., Moye, J.K., Cacela, D., Dean, K.M., Pritsos, C.A. 2017b. Low Level Exposure to Crude Oil Impacts Avian Flight Performance: The Deepwater Horizon Oil Spill Effect on Migratory Birds. Perez, C.R., Moye, J., Cacela, D., Dean, K., Pritsos, C.A., 2017c. Homing pigeons externally exposed to Deepwater Horizon Crude Oil Change Flight Performance and behavior. Environ. Pollut. 230, 530–539. http://dx.doi.org/10.1016/j.envpol.2017.07.008. Pritsos, K.L., Perez, C.R., Muthumalage, T., Dean, K.M., Cacela, D., Hanson-Dorr, K., Cunningham, F.L., Bursian, S.J., Link, J.E., Shriner, S.A., Horak, K.E., Pritsos, C.A., 2017. Dietary intake of Deepwater Horizon Oil-injected Live Food Fish by Double-crested Cormorants Resulted in Oxidative Stress. (Submitted for publication EES-16-1417).

Karen M. Dean Department of Neuroscience, University of Lethbridge, Lethbridge, AB, Canada Steven J. Bursian Department of Animal Science, Michigan State University, East Lansing, MI, USA

3 Ecotoxicology and Environmental Safety 146 (2017) 4–10

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Reprint of: Overview of avian toxicity studies for the Deepwater Horizon MARK Natural Resource Damage Assessment ⁎ S.J. Bursiana, , C.R. Alexanderb, D. Cacelac, F.L. Cunninghamd, K.M. Deanc, B.S. Dorrd, C.K. Ellise, C.A. Godard-Coddingb, C.G. Guglielmof, K.C. Hanson-Dorrd, K.E. Harrg, K.A. Healyh, M.J. Hooperi, K.E. Horake, J.P. Isanhartj, L.V. Kennedyf, J.E. Linka, I. Magginif, J.K. Moyek, C.R. Perezk, C.A. Pritsosk, S.A. Shrinere, K.A. Trustl, P.L. Tuttleh a Department of Animal Science, Michigan State University, East Lansing, MI, USA b The Institute of Environmental and Human Health, Texas Tech University, Lubbock, TX, USA c Abt Associates, Boulder, CO, USA d US Department of Agriculture, Wildlife Services, Mississippi Field Station, Mississippi State University, Starkville, MS, USA e US Department of Agriculture, Wildlife Services, Fort Collins, CO, USA f Department of Biology, Advanced Facility for Avian Research, University of Western Ontario, London, ON Canada g URIKA, LLC, Mukilteo, WA, USA h US Fish and Wildlife Service, Deepwater Horizon Natural Resource Damage Assessment and Restoration Office, Fairhope, AL, USA i US Geological Survey, Columbia Environmental Research Center, Columbia, MO, USA j US Department of the Interior, Denver, CO, USA k Department of Agriculture, Nutrition and Veterinary Sciences, University of Nevada, Reno, Reno, NV, USA l US Fish and Wildlife Service, National Wildlife Refuge System, Portland, OR, USA

ARTICLE INFO ABSTRACT

Keywords: The Oil Pollution Act of 1990 establishes liability for injuries to natural resources because of the release or threat Deepwater Horizon oil spill of release of oil. Assessment of injury to natural resources resulting from an oil spill and development and Avian toxicity studies implementation of a plan for the restoration, rehabilitation, replacement or acquisition of natural resources to Natural Resource Damage Assessment compensate for those injuries is accomplished through the Natural Resource Damage Assessment (NRDA) Oil toxicity process. The NRDA process began within a week of the Deepwater Horizon oil spill, which occurred on April 20, 2010. During the spill, more than 8500 dead and impaired birds representing at least 93 avian species were collected. In addition, there were more than 3500 birds observed to be visibly oiled. While information in the literature at the time helped to identify some of the effects of oil on birds, it was not sufficient to fully characterize the nature and extent of the injuries to the thousands of live oiled birds, or to quantify those injuries in terms of effects on bird viability. As a result, the US Fish and Wildlife Service proposed various assessment activities to inform NRDA injury determination and quantification analyses associated with the Deepwater Horizon oil spill, including avian toxicity studies. The goal of these studies was to evaluate the effects of oral exposure to 1–20 ml of artificially weathered Mississippi Canyon 252 oil kg bw-1 day-1 from one to 28 days or one to five applications of oil to 20% of the bird's surface area. It was thought that these exposure levels would not result in immediate or short-term mortality but might result in physiological effects that ultimately could affect avian survival, reproduction and health. These studies included oral dosing studies, an external dosing study, metabolic and flight performance studies and field-based flight studies. Results of these studies indicated changes in hematologic endpoints including formation of Heinz bodies and changes in cell counts. There were also effects on multiple organ systems, cardiac function and oxidative status. External oiling affected flight patterns and time spent during flight tasks indicating that migration may be affected by short-term repeated exposure to oil. Feather damage also resulted in increased heat loss and energetic demands. The papers in this special issue indicate that the combined effects of oil toxicity and feather effects in avian species, even in the case of relatively light oiling, can significantly affect the overall health of birds.

DOI of original article: http://dx.doi.org/10.1016/j.ecoenv.2017.03.046 ⁎ Corresponding author. E-mail address: [email protected] (S.J. Bursian). http://dx.doi.org/10.1016/j.ecoenv.2017.05.014 Received 13 July 2016; Received in revised form 24 March 2017; Accepted 28 March 2017 Available online 27 May 2017 0147-6513/ © 2017 Published by Elsevier Inc. S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 4–10

1. Introduction provide for the opportunity to settle damage claims without litigation (Hagerty and Ramseur, 2010; Barron, 2012; Vann and Meltz, 2013a, b; The Deepwater Horizon (DWH) oil spill that began on April 20, Deepwater Horizon Natural Resource Damage Assessment Trustees, 2010 initiated a Natural Resource Damage Assessment (NRDA) process 2016). under the Oil Pollution Act of 1990. As part of the DWH NRDA, a suite of avian toxicity studies was designed and implemented to inform 4. DWH oil spill NRDA NRDA injury determination and quantification analyses associated with the DWH oil spill. The purpose of the present paper in this special issue The NRDA process, as described above, began a week after the DWH of Ecotoxicology and Environmental Safety is to present a brief oil spill. On April 20, 2010, the DWH mobile drilling unit located in the description of the NRDA process under the Oil Pollution Act, general northern Gulf of Mexico (GOM) 66 km off the Louisiana coast in information about the DWH NRDA, a summary of DWH NRDA activities Mississippi Canyon 252 exploded, caught fire, and eventually sank, associated with the assessment of immediate effects of the spill on birds, resulting in a massive release of oil and other substances from British a brief summary of the literature as it relates to the toxicity of oil in 's (BP) Macondo well. The explosion and fire killed 11 avian species, and a description of the avian toxicity studies that were workers and injured 17 others. Initial efforts to cap the well following conducted to inform the DWH NRDA. The other papers that comprise the explosion were unsuccessful, and the uncontrolled discharge of oil this special issue describe in greater detail the methodologies, specific and gas continued for 87 days. Approximately 3.19 million barrels (507 endpoints assessed, findings and significance of the data. million liters [l]) of South Louisiana sweet crude oil were released into the ocean creating the largest offshore oil spill in the history of the 2. Oil Pollution Act . Cumulatively, oil on the water surface extended over an area of more than 43,300 square miles (112,100 square kilometers The Oil Pollution Act of 1990 (OPA; Public Law 101–380 [33 U.S.C. [km]) and oiled more than 1300 miles (2100 km) of shoreline or more 2701 et seq.; 104 Stat. 484]) establishes liability for injuries to natural than 10% of the GOM shoreline off the coasts of Louisiana, Mississippi, resources because of the release or threat of release of oil into navigable Alabama and western Florida. Virtually all living resources and waters used for navigation, adjoining shorelines, and the exclusive supporting habitats in the northern GOM were adversely affected. economic zone. The parties responsible for causing the oil spill are The overall event became known as the DWH oil spill (Corn and responsible for these injuries to natural resources. Trustees, consisting Copeland, 2010; Hagerty and Ramseur, 2010; Barron, 2012; Carriger of federal, state and tribal government officials, as well as representa- et al., 2015; Deepwater Horizon Natural Resource Damage Assessment tives of foreign governments, may act on the behalf of the public to Trustees, 2016). assess injury to natural resources resulting from an oil spill and to In response to the DWH oil spill, the US Coast Guard, on April 28, develop and implement a plan for the restoration, rehabilitation, 2010, notified BP Exploration and Production and seven other entities replacement or acquisition of natural resources to compensate for those that they were identified as potential responsible parties (Vann and injuries through the Natural Resource Damage Assessment (NRDA) Meltz, 2013a, b). The Trustee Council for the DWH NRDA is comprised process (Barron, 2012; Vann and Meltz, 2013a, b; Deepwater Horizon of representatives from the US Department of the Interior, Department Natural Resource Damage Assessment Trustees, 2016). of Commerce, US Environmental Protection Agency, Department of Agriculture and the five Gulf states of Alabama, Florida, Louisiana, 3. NRDA process Mississippi and Texas. Shortly after the DWH oil spill, the pre- assessment phase was initiated by the Trustees. Impacts of the oil spill, There are three main phases that constitute an NRDA under the including visibly oiled and dead birds, sea turtles and marine mammals OPA: the pre-assessment phase, the restoration planning phase and across 2100 km of GOM shoreline, had been documented by August 19, restoration implementation. During the pre-assessment phase, Trustees 2010. Because of this information, as well as the duration of the oil determine if there is jurisdiction under OPA and if restoration of injured release, Trustees issued a Notice of Intent to Conduct Restoration natural resources is feasible. If the Trustees determine injuries have Planning and invited potential responsible parties to participate in the resulted or are likely to result from the incident and response actions NRDA with BP being the only potential responsible party agreeing to will not address or are not expected to address the injuries, the Trustees participate in the process. From 2010–2015, the Trustees conducted may prepare a Notice of Intent to Conduct Restoration Planning injury assessment activities (including data collection during the pre- Activities, which is made publicly available and sent to the responsible assessment process) as components of the second phase of the NRDA, party. The second phase in the NRDA process is the restoration planning restoration planning. During this period, Trustees developed and phase, which consists of the injury assessment step and the restoration implemented numerous studies to evaluate the impacts of released oil selection step. During the injury assessment step, Trustees determine on natural resources to aid them in restoration planning. Injury the nature and extent of injuries to natural resources and the benefits assessment began immediately after the explosion and it continued as they provide due to the release of oil by determining a pathway linking a multi-phased iterative approach such that planning and design the incident and the natural resources (pathway) and establishing decisions were based on data that had been collected and evaluated exposure to (exposure assessment) and subsequent injury of natural at the beginning of the spill (Deepwater Horizon Natural Resource resources (injury evaluation) by the released oil. One approach to Damage Assessment Trustees, 2016). quantify the degree and spatial and temporal aspects of the injuries is by comparing resources and associated benefits before and after the 5. DWH NRDA assessment of short-term impact on birds release of oil. In the restoration selection step, information from the injury assessment is used to develop various alternatives for primary or As mentioned above, one of the immediate impacts of the DWH oil compensatory restoration of natural resources and the benefits they spill was visibly oiled and dead birds. The various habitats of the provide. The Trustees then evaluate the restoration alternatives and northern GOM serve as important resources for over 150 species of choose a preferred alternative that typically becomes a draft restoration birds. Avian species may spend their entire life cycle in the northern plan (Draft Damage Assessment and Restoration Plan), which is made GOM, or they may use the area as a migratory stopover or wintering available for public comment and subsequently finalized. The third ground. Seabirds, colonial waterbirds, coastal marsh birds and shore- phase, restoration implementation, begins after Trustees have finalized birds using such an area are particularly susceptible to oil spills because a restoration plan. Often, the final restoration plan is submitted to the of their use of open water where oil tends to concentrate or because of responsible party for implementation or funding of the plan, which may high density along coastlines and marshes where extensive oiling can

5 S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 4–10 take place. During the spill, more than 8500 dead and impaired birds exposed sulfhydryl groups on hemoglobin. This damage results in representing at least 93 avian species were collected. In addition, there alterations to the tertiary structure of hemoglobin and causes precipi- were more than 3500 birds observed to be visibly oiled. It is tates to form that can coalesce into Heinz bodies. This damage is conservatively estimated that avian mortalities ranged from 51,600 to irreversible and the body can only compensate by producing more red 84,500 and that lost reproduction accounted for an additional 4600 to blood cells to account for the loss of hemoglobin. If compensation is 17,900 fledglings (Deepwater Horizon Natural Resource Damage incomplete, then the bird suffers from the potentially life-threatening Assessment Trustees, 2016). effects of reduced oxygen carrying capacity. Hemolytic anemia may be As the enormity of the DWH oil spill became apparent, the US Fish particularly important during migration, which is a time when meta- and Wildlife Service (FWS) convened a team of experts to design an bolic oxygen demands are extremely high and hemoglobin concentra- approach to assess potential injuries to resources for which the FWS has tions are elevated (Piersma et al., 1996). Controlled studies have shown trust responsibilities, including migratory birds. Initial steps included that hatchling and nestling herring gulls (Larus argentatus) can suffer the evaluation of pathways of transport and exposure, mechanisms of from these effects within four days after oral dosing with 5 ml kg body injury, and habitats and resources likely to be injured. Subsequent steps weight (bw)-1 day-1 (Leighton et al., 1985; Leighton, 1986). While these included the identification and design of approaches and models to studies provide some basic information on potential dosages, it estimate injury to avian resources. For birds, three primary approaches represents only a small step in understanding the development and were identified to estimate mortality: Shoreline Deposition Model progression of hemolytic anemia in all the birds exposed to MC252 oil. (SDM), Offshore Exposure Model (OEM) and Live Oiled Bird Model In the summer of 2010, as part of the NRDA pre-assessment, FWS (LOBM). The SDM uses information on the recovery of birds on initiated the Blood Physiology Study (US Department of the Interior, shorelines combined with information such as the deposition and 2011) to determine whether hemolytic anemia was a key diagnostic persistence of birds on shorelines, the ability of survey crews to locate feature in birds oiled during the DWH oil spill and if biomarkers of and recover birds, the deposition of carcasses under non-spill condi- hemolytic anemia as well as other physiological and biochemical tions, and other factors to estimate mortality across the affected area. indicators of avian health were consistently related to exposure of The LOBM was developed to estimate injury to birds that were oiled, birds to oil. In this field study, black skimmers (Rynchops niger), brown but did not quickly die and were not sufficiently incapacitated to enable pelicans (Pelecanus occidentalis), great egrets (Ardea alba), and clapper capture by Wildlife Operations personnel. At the most basic level, the rails (Rallus crepitans) were collected from oiled and unoiled areas and LOBM relies on three primary inputs: the numbers of birds occurring in their blood was analyzed for signs of hemolytic anemia (as well as other areas affected by the oil spill (abundance); the incidence and degree to endpoints). Approximately 35 birds of each species were collected from which birds are oiled (oiling rates); the fate of oiled birds (i.e., the oiled areas, and approximately 18 to 35 birds of each species were likelihood a bird would die or suffer other adverse effects due to oil). collected from unoiled areas. The data suggested that birds with small The Trustees implemented several studies to generate these types of amounts of visible oil on their feathers as well as birds collected from data for dominant bird groups, or guilds, affected by the oil spill. Field oiled areas with no visible oiling were experiencing hemolytic anemia surveys were successful in documenting abundance and oiling rates for based on decreased packed cell volume, increased reticulocytes and birds across the extent of the spill-affected area. However, due to a increased incidence of Heinz bodies (Fallon et al., 2014). The authors of variety of reasons, definitive data required for fate estimation were this report concluded that while a dose-response relationship between difficult to obtain. Consequently, the Trustees evaluated other ap- exposure to oil and the frequency of Heinz bodies was not established, proaches to assess the fate of externally oiled birds. Key in this this pathology could be used for injury assessments of birds following evaluation was information in the literature as well as data on oil- oil spills. However, the limited scope of this pre-assessment study related health effects in field collected birds from some of the DWH (hemolytic anemia and associated hematological measures) was NRDA pre-assessment studies (Deepwater Horizon Natural Resource thought insufficient to establish a definitive causal link between avian Damage Assessment Trustees, 2016). exposure to MC252 oil and injury. In addition, it was recognized that there were many other potential effects that could result from exposure 6. Toxic effects of oil on birds to oil at doses that induced hemolytic anemia. Understanding the greater spectrum of toxicological effects at these doses was necessary to While the acutely lethal effects of heavy oiling to birds are well accurately characterize the injury assessment step of the restoration known from oil spills worldwide (e.g., Exxon Valdez, Nestucca, Apex planning phase. Houston, and various spills in the North Sea) (Burger, 1993), the physiological effects and ramifications of exposure to lesser amounts of 6.2. Thermoregulation oil are not as clear and may have greater effects on avian populations than acute mortality. When ingested by birds at less than acutely lethal One such potential effect of exposure to oil is disruption of dosages, oil can cause a wide range of adverse effects, including thermoregulation. Thermoregulation is of critical importance to home- anemia, decreased nutrient absorption, altered stress response, and othermic species with high energetic demands, such as migratory birds. decreased immune function (Miller et al., 1978; Szaro et al., 1978; Oil spills in colder climates have illustrated the detrimental effects of Holmes, 1984; Leighton et al., 1985; Leighton, 1985, 1986, 1993; extensive oil coverage on bird survival (Perry et al., 1978). Not only Peakall et al., 1983). does oil cause reduced buoyancy, which in turn increases the surface area of the bird exposed to cold water, but feather barbules become 6.1. Hemolytic anemia matted, further reducing the insulative properties of the feathers (Lambert et al., 1982; O'Hara and Morandin, 2010). In areas such as One of the toxic effects of ingested oil in birds is hemolytic anemia the sub-arctic, this results in rapid hypothermia and death. In more (Hartung and Hunt, 1966; Eastin and Rattner, 1982; Pattee and temperate regions, these changes in thermoregulatory ability have not Franson, 1982; Lee et al., 1986; Leighton et al., 1985; Leighton, been well studied. In areas such as the GOM, the changes in thermo- 1986; Hughes et al., 1990; Yamato et al., 1996; Walton et al., 1997; regulatory ability are more likely to increase energetic demands and Newman et al., 2000; Seiser et al., 2000; Troisi et al., 2007). Hemolytic cause behavioral modifications (Jenssen, 1994). Birds exposed experi- anemia occurs by damage to hemoglobin in red blood cells associated mentally to oil for only a few hours showed increases in heat production with the formation of epoxides of polycyclic aromatic hydrocarbons and thermal conductance even at room temperature (McEwan and (PAHs) by the action of cytochrome P450 (CYP450) monooxygenases in Koelink, 1973; Erasmus et al., 1981). Jackass penguins (Spheniscus the liver. These reactive oxygen species cause oxidative damage to demersus) with environmental oil exposure covering up to 70% of their

6 S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 4–10 bodies had a 2.5o C drop in body temperature after only 15 min in water Birds exposed to oil show increases in inflammatory responses, depres- at 19.5–20.5o C. In water temperatures of 4.5–6o C, common eiders sions in circulating lymphocyte numbers and immunosuppression that (Somateria mollissima) exposed to 70 ml of oil became hypothermic in results in secondary infections (Fry and Lowenstine, 1985; Briggs et al., only 70 min (Jenssen and Ekker, 1991). Sanderlings (Calidris alba) with 1997; McOrist and Lenghuas, 1992; Newman et al., 2000). The irritant 20% oil coverage of their feathers spent significantly more time effects of oil on the gastrointestinal tract, combined with observed preening and less time resting and were more aggressive (Burger and decreases in adrenal function, make it unsurprising that inflammatory Tsipoura, 1998). Environmentally exposed black-legged kittiwakes responses are upregulated. Decreased circulating lymphocyte numbers (Rissa tridactyla) had shorter, more frequent foraging trips following could be indicative of site-specific action; however, inflammatory oiling (Walton et al., 1997). The consequences of increased energetic processes are not well understood in avian responses to oil. Further, demands to maintain body temperature, combined with behavioral the ability of a bird to fight off secondary infections is likely to be changes such as increased time spent preening and changes in foraging diminished given its immunocompromised state. patterns, have the potential to cause weight loss, interfere with reproduction, affect the immune response and prevent optimal body 7. DWH NRDA avian toxicity studies condition for migration. However, there was question whether cur- rently available information was sufficient to assess these types of The information in the literature at the time helped to identify some injuries for live oiled birds exposed to MC252 oil that were observed of the effects of oil on individual systems or responses in birds. during and after the DWH oil spill. However, information from the literature and the field was not judged to be sufficient to fully characterize the nature and extent of the injuries 6.3. Food intake and nutrient absorption to the thousands of live oiled birds, or to quantify those injuries in terms of effects on bird viability. As a result, the FWS proposed various Food intake and nutrient absorption are of critical importance in assessment activities to inform NRDA injury determination and quanti- coping with changes in metabolic demand. Both increased and fication analyses associated with the DWH oil spill, including avian decreased food intake have been reported in experimentally oiled birds toxicity studies that are described here. The goal of these studies was to (Holmes et al., 1978a, b; Szaro et al., 1978, 1981; Gorsline et al., 1981; evaluate the effects of low to moderate oil exposure and potentially Rattner, 1981; Harvey et al., 1982; Miller et al., 1982; Pattee and repeated oil exposure that did not result in immediate or short-term Franson, 1982; Lee et al., 1985; Hughes et al., 1990; Evans and Keijl, mortality but might result in physiological effects that ultimately could 1993; Burger and Tsipoura, 1998). A bird's ability to absorb nutrients affect avian survival, reproduction and health. Phase 1 involved and produce sufficient fat stores can be affected by a combination of development of a work plan outlining the types of studies to be factors including impairments to gastrointestinal function (Crocker conducted based on consultation with experts including avian toxicol- et al., 1974; Hartung and Hunt, 1966; Beer, 1968; Eastin and Murray, ogists and pathologists, ecologists and ornithologists. During Phase 2, 1981; Miller et al., 1978, 1982), inflammatory responses (Briggs et al., the studies were implemented, data were analyzed and results were 1997; Newman et al., 2000) and increased metabolic rate (Butler et al., reported (US Department of the Interior, 2011). 1988). While loss of body condition is often reported in oiled birds, the There were five groups of avian toxicity studies that were initially extent of oiling, type of oil and length of exposure that cause these proposed by panels of academic, government and private sector experts changes are not well understood. Effects on nutrient and fat utilization convened by FWS. These studies included oral dosing studies, an could be particularly important for the many birds oiled in the GOM external dosing study, metabolic and flight performance studies, a during or prior to migration, as the ability of the birds to absorb energy reproductive effects study and field-based flight studies. from food is crucial during and immediately after the metabolically The species chosen for these studies included the double-crested taxing process of migration. cormorant (Phalacrocorax auritus), laughing gull (Leucophaeus atricilla), western sandpiper (Calidris mauri) and rock pigeon (Columba livia). 6.4. Homeostasis Double-crested cormorants were chosen because they were impacted by the DWH spill. They are common, primarily piscivorous seabirds that Environmental exposure to oil has also been shown to disrupt inhabit pelagic, coastal, and inland waterways (Glahn et al., 1995; homeostatic mechanisms. The hypothalamic-pituitary-adrenal (HPA) Johnson et al., 2002) and can be used as surrogates for other axis plays a critical role in maintaining homeostasis. Activation of the piscivorous species such as pelicans (Pelicanus sp.), terns, and skimmers. HPA axis results in secretion of corticosterone and aldosterone from the The laughing gull is a small black-headed gull that commonly nests in adrenal cortex. Corticosterone is typically described as being the large groups of up to 50,000 and was one of the most commonly oiled predominant stress hormone in birds, meaning that it coordinates species found in the GOM during the DWH oil spill. Its diet consists of physiological processes to allow the organism to respond to the stressful both terrestrial and aquatic invertebrates, and seasonal berries (Burger, event. Corticosterone is responsible for re-directing resources from 1988). The laughing gull's abundance and flexible diet makes the reproduction, reducing inflammatory responses and increasing catabo- species a useful potential model for studying the effects of DWH oil on a lism of energy reserves. Oil ingestion has a direct effect on the adrenal broad range of species. Western sandpipers are shorebirds that have cortex that results in cellular damage and affects the body's ability to long-distance migratory routes. They were chosen because they were mount an appropriate stress response. In birds, these effects include affected by the DWH spill and because of their applicability to methods increased adrenal gland weight and hypertrophy (Hartung and Hunt, used in metabolism and flight performance studies (Burns and 1966; Miller et al., 1978, 1982; Holmes et al., 1979; Peakall et al., 1983; Ydenberg, 2002; Nebel et al., 2013). The rock pigeon, although not a Leighton, 1986), changes in circulating corticosterone concentrations water-based bird such as those exposed to oil in the DWH spill, is a (Rattner and Eastin, 1981) and a blunted stress response (Gorsline and useful model to study the effects of oil exposure on flight dynamics and Holmes, 1982). However, it is less clear how this relates to other the metabolic challenges of migratory flight. The rock pigeon has been physiological changes following oil exposure, particularly metabolic used to assess the effects of other toxicants on flight activity (Moye and immune changes. et al., 2016). The objectives of the oral dosing studies were to develop oral dose- 6.5. Immune and inflammatory responses response relationships in multiple species that could be used to predict the toxicity of oil ingested by birds following the DWH oil spill, Immune and inflammatory responses are also upregulated in investigate endpoints that might be related directly or indirectly to response to exposure to oil (Briggs et al., 1996; Perez et al., 2010). the occurrence of hemolytic anemia to provide relevance to these

7 S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 4–10

findings and to determine oral doses to be used in the metabolic and publication, this issue). In the laughing gull study, in addition to reproductive studies. The objective of the external dosing study was to mortality there was a decrease in packed cell volume accompanied by develop quantitative relationships between external oiling of birds evidence of Heinz bodies, increases in hepatic antioxidant endpoints (percent cover, amount, location of oil) and internal dose. The and relative liver and kidney weights (Horak et al., submitted for objectives of the metabolic flight performance studies were to deter- publication, this issue). Double crested cormorants orally dosed with mine the adverse effects of external oil dosing on bird energetics and MC252 oil were the most severely affected of the species tested. metabolism including flight performance, thermoregulation, food/en- Oil–exposed birds ate less and weighed less compared to controls. ergy assimilation and body composition and to measure the effects of Clinical signs included reduced cloacal temperature, lethargy, feather feather oiling on flight performance. The objective of the reproductive damage, morbidity, and death (Cunningham et al., 2017, this issue). effects study, which was not conducted, was to establish a dose- There was evidence of hemolytic anemia as indicated by decreased response relationship between oral doses of MC252 oil and effects on packed cell volume, relative reticulocytosis with an inadequate regen- bird reproduction, including behavior, egg production, egg quality, egg erative response, and presence of Heinz bodies (Harr et al., 2017a, this viability and embryo/chick viability, health and growth. The objective issue). Additionally, oil-exposed birds had oil-induced changes in of the field-based flight studies was to determine if oral and external complete blood count estimates and plasma chemistry and electrophor- doses of MC252 oil affected performance in homing pigeons trained for esis endpoints (Dean et al., 2017b, this issue) and increases in CYP1A long-distance free flights between release sites and their loft. protein expression and catalytic activity (Alexander et al., submitted for publication, this issue). Liver and kidney weights were increased and 7.1. Oral dosing studies there was presence of lesions in kidney, liver, heart and thyroid gland (Harr et al., 2017b, this issue). Results related to changes in oxidative There were initially four oral dosing studies conducted with the stress endpoints in orally dosed DCCOs are presented in Pritsos et al. double-crested cormorant, laughing gull, western sandpiper and rock (2017, this issue). pigeon. The objective of each study was to determine the appropriate dosing method for administration of artificially weathered (Forth et al., 7.2. External dosing study 2016) MC252 oil for that species and if dosing with 10 and/or 20 ml of artificially weathered MC252 oil kg bw-1 day-1 for one or five An external dosing study was conducted with double-crested consecutive days resulted in the development of hemolytic anemia cormorants to provide a different exposure scenario than was achieved and associated endpoints. Oil was administered by gavage in combina- with the oral dosing studies (Cunningham et al., 2017, this issue). tion with a slurry of feed or immediately prior to gavage with feed. External dosing allows the bird to preen oil from their feathers Blood was collected at periodic intervals and processed for assessment providing continuous exposure to experimentally simulate field expo- of numerous endpoints including Heinz bodies, packed cell volume, sure. The purpose of the study was to determine if oiling of 20% of the hemoglobin concentration and complete blood counts, plasma chemis- bird's surface area every three days for 21 days resulted in oil toxicity. A tries, oxidative damage and cardiac biomarkers. Birds were necropsied subset of the endpoints found to be responsive during the oral dosing on day 6 with selective tissues being collected for evaluation of study with cormorants was chosen to determine if external oiling pathology, oxidative damage and/or CYP450 activity. Results of these methods produced oil toxicity. This subset of endpoints included the initial oral scoping studies indicated that a single dose of oil had no development of hemolytic anemia, cardiac abnormalities, and thermal effect within six days of dosing, but multiple doses caused significant stress. Results indicated hemolytic anemia was associated with Heinz changes in some hematology, plasma chemistry, acute phase protein bodies and reduced packed cell volume (Harr et al., 2017a, this issue), and oxidative stress endpoints. However, there were no obvious signs of as well as increased white blood counts, monocyte and lymphocyte oil-induced anemia and it was not possible to determine which plasma counts and changes in plasma chemistry endpoints (Dean et al., 2017b, endpoints could be used as indicators of oil toxicity. The appearance of this issue). Liver and kidney weights were greater in exposed birds oil in excreta within minutes of dosing in all species suggested that (Harr et al., 2017b, this issue) and there was evidence of oxidative there was insufficient time for absorption of toxicants when delivered damage (Pritsos et al., submitted for publication, this issue). Evaluation by oral gavage (Dean et al., 2017a, In this issue). of cardiac function showed decreased myocardial contractility and The decision was made to modify dosing methodology to more dysfunction and significant increases in ionized calcium in externally closely emulate field conditions and potentially increase absorption of oiled birds (Harr et al., 2017c, this issue). Thermography indicated oil. A second set of oral dosing studies was conducted with western greater heat loss in oiled birds with an associated increase in food sandpipers, laughing gulls and double-crested cormorants. For the consumption, presumably to meet the increased energetic demand sandpipers, oil continued to be administered in a feed slurry by gavage (Cunningham et al., 2017, this issue; Mathewson et al., submitted for but doses were decreased from 10 or 20 ml kg bw-1 day-1 to 1 or 5 ml kg publication). bw-1 day-1, and the days birds were dosed increased from four to 20. Both laughing gulls and double-crested cormorants were provided with 7.3. Metabolic and flight performance studies target doses of 5 or 10 ml kg bw-1 day-1 by provision of a daily allotment of oil-injected fish. Gulls were dosed for up to 28 days and The metabolic and flight performance studies were conducted with cormorants were dosed for up to 21 days. The objectives of the three western sandpipers. An external oil dosing/flight study was designed to oral dosing studies were to determine if dosing with artificially test the effects of external oiling on takeoff and endurance flight weathered MC252 oil at target doses of 1 or 5 (western sandpipers) performance. Birds were oiled externally on the wings and tail with or 5 or 10 ml kg bw-1 day-1 (gulls and cormorants) for 20–28 days artificially weathered MC252 oil. Takeoff performance determines the resulted in the development of hemolytic anemia and associated bird's ability to escape predators and was measured by inducing escape endpoints and to identify other anatomic, hematologic, and biochem- flights and measuring takeoff angle, speed and acceleration with high ical endpoints that warranted further investigation. The same endpoints speed video and accelerometers. Endurance performance determines that were evaluated in the oral scoping studies mentioned above were the bird's ability to make long distance flights, such as migration, and to evaluated in these oral dosing studies. sustain escape from predators. Flight endurance was measured in a Results of the western sandpiper study were not conclusive, but the wind tunnel under controlled conditions and energy costs of flight were numerical decrease in hemoglobin concentration, significant increase in measured using quantitative magnetic resonance. Results of the flight absolute liver weight and histological changes in the adrenal gland performance studies indicated that oil on the trailing edges of wings were consistent with exposure to oil (Bursian et al., submitted for and tail reduced takeoff speed by 30% and oil on wing, tail and breast

8 S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 4–10 feathers increased flight energy cost by 20–45% (Maggini et al., 2017a, stress on the immune system of seabirds. Regul. Toxicol. Pharmacol. 23, 145–155. Briggs, K.T., Gershwin, M.E., Anderson, D.W., 1997. Consequences of petrochemical this issue; Maggini et al., 2017b). A second metabolic study investigated ingestion and stress on the immune system of seabirds. ICES J. Mar. Sci. 54, 718–725. the effects of artificially weathered MC252 oil on nighttime resting Burger, A.E., 1993. Estimating the mortality of seabirds following oil spills: effects of oil metabolic rate and thermoregulatory costs. The objectives of this study volume. Mar. Pollut. Bull. 26, 140–143. ff Burger, J., 1988. Foraging behavior in gulls – differences in method, prey and habitat. were to determine the e ects of acute and three-day external oiling of Colonia. Waterbirds 11, 9–23. 20% of the body surface on nighttime resting metabolic rate and Burger, J., Tsipoura, N., 1998. Experimental oiling of sanderlings (Calidris alba): behavior thermal conductance, and to quantify adverse sub-lethal toxicological and weight changes. Environ. Toxicol. Chem. 17, 1154–1158. ff effects that may contribute to altered metabolic rates and thermogenic Burns, J.G., Ydenberg, R.C., 2002. The e ects of wing loading and gender on the escape flights of least sandpipers (Calidris minutilla) and western sandpipers (Calidris mauri). capacity under the different oil dosing regimens. The metabolic study Behav. Ecol. Sociobiol. 52, 128–136. indicated that while light feather oiling did not affect thermal Bursian, S.J., Dean, K.M., Harr, K.E., Kennedy, L., Link, J.E., Maggini, I., Pritsos, C.A., ff fi conductance, there was a reduction in body temperature and induction Pritsos, K., Schmidt, R., Guglielmo, C.G., 2017. E ect of oral exposure to arti cially weathered deepwater horizon crude oil on blood morphology and chemistries, of oxidative damage (Maggini et al., 2017c, this issue). hepatic antioxidant enzyme activities, organ weights and histopathology in western sandpipers (Calidris mauri). Ecotoxicol. Environ. Saf (submitted for publication, this 7.4. Field-based flight studies issue). Butler, R.G., Harfenist, A., Leighton, F.A., Peakall, D.B., 1988. Impact of sublethal oil and emulsion exposure on the reproductive success of Leach's storm-petrels: short and Three field-based flight studies were conducted with homing long-term effects. J. Appl. Ecol. 25, 125–143. pigeons to test the effects of externally applied MC252 artificially Carriger, J.F., Jordan, S.J., Kurtz, J.C., Benson, W.H., 2015. Identifying evaluation fl considerations for the recovery and restoration from the 2010 Gulf of Mexico oil spill: weathered crude oil on long distance ight performance. The objectives an initial appraisal of stakeholder concerns and values. Integr. Environ. Assess. of these studies to determine the effect of external oiling on flight from Manag. 11, 502–513. 50, 80 or 100 miles over time including flight speed and flight path and Corn M.L., Copeland C., 2010. The Deepwater Horizon oil spill: Coastal wetland and fl wildlife impacts and response. Congressional Research Service 7–5700, R41311. CRS whether external oiling combined with ying resulted in hemolytic Report for Congress, Prepared for Members and Committees of Congress. anemia and associated endpoints or weight loss in the homing pigeon. A Washington DC. third objective was to assess any damage to feather structure on the Crocker, A.D., Cronshaw, J., Holmes, W.N., 1974. The effect of crude oil on intestinal – homing pigeon due to external oiling of the feathers after multiple absorption in ducklings (Anas platyrhynchos). Environ. Pollut. 7, 165 177. Cunningham, F.L., Dean, K.M., Hanson-Dorr, K.C., Harr, K.E., Healy, K.A., Horak, K.E., flights. Oil was applied to cover 25% of the surface area of the tail Shriner, S.A., Bursian, S.J., Dorr, B.S., 2017. Development of methods for avian oil feathers and wings prior to release. Flight time and flight path were toxicity studies using the double crested cormorant (Phalacrocorax auritus). assessed with data loggers attached to the birds. Oiling resulted in Ecotoxicol. Environ. Saf. fl fl fl Dean, K.M., Cacela, D., Carney, M.W., Cunningham, F.L., Ellis, C., Gerson, A.R., altered ight paths, increased ight duration and increased ight Guglielmo, C.G., Hanson-Dorr, K.C., Harr, K.E., Healy, K.A., Horak, K.E., Isanhart, distance (Perez et al., submitted for publication, this issue). Oiling also J.P., Kennedy, L.V., Link, J.E., Lipton, I., McFadden, A.K., Moye, J.K., Perez, C.R., reduced the ability to regain body mass between flights (Perez et al., Pritsos, C.A., Pritsos, K.L., Muthumalage, T., Shriner, S.A., Bursian, S., 2017a. Testing of an oral dosing technique for double-crested cormorant, Phalacrocorax auritus, submitted for publication, this issue). laughing gull, Leucophaeus atricilla, homing pigeon, Columba livia, and western sandpiper, Calidris mauri, with artificially weathered MC252 oil. Ecotoxicol. Environ. 8. Conclusion Saf (submitted for publication, this issue). Dean, K.M., Bursian, S.J., Cacela, D., Carney, M.W., Cunningham, F.L., Dorr, B.S., Hanson-Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Lipton, I., McFadden, A.K., In the studies outlined above, exposure of selected avian species to McKernan, M.A., Harr, K.E., 2017b. Changes in white cell estimates and plasma weathered MC252 oil manifested itself in changes in hematologic chemistry measurements following oral or external dosing of double-crested cormorants, Phalacrocorax auritus, with artificially weathered MC252 oil. Ecotoxicol. parameters including formation of Heinz bodies and changes in cell Environ. Saf (submitted for publication, this issue). counts. There were also effects on multiple organ systems, cardiac Deepwater Horizon Natural Resource Damage Assessment Trustees, 2016. Deepwater function and oxidative status. Crude oil affected flight patterns and time Horizon oil spill: Final Programmatic Damage Assessment and Restoration Plan and 〈 spent during flight tasks indicating that migration may be affected by Final Programmatic Environmental Impact Statement. Retrieved from http://www. gulfspillrestoration.noaa.gov/restoration-planning/gulf-plan/〉. short-term repeated exposure to oil. Feather damage also resulted in Eastin, W.C., Murray, H.C., 1981. Effects of crude-oil ingestion on avian intestinal increased heat loss and energetic demands. The papers in this special function. Can. J. Physiol. Pharmacol. 59, 1063–1068. ff Eastin, W.C., Rattner, B.A., 1982. Effects of dispersant and crude oil ingestion on mallard issue indicate that the combined e ects of oil toxicity and feather – ff ducklings (Anas platyrhynchos). Bull. Environ. Contam. Toxicol. 29, 273 278. e ects in avian species, even in the case of relatively light oiling, can Erasmus, T., Randall, R.M., Randall, B.M., 1981. Oil pollution, insulation and body significantly affect the overall health of birds. Findings from these temperatures in the Jacass penguin Spheniscus demersus. Comp. Biochem. Physiol. – studies informed the assessment of damage due to the DWH oil spill and 69A, 169 171. Evans, M.I., Keijl, G.O., 1993. Impact of Gulf War oil spills on the wader populations of will do so for future spills as well. the Saudi Arabian Gulf coast. Sandgrouse 15, 85–105. Fallon J., Smith E.P., Hopkins W.A., 2014. Evaluating blood parameters as a measure of Acknowledgements the physiological injury to oiled birds from the Deepwater Horizon (MC252) oil spill. Technical Report for US Fish and Wildlife Service. DWH-AR0062366. 〈https://www. fws.gov/doiddata/dwh-ar-documents/788/DWH-AR0062366.pdf〉. The studies appearing in this special issue were funded by the US Forth, H.P., Mitchelmore, C.L., Morris, J.M., Lay, C.R., Suttles, S.E., Lipton, J., 2016. Fish and Wildlife Service as part of the Deepwater Horizon Natural Characterization of dissolved and particulate phases of water accommodated fractions used to conduct aquatic toxicity testing in support of the deepwater horizon Resource Damage Assessment. natural resource damage assessment. Environ. Toxicol. Chem. 〈http://onlinelibrary. wiley.com/doi/10.1002/etc.3672/full〉. References Fry, D.M., Lowenstine, L.J., 1985. Pathology of common murres and Cassin's auklets exposed to oil. Arch. Environ. Contam. Toxicol. 14, 725–737. Glahn, J.F., Brugger, K.E., 1995. The impact of double-crested cormorants on the Alexander, C., Hooper, M.J., Cacela, D., Smelker, K.D., Calvin, C.S., Dean, K.M., Bursian, Mississippi Delta catfish industry: a bioenergetics model. Colonia. Waterbirds 18, S.J., Cunningham, F.L., Hanson-Dorr, K.C., Horak, K.E., Isanhart, J.P., Link, J.E., 158–167. Shriner, S.A., Godard-Codding, C.A., 2017. CYP1A protein expression and catalytic Gorsline, J., Holmes, W.N., 1982. Suppression of adrenocortical activity in mallard ducks activity in double-crested cormorants experimentally exposed to Deepwater Horizon exposed to petroleum-contaminated food. Arch. Environ. Contam. Toxicol. 11, Mississippi Canyon 252 oil. Ecotoxicol. Environ. Saf (submitted for publication, this 497–502. issue). Gorsline, J., Holmes, W.N., Cronshaw, J., 1981. The effects of ingested petroleum on the Barron, M.G., 2012. Ecological impacts of the Deepwater Horizon oil spill: implications naphthalene-metabolizing properties of liver tissue in seawater-adapted mallard for immunotoxicity. Toxicol. Pathol. 40, 315–320. ducks (Anas platyrhynchos). Environ. Res. 24, 377–390. Beer, J.V., 1968. Post-mortem findings in oiled auks dying during attempted Hagerty C.L., Ramseur J.L., 2010 Deepwater Horizon oil spill: Selected issues for rehabilitation. In: Carthy, J.D., Arthur, D.R. (Eds.), The Biological Effects of Oil Congress. Congressional Research Service 7–5700, R41262. CRS Report for Congress, Pollution on Littoral Communities. Field Studies Council, London, pp. 123–129. Prepared for Members and Committees of Congress. Washington DC. Briggs, K.T., Yoshida, S.H., Gershwin, M.E., 1996. The influence of petrochemicals and Harr, K.E., Cunningham, F.L., Pritsos, C., Pritsos, K., Muthumalage, T., Dorr, B.S., Horak,

9 S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 4–10

K.E., Hanson-Dorr, K.C., Dean, K.M., Cacela, D., McFadden, A.K., Link, J.E., Healy, thermoregulatory costs from sublethal oil exposure in a colonial waterbird. Ecol. Appl K.A., Tuttle, P., Bursian, S.J., 2017a. Weathered MC252 crude oil-induced anemia (submitted for publication). and abnormal erythroid morphology in double-crested cormorants (Phalacrocorax McEwan, E.H., Koelink, A.F.C., 1973. The heat production of oiled mallards and scaup. auritus) with light microscopic and ultrastructural description of Heinz bodies. Can. J. Zool. 51, 27–31. Ecotoxicol. Environ. Saf (submitted for publication, this issue). McOrist, S., Lenghuas, C., 1992. Mortalities of little penguins (Eudyptula minor) following Harr, K.E., Bursian, S.J., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Hanson- exposure to crude oil. Vet. Rec. 130, 161–162. Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Reavill, D.R., Shriner, S.A., Schmidt, Miller, D.S., Peakall, D.B., Kinter, W.B., 1978. Ingestion of crude oil: sublethal effects in R.E., 2017b. Comparison of organ weights and histopathology between double- herring gull chicks. Science 199, 315–317. crested cormorants (Phalacrocorax auritus) dosed orally or externally with artificially Miller, D.S., Hallett, D.J., Peakall, D.B., 1982. Which components of crude oil are toxic to weathered Mississippi Canyon 252 crude oil. Ecotoxicol. Environ. Saf (submitted for young seabirds? Environ. Toxicol. Chem. 1, 39–44. publication, this issue). Moye, J.K., Perez, C.R., Pritsos, C.A., 2016. Effects of parenteral and methylmercury Harr, K.E., Rishniw, M., Rupp, T.L., Cacela, D., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C., exposure on flight activity in young homing pigeons (Columba livia). Environ. Pollut. Healy, K., Horak, K., Link, J.E., Reavill, D.R., Bursian, S.J., Cunningham, F.L., 2017c. 5, 23–30. Dermal exposure to weathered MC252 crude oil results in echocardiographically Nebel, S., Buehler, D.M., MacMillan, A., Guglielmo, C.G., 2013. Flight performance of identifiable systolic myocardial dysfunction in double-crested cormorants western sandpipers, Calidris mauri, remains uncompromised when mounting an acute (Phalacrocorax auritus). Ecotoxicol. Environ. Saf (submitted for publication, this phase immune response. J. Exp. Biol. 216, 2752–2759. issue). Newman, S.H., Anderson, D.W., Ziccardi, M.H., Trupkiewicz, J.G., Tseng, F.S., Hartung, R., Hunt, G.S., 1966. Toxicity of some oils to waterfowl. J. Wildl. Manag. 30, Christopher, M.M., Zinkl, J.G., 2000. An experimental soft-release of oil-spill 564–570. rehabilitated American coots (Fulica americana): ii. Effects on health and blood Harvey, S., Sharp, P.J., Phillips, J.G., 1982. Influence of ingested petroleum on the parameters. Environ. Pollut. 107, 295–304. reproductive performance and pituitary-gonadal axis of domestic ducks (Anas O'Hara, P.D., Morandin, L.A., 2010. Effects of sheens associated with offshore oil and gas platyrhynchos). Comp. Biochem. Physiol. 72C, 83–89. development on the feather microstructure of pelagic seabirds. Mar. Pollut. Bull. 60, Holmes, W.N., 1984. Petroleum pollutants in the marine environment and their possible 672–678. effects on seabirds. In: Hodgson, I.E. (Ed.), Reviews in Environmental Toxicology. Pattee, O.H., Franson, J.C., 1982. Short-term effects of oil ingestion on American kestrels Elsevier, New York, pp. 251–317. (Falco sparverius). J. Wildl. Dis. 18 (2), 235–241. Holmes, W.N., Cavanaugh, K.P., Cronshaw, J., 1978a. The effects of ingested petroleum Peakall, D.B., Miller, D.S., Kinter, W.B., 1983. Toxicity of crude oils and their fractions to on oviposition and some aspects of reproduction in experimental colonies of mallard nestling herring gulls: 1. Physiol. Biochem. Eff. Mar. Environ. Res. 8, 63–71. ducks (Anas platyrhynchos). J. Reprod. Fertil. 54, 335–347. Perez, C., Lores, M., Velando, A., 2010. Oil pollution increases plasma antioxidants but Holmes, W.N., Cronshaw, J., Gorsline, J., 1978b. Some effects of ingested petroleum on reduces coloration in a seabird. Oecologia 163, 875–884. seawater-adapted ducks (Anas platyrhynchos). Environ. Res. 17, 177–190. Perez, C., Moye, J., Cacela, D., Dean, K., Pritsos, C.A., 2017b. Body mass change in flying Holmes, W.N., Gorsline, J., Cronshaw, J., 1979. Effects of mild cold stress on the survival homing pigeons externally exposed to Deepwater Horizon crude oil. Ecotoxicol. of seawater-adapted mallard ducks (Anas platyrhynchos) maintained on food Environ. Saf (submitted for publication, this issue). contaminated with petroleum. Environ. Res. 20, 425–444. Perez, C.R., Moye, J.K., Cacela, D., Dean, K.M., Pritsos, C.A., 2017a. Low level exposure to Horak, K.E., Bursian, S.J., Ellis, C., Dean, K., Link, J., Hanson-Dorr, K., Cunningham, F., crude oil impacts avian flight performance: the Deepwater Horizon oil spill effect on Harr, K., Pritsos, C., Pritsos, K., Healy, K., Cacela, D., Shriner, S., 2017. Toxic effects migratory birds. Ecotoxicol. Environ. Saf (submitted for publication, this issue). of orally ingested oil from the deepwater horizon spill on Laughing Gulls. Ecotoxicol. Perry M.C., Ferrigno F., Settle F.H., 1978. Rehabilitation of birds oiled on two mid- Environ. Saf (submitted for publication, this issue). Atlantic estuaries. Proceedings of the Annual Conference of the Southeastern Hughes, M.R., Kasserra, C., Thomas, B.R., 1990. Effect of externally applied bunker fuel Association of Fish and Wildlife Agencies 32:318-325. on body mass and temperature, plasma concentration, and water flux of glaucous- Piersma, T., Everaarts, J.M., Jukema, J., 1996. Build-up of red blood cells in refueling bar- winged gulls, Larus glaucescens. Can. J. Zool. 68, 716–721. tailed godwits in relation to individual migratory quality. Condor 98, 363–370. Jenssen, B.M., 1994. Effects of oil pollution, chemically treated oil, and cleaning on the Pritsos, K.L., Perez, C.R., Muthumalage, T., Dean, K.M., Cacela, D., Hanson-Dorr, K., thermal balance of birds. Environ. Pollut. 86, 207–215. Cunningham, F.L., Bursian, S.J., Link, J.E., Shriner, S.A., Horak, K.E., Pritsos, C.A., Jenssen, B.M., Ekker, M., 1991. Dose dependent effects of plumage-oiling on 2017. Dietary intake of Deepwater Horizon oil-injected live food fish by double-crested thermoregulation of common eiders Somateria mollissima residing in water. Polar Res. cormorants resulted in oxidative stress. Ecotoxicol. Environ. Saf (submitted for 10, 579–584. publication, this issue). Johnson, J.H., Ross, R.M., McCullough, R.D., 2002. Little Galloo Island, Lake Ontario: a Rattner, B.A., 1981. Tolerance of adult mallards to subacute ingestion of crude petroleum review of nine year of double-crested cormorant diet and fish consumption oil. Toxicol. Lett. 8, 337–342. information. J. Gt. Lakes Res. 28, 182–192. Rattner, B.A., Eastin, W.C., 1981. Plasma corticosterone and thyroxine concentrations Lambert, G., Peakall, D.B., Philogene, B.J.R., Engelhardt, F.R., 1982. Effect of oil and oil during chronic ingestion of crude oil in mallard ducks (Anas platyrhynchos). Comp. dispersant mixtures on the basal metabolic rate of ducks. Bull. Environ. Contam. Biochem. Physiol. 68C, 103–107. Toxicol. 29, 520–524. Seiser, P.E., Duffy, L.K., McGuire, A.D., Roby, D.D., Golet, G.H., Litzow, M.A., 2000. Lee, Y.-Z., Leighton, F.A., Peakall, D.B., Norstrom, R.J., O'Brien, P.J., Payne, J.F., Comparison of pigeon guillemot, Cepphus columba, blood parameters from oiled and Rahimtula, A.D., 1985. Effects of ingestion of Hibernia and Prudhoe Bay crude oils on unoiled areas of eight years after the Exxon Valdez oil spill. Mar. Pollut. Bull. hepatic and renal mixed function oxidase in nestling herring gulls (Larus argentatus). 40, 152–164. Environ. Res. 36, 248–255. Szaro, R.C., Albers, P.H., Coon, N.C., 1978. Petroleum: effects on mallard egg Lee, Y.-Z., O'Brien, P.J., Payne, J.F., Rahimtula, A.D., 1986. Toxicity of petroleum crude hatchability. J. Wildl. Manag. 42, 404–406. oils and their effect on xenobiotic metabolizing enzyme activities in the chicken Szaro, R.C., Hensler, G., Heinz, G.H., 1981. Effects of chronic ingestion of No. 2 fuel oil on embryo in ovo. Environ. Res. 39, 153–163. mallard ducklings. J. Toxicol. Environ. Health 7, 789–799. Leighton, F.A., 1985. Morphological lesions in red blood cells from herring gulls and Troisi, G., Borjesson, L., Bexton, S., Robinson, I., 2007. Biomarkers of polycyclic aromatic Atlantic puffins ingesting Prudhoe Bay crude oil. Vet. Pathol. 22, 393–402. hydrocarbon (PAH)-associated hemolytic anemia in oiled wildlife. Environ. Res. 105, Leighton, F.A., 1986. Clinical, gross, and histological findings in herring gulls and Atlantic 324–329. puffins that ingested Prudhoe Bay crude oil. Vet. Pathol. 23, 254–263. US Department of the Interior, 2011. Interim, Partial Claim for Assessment Costs; 20 April Leighton, F.A., 1993. The toxicity of petroleum oils to birds. Environ. Rev. 1, 92–103. 2010 Deepwater Horizon (MC252) Spill. Retrieved from: 〈https://www.doi.gov/ Leighton, F.A., Lee, Y.Z., Rahimtula, A.D., O'Brien, P.J., Peakall, D.B., 1985. Biochemical sites/doi.gov/files/migrated/deepwaterhorizon/upload/Final-DOI-NRDA-AP-1-July- and functional disturbances in red blood cells of herring gulls ingesting Prudhoe Bay 2011-2.pdf〉. crude oil. Toxicol. Appl. Pharmacol. 81, 25–31. Vann A., Meltz R., 2013. The 2010 Deepwater Horizon oil spill: Natural Resource Damage Maggini, I., Kennedy, L.V., Elliott, K.H., Dean, K.M., MacCurdy, R., Macmillan, A., Pritsos, Assessment under the Oil Pollution Act. Congressional Research Service 7-5700, CRS C.A., Guglielmo, C.G., 2017a. Trouble on takeoff: crude oil on feathers reduces escape Report for Congress, Prepared for Members and Committees of Congress. performance of shorebirds. Ecotoxicol. Environ. Saf (submitted for publication, this Washington DC. issue). Vann A., Meltz R., 2013. The 2010 Deepwater Horizon oil spill: Natural Resource Damage Maggini, I., Kennedy, L.V., Macmillan, A., Elliott, K.H., Dean, K., Guglielmo, C.G., 2017b. Assessment under the Oil Pollution Act. Congressional Research Service 7-5700, CRS Like a ball and chain: light oiling of feathers increases flight energy expenditure in a Report for Congress, Prepared for Members and Committees of Congress. migratory shorebird. J. Exp. Biol (submitted for publication). Washington DC. Maggini, I., Kennedy, L.V., Bursian, S.J., Harr, K.E., Link, J.E., Pritsos, C.A., Pritsos, K.L., Walton, P., Turner, C.M.R., Austin, G., Burns, M.D., Monaghan, P., 1997. Sub-lethal Dean, K.M., Guglielmo, C.G., 2017c. Toxicological and thermoregulatory effects of effects of an oil pollution incident on breeding kittiwakes Rissa tridactyla. Mar. Ecol. feather contamination with artificially weathered MC 252 oil in western sandpipers Progress. Ser. 155, 261–268. (Calidris mauri). Ecotoxicol. Environ. Saf (submitted for publication, this issue). Yamato, O., Goto, I., Maede, Y., 1996. Hemolytic anemia in wild seaducks caused by Mathewson, P., Hanson-Dorr, K., Porter, W., Bursian, S., Dean, K., Healy, K., Horak, K., marine oil pollution. J. Wildl. Dis. 32, 381–384. Link, J., Harr, K., Dorr, B., 2017. Using a bioenergetics model to predict

10 Ecotoxicology and Environmental Safety 146 (2017) 11–18

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Testing of an oral dosing technique for double-crested cormorants, MARK Phalacocorax auritus, laughing gulls, Leucophaeus atricilla, homing pigeons, Columba livia, and western sandpipers, Calidris mauri, with artificially weather MC252 oil ⁎ K.M. Deana, , D. Cacelaa, M.W. Carneya, F.L. Cunninghamb, C. Ellisc, A.R. Gersond,e, C.G. Guglielmoe, K.C. Hanson-Dorrb, K.E. Harrf, K.A. Healyg, K.E. Horakc, J.P. Isanharth, L.V. Kennedye, J.E. Linki, I. Liptona, A.K. McFaddena, J.K. Moyej, C.R. Perezj, C.A. Pritsosj, K.L. Pritsosj, T. Muthumalagej, S.A. Shrinerc, S.J. Bursiani a Abt Associates, 1811 Ninth St., Suite 201, Boulder, CO 80302, USA b USDA/APHIS/WS/NWRC-MS Field Station, MS State University, P.O. Box 6099, Starkville, MS 39762, USA c USDA/APHIS/WS/NWRC, 4101 LaPorte Ave, Fort Collins, CO, USA d Biology Department, University of Massachusetts, Amherst, MA, USA e Department of Biology, University of Western Ontario, ON, Canada f Urika Pathology LLC, 8712 53rd Pl W., Mukilteo, WA 98275, USA g US Fish and Wildlife Service, Deepwater Horizon NRDAR Field Office, Fairhope, AL, USA h US Department of the Interior, Office of Restoration and Damage Assessment, DC, USA i Department of Animal Science, Michigan State University, East Lansing, MI, USA j Department of Agriculture, Nutrition and Veterinary Sciences, University of Nevada, Reno, NV, USA

ARTICLE INFO ABSTRACT

Keywords: Scoping studies were designed to determine if double-crested cormorants (Phalacocorax auritus), laughing gulls DWH (Leucophaues atricilla), homing pigeons (Columba livia) and western sandpipers (Calidris mauri) that were gavaged Deepwater Horizon oil spill with a mixture of artificially weathered MC252 oil and food for either a single day or 4–5 consecutive days Birds oil spill showed signs of oil toxicity. Where volume allowed, samples were collected for hematology, plasma protein Oil toxicity electrophoresis, clinical chemistry and electrolytes, oxidative stress and organ weigh changes. Double-crested Oil gavage cormorants, laughing gulls and western sandpipers all excreted oil within 30 min of dose, while pigeons re- PAH birds gurgitated within less than one hour of dosing. There were species differences in the effectiveness of the dosing technique, with double-crested cormorants having the greatest number of responsive endpoints at the comple- tion of the trial. Statistically significant changes in packed cell volume, white cell counts, alkaline phosphatase, alanine aminotransferase, creatine phosphokinase, gamma glutamyl transferase, uric acid, chloride, sodium, potassium, calcium, total glutathione, glutathione disulfide, reduced glutathione, spleen and liver weights were measured in double-crested cormorants. Homing pigeons had statistically significant changes in creatine phosphokinase, total glutathione, glutathione disulfide, reduced glutathione and Trolox equivalents. Laughing gulls exhibited statistically significant decreases in spleen and kidney weight, and no changes were observed in any measurement endpoints tested in western sandpipers.

1. Introduction important life cycle and breeding habitat for a wide range of bird species. The length of time that oil was present in these habitats and the During the Deepwater Horizon oil spill that began on April 20, 2010 wide area that it covered meant there was the potential for birds to be many important bird habitats in the northern Gulf of Mexico became oiled multiple times and to varying extents. contaminated and resulted in birds becoming oiled. The Gulf of Mexico The acute mortality caused by heavy oiling is well understood. Loss is not only an important migratory stopover, but it also represents of feather function and thermoregulatory abilities can be lethal within a

⁎ Correspondence to: Department of Neuroscience, University of Lethbridge, AB, Canada. E-mail address: [email protected] (K.M. Dean). http://dx.doi.org/10.1016/j.ecoenv.2017.07.003 Received 13 October 2016; Received in revised form 29 June 2017; Accepted 3 July 2017 Available online 03 August 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 11–18 matter of hours in cold waters (Burger, 1993). But the damage to in- 2. Methods dividual and population health caused by sub-lethal exposures are more difficult to quantify. Not only are there a range of conditions that may 2.1. Toxicant and general dosing be observed including, anemia, organ dysfunction, loss of feather function, decreased nutrient absorption, altered stress response, and Artificially weathered MC252 oil (DWH7937, batch# B030112) was decreased immune function (Szaro et al., 1978; Leighton et al., 1985; prepared from crude oil collected during the DWH oil spill. Oil char- Leighton, 1985, 1986; Peakall et al., 1982, 1989; Leighton, 1993), there acterization is described in Forth et al. (2017). are also be functional repercussions caused by the deficits. Reduced As described in Forth et al. (2017), source oil was artificially take off speed in western sandpipers due to a light oiling of wings weathered by TDI-Brooks using a modification of Carls et al. Approxi- (Maggini et al., 2017; this edition) may increase predation risk, al- mately 3.5 L of source oil was heated in a glass beaker to 90–105 °C though there may be no associated measurable health effects. The using a digital hot plate (model CMAGHP751, IKA) and stirred using a Deepwater Horizon oil spill is also made more complicated by the spill digital, top-loading mixer (model BDC250, Caframo) to mix but not duration. The spill occurred over a period of months, and oil persisted aerate the oil. The oil was stirred until a mass loss of approximately in the environment for considerably longer (PDARP, 2016). De- 33–38% was achieved which generally took about 20 h to complete. termining the oral oil exposure that results in adverse outcomes for This correlated to a BTEX (i.e., sum of benzene, toluene, ethylbenzene, avian species is of critical importance, but following the DWH spill and xylene) depletion of approximately 99.9% and a TPAH50 depletion confounding factors such as determining when the bird was oiled, by of approximately 22% relative to hopane. When not in use, the oil was how much and how many times, made it virtually impossible to esti- stored in a leaf-proof container in a flammables storage cabinet. mate dosage and relate it to health deficits. Rather than try to untangle Oil was mixed with species-specific diets for use in each test, aiming the complex nature of such events a simpler approach was attempted for dosing of between 5 and 20 ml/kg body weight/24 h. These doses for determining MC252 oil toxicity to birds. That was to determine if a correspond to literature reports of anemia development in response to single oiling event could in a measurable way, affect the health of an gavage with oil in mallards and herring gull chicks (Hartung and Hunt, individual bird. 1966; Leighton, 1986). These health impacts can be measured through the use of a number of clinically relevant plasma markers, blood cell counts, necropsy 2.2. Species-specific dosing and sampling methods findings and hepatic antioxidant enzymes. Hemolytic anemia is often a defining feature of oil toxicity in birds. It is reported in both field and 2.2.1. Double-crested cormorants laboratory studies (Hartung and Hunt, 1966; Eastin and Rattner, 1982; 2.2.1.1. Animal care. Adult and sub-adult (based on plumage Pattee and Franson, 1982; Lee et al., 1986; Leighton et al., 1985; development) double-crested cormorants (Phalacocorax auritus) were Leighton, 1986; Hughes et al., 1990; Yamato et al., 1996; Walton et al., collected from Little Mossy Lake, MS on November 2, 2012 (33.340, 1997; Newman et al., 2000; Seiser et al., 2000; Troisi et al., 2007). −90.423), according to NWRC approved Institutional Animal Care and Hemolytic anemia occurs by oxidative damage to hemoglobin during Use (IACUC) protocol QA1992. Birds were allowed acclimate for two detoxification and elimination of polycyclic aromatic hydrocarbons weeks in individual pens containing 190 L water-filled tanks for feeding (PAHs) from the body (Peakall et al., 1989; Troisi et al., 2007). In young and perching. Up to 600 g of fresh live fingerling channel catfish and growing birds such as herring gulls and Atlantic puffins, a single (Ictalurus punctatus;10–20 cm) were supplied and consumption oral dose of oil can result in hemolytic anemia within 4–5 days measured daily. (Leighton, 1986) Experimental oral exposure to oil in birds is generally achieved 2.2.1.2. Gavage. Birds were divided into three treatment groups through either feeding trials or gavage. Feeding trials have most com- (control (1 female, 6 males), single dose of 20 ml oil/kg BW (4 monly been used for longer duration studies in ducks (Holmes et al., females, 2 males and 1 indeterminate sex) and 5 day dosing of 20 ml 1978; Szaro et al., 1978; Harvey et al., 1981, 1982; Cavanaugh and oil/kg ((2 females, 5 males)). Human grade catfish fillets were Holmes, 1982; Cavanaugh et al., 1983), but there are a limited number homogenized in a commercial grade blender with an equivalent of studies where short-term adapted methods have been used. Alonso- water volume to form a1:1 slurry. This slurry was then mixed 1:1 Alvarez et al. (2007b) used bread covered in oil to feed to yellow-legged with artificially weather MC252 oil. A single batch of oil:food slurry gulls for a seven day feeding trial. However, gavage is more commonly was prepared on the day prior to first dose, then stored as aliquots as used to investigate acute exposure (Hartung and Hunt, 1966; Wootton −20 °C until use. Slurry was thawed to room temperature each day. et al., 1979; McEwan and Whitehead, 1978, 1980; Eastin and Rattner, Cormorants were fasted overnight before each gavage. Volume for 1982; Leighton et al., 1985; Leighton, 1985, 1986; Lee et al., 1985; gavage was adjusted according to body weight, with each bird receiving Peakall et al., 1989; Brausch et al., 2010). As such, gavage methods a total of 40 ml of oil:food slurry per kilogram body weight. That is were developed for one captive model species, the homing pigeon and approximately 80 ml in total, as a single daily bolus. Gavage was three Gulf of Mexico relevant species, present in the gulf for different achieved by filling a 60 ml BD syringe capped with a 40 cm long, 6 mm periods of time. The laughing gull, which is present during the warmer inner diameter flexible catheter with the oil:food slurry, then slowly spring and summer months, the double-crested cormorant which over- inserting into the esophagus until a moderate level of resistance was winters in the area and the western sandpiper which uses the area as a felt, indicating tube placement proximal to the stomach (approximately migratory stopover point were also tested to determine if short duration 35 cm). The oil:fish slurry was slowly expelled, then syringe removed oral gavage or single bolus dose of oil could result in anemia in the and re-filled to reach the 80 ml (adjusted for BW) total. The syringe and 4–5 day duration recorded by Leighton (1986). The homing pigeon, a feeding tube were weighed before and after dosing to verify the dose novel but highly useful laboratory model for flight and toxicity testing administered. Feeder fish were provided immediately in each pen fol- was also used for methods comparison. Unlike the other species tested, lowing gavage and birds were monitored for regurgitation and first sign the pigeon is primarily a seed eating bird with a large crop and mus- of oil in feces. Dosing continued for 5 days. All birds were handled cular gizzard that limit the rate at which food reaches the intestines. every day with sham gavage used for controls and single dose birds. Toxicity was monitored not only by development of anemia, but by the use of clinically relevant plasma markers, oxidative stress markers and 2.2.1.3. Blood collection. Blood was collected from the brachial vein blood counts. using a 25 G needle and syringe prior to dosing, then on dose days 2 and 4, with additional samples collected at necropsy on day 6. Heparinized blood was used to prepare two blood smears for complete blood count

12 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 11–18

(CBC) and two blood smears for Heinz bodies. For the latter, 20 µl of feed was mixed with 325 ml distilled water in a blender. For dosing, heparinized whole blood was incubated for 20 min at room 1.0 ml of feed slurry was drawn up into the syringe followed by oil temperature with 40 µl new methylene blue (new methylene blue N volume adjusted for bird weight. A 10 cm length of a 14-French [NMB]; Sigma-Aldrich, St. Louis, MO) that was prepared by mixing catheter tube (Kendall, Mansfield, MA) was attached to the Luer end 0.025 g NMB and 0.08 g potassium oxalate (Sigma-Aldrich, St. Louis, of the syringe and inserted through the bird's crop, until resistance was MO) with 5 ml water following the method of Leighton (1985). The met with, indicating the catheter was placed at the entrance to the blood/NMB mixture was then used to make the Heinz body smears. stomach. When the contents of the syringe were expelled, the dose of oil EDTA-treated whole blood was sent to the Mississippi State was followed by the feed slurry, effectively cleaning the oil out of the University College of Veterinary Medicine Diagnostic Laboratory syringe and gavage tube. Services (MSU CVM DLS; Starkville, MS) for plasma chemistries. Birds were fasted overnight (approximately 15 h) prior to the first Included in the panel were electrolyte concentrations (sodium, po- day of oil administration. The intended nominal daily oil dose was to be tassium, chloride, phosphorus and calcium), enzyme activities (alkaline 20 ml oil/kg body weight (BW) for a single day or for five consecutive phosphatase, alanine aminotransferase (ALT), aspartate amino- days. However, when this dose was administered to the birds (2 female, transferase (AST), gamma glutamyl transferase (GGT), creatine phos- 5 male) that would receive only a single bolus dose on Day 1 of testing, phokinase (CK), lactate dehydrogenase (LDH)) and blood analytes the majority of birds regurgitated immediately. As a result, the dose was (glucose, cholesterol, urea, uric acid, total protein, creatinine). Plasma reduced to 5 ml/kg BW (2 female, 5 male) administered twice daily for was obtained by centrifugation of blood at 2000g for 5 min. Plasma was a total of 10 mg/kg BW per day, with at least 3 h between doses for the removed and stored at −80 °C until shipping. Plasma samples were five day dose group. Regurgitation was considerably less, but was not analyzed by the University of Miami, Miller School of Medicine, Avian entirely eliminated. Control birds (1 female, 5 male, 3 indeterminate) and Wildlife Laboratory for plasma protein electrophoresis serum were sham gavaged by inserting the gavage tube to approximately the amyloid A and haptoglobin concentrations. same length and restraining the birds for an equivalent period of time. Between oil doses, birds were allowed ad libitum food access; how- 2.2.1.4. Necropsy. On day 6 a single brachial vein sample was collected ever, consumption was variable. Time between gavage and regurgita- for the above measurements, then birds were euthanized by cervical tion or presence of oily excreta was recorded for each bird. dislocation and necropsied immediately. Additional blood was collected via post-mortem cardiac puncture for analysis of 3-methyl histidine 2.2.2.3. Blood collection. One day prior to the start of the study a blood (Metabolic Technologies, Iowa). sample (up to 5.0 ml) was collected from the brachial vein using a 25 Assessment of organ gross abnormalities were undertaken at ne- gauge needle, then decanted into the appropriate size of BD lithium cropsy. The liver was removed first, weighed to the nearest 0.001 g, heparin or serum separate tube before centrifugation. Blood also was then five 1.0 g and one 2.0 g sections were flash frozen in liquid ni- collected from all birds on Days 2, 4 and 6 in the same way. CBC and trogen for measurement of oxidative damage (Dr. Chris Pritsos, Heinz body slides were prepared as per the double-crested cormorants. University of Nevada, Reno) and Cytochrome P450 enzyme activity (Dr. Plasma samples were analyzed by the University of Miami, Miller Celine Godard-Codding, Texas Tech University), respectively. Samples School of Medicine, Avian and Wildlife Laboratory. Included in the were sorted on dry ice, then transferred to a low temperature (−70 °C) panel were plasma protein electrophoresis serum amyloid A and hap- freezer until they were shipped on excess dry ice. Thyroid gland, spleen, toglobin concentrations, electrolyte concentrations (sodium, potassium, adrenal glands, brain, heart, lungs, remainder of liver, kidneys and chloride, phosphorus and calcium), enzyme activities (alkaline phos- gastrointestinal (GI) tract were removed and weighed, then placed in phatase, alanine aminotransferase (ALT), aspartate aminotransferase 10% neutral buffered formalin (NBF). If present, the thymus and bursa (AST), gamma glutamyl transferase (GGT), creatine phosphokinase of Fabricius were noted and removed and placed in 10% neutral buf- (CK)) and blood analytes (glucose, total protein, creatinine). Plasma fered formalin. Twenty-four hours later organs were transferred to fresh was obtained by centrifugation of blood at 2000g for 5 min. Plasma was 10% NBF before being shipped to Zoo Exotic Pathology Services (ZEPS) removed and stored at −80 °C until shipping. for standard hematoxylin and eosin staining by Dr. D. Reavill and Dr. R. Schmidt. 2.2.2.4. Necropsy. Necropsies were performed as per double-crested cormorants, but organs were weighed to the nearest 0.1 mg. 2.2.2. Homing pigeons 2.2.2.1. Animal care. Adult homing pigeons (Columba livia domestica) 2.2.3. Laughing gulls aged 2–7 years were purchased from a racing pigeon breeder (Foy's 2.2.3.1. Animal care. Adult laughing gulls (Leucophaeus atricilla) were Pigeon Supplies, Beaver Falls, PA) and maintained at the University of collected from Cameron Parrish, LA on August 14, 2012, according to Nevada Reno (UNR) Homing Pigeon Research Facility, following a 7 NWRC approved IACUC protocol QA1992. Upon arrival at the NWRC- days quarantine. Pigeons were maintained according to the University MS Field Station. For the first three days of quarantine birds were group of Nevada, Reno IACUC guidelines. housed in 3.3 m × 3.3 m × 2.0 m pens (five birds per pen) to improve Birds were maintained in lofts with outdoor fly pens attached. adaptation to captivity. Then they were transferred to individual During quarantine and maintenance, birds were provided Purina Mills stainless steel rabbit cages measuring 61 cm × 46 cm × 38 cm or Nutriblend Green (5454; St. Louis, MO) and water ad libitum. During 71 cm × 46 cm × 38 cm for the remainder of the 14-day quarantine testing, feed was withheld for approximately 15 h starting in the late period. Birds remained in this cages for the study. Unique leg bands afternoon/early evening of the day prior to dosing with MC252 oil. On were applied during the quarantine period. Food [frozen/thawed days when birds were not dosed with MC252 oil, feed was offered ad pogies (Brevoortia sp.) and Mazuri Fish Analog 50/10 mix (Purina libitum. All experimental birds were transferred daily (control, single Mills, St. Louis, Missouri)] and water were provided fresh daily. Birds dose, and 5 day dosed birds) to portable holding cages to ensure were randomly assigned to treatment groups identical handling. Following dosing each bird was placed in an in- dividual cage for observation. At completion of each day of dosing all 2.2.3.2. Gavage. Preparation of the oil:fish slurry was identical to that birds were returned to their lofts overnight. used for double-crested cormorants. Human grade catfish fillets were homogenized in a commercial grade blender with an equivalent water 2.2.2.2. Gavage. Oil and feed were separately aspirated into the gavage volume to form a1:1 slurry. This slurry was then mixed 1:1 with syringe because it was not possible to prepare an oil/feed slurry that artificially weather MC252 oil. A single batch of oil:food slurry was didn’t separate while loading the syringe. One hundred grams of pigeon prepared on the day prior to first dose, then stored as aliquots as

13 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 11–18

−20 °C until use. Slurry was thawed to room temperature each day. 3.7 m) under 12 L:12D light conditions at approximately 19 °C. One Birds were divided into three treatment groups (control (1 female, 5 week prior to dosing, birds were transferred into a large holding room male, 1 indeterminate), single dose of 20 ml oil/kg BW (2 female, 5 under the same photoperiod and temperature. Four days prior to male) and 5 day dosing of 20 ml oil/kg (2 female, 5 male)). The total dosing, dividers were placed in the room to create four 1.8 × 1.8 m volume for the day was divided into two equivalent doses. Birds were corrals. The five or six birds comprising each of the four dose groups gavaged with an 18-French (approximately 6 mm) polyvinyl chloride (see below) were maintained in their respective corrals throughout the feeding tube (20.5 cm long, 6.0 mm outer diameter) attached to a 10 ml trial. glass syringe. Just prior to dosing, birds were weighed. The syringe and Birds were fed an ad libitum diet of 6:1 mix of Aquamax fish diet feeding tube were loaded with the appropriate volume of oil or oil:fish (SD03 Fingerling Starter 300; Agribrands Purina Canada, Woodstock, slurry based on the bird's BW. The feeding tube was inserted into the Ontario) and Purinature Chick Starter (Agribrands Purina Canada). The bird's esophagus until a moderate level of resistance was felt, indicating diet was supplemented with mealworms every other day. tube placement proximal to the stomach (approximately 9 cm) and the oil or oil:fish slurry was expelled from the syringe and feeding tube. The syringe and feeding tube were weighed before and after dosing to verify 2.2.4.2. Gavage. Birds were divided into four groups for testing the dose administered. After each bird was dosed, it was placed in its (control (1 female, 4 males), 20 ml/kg BW for one day (1 female, 4 cage and monitored for regurgitation and first appearance of oily ex- males), 10 ml/kg BW for 4 days (1 female, 4 males) and 20 ml/kg BW creta. Water and half the daily allowance of food were offered at this for 4 days (2 females, 4 males)). Oil was administered as a 1:1 oil/ time. Approximately 4–6 h later, the birds dosed with an oil:fish slurry mealworm slurry. A mealworm homogenate was prepared at the ratio dose identical to the first dose of the day, resulting in a total of 20 ml of six mealworms per 2 ml water, placed in a 12 ml polypropylene tube oil/kg BW (40 ml oil: fish slurry/kg BW). Monitoring was continued, and homogenized with an Omni 2000 (Omni International, Kennesaw, and birds were provided with the remainder of their food allotment GA) variable speed tissue homogenizer. This mixture was centrifuged at until approximately one hour after sunset. Control birds were gavaged 2000g in a Galaxy Mini microcentrifuge (VWR International, St. identically to the treated birds, but with an empty syringe. Catharines, Ontario) to remove cuticle particles that would clog the gavage needle. 2.2.3.3. Blood collection. Blood (up to 1.0 ml) was drawn through the Birds were weighed daily to the nearest gram prior to dosing. The brachial or tarsal vein on Day 1 (just prior to first dose of oil:fish slurry), total volume for the day was divided into two equivalent doses, the first Day 2, Day 4 and Day 6. The size of these birds would only allow for of which was given after a one hour fast. The appropriate volumes of minimal sampling to be undertaken during the study, so on Days 1, 2 homogenate and oil for each bird's BW were combined in a micro- and 4 the only blood collected was for preparation of CBC and Heinz centrifuge tube and the mixture thoroughly vortexed for 30 s. The ap- body slides (as above). The brachial vein was punctured with a 25 G propriate volume was administered to a manually restrained bird needle and blood was drawn with a 1 ml syringe, and decanted into through a 5.08 cm, 20 G stainless steel gavage needle attached to a 1 ml heparinized tubes for processing. Samples were kept at 4 °C prior to Luer lock glass syringe. The second dose was administered approxi- processing, which took place within 4 h of collection. Plasma and whole mately one hour after the first dose for a total dose of 10 or 20 ml oil/kg blood were analyzed by Mississippi State University College of BW. Veterinary Medicine Diagnostic Laboratory Services. Birds were observed for the first 30 min after dosing to assess On day 6 additional blood was collected from the brachial vein for whether oil was regurgitated or excreted. After the first 30 min, birds measurement of packed cell volume, hemoglobin concentration, elec- were observed every 30 min for two hours to assess whether oil was trolyte concentrations (sodium, potassium, chloride, phosphorus and regurgitated or excreted, and the approximate amount. Birds were calcium), enzyme activities (alkaline phosphatase, alanine amino- provided food ad libitum approximately 30 min after the second dose of transferase (ALT), aspartate aminotransferase (AST), gamma glutamyl the day until one hour prior to receiving the next day's first dose. transferase (GGT), creatine phosphokinase (CK), lactate hydrogenase (LDH)) and blood analytes (glucose, total protein, creatinine). These samples were taken to the analytical laboratory within 90 min of col- 2.2.4.3. Blood collection. The small size of the western sandpipers lection. Additional blood was collected via cardiac puncture for plasma limited the blood volume collected during the study to preparation of protein electrophoresis, haptoglobin and serum amyloid A analyses at Heinz body slides. Approximately 30 µl of blood was collected on Day 3 the University of Miami following cervical dislocation. Cardiac blood from a pricked 27 G, 1.27 cm needle) brachial vein using a heparinized was obtained with a heparinized 22 G needle with a 3 ml syringe. microhematocrit tube. The blood was added to two times its volume of Plasma was obtained by centrifugation of blood at 2000g for 5 min. New Methylene Blue N stain (NMB, Sigma-Aldrich, St. Louis, MO; Plasma was removed and stored at −80 °C until shipping. 0.025 g NMB plus 0.08 g potassium oxalate in 5 ml of deionized water mixed fresh just prior to use) and incubated at room temperature for 2.2.3.4. Necropsy. Necropsies were performed as per the double- 20 min. Two smears were made from this preparation and allowed to crested cormorants. air dry. Following blood collection, birds were gavaged as above. On day five, birds were bled (26 G, 1.27-cm needle, 1-ml syringe) 2.2.4. Western sandpipers via jugular and/or brachial vein to collect as much blood as possible. 2.2.4.1. Animal care. Adult western sandpipers (Calidris mauri) that The limited blood volume of the birds precluded extensive analyses, had previously been held at the University of Western Ontario's however, blood smears for Heinz bodies, complete blood counts (CBCs) Advanced Facility for Avian Research (AFAR) were used in this study. and clinical chemistries were prioritized. Collection volumes ranged These birds were captured in British Columbia under the guidelines of from 500 to 1150 µl. Heinz body blood slides were prepared as de- the University of Western Ontario University Council on Animal Care scribed above. The slides and remaining whole blood were sent im- and according to permit CA-0256 from the Canadian Wildlife Service. mediately to the University of Guelph Animal Health Laboratory The birds were captured for a separate study on the effects of immune (Guelph, ON, Canada) for analysis. challenge on flight capacity that took place 11 months prior to this dosing trial. In this previous experiment the birds were treated with lipopolysaccharide to induce an immune response and a blood sample 2.2.4.4. Necropsy. Necropsies were performed as per the double- had been collected from each. crested cormorants, but without collection of liver for oxidative stress Birds were maintained a specialized shorebird room (2.4 m × or CYP450 measurements. Organs were weighed to the nearest 0.1 mg.

14 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 11–18

2.3. Statistical methods least following a short duration study, to result in initial increases in organ weight as part of detoxification processes. Alterations in kidney Body weights and hematologic and plasma clinical chemistry values and spleen weights are likely be an incidental finding in this study due determined over multiple time points were modeled as a repeated to the small sample sizes used, particularly since there were no ac- measures analysis of covariance (ANCOVA) with interaction, where companying changes in clinical chemistry markers. Homing pigeons treatment group is a categorical explanatory variable and elapsed days also showed very little response to the gavage method for oil dosing. is a continuous explanatory variable. Differences among treatment Creatine phosphokinase (CK) was higher in the five day (10 ml/kg BW/ groups on Day 0 were evaluated. 3-Methyl histidine, malondialdehyde day) dosed birds. While an increase in CK alone in the five day dosed + 4-hydroxylalkenals, hepatic oxidative stress, and organ weight birds could indicate some liver involvement, it could also be part of a endpoints were analyzed by one-way analyses of variance (ANOVAs) stress response to handling and dosing as the birds may be more prone utilizing the proc mixed procedure of SAS (v. 9.4). Means were com- to slight muscle injury. Although all birds were handled the same way, pared using the least squares means function and considered sig- since the five day dosed group were regurgitating, that the procedures nificantly different if p < 0.05. Relative organ weights were analyzed were more stressful to them as a whole. At no point during the study did by one-way analyses of variance (ANOVAs) utilizing the proc mixed there to appear to be any differences in the way any of the pigeons procedure of SAS following arc sin transformation. responded to handling; however, over the course of the study the five day dosed birds did become more lethargic following dosing. One im- 3. Results/discussion portant difference in homing pigeon relative to western sandpiper and laughing gull was that there were also changes in oxidative stress Of primary interest for each of the studies reported here was whe- markers for both the single dose (20 ml/kg BW) and the five day ther the gavage method for delivery of oil was appropriate, and second (10 ml/kg BW/day). Glutathione and superoxide dismutase (SOD) both if any of the species tested showed signs of oil toxicity. As such, the showed decreases, which could be construed as counter-intuitive fol- endpoint responses for all species are summarized in Table 1. Raw data lowing polycyclic aromatic hydrocarbon (PAH) exposure; however, the for avian toxicity studies conducted as part of the Deepwater Horizon declines are likely just indicative of the early stages of exposure, prior Damage Assessment are publicly available at https://www.diver.orr. to enzyme activity upregulation within the liver. Nonetheless, these noaa.gov/deepwater-horizon-nrda-data, while work plans and reports changes are indications that PAHs were absorbed by the homing pigeon can be accessed through https://www.doi.gov/deepwaterhorizon/ following the gavage procedure. adminrecord. Double-crested cormorants were the only species to show at least a It should be noted that for this study we were unable to accurately partial response to oil exposure in the gavage trial; however, there were count Heinz bodies or reticulocytes in any of the slides prepared, de- also a number of inconclusive results that will require further study to spite consultation with three different veterinarians. This was likely due elucidate. Both the single dose (20 ml/kg BW) and the five day dose to a combination of factors including, some the oil exposure methods (20 ml/kg BW/day) groups had drops in packed cell volume that typify perhaps not causing Heinz body formation, the interval between dosing avian response to oil (Hartung and Hunt, 1966; Eastin and Rattner, and blood collection being too short to detect Heinz bodies, variation in 1982; Pattee and Franson, 1982; Lee et al., 1986; Leighton et al., 1985; timing of stain exposure, and in some cases fading of the stain due over Leighton, 1986; Hughes et al., 1990; Yamato et al., 1996; Walton et al., time. Following consultation with Dr. Kendal Harr (Urika Pathology 1997; Newman et al., 2000; Seiser et al., 2000; Troisi et al., 2007). LLC) it was decided that further experiments would incorporate elec- White blood cells counts also declined for the five day dose birds but tron microscopy for detection and light microscopy for quantitation of not the single dose birds (df;2,51F = 5.00; p = 0.022), while monocyte Heinz bodies. counts increased more for single dosed birds than controls or five day Western sandpipers did not show any measurable changes in blood dosed birds (df = 2,15; F = 5.95; p = 0.013). Heterophils decreased endpoints or organ weights following gavage with the oil:food slurry. over the course of the study, but this was not dose-related (df = 1,51; F However, these birds excreted oil within 10 min of gavage. Rapid oil = 4.24; f = 0.045). Changes in white cells counts related to dose would excretion was also observed for laughing gulls and double-crested be expected to begin during the study as inflammatory processes are cormorants, but was substantially longer in homing pigeons. The stimulated; however, the complex nature of these responses would not minimum duration for detection of oily excreta in pigeons was 90 min; only be driven by oil exposure, but by a generalized stress response to however, there were some observations of oily excreta the following captivity and handling, as well as any pre-existing health conditions, morning before the next gavage. such as the presence of injuries, parasites or diseases. It is clear from Pigeons regurgitated more frequently than the other species, and the even our very basic white cell counts that more complex analyses would number of treated birds that regurgitated increased throughout the be required to understand how oil toxicity influences and itself affects study. Regurgitation was observed within a 10–120 min window fol- immune function in free-living birds. lowing gavage, but no attempt was made to determine specific volumes, A subset of the plasma clinical chemistry markers also had statisti- as the oil was often ejected with additional materials from the crop, cally significant dose-related changes. Potential endpoints indicative of including some bedding materials consumed by the birds. The pigeon is liver damage that could be affected by oil exposure are alkaline phos- the avian emesis model for testing of pharmaceutical agents such as phatase, ALT, AST, cholesterol, glucose, GGT, LDH and total protein cisplatin (Tanihata et al., 2000), exhibiting both early and delayed (Szaro et al., 1981; Eastin and Rattner, 1982; Newman et al., 2000; emetic responses. The immediate regurgitation observed for these pi- Seiser et al., 2000; Golet et al., 2002; Alonso-Alvarez et al., 2007a, geons indicates there may have been some element of gastro-intestinal 2007b). Total protein (df = 2,51; F = 10.90; p = 0.0001), glucose (df irritation caused by the oil, or through stimulation of central or per- = 2,51; F = 5.48; p = 0.007) and cholesterol (df = 1,51; F = 4.08; p ipheral nervous system in the pigeons; however, this was beyond the = 0.049) concentrations decreased with treatment for the group scope of these studies. Regurgitation was not observed in the control treated with oil for five days by day 6, indicative of liver dysfunction. birds or the single dose birds after day 1, so it is unlikely that stress or Albumin (df = 2, 51; F = 0.79; p = 0.4576), ALT, alkaline phospha- the gavage process itself was responsible. The pigeon was the only tase, GGT and LDH activities were inconclusive. ALT decreased in species for which food intake began to decrease, but given the short controls and 5 day dosed birds, but not in the single dosed birds (df = duration of the study there were no changes in body weight. 2,15; F = 7.61; p = 0.005) while alkaline phosphatase increased in the The only responsive endpoints measured in the laughing gulls were five day dosed birds, but did not change in controls or single dosed the kidney and spleen weights, both of which were lower in the five day birds (df = 2,51; F = 5.20; p = 0.02). GGT activity decreased in the (20 ml/kg BW/day) dosed birds. Generally, oil would be expected, at five day treated birds (df = 2,51; F = 8.70; p < 0.0005), but not in the

15 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 11–18

Table 1 Summary of results from Deepwater Horizon Natural Resources Damage Assessment avian toxicity scoping studies.

WESA Scopinga LAGU Scoping DCCO Scoping ROPI Scoping

Single dose Daily dose Single dose Daily dose Single dose Daily dose Single dose Daily dose

Analyte 20 ml oil/kg 10 ml oil/kg 20 ml oil/kg 20 ml oil/kg 20 ml oil/kg 20 ml oil/kg 20 ml oil/kg 20 ml oil/kg 10 ml oil/kg BW BW BW BW BW BW BW BW BW

Hematology Packed cell volume (%) NA NA NA NC NC ↓↓NC NC Hemoglobin NA NA NA NC NC NC NC NC NC Reticulocyte count IC IC IC NC NC NA NA NC NC Heinz bodies IC IC IC IC IC IC IC IC IC White blood cell count NA NA NA NA NA IC ↓ NC NC Basophil count NA NA NA NA NA NA NA NA NA Eosinophil count NA NA NA NA NA NC NC NC NC Heterophil count NA NA NA NA NA NC NC NC NC Lymphocyte count NA NA NA NA NA NC NC NC NC Monocyte count NA NA NA NA NA ↑↑NC NC Plasma proteins Pre-albumin NA NA NA NC NC NC NC NC NC Albumin NA NA NA NC NC NC NC NC NC α−1-globulins NA NA NA NC NC NC NC NC NC α−2-globulins NA NA NA NC NC NC ↓ NC NC β-globulins NA NA NA NC NC NC NC NC NC γ-globulins NA NA NA NC NC NC IC NC NC Albumin:globulin ratio NA NA NA NC NC NC NC NC NC Plasma clinical chemistries Alanine NA NA NA NC NC IC ↓ NC NC aminotransferase Alkaline phosphatase NA NA NA NC NC IC ↑ NC NC Aspartate NA NA NA NC NC NC NC NC NC aminotransferase Creatine phosphokinase NA NA NA NC NC ↑↑NC ↑ Creatinine NA NA NA NC NC IC IC NC NC Gamma glutamyl NA NA NA NC NC ↓↓NC NC transferase Haptoglobin NA NA NA NC NC IC IC NC NC Lactate dehydrogenase NA NA NA NA NA IC IC NA NA 3-methyl-histidine NA NA NA NA NA NC ↑ NA NA Serum amyloid A NA NA NA NC NC NC NC NC NC Total protein NA NA NA NC NC ↓↓NC NC Cholesterol NA NA NA NC NC IC IC NA NA Glucose NA NA NA NC NC ↓↓NC NC Urea NA NA NA NA NA NC NC ↓ NA Uric acid NA NA NA NA NA NC ↑ NA NA Plasma minerals Calcium NA NA NA NC NC ↓↓NC NC Chloride NA NA NA NC NC NA NA NC NC Phosphorus NA NA NA NC NC IC IC NC ↓ Potassium NA NA NA NC NC IC IC NC NC Sodium NA NA NA NC NC IC ↓ NC NC Hepatic antioxidant enzymes Total glutathione NC NC NC NC NC NC ↑↓↓ Glutathione disulfide NC NC NC NC NC ↑↑NC ↓ Reduced glutathione NC NC NC NC NC NC ↑↓NC Lipid peroxidation NA NA NA NC NC NC NC NC NC Superoxide dismutase NC NC NC NC NC NC NC NC ↓ Trolox NC NC NC NA NA NA NA NA NA Organ weightsb Adrenals NC NC NC NC NC NC NC NC NC Brain NA NA NA NC NC NC NC NC NC Heart NC NC NC NC NC NC NC NC NC Kidneys NC NC NC NC ↓ NC NC NC NC Liver NC NC NC NC NC NC ↑ NC NC Spleen NA NA NA NC ↓ NC NC NC NC Thyroids NA NA NA NC NC NC NC NC NC Relative organ weights Adrenal NC NC NC NC NC NC NC NC NC Brain NA NA NA NC NC NC NC NC NC Heart NC NC NC NC NC NC NC NC NC Kidneys NC NC NC NC ↓ NC ↑ NC NC Liver NC NC NC NC NC ↑↑NC NC Spleen NA NA NA NC ↓ NC NC NC NC Thyroid NA NA NA NC NC NC NC NC NC Body weight NC NC NC NC NC NC NC NC ↓ Feed intake NA NA NA NC NC NC NC NA NA (continued on next page)

16 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 11–18

Table 1 (continued)

WESA Scopinga LAGU Scoping DCCO Scoping ROPI Scoping

Single dose Daily dose Single dose Daily dose Single dose Daily dose Single dose Daily dose

Analyte 20 ml oil/kg 10 ml oil/kg 20 ml oil/kg 20 ml oil/kg 20 ml oil/kg 20 ml oil/kg 20 ml oil/kg 20 ml oil/kg 10 ml oil/kg BW BW BW BW BW BW BW BW BW

Regurgitation NO NO NO NO OC NO NO YES YES Approximate time to oil < 10 min < 10 min < 10 min 10–20 min 10–20 min 20–30 min 20–30 min > 90 min > 90 min excretion

Abbreviations: NA = not analyzed; NC = no change, relative to control; IC = inconclusive; OC = occasional. Only significant (p < 0.05) changes are indicated by up or down arrows. a Due to marked hemolysis of the plasma samples, the analyzing lab deemed the samples unusable for intended analyte analysis. b Only absolute and relative weights and body weights were analyzed for this study. control or single dose birds. LDH decreased in controls and single dose would be required before a conclusion could be made as to whether the birds, but remained constant in five day dosed birds (df=2,51; F=7.72; muscle wastage was due to captivity/handling stress, general lack of p=0.001). The measured changes in GGT, LDH and alkaline phospha- activity due to captivity or oil-induced wastage from lack of nutrient tase activities are somewhat confounded by statistically significant consumption or ingestion. differences among controls and dosed birds prior to the start of dosing, The use of an acute gavage study for oral dosing of birds was largely so interpretation is limited. AST did not change with treatment or over unsuccessful for adding to our understanding of the effects of oil on the course of the study. avian physiology. The most commonly observed consequence of in- Changes in albumin, glucose, cholesterol, total protein, ALT, GGT, gestion of sublethal volumes of oil is the development of hemolytic LDH, alkaline phosphatase and AST, are likely indicative of damage to anemia. However, over the course of the five day studies, the double- the liver and biliary tree, but changes in some of these endpoints such crested cormorant was the only species for which anemia was observed. as alkaline phosphatase, AST and LDH are non-specific birds (Harr, Due to issues with rapid excretion of oil, it was impossible to determine 2005; Hochleitner et al., 2005). Further experimentation with larger if the lack of response in the other species was due to the dose provided, sample sizes, longer dosing and/or sampling are needed for further lack of absorption of the dose or a short exposure time. While it was clarification. Though increases relative liver weight (p < 0.05) and possible, at least in the double-crested cormorant, to gain some insight liver oxidative stress markers such as total glutathione (p < 0.0001), into the types of responsive endpoints that should be measured, such as reduced glutathione (p < 0.0001) and oxidized glutathione (p = clinical markers of liver, kidney, gastro-intestinal and muscle damage, 0.0001) are all indications that the liver is responding to PAH exposure it is apparent that further study and method development is required in through upregulation of detoxification processes. So while oxidative order to fully elucidate the functional impacts. damage to the liver has not progressed sufficiently for manifestation of clinical disease state to be apparent, initial processes have begun even Acknowledgements after only a short exposure time, indicating that there is potential for double-crested cormorants to be a useful species for understanding the The studies appearing in this special issue were funded by the U.S. effects of Deepwater Horizon oil on adult Gulf of Mexico birds. Changes Fish and Wildlife Service (Order No. F12PD01069) as part of the in concentrations and activities of plasma clinical chemistry markers in Deepwater Horizon Natural Resource Damage Assessment. Special control birds and differences among groups prior to dosing for some thanks to Paul Fioranelli, Raleigh Middleton, Treena Ferguson, Scott clinical markers make it very difficult to draw conclusions as to which Lemmons, Kathryn McGlammery, Breanna Sage and Jesse Fallon, for are the most important for interpretation of the effects of oil on organ their technical assistance. Also, thanks to Dr. Carolyn Cray and Marilyn function in double-crested cormorants. Rodriguez at the University of Miami Comparative Pathology Additionally there was some evidence in the double-crested cor- Laboratory, Dr Kris Ruotsalo at the University of Guelph Animal Health morants the kidney, gastro-intestinal and muscle damage may also be Laboratory and the Mississippi State University College of Veterinary occurring. Uric acid, a marker of kidney function, was increased sig- Medicine Diagnostic Laboratory Services for all of their patience and nificantly during the study in the five day dosed birds, but did not excellent work on sample analysis. change in single dose birds (df = 2,51; F = 8.05; p = 0.0009). Decreases in plasma calcium in both single and five day dosed birds (df References = 2,51; F = 14.37; p < 0.0001) and a decrease in sodium in the five day dosed birds (df = 2,15; F = 5.74; p = 0.014) may be indicative of Alonso-Alvarez, C., Munilla, I., Lopez-Alonso, M., Velando, A., 2007a. Sublethal toxicity kidney dysfunction, or gastro-intestinal irritation/inflammation pre- of the Prestige oil spill on yellow-legged gulls. Environ. Int. 33, 773–781. Alonso-Alvarez, C., Perez, C., Velando, A., 2007b. Effects of acute exposure to heavy fuel venting absorption across microvilli. Decreases in food intake and/or oil from the Prestige spill on a seabird. Aquat. Toxicol. 84, 103–110. impaired intestinal transport of oil have been suggested as possible Brausch, J.M., Blackwell, B.R., Beall, B.N., Caudillo, C., Kolli, V., Godard-Codding, C., mechanisms for the decrease in the plasma concentrations of choles- Smith, P.N., 2010. Effects of polycyclic aromatic hydrocarbons in northern bobwhite quail (Colinus virginianus). J. Toxicol. Environ. Health Part A 73, 540–551. terol, glucose and total protein (Eastin and Rattner, 1982; Newman Burger, A.E., 1993. Estimating the mortality of seabirds following oil spills: effects of spill et al., 2000; Alonso-Alvarez et al., 2007a, 2007b), and this may also be volume. Mar. Pollut. Bull. 26 (3), 140–143. the case for minerals. CK activity was increased in dosed birds relative Cavanaugh, K.P., Goldsmith, A.R., Holmes, W.N., Follett, B.K., 1983. Effects of ingested to controls, over the course of the study (df = 2,50; F = 4.83; p = petroleum on the plasma prolactin levels during incubation and on the breeding success of paired mallard ducks. Arch. Environ. Contam. Toxicol. 12, 335–341. 0.012), suggesting that as suggested by Newman et al. (2000), the Cavanaugh, K.P., Holmes, W.N., 1982. Effects of ingested petroleum on plasma levels of muscular damage inducing this increased activity was more likely the ovarian steroid hormones in photostimulated mallard ducks. Arch. Environ. Contam. – result of the stress of captivity and handling rather than oil ingestion. 3- Toxicol. 11, 503 508. Eastin, W.C.J., Rattner, B.A., 1982. Effects of dispersant and crude oil ingestion on methyl histidine, a marker of muscle wastage, was also higher in the mallard ducklings (Anas platyrhynchos). Bull. Environ. Contam. Toxicol. 29, 273–278. five day dosed birds (p = 0.049). The larger blood volume required for Final Programmatic Damage Assessment and Restoration Plan (PDARP) and Final this measurement meant that it could only be measured at necropsy, Programmatic Environmental Impact Statement (PEIS), 2016. Chapter 4: Injury to natural resources. http://www.gulfspillrestoration.noaa.gov/restoration-planning/ somewhat hampering interpretation. However, further information gulf-plan.

17 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 11–18

Forth, H.P., Mitchemore, C.M., Morris, J.M., Lay, C.R., Lipton, J., 2017. Characterization Maggini, I., Kennedy, L.V., Elliott, K.H., Dean, K.M., MacCurdy, R., Macmillan, A., Pritsos, of dissolved and particulate phases of water accommodated fractions used to conduct C.A., Guglielmo, C.G., 2017. Trouble on Takeoff: Crude oil on feathers reduces escape aquatic toxiciting testing in support of the Deepwater Horizon oil spill natural re- performance of shorebirds. Ecotoxicology and Environmental Safety, 〈http://dx.doi. source damage assessment. Environ. Toxicol. Chem. 36 (6), 1450–1459. org/10.1016/j.ecoenv.2017.04.025〉. Golet, G.H., Seiser, P.E., McGuire, A.D., Roby, D.D., Fischer, J.B., Kuletz, K.J., Newman, McEwan, E.H., Whitehead, P.M., 1978. Influence of weathered crude oil on liver enzyme S.H., 2002. Long-term direct and indirect effects of the 'Exxon Valdez' oil spill on metabolism of testosterone in gulls. Can. J. Zool. 56, 1922–1924. pigeon guillemots in Prince William Sound, Alaska. Mar. Ecol. Progress. Ser. 241, McEwan, E.H., Whitehead, P.M., 1980. Uptake and clearance of petroleum hydrocarbons 287–304. by the glaucous-winged gull (Larus glaucescens) and the mallard duck (Anas pla- Harr, K.E., 2005. Diagnostic Values of Biochemistry. In: Clinical AvianMedicine 1. Spix tyrhynchos). Can. J. Zool. 58, 723–726. Publishing, pp. 611–630 Ch 23. Newman, S.H., Anderson, D.W., Ziccardi, M.H., Trupkiewicz, J.G., Tseng, F.S., Hartung, R., Hunt, G.S., 1966. Toxicity of some oils to waterfowl. J. Wildl. Manag. 30 (3), Christopher, M.M., Zinkl, J.G., 2000. An experimental soft-release of oil-spill re- 564–570. habilitated American coots (Fulica americana): II. effects on health and blood para- Harvey, S., Klandorf, H., Phillips, J.G., 1981. Reproductive performance and endocrine meters. Environ. Pollut. 107 (3), 295–304. Retrieved from. 〈http://www.vliz.be/ responses to ingested petroleum in domestic ducks (Anas platyrhynchos). Gen. Comp. imisdocs/publications/141418.pdf〉. Endocrinol. 45 (3), 372–380. Pattee, O.H., Franson, J.C., 1982. Short-term effects of oil ingestion on American kestrels Harvey, S., Sharp, P.J., Phillips, J.G., 1982. Influence of ingested petroleum on the re- (Falco sparverius). J. Wildl. Dis. 18 (2), 235–241. productive performance and pituitary-gonadal axis of domestic ducks (Anas pla- Peakall, D.B., Hallett, D.J., Bend, J.R., Foureman, G.L., Miller, D.S., 1982. Toxicity of tyrhynchos). Comp. Biochem. Physiol. 72C (1), 83–89. Prudhoe Bay crude oil and its aromatic fractions to nestling herring gulls. Environ. Hochleitner, M., Hochleitner, C., Harrison, L.D., 2005. Evaluating and Treating the Liver. Res. 27 (1), 206–215. In: Harrison, G., Lightfoot, T. (Eds.), Clinical Avian Medicine 1. Spix Publishing, pp. Peakall, D.P., Norstrom, R.J., Jeffrey, D.A., Leighton, F.A., 1989. Induction of hepatic 441–450. mixed function oxidases in the herringg Gull (Larus agrentatus) by Pruhoe Bay crude Holmes, W.N., Cavanaugh, K.P., Cronshaw, J., 1978. The effects of ingested petroleum on oil and its fractions. Comp. Biochem. Physiol. 94C (2), 461–463. oviposition and some aspects of reproduction in experimental colonies of mallard Seiser, P.E., Duffy, L.K., McGuire, A.D., Roby, D.D., Golet, G.H., Litzow, M.A., 2000. ducks (Anas platyrhynchos). J. Reprod. Fertil. 54, 335–347. Comparison of pigeon guillemot, Cepphus columba, blood parameters from oiled and Hughes, M.R., Kasserra, C., Thomas, B.R., 1990. Effect of externally applied bunker fuel unoiled areas of Alaska eight years after the Exxon Valdez oil spill. Mar. Pollut. Bull. on body mass and temperature, plasma concentration, and water flux of glaucous- 40 (2), 152–164. winged gulls, Larus glaucescens. Can. J. Zool. 68, 716–721. Szaro, R.C., Dieter, M.P., Heinz, G.H., 1978. Effects of chronic ingestion of south Lee, Y.-Z., O'Brien, P.J., Payne, J.F., Rahimtula, A.D., 1986. Toxicity of petroleum crude Louisiana crude oil on mallard ducklings. Environ. Res. 17 (3), 426–436. oils and their effect on xenobiotic metabolizing enzyme activities in the chicken Szaro, R.C., Hensler, G., Heinz, G.H., 1981. Effects of chronic ingestion of No. 2 fuel oil on embryo in Ovo. Environ. Res. 39 (1), 153–163. mallard ducklings. J. Toxicol. Environ. Health 7, 789–799. Lee, Y.-Z., Leighton, F.A., Peakall, D.B., Norstrom, R.J., O'Brien, P.J., Payne, J.F., Tanihata, S., Igarashi, I., Suzuki, M., Uchiyama, T., 2000. Cisplatin-induced early and Rahimtula, A.D., 1985. Effects of ingestion of Hibernia and Prudhoe Bay crude oils on delayed emesis in the pigeon. Br. J. Pharmacol. 130 (1), 132–138. hepatic and renal mixed function oxidase in nestling herring gulls (Larus argentatus). Troisi, G., Borjesson, L., Bexton, S., Robinson, I., 2007. Biomarkers of polycyclic aromatic Environ. Res. 36 (1), 248–255. hydrocarbon (PAH)-associated hemolytic anemia in oiled wildlife. Environ. Res. 105 Leighton, F.A., 1985. Morphological lesions in red blood cells from herring gulls and (3), 324–329. Atlantic puffins ingesting Prudhoe Bay crude oil. Vet. Pathol. 22, 393–402. Walton, P., Turner, C.M.R., Austin, G., Burns, M.D., Monaghan, P., 1997. Sub-lethal ef- Leighton, F.A., 1986. Clinical, gross, and histological findings in herring gulls and Atlantic fects of an oil pollution incident on breeding kittiwakes Rissa tridactyla. Mar. Ecol. puffins that ingested Prudhoe Bay crude oil. Vet. Pathol. 23, 254–263. Retrieved Progr. Ser. 155, 261–268. from. 〈http://vet.sagepub.com/content/23/3/254.full.pdf〉. Wootton, T.A., Grau, C.R., Roudybush, T.E., Hahs, M.E., Hirsch, K.V., 1979. Reproductive Leighton, F.A., 1993. The toxicity of petroleum oils to birds. Environ. Rev. 1, 92–103. responses of quail to bunker C oil fractions. Arch. Environ. Contam. Toxicol. 8, Leighton, F.A., Lee, Y.Z., Rahimtula, A.D., O'Brien, P.J., Peakall, D.B., 1985. Biochemical 457–463. and functional disturbances in red blood cells of herring gulls ingesting Prudhoe Bay Yamato, O., Goto, I., Maede, Y., 1996. Hemolytic anemia in wild seaducks caused by crude oil. Toxicol. Appl. Pharmacol. 81 (1), 25–31. marine oil pollution. J. Wildl. Dis. 32 (2), 381–384.

18 Ecotoxicology and Environmental Safety 146 (2017) 19–28

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Reprint of: Development of methods for avian oil toxicity studies using the ☆ MARK double crested cormorant (Phalacrocorax auritus) ⁎ Fred Cunninghama, , Karen Deanb, Katie Hanson-Dorra, Kendal Harrc, Kate Healyd, Katherine Horake, Jane Linkf, Susan Shrinere, Steven Bursianf, Brian Dorra a USDA/ APHIS/ Wildlife Services/National Wildlife Research Center, Starkville, MS, United States b Abt Associates, Boulder, CO, United States c Urika Pathology LLC, Mukilteo, WA, United States d US Fish and Wildlife Service, Deepwater Horizon NRDAR Field Office, Fairhope, AL, United States e USDA/ APHIS/Wildlife Services/National Wildlife Research Center, Ft. Collins, CO, United States f Department of Animal Science, Michigan State University, East Lansing, MI, United States

ARTICLE INFO ABSTRACT

Keywords: Oral and external dosing methods replicating field exposure were developed using the double crested cormorant Methods development (DCCO) to test the toxicity of artificially weathered Deepwater Horizon Mississippi Canyon 252 oil. The majority Oil toxicity of previous oil dosing studies conducted on wild-caught birds used gavage methods to dose birds with oil and Double crested cormorant determine toxicity. However, rapid gut transit time of gavaged oil likely reduces oil absorption. In the present Deepwater Horizon studies, dosing relied on injection of oil into live feeder fish for oral dosing of these piscivorous birds, or applying Internal oil exposure oil to body contour feathers resulting in transdermal oil exposure and oral exposure through preening. Both oral External oil exposure and external oil dosing studies identified oil-related toxicity endpoints associated with oxidative stress such as hemolytic anemia, liver and kidney damage, and immuno-modulation or compromise. External oil application allowed for controlled study of thermoregulatory stress as well. Infrared thermal images indicated significantly greater surface temperatures and heat loss in treated birds following external oil applications; however, measurements collected by coelomically implanted temperature transmitters showed that internal body temperatures were stable over the course of the study period. Birds exposed to oil externally consumed more fish than control birds, indicating metabolic compensation for thermal stress. Conversely, birds orally dosed with oil experienced hypothermia and consumed less fish compared to control birds.

1. Introduction 1978; Szaro et al., 1978; Harvey et al., 1981, 1982; Pattee and Franson, 1982; Cavanaugh et al., 1983; Cavanaugh and Holmes, 1987; Alonso- The acutely lethal effects of heavy oiling to birds are well known Alvarez et al., 2007b), while gavage is more commonly used to from oil spills worldwide, such as the Exxon Valdez, Nestucca, Apex investigate acute exposure (Hartung and Hunt, 1966; Wootton et al., Houston, and various spills in the North Sea (Burger, 1993). When 1979; McEwan and Whitehead, 1980; Eastin and Rattner, 1982; ingested by birds at less than acutely lethal dosages, oil can cause a Leighton et al., 1985; Leighton, 1985, 1986; Lee et al., 1985; Peakall wide range of adverse effects, including anemia, decreased nutrient et al., 1989; Brausch et al., 2010). absorption, altered stress response, and decreased immune function To assess the specific adverse effects of Mississippi Canyon 252 (Szaro et al., 1978; Leighton et al., 1985; Leighton, 1985, 1986, 1993; (MC252) oil that was released during the Deepwater Horizon (DWH) oil Peakall et al., 1989). spill on Gulf of Mexico-relevant avian species, the double crested Experimental oral exposure of avian species to oil is generally cormorant (Phalacrocorax auritus; DCCO) was chosen as one of four achieved through either feeding trials or gavage. Feeding trials have species for initial oral dosing studies conducted under Phase 2 of the most commonly been used for longer duration studies (Holmes et al., avian toxicity studies for the DWH Natural Resource Damage assess-

DOI of original article: http://dx.doi.org/10.1016/j.ecoenv.2017.03.025 ☆ A publisher's error resulted in this article appearing in the wrong issue. The article is reprinted here for the reader's convenience and for the continuity of the special issue. For citation purposes, please use; Ecotoxicology and Environmental Safety Volume 141 pp. 199-208. ⁎ Correspondence to: National Wildlife Research Center, Mississippi Field Station, PO Box 6099, Mississippi State, MS 39759, United States. E-mail address: [email protected] (F. Cunningham). http://dx.doi.org/10.1016/j.ecoenv.2017.05.017 Received 29 August 2016; Received in revised form 17 March 2017; Accepted 18 March 2017 Available online 29 May 2017 0147-6513/ Published by Elsevier Inc. F. Cunningham et al. Ecotoxicology and Environmental Safety 146 (2017) 19–28 ment (NRDA) (Bursian, submitted for publication et al., this issue). Use Committee (IACUC) approved procedures. Capture was conducted Double crested cormorants were chosen specifically for inclusion in under the authority of USFWS Migratory Bird Permit # MB019065-3, these tests because they were affected by the DWH spill. Double crested and Mississippi and Alabama state scientific collection permits. Twenty- cormorants are also common, primarily piscivorous waterbirds that seven DCCOs were collected on March 12, 2013 from McIntyre Scatters, inhabit pelagic, coastal, and inland waterways (Glahn et al., 1995; Leflore County, MS for the oral dosing study and 31 DCCOs were Johnson et al., 2002) and as such can be used as surrogates for other captured from Mississippi and Alabama from night roosts on December piscivorous species, such as pelicans (Pelicanus sp.), terns (Sterna sp.), 29 and 30, 2014 and January 11 and 12, 2015 for the external oiling and skimmers (Rynchops sp). study using a customized capture boat, flood lights, and dip nets (King An initial oral toxicity study conducted with DCCOs followed et al., 1994). Birds were transported from the field to captive facilities similar protocols described in the literature (Leighton, 1986) as well in an enclosed trailer. as recommendations from an expert panel. Artificially weathered MC252 oil (DWH7937, batch# B030112) was prepared from crude 2.2. Feeding and housing oil collected during the DWH oil spill as described by Forth et al. (2016). Birds were gavaged with a 1:1 mixture of artificially weathered Upon arrival, DCCOs were weighed to the nearest gram (g) and MC252 oil and fish slurry that provided a dose of 20 mL oil/kg body individually marked with a unique alphanumeric coded plastic leg weight (bw) once or on five consecutive days at daily doses of 20 mL band. In both dosing studies the DCCOs were held in a 23 m (m) x 10 m oil/kg bw. Results of this initial oral toxicity trial suggested an apparent mechanically ventilated building. DCCOs were individually housed in treatment-related decrease in packed cell volume (PCV) with mild to cages that measured 3.3 m x 1.5 m x 2.0 m (length x width x height) moderate anemia by day 3. There were significant changes in complete and contained shallow, 190-liter (L) plastic tanks filled with water blood count (CBC), plasma chemistry, plasma protein and acute phase containing an air stone to provide adequate dissolved oxygen for protein endpoints, although values were consistently within reference maintenance of live channel catfish (Ictalurus punctatus). Water was intervals. Some endpoints indicative of oxidative stress were signifi- changed daily to limit oil re-exposure. During the external dosing study cantly different in oil-dosed birds compared to controls. Specifically, perches made of 7.6 cm (cm) diameter polyvinyl chloride (PVC) were there were significant changes in white cell counts, activities of alkaline provided in each pen. phosphatase, alanine aminotransferase, creatine phosphokinase and Cormorants were fed live, farm-raised fingerling channel catfish. gamma glutamyl transferase, concentrations of uric acid, chloride, Each DCCO was offered approximately 600 g catfish/day placed into sodium, potassium, calcium, total glutathione, glutathione disulfide the water tank within each individual pen. The following day, all and reduced glutathione. Kidney and liver weights were increased in uneaten catfish were removed from individual tanks and weighed to birds administered oil, although exposure of DCCOs to oil did not result assess individual food consumption. in treatment-related pathology or other observable abnormalities at necropsy (Dean et al., 2017a, this issue). While the results of this initial 2.3. Quarantine and daily monitoring oral dosing study were inconclusive, they did suggest potential adverse effects after short internal exposure time to polycyclic aromatic Birds were allowed to acclimate to captivity for a minimum of 21 hydrocarbons (PAHs) that are principle components of oil. days prior to initiation of the study. All individuals were inspected once The decision was made to modify dosing methods to increase the daily during quarantine and twice daily during the trial for signs of pain exposure time of birds to oil and to better emulate the field conditions. or distress. Distressed animals (e.g., those exhibiting lethargy, emacia- Birds can be exposed to oil via multiple paths following an oil spill. tion, persistent recumbency) were evaluated and either treated and Exposure routes include consumption of contaminated food, exposure retained in the study or euthanized. If an individual bird exhibited signs of skin, feathers and mucus membranes to oiled water and inhalation of of cold stress such as hunched posture and “fluffed” feathers, then heat volatiles as they weather. Externally oiled birds devote long periods of lamps were provided to all birds at that time. Use of heat lamps was time to removal of oil and maintenance of feather integrity, making recorded on health and daily monitoring sheets. All animal care and consumption of oil via preening an important route of exposure monitoring procedures were approved by the IACUC under NWRC (Leighton, 1993). protocol QA-2107 for the oral dosing study and QA-2326 for the The direct application of oil to feathers, while it mimics conditions external dosing study. Health assessment checklists were completed in the wild whereby birds are able to preen as they deem necessary, and daily. therefore consume oil throughout the day, has the disadvantage for toxicological studies that the dose consumed may vary between 2.4. Screening for abnormal subjects individuals. Oral dosing by gavage should allow better control of dose, but the method has the disadvantage that dose can only be delivered A wild population of any species contains a cross-section of age over a prescribed period, potentially limiting the absorption of oil by classes and individuals of variable health status. To avoid miscalculat- the gastrointestinal tract. As such, modifications were implemented to ing or overestimating oil toxicity effects, only birds assessed as healthy improve oral and external dosing of DCCOs to determine the utility of based on physical exam, body weight and appetite were included in the each method and whether the two types of dosing provided different internal dosing study. This study was also used to establish reference insight into the effects of oil on birds. In the first method, the oil dose intervals for hematological and biochemical analysis values which were was delivered to DCCOs in live food catfish, consumed throughout the used in the external dosing study to further screen study candidates. To day, rather than by bolus gavage. The second method involved determine which birds were healthy and which were abnormal, application of a calculated dose of oil to the DCCO feathers every three proposed reference intervals were established in accordance with days. American Society for Veterinary Clinical Pathology (ASVCP) guidelines, using the Reference Value Advisor and a more stringent setting of the 2. Materials and methods Dixon Test, using confidence levels of 0.1, or Tukey's Outlier Test (Geffre et al., 2011; Friedrichs et al., 2012). If outlying values were 2.1. Animal collection and husbandry found that indicated an abnormality according to these proposed reference intervals, all blood values for that bird were deleted from National Wildlife Research Center Mississippi Field Station (NWRC- the dataset. In the oral dosing study, there were six birds in each of the MSFS) staff conducted all bird capture, transport, quarantine, feeding three treatment groups that were considered to be representative of a and maintenance according to standard Institutional Animal Care and healthy population at the beginning of the study, based on hematology

20 F. Cunningham et al. Ecotoxicology and Environmental Safety 146 (2017) 19–28 and clinical chemistry endpoints. In the external dosing study, 25 birds were randomly assigned to the treatment or control groups (12 birds and 13 birds respectively).

2.5. Oil source

The oil used in both the oral dosing and external dosing studies was artificially weathered MC252 oil (DWH7937, batch# B030112) pre- pared from crude oil collected during the DWH oil spill (Forth et al., 2016).

2.6. Dosing Fig. 1. Determination of the mean surface area of DCCO carcasses used as the basis for estimating the area of bird requiring oil application. 2.6.1. Oral DCCOs were randomly assigned to one of three treatment groups: a feathered surface of the legs. Wings remained attached to the skins and control group (n=8) that was fed catfish that had been lightly severed from the body at the midpoint of the humerus. Skins were anesthetized and allowed to revive; a group dosed daily with up to spread out on a flat surface and photographed with a Nikon D200 5 mL oil/kg bw/day through provision of oil-containing catfish as digital camera. Images were imported into Image J (National Institutes described below (n=9); a group dosed daily with up to 10 mL oil/kg of Health). Total surface area was calculated based on the entire bw/day through provision of oil-containing catfish as described below feathered surface of the body. All non-feathered areas of the carcass (n=9). were excluded (face, beak, bare leg, and feet). Wings remained folded Fingerling catfish (approximately 7.5–10 cm) were lightly anesthe- in a relaxed position to duplicate the posture of oiled birds observed in tized using MS222 (tricaine methanesulfonate), given in an intraper- the field. The surface area of the body excluded the tail due to high itoneal (IP) injection of 2.0 mL of oil using a 20-gauge needle on a 25- variation in tail feathers among individuals, and no clear pattern in mL stainless steel/glass barrel syringe. Each catfish was injected with molting of retrices in DCCOs (Dorr et al., 2014). The mean surface area the same volume of oil to ensure that oil consumption per bird could be of DCCO carcasses was 1403.1 ± 86.5 cm2 (n=7), and was used as the calculated easily based on the number of catfish that were consumed. basis for estimating the area of bird that required oil application Injected catfish were placed into a separate holding tank to ensure (Fig. 1). revival from anesthesia and retention of oil. Prior to initiation of the external dosing study, oil was applied to Oil-injected catfish were fed to cormorants in their water-filled DCCO carcasses to determine the most efficient method for application foraging tanks at a rate that offered DCCOs either 5 or 10 mL oil/kg bw. of oil to their feathers. Coverage of 20% was targeted, which was Oil-injected catfish were observed to survive for more than 24 h if not equivalent to 280 cm2 of the surface area of a DCCO while in the resting consumed by birds while foraging. The method by which DCCOs or loafing position. This oiling rate (20%) is the high limit of the light- consume catfish usually involves capture and handling of the catfish oiling category (6–20%) used by the U.S. Fish and Wildlife Service for by hooking the end of the bill into the catfish gills. This often results in the DWH NRDA (Deepwater Horizon Natural Resource Damage death of the catfish due to damage to the gills. If the handled catfish Assessment Trustees, 2016). To achieve this coverage, oil was applied was dropped by the bird, it usually died and was not consumed. Thus, to the breast (140 cm2) and back (140 cm2) feathers. Plastic stencils, cages were checked once or twice each afternoon to replace any which measured 8×17.5 cm and 7×20 cm for the breast and back, uneaten dead catfish with an equal number of oil-injected live catfish. respectively, were used as a guide for application of oil. The total This method provided the greatest opportunity for DCCOs to have live weight of oil applied to each DCCO during each application totaled catfish available to them and encourage maximum consumption of oil- 13 g, approximately 6.5 g to the breast and 6.5 g to the back (approxi- injected catfish. mately 13 mL). Once a cormorant had consumed its allotment of oil-injected catfish, A total of 25 DCCOs allocated to two groups were used in this trial. a subsequent feeding of non-oil injected catfish was offered to each bird DCCOs were assigned to either control or 20% coverage oiling groups to achieve a possible daily intake of greater than 600 g catfish per based on blood samples collected at the initiation of the three-week cormorant. If birds did not consume the total quantity of injected quarantine period. Complete blood count values were used to ensure catfish, they were not offered additional non-oiled injected catfish. If an equal division of birds with potential health concerns between groups. individual bird was observed to eat all of the food provided on a given Monocyte counts greater than 2.0x109 cells/L were considered abnor- day, its food ration was increased the next day to ensure that ingestion mal (severe monocytosis). Additionally, a small oil spill took place on volume was not restricted up to the 600 g whole catfish maximum. The November 8, 2013 not far from where some of the DCCOs were amount of catfish consumed daily by each bird was recorded. collected, which could potentially have affected some plasma values In addition to internal dosing with oil, there were instances of associated with oxidative stress, or liver and kidney function. These inadvertent external oiling. Although all tanks were cleaned daily with birds were also evenly distributed between groups. 100% water replacement, oil was present in the tanks of all oil-treated The initial sample size of the control (no oil applied) and treated birds due to defecation by the birds during normal daily activities. No groups were 12 and 13 individuals, respectively. During the course of oil was purposefully placed on the integument of the DCCOs during this the trial, one bird from the control group and two birds from the study. The inadvertent presence of oil in the water and the subsequent treatment group died and were not replaced. Therefore, the final external oiling of the birds is somewhat representative of the multiple number of birds in both the control and treated groups was 11. The exposure routes expected to occur in the field. Data related to the amount of oil applied to each bird on the days of application was degree of external oiling (light or moderate) of each bird were recorded. determined by subtracting the weight of the beaker, oil and brush after application of oil from the initial weight. The control group had 6.5 g of 2.6.2. External water applied to the breast and another 6.5 g applied to the back to In order to estimate the body surface area prior to external dosing, ensure similar treatment and handling as oiled birds. Oil or water was DCCO carcasses (n=7) that ranged in weight from 1.6 to 2.4 kg were applied every three days through day 15 of the trial and cumulative oil skinned by making a ventral cut from head to vent. A cut was also made ingested was calculated based on a study by Hartung (1963). on the ventral side of each leg to allow for complete flattening of the

21 F. Cunningham et al. Ecotoxicology and Environmental Safety 146 (2017) 19–28

2.7. Blood sample collection the liver, kidney and GI tract, and the remaining thyroid and adrenal glands were flash-frozen in liquid nitrogen for subsequent analysis of In the oral dosing study a blood sample was taken from each bird on cytochrome P450 activity (Alexander, submitted for publication et al., day 0 (i.e., the day before oil dosing began) as a baseline comparison, this issue) and oxidative damage assessment (Pritsos et al., 2017, this and then once weekly until the study ended or the bird died or was issue). euthanized. In the external dosing study, all birds had a blood sample taken during quarantine to provide baseline data. Once dosing began, 2.9. Additional external dosing study methods blood was collected every six days (on days 0, 6, 12 and 18), prior to external application of oil. At the end of the 21-day trial, birds were All birds in the external dosing study had an Advanced Telemetry sampled for blood before euthanasia and necropsy. Systems (ATS, Isanti, MN USA) F1815T implantable very high fre- To collect the sample, the brachial vein was punctured with a 25- quency (VHF) temperature transmitter surgically implanted in the gauge, 19-mm butterfly needle attached to a 300-mm tube (600-µL coelom prior to study initiation for daily determination of internal volume). On day 0 only, the needle and tube were flushed with a 100- body temperature. External body temperature was taken just prior to international unit (IU) sodium heparin solution before use; thereafter, initial oiling and every six days thereafter (on days 3, 9, 15 and 20) with the needle and tube were flushed with a 100-IU lithium heparin a handheld scanning thermograph camera (FLIR ThermoCAM T640; solution. Approximately 3–4 mL of blood were collected and trans- FLIR Systems, Boston, MA, USA) (Mathewson, submitted for publica- ferred to labeled lithium heparin Vacutainer™ tubes and kept on ice for tion et al.,). Echocardiograms were used to evaluate DCCO cardiac subsequent processing. In the external dosing study, serum samples structure and function (Harr et al., 2017c, this issue). (1 mL) were collected from the brachial vein first using syringes and tubes containing no anticoagulant. Following the 1 mL collection, 2.10. Statistical methods heparinized syringes were used to collect approximately 10 mL of blood from the jugular vein that was divided for hematology and The following statistical methods were used in the present studies in biochemical analyses. addition to other studies in this issue. Oxidative stress endpoints, 3- Each blood sample collected during the oral and external dosing methyl histidine, and organ weights were compared by one-way studies was aliquoted for the following assessments: Heinz bodies, analyses of variance (ANOVAs). Statistical significance was assessed CBCs, electron microscopy, PCV, hemoglobin concentration (Harr et al., based on p-values determined for each distinct ANOVA model (i.e., 2017a, this issue), total antioxidant capacity (Pritsos et al., 2017, this there was no attempt to control for experiment-wise Type I error). issue), and plasma clinical chemistries (Dean et al., 2017b, this issue). Hematologic and plasma clinical chemistry values, FLIR data, body In the external dosing study, a whole blood sample was collected using weight data, and body temperature data collected across multiple time no anticoagulant to determine activated clotting time. A glass tube points were compared using linear mixed effects regression models with containing 3 mg of diatomaceous earth was incubated at 37 °C prior to a repeated measures structure, where treatment was modeled as a fixed the addition of blood. A timer was started when 0.5 mL of whole blood effect and the individual bird within elapsed days was modeled as a was transferred to the tube, which was then inverted to several times to random effect. Regression models included effects for elapsed days, mix the diatomaceous earth with the whole blood. The tube was treatment, and a treatment*days interaction term. For the oral dosing incubated for 30 s at 37 °C and visually examined for microclot study, elapsed days and treatment (oil dose level as mL/kg bw/day) formation by inverting the tube. This last step was repeated until clots were modeled as continuous variables, where treatment was defined as formed. The time that elapsed from introduction of blood into the tube the average daily consumption determined from daily observations of until evidence of the first clot was activated clotting time. actual oil consumption by each individual bird. Graphical inspection of data distributions using boxplots and scatterplots overlaid with fitted 2.8. Necropsy regression models indicated that data for endpoints analyzed using ANOVA or regression were generally symmetric about their means and Depressed birds were euthanized, following multiple consultations did not span more than an order of magnitude; thus, transformations with the onsite veterinarian. A depressed bird was one that tucked its were deemed to be unnecessary. Differences among treatment groups head under its wings, was lethargic and unresponsive, experienced on day 0 were evaluated using Kruskal-Wallis test. weight loss, and had a cloacal temperature of 39.4 °C or less. Box plots were used to provide a quick visual summary of distribu- After birds were euthanized, they were necropsied as soon as tions for endpoints not analyzed over time and for endpoints for which possible to ensure fresh tissues adequate for analysis. The remaining the regression model was not appropriate, illustrating the range, shape, birds were necropsied on the last day of the respective studies. Birds and extremes of the distributions. Side-by-side box plots allow for a were weighed and euthanized by cervical dislocation according to visual comparison of these characteristics across treatment groups. IACUC-approved protocols. Necropsy blood samples were obtained by Calculations were performed using TIBCO Spotfire S-PLUS 8.2 for direct cardiac puncture with a non-heparinized needle (20-gauge, 25.4- Windows. mm) and 10-mL syringe for serum samples, followed by subsequent heparinized (100-IU lithium heparin) needles (20-gauge, 25.4-mm) and 3. Results 10-mL syringes for whole blood and plasma samples. An attempt was made to collect 30 mL of blood from each bird for a larger suite of Results associated with endpoints assessed in the oral and external endpoints than those evaluated at day 0. dosing studies with DCCOs that are not covered in the present report All organs were assessed grossly for abnormalities and documented include oil-induced increases in CYP1A protein expression and catalytic by digital images. The brain, heart, lungs, kidneys, thyroid gland, liver, activity (Alexander, submitted for publication et al., this issue), changes gastrointestinal (GI) tract, spleen, and adrenal glands were collected in CBC estimates and plasma chemistry and electrophoresis endpoints and weighed to the nearest mg. In addition, if gonads, thymus, and (Dean et al., 2017b, this issue), hemolytic anemia as indicated by bursa were present, they were collected and weighed. decreased PCV, relative reticulocytosis with an inadequate regenerative Organs were placed in an appropriately labeled specimen jar response, and presence of Heinz bodies (Harr et al., 2017a, this issue), containing 10% neutral buffered formalin; one adrenal and one thyroid and increases in liver and kidney weights and the presence of lesions in gland were each placed in individual micro-centrifuge tubes, also kidney, liver, heart and thyroid gland (Harr et al., 2017b, this issue). containing 10% neutral buffered formalin, for subsequent histopatho- Results related to a decrease in cardiac systolic function and internal logical assessment (Harr et al., 2017b, this issue). Two samples each of body temperature and heat loss in externally dosed DCCOs are

22 F. Cunningham et al. Ecotoxicology and Environmental Safety 146 (2017) 19–28

Table 1 of the study. The 5 mL/kg bw group increased intake until day 9 when Effect of daily oral dosing with artificially weathered MC252 oil on food intake and oil consumption peaked at 558 g of catfish consumed per day and then ingestion by double-crested cormorants (Phalacrocorax auritus). decreased reaching the lowest average consumption, 233 g of catfish, fi Control 5 mL/kg bw 10 mL/kg bw by day 12. Consumption rebounded to 601 g of cat sh by day 21 to approach control levels of intake. The 10 mL/kg bw group had a Mean SE Mean SE Mean SE reduced intake from day 1 and peaked on day 9 at 330 g of catfish per day. This treatment group was terminated on day 14 due to mortality Days 1–7 n6 8 6 and health of the remaining DCCOs in the group. Oil ingestion 0 0 5.3 0.1 8.8 0.4 All dosed and control groups experienced a similar decrease in body Food intake 361.8 38.9 290.9 26.7 267.2 17.8 weight over the 21-day trial. Body weight at necropsy was not Days 8–14 significantly different among treatment groups (1801.2 ± 63.5 g, n6 7 5 1680.8 ± 33.5 g, 1565.5 ± 75.7 g for the control, 5 mL/kg bw and Oil ingestion 0 0 5.1 0.1 8.5 0.3 10 mL/kg bw treatment groups, respectively). Food intake 574.1 13.0 398.4 21.1 211.6 18.4

Days 15–20 n6 7 . 3.1.2. Clinical signs and mortality Oil ingestion 0 0 5.2 0.2 . . Food intake 545.1 30.8 431.0 37.1 . . Of the 26 adult, mixed-sex DCCO used in this study, 16 were euthanized on day 21. A total of 10 treated DCCOs died or were All daysa euthanized within 17 days of the start of the study. Prior to day 21, n6 7 5 Oil ingestion 0 0 5.2 0.3 8.4 0.9 double crested cormorants were euthanized based on veterinary Food intake 484.9 58.5 367.7 58.5 247.5 33.0 assessment of severe distress, or when the animals were moribund, to ensure that necropsies could be performed on fresh carcasses and that a a – Days 1 11 for birds in 10 mg/kg bw group. complete suite of endpoints could be sampled. All birds in the 10 mL oil/kg bw group died or required euthanasia before the end of the presented in Harr et al. (2017c) and Mathewson et al. (2017), study. One DCCO in the 5 mL oil/kg bw group and 0 control birds died respectively. Results related to changes in oxidative stress endpoints before the end of the study. in orally dosed DCCOs are presented in (Pritsos et al., 2017, this issue). Clinical signs of toxicity in birds orally dosed with oil included reduced cloacal temperature, lower body weight, lethargy, feather 3.1. Oral damage, morbidity, and death. These signs were first observed on day 5 in birds dosed with 10 mL oil/kg bw and on day 7 in birds dosed with 3.1.1. Oil intake, food consumption and body weight 5 mL oil/kg bw (Table 2). There was a reduced ability of oiled birds to Mean oil consumption by treatment is presented in Table 1. The control their cloacal temperatures. The cloacal temperatures of the 5 mL/kg bw treatment group averaged 5.2 mL during the course of the control group were consistent during the study, ranging from 41.2 to study and the 10 mL/kg bw group never reached the target dose and 43.0 C. Cloacal temperatures of the 5 mL oil/kg bw group ranged from averaged 8.4 mL/kg bw due to reduced intake of oil-injected catfish. 39.5 to 41.7. The 10 mL oil/kg bw group's cloacal temperatures ranged The control birds increased food intake until day 8 when consump- from 40.0 to 41.2 on day 7 of the study to 38.3–39.7 C just prior to their tion peaked at an average of 621 g of catfish consumed per day and death or euthanasia. Abnormal excreta containing blood (hematoche- then ranged from 518 to 606 g of catfish consumed per day for the rest zia) was first observed in the 10 mL oil/kg bw birds on day 8 and in the

Table 2 Average and cumulative oil consumed and the progression of clinical signs and mortality in double-crested cormorants orally exposed to oil.

Day 5 mL oil/kg bw 10 mL oil/kg bw Dose Event Total oil at death

n Daily oil Cumulative oil intake n Daily oil Cumulative oil intake intake intake

1 9 8.8 8.8 9 13.4 13.4 Study initiation: first day of oil exposure 2 9 8.9 17.7 9 15.3 28.7 3 9 8.7 26.4 9 13.3 42.0 4 9 10.0 36.4 9 13.3 55.3 5 9 9.8 46.2 9 14.5 69.8 10 1 died 40 mL 6 9 10.0 56.2 8 16.3 86.1 7 8 10.0 66.2 8 14.0 100.1 5, 10 Lethargic, reduced body temp; 1 5 mL/kg bw died 120 mL 8 8 8.0 74.2 8 13.8 113.9 9 8 8.3 82.5 8 15.3 129.2 10 Lethargic, abnormal excreta, reduced body temp 10 8 8.3 90.8 7 15.7 144.9 10 Lethargic, abnormal excreta 11 8 8.3 99.1 7 14.0 158.9 10 Lethargic; 1 died 12 8 8.3 107.4 5 14.4 173.3 5, 10 Lethargic, abnormal excreta, reduced body temp; 139 mL 2−10 mL oil/kg bw euthanized 13 8 8.3 115.7 2 10.0 183.3 5, 10 Lethargic, reduced body temp 207 mL 14 8 8.3 124.0 2 10.0 193.3 5, 10 Lethargic, abnormal excreta, reduced body temp; 182 mL 1−10 mL oil/kg bw euthanized 15 8 8.3 132.3 10 Lethargic, reduced body temp; 2−10 mL oil/kg bw 223 mL euthanized 16 8 8.0 140.3 5, 10 Abnormal excreta 17 8 8.3 148.6 5, 10 Abnormal excreta 18 8 8.3 156.9 5, 10 Abnormal excreta 19 8 8.3 165.2 Control DCCO abnormal excreta 20 8 8.3 173.5 21 8 8.3 181.8 Necropsy remaining birds

23 .Cniga tal. et Cunningham F.

Table 3 Cumulative oil applied, estimated oil consumed and the progression of clinical signs and mortality in double-crested cormorants exposed to external oiling.

Study Day Cumulative oil applied Estimated cumulative oil ingested Treatment Event (mL) (mL)A Control Oiled n birds affected n birds affected

0 13 0 1 0 Study initiation: first day of exposure to oil; one control bird died prior to water application 1 3 0 13 Visible oil on feathers 2 3 0 13 Visible oil on feathers 3 26 4 0 13 Feathers matted and rough 4 7 0 13 Feathers matted and rough 5 8 0 13 Feathers matted and rough 6 39 9 0 13 Feathers matted and rough, feather damage observed, skin noticeably discolored 7 13 0 13 Feathers matted and rough, skin noticeably discolored 8 15 0 13 Feathers matted and rough, skin noticeably discolored 9 52 16 0 13 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating 10 20 0 13 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating 11 21 0 13 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating 12 65 22 0 13 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity 24 deteriorating (4 birds with abnormal excreta containing RBCs) 13 26 0 13 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating (6 birds with abnormal excreta containing RBCs) 14 28 0 13 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating (2 birds - feather plucking observed in pens, 7 birds with abnormal excreta containing RBCs) 15 78 29 0 12 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating (10 birds - feather plucking, 1 oil treated bird mortality) 16 33 0 12 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating, feather plucking 17 34 0 12 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating, feather plucking 18 35 0 12 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating, feather plucking (4 birds lethargic) Ecotoxicology andEnvironmental Safety146(2017)19–28 19 36 0 11 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating, feather plucking (1 oil treated bird mortality) 20 37 0 11 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating, feather plucking 21 38 0 11 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating, feather plucking (7 birds with abnormal excreta containing RBCs); necropsyB 22 38 0 11 Feathers matted and rough, skin noticeably discolored, oil covering much of surface area of birds, feather integrity deteriorating, feather plucking; necropsyC

A Calculation based on Hartung (1963), % ingestion of oil =1.0409x +0.1731. B Eleven DCCO necropsied on day 21. C Eleven DCCO necropsied on day 22. F. Cunningham et al. Ecotoxicology and Environmental Safety 146 (2017) 19–28

Fig. 2. Treated (dosed) and control group DCCOs daily feeding (g). Fig. 3. Treated (dosed) and control group DCCOs changes in body weights over time during the 21-day treatment period. 5 mL oil/kg bw birds on day 13. During blood collections, it was noted that in oil-dosed DCCOs the time required for blood to clot was longer very dark in color. The kidney appeared mottled. There was frothy fluid compared to controls. Upon necropsy of birds in the 5 mL oil/kg bw on the surface of the right lung and there were petechial hemorrhages group, the remaining blood was pooling and not forming clots, whereas on the cerebral hemispheres, cerebellum and brain stem. clot formation was observed in all control birds. A number of clinical signs were observed in birds externally dosed with oil including deterioration of feather integrity, abnormal feces, 3.2. External excessive preening, feather plucking, and lethargy (Table 3). Feathers of all oiled birds appeared matted and rough by day 3. As the study fi 3.2.1. Oil intake, food consumption and body weight progressed, two DCCOs rst began to pluck feathers on day 14 of the Hartung and Hunt (1966) estimated that a duck with 7 g of oil on its study and by day 16 all oiled birds were engaged in this activity. The feathers would ingest approximately 1.5 g of oil on the first day of oil skin on the breast and back of some oiled birds was noticeably exposure. Based on previous studies (Hartung, 1963), it was estimated discolored by day 6 and by day 9, oil covered much of the surface area that DCCOs would ingest 21% of the oil applied to their feathers within of all birds and the integrity of the feathers was deteriorating. Plucked the first day and 50% by day 8. In the present study, based on external feathers were largely downy material, but also included some contour application of 13 g oil to DCCO every three days over 15 days, a feathers from the breast and back. Skin in those areas was observed to cumulative ingestion of approximately 38 g of oil was estimated have become thickened, discolored and irritated. Abnormal excreta was (Table 3). observed in four oiled birds beginning on day 12 and, by the end of the Food consumption had a significant elapsed day (p=0.006) and trial, seven of 11 oiled birds that survived to necropsy had abnormal treatment*day interaction (p < 0.001) (Fig. 2). Birds treated with oil feces. Microscopic examination of opportunistic fecal samples revealed consumed more fish than control birds during the trial period beginning large numbers of RBCs, documenting GI hemorrhage. Only one control on day 7 of the trial. Average food consumption over the 21-day trial bird had abnormal excreta that consisted of green diarrhea with no for the control group was 403 g compared to 421 g in the oiled group. evidence of gelatinous protein or blood as noted in the oiled birds. Six From day 1 through day 6 of the trial, mean daily food consumption of the 11 birds that survived to necropsy were described as being was 410 g for the control group and 337 for the oiled group. From day 7 lethargic within four days of necropsy. through day 20 of the trial, daily food consumption averaged 399 g for the control group and 464 g for the oiled birds. 3.2.2.1. Blood coagulation. It was observed during the oral dosing study There was a significant treatment group effect (p=0.02), elapsed that blood in oil-dosed birds clotted very slowly or did not clot during day effect (p < 0.001), and treatment*day (p=0.04) interaction with necropsy. The decision was made to examine this apparent clotting respect to changes in DCCO body weights during the trial period. Body dysfunction at necropsy of externally oiled birds by determination of weights increased over the trial period in both oil treated birds and activated clotting time and time required for untreated blood to clot. control birds; however, birds treated with oil gained weight at a faster Clotting dysfunction was evidenced by a significant increase rate than control birds (Fig. 3). (p < 0.001) in activated clotting time in oiled birds (359 ± 90.5 s) compared to controls (172 ± 21.5 s) at necropsy. Blood of control birds 3.2.2. Clinical signs and mortality clotted within three to four minutes compared to no clotting after eight Three DCCOs died after the study was initiated. A control bird died minutes for several treated birds. on day 0 immediately after being sampled for blood. A dosed bird was found dead on day 14 and a second dosed bird was found dead on day 4. Discussion 18, both from unknown causes that may or may not have been related to oil exposure. Upon necropsy of the treated bird that died on day 14, Modifications in the oral dosing technique for DCCOs described here there were fish in the esophagus and stomach, indicating that the bird resulted in development of clinical signs and changes in a number of had eaten recently. The heart appeared slightly flaccid after removal. hematological, biochemical, and tissue endpoints consistent with Unclotted blood was present in the body cavity. Bile collected from the petroleum intoxication that are presented and discussed in other reports gall bladder was hemolytic. Packed cell volume determined from a comprising this special issue (Harr et al., 2017a, 2017b, Pritsos et al., blood sample collected from the heart ranged from 26% to 28%. Upon 2017, Alexander, submitted for publication et al., Dean et al., 2017b). necropsy of the bird that died on day 18, there were petechial The hemolytic anemia (Harr et al., 2017a, this issue) and changes hemorrhages on the heart and pancreas and there were hemorrhages indicative of liver and kidney damage (Harr et al., 2017b, this issue) in the omentum in the region of the small intestine. There were multiple observed in the oral study have been reported in other studies assessing beige focal lesions located on the surface of the liver and the bile was the effects of oil exposure on avian species. However, these results are

25 F. Cunningham et al. Ecotoxicology and Environmental Safety 146 (2017) 19–28 specific for the source of oil, the species, and age class of the birds, the trial, oiled birds consumed 38% more food compared to average intake dosing methodology, and the husbandry employed, and should be used over the first six days while control birds consumed 3% less food. For with caution. New findings from the oral dosing study included externally oiled birds, the expected range of food consumed on a daily cardiovascular abnormalities documented upon gross necropsy that basis was 356–407 g based on initial mean body weight of 1697 g and prompted further diagnostic evaluation in the external dosing study for controls the range was 382–437 g based on initial mean body weight (Harr et al., 2017c). Evidence of coagulopathy found on gross necropsy of 1821 g. Over the 21-day trial, control food consumption was within has not been reported before in oil-dosed birds and has been minimally the expected range and oiled bird food consumption was 3–18% greater investigated using MC252 oil in any species. than the predicted range of food consumed. From day 1 through day 6 The methods developed in the external dosing study of DCCO with of the trial, mean daily food consumption of control birds was within artificially weathered MC252 oil, also resulted in clinical signs and the expected range of food consumed and 5–17% less for the oiled changes in a number of endpoints consistent with petroleum intoxica- group. From day 7 through day 20 of the trial, daily food consumption tion. Hemolytic anemia and clotting dysfunction (Harr et al., 2017a, continued to be within the expected range of food consumed for the this issue), cardiac abnormalities (Harr et al., 2017c, this issue), and control group but was 14–30% greater than the predicted range for the changes indicative of liver and kidney damage (Harr et al., 2017b, this oiled birds. The greater food consumption in externally dosed birds issue) were similar to those observed in the DCCO oral dosing study, suggests metabolic compensation for thermal stress (Mathewson, sub- indicating that the method of oil application to DCCO feathers is an mitted for publication et al.,). Infrared thermal images indicated appropriate means of assessing oil toxicity in this species. significantly greater surface temperatures and heat loss in oiled birds following external oil applications, but measurements collected by 4.1. Oil intake, food consumption and body weight coelomically implanted temperature transmitters showed that internal body temperatures were stable over the course of the study period The mean daily intake of oil by orally dosed DCCOs over the course (Mathewson, submitted for publication et al.,). of the present 21-day study was a reflection of food consumption in In the oral exposure study, mean body weight of control and 5 mL each group. Those birds in the 5 mL oil/kg bw group consumed an oil/kg bw birds at day 21 was 0.6% and 9.2% less, respectively, average of 5.2 mL oil/kg bw per day, which was slightly more than the compared to mean body weight at day 0 and mean body weight of the targeted dose. Birds in the 10 mL oil/kg bw group consumed an average 10 mL oil /kg bw birds at day 14 was 11.4% less compared to day 0 of 8.5 mL oil/kg bw per day, which was 15% below the targeted dose mean body weight. Although body weights were not significantly and is a reflection of the 19% decrease in food consumption compared different among treatments at necropsy, the loss of body weight over to the quarantine period. The total quantity of oil ingested by the nine time in the birds dosed with oil was a reflection of the decrease in food DCCOs in the 10 mL/kg bw group that died during the course of the consumption. Herring gull (Larus argentatus) nestlings orally dosed with trial ranged from 40 (day 5) to 223 mL (day 15). 10 mL of Prudhoe Bay crude oil/kg bw/day or more consumed less food In the external dosing study, a cumulative ingestion of 38 g of oil beginning on the third day of dosing compared to controls and began to through preening was estimated based on Hartung (1963). This lose weight on the fifth day of dosing (Leighton, 1986). estimate is close to the minimum quantity of oil associated with Mean body weight of externally dosed birds increased by 10.3% mortality in the orally dosed DCCOs. over the 21-day trial (1697 g vs 1872 g) while mean body weight of Birds orally dosed with oil consumed less food than control birds, control birds increased 3.7% (1821 g vs 1889 g). The increase in body whereas food consumption of externally oiled birds was greater weight of oiled birds is a reflection of the increase in food consumption, compared to controls. Mean food consumption during the oral dosing which, as stated above, is thought to be metabolic compensation for study for the 5 mL oil/kg BW group was 76% of control consumption; thermal stress experienced by these birds (Mathewson, submitted for food consumption for the 10 mL oil/kg BW group was 51% of control publication et al.,). consumption. Mean daily food consumption for all three groups for the first five days of the trial (control =345 g/day, 5 mL oil/kg bw 4.2. Clinical signs and mortality =284 g/day, 10 mL oil/kg bw =279 g/day) was similar to the mean consumption reported for the quarantine period (298 g/day). After day Necropsy revealed underlying disease in dosed and control birds in 5, mean daily consumption of the 10 mL oil/kg bw birds for the both studies, including bacterial infection, which was not cultured or remainder of the trial (225 g/day) decreased by 19%, while average further elucidated. Disease may have contributed to some observed daily food consumption for the control and 5 mL oil/kg bw bird groups clinical signs. However, because disease was present in the control (559 g/day and 403 g/day, respectively) increased by 62% and 42%, population, the only variable to explain the differences in mortality and respectively. Using the mean body weight of DCCOs at the beginning of morbidity between oiled and control birds is oil intoxication. the oral dosing study (1728 g for the control; 1677 g for the 5 mL oil/kg In the oral dosing study, all of the 10 mL oil/kg bw birds died or bw; and 1657 g for the 10 mL oil/kg bw groups), and assuming that were euthanized before the end of the 21-day study, while one bird in food consumption is 21–24% of bw (Glahn and Brugger, 1995), the the 5 mL oil/kg bw group died. In the external dosing study, two oiled expected range of food consumed on a daily basis for the control, 5 mL birds died prior to the end of the study. Orally dosed birds that died or oil/kg bw, and 10 mL oil/kg bw groups was 363–415 g, 352–402 g, and were euthanized before the end of the study displayed clinical signs that 348–398 g, respectively. The actual percent mean food consumption included lethargy, reduced food consumption, hypothermia, loss of over the specified time period for the control, 5 mL oil/kg bw, and body weight, and abnormal excreta. Behaviorally, birds in the 10 mL 10 mL oil/kg bw groups was 35–54% greater, 0–14% greater, and oil/kg bw group held their heads under their wings, regardless of 35–43% less than the predicted food consumption, respectively. Similar stimulation, while control birds were bright, alert, and responsive. In to results reported here, herring gull (Larus argentatus) nestlings orally the external dosing study, oiled birds also were lethargic and had dosed with 10 mL of Prudhoe Bay crude oil/kg bw/day or more abnormal excreta, but they did not experience hypothermia, reduced consumed less food beginning on the third day of dosing (Leighton, food consumption or loss of body weight. Rather, the clinical signs 1986). It is probable that food aversion and gastrointestinal distress displayed by externally oiled birds were primarily related to feather and induced by multi-organ system dysfuction were contributory mechan- skin integrity, as would be expected. The difference in mortality and isms to the decreased food consumption in birds orally dosed with oil in apparent severity of clinical signs between the studies could be a the present study. reflection of the amount of oil ingested during the course of the study. It Externally oiled birds consumed more food compared to controls was calculated that externally oiled birds ingested less oil during the after the second application of oil. From day 7 through day 20 of the 21-day trial (38 g), when compared to orally dosed birds that consumed

26 F. Cunningham et al. Ecotoxicology and Environmental Safety 146 (2017) 19–28 between 40 and 223 g of oil prior to death or necropsy. helped carry out each study. Appreciation is also expressed to Dr. Tacy Mortality, weight loss, and some of the clinical signs reported here Rupp of Veterinary Cardiopulmonary Care Center in West Palm Beach, have been reported in other oral dosing studies. Pekin ducks fed diets Florida for developing a method for species-specific DCCO echocardio- that provided approximately 2.9 mL of South Louisiana crude oil or graphy. Funding for these studies was provided by the US Fish and 1.3 mL of No. 2 fuel oil/kg bw/day experienced 22% and 36% Wildlife Service. as part of the Deepwater Horizon Natural Resource mortality, respectively, at the end of 50 days at an ambient temperature Damage Assessment. of 27 °C; and 67% and 43% mortality, respectively, at the end of a second 50-day period at an ambient temperature of 3 °C (Holmes et al., References 1978). Ducks fed a diet that provided 2.5 mL of Kuwait crude oil/kg bw per day had no mortality at 27 °C but 67% mortality at 3 °C (Holmes Alexander, C.R., Hooper, M.J., Cacela, D., Smelker, K.D., Dean, K.M., Bursian, S.J., et al., 1978). Herring gull chicks administered daily oral doses of 10 or Cunningham, F.L., Hanson-Dorr, K.C., Horak, K.E., Isanhart, J.P., Shriner, S.A., Godard-Codding, C.A.J., 2017. CYP1A protenin expression and catalytic activity in 20 mL Prudhoe Bay crude oil/kg bw were lethargic on and after day 4 Double-crested cornorants experimentally exposed to Deepwater horizon Mississippi and began to lose weight on day 5 (Miller et al., 1978). Two gull chicks Canyon 252 oil. Ecotoxicol. Environ. Saf (submitted for publication, this issue). in the 20 mL oil/kg bw per day group were moribund on day 5 and Alonso-Alvarez, C., Munill, I., Lopez-Alonso, M., Velando, A., 2007b. Sublethal toxicity of the Prestige oil spill on yellow-legged gulls. Environ. Int. 33, 773–781. subsequently euthanized (Leighton, 1986). In the same study, Atlantic Brausch, J.M., Blackwell, B.R., Beall, B.N., Caudillo, C., Kolli, V., Godard-Codding, C., puffin(Fratercula arctica) nestlings that were administered daily oral Cox, S.B., Cobb, G., Smith, P.N., 2010. Effect of polycyclic aromatic hydrocarbons in doses of 10 mL oil/kg bw were weak by day 4 and two puffins died by northern bobwhite quail (Colinus virginianus). J. Toxicol. Environ. Health Part A 73, – day 5. Two of 16 American kestrels administered feed containing 3.0% 540 551. Burger, A.E., 1993. Estimating the mortality of seabirds following oil spills: effects of oil oil from the Mexican Ixtoc I well blowout died after two weeks on volume. Mar. Pollut. Bull. 26, 140–143. treatment when there was a 24 °C drop in ambient temperature (Pattee Bursian, S.J., Alexander, C.R., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Ellis, and Franson, 1982). Both birds had lost weight (21% and 31%) at the C.K., Godard-Codding, C.A., Guglielmo, C.G., Hanson-Dorr, K.C., Harr, K.E., Healy, K. A., Hooper, M.J., Horak, K.E., Isanhart, J.P., Kennedy, L.V., Link, J.E., Maggini, I., time they died; death was attributed to weight loss followed by cold Moye, J.K., Perez, C.R., Pritsos, C.A., Shriner, S.A., Trust, K.A., Tuttle, P.L., 2017. stress (Pattee and Franson, 1982). Because ambient temperatures in the Overview of avian aoxicity studies for the Deepwater Horizon Natural Resource present studies were cooler than those recorded in the Gulf of Mexico at Damage Assessment. Ecotoxicol. Environ. Saf. (submitted, this issue). Cavanaugh, K.P., Goldsmith, A.R., Holmes, W.N., Follett, B.K., 1983. Effects of ingested the time of the oil spill, and orally dosed birds had depressed body petroleum on the plasma prolactin levels during incubation and on the breeding temperatures, heat lamps were added to all cages to supply additional success of paired mallard ducks. Arch. Environ. Contam. Toxicol. 12, 335–341. heat. Regardless of heat lamp placement, significant and apparent dose- Cavanaugh, K.P., Holmes, W.N., 1987. Effects of ingested petroleum on the development of ovarian endocrine function in photostimulated mallard ducks (Anas platyrhynchos). dependent mortality occurred. As the temperature and heat lamp Arch. Environ. Contam. Toxicol. 16, 247–253. placement were identical in cages of both control and orally dosed Dean, K.M., Cacela, D., Carney, M.W., Cunningham, F.L., Ellis, C., Gerson, A.R., birds, any significant changes in endpoints were because of oil Guglielmo, C.G., Hanson-Dorr, K.C., Harr, K.E., Healy, K.A., Horak, K.E., Isanhart, J.P., Kennedy, L.V., Link, J.E., Lipton, I., McFadden, A.K., Moye, J.K., Perez, C.R., exposure. Hypothermia and thermal stress were considered to be a Pritsos, C.A., Pritsos, K.L., Muthumalage, T., Shriner, S.A., Bursian, S.J., 2017a. significant component of mortality in the present oral dosing study. Testing of an oral dosing technique for double crested cormorant, Phalacrocorax In the oral dosing study, all oil-dosed birds had oil on the plumage auritus, laughing gull, Leucophaeus atricilla, homing pigeon, Columba livia, and fi as a result of foraging for fish in the feed tanks that contained oil western sandpiper, Calidris mauri, with arti cially weather MC252 oil. Ecotoxicol. Environ. Saf (submitted for publication, this issue). excreted by the birds. External oiling can result in physical alteration of Dean, K.M., Bursian, S.J., Cacela, D., Carney, M.W., Cunningham, F.L., Dorr, B.S., the feathers, causing matting and loss of insulation and water-repellent Hanson-Dorr, K.C., Healy, K., Horak, K., Link, J.E., Lipton, I., McFadden, A.K., properties. The loss of insulation and water-repellency can lead to death McKernan, M.A., Harr, K.E., 2017b. Changes in white cell estimates and plasma chemistry measurements following oral or external dosing of Double-crested of oiled birds as a result of heat loss and drowning (Leighton, 1993). In cormorants, Phalacocorax auritus, with artificially weather MC252 oil. Ecotoxicol. the external dosing study, disruption of feather and skin integrity was Environ. Saf (submitted for publication, this issue). the most pronounced clinical sign. The oiled birds engaged in feather Deepwater Horizon Natural Resource Damage Assessment Trustees, 2016. Deepwater Horizon oil spill: final Programmatic Damage Assessment and Restoration Plan and plucking, which was not unexpected because it has been reported in Final Programmatic Environmental Impact Statement. Retrieved from http://www. other birds following oil spills (Snyder et al., 1973). Feather plucking gulfspillrestoration.noaa.gov/restoration-planning/gulf-plan/. will naturally result in greater total heat loss from the birds than Dorr, B.S., Hatch, J.J., Weseloh, D.V., 2014. Double-crested cormorant (Phalacrocorax auritus). In: Poole, A. (Ed.), The Birds of North America Online. Ithaca: Cornell expected. Laboratory of Ornithology; The Birds of North America Online. 〈http://bna.birds. The exposure methods used in the oral and external dosing studies cornell.edu/bna/species/441〉. resulted in similar signs of oil toxicity and can serve as appropriate Eastin, W.C.J., Rattner, B.A., 1982. Effects of dispersant and crude oil ingestion on mallard ducklings (Anas platyrhynchos). Bull. Environ. Contam. Toxicol. 29, 273–278. means of assessing oil toxicity in DCCOs. Birds orally dosed with oil Forth, H.P., Mitchelmore, C.L., Morris, J.M., Lay, C.R., Suttles, S.E., Lipton, J., 2016. were more severely affected based on the timeline, type and magnitude Characterization of dissolved and particulate phases of water accommodated of clinical signs and mortality. In both studies, birds consumed oil fractions used to conduct aquatic toxicity testing in support of the Deepwater horizon either as a result of eating oil-contaminated food (oral) or preening natural resource damage assessment. Environ. Toxicol. Chem(http://dx.doi.org/ http://onlinelibrary.wiley.com/10.1002/etc.3672). (external). Additionally, birds in both studies experienced feather Friedrichs, K.R., Harr, K.E., Freeman, K.P., Szladovits, B., Walton, R.M., Barnhart, K., damage as a result of intentional or unintentional external oiling. Blanco-Chavez, J., 2012. ASVCP reference interval guidelines: determination of de Orally dosed birds had difficulty maintaining internal body temperature novo reference intervals in veterinary species and other related topics. Vet. Clin. Pathol. 41, 441–453. as compared to externally doses birds that were losing heat but Geffre, A., Concordet, D., Braun, J.P., Trumel, C., 2011. Reference value Advisor: a new maintaining core temperature. It is possible that the greater mortality freeware set of macroinstructions to calculate reference intervals with Microsoft – in orally dosed birds was because of hypothermia, which in turn can be Excel. Vet. Clin. Path 40, 107 112. – Glahn, J.F., Brugger, K.E., 1995. The impact of double-crested cormorants on the attributed to a combination of ingestion of 40 233 mL oil that induced Mississippi Delta catfish industry: a bioenergetics model. Colonia. Waterbirds 18, toxicity and unintentional external oiling that exacerbated heat loss. 158–167. While externally dosed birds also ingested oil, the estimated amount Harr, K.E., Cunningham, F.L., Pritsos, C., Pritsos, K., Muthumalage, T., Dorr, B.S., Horak, K.E., Hanson-Dorr, K.C., Dean, K.M., Cacela, D., McFadden, A.K., Link, J.E., Healy, (38 mL) was at the lower end of the range associated with mortality. K.A., Tuttle, P., Bursian, S.J., 2017a. Weathered MC252 crude oil-induced anemia and abnormal erythroid morphology in double-crested cormorants (Phalacrocorax Acknowledgements auritus) with light microscopic and ultrastructural description of Heinz bodies. Ecotoxicol. Environ. Saf (submitted for publication, this issue). Harr, K.E., Bursian, S.J., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Hanson- Appreciation is expressed to the technicians of the National Wildlife Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Reavill, D.R., Shriner, S.A., Schmidt, Service's Mississippi Field Station, Paul Fioranelli, Alex Crain, Lanna R.E., 2017b. Comparison of organ weights and histopathology between double- fi Durst and Raleigh Middleton, who participated in DCCO capture and crested cormorants (Phalacrocorax auritus) dosed orally or externally with arti cially

27 F. Cunningham et al. Ecotoxicology and Environmental Safety 146 (2017) 19–28

weathered Mississippi Canyon 252 crude oil. Ecotoxicol. Environ. Saf (submitted for puffins that ingested Prudhoe Bay crude oil. Vet. Pathol. 23, 254–263. publication, this issue). Leighton, F.A., 1993. The toxicity of petroleum oils to birds. Environ. Rev. 1, 92–103. Harr, K.E., Rishniw, M., Rupp, T.L., Cacela, D., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C., Leighton, F.A., Lee, Y.Z., Rahimtula, A.D., O'Brien, P.J., Peakall, D.B., 1985. Biochemical Healy, K., Horak, K., Link, J.E., Reavill, D.R., Bursian, S.J., Cunningham, F.L., 2017c. and functional disturbances in red blood cells of herring gulls ingesting Prudhoe Bay Dermal exposure to weathered MC252 crude oil results in echocardiographically crude oil. Toxicol. Appl. Pharm. 81, 25–31. identifiable systolic myocardial dysfunction in double-crested cormorants Mathewson, P., Hanson-Dorr, K., Porter, W., Bursian, S., Dean, K., Healy, K., Horak, K., (Phalacrocorax auritus). Ecotoxicol. Environ. Saf (submitted for publication, this Link, J., Harr, K., Dorr, B., 2017. Using a bioenergetics model to predict issue). thermoregulatory costs from sublethal oil exposure in a colonial waterbird. Ecol. Appl Hartung, R., 1963. Ingestion of oil by water-fowl. Pap. Mich. Acad. Sci. Arts, Lett. 48, (submitted for publication). 49–55. McEwan, E.H., Whitehead, P.M., 1980. Uptake and clearance of petroleum hydrocarbons Hartung, R., Hunt, G.S., 1966. Toxicity of some oils to waterfowl. J. Wildl. Manag. 30, by the glaucous-winged gull (Larus glaucescens) and the mallard duck (Anas 564–570. platyrhynchos). Can. J. Zool. 58, 723–726. Harvey, S., Klandorf, H., Phillips, J.G., 1981. Reproductive performance and endocrine Pattee, O.H., Franson, J.C., 1982. Short-term effects of oil ingestion on American kestrels responses to ingested petroleum in domestic ducks (Anas platyrhynchos). Gen. Comp. (Falco sparverius). J. Wildl. Dis. 18, 235–241. Endocrinol. 45, 372–380. Peakall, D.B., Norstrom, R.J., Jeffrey, D.A., Leighton, F.A., 1989. Induction of hepatic Harvey, S., Sharp, P.J., Phillips, J.G., 1982. Influence of ingested petroleum on the mixed function oxidases in the herring gull (Larus aregentatus) by Prudhoe Bay crude reproductive performance and pituitary-gonadal axis of domestic ducks (Anas oil and its fractions. Comp. Biochem Physiol. 94C, 461–463. platyrhynchos). Comp. Biochem. Physiol. 72C, 83–89. Pritos, K.L., Perez, C.R., Muthumalage, T., Dean, K.M., Cacela, D., Hanson-Dorr, K.C., Holmes, W.N., Cavanaugh, K.P., Cronshaw, J., 1978. The effects of ingested petroleum on Cunningham, F.L., Bursian, S.J., Link, J.E., Shriner, S., Horak, K., Pritsos, C.A., 2017. oviposition and some aspects of reproduction in experimental colonies of mallard Ecotoxicol. Environ. Saf (submitted for publication, this issue). ducks (Anas platyrhynchos). J. Reprod. Fertil. 54, 335–347. Snyder, S.B., Fox, J.G., Soave, O.A., 1973. Mortalities in Waterfowl Following Bunker C Johnson, J.H., Ross, R.M., McCullough, R.D., 2002. Little Galloo Island, Lake Ontario: a Fuel Exposure: An Examination of the Pathological, Microbiological, and Oil review of nine year of double-crested cormorant diet and fish consumption Hydrocarbon Residue Findings in Birds That Died After the San Francisco Bay Oil information. J. Gt Lakes Res. 28, 182–192. Spill January 18, 1971. Divison of Laboratory Animal Medicine, Stanford Medical Lee, Y.-Z., Leighton, F.A., Peakall, D.B., Norstrom, R.J., O'Brien, P.J., Payne, J.F., Center, Stanford, CA, pp. 27. Rahimtula, A.D., 1985. Effects of ingestion of Hibernia and Prudhoe Bay crude oils on Szaro, R.C., Albers, P.H., Coon, N.C., 1978. Petroleum: effects on mallard egg hepatic and renal mixed function oxidase in nestling herring gulls (Larus argentatus). hatchability. J. Wildl. Manag 42, 404–406. Environ. Res. 36, 248–255. Wootton, T.A., Grau, C.R., Roudybush, T.E., Hahs, M.E., Hirsch, K.V., 1979. Reproductive Leighton, F.A., 1985. Morphological lesions in red blood cells from herring gulls and responses of quail to bunker C oil fractions. Arch. Environ. Contam Toxicol. 8, Atlantic puffins ingesting Prudhoe Bay crude oil. Vet. Pathol. 22, 393–402. 457–463. Leighton, F.A., 1986. Clinical, gross, and histological findings in herring gulls and Atlantic

28 Ecotoxicology and Environmental Safety 146 (2017) 29–39

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Weathered MC252 crude oil-induced anemia and abnormal erythroid MARK morphology in double-crested cormorants (Phalacrocorax auritus) with light microscopic and ultrastructural description of Heinz bodies ⁎ Kendal E. Harra, , Fred L. Cunninghamb, Chris A. Pritsosc, Karen L. Pritsosc, Thivanka Muthumalagec, Brian S. Dorrb, Katherine E. Horakd, Katie C. Hanson-Dorrb, Karen M. Deane, Dave Cacelae, Andrew K. McFaddene, Jane E. Linkf, Katherine A. Healyg, Pete Tuttleg, Steven J. Bursianf a URIKA, LLC. 8712 53rd Pl W, Mukilteo, WA 98275, USA b USDA/USDA/WS/NWRC, Mississippi Field Station, Mississippi State University, Starkville, MS, USA c University of Nevada-Reno, Max Fleischmann Agriculture Bldg. 210, Reno, NV 89557, USA d USDA/USDA/WS/NWRC, Fort Collins, CO, USA e Abt Associates, 1881 Ninth St., Ste 201, Boulder, CO 80302-5148, USA f Michigan State University, East Lansing, MI, USA g US Fish and Wildlife Service, Deepwater Horizon NRDAR Field Office, Fairhope, AL, USA

ARTICLE INFO ABSTRACT

Keywords: Injury assessment of birds following the Deepwater Horizon (DWH) oil spill in 2010 was part of the Natural Avian Resource Damage Assessment. One reported effect was hemolytic anemia with the presence of Heinz bodies (HB) Birds in birds, however, the role of route and magnitude of exposure to oil is unknown. The purpose of the present Deepwater Horizon gulf oil spill study was to determine if double-crested cormorants (Phalacocorax auritis; DCCO) exposed orally and dermally Denatured hemoglobin to artificially weathered crude oil would develop hemolytic anemia including HB and reticulocytosis. In the oral Oxidative damage experiment, sub-adult, mixed-sex DCCOs were fed control (n = 8) or oil–injected fish with a daily target dose of Petroleum Heinz bodies 5 (n = 9) or 10 (n = 9) ml oil/kg for 21 days. Then, subadult control (n = 12) and treated (n = 13) cormorant groups of similar sex-ratio were dermally treated with approximately 13 ml of water or weathered MC252 crude oil, respectively, every 3 days for 6 dosages approximating 20% surface coverage. Collected whole blood samples were analyzed by light (new methylene blue) and transmission electron microscopy. Both oral and dermal treatment with weathered DWH MC252 crude oil induced regenerative, but inadequately compensated, anemia due to hemolysis and hematochezia as indicated by decreased packed cell volume, relative increase in re- ticulocytes with lack of difference in corrected reticulocyte count, and morphologic evidence of oxidant damage at the ultrastructural level. Hemoglobin precipitation, HB formation, degenerate organelles, and systemic oxi- dant damage were documented. Heinz bodies were typically < 2 µm in length and smaller than in mammals. These oblong cytoplasmic inclusions were difficult to see upon routine blood smear evaluation and lacked the classic button appearance found in mammalian red blood cells. They could be found as light, homogeneous blue inclusions upon new methylene blue staining. Ultrastructurally, HB appeared as homogeneous, electron-dense structures within the cytosol and lacked membranous structure. Oxidant damage in avian red blood cells results in degenerate organelles and precipitated hemoglobin or HB with different morphology than that found in mammalian red blood cells. Ultrastructural evaluation is needed to definitively identify HB and damaged or- ganelles to confirm oxidant damage. The best field technique based on the data in this study is assessment of PCV with storage of blood in glutaraldehyde for possible TEM analysis.

⁎ Corresponding author. E-mail addresses: [email protected] (K.E. Harr), [email protected] (F.L. Cunningham), [email protected] (C.A. Pritsos), [email protected] (K.M. Dean), [email protected] (K.A. Healy), [email protected] (S.J. Bursian). http://dx.doi.org/10.1016/j.ecoenv.2017.07.030 Received 15 September 2016; Received in revised form 11 July 2017; Accepted 14 July 2017 Available online 20 July 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 29–39

1. Introduction including generation of HB, reticulocytosis, and a decrease in packed cell volume (PCV), which were reported by Fallon et al. (2014). Following the Deepwater Horizon (DWH) oil spill in 2010, many oiled and not visibly oiled bird carcasses were recovered from areas along the 2. Methods Gulf of Mexico. In addition, a large number of live birds was observed, but fi with lower levels of oiling de ned as < 40% body coverage (Burger, 2.1. Toxicant 1993). The fate of these oiled birds was an important component of the avian injury assessment for the DWH Mississippi Canyon 252 (MC252) Oil MC252 (DWH7937, batch# B030112) oil was collected during the Spill Natural Resource Damage Assessment (NRDA) as well as our un- 2010 Deepwater Horizon oil spill and artificially weathered prior to use ff derstanding of the e ects of oil on avian health. in the studies as previously described (Forth et al., 2017). Crude oil from different geographical regions varies in chemical composition, and therefore, may have different toxic effects resulting in 2.2. Animals and husbandry diverse clinical signs. MC252 is a south Louisiana sweet (low in sulfur) crude oil. Compared with other crude oils, MC252 has relatively high Animal acquisition, maintenance and use were approved by the concentrations of alkanes that microorganisms can use as a food source Institutional Animal Care and Use Committee of the US Department of and relatively low concentrations of polycyclic aromatic hydrocarbons Agriculture National Wildlife Research Center (NWRC). Complete de- (PAHs) (Faksness et al., 2015; Turner et al., 2014). Therefore, it has been tails of animal collection and husbandry and methodology of the oral purported to be biodegradable and less toxic than other oils (Kimes et al., and dermal exposure studies are in Cunningham et al. (2017, this issue). 2014). However, MC252 did also contain low levels of volatile organic Birds were allowed to acclimate to captivity in quarantine for a compounds such as benzene, toluene, and xylene. Further, the toxicity of minimum of 21 days prior to initiation of each study. an oil across taxa may vary with the degree of weathering (Bellas et al., 2013; Finch et al., 2011; Rial et al., 2013). The DWH spill was approxi- mately 80 km off the Louisiana coastline and the oil traveled over 1600 m 2.3. Oral dosing experiment through the water column to reach the surface (Kimes et al., 2014; fl Trustees, 2015). This resulted in exposure of avian, aquatic, and terrestrial Brie y, a total of 26 adult, mixed-sex, apparently healthy, wild species to weathered crude oil that had undergone loss of volatile organic caught DCCOs were randomly assigned to one of three treatment fi compounds such as benzene, toluene and xylene. Indeed, the weathered groups: a control group (n = 8) fed sh lightly anesthetized with MC252 tested here had a relatively high concentration of PAHs (Forth MS222; a treatment group dosed daily with up to 5 ml oil/kg BW oil by ff fi et al., 2017). Hence, the specific suite of toxic effects on wildlife caused by o ering oil-injected sh (n = 9); and a treatment group dosed daily ff fi these variously weathered subtypes of MC252 required further elucidation with up to 10 ml oil/kg BW by o ering oil-injected sh (n = 9). Daily fi as these weathered types of oil were more likely to be the cause of avian provision of oiled-injected sh was up to 21 days. exposure (Henkel et al., 2012). Blood samples were collected from each bird on day 0 (the day When birds are exposed to less than acutely lethal dosages, oil can cause a before oral dosing began) as a baseline comparison, then twice weekly wide range of adverse effects, including hemolytic anemia, renal, myeloid, until humane euthanasia, which occurred either during the study if and hepatic damage, decreased nutrient absorption, altered stress response, warranted by clinical signs or at the conclusion of the study prior to and decreased immune function (Alonso-Alvarez et al., 2007; Leighton, 1985, complete necropsy. Blood samples were collected in heparinized syr- 1986; Peakall et al., 1981, 1989; Szaro et al., 1978). Hemolytic anemia may inges via brachial veins while animals were manually restrained. be characterized by red blood cell (RBC) regeneration and therefore increased numbers of reticulocytes when the bone marrow is intact. When hemolytic 2.4. Dermal exposure experiment anemia is caused by a toxicant, oxidative damage to hemoglobin results in hemichrome formation, hemoglobin precipitation, and development of Heinz A total of 31 DCCO's were captured and retained in captivity. A total of bodies (HB) (Desnoyers, 2010). Heinz bodies are cytoplasmic inclusions 25 subadult DCCOs allocated to a control group (n = 12, 5 male, 7 female) composed of tightly associated, oxidized globin molecules generated during and an exposed group (n = 13, 6 males, 7 females) were used in this trial. the oxidative breakdown of hemoglobin. Red blood cells containing HB have DCCOs were assigned to treatment groups based on the results of blood decreasedabilitytodeliveroxygentocellsandareremovedfromcirculation. samples collected at the initiation of the three-week quarantine period. Heinz bodies are definitive evidence of oxidative damage to hemoglobin, Complete blood count (CBC) values were used to ensure equal division of whether due to toxic oxidants (such as oil) or predisposition to genetic ab- birds with potential health concerns between groups. DCCO's with normalities of hemoglobin or RBC reducing pathways (Desnoyers, 2010). Oil monocyte counts greater than 2.0 × 109 cells/l were considered abnormal and its oxidized metabolites may damage different stages of RBC maturation (severe monocytosis); and were divided between control (n = 4) and including rubriblasts to rubricytes in thebonemarrowaswellasmatureRBCs treatment (n = 3) groups. Additionally, a small oil spill took place one in peripheral blood (Olsgard, 2007). During the initial DWH MC252 oil spill yearpriortothestudy,notfarfromwhere6oftheDCCOswerecollected assessment, increased percentages of HB, reticulocytosis and anemia were and were evenly distributed between groups. During the course of the reported during field evaluation of live, oiled birds in the Gulf of Mexico trial, one bird from the control group and two birds from the treatment (Fallon et al., 2014). group died and were not replaced. Therefore, the final number of birds in Double-crested cormorants were chosen as a representative animal the control and exposed group was 11 birds each to total 22 in the study. model because they were impacted by the DWH spill, are common, pri- Oilonexposedbirds(13ml)andwateroncontrolbirds(13ml)wasap- marily piscivorous waterbirds that inhabit pelagic, coastal, and inland plied every three days through Day 15 of the trial (on Days 0, 3, 6, 9, 12, waterways (Reed et al., 2003),andassuch,couldbeusedassurrogatesfor and 15). Detailed description of application is available in Cunningham other piscivorous species such as pelicans (Pelicanus sp.), terns (Sternidae et al. (2017). Oil exposure may have been via preening, transdermally, or sp.), and skimmers (Rynchops sp.). Additionally, DCCO are listed by the through inhalation. International Union for Conservation of Nature (IUCN) as a species of least Blood samples were collected in heparinized syringes via jugular concern that is relatively easily managed in captivity. veins while animals were manually restrained. All birds had a blood One objective of the present study was to determine if double- sample taken during quarantine to provide baseline data (day −21). crested cormorants (Phalacocorax auritis; DCCOs) orally or dermally During the trial, blood was collected every six days just prior to external exposed to artificially weathered MC252 oil (DWH7937, batch# application of oil or water (days 0, 6, 12, and 18) and just prior to B030112) would develop clinical signs indicative of hemolytic anemia, euthanasia and necropsy (day 21).

30 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 29–39

Table 1 Effect of oral or dermal weathered MS252 crude oil exposure on hematologic values in double-crested cormorants (Phalacocorax auritus). 1Necopsied at 21 days; 2Necropsied at 14 days.

Orally dosed birds Dermally dosed birds

Analyte Control1 5 ml oil/kg BW1 10 ml oil/kg BW2 Control Exposed

PCV (%) Day 0 47 ± 1.2 43 ± 0.9 42 ± 1.1 40 ± 0.3 43 ± 0.3 Day 6 ––– 39 ± 0.4 37 ± 0.3 Day 7 40 ± 1.5 33 ± 1.4 34 ± 1.5 –– Day 12 ––– 40 ± 0.3A 34 ± 0.2B Day 14 39 ± 1.6A 32 ± 1.4A 24 ± 1.6B –– Day 18 ––– 41 ± 0.3A 31 ± 0.2B Day 21 39 ± 1.8A 28 ± 1.7B – 40 ± 0.3A 30 ± 0.3B Aggregate reticulocytes (%) Day 0 3 ± 0.4 2 ± 0.5 5 ± 0.5 4 ± 0.3 4 ± 0.4 Day 6 ––– 4 ± 0.3 4 ± 0.2 Day 7 2 ± 0.5 2 ± 0.5 3 ± 0.5 –– Day 12 ––– 5 ± 0.2 6 ± 0.5 Day 14 3 ± 1.1 5 ± 1.2 7 ± 1.4 –– Day 18 ––– 5 ± 0.4 7 ± 1.0 Day 21 4 ± 1.5 8 ± 1.6 – 4 ± 0.3 5 ± 0.3 Corrected reticulocytes (%) Day 0 3 ± 0.5 2 ± 0.5 4 ± 0.6 4 ± 0.2 4 ± 0.3 Day 6 ––– 3 ± 0.2 4 ± 0.2 Day 7 2 ± 0.4 1 ± 0.5 2 ± 0.4 –– Day 12 ––– 4 ± 0.2 5 ± 0.4 Day 14 3 ± 0.7 4 ± 0.7 4 ± 0.9 –– Day 18 ––– 4 ± 0.3 4 ± 0.6 Day 21 3 ± 1.2 6 ± 1.4 – 3 ± 0.3 3 ± 0.2 Heinz bodies (%) Day 0 1 ± 0.3 0 ± 0.3 2 ± 0.4 1 ± 0.1 1 ± 0.2 Day 6 ––– 1 ± 0.0A 2 ± 0.2B Day 7 1 ± 0.3 5 ± 2.1 3 ± 0.3 –– Day 12 ––– 1 ± 0.1 3 ± 0.6 Day 14 1 ± 0.2A 3 ± 0.7A 7 ± 2.2B –– Day 18 ––– 1 ± 0.1 3 ± 0.2 Day 21 1 ± 0.3 2 ± 0.8 – 1 ± 0.1A 5 ± 0.6B

ABMeans within study with different superscripts in same row differ significantly (p < 0.0007). 1 Based on 4 time points, 0, 7, 14 and 21 days. 2 Based on 3 time points, 0, 7 and 14 days.

2.5. Laboratory analysis ultramicrotome (RMC, Boeckeler Instruments, Tucson, AZ) were prepared on copper grids and stained with uranyl acetate and lead citrate. Sections Smears for HB evaluation were prepared by mixing heparinized whole were imaged using a JEOL 100CX transmission electron microscope blood with new methylene blue (NMB) N stain (Ricca Chemical Co., (JEOL, Tokyo, Japan) at an accelerating voltage of 100 kV. Representative Arlington, TX) in a 1:2 ratio, respectively, and incubating at room tem- images were taken without knowledge of treatment groups. perature for 20 min. After incubation, standard blood smears were pre- Hepatic tissue was excised rapidly following euthanasia and flash pared. Percentages of RBCs that contained HB were enumerated in all frozen in liquid nitrogen as subsamples until transferred to a −70 °C adequate samples by using 600× (high dry) magnification and counting freezer for subsequent assessment of oxidative stress markers following 200–500 RBCs. Adequate samples contained a fine monolayer of RBCs procedures described in Pritsos et al. (2017, this issue). stained such that the nuclei were dark blue (internal control). Aggregate reticulocytes and ring forms were enumerated using previously described methods (Johns et al., 2008). Relative reticulocyte counts were corrected 2.6. Statistical methods for anemia with the PCV normalized to a mean PCV of 45% in cormorants (Reference Interval = 34–53%) using the equation reticulocyte count × Hematologic values, including PCV, relative HB, and relative and (measured PCV/normal PCV). corrected reticulocytes, collected across multiple time points were com- Twenty µl of heparinized whole blood was placed in labeled 1.5 ml pared using linear mixed effects regression models with a repeated mea- polypropylene microcentrifuge tubes containing excess 2.5–3.0% glutar- sures structure. Regression models included elapsed days, treatment, and aldehyde buffered to pH 7.2 (Electron Microscopy Services, Hatfield, PA), treatment*days interaction as fixed effects and individual birds within gently inverted, sealed, and stored in the dark at 4 °C. All whole blood treatment as random effects. Elapsed days and treatment (oral or external samples from control and treated DCCOs from each experiment were oil dose as ml/kg/day) were modeled as continuous variables. Within-day processed and shipped to Michigan State University's Center for Advanced contrasts among treatment groups were assessed with the Kruskal-Wallis Microscopy (East Lansing, MI) in an identical manner at the same time. test. The criterion for statistical significance was p < 0.05. Calculations Aliquots were post-fixed in 1% osmium tetroxide, dehydrated in a graded were performed using TIBCO Spotfire S-PLUS 8.2 for Windows. Reference acetone series, and infiltrated and embedded in Poly/Bed 812 resin intervals were established in accordance with ASVCP guidelines using (Polysciences, Warrington, PA). Thin sections (70 nm) cut with a PTXL Reference Value Advisor (Friedrichs et al., 2012; Geffré et al., 2009).

31 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 29–39

3. Results

Details on total food consumption, weight loss, other clinical signs, and mortality for both studies can be found in Cunningham et al. (2017), this issue.

3.1. Oral dosing experiment

Orally dosed birds had incidental dermal exposure due to defecation of oil into the tanks. Although not quantified, oil was present on feathers of all dosed birds, especially in the high dose group. Clinical signs included reduced cloacal temperatures, apparent hypothermia based on observations of birds seeking supplemental heat, weight loss, lack of appetite, lethargy, feather damage, moribundity, anemia, ab- normal feces and hematochezia, and death. All dosed birds had at least some clinical signs including anemia at a total dose of approximately 80 ml/kg and all were dead prior to 200 ml/kg total dose. All birds with a measured PCV ≤ 24% died. Of the 26 adult, mixed-sex DCCO used in the oral dose study, 16 were euthanized on Day 21. A total of 10 treated DCCOs died or were euthanized within 17 days of the start of the study for humane reasons, including all 9 high dose animals. Control birds exhibited normal behavior, did not seek heat or lose weight and maintained normal cloacal temperatures throughout the study. No control birds died prior to necropsy on day 21. Packed cell volume was significantly decreased by exposure to oil in a dose-related manner (Fig. 1a and Table 1, p < 0.002). Packed cell volume decreased over time in all three groups, but the decrease was more pronounced in oil-dosed birds and was dose dependent. The PCV regression line for control birds was within the reference interval (34–53%), while the regression lines for birds in the 5 and 10 ml oil/ kg/day groups were below the reference interval by day 14 and day 7, respectively, indicating anemia. All birds in the high-dose group be- came moderately (< 31%) to severely (< 20%) anemic by day 14. All birds in the low-dose group became anemic by day 21 at the time of scheduled necropsy. At day 7 of oral oil administration, mean PCV was 40 ± 1.5%, 33 ± 1.4%, and 34 ± 1.5% (mean ± SE) for control, low- Fig. 1. a. Effect of daily oral dosing with artificially weathered MC252 oil on PCV in sub- dose, and high-dose groups, respectively. At day 14, mean PCV of adult, mixed-sex DCCO. Regression analysis showed a significant day effect (p = 0.0017) control, low-dose, and high-dose groups was 39 ± 1.6%, 32 ± 1.4%, and a significant treatment*day interaction (p < 0.001). Dotted lines indicate reference and 24 ± 1.6%, respectively. No control animals exhibited anemia at intervals for apparently healthy cormorants from this population (34 − 53%). b. ff any time during the study. Oil exposure also induced statistically sig- Regression analysis of the e ect of dermal oiling and elapsed days of treatment on PCV in in sub-adult, mixed-sex DCCO. By day 12 and continuing through the end of the study, % nificant changes in plasma clinical chemistries and gross findings at PCV was significantly lower in oiled birds (p < 0.001). Dotted lines indicate reference necropsy (Dean et al., 2017; Harr et al., 2017). intervals for apparently healthy cormorants from this population (34 − 53%). Aggregate reticulocyte quantification by light microscopy in DCCO blood indicated an increasing trend in a dose-related manner by day 14 (Table 1 and Figs. 2 and 3a). At day 7 of oral oil administration, mean Clinical signs reported in oil-exposed birds in this study included reticulocyte percentage of all groups was similar at approximately 2%. deterioration of feather integrity, abnormal feces, hematochezia, ex- At day 14, mean reticulocyte percentages were 3 ± 1.1%, 5.0 ± 1.2% cessive preening and feather plucking, shivering, cardiac arrhythmia, and 7 ± 1.4% (mean ± SE) for control, low-dose, and high-dose dyspnea, and lethargy. Feathers of all oiled birds appeared matted and groups, respectively. When the relative reticulocyte count was cor- rough by day 3. Although oil was only applied to feathers, preening rected for anemia, there was still a mild increase in the reticulocytes at resulted in the skin on the breast and back of some oiled birds being day 14, however, these values did not extend past the reference interval noticeably discolored by day 6 and by day 9, oil covered much of the (control, 3 ± 0.7%; low dose, 4 ± 0.7%; high dose, 4 ± 0.9% re- surface area of all birds with subjective compromised feather integrity. ticulocytes/PCV; mean ± SE; RI 0–6%, Fig. 4a) and the change was not Abnormal feces (Fig. 5a, b) were observed in four oiled birds beginning statistically significant. on day 12 and by the end of the trial seven of 11 oiled birds that sur- vived to necropsy had abnormal feces. Only one control bird had ab- 3.2. Dermal exposure experiment normal feces that consisted of green diarrhea with no evidence of ge- latinous protein or blood as noted in the oiled birds. Plucking of down fi Of the 25 adult, mixed sex DCCOs, one bird from the control group feathers was rst apparent on day 14 in two birds and by day 16 all died immediately after blood sampling on day 0 and two birds from the oiled birds were engaged in this activity. On the day of necropsy, only treatment group were found dead in their cages on days 14 and 18. The the treated birds were noted to be positioned by their heat lamps, oc- control bird had a severe monocytosis (> 4000 monocytes/µl) and casionally shivering. Control birds were typically perched in their en- multiple, pulmonary granulomas including one involving the heart apex. closures and did not exhibit abnormal clinical signs. s Granulomatous pneumonia with intralesional bacteria was diagnosed Packed cell volume of oiled DCCO declined throughout the 21-day fi ff upon histopathology. One exposed bird died with probable septicemia trial with the decrease being signi cantly (p < 0.001) di erent com- (underlying etiologic agent not identified). One exposed bird died with pared to controls by day 12 (Table 1, Fig. 1b). By day 21, all but two no significant lesions that could be assessed as a cause of death. treated birds were anemic, based on a lower reference value of 34%. At

32 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 29–39

Fig. 2. Aggregate to ring form cormorant reticulocyte stained with supravital New Methylene Blue at 1000×.

Fig. 4. a. Effect of daily oral dosing with artificially weathered MC252 oil on % re- ticulocytes normalized to PCV in sub-adult, mixed-sex. The lower and upper boundaries of the boxes indicate the 25th and 75th percentiles, respectively. The black line within the boxes is the median and the heavy black line within the boxes is the mean value. The lower and upper whiskers indicate the 10th and 90th percentiles, respectively. Dotted lines indicate reference intervals for untreated cormorants from this population (0 – 6%). There was no significant difference between doses for % normalized reticulocytes, how- ever, there was a significant time effect (p < 0.0004) and a significant dose*time inter- action (p < 0.0409). There is no data box for the d 21, 10 ml oil/kg BW treatment group because all cormorants in that group had either died or been euthanized by d 21 of the study. b. Effect of dermal exposure with artificially weathered MC252 oil on % re- ticulocytes normalized to PCV in sub-adult, mixed-sex. The lower and upper boundaries of the boxes indicate the 25th and 75th percentiles, respectively. The black line within the boxes is the median and the heavy black line within the boxes is the mean value. The lower and upper whiskers indicate the 10th and 90th percentiles, respectively. Dotted lines indicate reference intervals for untreated cormorants from this population (0 – 6%). There was no significant dose effect and no significant dose *time interaction, however, there was a significant effect of time (p < 0.01). Percent normalized reticulocytes were significantly greater on days 12 and 18 than on days 0, 6 and 21.

Fig. 3. a. Effect of daily oral dosing with artificially weathered MC252 oil on reticulocytes day 6 of dermal oil exposure, mean PCV was 39 ± 0.4% and 37 ± 0.3 (aggregat%) in sub-adult, mixed-sex. The lower and upper boundaries of the boxes in- (mean ± SE) for control and oiled groups, respectively. At day 12, dicate the 25th and 75th percentiles, respectively. The black line within the boxes is the mean PCV was 40 ± 0.3% and 34 ± 0.2% (mean ± SE) and at day 18, median and the heavy black line within the boxes is the mean value. The lower and upper mean PCV of control and oiled groups was 41 ± 4.45% and 31 ± 0.2%, whiskers indicate the 10th and 90th percentiles, respectively. Dotted lines indicate re- respectively. A single control bird with severe monocytosis exhibited a ference intervals for untreated cormorants from this population (0 – 6%). Regression analysis showed no significant main effect of treatment, day, or treatment *day interac- mild anemia of 32% throughout the study. All other control birds tion (p > 0.05). b. Effect of dermal oiling and elapsed days of treatment on percent re- maintained a normal erythron. ticulocytes in sub-adult, mixed-sex DCCO. The lower and upper boundaries of the boxes Relative reticulocytosis was apparent by day 12 in oiled birds (p = indicate the 25th and 75th percentiles, respectively. The black line within the boxes is the 0.01). (Figs. 2, 3b) Control birds did not exhibit moderate anemia (≤ median and the heavy black line within the boxes is the mean value. The lower and upper 31%), reticulocytosis, or Heinz body formation. (Table 1, Fig. 1b). At whiskers indicate the 10th and 90th percentiles, respectively. Dotted lines indicate re- day 12, mean reticulocyte percentages were 4 ± 0.2% and 5 ± 0.5% for ference intervals for untreated cormorants from this population (0 – 6%). On day 12, reticulocyte percentage was significantly lower in control than in oiled birds (p = 0.01). control and treated groups, respectively. At necropsy on day 21, mean

33 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 29–39

Fig. 6. NMB-stained preparation of cormorant blood at 1000x magnification using light microscopy. The white arrow indicates a cell with eight hemoglobin precipitates or HB that are homogeneous light blue cytoplasmic inclusions consistent in size and shape with those observed on transmission electron microscopy.

Fig. 5. a. Abnormal feces from a treated cormorant found on day 12 that shows watery diarrhea admixed with yellow urates and clear, tan and red gelatinous components. b. Abnormal feces imprint on slide at 200× magnification showing RBC admixed with degenerate epithelial cells, proteinaceous debris and a mixed population of bacteria, documenting hematochezia. (Platinum Quik-Dip™ Stain, Mercedes Medical, Inc., Sarasota, Florida, USA). reticulocyte percentages were 4 ± 0.3% and 5 ± 0.3% for control and treated groups, respectively and were not statistically different. When the relative reticulocyte count was corrected for anemia, there was no statistical difference in control and treated birds at any time point (Table 1, Fig. 4b).

3.3. Hemoglobin precipitation and Heinz body formation

Fig. 7. ff fi Hemoglobin damage was similarly documented in both oral and a and b. E ect of daily oral (a) and dermal (b) dosing with arti cially weathered MC252 oil on percent HB in sub-adult, mixed-sex DCCO. Results were similar for both dermal exposure groups and is described together. Heinz body identi- routes of exposure. Regression analysis showed a significant treatment*day interaction fi cation (Fig. 6) by light microscopy in blood from orally-dosed DCCOs (p < 0.05). Dotted lines indicate reference intervals for untreated cormorants from this revealed a significant treatment*day interaction (p = 0.01) as there population (0 − 3%). was a dose-related increase over time (Fig. 7a, Table 1). At day 7 of oral oil administration, there were 1 ± 0.3%, 5 ± 2.1%, and 3 ± 0.3% group. Similarly, in the dermally-exposed birds, there was a mild, but fi (mean ± SE) HB in the control, low-dose, and high-dose groups, re- signi cant (p < 0.001), increase in the percentage of HB in oil-treated spectively. At day 14, 1 ± 0.2% RBCs contained HB in the control birds beginning at day 6 that persisted throughout the study indicating group, 3 ± 0.7% in the low-dose group, and 7 ± 2.2% in the high-dose oxidant damage to RBCs (Fig. 7b). At day 6 and at necropsy on day 21,

34 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 29–39

Fig. 8. (a) Representative RBC ultrastructural image from a control cormorant lacking HB. The cytoplasm is uniform gray in all cells indicating a uniform density and protein concentration with almost no evidence of organelles or HB. Depending upon the TEM orientation, cells may appear ovoid or ellipsoid. There is some nuclear membrane separation in many cells due to processing. The mild background stain precipitation does not change the homogeneous gray appearance of the cytoplasm in samples from the normal control birds. (b) Relatively low magnifica- tion images of RBCs from exposed DCCO showing the frequency of dark opacities with no typical organelle structure (RED arrows), consistent with HB. (c) Higher magnification image of a cytoplasmic inclu- sion with lack of organelle structure and density consistent with a Heinz body (top) and a slightly lighter cytoplasmic inclusion. (d) Normal appearing mitochondria with distinct cristae from an untreated cormorant. (e) Probable degenerate mitochondria with cristae still visible in an RBC from an oil treated cormorant. (f) RBC from treated cormorant showing both probable degenerate mitochondria and inclu- sions which lack any organelle structure consistent with denatured hemoglobin. (g) Low magnification ultrastructural image of RBCs from a treated cor- morant of different orientation showing different size, shape, and frequency of degenerate organelles and hemoglobin.

35 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 29–39

Table 2 Effect of oral and dermal oil exposure on indicators of systemic oxidative damage in double-crested cormorants (Phalacocorax auritis)(Pritsos et al., 2017, this issue).

Orally dosed birds Dermally dosed birds

Controla 5 ml oil/kg BWa 10 ml oil/kg BWb Control 20% surface area exposure

RBC SOD (U/g hemoglobin) Day 0 0.25 ± 0.02 0.30 ± 0.02 0.26 ± 0.02 Day 6 Day 7 0.28 ± 0.03 0.29 ± 0.02 0.27 ± 0.02 Day 12 Day 14 0.29 ± 0.01 0.31 ± 0.007 0.31 ± 0.02 Day 18 Day 21 0.29 ± 0.02 0.31 ± 0.02 – Liver SOD (U/mg protein) 1.74 ± 0.16A 1.21 ± 0.106B 1.29 ± 0.16AB 0.76 0.64 Actual GSSG (nmol/mg) 0.72 ± 0.19B 3.57 ± 0.67A 1.96 ± 0.24AB 1.27 1.29 Reduced GSH (nmol/mg) 24.90 ± 1.25B 62.20 ± 7.42A 85.38 ± 8.63A 25.4A 33.9B Total GSH (nmol/mg) 25.94 ± 1.19B 76.75 ± 14.55A 82.88 ± 9.73A 131.00 36.5B MDA+HAE (nmol/mg) 0.79 ± 0.04 0.78 ± 0.12 0.90 ± 0.06 0.43A 0.32B Trolox Eq. (µmol/mg) 0.12 ± 0.004B 0.13 ± 0.009AB 0.16 ± 0.008A 0.11A 0.12B Kidney SOD (U/mg protein) 1.13 ± 0.042 1.26 ± 0.12 – 0.53 0.536 Actual GSSG (nmol/mg) 0.06 ± 0.03 0.007 ± 0.007 – 0.22 0.17 Reduced GSH (nmol/mg) 0.62 ± 0.08B 1.12 ± 0.12A – 0.56 0.95 Total GSH (nmol/mg) 0.69 ± 0.08B 1.13 ± 0.11A – 0.89 1.29 MDA+HAE (nmol/mg) 1.01 ± 0.05A 0.50 ± 0.04B – 0.50A 0.34B Trolox Eq. (µmol/mg) 0.23 ± 0.05 0.17 ± 0.02 – 0.18 0.21

a Necropsied at 21 days. b Necopsied by 14 days. respectively, 1 ± 0.0% and 1 ± 0.1 RBCs contained HB in the control study is clearly important, this is an underestimation of mortality and group upon light microscopic examination. Treated birds had 2 ± 0.2% lack of recruitment induced by chronic, low-level exposure to oil that and 3 ± 0.2% at day 6 and day 18, respectively. has ongoing effects impacting individuals, populations and ecosystems Heinz bodies were rarely (< 1% RBCs estimated to contain HB) (Camphuysen et al., 2002; Iverson and Esler, 2010). This study docu- found in control birds and frequently (10–40% RBCs estimated to ments that lasting effects to wildlife also occur through chronic, low- contain HB) found in both orally- and dermally-exposed birds on level exposure to oil. Sublethal injury induced by oil intoxication may transmission electron microscopy (TEM) evaluation (Fig. 8a, b, c). Er- induce intoxication, lethargy, and decrease feather integrity which ythrocytes in all oil-exposed birds tended to have smudged nuclei impairs the ability to migrate to feeding and breeding grounds, re- lacking chromatin detail, and frequent RBCs contained many dark cy- sulting in lack of reproductive success and decreased recruitment of a toplasmic inclusions with the same homogeneity and electron density species that results in decreased total animal numbers, similar to as denatured hemoglobin precipitates and HB, not typically associated mortality (Iverson and Esler, 2010). It should be noted that especially in with the RBC membrane. Degenerate organelles including mitochon- the oral dosing study, which resulted in higher mortality, discerning oil dria, ribosomes, and endoplasmic reticulum were also found in the on feathers at a distance was difficult in these dark brown birds. Oiling cytoplasm of RBCs from treated DCCOs (Fig. 8d, e). Similar organelles only became noticeable when it was accompanied by moderate to se- were rarely found in control birds but were more distinct with better vere disruption of feather integrity and feather plucking. Therefore, ultrastructural detail, and therefore, likely not degenerate (Fig. 8f). Low birds suffering from oil intoxication may go unnoticed by even the magnification TEM figures were used to assess the frequency of HB experienced field observer. (Fig. 8g). Membrane pitting (not pictured) was found in RBC from both Exposure to oil can result in pathologic damage to multiple organ control and treated birds. systems due to oxidative injury from oil components and metabolites All of the antioxidant endpoints measured in hepatic tissue at ne- produced by the cyp1a detoxification pathways. Here we focus on the cropsy of orally-dosed DCCOs were significantly affected by exposure to effects on the erythron that was directly effected by oxidative damage oil (p < 0.05). Reduced and total glutathione concentrations were sig- and hemorrhage through the gastrointestinal tract (Pritsos et al., 2017; nificantly increased and decreased respectively (Table 2)(Pritsos et al., Harr et al., 2017). Anemia is a generalized term for decreased erythron 2017, in this issue). mass or decreased PCV and may be caused by a myriad of etiologies. Anemia, while a result of other injury or disease, may in turn cause 4. Discussion decreased oxygen perfusion to tissues resulting in anaerobic metabo- lism, altered cell membrane permeability, cell and tissue dysfunction In this study, we demonstrated that both dermal application of and and, if severe, organ failure. These changes result in clinical signs such provision of food items containing MC-252 oil to double-crested cor- as lethargy and dyspnea that would contribute to a lack of migratory morants resulted in anemia, degenerate organelles, Heinz body for- and reproductive success. Therefore, anemia itself, as a pathologic mation, extravascular hemorrhage, probable bone marrow damage, and state, may cause damage to the population. mortality due to oxidative damage. Wildlife mortality has previously Oxygen carried by hemoglobin is a strong oxidant because it can been used to assess impact to natural resources caused by oil spills (US generate highly reactive derivatives such as the superoxide free radical Department of the Interior, 2011). Acute avian mortality has been well and hydrogen peroxide, and because, by reacting with iron, it forms the described for many large scale oil spills including the Exxon Valdez, reactive hydroxyl radical. These oxidants are constantly being pro- Prestige, and Deepwater Horizon spills (Alonso-Alvarez et al., 2007; duced, and RBCs have several mechanisms to prevent oxidation Barron, 2012; Finch et al., 2011; Piatt and Ford, 1996; Zuberogoitia of hemoglobin through the use of reduced glutathione and enzymes et al., 2006). Although quantifying acute mortality as we saw in this such as superoxide dismutase, glutathione reductase and glutathione

36 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 29–39 peroxidase. This was measured and proven in the orally dosed popu- birds (Olah et al., 2014; Simpson, 1971; Sugawara et al., 2010). When lation (Table 2). Erythrocytes are subject to oxidative injury when in- HB are removed by histiocytic cells the entire RBC may be removed tracellular reducing pathways are insufficient to meet the oxidant from circulation, resulting in an extravascular hemolytic anemia. Ad- challenge. When oxidant injury occurs, denatured hemoglobin forms ditionally, in this study we documented hematochezia in oil-treated hemichromes, which may bind to membranes or aggregate to form birds only. The amount of blood noted in feces likely also contributed to larger hemoglobin precipitates (Waugh and Low, 1985). Oxidative the documented anemia in oil-treated birds. The body's response to damage may result in hemolytic anemia due to direct cell membrane anemia is production of RBCs in hematopoietic tissue. This will result in damage but may also shorten the life span of RBCs by the formation of both mature and immature RBCs (reticulocytes and rubricytes) release HB and eccentrocytes (Desnoyers, 2010). Eccentrocytes are RBCs that into the blood. Hence, increased numbers of reticulocytes (re- have had their membranes partially fused by oxidative damage, re- ticulocytosis), as found in the oil-dosed birds in this study, are used to sulting in their hemoglobin (and all cell contents) being shifted to one document RBC regeneration. When reticulocyte counts from both der- side of the cell. Eccentrocytes have been reported in many mammalian mally- and orally-exposed cormorants were corrected for the severity of species but have never been documented in birds. In this study, some the anemia, they were not above the reference interval, indicating a RBCs with HB appeared to have irregular borders with pits in the cel- lack of compensatory regeneration. Additionally, there was a reduction lular membrane, but nothing similar to a classic eccentrocyte was in reticulocytosis over time that suggests the absence of production of identified upon light or electron microscopy (Caldin et al., 2005). the erythroid line by the bone marrow repeatedly exposed to crude oil. Heinz bodies are named for Robert Heinz, who first described ag- This is consistent with oil-induced bone marrow damage previously gregations of protoplasm in cells exposed to oxidizing agents in 1890 documented in mammals (Meyne and Deaven, 1982). (Heinz, 1890). Denatured hemoglobin, documented by the gold stan- Hemolytic anemia has been demonstrated in several species of birds dard of electron microscopy in the present study, is indicative of oxi- exposed to crude oil (Fry and Lowenstine, 1985; Leighton, 1985; Troisi dative damage to hemoglobin in RBCs and consistent with oil-induced et al., 2006, 2007). It is believed that oxidative damage is mediated by hemolytic anemia. The changes in hepatic oxidative stress endpoints metabolites of polycyclic aromatic hydrocarbons (PAH) generated from provide further evidence of systemic oxide radical damage in the body metabolic actions of cytochrome P450 enzymes in birds as well as of DCCOs (Table 2). Oxidants are used by the body in immune response mammals (Troisi et al., 2006, 2007). Controlled dosing studies of to combat pathogens and may also be end products of ongoing phy- Atlantic puffins (Fratercula arctica) and herring gull chicks (Larus ar- siologic processes (Waugh and Low, 1985). Therefore, occasional he- gentatus) demonstrated a dramatic reduction in the number of circu- moglobin precipitates and HB would be expected in RBCs from control lating RBCs in birds orally dosed with large volumes of crude oil animals as documented in this study. There were few HB (< 1% RBCs (Leighton, 1985). Further, the presence of damaged hemoglobin in the contained HB) identified in the control population as confirmed using erythrocytes from these birds, as evidenced by Heinz body inclusions, electron microscopy. Control birds also had significantly lower oxida- points to oxidative damage as a mechanism of anemia. In heavily oiled tive stress endpoints, indicating that these statistically significant birds admitted to rehabilitation facilities, Troisi et al. (2007) further variables were independent of disease that was found in both the demonstrated a correlation between the percentage of HB and circu- control and oil-exposed groups of wild caught DCCOs. Reviewers of the lating PAH concentrations in plasma from heavily-oiled Common ultrastructural samples could easily distinguish control from treated Guillemots (Uria aalge), suggesting a dose-response relationship. In the samples based on the numbers of inclusions (HB and denatured orga- current study, PAH concentrations in cormorant tissues were not nelles) in RBCs. The HB relative count produced using light micro- measured and so comparisons are not possible. Newman et al. (2000) scopsy was significantly lower than those counted in the same samples found a mild to moderate, regenerative hemolytic anemia induced by using electron microscopy. This is likely due to conservative counting of administration of 2.5 and 10 ml/kg Prudhoe Bay crude oil. This study samples where the small blue inclusions could not be discerned from documented reticulocytosis but not HB detected by light microscopy of methylene blue stain precipitate. In comparison, stain precipitate RBCs stained supravitally with new methylene blue (Newman et al., versus damaged organelles or HB could be easily differentiated in 2000). However, electron microscopy was never performed on those electron micrographs (Fig. 8). samples, so the absence of electron dense inclusions consistent with HB Heinz bodies found in avian RBCs exhibit some morphologic dif- and damaged mitochondria, i.e. evidence of oxidant damage, was never ferences compared to those found in mammals including decreased confirmed. In a field study conducted as part of the Deepwater Horizon total numbers, decreased size and a lack of membrane association. The NRDA, Fallon et al. (2014) reported that American oystercatchers, lower numbers and smaller size of HB in cormorant erythrocytes upon black skimmers, brown pelicans and great egrets with small amounts of examination by electron microscopy in the present study is consistent visible oil present on their feathers suffered from oxidative injury to with previous findings in turkeys (Simpson, 1971). The classic cyto- RBCs as indicated by the presence of HBs, had PCVs that were 4–19% plasmic button typical of the mammalian Heinz body was rarely found less compared to birds from reference sites and had 27–40% more re- in NMB preparations from cormorants, but rather small, oval to irre- ticulocytes compared to birds from reference sites. Additionally, birds gularly shaped, light blue inclusions were found throughout the cyto- with no visible oiling sampled in areas potentially affected by the oil plasm (Fig. 6). Although the dark cytoplasmic inclusions found on TEM spill had evidence of HBs, a decrease in PCV and an increase in re- were not typically membrane-associated as HB classically are in mam- ticulocytes compared to birds from reference sites. The findings in this mals, they are consistent in structure and electron density with HB. study support previous literature and further determine that significant Binding of hemichromes and hemoglobin precipitates has been found to hemolytic anemia may be induced in birds receiving as little as 5 ml/kg be biphasic and exhibit low and high affinity sites of binding in mam- body weight dermally applied chronically (to total 65 ml dermal dose mals. High affinity sites of hemoglobin precipitate binding have been and estimated 22 ml ingested). shown to be on the cytoplasmic domain of band 3 transmembrane While HB may be visualized using numerous supravital staining protein in rodents (Waugh and Low, 1985; Zhang et al., 2003). It is techniques including new methylene blue (NMB), methyl violet, Nile unknown if band 3 exists in bird membranes and a lack of band 3 would blue sulfate, brilliant cresyl blue, Janus green, neutral red, Victoria result in decreased Heinz body binding to the membrane. Further cy- blue, Bismark brown, or gentian violet, they are most commonly tochemical investigation of the avian erythrocyte is warranted to better identified using new methylene blue. The supravital technique must be understand why avian HB appear to be morphologically different than used, as opposed to NMB staining of air-dried blood smears because mammalian HB. blood smears dipped in new methylene blue create a substandard pre- Heinz bodies are removed from RBCs by the reticuloendothelial paration and HB are frequently not visualized as they appear as un- system, especially in the spleen by histiocytic cells in mammals and stained refractile bodies (Jain, 1973). Due to the lack of distinct button

37 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 29–39 morphology found in oxidant damaged mammalian erythrocytes, Fallon, J.A., Smith, E.P., Hopkins, W., 2014. Determining physiological injury to birds fi ffi fi exposed to oil from the Deepwater Horizon (Mississippi Canyon 252) spill. USFWS identi cation of HB by light microscopy is di cult and not de nitive. Rep. 1–127. Based on findings in this study, it also underestimates the number of Finch, B.E., Wooten, K.J., Smith, P.N., 2011. Embryotoxicity of weathered crude oil from the Gulf of Mexico in mallard ducks (Anas platyrhynchos). Environ. Toxicol. Chem. damaged RBC in the sample. Transmission electron microscopy can be 30, 1885–1891. http://dx.doi.org/10.1002/etc.576. very helpful in characterizing or identifying RBC cytoplasmic inclu- Forth, H.P., Mitchelmore, C.L., Morris, J.M., Lay, C.R., Lipton, J., 2017. Characterization sions, and is recommended by these authors as a confirmatory test for of dissolved and particulate phases of water accommodated fractions used to conduct aquatic toxicity testing in support of the deepwater horizon natural resource damage the presence of HB in avian RBCs (Desnoyers, 2010). assessment. Environ. Toxicol. Chem. 36 (6), 1460–1472. Friedrichs, K.R., Harr, K.E., Freeman, K.P., Szladovits, B., Walton, R.M., Barnhart, K.F., 5. Summary Blanco-Chavez, J., 2012. ASVCP reference interval guidelines: determination of de novo reference intervals in veterinary species and other related topics. Vet. Clin. Pathol. 41, 441–453. http://dx.doi.org/10.1111/vcp.12006. Oral and dermal exposure of double-crested cormorants to weath- Fry, D.M., Lowenstine, L.J., 1985. Pathology of common murres and Cassin's auklets exposed to oil. Arch. Environ. Contam. Toxicol. 14, 725–737. http://dx.doi.org/10. ered DWH MC 252 crude oil induced hemolytic anemia as indicated by 1007/BF01055780. decreased PCV, relative reticulocytosis with an inadequate regenerative Geffré, A., Friedrichs, K., Harr, K., Concordet, D., Trumel, C., Braun, J.P., 2009. Reference response, and presence of HB and degenerate organelles, which is values: a review. Vet. Clin. Pathol. 38, 288–298. http://dx.doi.org/10.1111/j.1939- 165X.2009.00179.x. consistent with other reports of oil-exposed birds. Additionally, this Harr, K.E., Reavill, D.R., Bursian, S.J., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, study documents extravascular blood loss through hematochezia con- B.S., Hanson-Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Shriner, S.A., Schmidt, tributing to the severity of anemia, potentially due to coagulopathy. R.E., 2017. Organ weights and histopathology of double-crested cormorants (Phalacrocorax auritus) dosed orally or dermally with artificially weathered Avian HB differ from those found in mammalian RBC in that they are a Mississippi Canyon 252 crude oil. Ecotoxicol. Environ. Saf (in press). relatively consistent small size and are located within the cytoplasm, Heinz, R., 1890. Morphologische veränderungen des roten blutkörperchens durch gifte. ff [Virchows] archiv für pathologische anatomie und physiologie und für klinische possibly due to the nucleated cell and di erent cell cytoskeleton. Medizin, Berlin,122, pp. 112–116. Hematologists should be aware of these differences in Heinz body ul- Henkel, J.R., Sigel, B.J., Taylor, C.M., 2012. Large-scale impacts of the Deepwater ff trastructure when attempting to identify them by light or electron mi- Horizon oil spill: Can local disturbance a ect distant ecosystems through migratory shorebirds? Bioscience 62, 676–685. http://dx.doi.org/10.1525/bio.2012.62.7.10. croscopy. Ultrastructural assessment of suspected HB in birds is re- Iverson, S.A., Esler, D., 2010. Harlequin duck population injury and recovery dynamics commended to confirm identification by light microscopy which is following the 1989 Exxon Valdez oil spill. Ecol. Appl. 20, 1993–2006. http://dx.doi. org/10.1890/09-1398.1. challenging and underestimates the damage of the erythron. The best Jain, N., 1973. Demonstration of Heinz bodies in erythrocytes of the cat. Bull. Am. Soc. field technique based on the data in this study is assessment of PCV with Vet. Clin. Pathol. 2, 13–23. storage of blood in glutaraldehyde for possible TEM analysis. Johns, J.L., Shooshtari, M.P., Christopher, M.M., 2008. Development of a technique for quantification of reticulocytes and assessment of erythrocyte regenerative capacity in birds. Am. J. Vet. Res. 69, 1067–1072. http://dx.doi.org/10.2460/ajvr.69.8.1067. Acknowledgements Kimes, N.E., Callaghan, A.V., Suflita, J.M., Morris, P.J., 2014. Microbial transformation of the Deepwater Horizon oil spill: past, present, and future perspectives. Front. Microbiol. 5, 1–11. http://dx.doi.org/10.3389/fmicb.2014.00603. The studies appearing in this special issue were funded by the U.S. Leighton, F.A., 1985. Morphological lesions in red blood cells from herring gulls and Fish and Wildlife Service (Order No. F12PD01069) as part of the Atlantic puffins ingesting Prudhoe Bay crude oil. Vet. Pathol. 22, 393–402. http://dx. doi.org/10.1177/030098588502200415. Deepwater Horizon Natural Resource Damage Assessment. Leighton, F.A., 1986. Clinical, gross, and histological findings in herring gulls and Atlantic Authors would like to acknowledge Alicia Withrow for her expertise puffins that ingested Prudhoe Bay crude oil. Vet. Pathol. 23, 254–263. http://dx.doi. in quality processing of samples and TEM and Drs. Michael Scott, Rose org/10.1177/030098588602300305. Meyne, J., Deaven, L.L., 1982. Cytogenetic effects of shale-derived oils in murine bone Raskin and John Harvey for their assistance in reviewing this manu- marrow. Environ. Mutagen. 4, 639–645. script. Newman, S.H., Anderson, D.W., Ziccardi, M.H., Trupkiewicz, J.G., Tseng, F.S., Christopher, M.M., Zinkl, J.G., 2000. An experimental soft-release of oil-spill re- habilitated American coots (Fulica americana): II. Effects on health and blood para- References meters. Environ. Pollut. 107, 295–304. http://dx.doi.org/10.1016/S0269-7491(99) 00171-2. Olah, I., Nagy, N., Vervelde, L., 2014. Avian immunology. In: Schat, K., Kaspers, B., Alonso-Alvarez, C., Pérez, C., Velando, A., 2007. Effects of acute exposure to heavy fuel – Kaiser, P. (Eds.), Avian Immunology. Academic Press (Elsevier), San Diego, CA, pp. oil from the Prestige spill on a seabird. Aquat. Toxicol. 84, 103 110. http://dx.doi. 11–41. org/10.1016/j.aquatox.2007.06.004. Olsgard, M.L., 2007. Toxicological Evaluation of Inhalation Exposure to Benzene and Barron, M.G., 2012. Ecological impacts of the Deepwater Horizon Oil Spill: Implications – Toluene in a Raptorial Bird, the American kestrel, Falco sparverius. University of for immunotoxicity. Toxicol. Pathol. 40, 315 320. http://dx.doi.org/10.1177/ Saskatchewan. 0192623311428474. Peakall, D., Norstrom, R., Jeffrey, D.A., Leighton, F., 1989. Induction of hepatic mixed Bellas, J., Saco-Álvarez, L., Nieto, Ó., Bayona, J.M., Albaigés, J., Beiras, R., 2013. function oxidases in the herring gull (Larus argentatus) by Prudhoe Bay crude oil and Evaluation of artificially-weathered standard fuel oil toxicity by marine invertebrate – – its fractions. Comp. Biochem. Physiol. Part C 94C, 461 463. embryogenesis bioassays. Chemosphere 90, 1103 1108. http://dx.doi.org/10.1016/ Peakall, D.B., Tremblay, J., Kinter, W.B., Miller, D.S., 1981. Endocrine dysfunction in j.chemosphere.2012.09.015. seabirds caused by ingested oil. Environ. Res. 24, 6–14. http://dx.doi.org/10.1016/ Burger, A.E. Estimating the mortality of seabirds following oil spills: Effects of spill – 0013-9351(81)90126-2. volume. Mar. Pollut. Bull. 26, 140 143. http://dx.doi.org/10.1016/0025-326X(93) Piatt, J., Ford, R., 1996. How Many Seabirds Were Killed by the Exxon Valdez Oil Spill. 90123-2. American Fisheries Society Symposiumpp. 712–719. Caldin, M., Carli, E., Furlanello, T., Solano-Gallego, L., Tasca, S., Patron, C., Lubas, G., Pritsos, K.L., Perez, C.R., Muthumalage, T., Dean, K.M., Cacela, D., Hanson-Dorr, K., 2005. A retrospective study of 60 cases of eccentrocytosis in the dog. Vet. Clin. – Cunningham, F.L., Bursian, S.J., Link, J.E., Shriner, S.A., Horak, K.E., 2017. Dietary Pathol. 34, 224 231. http://dx.doi.org/10.1111/j.1939-165X.2005.tb00045.x. intake of Deepwater Horizon oil-injected live food fish by double-crested cormorants Camphuysen, C.J., Heubeck, M., Cox, S.L., Bao, R., Humple, D., Abraham, C., A.S., 2002. – resulted in oxidative stress. Ecotoxicol. Environ. Saf (submitted for publication, this The Prestige oil spill in Spain. Atl. Seabirds 4, pp. 131 140. issue). Cunningham, F., Dean, K., Hanson-Dorr, K., Harr, K.E., Healy, K., Horak, K., Shriner, S., Reed, J.M., Causey, D., Hatch J.J., Cooke, F., Crowder, L., 2003. Review of the Double- Link, J., Bursian, S., Dorr, B., 2017. Development of Methods for the study of Avian Crested Cormorant Management Plan. Oil Toxicity Studies using the Double Crested Cormorant (Phalacrocorax auritus). Rial, D., Radović, J.R., Bayona, J.M., Macrae, K., Thomas, K.V., Beiras, R., 2013. Effects of Ecotoxicol. Environ. Saf., in press. simulated weathering on the toxicity of selected crude oils and their components to Dean, K.M., Bursian, S.J., Cacela, D., Carney, M.W., Cunningham, F.L., Dorr, B., Hanson- sea urchin embryos. J. Hazard. Mater. 260, 67–73. http://dx.doi.org/10.1016/j. Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Lipton, I., McFadden, A.K., McKernan, jhazmat.2013.05.004. M.A., Harr, K.E., 2017. Changes in white cell estimates and plasma chemistry mea- Simpson, C.F., 1971. Fate of erythrocytes containing Heinz bodies in the spleen and liver surements following oral or dermal dosing of double crested cormorants, – fi of dogs and turkeys. Vet. Pathol. 8, 118 129. Phalacocorax auritus, with arti cially weathered MC252 oil. Ecotoxicol. Environ. Saf Sugawara, Y., Hayashi, Y., Shigemasa, Y., Abe, Y., Ohgushi, I., Ueno, E., Shimamoto, F., (in this issue). 2010. Molecular biosensing mechanisms in the spleen for the removal of aged and Desnoyers, M., 2010. Anemia associated with oxidative injury. In: Weiss, D.J., Wardrop, damaged red cells from the blood circulation. Sensors 10, 7099–7121. http://dx.doi. K. (Eds.), Schalm's Veterinary Hematology. Blackwell Publishing, Ames, IA, pp. – org/10.3390/s100807099. 239 245. Szaro, R.C., Dieter, M.P., Heinz, G.H., Ferrell, J.F., 1978. Effects of chronic ingestion of Faksness, L.G., Altin, D., Nordtug, T., Daling, P.S., Hansen, B.H., 2015. Chemical com- – fi south Louisiana crude oil on mallard ducklings. Environ. Res. 17, 426 436. http://dx. parison and acute toxicity of water accommodated fraction (WAF) of source and eld doi.org/10.1016/0013-9351(78)90046-4. collected Macondo oils from the Deepwater Horizon spill. Mar. Pollut. Bull. 91, – Troisi, G., Borjesson, L., Bexton, S., Robinson, I., 2007. Biomarkers of polycyclic aromatic 222 229. http://dx.doi.org/10.1016/j.marpolbul.2014.12.002. hydrocarbon (PAH)-associated hemolytic anemia in oiled wildlife. Environ. Res. 105,

38 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 29–39

324–329. http://dx.doi.org/10.1016/j.envres.2007.06.007. sites/doi.gov/files/migrated/deepwaterhorizon/upload/Final-DOI-NRDA-AP-1-July- Troisi, G.M., Bexton, S., Robinson, I., 2006. Polyaromatic hydrocarbon and PAH meta- 2011-2.pdf〉. bolite burdens in oiled common guillemots (Uria aalge) stranded on the East Coast of Waugh, S.M., Low, P.S., 1985. Hemichrome binding to band 3: nucleation of Heinz bodies England (2001–2002). Environ. Sci. Technol. 40, 7938–7943. http://dx.doi.org/10. on the erythrocyte membrane. Biochemistry 24, 34–39. http://dx.doi.org/10.1021/ 1021/es0601787. bi00322a006. Trustees, D.W.H., 2015. Deepwater Horizon oil spill programmatic damage assessment Zhang, Y., Manning, L.R., Falcone, J., Platt, O., Manning, J.M., 2003. Human erythrocyte and restoration plan and programmatic environmental impact statement. October membrane band 3 protein influences hemoglobin cooperativity. Possible effect on Draft. oxygen transport. J. Biol. Chem. 278, 39565–39571. http://dx.doi.org/10.1074/jbc. Turner, R.E., Overton, E.B., Meyer, B.M., Miles, M.S., Hooper-Bui, L., 2014. Changes in M303352200. the concentration and relative abundance of alkanes and PAHs from the Deepwater Zuberogoitia, I., Martínez, J.A., Iraeta, A., Azkona, A., Zabala, J., Jiménez, B., Merino, R., Horizon oiling of coastal marshes. Mar. Pollut. Bull. 86, 291–297. http://dx.doi.org/ Gómez, G., 2006. Short-term effects of the Prestige oil spill on the peregrine falcon 10.1016/j.marpolbul.2014.07.003. (Falco peregrinus). Mar. Pollut. Bull. 52, 1176–1181. http://dx.doi.org/10.1016/j. US Department of Interior, 2011. Interim, partial claim for assessment costs; 20 April marpolbul.2006.02.016. 2010 Deepwater Horizon (MC252) spill. Retrieved from: 〈https://www.doi.gov/

39 Ecotoxicology and Environmental Safety 146 (2017) 40–51

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Changes in white cell estimates and plasma chemistry measurements MARK following oral or external dosing of double-crested cormorants, Phalacocorax auritus, with artificially weathered MC252 oil

⁎ Karen M. Deana, , Steven J. Bursianb, Dave Cacelaa, Michael W. Carneya, Fred L. Cunninghamc, Brian Dorrc, Katie C. Hanson-Dorrc, Kate A. Healyd, Katherine E. Horake, Jane E. Linkf, Ian Liptona, Andrew K. McFaddena, Moira A. McKernang, Kendal E. Harrg a Abt Associates, 1811 Ninth St., Suite 201, Boulder, CO 80302, USA b Department of Animal Science, Michigan State University, East Lansing, MI 48824, USA c USDA/APHIS/WS/NWRC-MS Field Station, MS State University, P.O. Box 6099, Starkville, MS 39762, USA d US Fish and Wildlife Service, Deepwater Horizon NRDAR Field Office, Fairhope, AL, USA e USDA/APHIS/WS/NWRC, 4101 LaPorte Ave, Fort Collins, CO 80521, USA f US Fish and Wildlife Service, Ecological Services, Falls Church, VA, USA g Urika Pathology LLC, 8712 53rd Pl W., Mukilteo, WA 98275, USA

ABSTRACT

Scoping studies were designed whereby double-crested cormorants (Phalacocorax auritus) were dosed with ar- tificially weathered Deepwater Horizon (DWH) oil either daily through oil injected feeder fish, or by application of oil directly to feathers every three days. Preening results in oil ingestion, and may be an effective means of orally dosing birds with toxicant to improve our understanding of the full range of physiological effects of oral oil ingestion on birds. Blood samples collected every 5–6 days were analyzed for a number of clinical endpoints including white blood cell (WBC) estimates and differential cell counts. Plasma biochemical evaluations were performed for changes associated with oil toxicity. Oral dosing and application of oil to feathers resulted in clinical signs and statistically significant changes in a number of biochemical endpoints consistent with petro- leum exposure. In orally dosed birds there were statistically significant decreases in aspartate amino transferase (AST) and gamma glutamyl transferase (GGT) activities, calcium, chloride, cholesterol, glucose, and total protein concentrations, and increases in plasma urea, uric acid, and phosphorus concentrations. Plasma electrophoresis endpoints (pre-albumin, albumin, alpha-2 globulin, beta globulin, and gamma globulin concentrations and al- bumin: globulin ratios) were decreased in orally dosed birds. Birds with external oil had increases in urea, creatinine, uric acid, creatine kinase (CK), glutamate dehydrogenase (GLDH), phosphorus, calcium, chloride, potassium, albumin, alpha-1 globulin and alpha-2 globulin. Decreases were observed in AST, beta globulin and glucose. WBC also differed between treatments; however, this was in part driven by monocytosis present in the externally oiled birds prior to oil treatment.

1. Introduction more became oiled (PDARP, 2016). Still, these numbers are unlikely to reflect the true cost of the DWH spill on bird life as they do not account On April 10, 2010 when the Deepwater Horizon (DWH) well MC252 for the sub-lethal health related effects of oil. exploded, the temperate waters of the Gulf of Mexico gradually became At sub-lethal oral doses the physiological effects of oil exposure contaminated with approximately 3.19 million barrels (507 million li- include anemia, organ dysfunction, decreased nutrient absorption, al- ters) of South Louisiana sweet crude oil (PDARP, 2016). Some of this oil tered stress response, and decreased immune function (Szaro et al., made its way to beaches, marshes and shallower fishing grounds used 1978; Leighton et al., 1985; Leighton, 1985, 1986, 1993; Peakall et al., by birds. As a result, thousands of the 150 species of birds that occur in 1989). Hemolytic anemia is one of the most commonly reported effects the waters and wetlands of the Gulf of Mexico died, and thousands of oil ingestion in birds (Hartung and Hunt, 1966; Eastin and Rattner,

⁎ Corresponding author: Department of Neuroscience, University of Lethbridge, Alberta, Canada. E-mail address: [email protected] (K.M. Dean). http://dx.doi.org/10.1016/j.ecoenv.2017.08.007 Received 21 September 2016; Received in revised form 31 July 2017; Accepted 2 August 2017 Available online 25 August 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51

1982; Pattee and Franson, 1982; Lee et al., 1986; Leighton et al., 1985; compare and contrast the changes in plasma chemistry measurements Leighton, 1986; Hughes et al., 1990; Yamato et al., 1996; Walton et al., often used for diagnostic purposes to determine if there are changes 1997; Newman et al., 2000; Seiser et al., 2000; Troisi et al., 2007), but that indicate oil toxicity, and if those changes are of similar scope and itself a marker of oxidative damage. Detoxification and elimination of magnitude between the two dosing systems. polycyclic aromatic hydrocarbons (PAHs) from the body occurs through activation of cytochrome P450 (CYP450) mono-oxygenases in the liver 2. Methods to metabolize lipophilic PAHs into more hydrophilic and reactive oxides and epoxides (Peakall, 1989; Troisi, 2006 }. These reactive Full method details for all aspects of this study are available in oxygen species can cause oxidative damage to red blood cells and organ Cunningham et al. (2017). General methods are summarized below. systems including damage to liver, kidney, gastro-intestine, adrenal glands, muscle tissue and salt glands in birds (Leighton, 1986). Mea- 2.1. Animal collection and husbandry surement of plasma biochemistry markers of organ damage such as enzyme activities, blood urea nitrogen, urea, uric acid, creatinine, Twenty-six double-crested cormorants were collected for the oral chloride, cholesterol have been reported in a variety of bird species dosing study and 31 were collected for external oil application. All birds following oil spills (Fleming et al., 1982; Pattee and Franson, 1982). were collected in Mississippi and Alabama, then transported to the Exposure to oil can also cause an upregulation of immune and in- National Wildlife Research Center Mississippi Field Station. Upon ar- flammatory responses, and cause endocrine disruption (Briggs et al., rival birds were given a unique identifier quarantined in individual 1996; Perez et al., 2010). Oiled birds show increases in inflammatory pens for 2–3 weeks. Quarantine and experimental procedures con- responses, depressions in lymphocyte concentrations and im- formed and were approved by the NWRC Institutional Animal Care and munosuppression that results in increased susceptibility to secondary Use Committee (IACUC; protocols QA2326 and QA2107). infections (Fry and Lowenstine, 1985; Briggs et al., 1997; McOrist and Each pen contained a 190 L water tank used for daily feeding with Lenghuas, 1992; Newman et al., 2000). The irritant effects of oil on the live fingerling channel catfish (Ictalurus punctatus) that were maintained gastrointestinal tract, combined with observed decreases in adrenal under standard feeding, temperature and aeration settings. At least function (Rattner et al., 1981; Gorsline et al., 1981; Gorsline et al., 600 g of fish were provided to each bird daily. Actual food consumption 1982; Fry and Lowenstine, 1985; Seiser et al., 2000; Newman et al., was calculated per bird based on weight of fish remaining the next day. 2000), make it unsurprising that inflammatory responses are upregu- Water was changed manually every other day and dead fish were re- lated. Decreases or increases in lymphocyte and eosinophil concentra- moved as soon as possible. Birds were monitored daily. tions could be indicative of direct effects of oil or reactive oxygen species on white blood cell production; however, inflammatory pro- 2.2. Toxicant cesses are not fully understood in avian responses to oil (Seiser et al., 2000; Newman et al., 2000; Garcia et al., 2010). Artificially weathered MC252 oil (DWH7937, batch# B030112) was Generally, changes in plasma clinical chemistries differ amongst prepared from crude oil collected during the DWH oil spill (Forth et al., species studied with exposure route, duration, oil type and species 2017). When not in use, the oil was stored in a leak-proof container in a sensitivity; however, few studies have been able to combine clinical flammable storage cabinet. measurements with mechanistic toxicological approaches to define a suite of endpoints that fully described the adverse effects of crude oil on 2.3. Oral dosing birds. Measurements of clinically-relevant plasma endpoints are often collected opportunistically following an oil spill, making them reliant Double-crested cormorants were randomly assigned to one of three on availability of individuals for which no clinical history is available, treatment groups: control that was administered fish that had been including dose. Further, elucidation of the toxicological mechanisms lightly anesthetized and allowed to revive (n = 8); a group dosed daily responsible for the clinical health effects observed in wild birds is dif- with the target 5 ml oil/kg body weight (BW) (n = 9); and a group ficult because experimental dose manipulations have been hampered by dosed daily with the target 10 ml oil/kg BW (n = 9). Actual doses as a paucity of effective dosing techniques for birds. One of the effects of calculated by Cunningham et al. (2017) were 5.2 ± 0.3 ml oil/kg BW oil is as a gastro-intestinal irritant, and as such birds given a bolus oral and 8.4 + 0.9 ml oil/kg BW respectively. dose will either regurgitate oil or the oil will have a rapid transit time Fingerling channel catfish were lightly anesthetized using tricaine through the gastro-intestinal tract resulting in little PAH absorption. methanesulfonate (MS222) and given an intraperitoneal injection of External application of oil, while effective for oral dosing due to the 2.0 ml of oil. Each fish was injected with the same volume to ensure natural tendencies of birds to maintain feather integrity by preening, is that per bird oil consumption could be calculated based on number of still not optimal because dose can only be approximated. Hartung fish consumed. Injected fish were placed into a holding tank to monitor (1963) estimated that 25% of a 20 ml moderate density oil applied to recovery from anesthesia and ensure oil retention. Oil-injected fish the feathers of a black duck would be preened by day 3% and 50% by survived for more than 24 h if not killed by foraging birds. Once birds day 8. This varied slightly depending on the volume applied, but the had consumed all oil-injected fish, additional uninjected fish were of- utility of this study is limited because only applied once, so estimates fered (up to 600 g per bird). Consumption of both dosed and clean fish for studies in which oil is applied multiple times will only roughly were monitored daily for each bird. approximate ingestion. Despite issues with dose determination, external oil application has 2.4. External oil application the potential to be an effective means of understanding the adverse effects of oil on avian health and physiology, but the link needs to be The external oiling study took place after the oral dosing study. made between definitive oral dosing and external oil application to il- Baseline differences in plasma markers were measured in pre-study lustrate the effectiveness of the latter technique. As such, oral and ex- testing for the external oiling study indicated the presence of pre-ex- ternal dosing methods were developed for the double-crested cor- isting disease in some individuals. Diseased animals were not excluded morant (Phalacocorax auritus,) (Cunningham et al., 2017). The potential so that the population would accurately represent wild bird popula- differences between the two dosing methods are likely to related to the tions. Birds were assigned to either control or moderate (16–40%) absolute dose rates, as oral delivery in feed can potentially result in a oiling groups based on blood samples collected at the beginning of higher dose that what could be consumed from feather preening, and quarantine. High monocyte counts, greater than 2.0 × 109 cell/L, were actual absorption of that dose from the gastrointestinal tract. Here we considered abnormal so these birds were divided evenly between

41 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51 control and treatment groups. Additionally, a small oil spill took place birds due a sampling error. Plasma biochemical analysis was performed in November 8, 2013 near where some birds were collected. As such, using the Roche Cobas Mira Plus chemistry analyzer (Roche these birds were also evenly distributed between groups. Initial sample Diagnostics, Indianapolis, Indiana 46250, USA). Protein electrophoresis sizes for control group and moderate externally exposed group were 12 was conducted as per Delk et al. (2014). Protein fractions were de- and 13 respectively. One control bird and two treated birds died during termined via the Helena SPIFE 3000 system using Split Beta gels (He- the study leaving a sample size of 11 for each group. lena Laboratories, Inc., Beaumont, Texas 77707, USA). Percentages and Oil exposed birds had a total of approximately 13 g applied to breast absolute values (g/dl) for each fraction were obtained by multiplying (6.5 g) and back (6.5 g) feathers by brush. Templates were used to the percentage by total protein concentration. The albumin:globulin ensure consistent area of application and provide a total surface area (A:G) ratio was calculated by dividing albumin by the sum of the glo- coverage equivalent to “moderate” oil coverage. Sham treatment of the bulin fractions. control birds included the same handling procedures and equivalent Bile acid concentrations were determined following validation ac- amount of water applied to the breast and back. Sham and oil appli- cording to Rayhel et al. (2015), by radioimmunoassay using the Con- cations took place every three days through day 15 of the trial (days 0, jugated Bile Acids Component System (MP Biomedicals, Santa Ana, 3, 6, 8, 12, and 15). By day 15 the dosed birds had become heavily California 92707, USA) according to manufacturer directions. Samples oiled, were showing clinical signs of oil toxicity such as postural and standardized controls were added to tubes coated in rabbit bile acid changes and abnormal feces, and had engaged in significant feather antiserum and incubated for 1 h with 125I-labeled bile acid. Afterward, plucking. No further oil application occurred at this time as we judged tubes were rinsed and read on a gamma counter (Laboratory Technol- that preening would continue to result in similar rates oil ingestion in ogies Inc., Maple Park, Illinois 60151, USA). A standard curve was the dose birds. Necropsy and sampling occurred on day 21 and 22. formulated from the controls and compared to sample results for quantification 2.5. Blood sampling 2.7. Statistical methods For the external oiling study, a single blood sample was collected at the beginning of quarantine. This collection did not take place for the 2.7.1. Screening for abnormal subjects orally dosed birds as there were initial concerns that capture and Reference intervals for clinical hematology and chemistry endpoints handling stress may be too great to warrant an additional stressor. (Harr et al., 2017b) were verified in accordance with the American Blood samples began with a sample on Day 0 (prior to dosing) and Society for Veterinary Clinical Pathology (ASVCP) guidelines using then every 7 days (oral dosing) or 6 days (external dosing). The slight MedCalc (Version 14.12.0 64 bit; MedCalc Software, Ostend, Belgium) difference between the two sampling protocols was necessary to ensure and a more stringent setting of the Dixon Test using confidence levels of that blood sampling was timed to occur on a day when oil application 0.1 or Tukey's Outlier Test (Geffre et al., 2011; Friedrichs et al., 2012). was also being undertaken. Oil was applied every three days (Day 0, 3, If outlying values were measured that indicated an abnormality, then 6, 9, 12 and 15) with blood being collected on days 0, 6, 12, 18 and 21. data collected from these individuals was deleted from the dataset. As Approximately 3–4 ml of blood was collected from either the bra- described in Cunningham et al. (2017) only birds assessed as healthy chial or jugular veins using a 25 G butterfly needle, then transferred to based on physical examination, body weight and appetite were in- labeled lithium heparin Vacutainer™ tubes and kept on ice for sub- cluded for the oral study. This screen for abnormal subjects at the be- sequent processing. Plasma was separated by centrifugation of whole ginning of the study removed outliers from analyses so only individuals blood at 2000 g for 5 min and stored at −80 °C until shipping, which representative of a healthy population were included in statistical generally occurred within 24–48 h of collection. Samples were scored analyses. In the external study, all birds were used regardless of severe for hemolysis using a 3-point scale, Samples receiving a score of “3” monocytosis to maintain animal number and better represent the field indicating severe hemolysis were excluded from analysis. setting, though animals with severe disease were separated between control and treated groups. 2.6. Sample analysis 2.7.2. Mortality and sample sizes All laboratory personnel were blinded to sample origin. Blood Samples sizes varied amongst the endpoints tested following both smears for WBC estimates were prepared using a standard push tech- testing for abnormal subjects (above) and mortality. As described by nique”. The slide was allowed to air dry, fixed in methanol, and sent to Cunningham et al. (2017), all group B dosed birds (10 ml/kg BW/day) a clinical pathology laboratory. Blood smears were stained using in the oral dosing study died or were euthanized by Day 14. Mortality Platinum Quik-Dip™ Stain (Mercedes Medical, Inc. 7590 Commerce related to the 5 ml/kg BW/day dosing schedule was limited to a single Court - Sarasota, FL 34243) in an automated slide stainer (Midas III individual, with no mortality of control birds. Mortality during the slide stainer, EMD Millipore, 290 Concord Road, Billerica, MA 01821). external dosing study was limited to a single control and two dosed Samples with >25% WBC lysis, as assessed on blood smear, were not birds. Each of the Figs. 1–5 reported here show individual points for all analyzed further. WBC count was estimated from the blood smear by of the birds from which samples were collected on that day. averaging number of WBC in 10 fields and multiplying by the square of the objective. Manual differentials quantified cell types based on a 200 2.7.3. Statistical analyses total cell count. White blood cell morphology including presence of Hematologic and plasma clinical chemistry values collected across thrombocyte clumping and adequacy were determined. Clumped multiple time points (during the experimental period) were compared thrombocytes were not enumerated but clumping subjectively indicated using linear mixed effects regression models with a repeated measures adequate numbers. structure. Regression models included effects for elapsed days, treat- Plasma samples were analyzed by a clinical pathology reference ment and a treatment*days interaction term. Elapsed days and treat- laboratory (University of Miami, Miller School of Medicine, Avian and ment were modeled as continuous variables, where treatment was de- Wildlife Laboratory). Analyses included plasma protein electrophoresis, fined as the average daily consumption determined from daily bile acids, sodium, potassium, chloride, phosphorus, calcium, alkaline observations of actual oil consumption by each individual bird or as the phosphatase (AP), alanine aminotransferase (ALT), aspartate amino- precise amount of externally applied oil. Statistically significant dif- transferase (AST), gamma glutamyl transferase (GGT), creatine kinase ferences (see Supplementary materials) between control and treatment (CK), glucose, cholesterol, urea, uric acid, total protein, creatinine. Uric groups were identified by significant main effects of treatment (p < acid concentrations were not measured on Day 0 for the orally treated 0.05) or by a significant interaction between treatment and elapsed

42 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51

Fig. 1. Regression analyses showing changes in leucocyte counts over time in double-crested cormorants exposed to oil via ingestion of oil-injected fish (LEFT panel) and via application of oil to feathers (RIGHT panel). Solid lines are controls, dashed lines are oil treated, dotted lines are boundaries of reference intervals. White blood cell counts (A & B); heterophils (C & D), eosinophils (E & F), lymphocytes (G & H) and monocytes (I & J) are all shown as cell ×109/L. Heterophil:lymphocyte ratio is shown in panels K & L. days (p < 0.05). Data distributions for all endpoints were generally 3. Results symmetric about their means and did not span more than an order of magnitude, thus data transformations to meet normality assumptions Raw data for avian toxicity studies conducted as part of the were deemed to be unnecessary. Differences among treatment groups Deepwater Horizon Damage Assessment are publicly available at on Day 0 were evaluated using the Kruskal-Wallis test (see https://www.diver.orr.noaa.gov/deepwater-horizon-nrda-data, while Supplementary materials). work plans and reports can be accessed through https://www.doi.gov/ deepwaterhorizon/adminrecord.

43 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51

Fig. 2. Regression analyses showing changes in enzyme activities over time in double-crested cormorants exposed to oil via ingestion of oil-injected fish (LEFT panel) and via application of oil to feathers (RIGHT panel). Solid lines are controls, dashed lines are oil treated. Aspartate aminotransferase (A & B), gamma glutamyl transferase (C & D), alanine aminotransferase (E & F) and creatinine phosphokinase (G & H)) are all shown as are all shown as U/l, dotted lines are boundaries of reference intervals.

3.1. Hematology heterophil:lymphocyte ratio among controls birds was relatively con- stant through time while the ratio among dosed birds declined, al- In the oral treatment study, total WBC, heterophil and eosinophil though not significantly (Fig. 1K). counts decreased through the course of the trial and the rate of decrease In the external oiling study total WBC, heterophil and eosinophil was greater among the dosed birds than control birds (p < 0.02, counts, decreased through the course of the trial in a manner similar to Fig. 1A; p < 0.001, Fig. 1C; p < 0.04, Fig. 1E, respectively). Lym- the results of the oral dosing study. However, in contrast to the oral phocyte and monocyte counts also decreased through time, but the treatment study, the rates of decline in WBC and eosinophil counts were rates of decrease among the control birds and dosed birds were not slightly greater among control birds than among dosed birds (p < 0.02 significantly different (Fig. 1G, Fig. 1I, respectively). The and p < 0.05, respectively) and significant differences in heterophil

44 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51

Fig. 3. Regression analyses showing changes in plasma analytes over time in double-crested cormorants exposed to oil via ingestion of oil-injected fish (LEFT panel) and via application of oil to feathers (RIGHT panel). Solid lines are controls, dashed lines are oil treated. Cholesterol (A & B), glucose (C & D), total protein(E & F), creatinine (G & H), urea (I & J) and uric acid (K & L) are all shown as mg/dl; dotted lines are boundaries of reference intervals. counts were not seen (Fig. 1A, Fig. 1B, Fig. 1C). Of clinical significance 3.2. Plasma clinical chemistry was that only treated birds had monocyte counts greater than 3.7 × 109/L. Through the course of the trial lymphocyte counts and monocyte 3.2.1. Enzymes counts increased slightly among dosed birds and decreased slightly In the oral exposure trial, AST, GGT, and ALT activities each de- among control birds (p < 0.006, Fig. 1H and p < 0.005, Fig. 1J, re- creased at a greater rate among dosed birds than among control birds (p spectively). The heterophil:lymphocyte ratio declined through the < 0.02, Fig. 2A; p < 0.01, Fig. 2C; and p < 0.03, Fig. 2E, respectively). course of the trial (p < 0.02), but the rate of decline was similar among In contrast, CK activity increased at a greater rate among dosed birds control birds and dosed birds (Fig. 1L). than among control birds, although the difference was not statistically significant (Fig. 2G).

45 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51

Fig. 4. Regression analyses showing changes in plasma mineral and salt concentrations over time in double-crested cormorants exposed to oil via ingestion of oil-injected fish (LEFT panel) and via application of oil to feathers (RIGHT panel). Solid lines are controls, dashed lines are oil treated. Calcium (A & B), chloride (C & D), sodium (E & F), potassium (G & H) and phosphorus (I & J) are all shown as mg/dl; dotted lines are boundaries of reference intervals.

Externally oiled birds showed a significant decrease in AST activity Although there were some significant changes, enzyme activities for through the course of the trial (p < 0.05, Fig. 2B). There was a slight most individual birds were within the reference intervals with the ex- decrease in GGT activity through time among dosed and control birds ception of ALT activity which was occasionally markedly increased. but there was no statistically significant difference between groups (Fig. 2D). ALT activity increased through time among dosed birds while ALT decreased among control birds (p < 0.003, Fig. 2F). CK activity 3.2.2. Metabolites was significantly greater amongst the dosed birds relative to the control Among orally dosed birds, cholesterol, glucose and total protein group throughout the study activity (p < 0.001; Fig. 2H), and activities concentrations each decreased through time while concentrations re- in both groups were relatively constant through the course of the study. mained relatively constant in control birds (p < 0.001, Fig. 3A; p < 0.001, Fig. 3C; and p < 0.001, Fig. 3E, respectively). Creatinine

46 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51

Fig. 5. Regression analyses showing changes in plasma protein concentrations over time in double-crested cormorants exposed to oil via ingestion of oil-injected fish (LEFT panel) and via application of oil to feathers (RIGHT panel). Solid lines are controls, dashed lines are oil treated. Pre-albumin (A & B), albumin (C & D), alpha −1-globulins (E & F), alpha-2-globulins (G & H), beta globulins (I & J) and gamma globulins (K & L) are all shown as g/dl. Albumin: globulin ratio is depicted in M and N; dotted lines are boundaries of reference intervals. concentration remained approximately constant in all treatment groups significantly lower in dosed birds (p<0.01) and both groups decreased (Fig. 3G), while urea concentration increased through time at a sig- slightly through time (p < 0.001, Fig. 3D). Total protein was sig- nificantly faster rate among dosed birds than in control birds (p < nificantly lower among dosed birds (p < 0.01) and concentrations 0.001, Fig. 3I). Uric acid concentration remained relatively constant decreased slightly through time among dosed birds (p < 0.02, Fig. 3F). among control birds but increased in the dosed groups (p < 0.002, Creatinine concentrations among control birds were stable through Fig. 3K) time but among dosed birds concentrations increased slightly through Among externally exposed birds, cholesterol concentration in- time (p < 0.001, Fig. 3H). Urea and uric acid concentrations each in- creased slightly through the study in both groups, although it increased creased significantly among dosed birds, while concentrations among faster in dosed birds (p < 0.02, Fig. 3B). Glucose concentration was control birds remained relatively constant (p < 0.001, Fig. 3J and p <

47 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51

0.001 Fig. 3L, respectively). Measurements for urea and uric acid ex- birds (p < 0.004, Fig. 5N). ceeded reference interval concentrations in dosed birds. 4. Discussion 3.2.3. Minerals and electrolytes In the oral ingestion trial, calcium concentrations in control birds Measurement of clinically relevant plasma endpoints are important in were relatively constant through time, while they decreased through determining not only the health status of an individual bird during the course of the trial among the dosed birds (p < 0.04, Fig. 4A). treatment and rehabilitation, but as research tools for determining the Chloride concentrations among control birds and dosed birds were both full extent of oil toxicity on a species, and assessment of potential po- relatively constant, although concentrations were slightly lower among pulation-wide effects of an oil spill. Changes in clinically relevant bio- the dosed groups (p<0.001, Fig. 4C). Sodium concentrations decreased chemical endpoints associated with toxicity of artificially weathered significantly through time (p < 0.001) in all groups, with no significant MC252 oil were observed using both dosing methods. Changes including differences between treatment groups (Fig. 4E). In contrast to sodium, increased phosphorus, urea and uric acid concentrations and decreased potassium concentrations increased significantly through time (p < total protein, albumin, and A:G ratio indicative of organ dysfunction 0.001) in all groups, with no significant differences among treatment were observed in both experiments and are consistent with previous groups (Fig. 4G). Phosphorus concentrations increased slightly among studies (Hartung and Hunt, 1966; Szaro et al., 1978; Patton and Dieter, dosed birds while they were relatively constant among control birds, 1980; Eastin and Rattner, 1982; Fleming et al., 1982; Pattee and Franson, but the distinction was not statistically significant (p < 0.06, Fig. 4I). 1982; Fry and Lowenstein, 1985; Lee et al., 1986; Leighton et al., 1985; Trends were different in the external exposure trial. Calcium con- Leighton, 1986; Hughes et al., 1990; Yamato et al., 1996; Walton et al., centrations among control birds were relatively constant, while they 1997; Newman et al., 2000; Seiser et al., 2000; Troisi et al., 2007). As increased slightly through the course of the trial among the dosed birds such it is feasible that there are clinically relevant changes in plasma (p < 0.03, Fig. 4B). Chloride concentrations increased through time biochemistries that can be identified for this species, that may be ap- among controls and dosed birds (p < 0.001), but there were no sig- plicable to other oil spills and species. While the volume of oil spilled and nificant differences between groups (Fig. 4D). Sodium concentrations the environmental conditions play a major role in avian mortality, we increased through time among controls and dosed birds (p < 0.001), cannot understand the full impact of an oil spill on a species until we can but the rate of increase was significantly greater among dosed birds (p determine how toxic it is compared to other spills. The Exxon Valdez < 0.03, Fig. 4F). Potassium concentrations among control and dosed killed hundreds of thousands of birds rapidly due to extreme tempera- groups were both relatively stable through the course of the trial, but tures and large volume of oil within a small area, but the toxicity of the mean concentrations were slightly higher among the dosed birds (p < oil itself, particularly to harlequin ducks (Histrionicus histrionicus), af- 0.03, Fig. 4F). Phosphorus concentrations increased markedly through fected survival rates for up to 14 years following the spill (Iverson and time among dosed birds, while they decreased slightly among control Esler, 2010). It is clear that although there will always be a range of birds (p < 0.001, Fig. 4J). species affected by a single oil spill event, that we need to have a better understanding of how to use standard clinical measurements as a means 3.2.4. Plasma protein electrophoresis of assessing damage to a species. In the oral ingestion trial, consistent with the overall decrease in The magnitude and number of changes that were measured in total protein (Fig. 3E) most of the protein fractions decreased. Pre-al- clinical plasma chemistry values indicate that the oral dosing method bumin concentrations decreased through time in control group and the had a greater degree of toxicity than the external dosing method. This is dosed groups (p < 0.001), but the rate of decrease was significantly likely due to relative dose effects. Orally dosed birds received 5.2 ± greater among the dosed groups (p < 0.03, Fig. 5A). Concentrations of 0.3 ml oil/kg BW and 8.4 + 0.9 ml oil/kg BW per day totaling 208 and albumin, alpha-1-globulins, alpha-2-globulins, beta globulins, gamma 235 ml respectively, while the externally dosed birds likely only con- globulins and the A:G ratio were each relatively constant among control sumed (based on Hartung, 1963) a total of 38–38.5 g over 21 days birds, while each of these endpoints in dosed birds decreased sig- (Cunningham et al., 2017), or approximately 0.9 ml/kg BW daily. There nificantly through the course of the trial (p < 0.001, Fig. 5C; p < were additional signs and symptoms of this toxicity reported in other 0.004, Fig. 5E; p < 0.02, Fig. 5G; p < 0.005, Fig. 5I; p < 0.001, manuscripts describing these studies (Alexander et al., 2017; Fig. 5K; p < 0.001, Fig. 5M, respectively). Cunningham et al., 2017; Harr et a, 2017a,c,d;Pritsos et al., 2017). While total protein concentration decreases were both clinically and These include increases in relative organ weight, hypertrophy and statistically significant, patterns of change in the plasma protein end- histopathological changes at necropsy, development of hemolytic an- points in the external exposure trial differed from those in the oral in- emia, inflammation and atrophy (Harr et al., 2017a, 2017c), as well as gestion trial and were not consistent amongst protein fractions. Pre- newly documented clotting dysfunction, cardiomyopathy and asso- albumin concentrations decreased slightly among both control birds ciated functional losses (Harr et al., 2017d). However, there was a and dosed birds (p < 0.04), with the rate of decrease slightly faster confounding variable for the orally dosed birds in that they developed among the dosed group (p < 0.04, Fig. 5B). Albumin concentrations some food aversion or gastrointestinal irritation that reduced their food decreased faster through the course of the trial among dosed birds than intake overall and caused weight loss, while externally dosed birds did among control birds, although the amount of change in both groups was not lose weight, and in fact increased food consumption (Cunningham relatively minor (p < 0.001, Fig. 5D). Alpha-1-globulin concentration et al., 2017). Reduced food intake and weight loss could exacerbate the increased faster through the course of the trial among dosed birds than problems faced by organ systems as they endeavor to detoxify the in- among control birds (p < 0.001, Fig. 5F). No significant treatment gested oil. While these two studies are useful in that they show similar effects or trends through time were evident in Alpha-2-globulin con- trends in clinically relevant plasma endpoints, further investigation is centrations (Fig. 5H). Beta globulin concentration were relatively stable required to conduct a direct comparison with similar doses. through time, but mean concentrations were significantly lower among the dosed birds (p < 0.002, Fig. 5J). Mean gamma globulin con- 4.1. Organ function centrations were also significantly lower among the dosed birds (p < 0.001) and concentrations among both groups decreased slightly Oxidative damage resulting from metabolism of PAHs is also linked through the course of the trial (p < 0.007, Fig. 5K). Trends in A:G ratio to organ damage and dysfunction in juvenile and adult birds. Effects differed significantly between the control and dosed groups, where the can include liver, kidney, adrenal, salt gland, spleen and gastro-in- ratio decreased among dosed birds but increased slightly among control testinal damage (Hartung and Hunt, 1966; Snyder et al., 1973; Gorman

48 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51 and Sims, 1978; Szaro, 1977; Holmes et al., 1970; Szaro et al., 1978; (Eastin and Rattner, 1982; Newman et al., 2000; Alonso-Alvarez et al., Gorsline and Holmes, 1981; Eastin and Murray, 1981; Fleming et al., 2007a, 2007b). Decreased A:G ratio found in both the oral and external 1982; Miller et al., 1979, 1982; Pattee and Franson, 1982; Fry and exposure studies supports loss either through the kidney or GI tract as Lowenstein, 1985; Leighton, 1986; Couillard and Leighton, 1990a, the smaller proteins such as albumin will be lost preferentially to the 1990b, Couillard and Leighton, 1991; Stubblefield et al., 1995; larger globulin molecules. While malabsorption of nutrients was in- Newman et al., 2000; Alonso-Alvarez et al., 2007a, 2007b; Duerr, dicated at necropsy and through anecdotal observation of abnormal 2013). excreta (Harr et al., 2017c), there were no specific changes in plasma Studies examining the effects of oil on birds have suggested that oil analytes that specifically assessed the gastrointestinal tract in the panel exposure can lead to hepatic damage or dysfunction, reporting de- that was tested. Malabsorption testing with xylose or similar GI func- creases in plasma cholesterol, glucose, albumin, uric acid and total tional assays could be used to further assess the gastrointestinal com- protein concentrations, and increases in ALP, ALT, AST, GGT, and bile promise in oral versus external exposure. acids (Eastin and Rattner, 1982; Stubblefield et al., 1995; Briggs et al., There is some indication that oil exposure also results in muscle 1997; Newman et al., 2000; Seiser et al., 2000; Golet et al., 2002; damage possibly due to direct oxidative damage or secondary to be- Alonso-Alvarez et al., 2007a, 2007b). Generally, birds with hepatic havioral or physical causes. Newman et al. (2000) suggest that recent dysfunction or disease will exhibit declines in albumin, cholesterol, net capture may cause elevations in AST and creatine kinase activities glucose, total protein, uric acid and increases in bile acids, alkaline as would be expected. Elevated creatine kinase activity was found in phosphatase, AST, GGT and LDH (Harr, 2005; Hochleitner et al., 2005). orally and externally oiled birds but not in control birds. It increased Orally dosed birds had significant declines in cholesterol, glucose, total over time in captivity and did appear to increase with dose in the oral protein concentrations as well as all protein fractions over the dosing experiment. While this could be related to the direct impacts of in- period, while externally oiled birds showed reductions in glucose, al- creased dosage, other factors such as excessive preening, poorly con- bumin, and total protein levels, and an increase in ALT. These func- trolled movements due to poor feather integrity, or muscular injuries tional differences between the two dosing methods likely reflect the may have contributed to the increased activities of these enzymes found extent of damage to the liver caused by differences in daily and absolute within the myocyte cytosol. dosing. Orally dosed birds actually exhibited declines in GGT and AST that could be indicative of later stages of liver damage, whereby pro- 4.1.1. Potential immune and adrenal changes duction capabilities are lost. AST is also produced in muscle and other Immunosuppression has been reported in both oil dosing studies organs. As the body condition and muscle mass also declined in the and during rehabilitation of oiled birds (Rocke et al., 1984; Leighton, orally dosed group, this may have also contributed to the decreased AST 1986; Khan and Ryan, 1991; Anderson et al., 1996; Newman et al., activity (Cunningham et al., 2017; this edition). 2000); however, data interpretation is often confounded by capture/ Increases in plasma urea, uric acid, and phosphorus concentrations handling stress and presence of pre-existing infections or injuries in found in both orally and externally exposed birds are indicative of renal wild birds, particularly in the short term assessments (Harvey et al., insufficiency at both the glomerular and tubular levels (Braun, 2015) 1982; Leighton et al., 1986; McOrist and Lenghaus, 1992; Briggs et al., and have also been reported in avian oil exposure studies (Hartung and 1997; Newman et al., 2000). In these experiments, immune measure- Hunt, 1966; Eastin and Murray, 1981). Both orally and externally dosed ments were limited due to the nature of the study design. There were no birds had significant increase in both urea and uric acid concentrations increases in gamma globulins in either the oral or external dosing ex- that were above the reference intervals. Urea concentration can also periments, so we used WBC and differential cell counts as generalized increase with water loss, which was supported by increased sodium indicators of immune system function. Results were not strongly sup- concentration in this study. Birds were housed with pools of fresh water portive of oil-induced immunocompromise. Monocyte counts in the changed regularly. Thus, dehydration in the face of water ad libitum external dosing study (controls and treated birds) were an order of further supports renal insufficiency. Specific gravity of urates was not magnitude higher than the oral dosing study, but all other leucocyte possible due to the aquatic nature of the animal and fecal mixing which counts were similar. As stated in the Methods section, a blood sample precluded confirmation of lack of urinary concentration. While in- was collected immediately post-capture for the external dosing study to creased sodium was measured in the externally oiled birds, control determine if there were any pre-existing conditions present that could birds in that experiment also had slightly increased sodium con- affect data interpretation. This was in fact the case for the monocyte centration, indicating that there may have been some kind of stress- counts, so any counts greater than 2.0 × 109 cell/L were considered to related impact on electrolyte balance, through ACTH release be abnormal and a chronic infection was suspected. Since both these (Olanrewaju et al., 2007). Orally dosed birds, including controls, all experiments were considered as pilot or scoping studies, birds with high showed reductions in plasma sodium concentration over the course of monocyte counts were divided evenly between the control and treated the study, indicating improved hydration when housed with freshwater groups as comparisons between the two types of dosing were not an only, and potential adaptation to captivity. essential component of the study design. Histopathology following ne- The gastrointestinal irritation caused by oil can be more difficult to cropsy identified parasitic, bacterial or fungal infections in many in- measure, but usually manifests as a deterioration in nutritional state dividuals in the external oiling study (Harr et al., 2017c), including including reduced weight gain in chicks, hyperphagia and weight loss those with high monocyte counts, but there did not appear to be a trend in adults, as well as abnormal excreta (Hartung and Hunt, 1966; that set apart those with high monocyte counts. Holmes et al.; Snyder et al., 1973; Snyder et al., 1978; Patton and There were statistically significant decreases in some leucocytes; Dieter, 1980; Szaro et al., 1981; Fleming et al., 1982; Pattee and however, the decreases also occurred in control birds. In the oral dosing Franson, 1982; Trivelpiece et al., 1984; Hughes et al., 1990; Evans and study, WBC, heterophil, and eosinophil counts decreased over time, Keijl, 1993). Malabsorption of nutrients can manifest as low plasma with the greatest decreases being in oil-dosed birds, while lymphocyte protein and decreased chloride concentration, which were observed in counts decreased similarly in all three groups over the course of the these dosing studies. As stated above, decreased total protein and study. Control birds in the external oiling experiment showed a greater protein fractions may be due to decreased production by the liver but decline in white cell estimate and lymphocyte counts than treated birds, loss through either the gastrointestinal tract or kidney may contribute while heterophil numbers declined similarly between control and to statistically significant cumulative decrease found. Additionally, re- treated birds. It is possible that decreases in lymphocyte numbers are duced food intake and/or impaired intestinal transport were also sug- indicators of a captivity-induced stress response (Leighton, 1986; Briggs gested as possible mechanisms for the decreased plasma measurements et al., 1996), particularly since they were observed in control birds. As though this has only been proven in the most cachectic individuals such, we also looked at the heterophil/Lymphocyte ratio, increases in

49 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51 which are often used as an estimate of activation of the hypothalamic- patience and excellent work on sample analysis. pituitary-adrenal (HPA) axis, due to decreases in lymphocyte numbers caused by glucocorticoids (Martin, 2009). There were statistically sig- Appendix A. Supplementary material nificant decreases in the heterophil/Lymphocyte ratio in the treated birds in both experiments, but only in controls for the oral dosing study. Supplementary data associated with this article can be found in the While oil exposure has been linked to adrenal dysfunction and hyper- online version at http://dx.doi.org/10.1016/j.ecoenv.2017.08.007. trophy in birds (Peakall et al., 1981; Rattner and Eastin, 1981; Gorsline and Holmes, 1982; Lattin et al., 2014, Lattin and Romero, 2014), due to References the limited scope of these studies adrenocorticotrophic hormone sti- mulation of the HPA axis could not be conducted to determine baseline Alexander, C.R., Hooper, M.J., Cacela, D., Smelker, K.D., Calvin, C.S., Dean, K.M., or HPA activation in response to captivity and oil exposure. Bursian, S.J., Cunningham, F.L., Hanson-Dorr, K.C., Horak, K.E., Isanhart, Link, J., ff Shriner, S.A., Godard-Codding, C.A.J., 2017. CYP1A protein expression and catalytic While the e ects of captivity-induced stress and pre-existing are activity in double-crested cormorants experimentally exposed to deepwater Horizon likely to contribute to changes in leucocyte counts, the anemia observed Mississippi Canyon 252 oil. Ecotoxicol. Environ. Saf. http://dx.doi.org/10.1016/j. in treated birds in both experiments (Harr et al., 2017a) is also likely to ecoenv.2017.05.015. Alonso-Alvarez, C., Munilla, I., Lopez-Alonso, M., Velando, A., 2007a. Sublethal toxicity have caused a shift by the bone marrow towards erythropoiesis as ob- of the Prestige oil spill on yellow-legged gulls. Environ. Int. 33, 773–781. served by Leighton (1986) in herring gulls and Atlantic puffins dosed Alonso-Alvarez, C., Perez, C., Velando, A., 2007b. Effects of acute exposure to heavy fuel with Prudhoe Bay crude. Thymus and bursa of Fabricius regress in most oil from the Prestige spill on a seabird. Aquat. Toxicol. 84 (1), 103–110. http://dx. avian species during development, and whilst involuted bursas were doi.org/10.1016/j.aquatox.2007.06.004. Anderson, D.W., Gress, F., Fry, D.M., 1996. Survival and dispersal of oiled brown pelicans collected from a small number of individuals during necropsy, histo- after rehabilitation and release. Mar. Pollut. Bull. 32 (10), 711–718. pathology did not report anything of note (Harr et al., 2017c). It has Braun, E.J., 2015. Chapter 12: osmoregulatory systems of birds. In: Scanes, C.G. (Ed.), – been suggested that oil toxicity has more impact on cell-mediated im- Sturkies AvianPhysiology 6. Academic Press Elsevier, pp. 285 299 (ed). Briggs, K.T., Yoshida, S.H., Gershwin, M.E., 1996. The influence of petrochemicals and mune responses than antibody-mediated responses (Briggs et al., 1997); stress on the immune system of seabirds. Regul. Toxicol. Pharmacol. 23, 145–155. however, in this study the only measurements of cell-mediated immune Briggs, K.T., Gershwin, M.E., Anderson, D.W., 1997. Consequences of petrochemical in- function were WBC and differential counts, so interpretation of the gestion and stress on the immune system of seabirds. ICES J. Mar. Sci. 54, 718–725. fi ff Couillard, C.M., Leighton, F.A., 1990a. Sequential study of the pathology of Prudhoe Bay speci ce ects of oil are somewhat limited. While albumin, pre-al- crude oil in chicken embryos. Ecotoxicol. Environ. Saf. 19, 17–23. bumin and gamma globulin decreased during the study indicating Couillard, C.M., Leighton, F.A., 1990b. The toxicopathology of Prudhoe Bay crude oil in malabsorption and liver disease, alpha-1-globulin increased potentially chicken embryos. Fundam. Appl. Toxicol. 14, 30–39. fl Cunningham, F.L., Dean, K., Hanson-Dorr, K.C., Harr, K., Healy, K., Horak, K., Link, J., indicating in ammation. Shriner, S., Bursian, S., Dorr, B., 2017. Development of methods for avian oil toxicity studies using the double-crested cormorant (Phalacocorax auritus). Ecotoxicol. 4.2. Conclusions Environ. Saf. 141, 199–208. http://dx.doi.org/10.1016/j.ecoenv.2017.03.025. Delk, K.W., Wach, F.F., Burgdorf-Moisuk, A., Kass, P.H., Cray, C., 2014. Acute phase protein and electrophoresis protein fraction values for captive American flamingos. J. Following an oil spill the first measurements made of oil coated Zoo. Wildl. Med. 46 (4), 929–933. birds are naturally used to determine their health status. While sam- Duerr, R.S., 2013. Investigation into nutritional condition and digestive capabilities of pling oiled birds for health assessment is paramount in ensuring the seabirds during rehabilitation in California (PhD). University of California Davis. Eastin, W.C.J., Murray, H.C., 1981. Effects of crude oil ingestion on avian intestinal highest likelihood of survival for each individual, it somewhat limits function. Can. J. Physiol. Pharmacol. 59, 1063–1068. our understanding of oil toxicity, and ability extrapolate of potential Eastin, W.C.J., Rattner, B.A., 1982. Effects of dispersant and crude oil ingestion on – population effects. Additionally, our understanding of the toxicological mallard ducklings (Anas platyrhynchos). Bull. Environ. Contam. Toxicol. 29, 273 278. ff Evans, M.I., Keijl, G.O., 1993. Impact of Gulf War oil spills on the wader populations of e ects of oil on avian physiology have been somewhat hampered by the the Saudi Arabian Gulf coast. Sandgrouse 15, 85–105. difficulties associated with developing reliable and reproducible dosing Fleming, W.J., Sileo, L., Franson, J.C., 1982. Toxicity of Prudhoe Bay crude oil to sandhill – paradigms. In this study, we tested two oil dosing methods on double- cranes. J. Wildl. Manag. 46 (2), 474 478. Fry, D.M., Lowenstine, L.J., 1985. Pathology of common murres and Cassin's auklets crested cormorants that would result in oral ingestion, and incorporated exposed to oil. Arch. Environ. Contam. Toxicol. 14, 725–737. both clinical health measurements and toxicological assessments to Garcia, M.T.A., Hermosa, Y., Aguirre, J.I., 2010. Does breeding status influence haema- determine the overall impact of artificially weather MC252 on cor- tology and blood biochemistry of yellow-legged gulls? Acta Biol. Hung. 61 (4), 391–400. morant physiology. Although limited dose ranges were used, both Golet, G.H., Seiser, P.E., McGuire, A.D., Roby, D.D., Fischer, J.B., Kuletz, K.J., Newman, methods resulted in impacts to clinical plasma chemistry measurements S.H., 2002. Long-term direct and indirect effects of the 'Exxon Valdez' oil spill on that are supported by necropsy findings and the broader avian oil pigeon guillemots in Prince William Sound, Alaska. Mar. Ecol. Prog. Ser. 241, fi 287–304. toxicity literature. We concluded that clinically signi cant increases in Gorman, M.L., Simms, C.E., 1978. Lack of effect of ingested Forties Field crude oil on urea, uric acid, and phosphorus concentrations combined with de- avian growth. Mar. Pollut. Bull. 9 (10), 273–276. creases in total protein, albumin, and A:G ratio were consistently found Gorsline, J., 1982. The Effects of South Louisiana Crude Oil on Adrenocortical Function. in orally and externally exposed double-crested cormorants using arti- Aspects of AvianEndocrinology: Practical and Theoretical Implications. Texas Tech Press, Lubbock, TX, pp. 359–364. ficially weathered MC252 oil. Further experimentation is needed is to Gorsline, J., Holmes, W.N., 1981. Effects of petroleum on adrenocortical activity on he- more precisely determine the doses and exposure durations that have patic naphthalene-metabolizing activity in mallard ducks. Arch. Environ. Contam. – the potential to cause adverse physiological outcomes and to fine-tune Toxicol. 10, 765 777. Harr, K.E., 2005. Ch 23: diagnostic values of biochemistry. In: Harrison, G., Lightfoot, T. the clinical measurements indicative of organ damage and function loss (Eds.), Clinical AvianMedicine 1. Spix Publishing, pp. 611–630. that can be extrapolated beyond the MC252 oils. Harr, K.E., Cunningham, F.L., Pritsos, C.A., Pritsos, K.L., Muthumalage, T., Dorr, B.S., Horak, K.E., Hanson-Dorr, K.C., Dean, K.M., Cacela, D., Link, J.E., Healy, K., Tuttle, Acknowledgements P., Bursian, S.J., 2017a. Weathered MC252 crude oil-induced anemia and abnormal erythroid morphology in double-crested cormorants (Phalacrocorax auritus) with light microscopic and ultrastructural description of Heinz bodies. Ecotoxicol. Environ. Saf. The studies appearing in this special issue were funded by the U.S. http://dx.doi.org/10.1016/j.ecoenv.2017.07.030. Harr, K.E., Dean, K., Hanson-Dorr, K., Healy, K., McFadden, A., Cunningham, F.L., 2017b. Fish and Wildlife Service (Order No. F12PD0106) as part of the Hematologic and biochemical reference intervals in double-crested cormorants Deepwater Horizon Natural Resource Damage Assessment. Special (Phalacrocorax auritis) from the southeastern United States. Vcp. Prog. thanks to Paul Fioranelli, Alexander Crain, Lanna Durst, Raleigh Harr, K.E., Reavill, D.R., Bursian, S.J., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Shriner, S.A., Schmidt, Middleton, Christine Ellis, Nicole Mooers and Jeremy Ellis for their R.E., 2017c. Organ weights and histopathology of double-crested cormorants technical assistance with animal care and sampling on this project. (Phalacrocorax auritus) dosed orally or dermally with artificially weathered Also, thanks to Dr. Carolyn Cray and Marilyn Rodriguez at the Mississippi Canyon 252 crude oil. Ecotoxicol. Environ. Saf. http://dx.doi.org/10. University of Miami Comparative Pathology Laboratory for all of their 1016/j.ecoenv.2017.07.011.

50 K.M. Dean et al. Ecotoxicology and Environmental Safety 146 (2017) 40–51

Harr, K.E., Rishniw, M., Rupp, T.L., Cacela, D., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C., Newman, S.H., Anderson, D.W., Ziccardi, M.H., Trupkiewicz, J.G., Tseng, F.S., Healy, K., Horak, K., Link, J.E., Reavill, D., Bursian, S.J., Cunningham, F.L., 2017d. Christopher, M.M., Zinkl, J.G., 2000. An experimental soft-release of oil-spill re- Dermal exposure to weathered MC252 crude oil results in echocardigraphically habilitated American coots (Fulica americana): II. Effects on health and blood para- identifiable systolic myocardial dysfunction in double crested cormorants meters. Environ. Pollut. 107 (3), 295–304. (Phalacrocorax auritus). Ecotoxicol. Environ. Saf. Accept. http://dx.doi.org/10.1016/ Olanrewaju, H.A., Thaxton, J.P., Dozier 3rd, W.A., Branton, S.L., 2007. Electrolyte diets, j.ecoenv.2017.04.010. stress and acid-base balance in broiler chickens. Poult. Sci. 86 (7), 1363–1371. Hartung, R., 1963. Ingestion of oil by waterfowl. Pap. Mich. Acad. Sci. Arts Lett. 48, Pattee, O.H., Franson, J.C., 1982. Short-term effects of oil ingestion on American kestrels 49–55. (Falco sparverius). J. Wildl. Dis. 18 (2), 235–241. Hartung, R., Hunt, G.S., 1966. Toxicity of some oils to waterfowl. J. Wildl. Manag. 30 (3), Patton, J.F., Dieter, M.P., 1980. Effects of petroleum hydrocarbons on hepatic function in 564–570. the duck. Comp. Biochem. Physiol. 65C, 33–36. Harvey, S., Phillips, J.G., Sharp, P.J., 1982. Reproductive Performance and Endocrine PDARP, 2016. Deepwater Horizon Oil Spill: Final programmatic damage assessment and Responses to Ingested North Sea Oil. Aspects of AvianEndocrinology: Practical and restoration plan and final programmatic environmental impact statement. http:// Theoretical Implications. Texas Tech Press, Lubbock, TX, pp. 379–395. www.gulfspillrestoration.noaa.gov/sites/default/files/wp-content/uploads/Front- Hochleitner, M., Hochleitner, C., Harrison, L.D., 2005. Evaluating and treating the liver. Matter-and-Chapter-1_Introduction-and-Executive-Summary_508.pdf. In: Harrison, G., Lightfoot, T. (Eds.), Clinical Avian Medicine 1. Spix Publishing, pp. Peakall, D.B., Tremblay, J., Kinter, W.B., Miller, D.S., 1981. Endocrine dysfunction in 441–450. seabirds caused by ingested oil. Environ. Res. 24 (1), 6–14. Holmes, W.N., Gorsline, J., Cronshaw, J., 1970. Effects of mild cold stress on the survival Peakall, D.P., Norstrom, R.J., Jeffrey, D.A., Leighton, F.A., 1989. Induction of hepatic of seawater-adapted mallard ducks (Anas platyrhynchus) maintained on food con- mixed function oxidases in the herringg gull (Larus agrentatus) by Pruhoe Bay crude taminated with petroleum. Environ. Res. 20, 425–444. oil and its fractions. Comp. Biochem. Physiol. 94C (2), 461–463. Hughes, M.R., Kasserra, C., Thomas, B.R., 1990. Effect of externally applied bunker fuel Perez, C., Lores, M., Velando, A., 2010. Oil pollution increases plasma antioxidants but on body mass and temperature, plasma concentration, and water flux of glaucous- reduces coloration in a seabird. Oecologia 163, 875–884. winged gulls, Larus glaucescens. Can. J. Zool. 68, 716–721. Pritsos, K.L., Perez, C.R., Muthumalage, T., Dean, K.M., Cacela, D., Hanson-Dorr, K.C., Iverson, S.A., Esler, D., 2010. Harlequin duck population injury and recovery dynamics Cunningham, F.L., Bursian, S.J., Link, J.E., Shriner, S.A., Horak, K.E., Pritsos, C.A., following the 1989 Exxon Valdez oil spill. Ecol. Appl. 20 (7), 1993–2006. 2017. Dietary intake of Deepwater Horizon oil-injected live food fish by double- Lattin, C.R., Romero, L.M., 2014. Chronic exposure to a low dose of ingested petroleum crested cormorants resulted in oxidative stress. Ecotoxicol. Environ. Saf. http://dx. disrupts corticosterone receptor signalling in a tissue-specific manner in the house doi.org/10.1016/j.ecoenv.2017.06.067. sparrow (Passer domesticus). Conserv Physiol. 2 (1), cou058. http://dx.doi.org/10. Rayhel, L., Aitken-Palmer, C., Joyner, P., Cray, C., Lizarraga, C.A., Ackermans, B., Crowe, 1093/conphys/cou058. C., 2015. Hematology and biochemistry in captive white-naped cranes (Grus vipio). J. Lattin, C.R., Ngai, H.M., Romero, L.M., 2014. Evaluating the stress response as a bioin- Zoo. Wildl. Med. 46 (4), 747–754. dicator of sub-lethal effects of crude oil exposure in wild house sparrows (Passer Rocke, T.E., Yuill, T.M., Hinsdill, R.D., 1984. Oil and related toxicant effects on mallard domesticus). PLoS ONE 9 (7), e102106. http://dx.doi.org/10.1371/journal.pone. immune defenses. Environ. Res. 33 (2), 343–352. 0102106. Seiser, P.E., Duffy, L.K., McGuire, A.D., Roby, D.D., Golet, G.H., Litzow, M.A., 2000. Lee, Y.-Z., O'Brien, P.J., Payne, J.F., Rahimtula, A.D., 1986. Toxicity of petroleum crude Comparison of pigeon guillemot, Cepphus columba, blood parameters from oiled and oils and their effect on xenobiotic metabolizing enzyme activities in the chicken unoiled areas of Alaska eight years after the Exxon Valdez oil spill. Mar. Pollut. Bull. embryo in Ovo. Environ. Res. 39 (1), 153–163. 40 (2), 152–164. Leighton, F.A., 1986. Clinical, gross, and histological findings in herring gulls and Atlantic Snyder, S.B., Fox, J.G., Soave, O.A., 1973. Mortalities in Waterfowl Following a Bunker C puffins that ingested Prudhoe Bay crude oil. Vet. Pathol. 23, 254–263. Fuel Exposure. Retrieved from. Leighton, F.A., 1993. The toxicity of petroleum oils to birds. Environ. Rev. 1, 92–103. Stubblefield, W.A., Hancock, G.A., Prince, H.H., Ringer, R.K., 1995. Effects of naturally Leighton, F.A., Lee, Y.Z., Rahimtula, A.D., O'Brien, P.J., Peakall, D.B., 1985. Biochemical weathered Exxon Valdez crude oil on mallard reproduction. Environ. Toxicol. Chem. and functional disturbances in red blood cells of herring gulls ingesting Prudhoe Bay 14 (11), 1951–1960. crude oil. Toxicol. Appl. Pharmacol. 81 (1), 25–31. Szaro, R.C., Dieter, M.P., Heinz, G.H., 1978. Effects of chronic ingestion of south Martin, L.B., 2009. Stress and immunity in wild vertebrates: timing is everything. Gen. Louisiana crude oil on mallard ducklings. Environ. Res. 17 (3), 426–436. Comp. Endocrinol. 163 (1–2), 70–76. http://dx.doi.org/10.1016/j.ygcen.2009.03. Szaro, R.C., Hensler, G., Heinz, G.H., 1981. Effects of chronic ingestion of No. 2 fuel oil on 008. mallard ducklings. J. Toxicol. Environ. Health 7, 789–799. McOrist, S., Lenghuas, C., 1992. Mortalities of little penguins (Eudyptula minor) following Troisi, G., Borjesson, L., Bexton, S., Robinson, I., 2007. Biomarkers of polycyclic aromatic exposure to crude oil. Vet. Rec. 130, 161–162. hydrocarbon (PAH)-associated hemolytic anemia in oiled wildlife. Environ. Res. 105 Miller, D.S., Kinter, W.B., Peakall, D.B., 1979. Effects of Crude Oil Ingestion on Immature (3), 324–329. Pekin Ducks (Anas platyrhynchos) and Herring Gulls (Larus argentatus) Animals as Walton, P., Turner, C.M.R., Austin, G., Burns, M.D., Monaghan, P., 1997. Sub-lethal ef- Monitors of Environmental Pollutants. National Academy of Sciences, Washington, fects of an oil pollution incident on breeding kittiwakes Rissa tridactyla. Mar. Ecol. D.C., pp. 27–40. Prog. Ser. 155, 261–268. Miller, D.S., Hallett, D.J., Peakall, D.B., 1982. Which components of crude oil are toxic to Yamato, O., Goto, I., Maede, Y., 1996. Hemolytic anemia in wild seaducks caused by young seabirds? Environ. Toxicol. Chem. 1 (1), 39–44. marine oil pollution. J. Wildl. Dis. 32 (2), 381–384.

51 Ecotoxicology and Environmental Safety 146 (2017) 52–61

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Organ weights and histopathology of double-crested cormorants MARK (Phalacrocorax auritus) dosed orally or dermally with artificially weathered Mississippi Canyon 252 crude oil ⁎ Kendal E. Harra, , Drury R. Reavillb, Steven J. Bursianc, Dave Cacelad, Fred L. Cunninghame, Karen M. Deand, Brian S. Dorre, Katie C. Hanson-Dorre, Kate Healyf, Katherine Horakg, Jane E. Linkc, Susan Shrinerg, Robert E. Schmidtb a Urika Pathology, Mukilteo, WA 98275 USA b Zoo/Exotic Pathology Service, 6020 Rutland Drive #14, Carmichael, CA 95608, USA c Department of Animal Science, Michigan State University, 474 South Shaw Lane, East Lansing, MI 48824, USA d Abt Associates, 1881 Ninth St., Ste 201, Boulder, CO 80302-5148, USA e USDA WS National Wildlife Research Center, P.O. Box 6099, Mississippi State University, Starkville, MS 39762, USA f DWH NRDAR Field Office, USFWS, 341 Greeno Road North, Suite A, Fairhope, AL 36532, USA g USDA WS National Wildlife Research Center, Fort Collins, CO 80521 USA

ARTICLE INFO ABSTRACT

Keywords: A series of toxicity tests were conducted to assess the effects of low to moderate exposure to artificially Deepwater Horizon weathered Deepwater Horizon Mississippi Canyon 252 crude oil on representative avian species as part of the Avian Natural Resource Damage Assessment. The present report summarizes effects of oral exposure (n=26) of double- − − Oil crested cormorants (DCCO; Phalacrocorax auritus) to 5 or 10 ml oil kg 1 day 1 for up to 21 days or dermal Organ weights application (n=25) of 13 ml oil to breast and back feathers every three days totaling 6 applications in 21 days on Histopathology organ weights and histopathology. Absolute and relative kidney and liver weights were increased in birds ex- posed to oil. Additionally, gross and/or histopathologic lesions occurred in the kidney, heart, pancreas and thyroid. Clinically significant renal lesions in the orally dosed birds included squamous metaplasia and increased epithelial hypertrophy of the collecting ducts and renal tubules and mineralization in comparison to controls. Gross cardiac lesions including thin walls and flaccid musculature were documented in both orally and dermally dosed birds and myocardial fibrosis was found in low numbers of dermally dosed birds only. Cytoplasmic va- cuolation of the exocrine pancreas was noted in orally dosed birds only. Thyroid follicular hyperplasia was increased in dermally dosed birds only possibly due to increased metabolism required to compensate damaged feather integrity and thermoregulate. Gastrointestinal ulceration was found in orally dosed birds only. There were no significant hepatic histopathologic lesions induced by either exposure route. Therefore, hepatic histo- pathology is likely not a good representation of oil-induced damage. Taken together, the results suggest that oral or dermal exposure of DCCOs to artificially weathered MC252 crude oil induced organ damage that could potentially affect survivability.

1. Introduction interest to examine the effects of these lower amounts of oil (less than 30% of body coverage) in order to assess avian injury as part of the During the Deepwater Horizon (DWH) oil spill in 2010, many live DWH Mississippi Canyon 252 (MC252) Oil Spill Natural Resource Da- birds representing at least 93 species were found visibly oiled but not to mage Assessment (NRDA). the extent to cause immediate mortality (Deepwater Horizon Natural Diving birds such as the double-crested cormorant (DCCO; Resource Damage Assessment Trustees, 2016). It was of particular Phalacrocorax auritus) are particularly susceptible to oil exposure

⁎ Corresponding author. E-mail addresses: [email protected] (K.E. Harr), [email protected] (D.R. Reavill), [email protected] (S.J. Bursian), [email protected] (D. Cacela), [email protected] (F.L. Cunningham), [email protected] (K.M. Dean), [email protected] (B.S. Dorr), [email protected] (K.C. Hanson-Dorr), [email protected] (K. Healy), [email protected] (K. Horak), [email protected] (J.E. Link), [email protected] (S. Shriner), [email protected] (R.E. Schmidt). http://dx.doi.org/10.1016/j.ecoenv.2017.07.011 Received 14 September 2016; Received in revised form 3 July 2017; Accepted 3 July 2017 Available online 19 July 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 52–61 during a spill because they dive through the water, and thus the oil, in neutral buffered formalin for histological examination. order to feed (Szaro et al., 1978). They are therefore likely to be ex- posed not only dermally, but also by incidental ingestion through at- 2.4. Histopathology tempts to rid their feathers of oil through preening and possibly by ingesting oil through contaminated feedstuffs. In birds, anemia, dis- Tissues were paraffin embedded, sectioned at approximately 5 µm, rupted feather function, hypothermia, respiratory distress, seizures, affixed to glass microscope slides and stained with hematoxylin and diarrhea, hepatic disease and renal disease have all been reported eosin. Organs were examined by board certified veterinary pathologists secondary to exposure to petroleum products (Mazet et al., 2002). in the oral (RES) and dermal (DRR) study. Lesions were graded using The two sets of data reported here address the effects of both oral the scale 1 = minimal, 2 = mild, 3 = moderate, 4 = severe that both and dermal oiling on the relevant oil exposure endpoints of organ pathologists developed. Hepatic iron was assessed using standard he- weights and histopathology in the DCCO. matoxylin and eosin staining and Prussian blue staining, which is spe- cific for iron, using the same grading used for lesions (Khan and Nag, 2. Methods 1993) and was examined by RES only. Quality assurance was per- formed by a third boarded pathologist, Dr. Jennifer Brazzell, to insure This study was performed under the authority of USFWS MBPO consistent results. Federal Permit # MB019065-3, Mississippi and Alabama state (#8017) scientific collection permits, and Institutional Animal Care and Use 3. Results Committee (IACUC) under NWRC protocol QA-2326. Cunningham et al. (2017, this issue) provides a detailed description of animal capture and 3.1. Mortality handling for the experimental oral and dermal exposure of DCCO to DWH artificially weathered MC252 oil. Of the 26 adult, mixed-sex DCCO used in the oral dose study, 16 were euthanized on Day 21. A total of 10 treated DCCOs died or were 2.1. Oral dosing study euthanized within 17 days of the start of the study for humane reasons, including all 9 high dose animals. DCCO began exhibiting clinical signs Captured DCCOs were randomly assigned to one of three treatment such as anemia, abnormal feces, lethargy, and behavioral thermogen- groups: a control group (n = 8, 7 male, 1 female) that was fed catfish esis (shivering under a heat lamp) at a total dose of approximately that had been lightly anesthetized with MS222 and allowed to revive; a 80 ml/kg and all were dead prior to 200 ml/kg total dose. group dosed daily with up to 5 ml oil/kg bw/day through provision of Of the 25 subadult, mixed-sex DCCO used in the dermal exposure oil-containing, lightly anesthetized catfish (n = 9, 6 male, 3 female); a study one control bird and two treated birds died prior to Day 21 of the group dosed daily with up to 10 ml oil/kg bw/day through provision of study. DCCO began exhibiting clinical signs such as anemia, hema- oil-containing, lightly anesthetized catfish as described below (n = 9, 7 tochezia, behavioral thermogenesis, decreased appetite, and lethargy male, 2 female). by day 10. One exposed bird died with probable septicemia (underlying etiologic agent not identified). One exposed bird died with no sig- 2.2. Dermal dosing study nificant lesions that could be assessed as a cause of death. A chronic, necrotizing granuloma was found at the heart base of the control bird at A total of 31 DCCO's were captured and retained in captivity. Birds necropsy. were allowed to acclimate to captivity for a minimum of 21 days prior to initiation of the study. A total of 25 subadult DCCOs allocated to a 3.2. Absolute and relative organ weights control group (n=12, 5 male, 7 female) and an exposed group (n=13, 6 males, 7 females) were used in this trial. DCCOs were assigned to Absolute and relative kidney weights were significantly greater in treatment groups based on the results of blood samples collected at the orally and dermally dosed DCCOs compared to their respective controls initiation of the three-week quarantine period. Complete blood count (Fig. 1a and b). Absolute and relative liver weights were significantly − − (CBC) values were used to ensure equal division of birds with potential greater in DCCOs orally dosed with 5 ml oil kg bw 1 day 1 and in health concerns between groups. DCCO's with monocyte counts greater DCCOs that were dermally oiled compared to controls (Fig. 2a and b). than 2.0×109 cells/l were considered abnormal (severe monocytosis); There was no significant difference in absolute or relative (expressed as and were divided between control (n=4) and treatment (n=3) groups. % body weight) brain, heart or spleen weight in orally or dermally Additionally, a small oil spill took place one year prior to the study, not dosed DCCOs compared to controls. far from where 6 of the DCCOs were collected and were evenly dis- tributed between groups. During the course of the trial, one bird from 3.3. Histopathology the control group and two birds from the treatment group died and were not replaced. Therefore, the final number of birds in the control A number of histological lesions were found in tissues of DCCOs and exposed group was 11 birds each to total 22 in the study. Oil on dosed orally or dermally with artificially weathered MC252 oil exposed birds (13 ml) and water on control birds (13 ml) was applied (Table 1). Organ weights from both exposure groups are summarized in every three days through Day 15 of the trial (on Days 0, 3, 6, 9, 12, and Table 2. 15). Detailed description of application is available in Cunningham et al. (2017). 3.3.1. Oral dosing study Gross pathologic findings in orally dosed birds included enlarged 2.3. Necropsy kidneys, hearts that had flaccid musculature, proventricular ulcerative lesions, intestinal edema, yellow bile, and large numbers of intestinal The oral dosing study was terminated at 21 days of exposure and the parasite numbers in the low dose group with no to few intestinal dermal oiling study was terminated on days 21 and 22 of exposure. All parasites in the high dose group. During necropsy of orally dosed birds, members of the high dose group of the oral study were euthanized by blood pooled in the cavities and did not clot after several minutes day 18 for humane reasons and so the day 21 sampling was not per- whereas blood did not pool in control birds. Gonad identification re- formed. Final blood samples were obtained (Cunningham et al., 2017), vealed 20 males and 6 females with immature gonads, divided as de- then DCCOs were euthanized by cervical dislocation. Organs were ex- scribed in the materials and methods. posed, photographed, then removed, weighed and preserved in 10% Inflammatory renal lesions were common in all groups; however,

53 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 52–61

Fig. 1. a) Effect of oral or dermal dosing with artificially weathered MC252 oil on kidney weight in double-crested cormorants (Phalacrocorax auritus). The lower and upper boundaries of the boxes indicate the 25th and 75th percentiles, respectively. The black line within the boxes is the median and the white line within the boxes is the mean value. The lower and upper whiskers indicate the 10th and 90th percentiles, respectively. Kidney weight was significantly greater in orally dosed birds (p = 0.0092) and dermally oiled birds (p < 0.0001) than in controls. b) Effect of oral or dermal dosing with artificially weathered MC252 oil on kidney weight relative to body weight in double-crested cormorants (Phalacrocorax auritus). The lower and upper boundaries of the boxes indicate the 25th and 75th percentiles, respectively. The black line within the boxes is the median and the white line within the boxes is the mean value. The lower and upper whiskers indicate the 10th and 90th percentiles, respectively. Kidney weight relative to body weight was significantly greater in orally dosed birds (p = 0.0017) and in dermally oiled birds (p = 0.0005) than in controls.

− − collecting duct and renal tubule hypertrophy, squamous metaplasia, dosed with 5 ml kg bw 1 day 1, although these results are anecdotal as and mineralization were seen more frequently in histopathology sam- parasite numbers were not quantified. Cytoplasmic vacuolation of ples from treated birds versus control birds, with the most severe and exocrine cells was the most common lesion seen in the pancreas, and frequent lesions found in the high dose group. Collecting duct and renal increased in the oil-dosed birds in comparison to the control birds. − − tubule lesions in birds dosed with 10 ml oil kg bw 1 day 1 and 5 ml oil Inflammatory liver lesions were mild, non-specific, and included − − kg bw 1 day 1 increased in frequency, distribution, and severity. mild lymphocyte predominant cholangitis, cholangiohepatitis, and he- Periductal inflammation was most commonly described and appeared patitis in all groups. In addition, in many of the birds, hemosiderin- to involve a major portion of the collecting duct and renal tubule within containing macrophages were present in the liver and spleen. Iron medullary cones. The type of inflammatory infiltrate varied but was staining of the livers revealed minimal hepatic iron staining (graded as primarily associated with lymphocytes and plasma cells, although 1) in most birds. Moderate staining (graded as 2) was noted in two of − − heterophils were occasionally seen. Renal coccidiosis was noted with the 5 ml oil kg bw 1 day 1 group birds and in one control bird. There similar frequency in the treated and control groups. was no observable difference in severity of iron staining of the liver Increased exocrine pancreatic cytoplasmic vacuolation was noted in between groups. The splenic lesions seen were most commonly in the − − both the low dose and high dose groups. Amyloidosis was noted in the 5 ml oil kg bw 1 day 1 birds. Some of the splenitis was obviously as- pancreas of one high dose bird. sociated with sepsis; however, in some birds there was a more chronic Enteric lesions occurred more commonly in birds orally dosed with inflammatory infiltrate of undetermined cause. − − 10 ml oil kg bw 1 day 1 compared with the other groups, but similar All other lesions identified in numerous organ systems were similar lesions were found in all groups. Involution of the bursa of Fabricius between the treatment groups. Almost all examined lungs were con- was noted in both treated and control groups. Gastrointestinal lesions gested and had airway hemorrhage, which was considered a necropsy across groups were common and involved primarily low-level para- artifact. One bird had lesions consistent with bacterial septicemia. One sitism with nematodes (proventriculus, ventriculus, cecum, and cloaca), bird in the control group had a focus of mineralization in the brain, trematodes (proventriculus), and cestodes (colon). There were few in- which is likely to have been an incidental finding. Additionally, one stances of variably severe inflammation of unknown etiology at all le- bird in each group had an esophagitis, which appeared to be a bacterial vels of the gastrointestinal tract. Gross necropsy reports noted that infection. enteric parasites seemed to be more prevalent in the intestines of birds

54 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 52–61

Fig. 2. a)Effect of oral or dermal dosing with artificially weathered MC252 oil on liver weight in double-crested cormorants (Phalacrocorax auritus). The lower and upper boundaries of the boxes indicate the 25th and 75th percentiles, respectively. The black line within the boxes is the median and the white line within the boxes is the mean value. The lower and upper − − whiskers indicate the 10th and 90th percentiles, respectively. Liver weight was significantly greater in birds orally dosed (p = 0.0012) with 5 ml oil kg body weight (bw) 1 day 1 and in dermally oiled birds (p < 0.0001) than in controls. b) Effect of oral or dermal dosing with artificially weathered MC252 oil on liver weight relative to body weight in double-crested cormorants (Phalacrocorax auritus). The lower and upper boundaries of the boxes indicate the 25th and 75th percentiles, respectively. The black line within the boxes is the median and the white line within the boxes is the mean value. The lower and upper whiskers indicate the 10th and 90th percentiles, respectively. Liver weight relative to body weight was − − significantly greater (p = 0.0006) in birds orally dosed with 5 ml oil kg body weight (bw) 1 day 1 compared to controls. Liver relative to body weight was significantly greater (p < 0.0001) in dermally oiled birds than in controls.

3.3.2. Dermal dosing study Both the control and oiled groups of DCCOs had intra-epithelial Gonad identification revealed 5 males and 7 females in the control nematodes recognized within the cloaca, although they were not group and 6 males and 7 females in the treated group. No developed quantified. In addition, both groups had equal numbers of birds that ovaries were found in any of the females at necropsy indicating a were supporting a moderate to severe nematode related lesion of the subadult classification. In the dermal dosing trial, there did not appear ventriculus with granulomas forming within the submucosa, as well as to be any difference in the number or severity of lesions between the nematode free lumens. The proventriculus was similar between the control and treated groups with the exception of the heart and the control and oil-dosed group in terms of supporting intraluminal ne- thyroid gland, which were somewhat increased in treated birds. matodes and variable amounts of inflammation. Renal inflammatory lesions were evenly distributed between both A reactive spleen, seen in approximately half of the control and groups as were the four of birds with intraepithelial coccidia. Lesions treated birds, was based on the identification of lymphoid follicles were characterized by a lymphoplasmacytic inflammatory infiltrate. identified within the splenic architecture. The liver lesions as well as One control bird had myocarditis and sequelae while four dermally the reactive spleen could easily be accounted for by the nonspecific oiled birds had cardiac lesions including myocardial fibrosis and bac- enteritis as well as the nematode parasitism recognized within the terial granuloma. The thyroid gland did seem to have some differences gastric sections. between the two treatment groups, with the dermally oiled birds being There were more severe lung lesions noted in the control group of more affected compared to controls. A diagnosis of hyperplastic goiter which at least one was a bronchopneumonia with granulomas. In one of (follicular hyperplasia) was based on the presence of multiple small the oil-dosed birds, there were multiple fungal granulomas as well as thyroid follicles, many of which were lined by cuboidal follicular epi- granulomatous inflammation in an air sac of a second oiled bird. thelium and were devoid of colloid. While the occurrence of this con- dition could depend on season and/or reproductive activity of the birds, 4. Discussion there were only two control birds with thyroid changes compared to eight in the dermally oiled group. The lesions are most suggestive of Many lesions, especially inflammatory lesions, recognized at gross goiter and could be due to a diet having either goitrogenic material and histologic examination were similar morphologically in both con- within it or low iodine levels. trol and treated birds and most likely represented background disease

55 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 52–61

Table 1 Summary of histological lesions found in tissues of double-crested cormorants (Phalacrocorax auritus) dosed orally or dermally with artificially weathered MC252 oil. Lesions were graded using the scale 1 = minimal, 2 = mild, 3 = moderate, 4 = severe.

Tissue/Treatment Lesion description Lesion distribution - Lesion grade - Mean lesion grade - Number of animals qualitative qualitative quantitative affected

Kidney Oral control lymphoplasmacytic interstitial nephritis multifocal minimal to 1.9 4/8 moderate periductal inflammation diffuse minimal to 1.9 4/8 moderate interstitial abscess formation multifocal minimal 1.0 1/8 tubular epithelial hypertrophy multifocal minimal to 2.0 1/8 moderate ureteral mucosal hyperplasia, uereteral inflammation, focal moderate 3.0 1/8 with intraluminal bacteria and inflammatory cells − 5 ml oil kg bw 1 lymphoplasmacytic interstitial nephritis multifocal to confluent moderate to severe 3.5 1/9 periductal inflammation multifocal to diffuse minimal to 1.8 3/9 moderate bacteremia multifocal to confluent moderate 3.0 1/9 ductal epithelial hypertrophy focal mild 2.0 1/9 fibrin thrombi/emboli multifocal moderate to severe 3.5 1/9 hypertrophy and squamous metaplasia multifocal severe 4.0 1/9 mineralization focal minimal to mild 1.3 2/9 peritubular inflammation multifocal minimal to mild 1.7 3/9 renal coccidiosis multifocal to confluent moderate 3.0 1/9 − 10 ml oil kg bw 1 periductal inflammation focal to multifocal mild to severe 3.0 2/9 Renal tubule epithelial hypertrophy focal to multifocal moderate to severe 3.5 2/9 ductal epithelial hypertrophy focal severe 4.0 1/9 mineralization focal to multifocal minimal to mild 1.3 4/9 squamous metaplasia focal to multifocal mild 2.0 5/9 tubular epithelial hypertrophy focal to multifocal mild to severe 3.0 3/9 Dermal control lymphoplasmacytic interstitial nephritis multifocal minimal to 2.1 8/11 moderate renal coccidiosis – 2/11 Treated lymphoplasmacytic interstitial nephritis focal - multifocal minimal to mild 1.8 8/11 renal coccidiosis focal mild 2.0 1/11

Heart Oral control chronic myodegeneration and fibrosis focal moderate 3.0 1/8 myocarditis focal minimal 1.0 2/8 − 5 ml oil kg bw 1 bacteremia multifocal mild 2.0 1/9 endocarditis multifocal to confluent mild 2.0 1/9 mineralization multifocal mild 2.0 1/9 myocarditis multifocal mild 2.0 1/9 myofiber degeneration/hemorrhage multifocal mild to moderate 2.5 1/9 − 10 ml oil kg bw 1 no lesions –––0/9 Dermal control epicardial hemorrhage focal minimal 1.0 1/11 myocardial hemorrhage multifocal mild 2.0 1/11 myocarditis, lymphoplasmacytic focal minimal 2.0 1/11 Treated bacterial granulomas multifocal moderate 3.0 1/11 myocardial fibrosis multifocal mild to moderate 2.3 3/11 septic and suppurative thrombi focal mild 2.0 1/11

Pancreas Oral control no lesions –––0/8 − 5 ml oil kg bw 1 cytoplasmic vacuolization focal to multifocal minimal to 1.4 4/9 moderate interstitial pancreatitis multifocal to confluent moderate 3.0 1/9 − 10 ml oil kg bw 1 amyloidosis multifocal to confluent moderate 3.0 1/9 cytoplasmic vacuolization multifocal to diffuse mild to moderate 3.0 3/9 zymogen depletion diffuse mild 2.0 1/9 Dermal control no lesions –––0/11 Treated no lesions –––0/11

Thyroid gland Oral control follicular hyperplasia multifocal moderate 3.0 1/8 colloid atrophy multifocal to confluent moderate 3.0 1/8 − 5 ml oil kg bw 1 follicular colloid atrophy diffuse mild 2.0 1/9 follicular hyperplasia diffuse mild 2.0 1/9 colloid atrophy diffuse moderate 3.0 1/9 − 10 ml oil kg bw 1 no lesions –––0/9 Dermal control follicular hyperplasia – mild 2.0 2/11 Treated follicular hyperplasia – minimal to 2.2 6/11 moderate

Esophagus Oral control no lesion –––0/8 (continued on next page)

56 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 52–61

Table 1 (continued)

Tissue/Treatment Lesion description Lesion distribution - Lesion grade - Mean lesion grade - Number of animals qualitative qualitative quantitative affected

− 5 ml oil kg bw 1 esophagitis multifocal mild to moderate 2.5 1/9 10 ml oil kg bw−1 no lesion –––0/9 Dermal control no lesion –––0/11 Treated no lesion –––0/11

Proventriculus Oral control nematodiasis focal minimal 1.0 2/8 proventriculitis focal to multifocal moderate to severe 3.3 2/8 serositis focal to multifocal mild to moderate 2.7 3/8 − 5 ml oil kg bw 1 nematodiasis focal to multifocal mild to severe 3.2 3/9 proventriculitis focal severe 4.0 2/9 serositis multifocal moderate to severe 3.3 2/9 glandular dilatation multifocal moderate to severe 3.8 2/9 mucosal hypertrophy/hyperplasia focal severe 4.0 1/9 trematodiasis focal mild to moderate 2.8 2/9 − 10 ml oil kg bw 1 nematodiasis focal mild to severe 3.5 4/9 proventriculitis focal to multifocal mild to severe 3.1 4/9 glandular dilatation focal severe 4.0 2/9 trematodiasis focal moderate 3.0 1/9 mucosal erosion multifocal to diffuse severe 4.0 2/9 myositis multifocal moderate to severe 3.5 1/9 Dermal control nematodiasis – moderate 3.0 3/11 proventriculitis diffuse to multifocal mild to moderate 2.7 3/11 interstitial granulomas multifocal mild 2.0 1/11 mucosal granulomas multifocal moderate 3.0 1/11 Treated nematodiasis focal minimal to mild 1.7 3/11 proventriculitis focal mild to moderate 2.5 2/11 serositis, lymphoplasmacytic – mild 2.0 1/11 gastritis focal – 2.0 1/11

Ventriculus Oral control nematodiasis multifocal moderate 3.0 2/8 ventriculitis multifocal to confluent mild to severe 3.3 6/8 − 5 ml oil kg bw 1 nematodiasis multifocal mild to severe 2.8 6/9 ventriculitis multifocal to confluent mild to severe 2.8 8/9 myositis multifocal to confluent severe 4.0 1/9 − 10 ml oil kg bw 1 ventriculitis multifocal moderate 3.0 1/9 serositis multifocal mild 2.0 1/9 Dermal control nematodiasis – moderate 3.0 2/11 ventriculitis diffuse severe 3.3 4/11 leiomyositis multifocal to diffuse mild to severe 3.0 4/11 lymphoplasmacytic leiomyositis multifocal mild 2.7 3/11 Treated nematodiasis moderate 3.2 5/11 ventriculitis multifocal severe 3.3 4/11 submucosal granulomas multifocal moderate 3.0 2/11 granulomatous leiomyositis multifocal severe 4.0 2/11

Cecum Oral control typhlitis diffuse moderate to severe 3.5 1/8 − 5 ml oil kg bw 1 nematodiasis multifocal moderate 3.0 1/9 typhlitis multifocal to confluent moderate to severe 3.5 2/9 − 10 ml oil kg bw 1 no lesions –––0/9

Small intestine Oral control enteritis multifocal minimal to mild 1.5 1/8 − 5 ml oil kg bw 1 no lesions –––0/9 − 10 ml oil kg bw 1 crypt necrosis multifocal minimal to mild 1.2 3/9 crypt necrosis/mineralization multifocal mild 2.0 1/9 enteritis multifocal to diffuse minimal to severe 2.8 2/9 villar atrophy/effusion diffuse mild 2.0 1/9 Dermal control enteritis, lymphoplasmacytic diffuse mild 2.0 2/11 enteritis, subacute diffuse mild to moderate 2.3 7/11 nematodiasis – mild 2.0 1/11 Treated enteritis, lymphoplasmacytic diffuse mild 2.0 2/11 enteritis, subacute diffuse mild to moderate 2.5 4/11

Colon Oral control cestodiasis multifocal mild to moderate 2.5 1/8 colitis diffuse moderate to severe 3.5 1/8 − 5 ml oil kg bw 1 cestodiasis focal mild 2.0 1/9 − 10 ml oil kg bw 1 crypt necrosis multifocal minimal 1.0 1/9

Cloaca/Bursa of Fabricius (continued on next page)

57 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 52–61

Table 1 (continued)

Tissue/Treatment Lesion description Lesion distribution - Lesion grade - Mean lesion grade - Number of animals qualitative qualitative quantitative affected

Oral control nematodiasis multifocal minimal to severe 2.8 6/8 bursal depletion diffuse minimal to severe 2.5 2/8 − 5 ml oil kg bw 1 nematodiasis multifocal to diffuse minimal to severe 2.4 7/9 bursal depletion diffuse severe 4.0 2/9 bursal involution diffuse moderate to severe 3.5 2/9 − 10 ml oil kg bw 1 nematodiasis multifocal minimal to 1.8 3/9 moderate bursal involution diffuse moderate to severe 3.8 2/9 Dermal control nematodiasis multifocal minimal to 2.1 11/11 moderate bursal depletion – mild to moderate 2.5 2/11 Treated nematodiasis multifocal minimal to 2.1 9/11 moderate no lesions –––0/11

Brain Oral control mineralization focal mild 2.0 1/8 − 5 ml oil kg bw 1 no lesions –––0/9 − 10 ml oil kg bw 1 no lesions –––0/9 Dermal control no lesions –––0/11 Treated hemorrhage multifocal moderate 3.0 1/11

Liver Oral control hemosiderin - containing macrophage accumulation multifocal minimal to mild 1.1 4/8 cholangiohepatitis multifocal minimal to mild 1.5 6/8 cholangitis multifocal minimal to mild 1.3 2/8 − 5 ml oil kg bw 1 hemosiderin - containing macrophage accumulation multifocal minimal to mild 1.1 5/9 cholangiohepatitis multifocal minimal to mild 1.5 7/9 cholangitis multifocal mild 2.0 1/9 embolization/thrombosis multifocal severe 4.0 1/9 hepatitis multifocal minimal to mild 1.5 1/9 macrophage accumulation multifocal minimal 1.0 1/9 − 10 ml oil kg bw 1 hemosiderin - containing macrophage accumulation multifocal minimal to mild 1.5 1/9 cholangitis multifocal minimal to mild 1.3 4/9 hepatitis multifocal mild 2.0 1/9 hepatocyte karyomegaly multifocal mild to moderate 2.5 1/9 vacuolar hepatopathy diffuse mild 2.0 1/9 Dermal control cholangiohepatitis – moderate 3.0 1/11 cholangitis multifocal mild to moderate 2.3 8/11 Treated cholangiohepatitis – moderate 3.0 1/11 cholangitis multifocal mild 2.0 4/11 hepatitis multifocal mild 2.0 3/11

Spleen Oral control hemosiderin - containing macrophage accumulation multifocal minimal to mild 1.5 1/8 − 5 ml oil kg bw 1 embolization/thrombosis multifocal severe 4.0 1/9 lymphoid depletion diffuse mild 2.0 1/9 serositis multifocal mild to moderate 2.5 1/9 splenitis diffuse mild to severe 2.8 3/9 vasculitis multifocal moderate 3.0 1/9 − 10 ml oil kg bw 1 no lesions –––0/9 Dermal control reactive spleen – mild to moderate 2.5 6/11 plasmacytic histiocytic splenitis – moderate 3.0 1/11 Treated reactive spleen – mild to moderate 2.0 5/11

Lung Oral control congestion diffuse mild to severe 2.9 8/8 hemorrhage multifocal moderate to severe 3.6 8/8 − 5 ml oil kg bw 1 congestion diffuse mild to severe 2.6 9/9 embolic pneumonia diffuse severe 4.0 1/9 hemorrhage multifocal mild to severe 3.5 8/9 − 10 ml oil kg bw 1 congestion diffuse mild 2.0 1/9 Dermal control nodular lymphoplasmacytic peribronchitis multifocal mild 2.0 2/11 Treated granulomatous pneumonia multifocal to moderate 3.0 1/11 coalescing in a wildlife population. While the presence of active disease in a (gastroenteritis, hepatitis, and various lesions in the kidneys) in both wildlife population accurately reflects field conditions, it makes dif- control and treatment groups. While there was a subjective evaluation ferentiation of subtle oil-induced damage impossible. Some of the that oil may impact the number of parasites, this was never quantified. confounding factors that may preclude oil-induced lesion evaluation include parasite loads in the birds (metazoans in the digestive tract and protozoans in the kidneys) that can account for some lesions

58 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 52–61

Table 2 Absolute organ weights in double crested cormorants (Phalacrocorax auritus) orally or dermally dosed with artificially weathered MC252 oil.

Dose Adrenals (g) SE Brain (g) SE Heart (g) SE Kidneys (g) SE Liver (g) SE Spleen (g) SE Thyroids (g) SE

Orally dosed cormorants Control 0.584 0.045 10.47 0.22 18.59A 0.90 17.75A 1.03 48.26A 2.90 0.87 0.08 0.184 0.034 n6 5 6 5 6 6 5 5 ml oil/kg BW 0.900 0.072 9.39 0.40 21.58AB 2.20 22.07B 0.76 75.47B 4.08 0.81 0.13 0.233 0.035 n6 6 6 6 6 5 6 10 ml oil/kg BW 0.700 0.163 9.79 0.27 18.90A 1.13 21.41B 0.74 62.65AB 4.92 0.58 0.17 0.149 0.014 n6 6 6 6 6 6 6 p-value 0.3032 0.1828 0.0381 0.0009 0.0002 0.2075 0.2330 Dermally dosed cormorants Control 0.664A 0.029 9.07 1.02 20.2 0.82 17.6A 0.58 45.3A 1.47 1.00 0.10 0.195 0.020 n11111111111111 Oiled 0.808B 0.047 7.59 0.48 21.4 1.37 21.6B 0.44 74.8B 3.91 0.80 0.05 0.185 0.027 n11111111111111 p-value 0.0221 0.2264 0.4902 < 0.0001 < 0.0001 0.0951 0.7793

ABWithin study columns means with different superscript letter are significantly different at stated p value.

− − 4.1. Oral dosing study 3), amyloidosis (10 ml oil kg bw 1 day 1 group = 1) and zymogen − − depletion (10 ml oil kg bw 1 day 1 group = 1). In avian species, cy- The increased collecting renal tubule epithelial hypertrophy, peri- toplasmic vacuolation have been reported in association with selenium ductular and peritubular inflammation, and squamous metaplasia in the and vitamin E deficiency, and zinc and cadmium toxicity. Zinc con- − − 10 ml oil kg bw 1 day 1, as well as renal mineralization that would be centration in MC252 weathered oil was 0.92 mg/kg and cadmium expected with an irritant toxicant, provides at least a partial explana- concentration was 0.05 mg/kg (Forth et al., 2017). Tissue concentra- tion for the increase in absolute and relative kidney weights. Squamous tions were unfortunately not analyzed. The authors hypothesize that metaplasia can be associated with chronic irritation and inflammation pancreatic changes were due to these compounds as postulated in the as well as vitamin A deficiency. The significant increase in absolute and literature but other components of the crude oil may have contributed relative kidney weights in the present oral dosing study supports renal to the cytoplasmic vacuolation in the pancreas. Zymogen depletion has dysfunction from structural changes, including edema, inflammation, been associated with zinc toxicity in waterfowl as well as degeneration, necrosis, or other damage. Statistically significant uratemia (Dean et al., necrosis or ductular hyperplasia (Sileo et al., 2003). However, ductular 2017, this issue) found in a dose response pattern in orally dosed birds change was not noted in the cormorant pancreatic tissues in this study. indicates that the tubular architectural changes had functional sig- Amyloidosis encompasses a collection of diseases characterized by the nificance and were a cause of injury. Dean et al. also documented ur- abnormal accumulation of amyloid proteins in tissues. In birds, ac- emia and hyperphosphatemia at levels indicating renal dysfunction. quired amyloidosis may occur subsequent to chronic inflammation or While no glomerular lesions were documented upon histopathology, it infection, enteric parasitism, aging, and stress (Landman et al., 1998). is likely that function decreased before architecture changed in this Because the birds in this study were wild caught, variations in extent of short 21 day study. The presence of inflammatory and proliferative infections, enteric parasitism, age and stress are likely. Therefore, dif- lesions localized to the renal tubules and ductules was consistent with ferences in these endpoints may be artifacts of life history rather than that previously reported in oil exposed birds and was consistent with a oil exposure. toxic insult. This architectural evidence of damage to the tubules and There were significant differences in absolute and relative liver collecting ducts could also contribute to the decrease in sodium and weight with minimal histopathologic changes which generally occurred chloride found in these birds, similar to the Western Sandpiper elec- in birds from all three groups; despite statistically significant changes in trolyte dyscrasia (Dean et al., 2017, this issue; Maggini et al., 2017, this hepatic biochemical endpoints including decreased AST, ALT, and GGT issue) as the ducts and tubules are the resorption site for these elec- activities (Dean et al., 2017). The increases in liver weight could in- trolytes. Western Sandpipers (Calidris mauri) also exposed to the same dicate edema, inflammatory infiltrate, hypertrophy/hyperplasia, or si- weathered MC 252 crude oil had evidence of interrenal cell hyper- milar potential changes because of oxidative damage (Duerr, 2013). In trophy consistent with that found in the literature (Bursian et al., 2017, the oral dosing study, when histopathologic lesions were compiled, this issue; Mazet, 2002). Cassin's auklets (Ptychoramphus aleuticus) ex- there was no microscopic explanation for the marked increase in gross posed to oil via dermal application and common murres (Uria aalge) weight of the liver normalized to body weight. Western Sandpipers that were recovered from an area affected by a spill of bunker C fuel oil (Calidris mauri) also exposed to the same weatherer MC 252 crude oil had renal tubular necrosis (Fry and Lowenstine, 1985), and mallard had increased absolute liver weight, no evidence of hemosiderosis in (Anas platyrhynchos) ducklings fed a diet containing 5.0% South treated birds, and significant evidence of hepatic oxidative damage Louisiana crude had tubular inflammation and degeneration in the (Bursian et al., 2017, this issue; Pritsos et al., 2017, this issue). kidney (Szaro et al., 1978). Exposure to oil and its metabolic products Increases in liver weight have been reported in prior avian oil ex- induces an immune and renal cellular response to damage in the kidney posure studies. Holmes et al. (1978) reported that adult Pekin ducks − while the bone marrow is intact. The inflammatory and cellular changes consuming approximately 6 ml day 1 of South Louisiana crude had in this study were consistent with those reported in the literature. increased relative liver weights compared to controls, but relative liver Gross necropsy findings of “flabby” or “flaccid” hearts suggested weights of ducks consuming 6 ml of Kuwait crude were comparable to cardiac damage. However, as expected, standard fixation procedures control weights. Herring gull chicks administered a single oral dose of − and hematoxylin and eosin staining were insufficient to determine the 0.3 ml kg bw 1 of Kuwait or South Louisiana crude oil had increased full extent of damage, resulting in false negative results. For further liver weights when necropsied nine days later (Miller et al., 1978). An discussion of cardiac dysfunction, please see Harr (2017b, this issue). increase in liver weight of herring gull chicks receiving five daily oral − − In this oral dosing study, pancreatic lesions were present only in oil- doses of 10 ml kg bw 1 day 1 of Prudhoe Bay crude oil was reported dosed birds and included cytoplasmic vacuolation of exocrine cells by Peakall et al. (1989), and mallard ducklings fed diets containing − − − − (5 ml oil kg bw 1 day 1 group = 5; 10 ml oil kg bw 1 day 1 group = 2.5% and 5.0% South Louisiana crude oil for eight weeks had

59 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 52–61 significant increases in liver weight (Szaro et al., 1978). Both the Miller fibrosis is discussed in Harr et al. (2017b) and highlights the need for et al. (1978) and Peakall et al. (1989) reported that the hepatic activity evaluation of collagen structure in the hearts of oil exposed birds. Un- of mixed function oxidase enzymes was significantly increased in the fortunately, no funding was allocated by USFWS for special stains for absence of hepatic pathology, suggesting that the increase in liver collagen at the time of this study. Additionally, as collagen changes and weight was a compensatory metabolic response. fibrosis represent very chronic changes, it is possible that the severity Szaro et al. (1978) reported that liver lesions in ducklings fed oil- and numbers of animals with histopathologic evidence of structural containing feed were subtle, consisting of generally minimal hyper- changes would increase in a longer duration study. trophy and vacuolation of hepatocytes and bile duct proliferation which More birds with thyroid gland follicular hyperplasia were present in is consistent with that found in this study. Leighton (1986) reported the dermally dosed group than in the control group. Hypertrophy and that the most predominant lesion in the livers of herring gull chicks and hyperplasia of the follicular epithelial cells may be in response to direct − Atlantic puffin nestlings dosed daily with 10 ml kg bw 1 of Prudhoe effects of crude oil constituents on the thyroid gland (BaJaJ et al., 2016) Bay crude oil consisted of enlarged Kupffer cells that were filled with or increased need for thyroid hormones. Feather damage reduces in- gold-brown pigment indicative of hemosiderin and phagocytized ery- sulation quality and increases metabolic heat production. Energy intake throcytes. Necrosis of individual hepatocytes and apoptosis were pre- by feather damaged birds is higher than normal feathered birds to valent in the gulls. Hepatic hemosiderosis in oil-exposed birds was also compensate for the increased metabolic heat production (Mazet et al., reported by Pattee and Franson (1982), Fry and Lowenstine (1985), and 2002). The thyroid gland is a critical organ for maintaining general Yamato et al. (1996). In the present oral dosing study, although anemia metabolic rate (Olson et al., 1999). Thyroid hormones stimulate general was present in most treated birds and packed cell volume (PCV) de- metabolism and new feather growth (Webster et al., 2016). Additional creased by 50% in six birds in the study (Harr et al., 2017b), hemosi- causes for the change in the thyroid gland including normal molting derosis was minimal. There was a mild increase in probable hemosi- and/or exposure to environmental goitrogenic substances should have − − derin in the 5 ml kg bw 1 day 1 group only. The intestinal erosion resulted in similar changes across both groups (BaJaJ et al., 2016). noted in the gastrointestinal tract was consistent with clinical ob- servations of abnormal feces. It is possible that iron was not accumu- 5. Conclusions lating in hepatocytes as much of the iron in RBC were lost due to ex- ternal hemorrhage in hematochezia or hematuria. Coagulopathy was Significant gross and/or histopathologic lesions were induced in the also documented in the cormorants using an activated clotting time kidney, heart, pancreas, thyroid, and possibly liver by either oral or analysis (Harr et al., 2017a) and so erosion did not have to be present dermal exposure of DCCO to weathered MC252 crude oil. The renal for blood loss to occur. lesions (collecting duct and renal tubule epithelial hypertrophy, peri- It should be noted that the oral high dose group consistently had ductular and peritubular inflammation, squamous metaplasia and renal smaller and fewer inflammatory lesions throughout the body than did mineralization) expected in the orally-dosed birds based on published either the control or low dose group (Table 1). This decrease in in- literature, were recognized in this study. These changes over time could flammatory cells and lesions was evidenced in the kidney, liver, heart, significantly impair the kidneys’ ability to excrete waste and maintain pancreas, and throughout the GI tract. Oil-induced oxidative damage of electrolyte and solute homeostasis as evidenced by uremia and ur- proliferative hemoprogenitors in the bone marrow has long been atemia, thereby causing morbidity or mortality. The significance of the documented in mammals and more recently decreased cell mediated cardiac changes found upon gross examination of the orally and der- immune response has been documented in birds (Olsgard et al., 2008). mally exposed birds was confirmed by antemortem functional testing. This finding indicates that immunocompromise when fighting disease Minimal histologic change was detected in the heart indicating that expected in wild populations may contribute to morbidity and mortality standard histologic techniques may not be an appropriate diagnostic in oil intoxicated birds. In addition, a decrease in RBC hemoprogenitors tool for determining oil-induced cardiac damage. Cytoplasmic vacuo- in the bone marrow would also contribute to anemia including the lation of pancreatic exocrine cells and zymogen depletion increased in severe anemia found in the high dose group in this study (Harr, 2017a, the orally exposed birds in comparison to the control birds supporting this issue). Indeed, the moderate to severe anemia in these birds was toxic exposure. The functional significance of the pancreatic lesions and found to be poorly compensated when reticulocytes and rubricytes potential to impact health and induce mortality were not explored in were evaluated adding further evidence that oil-induced bone marrow this study. In the dermally exposed birds, an increased number of birds damage contributes to not only immunocompromise but decreased had thyroid gland lesions which may be directly induced by compo- oxygen transport due to anemia. Unfortunately, bone marrow was not nents of weathered MC252 crude oil or secondary to feather damage evaluated in either study. Decreased inflammatory response and de- and increased need for metabolic heat production. While the histologic creased hemoglobin was also found in Western Sandpipers (Calidris changes in the liver were minimal and likely represent a false negative, mauri) exposed to the same weathered MC252 crude oil (Bursian et al., the change in organ weight combined with biochemical changes (Dean 2017) indicating that this response is consistent across bird species et al., 2017; Pritsos et al., 2017) indicate that the liver was also da- tested. maged. Results of the present study suggest that oral or dermal ex- posure of DCCOs to artificially weathered MC252 crude oil induced 4.2. Dermal dosing study organ damage that could adversely affect survivability.

Dermal dosing primarily resulted in feather damage and presumably Acknowledgements less internal oil exposure compared to the orally dosed birds (Cunningham et al., 2017; Dean et al., 2017, this issue). There were two The studies appearing in this special issue were funded by the U.S. organs which had histologic differences which should be further ex- Fish and Wildlife Service (Order No. F12PD01069) as part of the plored: the heart and the thyroid gland. Deepwater Horizon Natural Resource Damage Assessment. Thanks to While three exposed birds were found to have myocardial fibrosis, Dr. Jennifer Brazzell for quality assurance of the histopathology and Dr. this was documented in only one control bird with myocarditis. We Allan Pessier for critical review of this manuscript. postulate that fibrosis (scar tissue) was beginning to form as a healing response to oxidative damage documented in the bird's liver and RBC References (Pritsos et al., 2017; Harr et al., 2017a, 2017b). Cardiomyocytes have a high concentration of mitochondria due to their high metabolic state BaJaJ, J.K., Salwan, P., Salwan, S., 2016. Various possible toxicants involved in thyroid and are sensitive to oxidative damage.(Giordano, 2005) Myocardial dysfunction: a review. J. Clin. Diagn. Res.: JCDR 10 (1), FE01.

60 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 52–61

Bursian, S.J., Dean, K.M., Harr, K.E., Kennedy, L., Link, J.E., Maggini, I., Pritsos, C., ducks (Anas platyrhynchos). J. Reprod. Fertil. 54, 335–347. Pritsos, K.L., Schmidt, R.E., Guglielmo, C.G., 2017. Effect of oral exposure to artifi- Khan, R.A., Nag, K., 1993. Estimation of hemosiderosis in seabirds and fish exposed to cially weathered Deepwater Horizon crude oil on blood chemistries, hepatic anti- petroleum. Bull. Environ. Contam. Tox. 50, 125–131. oxidant enzyme activities, organ weights and histopathology in western sandpipers Landman, W.J.M., Gruys, E., Gielkens, A.L.J., 1998. Avian amyloidosis. Avian Pathol. 27, (Calidris mauri). Ecotoxicol. Environ. Saf (this issue). 437–449. Cunningham, F.L., Dean, K.M., Hanson-Dorr, K.C., Harr, K.E., Healy, K., Horak, K.E., Link, Leighton, F.A., 1986. Clinical, gross, and histological findings in herring gulls and Atlantic J.E., Shriner, S., Rupp, T.L., Bursian, S.J., Dorr, B.S., 2017. Development of methods puffins that ingested Prudhoe Bay crude oil. Vet. Pathol. 23, 254–263. for avian oil toxicity studies using the double crested cormorant (Phalacrocorax Maggini, I., Kennedy, L.V., Bursian, S.J., Dean, K.M., Gerson, A.R., Harr, K.E., Link, J.E., auritus). Ecotoxicol. Environ. Saf (this issue). Pritsos, C.A., Pritsos, K.L., Guglielmo, C.G., 2017. Toxicological and thermo- Dean, K.M., Bursian, S.J., Cacela, D., Carney, M.W., Cunningham, F.L., Dorr, B., Hanson- regulatory effects of feather contamination with artificially weathered MC 252 oil in Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Lipton, I., McFadden, A.K., McKernan, western sandpipers (Calidris mauri). Ecotoxicol. Environ. Saf (this issue). M.A., Harr, K.E., 2017. Changes in white cell estimates and plasma chemistry mea- Mazet, J.A.K., Newman, S.H., Gilardi, K.V.K., Tseng, F.S., Holcomb, J.B., Jessup, D.A., surements following oral or dermal dosing of double crested cormorants, Ziccardi, M.H., 2002. Advances in oiled bird emergency medicine and management. Phalacocorax auritus, with artificially weathered MC252 oil. Ecotoxicol. Environ. Saf J. Avian Med. Surg. 16, 146–149. (this issue). Miller, D.S., Peakall, D.B., Kinter, W.B., 1978. Ingestion of crude oil: sublethal effects in Deepwater Horizon Natural Resource Damage Assessment Trustees, 2016. Deepwater herring gull chicks. Science 199, 315–317. Horizon Oil Spill: Final Programmatic Damage Assessment and Restoration Plan and Olson, J.M., McNabb, F.M., Jablonski, M.S., Ferris, D.V., 1999. Thyroid development in Final Programmatic Environmental Impact Statement. Retrieved from 〈http://www. relation to the development of endothermy in the red-winged blackbird (Agelaius gulfspillrestoration.noaa.gov/restoration-planning/gulf-plan/〉. phoeniceus). Gen. Comp. Endocrinol. 116, 204–212. Duerr, R.S., 2013. Investigation into Nutritional Condition and Digestive Capabilities of Olsgard, M.L., Bortolotti, G.R., Trask, B.R., Smits, J.E., 2008. Effects of inhalation ex- Seabirds during Rehabilitation in California (Ph.D. Dissertation). University of posure to a binary mixture of benzene and toluene on vitamin A status and humoral California, Davis. and cell-mediated immunity in wild and captive American kestrels. J. Toxicol. Forth, H.P., Mitchelmore, C.L., Morris, J.M., Lay, C.R., Lipton, J., 2017. Characterization Environ. Health A 71 (16), 1100–1108. of dissolved and particulate phases of water accommodated fractions used to conduct Pattee, O.H., Franson, J.C., 1982. Short-term effects of oil ingestion on American kestrels aquatic toxicity testing in support of the Deepwater Horizon natural resource damage (Falco sparverius). J. Wildl. Dis. 18, 235–241. assessment. Environ. Toxicol. Chem. 36 (6), 1460–1472. Peakall, D.B., Norstrom, R.J., Jeffrey, D.A., Leighton, F.A., 1989. Induction of hepatic Fry, D.M., Lowenstine, L.J., 1985. Pathology of common murres and Cassin's auklets mixed function oxidases in the herring gull (Larus aregentatus) by Prudhoe Bay crude exposed to oil. Arch. Environ. Contam. Toxicol. 14, 725–737. oil and its fractions. Comp. Biochem. Physiol. 94C, 461–463. Giordano, F.J., 2005. Oxygen, oxidative stress, hypoxia, and heart failure. J. Clin. Pritsos, K.L., Perez, C.R., Muthumalage, T., Dean, K.M., Cacela, D., Hanson-Dorr, K.C., Investig. 115 (3), 500–508. Cunningham, F.L., Bursian, S.J., Link, J.E., Shriner, S., Horak, K.E., Pritsos, C.A., Harr, K.E., Cunningham, F.L., Pritsos, C.A., Pritsos, K.L., Muthumalage, T., Dorr, B.S., 2017. Dietary intake of deepwater horizon oil-injected live food fish by double- Horak, K.E., Hanson-Dorr, K.C., Dean, K.M., Cacela, D., Link, J.E., Healy, K., Tuttle, crested cormorants resulted in oxidative stress. Ecotoxicol. Environ. Saf (this issue). P., Bursian, S.J., 2017a. Weathered MC252 crude oil-induced anemia and abnormal Sileo, L., Beyer, W.N., Mateo, R., 2003. Pancreatitis in wild zinc-poisoned waterfowl. erythroid morphology in double-crested cormorants (Phalacrocorax auritus) with light Avian Pathol. 32, 655–660. microscopic and ultrastructural description of Heinz bodies. Ecotoxicol. Environ. Saf Szaro, R.C., Albers, P.H., Coon, N.C., 1978. Petroleum: effects on mallard egg hatch- (this issue). ability. J. Wildl. Manag. 42, 404–406. Harr, K.E., Rishniw, M., Rupp, T.L., Cacela, D., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C., Webster, R.K., Aguilar, R.F., Argandona-Gonzalez, A.-K., Conayne, P., De Sousa, D., Healy, K., Horak, K., Link, J.E., Reavill, D., Bursian, S.J., Cunningham, F.L., 2017b. Sriram, A., Svensson, C.M., Gartrell, B.D., 2016. Forced molt in four juvenile yellow- Dermal exposure to weathered MC252 crude oil results in echocardigraphically eyed penguins (Megadyptes antipodes). J. Wildl. Dis. 52 (4), 809–816. http://dx.doi. identifiable systolic myocardial dysfunction in double crested cormorants org/10.7589/2015-11-305. (Phalacrocorax auritus). Ecotoxicol. Environ. Saf (in press). Yamato, O., Goto, I., Maede, Y., 1996. Hemolytic anemia in wild seaducks caused by Holmes, W.N., Cavanaugh, K.P., Cronshaw, J., 1978. The effects of ingested petroleum on marine oil pollution. J. Wildl. Dis. 32, 381–384. oviposition and some aspects of reproduction in experimental colonies of mallard

61 Ecotoxicology and Environmental Safety 146 (2017) 62–67

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Dietary intake of Deepwater Horizon oil-injected live food fish by double- MARK crested cormorants resulted in oxidative stress

Karen L. Pritsosa, Cristina R. Pereza, Thivanka Muthumalagea, Karen M. Deanb, Dave Cacelab, Katie Hanson-Dorrc, Fred Cunninghamc, Steven J. Bursiand, Jane E. Linkd, Susan Shrinere, ⁎ Katherine Horake, Chris A. Pritsosa, a Department of Agriculture, Nutrition, and Veterinary Sciences, University of Nevada, Reno, United States b Abt Associates, United States c US Department of Agriculture, APHIS/Wildlife Services’ National Wildlife Research Center, MS, United States d Department of Animal Science, Michigan State University, East Lansing, MI, United States e US Department of Agriculture, APHIS/Wildlife Services, National Wildlife Research Center, Fort Collins, CO, United States

ARTICLE INFO ABSTRACT

Keywords: The Deepwater Horizon oil spill released 134 million gallons of crude oil into the Gulf of Mexico making it the Double-crested cormorants largest oil spill in US history and exposing fish, birds, and marine mammals throughout the Gulf of Mexico to its Oil spill toxicity. Fish eating waterbirds such as the double-crested cormorant (Phalacrocorax auritus) were exposed to the Deepwater Horizon oil both by direct contact with the oil and orally through preening and the ingestion of contaminated fish. This Oxidative stress study investigated the effects of orally ingestedMC252 oil-contaminated live fish food by double-crested cor- Glutathione morants on oxidative stress. Total, reduced, and oxidized glutathione levels, superoxide dismutase and glu- Antioxidant enzymes tathione peroxidase activities, total antioxidant capacity and lipid peroxidation were assessed in the liver tissues of control and treated cormorants. The results suggest that ingestion of the oil-contaminated fish resulted in significant increase in oxidative stress in the liver tissues of these birds. The oil-induced increase in oxidative stress could have detrimental impacts on the bird's life-history.

1. Introduction crude oil. It occurs when the balance between pro-oxidants and anti- oxidants is perturbed in favor of the former. Oxidative stress is mani- The Deepwater Horizon (DWH) oil spill in 2010 released more than fested as oxidative damage and leads to the oxidation of key biological 3 million barrels of South Louisiana Sweet Crude oil which was widely constituents such as proteins, lipids, and DNA. PAH compounds are dispersed by ocean currents, exposing countless organisms as the oil metabolized by cytochrome P-450 oxidases (Nebert and Dalton, 2006). dispersed throughout the Gulf of Mexico (Chang, 2011). This oil spill Cytochrome P4501A has been shown to be able to specifically meta- resulted in a large number of dead birds as well as a large number of bolize some of the PAHs found in oil (Sarasquete and Segner, 2000). live birds that were exposed to sublethal amounts of oil (USFWS, 2011). The metabolism of PAHs can produce toxic secondary metabolites that Crude oils, including South Louisiana Sweet crude oil, contain toxic can react with molecular oxygen to generate many reactive oxygen compounds that are readily accumulated by organisms, such as poly- species (ROS) in a redox cycling reaction. This redox cycling is the basis cyclic aromatic hydrocarbons (PAHs) (Leighton, 1993). PAH metabo- for chemical-induced oxidative damage. Thus, organisms that ingest lites are known toxicants that have been found to cause a wide range of these chemicals would be subject to potential oxidative damage. Oxi- adverse effects to oiled birds including liver and kidney damage, im- dative stress has been implicated in the etiology of a great number of munosuppression, suppressed growth, reduced hormone function, gas- diseases and conditions (Limón-Pacheco and Gonsebatt, 2009). trointestinal irritation, failed reproduction, behavioral changes, and One of the most commonly reported clinical indications of oxidative hemolytic anemia (Leighton, 1993; Briggs et al., 1997; Hartung and damage in birds is hemolytic anemia, which is damage to erythrocytes Hunt, 1966). mediated by oxidative PAH metabolites (Troisi et al., 2007). These Oxidative stress is one of the fundamental toxic mechanisms that metabolites oxidize hemoglobin resulting in hemichrome formation and can occur from exposure to xenobiotics, such as the PAHs found in hemoglobin precipitation to form dense granular bodies called Heinz

⁎ Corresponding author. E-mail address: [email protected] (C.A. Pritsos). http://dx.doi.org/10.1016/j.ecoenv.2017.06.067 Received 12 August 2016; Received in revised form 22 June 2017; Accepted 27 June 2017 Available online 05 July 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. K.L. Pritsos et al. Ecotoxicology and Environmental Safety 146 (2017) 62–67 bodies (Harr et al., in this issue). Heinz body hemolytic anemia has 2. Methods been reported in several species of oiled birds (Leighton et al., 1983, 1985; Leighton, 1986; Yamato et al., 1996; Troisi et al., 2007). In ad- 2.1. Animal collection and husbandry dition, field studies during the DWH oil spill assessment found in- creased percentages of Heinz bodies, reticulocytosis and anemia in live Double-crested cormorants (Phalacrocorax auritus; DCCO) were oiled birds captured and assessed in the Gulf of Mexico (Fallon et al., chosen to test forMC252-induced oxidative stress as part of the 2014). These findings thereby suggest that exposure to oil during the Deepwater Horizon Natural Resource Damage Assessment avian toxi- DWH oil spill resulted in oxidative stress that resulted in hemoglobin city studies because they were affected by the DWH spill. All animal damage and led to the development of hemolytic anemia in these birds. procedures were approved by the National Wildlife Research Center Anemia is a particularly important oxidative outcome because it causes (NWRC) Institutional Animal Care and Use Committee (IACUC, ap- reduced availability of oxygen to tissues (Fallon et al., 2013). Birds proval #2107). The specific details of the capture, transport quarantine suffering from hemolytic anemia are often fatigued and have reduced feeding and maintenance of these birds are provided in this issue energy available for metabolic processes, which could have profound (Cunningham et al., 2017). impacts on exercising birds that rely on high energy to sustain energy intensive flights. 2.2. Oral dosing study Biomarkers of oxidative stress have been used as valuable indicators of environmental change and the physiological responses to that Cormorants were randomly assigned to one of three treatment change. In addition, they may also be used as indicators of individual groups: a control group (n = 8); a group dosed daily with up to 5 ml health, because they provide quantification of tissue damage oil/kg bw/day (n = 9); a group dosed daily with up to 10 ml oil/kg bw/ (Constantini and Dell’Omo, 2015). Oxidative stress biomarkers may be day (n = 9). Fingerling catfish were given an intraperitoneal injection particularly useful as a tool to assess the health status of individuals of 2.0 ml of artificially weatheredMC252 (DWH7937, batch# B030112) after an environmental pollution event, such as the DWH oil spill. An- oil using a 20-gauge needle on a 25 ml stainless steel/glass barrel ® imals have multiple mechanisms of defense against oxidative damage, Hauptner syringe as described in this issue (Cunningham et al., 2017). including antioxidants and catalytic enzymes. The concentration of The oil-injected catfish were subsequently fed to cormorants in the oil antioxidants and antioxidant enzymes in the body are indicators of the treatment groups in their water-filled foraging tanks at a dosage of ei- amount of oxidative stress in the body (Koivula and Eeva, 2010). In a ther 5 or 10 ml oil/kg body weight (bw) as described in this issue polluted environment, antioxidant systems can be induced as an (Cunningham et al., 2017). The birds were necropsied on the last day of adaptive response to allow organisms to combat oxidative stress the study (day 21). Birds were weighed and euthanized by cervical (Marasco and Constantini, 2016). In contrast, the antioxidant system dislocation according to IACUC-approved protocols. Liver tissues were may also be inhibited, allowing for reduced protection and a greater extracted, cut up for each individual analysis, placed in separate susceptibility of cell damage. The activities (induction or inhibition) of cryovials, flash frozen with liquid nitrogen and shipped on dry ice to the antioxidant defense components have been used as biomarkers of oil Pritsos laboratory at the University of Nevada, Reno for oxidative stress exposure (Cossu et al., 1997; Cheung et al., 2000; Doyotte et al., 1997). marker analyses. Upon arrival at the lab in Reno, samples were re- Multiple biomarkers are commonly needed to reliably assess oxidative corded and stored into a −70 °C freezer until analyzed. Although blood stress, including both enzymatic and non-enzymatic antioxidants. Su- and liver tissue were both collected during this study, liver tissue was peroxide dismutase (SOD) and glutathione peroxidase (GPx) are im- chosen as the matrix to assess oxidative stress due to greater confidence portant antioxidant enzymes responsible for the removal of free radicals in sample viability of the liver tissues and limited quantities of blood and reactive species (Limón-Pacheco and Gonsebatt, 2009). Together tissue for analyses. with the antioxidant enzymes, antioxidant compounds that quench free radicals are also used as endpoints of oxidative stress, specifically glu- 2.3. Oxidative stress sample preparation and analyses tathione. Additionally, levels of lipid peroxidation products are com- monly used as a measure of oxidative cellular damage (Peréz et al., Oxidative damage in liver was assessed on liver homogenates pre- 2010). However, to date, this multiple oxidative stress marker assess- pared from the individual liver subsamples described above. For total, ment strategy has not been employed in migratory bird studies. oxidized and reduced glutathione (TGSH, GSSG, and RGSH, respec- The objective of this study was to determine if double-crested cor- tively) liver tissue was homogenized in ice-cold 50 mM 2-(N morpho- morants (Phalacrocorax auritis) fed fish injected with artificially lino) ethanesulphonic acid (MES), 1 mM EDTA buffer (pH 6)/g of weathered MC252 oil would develop signs of oxidative stress, con- tissue. The homogenate was centrifuged at 10,000 × g for 15 min at sistent with the red blood cell oxidative stress observed in the wild, as 4 °C and the supernatant was collected and kept on ice. The supernatant reported by Fallon et al. (2014). Double-crested cormorants are easily was deproteinated by addition of an equal volume of 0.1% metapho- managed in captivity, and were chosen as a model species that could be sphoric acid and then vortexed. After allowing the mixture to stand at used as a surrogate species for other piscivorous waterbirds that were room temperature for 5 min, it was centrifuged at 5000 × g for 5 min at impacted by the DWH oil spill. To assess oxidative stress in this study, room temperature. The supernatant was collected and used in the assay we used the multiple biomarker approach. We measured the effect of for TGSH and GSSG (kit #70300, Cayman Chemical). Reduced glu- sublethal oil exposure on the levels of hepatic enzymatic antioxidants tathione was calculated as the difference between TGSH and GSSG. To by analysis of superoxide dismutase and glutathione peroxidase activ- determine glutathione peroxidase activity, liver tissue was homo- ities. We also measured the levels of hepatic antioxidants by analysis of genized in ice-cold 50 mM Tris-HCl, 5 mM EDTA, 1 mM DTT buffer (pH total glutathione, reduced glutathione (GSH), oxidized glutathione 7.5)/g tissue. The homogenate was centrifuged at 10,000 × g for (GSSG), and total antioxidant capacity. In addition, we measured levels 15 min at 4 °C and the supernatant was collected and kept on ice. The of hepatic lipid peroxidation products as a marker of oxidative damage assay was performed on appropriately diluted supernatant with GPx in the liver. We predicted that oil-exposed birds would exhibit elevated assay kit #703102 (Cayman Chemical Co.). To determine total anti- oxidative stress in a dose-dependent manner as indicated by multiple oxidant capacity (Trolox), liver tissue was homogenized in 5–10 ml of oxidative stress markers. ice-cold phosphate buffered saline (PBS)/g tissue. The homogenate was centrifuged at 3000 × g for 12 min at 4 °C and the supernatant was collected for the assay. Trolox was determined using assay kit #TA02 (Oxford Biomedical Research) according to manufacturer's directions. For lipid peroxidation (LPO) assessment, 1 g of liver tissue was

63 K.L. Pritsos et al. Ecotoxicology and Environmental Safety 146 (2017) 62–67 homogenized in 5–10 ml of ice-cold 20 mM Tris buffer, pH 7.4, con- taining 5 mM BHT (to prevent sample oxidation). The homogenate was centrifuged at 3000 × g for 10 min at 4 °C and the resulting supernatant was diluted appropriately and used in assay kit #FR22 (Oxford Biomedical Research) according to manufacturer's directions. Liver peroxidation was chosen as a measure of oxidative damage over other endpoints, such as DNA damage and protein oxidation because of its more accessible assay methods and widespread use in environmental toxicology allowing for more direct comparisons between studies.

2.4. Statistical methods

For each oxidative stress parameter, a mean value per treatment group was calculated. Mean values of the two oil treated groups were contrasted with the mean values of the control group for all parameters using Dunnett's test. Data transformation was deemed unnecessary after visual inspection of data distributions. Calculations were performed using TIBCO Spotfire S-PLUS 8.2 for Windows. A p-value of less than 0.05 was interpreted as statistically significant.

3. Results

Double-Crested Cormorants were treated with artificially weathered MC252 crude oil internally through their food as previously described. Liver tissues were assessed for several markers of oxidative stress in- cluding; total glutathione, reduced glutathione, oxidized glutathione, glutathione peroxidase, superoxide dismutase, total antioxidant capa- city and lipid peroxidation. The effects of oral exposure to MC252 crude oil in double-crested cormorants on hepatic total glutathione, reduced glutathione and oxidized glutathione are illustrated in Fig. 1a, b and c. Total glutathione levels increase in a dose-dependent manner from 26 nmol GSH/mg protein in the controls to 79.9 nmol GSH/mg protein in the highest dosed group. Similarly, reduced glutathione levels in- creased dose-dependently from 24 nmol GSH/mg protein in controls to 76.8 nmol GSH/mg protein in the highest dosed group. Hepatic oxi- dized glutathione levels were also increased in the groups receiving oil, however the increases were not dose-dependent. Overall, hepatic glu- tathione levels in birds exposed to oil orally were greatly increased. Superoxide dismutase and glutathione peroxidase are considered front line enzymatic defenses against oxidative stress in biological or- ganisms. Hepatic superoxide dismutase activity decreased in both groups receiving oil treatment compared to the control group (Fig. 2). Glutathione peroxidase activity in hepatic tissues of orally treated birds significantly decreased compared to control birds (Fig. 3). During an oxidative insult due to a xenobiotic exposure, some an- tioxidants may be induced while others may be inhibited. Total anti- oxidant capacity is a measure of the total amount of antioxidant pro- tection, both enzymatic and non-enzymatic, available to a tissue. The total antioxidant capacity of hepatic tissues in these birds increased with increasing exposure to oil (Fig. 4). Statistical significance was Fig. 1. a. Activity levels of hepatic total glutathione in two groups of double-crested observed between the highest dosage group and controls. cormorants orally exposed to artificially weathered MC252 crude oil compared to control Lipid peroxidation is a measure of oxidative damage that results birds. The lower and upper boundaries of the boxes indicate the 25th and 75th percen- when pro-oxidation events overwhelm the antioxidant defenses of an tiles, respectively. The black line within the box is the median and the dots are individual values. * p < 0.05 vs control, Dunnett's test. b. Activity levels of hepatic reduced glu- organism. Polyunsaturated fatty acid decomposition generates both tathione in two groups of double-crested cormorants orally exposed to artificially malondialdehyde and 4-hydroxyalkenals. The measurement of both of weathered MC252 crude oil compared to control birds. The lower and upper boundaries these breakdown products is considered an excellent assessment of the boxes indicate the 25th and 75th percentiles, respectively. The black line within the method for lipid peroxidation. Lipid peroxidation was determined in box is the median and the dots are individual values. * p < 0.05 vs control, Dunnett's test. these studies by assaying for malondialdehyde +4-hydroxyalkenals. No c. Activity levels of hepatic glutathione disulfide in two dosing groups of double-crested fi statistically significant difference was observed between control and cormorants orally exposed to arti cially weathered crude oil compared to control birds. The lower and upper boundaries of the boxes indicate the 25th and 75th percentiles, treatment groups (Fig. 5). respectively. The black line within the box is the median and the dots are individual values. * p < 0.05 vs control, Dunnett's test. 4. Discussion hepatic glutathione disulfide, reduced glutathione, and total glu- fi In this study, we con rmed that oral exposure of double-crested tathione concentrations, as well as decreases in hepatic superoxide cormorants to weathered DWH MC252 crude oil-induced hepatic oxi- dismutase and glutathione peroxidase activities compared to control dative stress. Birds ingesting oil injected fish exhibited increases in

64 K.L. Pritsos et al. Ecotoxicology and Environmental Safety 146 (2017) 62–67

Fig. 2. Activity levels of hepatic superoxide dismutase in two dosing groups of double- Fig. 5. Levels of lipid peroxidation in two dosing groups of double-crested cormorants fi crested cormorants orally exposed to arti cially weathered crude oil compared to control orally exposed to artificially weathered crude oil compared to control birds. The lower birds. The lower and upper boundaries of the boxes indicate the 25th and 75th percen- and upper boundaries of the boxes indicate the 25th and 75th percentiles, respectively. tiles, respectively. The black line within the box is the median and the dots are individual The black line within the box is the median and the dots are individual values. values. * p < 0.05 vs control, Dunnett's test. oxidative conditions in oil exposed birds. Superoxide dismutase and glutathione peroxidase are important antioxidant enzymes that are considered part of the first line of defense against oxidative stress. These enzymes work to catalyze reactions that reduce toxic compounds to nontoxic metabolites. The decreased levels of hepatic SOD and GPx activity in this study likely reflect inhibition by organic electrophilic chemical constituents in oil (Staimer et al., 2012). Inhibition of both of these front line antioxidant enzymes would increase oxidative stress in the oil-exposed double-crested cormorants. The double-crested cormorants in this study showed evidence of increased hepatic tissue oxidative stress, however, it appears that the birds were able to compensate for this stress primarily by increasing hepatic glutathione levels and thus did not exhibit significantly in- creased lipid peroxidation levels. Glutathione is the most abundant cellular antioxidant (Limón-Pacheco and Gonsebatt, 2009) and is a primary detoxification pathway for xenobiotics in the liver. Induction of de novo synthesis of GSH occurs as an adaptive response to oxidative Fig. 3. Activity levels of hepatic glutathione peroxidase in two dosing groups of double- crested cormorants orally exposed to artificially weathered crude oil compared to control stress (Biswas and Rahman, 2009). Increased levels of ROS result in an birds. The lower and upper boundaries of the boxes indicate the 25th and 75th percen- initial decrease in GSH levels, which can elicit activation of transcrip- tiles, respectively. The black line within the box is the median and the dots are individual tion factors and result in increased production of GSH in the tissue (Itoh values. * p < 0.05 vs control, Dunnett's test. et al., 1997). GSH becomes oxidized during the reduction of peroxides via glutathione peroxidase. Thus, increased amounts of the oxidized form of glutathione, GSSG, are an indicator of increased oxidative stress, as GSSG normally only represents less than 1% of total cellular GSH (Biswas and Rahman, 2009). GSSG is recycled via glutathione reductase to regenerate GSH. Thus, increased concentrations of hepatic GSH are also consistent with oxidative stress. In this study, we found both total and reduced glutathione levels to be dose-dependent, with the highest levels occurring in the highest dosed group. This indicates that the highest dosed birds suffered greater oxidative stress pre- sumably due to the higher concentration of crude oil constituents and consequently needed greater protection of cells from oxidative damage. The increase in total antioxidant capacity observed in the oil-exposed birds is a reflection of the adaptive increase in glutathione observed in the oil groups. The failure to observe a significant increase in lipid peroxidation in these birds is very likely due to this robust antioxidant response. The DWH NRDA toxicity testing program used four different oil Fig. 4. Levels of total antioxidant capacity in two dosing groups of double-crested cor- samples of varying degrees of weathering in all of the studies con- morants orally exposed to artificially weathered crude oil compared to control birds. The fi lower and upper boundaries of the boxes indicate the 25th and 75th percentiles, re- ducted. The preparation and chemical composition of the DWH arti - spectively. The black line within the box is the median and the dots are individual values. cially weathered MC252 oil used in this study are described in Forth * p < 0.05 vs control, Dunnett's test. et al. (2016). Weathering of the oil resulted in the loss of the volatile, lighter weight PAH's, including BTEX compounds. Thus, the resulting birds. The metabolism of PAHs found in crude oil by cytochrome P450 oil had a compositional shift to the heavier compounds in the oil. enzymes produce oxides and reactive oxygen species that create Compared to the unweathered source oil, the percentage of PAH

65 K.L. Pritsos et al. Ecotoxicology and Environmental Safety 146 (2017) 62–67 depletion in the artificially-weathered source was 27% (27% weath- addition, it has been well established that oiled birds have difficulty ered). The toxicity of oil samples vary depending on chemical compo- maintaining thermoregulatory processes. Feathers contaminated with sition. The increased concentration of heavy PAH compounds in the oil crude oil lose their insulating and water-repellant properties, resulting sample used in this study, likely contributed to the toxicity observed in in body heat loss and an increase in metabolic rate in attempts to the treated cormorants. regulate body temperature, exhausting energy stores (Leighton, 1993). Changes in the oxidative endpoints of the double-crested cormor- Birds exposed to temperatures outside of their thermoneutral zone ex- ants described here provide supporting evidence of systemic oxide ra- perience increases in reactive oxygen species production (Lin et al., dical damage in the tissues of these birds. Heinz body hemolytic anemia 2008; Beaulieu et al., 2014). In this study, oil-dosed birds lost weight was observed in the birds of this study (Harr et al., in this issue) as well over the course of study due to lack of appetite and lethargy and all oil- as in birds collected from the field after the DWH oil spill (Fallon et al., dosed birds had oil on their plumage as a result of foraging for fish in 2014). This indicates that both wild and captive DWHMC252 oil-ex- the feed tanks that contained oil excreted by the birds. Consequently, posed birds are capable of suffering oxidative injury, given that de- the oil-dosed birds had reduced cloaca temperatures and were observed tection of Heinz bodies provides evidence of direct red blood cell da- to seek supplemental heat under the heat lamps provided (Cunningham mage. Consistent with these findings, Leighton et al. (1985) found that et al., 2017), suggesting that the birds were losing substantial body heat oral exposure of herring gulls (Larus argentatus) to Prudhoe Bay oil for as a result of oil contamination. Thus, the adverse effects of oil exposure two days resulted in oxidant stress on red blood cells as indicated by on the body condition and thermoregulation capabilities of birds would presence of Heinz bodies and elevated red blood cell GSH. Hemolytic further contribute to their overall oxidative stress load. anemia can have significant effects on avian migration and life-history. The life-histories of birds vary among species and thus so does the The destruction of red blood cells through oxidative damage decreases complex system underlying oxidative stress. This makes measuring the bird's ability to transport oxygen. Flying birds, particularly mi- oxidative stress in individuals complicated. Studies suggest that the best grating birds undergoing long distance flights, require increased oxygen way to assess oxidative stress is by use of multiple markers of oxidative to fuel the increased aerobic demands of the flight muscles. Thus, it stress including the use of at least one antioxidant capacity marker and would be presumed that birds suffering from oil-induced anemia would one marker of oxidative damage (Constantini, 2008; Skrip and exhibit a reduction in the ability to sustain flight. In addition to flight McWilliams, 2016). Measuring both aspects of the system are needed to ability, Heinz body anemia also adversely affects reproductive success, better understand how the bird is coping with the oxidative challenge and immune function (Henkel et al., 2012; Leighton, 1993; Briggs et al., and to what extent. Marker choice, as well as matrix choice, may pro- 1997). duce differing results given that activity levels differ with tissues, and Managing oxidative stress is a particularly important physiological thus should be selected based on research goals. Blood-based markers function for migrating birds and although the birds in this study were are often the preferred choice of researchers given the non-invasive not actively flying, flight is a major component of a bird's life history. nature of blood draw and accessibility in the field. However, the en- Therefore, the combination of flight stress and oil exposure stress zyme activity in the blood is relatively low (Constantini, 2008), and should be addressed when considering an individual's long term chal- suggests that tissue concentrations would provide a better under- lenges and probability of survival. Oiled birds are likely to experience standing of the changes occurring in the system. In this study, we great difficulties in trying to combat oxidative stress. In this study, we looked at a total of seven different biomarkers in the liver tissue of found that inactive double-crested cormorants fed oil-contaminated fish double-crested cormorants. We found significant differences in six out experienced increased oxidative stress. Although the cormorants in this of the seven biomarkers measured in the dietary-dosed birds. Given the study were able to induce glutathione levels high enough to avoid thorough use of oxidative biomarkers in this study, we can confidently oxidative damage (as measured by lipid peroxidation), it would be conclude that dietary oil exposure of double-crested cormorants caused expected that in combination with ecologically relevant inducers of increased oxidative stress compared to control birds. However, it oxidative stress such as physical activity, immune response, and re- should be noted that choosing the appropriate marker for oxidative production (Constantini, 2008), oiled birds would be unable to main- damage is also important. In this study, we used products of lipid tain redox balance and suffer exacerbated levels of oxidative damage. peroxidation as a measure for damage. We did not detect any significant Studies have shown that all flying birds experience oxidative stress changes in liver lipid peroxidation, suggesting that induction of the (Constantini et al., 2007, 2008; Jenni-Elermann et al., 2014; Eikenaar defense system was sufficient to prevent cellular damage. Yet, Heinz et al., 2016; Skrip et al., 2015; Skrip and McWilliams, 2016); however, body formations in red blood cells were detected in birds in both of the the preparation and repair ability of an individual determines the extent oil-exposed groups in a dose-dependent manner (Harr et al., in this of damage. Individuals in good condition are better able to maintain issue). This indicates that the oiled cormorants in this study suffered redox homeostasis, experiencing less oxidative stress than individuals cellular oxidative damage. Red bloods cells may be more susceptible to in poor body condition (Constantini et al., 2007). For migratory birds oxidation due to the oxygen-hemoglobin interactions and low levels of affected by oil spills, the body condition of the individual at the onset of antioxidant protection circulating in the blood compared to tissue oil exposure is probably a determining factor in its ability to up-regulate concentrations (Pandey and Rizvi, 2011). defense systems and cope with the oxidative challenge of both oil ex- Acute mortality has proven to be an important endpoint for avian posure and flight. species after oil spill events (Burger, 1993). This study helps demon- Over time, oil exposure reduces body condition due to changes in strate that sub-chronic effects due to low-level exposure to oil are also foraging behavior, loss of motivation to eat, lethargy, and reduced important endpoints for avian species after oil spill events and should nutrient absorption from oil ingestion (Perez et al., in this issue; Burger not be overlooked when estimating mortality. Among other clinical and Tsipoura, 1998; Burger, 1997). Ideally, birds experiencing un- effects, double-crested cormorants fed oil-injected fish at target doses of expected oxidative conditions could easily ingest antioxidant-rich food 5 ml oil/kg/day and 10 ml oil/kg/day for up to 21 days showed ele- to quickly re-establish redox balance (Catoni et al., 2008). However, vated oxidative stress in a dose-dependent manner compared to control birds affected by an oil spill would be unlikely to do this given their birds. Birds face oxidative challenges throughout their life, evolving immobile, lethargic state and the contaminated nature of their en- adaptive mechanisms to maintain health status. However, unexpected vironment. Thus, oiled birds would likely have to rely on the synthesis oxidative challenges such as high PAH exposure may push their adap- of de novo antioxidants. Synthesis of endogenous antioxidants is timely tive mechanisms past their limit and individuals may not be able to and often condition-dependent (Garratt and Brooks, 2012). Birds in compensate for the stress, leading to unrepairable damage. The ability poor condition are energy deficient and would be less likely to allocate of an individual to withstand toxicant-induced oxidative stress may sufficient energy and resources towards antioxidant production. In depend on its early life experiences, body condition, geographic

66 K.L. Pritsos et al. Ecotoxicology and Environmental Safety 146 (2017) 62–67 location of wintering and breeding sites, reproductive effort, health Doyotte, A., Cossu, C., Jacquin, M.C., Babut, M., Vasseur, P., 1997. Antioxidant enzymes, glutathione and lipid peroxidation as relevant biomarkers of experimental or field status, age, migratory status, food choice, sex, and genetics (Constantini exposure in the gills and digestive gland of the freshwater bivalve Unio tumidus. et al., 2012; Metcalfe and Alonso-Alvarez, 2010; Constantini et al., Aquat. Toxicol. 39, 93–110. 2010); and thus it would be expected to observe substantial individual Eikenaar, C., Jönsson, J., Fritzsch, A., Wang, H.-L., Isaksson, C., 2016. Migratory refueling affects non-enzymatic antioxidant capacity, but does not increase lipid peroxidation. differential responsiveness to low-levels of oil exposure. The inability to Physiol. Behav. 158, 26–32. cope with all sources of oxidative stress may force trade-offs in avian Fallon, J.A., Hopkins, W.A., Fox, L., 2013. A practical quantification method for Heinz bodies in birds applicable to rapid response field scenario. Environ. Toxicol. Chem. life-histories. While it is important to accurately estimate the number of 12, 401–405. deaths that result from oil spills, it is also important to understand how Fallon, J.A., Smith, E.P., Hopkins W., 2014. Determining physiological injury to birds non-lethal oil exposure affects life-history decisions that impact overall exposed to oil from The Deepwater Horizon (Mississippi Canyon 252) Spill. USFWS Rep., pp. 1–127. fitness. Forth, H.P., Mitchelmore, C.L., Morris, J.M., Lipton, J., 2016. Characterization of oil and water accommodated fractions used to conduct aquatic toxicity testing in support of Author contributions the Deepwater Horizon oil spill natural resource damage assessment. Environ. Toxicol. Chem. 999, 1–10. http://dx.doi.org/10.1002/etc.3672. Garratt, M., Brooks, R.C., 2012. Oxidative stress and condition-dependent sexual signals: K.M.D., F.C., S.S., K.H., C.A.P, S.J.B., J.E.L. study design; K.H-D., more than just seeing red. Proc. Biol. Sci. 279, 3121–3130. Harr, K.E., et al., 2017. Weathered MC252 crude oil-induced anemia and abnormal ery- F.C. animal care and experimentation; K.L.P. T.M. oxidative stress throid morphology in double-crested cormorants (Phalacrocorax auritus) with light analyses; K.L.P., D.C. data analyses; C.R.P., C.A.P. manuscript pre- microscopic and ultrastructural description of Heinz bodies. Ecotoxicol. Environ. Saf paration (in this issue). Hartung, R., Hunt, G.S., 1966. Toxicity of some oils to waterfowl. J. Wildl. Manag. 30, 564–570. Competing financial interests Henkel, J.R., Sigel, B.J., Taylor, C.M., 2012. Large-scale impacts of the Deepwater Horizon oil spill: can local disturbance affect distant ecosystems through migratory shore- birds. BioScience 62, 676–685. The authors declare that they have no competing financial, profes- Itoh, K., Chiba, T., Takahashi, S., Ishii, T., Igarashi, K., Katoh, Y., Oyake, T., Hayashi, N., sional or personal interests that might have influenced the performance Satoh, K., Hatayama, I., Yamamoto, M., Nabeshima, Y., 1997. An Nrf2/Small Maf heterodimer mediates the induction of phase II detoxifying enzyme genes through or presentation of the work described in this manuscript. antioxidant response elements. Biochem. Biophys. Res. Commun. 236, 313–322. https://doi.org/10.1006/bbrc.1997.6943. Acknowledgements Jenni-Elermann, S., Jenni, L., Smith, S., Constantini, D., 2014. Oxidative stress in en- durance flight: an unconsidered factor in bird migration. PLoS One 9 (5), e97650. http://dx.doi.org/10.1371/journal.pone.0097650. This project was funded in part by the Deepwater Horizon Natural Koivula, M.J., Eeva, T., 2010. Metal-related oxidative stress in birds. Environ. Pollut. 158 (7), 2359–2370. Resource Damage Assessment Trustees and the Nevada Agricultural Leighton, F.A., 1986. Clinical gross and histological findings in Herring Gulls and Atlantic Experiment Station. Puffins that ingested prudhoe bay oil. Vet. Pathol. 23, 254–263. Leighton, F.A., 1993. The toxicity of petroleum oils to birds. Environ. Rev. 1, 92–103. Leighton, F.A., Peakall, D.B., Ronald, G., Butler, R.G., 1983. Heinz- body from the in- References gestion of crude oil: primary toxic effect in marine birds. Science 220, 871–873. Leighton, F.A., Lee, Y.Z., Rahimtula, A.D., O’Brien, P.J., Peakall, D.B., 1985. Biochemical and functional disturbances in red blood cells of herring gulls ingesting Prudhoe Bay Beaulieu, M., Haas, A., Schaefer, H.M., 2014. Self-supplementation and effects of dietary crude oil. Toxicol. Appl. Pharm. 81, 25–31. antioxidants during acute thermal stress. J. Exp. Biol. 217, 370–375. http://dx.doi. Limón-Pacheco, J., Gonsebatt, M.E., 2009. The role of antioxidants and antioxidant-re- org/10.1242/jeb.092825. lated enzymes in protective responses to environmentally induced oxidative stress. Biswas, S.K., Rahman, I., 2009. Environmental toxicity, redox signaling and lung in- Mutat. Res. 674, 137–147. flammation: the role of glutathione. Mol. Aspects of Med. 20, 60–76. https://doi.org/ Lin, H., De Vos, D., Decuypere, E., Buyse, J., 2008. Dynamic changes in parameters of 10.1016/j.mam.2008.07.001. redox balance after mild heat stress in aged laying hens (Gallus gallus domesticus). Briggs, K.T., Gershwin, M.E., Anderson, D.W., 1997. Consequences of petrochemical in- Comp. Biochem. Physiol. 147, 30–35. http://dx.doi.org/10.1016/j.cbpc.2007.07. gestion and stress on the immune system of seabirds. J. Mar. Sci. 54, 718–725. 005. Burger, A.E., 1993. Estimating the mortality of seabirds following oil spills; effects of spill Marasco, V., Constantini, D., 2016. Signaling in a polluted world: oxidative stress as an volume. Mar. Pollut. Bull. 26, 140–143. overlooked mechanism linking contaminants to animal communication. Front. Ecol. Burger, J., 1997. Effects of oiling feeding behavior of sanderlings and semipalmated Evol. 4, 95. http://dx.doi.org/10.3389/fevo.2016.00095. plovers in New Jersey. Condor 99, 290–298. Metcalfe, N.B., Alonso-Alvarez, C., 2010. Oxidative stress as a life-history constraint: the Burger, J., Tsipoura, N., 1998. Experimental oiling of sanderlings (Calidris alba): behavior role of reactive oxygen species in shaping phenotypes from conception to death. and weight changes. Environ. Toxicol. Chem. 17, 1154–1158. Funct. Ecol. 24, 984–996. Catoni, C., Peters, A., Schaefer, H.M., 2008. Life history trade-offs are influenced by di- Nebert, D.W., Dalton, T.P., 2006. The role of cytochrome P450 enzymes in endogenous versity, availability, and interactions of dietary antioxidants. Anim. Behav. 76, signaling pathways and environmental carcinogenesis. Nat. Rev. Cancer 6 (12), 1107–1119. 947–960. http://dx.doi.org/10.1038/nrc2015. Chang, Y.-L., Oey, L., Xu, F.-H., Lu, H.-F., Fujisaki, A., 2011. 2010 oil spill: trajectory Pandey, K.B., Rizvi, S.I., 2011. Biomarkers of oxidative stress in red blood cells. Biomed. projections based on ensemble drifter analyses. Ocean Dyn. 61, 829–839. Pap. Med. Fac. Univ. Palacky. Olomouc Czech Repub. 155 (Suppl 2), 131–136. Cheung, C.C.C., Zheng, G.J., Li, A.M.Y., Richardson, B.J., Lam, P.K.S., 2000. Relationship Perez, C., Moye, J., Cacela, D., Dean, K., Pritsos, C., 2017. Body mass change in flying between tissue concentrations of polycyclic aromatic hydrocarbons and antioxidative homing pigeons externally exposed to Deepwater Horizon crude oil. Ecotoxicol. responses of marine mussels, Perna viridis. Aquat. Toxicol. 52, 189–203. Environ. Saf (in this issue). Constantini, D., 2008. Oxidative stress in ecology and evolution: lessons from avian Peréz, C., Lores, M., Veland, A., 2010. Oil pollution increases plasma antioxidants but studies. Ecol. Lett. 11, 1238–1251. reduces coloration in a seabird. Oecologia 163, 875–884. http://dx.doi.org/10.1007/ Constantini, D., Dell’Omo, G., 2015. Oxidative stress predicts long-term resight prob- s00442-010-1677-2. ability and reproductive success in Scopoli's shearwater (Calonectris diomedea). Sarasquete, C., Segner, H., 2000. Cytochrome P4501A (CYP1A) in teleostean fishes. A Conserv. Physiol. 3 (1), cov024. http://dx.doi.org/10.1093/conphys/con024. review of immunohistochemical studies. Sci. Total Environ. 247, 313–332. Constantini, D., Cardinale, M., Carere, C., 2007. Oxidative damage and anti-oxidant ca- Skrip, M.M., McWilliams, S.R., 2016. Oxidative balance in birds: an atoms-to-organisms- pacity in two migratory bird species at a stop-over site. Comp. Biochem. Physiol. C to-ecology primer for ornithologists. J. Field Ornithol. 87 (1), 1–20. http://dx.doi. Toxicol. Pharmacol. 144, 363–371. org/10.1111/jofo.12135. Constantini, D., Dell’Ariccia, G., Lipp, H.-P., 2008. Long flights and age affect oxidative Skrip, M.M., Bauchinger, U., Goymann, W., Fusani, L., Cardinale, M., Alan, R.R., status of homing pigeons (Columba livia). J. Exp. Biol. 211, 377–381. McWilliams, S.R., 2015. Migrating songbirds on stopover prepare for, and recover Constantini, D., Rowe, M., Butler, M.W., McGraw, K.J., 2010. From molecules to living form, oxidative challenges posed by long-distance flights. Ecol. Evol. 5 (15), systems; historical and contemporary issues in oxidative stress and antioxidant 3198–3209. http://dx.doi.org/10.1002/ece3.1601. ecology. Funct. Ecol. 24, 950–959. Staimer, N., Nguyen, T.B., Nizkorodov, S.A., Delfino, R.J., 2012. Glutathione peroxidase Constantini, D., Monaghan, P., Metcalfe, N.B., 2012. Early life experience primes re- inhibitory assay for electrophilic pollutants in diesel exhaust and tobacco smoke. sistance to oxidative stress. J. Exp. Biol. 215, 2820–2826. Anal. Bioanal. Chem. 403 (2), 431–441. Cossu, C., Doyotte, A., Jacquin, M.C., Babut, M., Exinger, A., Vasseur, P., 1997. Troisi, G., Borjesson, L., Bexton, S., Robinson, I., 2007. Biomarkers of polycyclic aromatic Glutathione reductase, selenium-dependent glutathione peroxidase, glutathione le- hydrocarbon (PAH)-associated hemolytic anemia in oiled wildlife. Environ. Res. 105, vels, and lipid peroxidation in freshwater bivalves, Unio tumidus, as biomarkers of 324–329. aquatic contamination in field studies. Ecotoxicol. Environ. Saf. 38, 122–131. USFWS, 2011. Deepwater Horizon Bird Impact Data from the DOI‐ERDC NRDA Database Cunningham, F., Dean, K., Hanson-Dorr, K., Harr, K., Healy, K., Horak, K., Shriner, S., 12 May 2011. Rupp, T., Bursian, S., Dorr, B., 2017. Development of methods for avian oil toxicity Yamato, O., Koto, I., Maede, Y., 1996. Hemolytic anemia in wild seaducks caused by studies using the double crested cormorant (Phalacrocorax auritus). Ecotoxicol. marine oil pollution. J. Wildl. Dis. 32 (2), 381–384. Environ. Saf. 141, 199–208. http://dx.doi.org/10.1016/j.ecoenv.2017.03.025.

67 Ecotoxicology and Environmental Safety 146 (2017) 68–75

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Reprint of: CYP1A protein expression and catalytic activity in double-crested MARK cormorants experimentally exposed to Deepwater Horizon Mississippi Canyon 252 oil

Courtney R. Alexandera, Michael J. Hooperb, Dave Cacelac, Kim D. Smelkera, Caleshia S. Calvina, Karen M. Deanc, Steve J. Bursiand, Fred L. Cunninghame, Katie C. Hanson-Dorre, Katherine E. Horakf, John P. Isanhartg, Jane Linkd, Susan A. Shrinerf, ⁎ Céline A.J. Godard-Coddinga, ,1 a The Institute of Environmental and Human Health, Texas Tech University, Lubbock, TX, USA b U.S. Geological Survey, Columbia Environmental Research Center, Columbia, MO, USA c Abt Associates, Boulder, CO, USA d Department of Animal Science, Michigan State University, East Lansing, MI, USA e U.S. Department of Agriculture, National Wildlife Research Center-Mississippi Field Station, Mississippi State University, Starkville, MS, USA f U.S. Department of Agriculture, National Wildlife Research Center, Fort Collins, CO, USA g U.S. Department of the Interior, Denver, CO, USA

ARTICLE INFO ABSTRACT

Keywords: Double-crested cormorants (Phalacrocorax auritus, DCCO) were orally exposed to Deepwater Horizon Mississippi Cytochrome P4501A Canyon 252 (DWH) oil to investigate oil-induced toxicological impacts. Livers were collected for multiple Double-crested cormorant analyses including cytochrome P4501A (CYP1A) enzymatic activity and protein expression. CYP1A enzymatic Oil spill activity was measured by alkoxyresorufin O-dealkylase (AROD) assays. Activities specific to the O-dealkylation Deepwater Horizon of four resorufin ethers are reported: benzyloxyresorufin O-debenzylase (BROD), ethoxyresorufin O-deethylase (EROD), methoxyresorufin O-demethylase (MROD), and pentoxyresorufin O-depentylase (PROD). CYP1A protein expression was measured by western blot analysis with a CYP1A1 mouse monoclonal antibody. In study 1, hepatic BROD, EROD, and PROD activities were significantly induced in DCCO orally exposed to 20 ml/ kg body weight (bw) oil as a single dose or daily for 5 days. Western blot analysis revealed hepatic CYP1A protein induction in both treatment groups. In study 2 (5 ml/kg bw oil or 10 ml/kg bw oil, 21 day exposure), all four hepatic ARODs were significantly induced. Western blots showed an increase in hepatic CYP1A expression in both treatment groups with a significant induction in birds exposed to 10 ml/kg oil. Significant correlations were detected among all 4 AROD activities in both studies and between CYP1A protein expression and both MROD and PROD activities in study 2. EROD activity was highest for both treatment groups in both studies while BROD activity had the greatest fold-induction. While PROD activity values were consistently low, the fold- induction was high, usually 2nd highest to BROD activity. The observed induced AROD profiles detected in the present studies suggest both CYP1A4/1A5 DCCO isoforms are being induced after MC252 oil ingestion. A review of the literature on avian CYP1A AROD activity levels and protein expression after exposure to CYP1A inducers highlights the need for species-specific studies to accurately evaluate avian exposure to oil.

1. Introduction days later, of the Deepwater Horizon drilling rig and the deaths of 11 rig workers (BP, 2010). Before it was capped, the uncontrolled well in The Deepwater Horizon (DWH) oil spill in the northern Gulf of the Macondo Prospect of Mississippi Canyon block 252 (MC252) Mexico began April 20th, 2010, with the explosion and sinking, two released more oil than any other oil spill in the history of the United

DOI of original article: http://dx.doi.org/10.1016/j.ecoenv.2017.02.049 ⁎ Corresponding author. E-mail address: [email protected] (C.A.J. Godard-Codding). 1 Associate Professor of Environmental Toxicology, The Institute of Environmental and Human Health, Texas Tech University and TTU Health Sciences Center, Box 41163, Lubbock, Texas 79409-1163. http://dx.doi.org/10.1016/j.ecoenv.2017.05.015 Received 15 September 2016; Received in revised form 25 February 2017; Accepted 27 February 2017 Available online 30 May 2017 0147-6513/ © 2017 Published by Elsevier Inc. C.R. Alexander et al. Ecotoxicology and Environmental Safety 146 (2017) 68–75

States (Camilli et al., 2010). Many birds died from the overwhelming reactions. Activities of the CYP1A-catalyzed O-dealkylation of the effects of direct oiling, while many more were observed alive with methoxy-, ethoxy-, pentoxy-, and benzoxy-substituted resorufin ether varying amounts of oil on skin and plumage, exposures whose substrates are known, respectively, as methoxyresorufin O-demethylase consequences are not well established (Haney et al., 2014). (MROD), ethoxyresorufin O-deethylase (EROD), pentoxyresorufin O- Birds frequently come in contact with oil after a spill due to their depentylase (PROD) and benzyloxyresorufin O-debenzylase (BROD). feeding and behavioral habits (Haney et al., 2014). Contaminated food AROD assays are a widely recognized method for studying CYP1A sources can cause increased oil ingestion while oiled feathers can lead activity and its induction in response to chemical exposure (Bucheli and to increased ingestion of oil through preening, reduced buoyancy and Fent, 1995; Burke et al., 1985; Kubota et al., 2009). On a species-to- thermoregulation, and disruption of flight. Two different routes of species basis, CYP1A varies in its forms and tissue distributions. There is exposure (oral ingestion and oiled feathers) were recently explored in a a resultant variability in the measured activities and induction respon- series of studies to investigate the health, behavioral, and mechanical siveness of the four AROD assays (Brunstrom and Halldin, 1998; Fossi effects of MC252 oil in birds at acute, sub-lethal levels (Bursian et al., et al., 1995; Giorgi et al., 2000; Helgason et al., 2010; Herve et al., 2017, this volume). Double-crested cormorants (Phalacrocorax auritus 2010; Kubota et al., 2009; Walker, 1998). Therefore, all four AROD DCCO) are migratory birds, travelling from the Great Lakes region to activities were explored to assess which of the dealkylase activities are the Gulf of Mexico during winter. This particular piscivorous species of relevant in each exposure scenario and in the particular DCCO species. water bird is commonly chosen as a model bird in toxicology because of Hepatic expression of CYP1A was also studied by directly measuring the its widespread population and habitats (Dorr et al., 2012; Lavoie et al., amount of protein present using semi-quantitative western blotting. 2015; Ofukany et al., 2012). DCCO was one of the many avian species Differences in band intensity allow direct quantification of changes in affected by the DWH oil spill, resulting in fledglings lost due to death of protein concentration following MC252 oil-exposures and the resulting reproductively mature birds (USFWS, 2015). MC252 oil was fed to birds induction of protein synthesis. to explore the toxicity of ingested spilled oil, a common route of Though a considerable amount of research has been done on exposure for birds inhabiting the contaminated area. Feather oiling, a CYP1As in mammals, less is known regarding other vertebrates such second route of exposure, was studied to evaluate the resulting as birds (Heubeck et al., 2003; Kubota et al., 2009; Walker and Ronis, metabolic and thermoregulatory effects and lab- and field-based flight 1989). The acute and sub-chronic sub-lethal responses to oil exposure effects. The profile of health effects endpoints in these studies included in the double-crested cormorant were examined in two separate studies analysis of cytochrome P450 and its induction in liver tissues. in order to help close this gap in the specific case of the DWH spill. We The cytochrome P450 superfamily of phase I metabolism enzymes report here on both CYP1A catalytic activity and protein expression in catalyzes the oxidative biotransformation of a vast array of exogenous DCCO exposed orally to DWH oil in various exposure scenarios. and endogenous compounds (Kubota et al., 2009; Nebert et al., 2006). The cytochrome P4501A (CYP1A) subfamily is of particular interest 2. Materials and methods because it biotransforms polycyclic aromatic hydrocarbons (PAHs) present in oil, leading to both activation and detoxication of these 2.1. Dosing studies compounds (Denison and Whitlock, 1995; Sarasquete and Segner, 2000). Additionally, the metabolism of xenobiotics by CYP1A can CYP1A enzymes and induction were investigated in two studies of result in the production of harmful reactive oxygen species (ROS) that captive double-crested cormorants. Wild caught double-crested cor- have the ability to alter an organism's immune, erythroid, reproductive, morants were young of the year or sub-adults captured and maintained and endocrine systems (Jennifer et al., 2006; Stegeman et al.al., 1992). on a clean diet of captive-reared fingerling channel catfish (Ictalurus The aryl hydrocarbon receptor (AhR), a multiprotein transcription punctatus) from 14 to 21 days prior to dosing. Study 1 DCCO (13 males, factor, regulates the expression of CYP1A (Walker et al., 2000; 7 females) were collected November 6, 2012 from Little Mossy Lake, Whitlock, 1999). AhR agonists include PAHs and planar halogenated Mississippi. Study 2 DCCO (19 males, 1 female) were collected March aromatic hydrocarbons (PHAH) such as planar polychlorinated biphe- 12, 2013 from McIntyre Scatters, Lefore County, Mississippi. In study 1, nyls, dioxins and furans (Safe, 1986, 1994). Exposure to these agonists artificially aged MC252 oil, originating from the DWH oil spill results in CYP1A induction and this highly sensitive process is the basis (DWH7937, batch # B030112; Forth et al., 2016), was mixed with for the wide use of CYP1A expression as a biomarker of exposure to ground catfish filets (1:1) and the resulting slurry was gavaged directly these compounds (Bucheli and Fent, 1995; Stegeman and Lech, 1991; to the birds’ stomach twice each dosing day. Control birds received a Stegeman et al., 1992; van der Oost et al., 2003; Whyte et al., 2000). gavage dose of only ground catfish fillets at the same times as the dosed CYP1A induction is a recognized and widely used measure of birds. To optimize dose absorption and retention in study 2, cormorants exposure to, and molecular effects of, oil and PAHs in humans, were fed catfish that were lightly anesthetized (tricaine methanesulfo- laboratory animals and wildlife species (Godard et al., 2004, 2006; nate) and injected with artificially aged MC252 oil (same source as Stegeman et al., 1992; Webb et al., 2014) including birds (Brausch Study 1). Control catfish received similar anesthesia but no oil. All feed et al., 2010; Brunström et al., 1991, Custer et al., 2000; Lee et al., 1985; fish were allowed to recover from anesthesia before being fed to Lee et al., 1986; Peakall et al., 1989). Avian species express two forms cormorants. A detailed description of all animal care and experimental of CYP1A: CYP1A4 and CYP1A5 (Gilday et al., 1996). Evolutionary procedures can be found elsewhere (Cunningham et al., 2017, this studies indicate the avian CYP1As, CYP1A4 and CYP1A5, are ortholo- issue). In study 1, the birds received 20 ml/kg bw/day of artificially gous to the mammalian CYP1As, CYP1A1 and CYP1A2, respectively; weathered MC252 oil on a single day, day 1, of the study or for 5 days, the pairs diverging after a single duplication event of CYP1A (Goldstone on days 1 through 5. All birds were euthanized on day 6 and liver and Stegeman, 2006; Kubota et al., 2006a). Constitutive expression of samples (approximately 1 g) were placed in cryovials, flash frozen in CYP1A is low, yet highly inducible in the presence of certain xenobio- liquid nitrogen and maintained at -80C until analyzed. Final number of tics, strengthening the utility of this enzyme subfamily as a biomarker birds for the AROD and Western blot analyses are reported in Table 1. (Whitlock, 1999). Mean body weights of Study 1 birds were 1977.3, 1792.5, and 1854.7 Cytochrome P4501A analysis is performed by measuring the rate of before exposure and 1957.1, 1762.7, and 1921.9 after exposure for the enzyme's metabolic degradation of synthetic substrates (its catalytic control, 1-day, and 5 day treatments, respectively. Additional informa- activity) or by directly quantifying the amount of the specific CYP1A tion on DCCO organ weights can be found in Harr et al. (2017) (this protein. Activity measurement of the CYP1A subfamily of enzymes is issue). based on the rates of O-dealkylation of four substituted resorufin ethers In study 2, the cormorants received a nominal dose of either 5 or by what are generally known as alkoxyresorufin O-dealkylase (AROD) 10 ml/kg bw/day of the artificially weathered MC252 oil, daily, for 21

69 C.R. Alexander et al. Ecotoxicology and Environmental Safety 146 (2017) 68–75

Table 1 CYP1A protein expression and activities in livers of double-crested cormorants dosed with artificially aged MC252 petroleum associated with the Deepwater Horizon oil spill. Values are means ( ± standard deviation) of CYP1A protein or activity. Lower case letters indicate significant differences among treatment groups within each enzyme measure. Induction over control values are fold-increases of enzyme measured in petroleum-dosed versus control cormorants.

Treatment Duration N CYP1A Protein N Activity (pmol/min/mg protein)

Intensity BROD EROD MROD PROD

Study 1 Control 3 0.03a ± 0.0001 3 1.77a ± 1.57 11.80a ± 19.07 2.41 ± 1.48 0.47a ± 0.44 20 ml/kg 1 day 5 1.93b ± 0.0018 5 26.00b ± 14.72 56.40b ± 39.79 10.20 ± 6.53 1.09b ± 0.46 20 ml/kg 5 days 7 1.88b ± 0.0022 7 25.30b ± 11.58 39.30a,b ± 18.48 12.80 ± 6.94 2.69b ± 1.09

Induction over control 20 ml/kg 1 day 71.5 14.7 4.8 4.2 2.3 20 ml/kg 5 days 69.6 14.3 3.3 5.3 5.7

Study 2 Control 4 0.07a ± 6.61E-05 6 4.01a ± 3.97 15.20a ± 18.98 5.68a ± 7.44 0.77a ± 0.49 5 ml/kg 21 days 5 2.07a,b ± 0.0030 5 37.29b ± 14.02 39.69a,b ± 12.73 16.89b ± 5.74 3.25b ± 0.85 10 ml/kg 12–15 days 4 1.68b ± 0.0019 6 21.29a,b ± 21.70 56.89b ± 27.98 25.69b ± 11.39 1.92a,b ± 1.19

Induction over control 5 ml/kg 21 days 30.4 9.3 2.6 3.3 4.3 10 ml/kg 12–15 days 24.7 5.3 3.7 4.5 2.5 days in the form of oil-injected catfish. Doses based on actual fish 530 nm excitation and 590 nm emission wavelengths. Resorufin stan- consumption data were 5.2 ± 0.2 and 8.5 ± 0.4 ml/kg bw/day dard and all resorufin substrates were purchased from Sigma. The (mean ± SE). Because of deteriorating physical condition, high dose standard curve was created using 8 resorufin concentrations (0, 0.02, birds were euthanized and tissues collected on days 12 through 15. 0.05, 0.2, 1, 2, 20, and 40 pmol). Rat liver S9 induced with Aroclor Signs of deteriorating physical condition include lack of alertness, 1254 (Celsis) was the positive control. Each well had final assay tucking of head under wings, and a cloacal temperature of 39.4 °C or concentrations of 0.01 mM dicumarol, 1.42 mM NADPH, 40 µg of less. Control and low dose birds were euthanized on day 21 and tissues sample protein, and 2 µM of either 7-benzyloxy, ethoxy, methoxy, or collected and processed as in Study 1. Final numbers of birds for the pentoxyresorufin substrate in a final well volume of 200 µl. NADPH was AROD and Western blot analyses are reported in Table 1. Mean body added to the wells last to initiate the reaction. Limit of detection for the weights of Study 2 birds were 2001, 1880, and 1968 g before exposure plate reader was 0.0005 pmol resorufin/µg of protein (Pezdek, 2014). and 1839, 1673, and 1545 g after exposure for control, 5 ml/kg bw/ An R2 value greater than 0.47 indicated positive activity. A 10% day, and 10 ml/kg bw/day treatments, respectively. Additional infor- coefficient of variance (CV) was acceptable for the positive control mation on DCCO organ weights can be found in Harr et al. (2017) (this and a CV of 50% was acceptable for samples. If the above mentioned special issue). criteria were not met, values were discarded before final calculations Both studies received IACUC approval from USDA/APHIS/WS/ for the overall AROD activity. Final AROD activity was expressed as NWRC. pmol/min/mg protein and the following equation was used: ((Sample slope–Blank slope)/RR standard slope) 2.2. Sample preparation Final AROD activity = Protein mg

2.2.1. Microsome or S9 preparation Microsomes and S9 fractions were prepared as described previously 2.3.2. Western Blotting technique (Trust et al., 2000). Samples were homogenized and then centrifuged at A sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS- 10,000×g for 10 min at 4 °C. The supernatant was centrifuged at PAGE) method was used to separate either microsomal or S9 samples. A 20,000×g for 20 min at 4 °C. The resulting supernatant, the S9 fraction, 4% SDS-PAGE stacking gel and a 12% SDS-PAGE resolving gel were was saved. Microsomes were prepared from larger S9 fractions by used. Microsomal samples in amounts of 16.6 µg protein for high centrifuging at 105,000×g for 1 h and 10 min at 4 °C in an ultracen- treatment groups (to avoid saturation of the total protein stain) and trifuge and resuspending the pellet (Beckman, Optima XL-100K ultra- 33.3 µg protein for medium and control treatment groups were run in centrifuge). the SDS-PAGE. A semi-dry transfer instrument (Bio-Rad) transferred the protein from the polyacrylamide gel to a PVDF membrane. Total 2.2.2. Protein analysis protein staining was carried out with SYPRO Ruby total protein stain Protein concentration was determined using the Bradford assay according to the manufacturer's (Thermo Fisher Scientific) manual. The (Bradford, 1976)on10μl of liver microsomes. Bovine serum albumin membrane was exposed to a UV-CCD camera (Alpha Innotech, (BSA) was used to create the standard curve. All samples were FluorChem SP) to reveal the total protein bands that were quantified measured in triplicate. Protein assay dye (Bio-Rad) was added to the using Quantity One software (Bio-Rad). The PVDF membrane was then sample and standard wells. BioTek Synergy 4 plate reader read the blocked in 5% TBS-Blotto. Mouse monoclonal CYP1A1 antibody (Santa absorbance of all wells at 595 nm. Cruz Biotechnology, sc-393979) was used as primary antibody (1:500 dilution) and followed by a goat anti-mouse secondary antibody (Santa 2.3. Sample analysis Cruz Biotechnology, sc-2005, 1:2000 dilution). ECL A and ECL B (ThermoFisher) photo developers were applied to the membrane to 2.3.1. Alkoxyresorufin-O-dealkylation analysis reveal the CYP1A bands. These were revealed by exposing the Catalytic activity of cytochrome P4501A was assessed as previously membrane to a CCD camera (Alpha Innotech, FluorChem SP) for described in birds (Brausch et al., 2010) but with four AROD activities 20 min. Intensities of the bands were quantified in arbitrary units using (BROD, EROD, MROD, and PROD) measured using a fluorescence plate Quantity One software 4.6.9 (Bio-Rad) after normalization using reader (BioTek Synergy 4) that took 16 readings over 15 min using SYPRO Ruby total protein stain.

70 C.R. Alexander et al. Ecotoxicology and Environmental Safety 146 (2017) 68–75

2.4. Statistical analysis and boxplot design MROD > BROD > PROD in both studies. This profile changed to EROD > BROD > MROD > PROD in study 1 treated birds and in study Differences among treatment groups were assessed with the 2 birds treated with 5 ml/kg MC252 oil daily. Hepatic AROD profiles Kruskal-Wallis test. If statistically significant differences (defined as were similar in both dosing studies: EROD and PROD activities were the p < 0.05) were detected, pairwise comparisons among treatment highest and the lowest, respectively, in both control and treated birds, groups were assessed with Dunn's test using the Bonferroni correction and BROD activity had the highest fold increase. for multiple comparisons. Correlations among CYP1A AROD activities AROD activity levels and profiles from untreated animals have been and band intensity were assessed using Spearman's rank correlation reported previously in double-crested cormorants as well as a variety of (rho). Statistical calculations were performed with R software version avian species and appear to vary widely, likely reflecting the different 3.2.3 (R Core Team, 2015). age groups and provenance of the birds (Custer et al., 2001; Davis et al., For boxplots: letter codes indicate statistically significant pairwise 1997; Esler et al., 2010; Hofius, 1992, Sanderson et al., 1994, comparisons (Dunn's test; p < 0.05, α=0.05). The lower and upper Verbrugge et al., 2001). Basal AROD activities and CYP1A protein limits of the boxes represent the 25th and 75th percentiles, respectively. expression level were 1.3–2.3 fold higher in study 2 than in study 1 but The horizontal black line within the boxes represents the median. The low in both studies (Table 1). Study 2 had a greater proportion of sub- lower and upper whiskers represent the data range and dots depict the adults and birds of the year than study 1, which may explain the individual measured values. observed difference in basal CYP1A expression. Basal hepatic EROD activities in the present study were within a similar range to findings in 3. Results and discussion wild DCCO embryos at pipping from reference colonies in South Dakota and Minnesota (16–18 pmol/min/mg microsomal protein, Custer et al., Crude oil is known to differ among sources in its hydrocarbon 2001) but lower than wild embryos at pipping from a reference site in content and composition. A series of studies was specifically conducted Oregon (54 pmol/min/mg protein, Davis et al., 1997), wild 10-day old to understand possible physiological, enzymatic, phenotypic and meta- chicks from a control site (31–56 pmol/min/mg microsomal protein, bolic changes in birds after exposure to MC252 oil through multiple Custer et al., 2001), untreated nestlings from Gull Island, Lake Huron routes (this special issue). Here, we report exclusively on CYP1A (around 80 pmol/min/mg microsomal protein according to publication catalytic activity and protein expression in captive double-crested figure, Verbrugge et al., 2001), and day-old hatchlings from a Canadian cormorants exposed orally to MC252 oil. Crude oil studies have been reference site (283 pmol/min/mg microsomal protein Sanderson et al., performed on a variety of species, especially fish and seaducks, to 1994). Some of the above basal EROD values are comparable or higher explore potential association between exposure and CYP1A catalytic than EROD activities found after treatment in the present study. activity or protein induction (Holth et al., 2014; Miles et al., 2007; Verbrugge and coauthors (2001) measured additional hepatic ARODs Stagg et al., 2000; Esler et al., 2010; Trust et al., 2000). To our in untreated nestlings from Gull Island and reported higher BROD and knowledge, this is the first study reporting CYP1A catalytic activity and MROD levels than in our studies (around 70 and 20 pmol/min/mg protein induction in double-crested cormorant after exposure to crude microsomal protein, respectively, according to publication figure) but oil. did not detect PROD. EROD and PROD activities (the only two studied) stayed at similar activities from hatching to fledgling (Hofius, 1992). 3.1. CYP1A catalytic activity Similar results were found in wild birds collected from reference sites (Esler et al., 2010). CYP1A AROD activity in birds is a widely used and accepted Many studies have reported CYP1A induction after crude oil or PAH biomarker of exposure to toxicants that are AhR agonists. Significant exposure in duck species and other bird species. Herring gull chicks positive correlations between such toxicant concentration/toxic equiva- orally exposed to either Prudhoe Bay or Hibernia crude oil had elevated lents (TEQs) and AROD activities have been found previously in double- hepatic microsomal EROD activities of approximately 19-fold over the crested cormorants and common cormorants exposed to CYP1A in- control birds (Lee et al., 1985). Rates of EROD activity in herring gull ducers (Kubota et al., 2005; Sanderson et al., 1994; van den Berg et al., chicks exposed to Prudhoe Bay crude oil revealed greater induction by 1994; Guruge and Tanabe, 1997) as well as in many other avian species the aromatic fraction than the aliphatic fraction when compared to exposed to PAHs or crude oil such as lesser scaup (Aythya affinis, Custer control birds (6.6-fold versus 2-fold, Peakall et al., 1989). Studies et al., 2000), and herring gull (Larus-Argentatus, Peakall et al., 1989). In performed 10 years after the Exxon-Valdez oil spill revealed EROD addition, EROD activity in Steller's eiders (Polysticta stelleri) was activity was still significantly higher than in birds from unoiled sites: positively correlated with PAH concentration in blue mussels (Mytilus pigeon guillemots (Cepphus columba, 1.6-fold), harlequin duck (Histrio- edulis), a common prey item (Miles et al., 2007). nicus histrionicus, 2.9-fold), and Barrow's goldeneyes (Bucephala islandi- Significant induction in AROD activity was detected in the livers of ca, 1.9-fold) (Golet et al., 2002; Trust et al., 2000). Twenty years after double-crested cormorants in both studies (Table 1, Figs. 1 and 2). In the spill, harlequin ducks from oiled sites had EROD activity levels 3.7- study 1, oil exposure led to significant BROD, EROD, and PROD fold greater than unoiled sites (Esler et al., 2010). Lesser scaup collected induction and a trend of increased MROD activity. EROD was found from areas with widespread petroleum contamination had higher to have the highest basal and induced activities of all four ARODS in BROD, EROD, and MROD levels (11.8-, 10.3-, 3.5-fold) than those both single and multiple Study 1 treatment groups. BROD had the from reference sites (Custer et al., 2000). Both Harlequin ducks and highest fold increase (over 14-fold) seen in both treatment groups when Steller's eiders from a petroleum-contaminated site exhibited induced compared to the control. Study 1 EROD, MROD and PROD inductions EROD levels when compared to birds from a control site (Miles et al., ranged from 3- to 6-fold. In study 2, all four ARODs were significantly 2007). induced in liver microsomes upon oil exposure. EROD was found to Induced AROD profiles of cormorants and other avian species have the highest activity in both the 5 ml/kg and 10 ml/kg study 2 exposed to a variety of CYP1A inducers highlight species differences. treatment groups. One bird in the 10 ml/kg study 2 treatment group Double-crested cormorant chicks exposed twice to β-Napthoflavone had a low EROD activity but that outlying datapoint could not be (BNF) and sampled 48 h after the last intraperitoneal injection had an excluded after review of raw data or pathology findings. BROD had the AROD profile (EROD > BROD≥MROD > PROD) similar to that in our highest fold-induction (over 9-fold and 5-fold) in both the 5 ml/kg and two studies (Verbrugge et al., 2001). Contrasting with our study, an 10 ml/kg study 2 treatment groups when compared to the control AROD profile of EROD > MROD > BROD > PROD was found in male group. Study 2 EROD MROD and PROD inductions ranged from 2.5-fold and female, juvenile and adult, common cormorants (Phalacrocorax to 4.5-fold. The AROD activity profile of control birds was EROD > carbo) environmentally exposed to PHAHs (Kubota et al., 2005). A

71 C.R. Alexander et al. Ecotoxicology and Environmental Safety 146 (2017) 68–75

Fig. 1. CYP1A western blot band intensity and AROD activity in study 1 liver microsomes after oral exposure to 20 ml/kg MC252 oil per day for one day (single) or 5 days (multiple). The superscript lowercase indicates significant differences within each boxplot among treatment groups. See Table 1 for sample sizes and mean values. profile of EROD > PROD > BROD > MROD was induced in black- min/mg (embryos environmentally exposed to dioxin-like compounds, eared kites (Milvus migrans lineatus) contaminated with dioxin-like Davis et al., 1997) and a range of 21.2–68.3 pmol/min/mg microsomal compound exposure (Kubota et al., 2006b). CYP1A4 and CYP1A5 have protein (environmentally exposed to p,p′-DDE and PCBs, Custer et al., been shown to preferentially bind to different resorufins and have 2001). uneven catalytic potential, likely contributing to the array of AROD BROD has been cited as a better indicator of CYP1A activity in birds profiles found in different studies reporting CYP1A activity in birds than in mammals (Elliott et al., 1996; Verbrugge et al., 2001). BROD (Kubota et al., 2009). Common cormorant CYP1A4 and CYP1A5 activity, in both studies 1 and 2, was found to be induced to the greatest expressed individually in yeast microsomes and exposed to dioxin-like degree compared to controls. Comparable results were found in lesser compounds, revealed EROD activity values were highest for both scaup exposed to PAHs (Custer et al., 2000). In contrast, MROD activity enzymes followed by BROD for CYP1A4 and MROD for CYP1A5 and was the most highly induced activity in double-crested cormorant that CYP1A4 showed a higher binding affinity and catalytic potential chicks exposed to BNF (19.5-fold, Verbrugge et al., 2001). Wild for BROD than CYP1A5 for MROD (Kubota et al., 2009). In the present black-footed albatrosses (Phoebastria nigripes) and jungle crows (Corvus study, all 4 ARODs were significantly induced after oil ingestion macrorhynchos) exposed to dioxin-like compounds, were reported to suggesting induction of both CYP1A isoforms in DCCO. have low BROD activity and high MROD activity (Kubota et al., 2010; EROD activity was highest for both dosing scenarios in the present Watanabe et al., 2005). PROD activities were the lowest AROD study. EROD activity (599 pmol/min/mg microsomal protein) was also activities in both studies reported here, which was consistent with highest out of all four ARODs in double-crested cormorant chicks studies in other species of cormorant (Kubota et al., 2005; Verbrugge intraperitoneally exposed to BNF (Verbrugge et al., 2001). Studies on et al., 2001; Guruge and Tanabe, 1997; van den Berg et al., 1994). This male, female, juvenile, and adult wild common cormorants exposed to is not consistent across all bird species, as studies of black-eared kites PHAHs all reported EROD activity as highest (110–470 pmol/min/mg) and black-footed albatrosses with environmental exposures to dioxin- (Kubota et al., 2005). Studies that focused on EROD activity alone in like compounds demonstrated PROD activities higher than MROD and/ double-crested cormorant chicks reported median values at 193 pmol/ or BROD activity (Kubota et al., 2006, Kubota et al., 2010). Despite the

Fig. 2. CYP1A western blot band intensity and AROD activity in study 2 liver microsomes after oral exposure to 5 or 10 ml/kg MC252 oil daily for up to 21 days. The superscript lowercase indicates significant differences within each boxplot among treatment groups. See Table 1 for sample sizes and mean values.

72 C.R. Alexander et al. Ecotoxicology and Environmental Safety 146 (2017) 68–75

Table 2 PCB contaminated sites when compared to reference sites. The fold- Correlative relationships among hepatic microsomal AROD activities in double-crested increase found in the present study is greater than those found in the fi cormorants orally dosed with arti cially aged MC252 oil. Rho values from Spearman's previously mentioned studies. rank correlation analysis. Italics represent statistically significant values (p < 0.05). For the EROD-PROD correlation in Study 2, p=0.05. 3.3. Correlation analyses among ARODs and between AROD and CYP1A EROD MROD PROD protein induction

Study 1 fi 0.79 0.51 0.60 BROD Signi cant positive correlations were found among all four hepatic 0.84 0.52 EROD AROD activities in both studies (Table 2). Similar findings have been 0.76 MROD reported in black-eared kites, common cormorants, and lesser scaups Study 2 exposed to PAHs (Kubota et al., 2006, Kubota et al., 2005; Custer et al., 0.72 0.65 0.87 BROD 2000). In study 2 DCCO livers, a significant positive correlation was 0.67 0.48 EROD detected between CYP1A band intensity and MROD (rho=0.65, 0.69 MROD p=0.018) and PROD activity (rho=0.62, p=0.027). A previous study conducted on wild common cormorants environmentally exposed to AhR agonists reported significant positive correlations between hepatic AROD activity and CYP1A but not CYP2B, CYP2C, nor CYP3A protein fi low PROD activities found in the present study, fold increase was high, (Kubota et al., 2005). No signi cant correlation between DCCO liver usually the second highest, and therefore, cannot be dismissed as CYP1A band intensity and EROD or BROD in study 2 nor any of the biologically irrelevant. AROD activities in study 1 were detected, possibly due to the small fi AROD induction pattern and levels vary among bird species and sample size of the studies or the catalytic speci cities of the DCCO between inducers in the same species. Thus AROD monitoring for CYP1As. Further studies are needed to elucidate in greater details CYP1A induction in wild birds should include all four of the resorufin correlation patterns between AROD and CYP1A protein in birds substrates to ensure capturing these species-, tissue-, and inducer- exposed to oil. specific variations. Induction profiles of CYP1A catalytic activity in Further analyses will investigate potential correlations among these the present study occurred in response to oral exposures of artificially detected CYP1A changes and other endpoints of oxidative stress aged oil from the Deepwater Horizon oil spill. Relative to crude oil from examined in the birds. Oxidative stress endpoints include measure- other spills, MC252 DWH oil was found to be high in alkanes and low in ments of antioxidants, antioxidant enzymes, and % Heinz bodies. PAH composition (Faksness et al., 2015; Turner et al., 2014). As 4. Conclusion petroleum oils are complex mixtures of CYP inducers and inhibitors, variability in induction responses should be anticipated in other DWH- associated species. Exposures to spills of other, non-DWH petroleum Our data show that acute and subchronic ingestion of MC252 oil oils might lead to different AROD response patterns, even in double- induces a hepatic CYP1A response in double-crested cormorants. fi crested cormorants, the species studied here. Signi cant induction was found through both CYP1A catalytic activity (AROD) and protein expression (western blot). Based on the AROD profiles observed in treated birds and the known catalytic specificity of 3.2. CYP1A protein expression CYP1A isoforms in common cormorant, both CYP1A4 and CYP1A5 were likely induced in DCCO in the present studies. Exposure to MC252 CYP1A protein expression has been used as a biomarker of PAH oil resulted in significant induction of all four ARODs. EROD had the exposure in many organisms including seabirds, fish, and sea turtles highest activity and BROD the highest fold induction in both studies (Verbrugge et al., 2001; Nakayama et al., 2008; Giannetti et al., 2012). and MROD and PROD correlated with CYP1A protein expression in CYP1A1 mouse monoclonal antibody was used to probe CYP1A4/1A5 study 2. We suggest the measurements of all four AROD activities as in the double-crested cormorants. While the antibody of choice was well as CYP1A protein expression to ensure a thorough evaluation of oil monoclonal and a single band was detected for each liver microsome exposure in birds. AROD activities and induction profiles are known to sample, we choose to refer to the protein detected as CYP1A rather than vary between species indicating species-specific studies are needed to CYP1A4. A single band may result because the two bird isoforms (1A4/ accurately evaluate avian exposure to oil. The biomarkers of exposure 1A5) exhibit high molecular similarities and have similar weights to oil used in this study should be included as part of a weight-of- differing by less than 1 kDa (Kubota et al., 2009), thus both isoforms evidence approach to assessing oil-induced injury to birds. could possibly be contributing to band intensity. Western blot analysis revealed significant CYP1A protein induction Declaration of interest in livers of DCCO exposed to oil in both of our studies. A 71-fold and 69- fold induction was noted in study 1 DCCO from the single and multiple There were no financial, professional, or personal conflicts of dose treatment groups, respectively (Table 1). Study 2 showed a interest with any author listed on this manuscript. significant 30-fold and 24-fold increase in CYP1A band intensity in the liver for the 5 and 10 ml/kg MC252 oil treatment groups, Acknowledgments respectively, when compared to the control group (Table 1). CYP1A protein levels have been shown to increase in birds after This project was funded in part by the U.S. Fish and Wildlife Service, exposure to various AhR agonists. Induction of CYP1A protein expres- Natural Resource Damage Assessment Trustees. AROD and western blot sion varies among species and contaminant exposure. For example, analyses were performed at Texas Tech University. Any use of trade, CYP1A band intensity increased 2–3-fold in double-crested cormorant firm, or product names is for descriptive purposes only and does not chicks exposed to halogenated aromatic hydrocarbon contamination imply endorsement by the U.S. Government. when compared to reference sites (Sanderson et al., 1994). In the same study, TEQs and CYP1A protein concentration were correlated at 5 References different sites across Canada. Double-crested cormorant chicks exposed to BNF resulted in induced CYP1A-like protein (Verbrugge et al., 2001). BP, 2010. Deepwater Horizon Accident Investigation Report. Viewed at: 〈http://www.bp. Bald eagle chicks exhibited 6-fold higher CYP1A protein expression at com/content/dam//pdf/sustainability/issue-reports/Deepwater_Horizon_

73 C.R. Alexander et al. Ecotoxicology and Environmental Safety 146 (2017) 68–75

Accident_Investigation_Report.pdf〉. taap.2000.8994. Bradford, M.M., 1976. Rapid and sensitive method for quantitation of microgram Godard, C., Smolowitz, R., Wilson, J., Payne, R., Stegeman, J.J., 2004. Induction of quantities of protein utilizing principle of protein-dye binding. Anal. Biochem. 72 cetacean cytochrome P4501A1 by b-naphthoflavone exposure of skin biopsy slices. (1–2), 248–254. http://dx.doi.org/10.1006/abio.1976.9999. Toxicol. Sci. 80, 268–275. http://dx.doi.org/10.1093/toxsci/kfh124. Brausch, J.M., Blackwell, B.R., Beall, B.N., Caudillo, C., Kolli, V., Godard-Codding, C., Godard, C., Wise, S.S., Kelly, R.S., Goodale, B., Kraus, S., Romano, T., et al., 2006. Benzo et al., 2010. Effects of polycyclic aromatic hydrocarbons in Northern bobwhite quail [a]pyrene cytotoxicity in right whale (Eubalaena glacialis) skin, testis and lung cell (Colinus virginianus). J. Toxicol. Environ. Health-Part A 73, 540–551. http://dx.doi. lines. Mar. Environ. Res. 62, S20–S24. http://dx.doi.org/10.1016/j.marenvres.2006. org/10.1080/15287390903566559. 04.016. Brunström, B., Broman, D., Naf, C., 1991. Toxicity and EROD-inducing potency of 24 Goldstone, H.M.H., Stegeman, J.J., 2006. A revised evolutionary history of the CYP1A polycyclic aromatic-hydrocarbons (PAHs) in chick-embryos. Arch. Toxicol. 65 (6), subfamily: gene duplication, gene conversion, and positive selection. J. Mol. Evol. 62 485–489. http://dx.doi.org/10.1007/BF01977361. (6), 708–717. http://dx.doi.org/10.1007/s00239-005-0134-z. Brunstrom, B., Halldin, K., 1998. EROD induction by environmental contaminants in Golet, G.H., Seiser, P.E., McGuire, A.D., Roby, D.D., Fischer, J.B., Kuletz, K.J., Irons, D.B., avian embryo livers. Comp. Biochem. Physiol. C.-Pharmacol. Toxicol. Endocrinol. 21 Dean, T.A., Jewett, S.C., Newman, S.H., 2002. Long-term direct and indirect effects of (1–3), 213–219. http://dx.doi.org/10.1016/S0742-8413(98)10042-7. the 'Exxon Valdez' oil spill on pigeon guillemots in Prince William Sound, Alaska. Bucheli, T.D., Fent, K., 1995. Induction of cytochrome-P450 as a biomarker for Mar. Ecol. Prog. Ser. 241, 287–304. http://dx.doi.org/10.3354/meps241287. environmental contamination in aquatic ecosystems. Crit. Rev. Environ. Sci. Technol. Guruge, K.S., Tanabe, S., 1997. Congener specific accumulation and toxic assessment of 25 (3), 201–268. http://dx.doi.org/10.1080/10643389509388479. polychlorinated biphenyls in common cormorants, Phalacrocorax carbo, from Lake Burke, M.D., Thompson, S., Elcombe, C., Halpert, J., Haaparant, T., Mayers, R.T., 1985. Biwa, Japan. Environ. Pollut. 96 (3), 425–433. http://dx.doi.org/10.1016/S0269- Ethoxy-, pentoxy- and benzyloxphenoxazones and homologues: a series of substrates 7491(97)00045-6. to distinguish between different induced cytochromes P-450. Biochem. Pharmacol. Haney, J.C., Geiger, H.J., Short, J.W., 2014. Bird mortality from the deepwater Horizon 34 (18), 3337–3345. oil spill. I. Exposure probability in the offshore Gulf of Mexico. Mar. Ecol. Prog. Ser. Bursian, S.J., Cacela, D., Cunningham, F., Dean, K.M., Dorr, B.S., Codding, C., Ellis, C., 513, 225–237. http://dx.doi.org/10.3354/meps10991. Guglielmo, C., Hanson-Dorr, K.C., Harr, K.E., Healy K., Hooper, M., Horak, K., et al. Harr, K.E, Bursian, S.J., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Hanson- 2017 (in this issue). Overview of Avian Toxicity Studies for the Deepwater Horizon Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Reavill, D.R., Shriner, S.A., Schmidt, Natural Resource Damage Assessment. R.E., 2017. (in this issue). Comparison of Organ Weights And Histopathology Camilli, R., Reddy, C.M., Yoerger, D.R., Van Mooy, B.A.S., Jakuba, M.V., Kinsey, J.C., Between Double-crested Cormorants (Phalacrocorax auritus) Dosed Orally Or McIntyre, C.P., Sylva, S.P., Maloney, J.V., 2010. Tracking hydrocarbon plume Externally with Artificially Weathered Mississippi Canyon 252 Crude Oil. transport and biodegradation at Deepwater Horizon. Science 330 (6001), 201–204. Helgason, L.B., Arukwe, A., Gabrielsen, G.W., Harju, M., Hegseth, M.N., Heimstad, E.S., http://dx.doi.org/10.1126/science.1195223. et al., 2010. Biotransformation of PCBs in Arctic seabirds: characterization of phase I Cunningham, F, Dean, K.M., Hanson-Dorr, K.C., Harr, K.E., Healy K., Horak, K., Link, J., and II pathways at transcriptional, translational and activity levels. Comp. Biochem. Shriner, S., Bursian, S., Dorr, B., 2017. in this issue. Development of methods for Physiol. C.-Toxicol. Pharmacol. 152 (1), 34–41. http://dx.doi.org/10.1016/j.cbpc. avian oil toxicity studies using the double crested cormorant (Phalacrocorax auritus). 2010.02.009. Custer, T.W., Custer, C.M., Hines, R.K., Sparks, D.W., Melancon, M.J., Hoffman, D.J., Herve, J.C., Crump, D., Jones, S.P., Mundy, L.J., Giesy, J.P., Zwiernik, M.J., et al., 2010. et al., 2000. Mixed function oxygenases, oxidative stress, and chromosomal damage Cytochrome P4501A induction by 2,3,7,8-tetrachlorodibenzo-p-dioxin and two measured in lesser scaup wintering on the Indiana Harbor Canal. Arch. Environ. chlorinated dibenzofurans in primary hepatocyte cultures of three avian species. Contam. Toxicol. 38 (4), 522–529. http://dx.doi.org/10.1007/s002449910068. Toxicol. Sci. 113 (2), 380–391. http://dx.doi.org/10.1093/toxsci/kfp270. Custer, T.W., Custer, C.M., Hines, R.K., Stromborg, K.L., Allen, P.D., Melancon, M.J., Heubeck, M., Camphuysen, K.C.J., Bao, R., Humple, D., Rey, A.S., Cadiou, B., Brager, S., Henshel, D.S., 2001. Organochlorine contaminants and biomarker response in Thomas, T., 2003. Assessing the impact of major oil spills on seabird populations. double-crested cormorants nesting in Green Bay and Lake Michigan, Wisconsin, USA. Mar. Pollut. Bull. 46 (7), 900–902. http://dx.doi.org/10.1016/S0025-326X(03) Arch. Environ. Contam. Toxicol. 40 (1), 89–100. http://dx.doi.org/10.1007/ 00098-5. s002440010151. Hofius, 1992. Characterization and Induction of Hepatic and Renal Detoxification Davis, J.A., Fry, D.M., Wilson, B.W., 1997. Hepatic ethoxyresorufin-o-deethylase activity Enzymes in Nestling European Starlings (Sturnus vulgaris). (Master's Thesis) Clemson and inducibility in wild populations of double-crested cormorants (Phalacrocorax University. auritus). Environ.Toxicol. Chem. 16, 1441–1449. Holth, T.F., Eidsvoll, D.P., Farmen, E., Sanders, M.B., Martinez-Gomez, C., Budzinski, H., Denison, M.S., Whitlock, J.P., 1995. Xenobiotic-inducible transcription of cytochrome Burgeot, T., Guilhermino, L., Hylland, K., 2014. Effects of water accommodated P450 genes. J. Biol. Chem. 270 (31), 18175–18178. http://dx.doi.org/10.1074/jbc. fractions of crude oils and diesel on a suite of biomarkers in Atlantic cod (Gadus 270.31.18175. morhua). Aquat. Toxicol. 154, 240–252. http://dx.doi.org/10.1016/j.aquatox.2014. Dorr, B.S, Taylor, J.D., Werner, S.J., King, D.T., Farquhar, J.F., Mazzocchi, I.M., 05.013. Mccullough, R.D., 2012. Summer and migrational movements of satellite-marked Kubota, A., Iwata, H., Goldstone, H.M.H., Kim, E.Y., Stegeman, J., Tanabe, S., 2006a. double-crested cormorants from a breeding colony managed by egg-oiling in Lake Cytochrome P450 1A4 and 1A5 in common cormorant (Phalacrocorax carbo): Ontario, USA. Waterbirds 35(Special Publication 1): 114-123. Evolutionary relationships and functional implications associated with dioxin and Elliott, J.E., Norstrom, R.J., Lorenzen, A., Hart, L.E., Philibert, H., Kennedy, S.W., related compounds. Toxicol. Sci. 92 (2), 394–408. http://dx.doi.org/10.1093/toxsci/ Stegeman, J.J., Bellward, G.D., Cheng, K.M., 1996. Biological effects of kfl001. polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls in bald eagle Kubota, A., Iwata, H., Tanabe, S., Yoneda, K., Tobata, S., 2005. Hepatic CYP1A induction (Haliaeetus leucocephalus) chicks. Environ. Toxicol. Chem. 15 (5), 782–793. http://dx. by dioxin-like compounds, and congener-specific metabolism and sequestration in doi.org/10.1897/1551-5028(1996)015<0782:BEOPDP>2.3.CO;2. wild common cormorants from Lake Biwa, Japan. Environ. Sci. Technol. 39 (10), Esler, D., Trust, K.A., Ballachey, B.E., et al., 2010. Cytochrome P4501A biomarker 3611–3619. http://dx.doi.org/10.1021/es048771g. indication of oil exposure in harlequin ducks up to 20 years after the Exxon Valdez oil Kubota, A., Iwata, H., Tanabe, S., Yoneda, K., Tobata, S., 2006b. Congener-specific spill. Environ. Toxicol. Chem. 29 (5), 1138–1145. http://dx.doi.org/10.1002/etc. toxicokinetics of polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans, 129. and coplanar polychlorinated biphenyls in black-eared kites (Milvus migrans): Faksness, L.-G., Altin, D., Nordtug, T., Daling, P.S., Hansen, B.H., 2015. Chemical Cytochrome P4501A-dependent hepatic sequestration. Environ. Toxicol. Chem. 25 comparison and acute toxicity of water accommodated fraction (WAF) of source and (4), 1007–1016. http://dx.doi.org/10.1897/05-224R.1. field collected Macondo oils from the Deepwater Horizon spill. Mar. Pollut. Bull. 91 Kubota, Akira, Kim, Eun-Young, Iwata, Hisato, 2009. Alkoxyresorufin (methoxy-, ethoxy-, (1), 222–229. pentoxy- and benzyloxyresorufin) O-dealkylase activities by in vitro-expressed Forth, H.P., Mitchelmore, C.L., Morris, J.M., Lay, C.R., Suttles, S.E., Lipton, J., 2016. cytochrome P450 1A4 and 1A5 from common cormorant (Phalacrocorax carbo). Characterization of dissolved and particulate phases of water accommodated Comp. Biochem. Physiol. C.-Toxicol. Pharmacol. 149 (4), 544–551. http://dx.doi. fractions used to conduct aquatic toxicity testing in support of the Deepwater Horizon org/10.1016/j.cbpc.2008.12.004. natural resource damage assessment. Environ. Toxicol. Chem. 36 (6), 1460–1472. Kubota, A., Watanabe, M., Kunisue, T., Kim, E., Tanabe, S., Iwata, H., 2010. Hepatic http://dx.doi.org/10.1002/etc.3803. CYP1A induction by chlorinated dioxins and related compounds in the endangered Fossi, M.C., Massi, A., Lari, L., Marsili, L., Focardi, S., Leonzio, C., et al., 1995. black-footed albatross from the North Pacific. Environ. Sci. Technol. 44 (9), Interspecies differences in mixed-function oxidase activity in birds – Relationship 3559–3565. http://dx.doi.org/10.1021/es1000177. between feeding-habits, detoxication activities and organochlorine accumulation. Lavoie, R.A., Kyser, T.K., Friesen, V.L., Campbell, L.M., 2015. Tracking overwintering Environ. Pollut. 90 (1), 15–24. http://dx.doi.org/10.1016/0269-7491(94)00098-X. areas of fish-eating birds to identify mercury exposure. Environ. Sci. Technol. 49 (2), Giannetti, M., Casini, S., Marsili, L., Maltese, S., Campani, T., Carletti, L., Fossi, M.C., 863–872. http://dx.doi.org/10.1021/es502813t. Fossi, M.C., 2012. First evidence of induction of CYP1A in Caretta caretta skin biopsy Lee, Y.Z., Leighton, F.A., et al., 1985. Effects of ingestion of Hibernia and Prudhoe Bay slices treated with polycyclic aromatic hydrocarbons, organochlorines and crude oils on hepatic and renal mixed-function oxidase in nestling herring-gulls polybrominated diphenyl ethers. Comp. Biochem. Physiol. A-Mol. Integr. Physiol. (Larusargentatus). Environ. Res. 36 (1), 248–255. http://dx.doi.org/10.1016/0013- 163 (1), S18–S19. http://dx.doi.org/10.1016/j.cbpa.2012.05.059. 9351(85)90022-2. Gilday, D., Gannon, M., Yutzey, K., Bader, D., Rifkind, A.B., 1996. Molecular cloning and Lee, Y.Z., Obrien, P.J., et al., 1986. Toxicity of petroleum crude oils and their effect on expression of two novel avian cytochrome p450 1A enzymes induced by 2,3,7,8- xenobiotic metabolizing enzyme-activities in the chicken-embryo In ovo. Environ. tetrachlorodibenzo-p-dioxin. J. Biol. Chem. 271 (51), 33054–33059. http://dx.doi. Res. 39 (1), 153–163. http://dx.doi.org/10.1016/S0013-9351(86)80017-2. org/10.1074/jbc.271.51.33054. Miles, A.K., Flint, P.L., Trust, K.A., Ricca, M.A., Spring, S.E., Arrieta, D.E., Hollmen, T., Giorgi, M., Marini, S., Longo, V., Mazzaccaro, A., Amato, G., Gervasi, P.G., 2000. Wilson, B.W., 2007. Polycyclic aromatic hydrocarbon exposure in Steller's eiders Cytochrome P450-dependent monooxygenase activities and their inducibility by (Polysticta stelleri) and harlequin ducks (Histronicus histronicus) in the eastern Aleutian classic P450 inducers in the liver, kidney, and nasal mucosa of male adult ring-necked Islands, Alaska, USA. Environ. Toxicol. Chem. 26 (12), 2694–2703. http://dx.doi. pheasants. Toxicol. Appl. Pharmacol. 167 (3), 237–245. http://dx.doi.org/10.1006/ org/10.1897/07-259.1.

74 C.R. Alexander et al. Ecotoxicology and Environmental Safety 146 (2017) 68–75

Nakayama, A., Riesen, I., Kollner, B., Eppler, E., Segner, H., 2008. Surface marker-defined species: carcinogen metabolism and biomarkers for carcinogen and pollutant head kidney granulocytes and B lymphocytes of rainbow trout express benzo[a] exposure. Environ. Health Perspect. 90, 101–109. http://dx.doi.org/10.2307/ pyrene-inducible cytochrome P4501A protein. Toxicol. Sci. 103 (10), 86–96. http:// 3430851. dx.doi.org/10.1093/toxsci/kfn024. Trust, K.A., Esler, D., Woodin, B.R., Stegeman, J.J., 2000. Cytochrome P450 1A induction Nebert, Daniel W., Dalton, Timothy, P., 2006. The role of cytochrome P450 enzymes in in sea ducks inhabiting nearshore areas of Prince William Sound, Alaska. Mar. Pollut. endogenous signalling pathways and environmental carcinogenesis. Nat. Rev. Cancer Bull. 40 (5), 397–403. 6 (12), 947–960. http://dx.doi.org/10.1038/nrc2015. USFWS (U.S. Fish and Wildlife Service). 2015. Technical Report: Quantification of fledglings Ofukany, A.F.A., Hobson, K.A., Wassenaar, Leonard, I., 2012. Connecting breeding and lost in 2011. DWH/MC252 Oil Spill Natural Resource Damage Assessment. August 31, wintering habitats of migratory piscivorous birds: implications for tracking 2015. 24p. Viewed here: https://www.fws.gov/doiddata/dwh-ar-documents/788/ contaminants (Hg) using multiple stable isotopes. Environ. Sci. Technol. 46 (6), DWH-AR0152670.pdf. 3263–3272. http://dx.doi.org/10.1021/es204135s. van den Berg, M., Craane, B.L.H.J., Sinnige, T., Vanmourik, S., Dirksen, S., Boudewijn, T., Peakall, D.B., Norstrom, R.J., Jeffrey, D.A., Leighton, F.A., 1989. Induction of hepatic van der Gaag, M., Lutke-Schipholt, I.J., Spenkelink, B., Brouwer, A., 1994. mixed-function oxidases in the herring Gull (Larusargentatus) by Prudhoe Bay crude- Biochemical and toxic effects of Polychlorinated-biphenyls (PCBs), Dibenzo-p-dioxins oil and its fractions. Comp. Biochem. Physiol. C.-Pharmacol. Toxicol. Endocrinol. 94 (PCDDs) and Dibenzofurans (PCDFs) in the cormorant (Phalacrocorax carbo) after in (2), 461–463. http://dx.doi.org/10.1016/0742-8413(89)90098-4. ovo exposure. Environ. Toxicol. Chem. 13 (5), 803–816. http://dx.doi.org/10.1897/ Pezdek, R., 2014. Alkoxyresorufin-o-dealkylation Assay Optimization in Reptilian and 1552-8618(1994)13[803:BATEOP]2.0.CO;2. Avian Chorioallantropic Membranes. (Master's Thesis) Texas Tech University. van der Oost, R., Beyer, J., Vermeulen, N.P.E., 2003. Fish bioaccumulation and R Core Team, 2015. R: A language and environment for statistical computing. R biomarkers in environmental risk assessment: a review. Environ. Toxicol. Pharmacol. Foundation for Statistical Computing, Vienna, Austria. URL 〈https://www.R-project. 13 (2), 57–149. http://dx.doi.org/10.1016/S1382-6689(02)00126-6. org/〉. Verbrugge, L.A., Giesy, J.P., Verbrugge, D.A., Woodin, B.R., Stegeman, J.J., 2001. Safe, S.H., 1986. Comparative toxicology and mechanism of action of polychlorinated Catalytic and immunochemical properties of hepatic cytochrome P450 1A in three dibenzo-pdioxins and dibenzofurans. Annu. Rev. Pharmacol. Toxicol. 26, 371–399. avian species treated with beta-naphthoflavone or isosafrole. Comp. Biochem. http://dx.doi.org/10.1146/annurev.pa.26.040186.002103. Physiol. C.-Toxicol. Pharmacology 130 (1), 67–83. http://dx.doi.org/10.1016/ Safe, S.H., 1994. Polychlorinated biphenyls (PCBs): environmental impact, biochemical S1532-0456(01)00221-6. and toxic responses, and implications for risk assessment. Crit. Rev. Toxicol. 24 (2), Walker, C.H., 1998. Avian forms of cytochrome P450. Comp. Biochem. Physiol. C.- 87–149. http://dx.doi.org/10.3109/10408449409049308. Pharmacol. Toxicol. Endocrinol. 121 (1–3), 65–72. Sanderson, J.T., Norstrom, R.J., Elliot, J.E., Hart, L.E., Cheng, K.M., Bellward, G.D., 1994. Walker, M.K., Heid, S.E., Smith, S.M., Swanson, H.I., 2000. Molecular characterization Biological effects of polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls and developmental expression of the Aryl hydrocarbon receptor from the chick in double-crested cormorant chicks (Phalacrocoraxauritus). J. Toxicol. Environ. embryo. Comp. Biochem. Physiol. C.-Pharmacol. Toxicol. Endocrinol. 126 (3), Health 41 (2), 247–265. 305–319. http://dx.doi.org/10.1016/S0742-8413(00)00119-5. Sarasquete, C., Segner, H., 2000. Cytochrome P4501A (CYP1A) in teleostean fishes. A Walker, C.H., Ronis, M.J.J., 1989. The monooxygenases of birds, reptiles, and review of immunohistochemical studies. Sci. Total Environ. 247 (2–3), 313–332. amphibians. Xenobiotica 19 (10), 1111–1121. http://dx.doi.org/10.1016/S0048-9697(99)00500-8. Watanabe, M.X., Iwata, H., Okamoto, M., Kim, E., Yoneda, K., Hashimoto, T., Tanabe, S., Schlezinger, Jennifer J., Struntz, William, D.J., Goldstone, Jared, V., Stegeman, John, J., 2005. Induction of cytochrome P450 1A5 mRNA, protein and enzymatic activities by 2006. Uncoupling of cytochrome P450 1A and stimulation of reactive oxygen species dioxin-Like compounds, and congener-specific metabolism and sequestration in the production by co-planar polychlorinated biphenyl congeners. Aquat. Toxicol. 77 (4), liver of wild jungle crow (Corvus macrorhynchos) from Tokyo, Japan. Toxicol. Sci. 88 422–432. http://dx.doi.org/10.1016/j.aquatox.2006.01.012. (2), 384–399. http://dx.doi.org/10.1093/toxsci/kfi326. Stagg, R.M., Rusin, J., McPhail, M.E., McIntosh, A.D., Moffat, C.F., Craft, J.A., 2000. Webb, S.J., Zychowski, G.V., Bauman, S.W., Higgins, B.M., Raudsepp, T., Gollahon, L.S., Effects of polycyclic aromatic hydrocarbons on expression of CYP1A in salmon (Salmo Wooten, K.J., Cole, J.M., Godard-Codding, C., 2014. Establishment, characterization, salar) following experimental exposure and after the Braer oil spill. Environ. Toxicol. and toxicological application of loggerhead sea turtle (Caretta caretta) primary skin Chem. 19 (11), 2797–2805. http://dx.doi.org/10.1897/1551-5028(2000) fibroblast cell cultures. Environ. Sci. Technol. 48 (24), 14728–14737. http://dx.doi. 019<2797:EOPAHO>2.0.CO;2. org/10.1021/es504182e. Stegeman, J., Brouwer, M., Di Gulio, R., Forlin, L., Fowler, B., Sanders, B., et al., 1992. Whitlock, J.P., 1999. Induction of cytochrome P4501A1. Annu. Rev. Pharmacol. Toxicol. Molecular responses to environmental contamination: enzyme and protein systems as 39, 103–125. http://dx.doi.org/10.1146/annurev.pharmtox.39.1.103. indicators of chemical exposure and effect. In: Huggett, R., Kimberly, R., Mehrle, P., Whyte, J., Jung, R., Schmitt, C., Tillitt, D., 2000. Ethoxyresorufin-O-deethylase (EROD) Bergman, H. (Eds.), Biomarkers Biochemical, Physiological, and Histological Markers activity in fish as a biomarker of chemical exposure. Crit. Rev. Toxicol. 30 (4), of Anthropogenic Stress. Lewis Publishers, Boca Raton, pp. 235–335. 347–570. http://dx.doi.org/10.1080/10408440091159239. Stegeman, J., Lech, J., 1991. Cytochrome P-450 monooxygenase systems in aquatic

75 Ecotoxicology and Environmental Safety 146 (2017) 76–82

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Dermal exposure to weathered MC252 crude oil results in MARK echocardiographically identifiable systolic myocardial dysfunction in double-crested cormorants (Phalacrocorax auritus) ⁎ K.E. Harra, ,1, M. Rishniwb, T.L. Ruppa, D. Cacelac, K.M. Deanc, B.S. Dorrd, K.C. Hanson-Dorrd, K. Healye, K. Horakf, J.E. Linkg, D. Reavillh, S.J. Bursiang, F.L. Cunninghamd a URIKA, LLC. Mukilteo, WA, USA b Veterinary Information Network, Davis, CA, USA c Abt Associates, Boulder, CO, USA d USDA/APHIS/Wildlife Services/National Wildlife Research Center, Mississippi Field Station Center, Mississippi State, MS, USA e US Fish and Wildlife Service, Deepwater Horizon NRDAR Field Office, Fairhope, AL, USA f USDA/APHIS/Wildlife Services/National Wildlife Research Center, Mississippi Field Station Center, Fort Collins, CO, USA g Michigan State University, East Lansing, MI, USA h Zoo/Exotic Pathology Service, Carmichael, CA, USA

ARTICLE INFO ABSTRACT

Keywords: During the Deepwater Horizon Natural Resource Damage Assessment, gross morphologic cardiac abnormalities, Arrhythmia including softer, more distensible musculature, were noted upon gross necropsy in hearts from laughing gulls Naphthalene and double-crested cormorants exposed to weathered MC252 crude oil. A species specific, echocardiographic Cardiomyopathy technique was developed for antemortem evaluation of function that was used to evaluate and better char- Petroleum crude oil acterize cardiac dysfunction. Control (n=12) and treated (n=13) cormorant groups of similar sex-ratio and ages Polycyclic aromatic hydrocarbons (PAH) were dermally treated with approximately 13 ml of water or weathered MC252 crude oil, respectively, every 3 Deepwater Horizon oil spill days for 6 dosages. This resulted in a low to moderate external exposure. Upon visualization and clinical as- sessment of the hearts of all test subjects, comprehensive diagnostic cardiographic measurements were taken twice, prior to oil application and after a 21 day dermal oil exposure. Oil-treated birds showed a decrease in cardiac systolic function, as characterized by an increased left ventricular internal dimension-systole and left ventricular stroke volume as well as concurrent decreased left ventricular ejection fraction and left ventricular fractional shortening when compared to both control birds’ and the treated birds’ time zero values. These changes are indicative of a possible dilative cardiomyopathy induced by oil exposure, although further eluci- dation of possible collagen damage is recommended. Arrhythmias including tachycardia in two treated birds and bradycardia in all treated birds were documented, indicating further clinically significant abnormalities induced by MC252 oil that warrant further investigation. A statistically significant increase in free calcium concentration, important to muscular and neurologic function in treated birds was also noted. This study documents that weathered MC252 oil caused clinically significant cardiac dysfunction that could result in mortality and decrease recruitment.

1. Introduction Assessment (NRDA) Trustees to characterize ecosystem damages that encompasses damage to animal health (Bursian et al., 2017; DWH The Deepwater Horizon (DWH) oil spill in 2010 released 3.2 million Trustees - Deepwater Horizon Natural Resource Damage), 2015). While barrels of crude oil into the northern Gulf of Mexico, exposing nu- the cardiotoxic effects of oil are well documented in fish, especially merous species of animals to the toxic components of oil. A compre- during embryonic development, there is minimal information of the hensive assessment of morbidity and mortality caused by DWH oil ex- cardiotoxic effects of crude oil in other vertebrates (Collier et al., 2014; posure was undertaken by the DWH Natural Resource Damage Incardona et al., 2014). In birds, anemia, disrupted feather function,

⁎ Corresponding author. E-mail address: [email protected] (K.E. Harr). 1 Postal address: URIKA, LLC, 8712 53rd Pl W, Mukilteo, WA 98275, USA. http://dx.doi.org/10.1016/j.ecoenv.2017.04.010 Received 6 September 2016; Received in revised form 4 April 2017; Accepted 5 April 2017 Available online 27 June 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 76–82 hypothermia, respiratory distress, seizures, diarrhea, hepatic disease and renal disease have all been reported secondary to exposure to petroleum products (Mazet et al., 2002). However, to the best of the authors’ knowledge, cardiac dysfunction has never been documented in birds post-petroleum exposure. Necropsies performed as part of a pre- liminary oral exposure study using weathered MC252 oil uncovered as yet undocumented gross morphological cardiac changes, consisting of softer, more distensible ventricular walls, in exposed double-crested cormorants (DCCOs) compared to control birds. This prompted further evaluation of live DCCOs for the presence of systolic myocardial dys- function. Therefore, we sought to evaluate cardiac function echo- cardiographically in DCCOs to determine if cardiac dysfunction could be identified in live DCCOs subjected to dermal oil exposure.

2. Materials and methods

2.1. Toxicant

Fig. 1. Echocardiographic view of left atrium (LA), left ventricle (LV), right ventricle MC252 (DWH7937, batch# B030112) oil was collected during the (RV), and aorta (Ao). 2010 Deepwater Horizon oil spill and artificially weathered by TDI- Brooks International, (College Station, TX) prior to use in the studies as allowing sufficient space (~1 cm) for open-mouth breathing. Birds were previously described (Forth et al., 2015). restrained in sternal position, supported at mid thorax holding the wings with the head well restrained. The keel remained free and was never 2.2. Animal husbandry compressed and feet were retracted from the coelomic window. The head was angled up with the tail pointing towards the floor so that the bird lay A total of 31 DCCO's were captured and retained in captivity under at approximately a 75° angle. Ultrasound gel and alcohol were used to part the authority of USFWS MBPO Federal Permit #MB019065-3, the feathers at the coelomic window caudal to the keel just to the right of fi Mississippi and Alabama state (#8017) scienti c collection permits, ventral midline. The probe was placed at the coelomic window and angled and Institutional Animal Care and Use Committee (IACUC) under towards the head or to the shoulder/thoracic vertebra to obtain the op- NWRC protocol QA-2326. Cunningham et al. (2017) in this issue pro- timal cardiac image. This positioning provided an apical 3-chamber or 4- vides a detailed description of animal capture and handling in March chamber view (Fig. 1). An axillary view was also used and images could be 2014. obtained; but, during the quarantine echocardiography, our cardiologist's Birds were allowed to acclimate to captivity for a minimum of 21 impressions were confirmed by repetitive quantitative measurements. The days prior to initiation of the study. A total of 25 subadult DCCOs al- coelomic window provided access that produced more precise repetitive located to a control group (n=12, 5 male, 7 female) and an exposed measurements. Therefore, the coelomic window was considered the pre- group (n=13, 6 males, 7 females) were used in this trial. DCCOs were ferable technique. assigned to treatment groups based on the results of blood samples All imaging was performed using a portable ultrasound unit with a collected at the initiation of the three-week quarantine period. 12-4 phased array probe (Model CX50, Philips, Andover, MA). Complete blood count (CBC) values were used to ensure equal division The echocardiographic evaluations followed a consistent protocol in- of birds with potential health concerns between groups. DCCO's with 9 cluding an apical three chamber view. First, the right atrium and tricuspid monocyte counts greater than 2.0 x 10 cells/l were considered ab- valve were examined using two-dimensional imaging. Color Doppler was normal (severe monocytosis); and were divided between control (n=4) then applied to interrogate the tricuspid valve for tricuspid regurgitation. and treatment (n=3) groups. Additionally, a small oil spill took place Next, left ventricular outflow and transvalvular aortic velocities were one year prior to the study, not far from where 6 of the DCCOs were obtained from the coelomic window with care to align with the left ven- collected and were evenly distributed between groups. During the tricular outflow tract so that optimal outflow velocities were obtained course of the trial, one bird from the control group and two birds from using spectral Doppler. Third, an apical three chamber image was ob- fi the treatment group died and were not replaced. Therefore, the nal tained, optimized for the left ventricle and left atrium and allowing clear number of birds in the control and exposed group was 11 birds each to visualization of the ascending aorta and aortic valve. Color Doppler was total 22 in the study. Oil on exposed birds (13 ml) and water on control applied to interrogate the mitral valve for mitral regurgitation. If regur- birds (13 ml) was applied every three days through Day 15 of the trial gitation of any valve was detected with color Doppler, spectral Doppler (on Days 0, 3, 6, 9, 12, and 15). Detailed description of application is was applied to confirm the observation. available in Cunningham et al. (2017). Following the live data collection, the video loops of 2-dimensional Plasma ionized calcium was measured using heparinized whole images were reviewed utilizing analysis software (Philips Xcelera blood by the potentiometric method using an iSTAT (Abaxis PACS). Dimensions and measurements obtained included: Diagnostics, Co, Union City, CA) at necropsy on days 23 and 24 of this study. I. heart rate (HR) II. interventricular septal dimension diastole (IVSd) 2.3. Echocardiography III. left ventricular internal dimension diastole (LVIDd) IV. left ventricular posterior wall dimension diastole (LVPWd) − During the quarantine period (Day = 3) and prior to oil application, V. interventricular septal dimension systole (IVSs) fi aboard-certi ed veterinary cardiologist (TLR) developed a method for VI. left ventricular internal dimension systole (LVIDs) fi fi species-speci c DCCO echocardiography modi ed for DCCO anatomy. VII. left ventricular posterior wall dimension systole (LVPWs) DCCO hearts were visualized in real time and chambers and wall thickness VIII. left atrial (LA) diameter were measured using a standardized, quantitative technique to evaluate IX. left atrial volume the four chambers of the heart during systole and diastole. Birds were X. aortic root diameter hooded using a towel with a band around bill, avoiding the nares, but

77 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 76–82

The following variables were calculated from the measurements as follows:

1) left ventricular fractional shortening ()LVIDd− LVIDs LVFS (%) = *100 LVIDd

2) left ventricular ejection fraction (LVEF) ()LVVol− LVVol LVEF (%) = ds*100 LVVold where

7 3 oLVVold = ⎛ ⎞*LVIDd ⎝ 2. 4+LVIDd ⎠ and

7 3 LVVols = ⎛ ⎞*LVIDs ⎝ 2. 4+LVIDs ⎠ Standard gross necropsy was performed 24 days after baseline echocardiograph measurements and a complete set of tissues was col- lected for histopathologic analysis by a boarded anatomic pathologist (DRR).

2.3.1. Statistical analysis Plots of echocardiographic variables were examined visually. Because of small sample sizes, data were analyzed non-parametrically. The post-exposure values of echocardiographic variables between ex- posed and unexposed DCCO were compared with Mann Whitney U Tests. Because of the small sample size and no prior evidence of cardiac injury in other species, we made no adjustment of the alpha values for experiment-wise error. All comparisons were considered significantly different at p < 0.05. Dot and box-and-whisker plots were used to display individual data for all endpoints. Calculations were performed using TIBCO Spotfire S-PLUS 8.2 for Windows and MedCalc version 14.12.0 (Ostend, Belgium).

3. Results

Birds XU11 (control), CU09 (exposed) and CU18 (exposed) died prior to Day 21 of the study. A chronic, necrotizing granuloma was found at the heart base of the control bird at necropsy. Upon complete necropsy, exposed bird (CU09) died with probable septicemia (under- lying etiologic agent not identified). Exposed bird (CU18), died with no significant lesions that could be assessed as a cause of death. All pretreatment values measured in the control and exposed DCCOs had no significant difference based on ANOVA and were similar at baseline based on visual inspection of the dot plots. At necropsy, gonad identification revealed 5 males and 7 females in the control group and 6 males and 7 females in the treated group. No developed ovaries were found in any of the females at necropsy indicating a subadult classifi- cation. After 21 day dermal exposure, DCCOs showed a decrease in cardiac systolic function, as characterized by an increased left ventricular in- ternal dimension in systole (LVIDs, P=0.02) and left ventricular sys- tolic volume (LVSV, P=0.02); consequently, left ventricular fractional shortening (LVFS, P=0.04) and left ventricular ejection fraction (LVEF, P=0.04) were decreased when compared to unexposed DCCO (Fig. 2a- Fig. 2. a. Pre-exposure and post- 21 day exposure left ventricular internal dimension in d, Table 1). After exposure, DCCOs showed an increase in inter- systole (LVIDs) in control and exposed DCCO. b. Pre-exposure and post-exposure left ventricular wall thickness in both diastole (IVSd, P=0.002) and systole ventricular fractional shortening (LVFS) in control and exposed DCCO. c. Pre-exposure (IVSs, P=0.0006), when compared to unexposed DCCOs. Conversely, and post-exposure interventricular septal thickness in diastole (IVSd) in control and ex- left ventricular internal dimension in diastole (LVIDd, P=0.1) and posed DCCO. cardiac weight (P=0.4) did not differ between control and exposed

78 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 76–82

Table 1 Summary statistics for measures cardiac variables in DCCO.

Measurement Treatment n median min max p value

Aortic root dimension (AoR) (cm) exposed 11 1.0 1.0 1.1 0.7 control 11 1.0 0.9 1.0

End diastolic volume (EDV 2D-Teich) (ml) exposed 11 6.2 4.2 8.2 0.1 control 11 4.9 3.3 9.6

Heart rate mean (BPM) exposed 11 140 80 263 0.8695 control 11 135 108 173

Interventricular septal dimension diastole (IVSd) (cm) exposed 11 0.38 0.36 0.43 0.002 control 11 0.36 0.36 0.38

Interventricular septal dimension systole (IVSs) (cm) exposed 11 0.49 0.47 0.55 0.0006 control 11 0.46 0.42 0.5

Left atrial area (cm2) exposed 11 3.1 2.4 3.5 0.2 control 11 2.6 2.1 3.5

Left atrial circumference (cm) exposed 11 6.4 5.3 6.9 0.2 control 11 5.9 5.6 6.8

Left atrial dimension (LA Dimen) (cm) exposed 11 1.4 1.3 1.6 0.02 control 11 1.5 1.3 1.8

Left ventricular ejection fraction (EF 2D-Teich) (%) exposed 11 62 56 70 0.04 control 11 69 57 81

Left ventricular fractional shortening (FS 2D-Teich) (%) exposed 11 30 27 36 0.04 control 11 36 27 46

Left ventricular internal dimension diastole (LVIDd) (cm) exposed 11 1.51 1.31 1.69 0.1 control 11 1.38 1.19 1.79

Left ventricular internal dimension systole (LVIDs) (cm) exposed 11 1.03 0.92 1.15 0.02 control 11 0.88 0.67 1.30

Left ventricular posterior wall dimension diastole (LVPWD) (cm) exposed 11 0.45 0.43 0.55 0.045 control 11 0.43 0.40 0.46

Maximum aortic valve outflow velocity (AV Vmax) (cm/sec) exposed 11 113 87 149 0.017 control 11 137 101 191

Systolic volume (ESV 2D-Teich) (ml) exposed 11 2.26 1.63 2.97 0.02 control 11 1.40 0.69 4.15

Heart weight (g) exposed 12 20 16 25 0.4 control 11 21 15 32

DCCOs. tachycardia was documented in two exposed DCCOs (Fig. 4b). Elec- Heart rates did not differ between exposed and unexposed DCCOs trocardiography was unavailable, so the exact nature of the tachyar- (P=0.7). However, exposed DCCOs had clinically significant ar- rhythmia could not be determined. rhythmia that included marked bradycardia, defined as < 70 BPM, Pericardial effusion was documented echocardiographically and (100%, 11/11 exposed DCCOs) and tachycardia, defined as > 200BPM confirmed at gross necropsy in two exposed DCCOs (Addendum Video 3 (18%, 2/11 exposed DCCOs) (Fig. 3). – Echocardiograph video of a heart from a treated bird exhibiting Control DCCOs had a sinus arrhythmia, which is typical in diving pericardial effusion. Control DCCOs did not have evidence of any excess animals (Fig. 4a). No other arrhythmia was found in control DCCOs. A fluid around the heart at necropsy. When microscopically evaluated, the fluid contained within the pericardial sac differed slightly between the two exposed DCCOs in that one was a low protein transudate while the second, classified as an exudate, had a component of suppurative inflammation without evidence of an infectious agent. Dyspnea (labored breathing) was also noted in exposed DCCOs but not controls. During the baseline evaluation, all DCCOs were able to breathe comfortably with a band placed around the beak to avoid the DCCOs biting the handlers. Comparison of the same population of DCCOs after oil exposure yielded notable differences between the control and exposed DCCOs. The oil-exposed DCCOs demonstrated se- vere dyspnea when handled. When the band was placed on the beak of several exposed DCCOs for short periods of time, the breathing was so labored that the safety of the DCCOs was threatened. We modified the beak restraint for the completion of the exams so that the DCCOs had their mouths open to breath at all times. All control DCCOs tolerated the beak band with mouths restrained with no breathing distress. Gross necropsy revealed several hearts from dermally exposed DCCOs appeared to have softer cardiac musculature than the hearts Fig. 3. Average heart rate of control and exposed DCCOs. No statistically significant from control DCCOs as in the oral exposure (Fig. 5a, b). difference was present as outliers in the treated group were on both ends of the dataset.

79 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 76–82

exhibited arrhythmias other than a sinus arrhythmia, including marked bradycardia in all exposed DCCOs and marked tachycardia in two ex- posed DCCOs. Additionally, a statistically significant increase in ionized calcium in the treated birds was present. In comparison to other avian species, the DCCOs have a larger keel and therefore a smaller coelomic viewing window where the ultrasound probe may be placed to view the cardiac silhouette. As the probe is more caudal, the angle to the head is more acute and more closely parallels the spine. Psittacines can be assessed in dorsal recumbency, but the preferred position to consistently view the cardiac silhouette in DCCOs was ventral recumbency. In psittacines, all four chambers of the heart can be seen in one window, but in DCCOs, the right atrium and ventricle were imaged with a slightly more cranial angle than the left atrium and left ventricle. To the authors’ knowledge, oil exposure has not been reported to cause cardiac dysfunction in birds, either in adults or developing life stages. The data in our study are strongly suggestive of cardiotoxicity and our findings mirror those observed in other species. Oil exposure has been shown to cause cardiac damage in a broad range of developing fish species including sole, herring, zebrafish, seabass and tuna (Brette et al., 2014; Claireaux and Davoodi, 2010; Marty et al., 2011; Jung et al., 2013; Tissier et al., 2015). The cardiac damage is theorized to occur by an impairment of the cardiac excitation-contraction coupling mechanism via blocking of the delayed rectified potassium current and a decrease in calcium current and calcium cycling (Brette et al., 2014). In this study, there was a statistically significant increase in plasma ionized calcium which warrants further investigation as a pathogenic mechanism of disease. Severe hypercalcemia may induce arrhythmia but this has not been shown to be causative in this study. All of the gonads in this study were assessed as immature at necropsy indicating that the bird's sex did not play a role in calcium values. In other studies, activation of the aryl hydrocarbon receptor is listed as the cause of ventricular remodeling and cardiac damage (Incardona et al., 2004). Similarly, developing birds exposed to dioxin-like compounds, which also activate the aryl hydrocarbon receptor, undergo changes in cardiac modeling (DeWitt et al., 2006; Kopf and Walker, 2009; Carro et al., 2013). Exposure to weathered MC252 oil and unweathered slick A oil collected during the Deepwater Horizon spill induced pericardial Fig. 4. – a. Example of typical heart rate in a control DCCO. Addendum 1 DCCO echo- edema, atrial arrhythmia, dose dependent bradycardia, and an inver- cardiograph video illustrating a normal heart from a control DCCO. This figure represents sion of the systolic and diastolic phases in isolated tuna ventricular 4 beats in 2 s or 120 beats per minute. b. Example of tachycardia found in two oil-exposed DCCOs. Addendum 2 – DCCO echocardiograph video illustrating tachycardia myocytes (Brette et al., 2014). It is interesting to note that very similar (> 200BPM). Note the disorganized electrical stimulation and contractions only found in cardiac pathology including arrhythmia, bradycardia, pericardial treated birds. edema/effusion, and decreased myocardial contractility are found in both cold and warm water species including tuna, amberjack, herring, Upon histopathologic examination, three hearts from exposed salmon, and zebra fish when exposed to either source or weathered DCCOs had myocardial fibrosis, while no myocardial fibrosis was found Louisiana sweet crude oil or source or weathered Exxon Valdez crude in the control DCCOs. Other histologic lesions, including septicemia and oil or Iranian heavy crude. While there is some variability in the pro- inflammation, were considered to be representative of background minence of effect between species and possibly between forms of oil, disease in the population. Special stain for collagen have not been the effects themselves are consistent. It is also striking that not only completed at the time of this report (Table 2). does this crude oil cause cardiac abnormalities in delicate developing life stages but did produce clinically significant changes in adult birds after only 3 weeks of moderate exposure. 4. Discussion Similar findings have also been found in mammals exposed to crude oil. Flaccid hearts were observed upon gross examination and edema In this study, a reproducible echocardiographic technique to eval- and coagulative necrosis of cardiac myofibers were found on histo- uate cardiac morphology and function was developed in DCCOs. The pathologic examination of sheep necropsied after crude oil exposure in DCCOs dermally exposed to weathered MC252 oil had echocardio- Brazil (Batista et al., 2013). Mostrom and Campbell reported enlarged graphic findings indicating decreased ventricular myocardial con- and “flabby” hearts upon gross necropsy of cattle exposed to sour tractility, as evidenced by increased ventricular dimensions during multiphase crude petroleum mixed with river water and soil during a systole; but no change in atrial dimensions was detected in this low pipeline spill and then burned during cleanup in the Red Deer River, sample number and low statistical power, short term study. This was Alberta, Canada. It should be noted that this crude petroleum (sour gas consistent with necropsy findings of soft, flaccid, and enlarged hearts and sour gas condensate) contains hydrogen sulfide and has a higher although the weight of the hearts of control and treated birds were not percentage of volatiles than the weathered MC252 oil in our study. statistically different and myocardial fibrosis was observed in only some Upon histopathologic evaluation, exposed cattle were found to have hearts from exposed DCCOs. Further study is needed to confirm possible hemorrhage at valves and on epicardium and endocardium. Hemor- dilative cardiomyopathy postulated by clinicians. Only exposed DCCOs rhagic lesions throughout the body were reported by the anatomic

80 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 76–82

Fig. 5. a. Heart from dermally exposed DCCO. b. Heart from a normal control DCCO.

pathologist as being consistent with Salmonella infection, however, volatile components as well as polycyclic aromatic hydrocarbons Salmonella was not cultured from any cattle (Mostrom and Campbell, (PAHs) found in oil and oil byproducts. The tissue concentrations of 1994). The generalized hemorrhagic lesions found in many tissues in- these components in an ingested complex mixture required to produce cluding the heart in the exposed cattle indicated that an induced coa- arrhythmias is unknown (Bradley et al., 2013). gulopathy was possible, similar to that which was documented in the Many of the piscine studies suggested that humans might also suffer cormorants (Harr et al., 2017a). Depressed or rapid, weak pulses were cardiac myocyte damage secondary to petroleum exposure however noted clinically in cattle exposed to crude oil and upon complete ne- this has not been well-documented in studies of people exposed to cropsy fibrinous pericarditis was noted in occasional exposed animals crude oil products (D’andrea and Reddy, 2014; Goldstein et al., 2011; only. No lesions were noted in the heart itself upon histopathology and Solomon and Janssen, 2010). In humans, diesel exhaust particles, heart rate/pulse were the only clinical measurement of the heart as- which also contain concentrated PAHs similar to weathered MC252 oil, sessed in this study (Bystrom, 1989). This is consistent with the findings increases the risk of significant arrhythmias and cardiac failure in urban in the DCCOs in this study where average heart rates did not differ residents (Brook et al., 2010). In Sprague Dawley rats, diesel exhaust between exposed and control DCCOs, but clinically significant, and particles resulted in eccentric left ventricular dilation with systolic possibly life-threatening arrhythmias, including bradycardia (n=11/ dysfunction similar to that which we found in the DCCOs (Bradley 11) and tachycardia (n=2/11), were noted only in exposed DCCOs et al., 2013). Bradley et al. (2013) postulated that induction of aryl upon echocardiographic evaluation. Cardiac arrhythmias in humans hydrocarbon reductase induces left ventricular dilation due to loss of have been observed in cases of oil exposure and are attributed to the collagen. Further histopathologic evaluation of the cormorant heart

Table 2 Histopathologic findings in cormorants dermally exposed to weathered MC252 oil. Harr et al. (2017b), this edition.

Tissue Treatment Lesion description Lesion distribution Grade Lesion grade Animals affected

Heart Control Epicardial hemorrhage focal minimal 1 1/11 Myocardial hemorrhage multifocal mild 2 1/11 Myocarditis, lymphoplasmacytic focal minimal 2 1/11 Externally oiled Bacterial granulomas multifocal moderate 3 1/11 Myocardial fibrosis multifocal mild to moderate 2.3 3/11 Septic and suppurative thrombi focal mild 2 1/11

81 K.E. Harr et al. Ecotoxicology and Environmental Safety 146 (2017) 76–82 samples using special stains for collagen are warranted to assess this matter air pollution and cardiovascular disease an update to the scientific statement from the American Heart Association. Circulation 121 (21), 2331–2378(Retrieved potential pathogenic mechanism of disease. from: 〈http://circ.ahajournals.org/content/121/21/2331.long/feed〉). The pathogenesis of the cardiotoxicity may also be by direct oxi- Bursian, S., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Ellis, C., Guglielmo, dative damage to the cardiomyocytes causing decreased cellular func- C.G., Hanson-Dorr, K.C., Harr, K.E., Healy, K.A., Hooper, M.J., Horak, K.E., Isanhart, J., Kennedy, L., Link, J.E., Maggini, I., Moye, J., Perez, C.R., Pritsos, C.A., Shriner, S., tion. Studies of specific components of crude oil have shown that sig- Trust, K., Tuttle, P., 2017. Overview of avian toxicity studies for the Deepwater nificant concentrations can be found in cardiac tissue, although this is Horizon Natural Resource Damage Assessment. Ecotoxicol. Environ. Saf (in press) fi ff (Citation to be nalized during editorial review). time-dependent. The chronic e ects of oxidant damage caused by these Bystrom, J.M., 1989. Study of the acute toxicity of ingested crude petroleum oil to cattle. components is not well studied, though mitochondrial damage in car- Masters of Science Thesis accepted by University of Saskatchewan, Saskatoon, diac myocytes has been postulated. Aromatics and 14C-naphthalene are Canada. Carro, T., Dean, K., Ottinger, M.A., 2013. Effects of an environmentally relevants poly- detectable in cardiac tissue at 72 h (the last time point of the study) chlorinated biphenyl (PCB) mixture on embryonic survival and cardiac development after a single administration of aromatic hydrocarbons in hens and pigs in the domestic chicken. Environ. Toxicol. Chem. 32 (6), 1317–1324. Claireaux, G., Davoodi, F., 2010. Effect of exposure to petroleum hydrocarbons upon (Eisele et al., 1985). Aromatics have also been found in cardiac tissue cardio-respiratory function in the common sole (Solea solea). Aquat. Toxicol. 98 (2), from ducks dosed singly and multiple times with sweet Louisiana crude 113–119(Retrieved from: 〈http://www.sciencedirect.com/science/article/pii/ oil (Gay et al., 1980). It was beyond the scope of this study to assess S0166445X10000469〉). Collier, T.K., Anulacion, B.F., Arkoosh, M.R., Dietrich, J.P., Incardona, J.P., Johnson, L.L., PAH concentrations in all tissues. Ylitalo, G.M., Myers, M.S., 2014. Effects on Fish of Polycyclic Aromatic Hydrocarbons Other pathologies documented at necropsy in other organ systems (PAHS) and Naphthenic Acid Exposures. In: Tierney, C.J., Farrell, K.B., Brauner, A.P. including the kidney, liver, and hematopoietic tissues likely com- (Eds.), Organic Chemical Toxicology of Fishes: Fish Physiology Vol. 33. Elsevier Inc, London, pp. 195–255. http://dx.doi.org/10.1016/B978-0-12-398254-4.00004-2. pounded direct cellular damage as well as aryl hydrocarbon receptor Cunningham, F., Dean, K., Hanson-Dorr, K., Harr, K.E., Healy, K., Horak, K., Shriner, S., binding to contribute to cardiac dysfunction. The treated DCCO in this Link, J., Bursian, S., Dorr, B., 2017. Development of Methods for the study of Avian ff Oil Toxicity Studies using the Double Crested Cormorant (Phalacrocorax auritus). study were found to be su ering from hemolytic anemia resulting from Ecotoxicol. Environ. Saf (in press) (Citation to be finalized during editorial review.). oxidative damage to the red blood cells (Harr et al., 2017a) and were D’andrea, M.A., Reddy, G.K., 2014. Hematological and hepatic alterations in nonsmoking fl likely to be under thermoregulatory stress due to feather disruption and residents exposed to benzene following a aring incident at the British Petroleum plant in Texas City. Environ. Health 13, 115–132. http://dx.doi.org/10.1186/1476- feather plucking (Cunningham et al., 2017). Therefore, stresses on the 069X-13-115. heart prior to necropsy were multifactorial and could have synergisti- DeWitt, J.C., Millsap, D.S., Yeager, R.L., Heise, S.S., Sparks, D.W., Henshel, D.S., 2006. External heart deformities in passerine birds exposed to environmental mixtures of cally resulted in decompensation. polychlorinated biphenyls during development. Environ. Toxicol. Chem. 25 (2), In summary, external exposure of double-crested cormorants to ar- 541–551. tificially weathered MC252 crude oil collected during the Deepwater DWH Trustees (Deepwater Horizon Natural Resource Damage), 2015. Deepwater Horizon oil spill draft programmatic damage assessment and restoration plan and draft pro- Horizon oil spill resulted in cardiac damage and dysfunction manifested grammatic environmental impact statement. 〈Http://www.gulfspillrestoration.noaa. as decreased ventricular contractility and arrhythmia, including both gov/restoration-planning/gulf-plan〉, pp. 1–685 (accessed 1 March 2017). bradycardia and tachycardia. To our knowledge, this is the first report Eisele, G.R., Tsai, S.C., Payne, J., Kyle, B., Urso, I., 1985. The distribution of orally ad- ministered C-Aniline HCl in tissues of dairy cattle, swine and laying pullets. J. Anim. of oil-induced cardiac damage in oil-exposed birds. Dilative cardio- Sci. 61 (6), 1492–1497. http://dx.doi.org/10.2527/JAS1985.6161492X. myopathy was postulated by clinicians and supported by echocardio- Forth, H.P., Mitchelmore, C.L., Morris, J.M., Lipton, J., 2015. Characterization of oil and fi water accommodated fractions used to conduct aquatic toxicity testing in support of graphy but further investigation is warranted to con rm this diagnosis. the Deepwater Horizon oil spill natural resource damage assessment. Environ. Further investigation to define the type of arrhythmia is warranted to Toxicol. Chem. 36 (6), 1450–1459. http://dx.doi.org/10.1002/etc.3672. (Citation to understand the pathogenesis of oil-induced cardiac damage. The find- be editorial review) (Epub 2016 Dec 9). Gay, M.L., Belisle, A.A., Patton, J.F., 1980. Quantification of petroleum-type hydro- ings are similar to those reported in humans, rats, and other mamma- carbons in avian tissue. J. Chromatogr. A 187 (1), 153–160. http://dx.doi.org/10. lian species typically attributed to the PAH component of oil. The 1016/S0021-9673(00)87881-4. changes in cardiac function reported here could impact acute mortality, Goldstein, B.D., Osofsky, H.J., Lichtveld, M.Y., 2011. The Gulf Oil Spill. N. Engl. J. Med. 364 (14), 1334–1348. http://dx.doi.org/10.1056/NEJMra1007197. long-term survivability, and recruitment of individuals and populations Harr, K.E., Cunningham, F.L., Pritsos, C., Pritsos, K., Muthumalage, T., Dorr, B., Horak, K., of birds. Hanson-Dorr, K., Dean, K., Cacela, D., Link, J.E., Healy, K., Tuttle, P., Bursian, S.J., 2017a. Weathered MC252 crude oil-induced anemia and abnormal erythroid mor- phology in double-crested cormorants (Phalacrocorax auritus) with light microscopic Addendum and ultrastructural description of Heinz bodies. Ecotoxicol. Environ. Saf (in Press) (Citation to be finalized during editorial review.). Harr, K.E., Bursian, S.J., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Hanson- Three videos are listed in the text. Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Reavill, D.R., Shriner, S.A., Schmidt, R.E., 2017b. Comparison of organ weights and histopathology between double- Acknowledgements crested cormorants (Phalacrocorax auritus) dosed orally or externally with artificially weathered Mississippi Canyon 252 crude oil. Ecotoxicol. Environ. Saf (in press) (Citation to be finalized during editorial review.). Funding: This work was supported by the United States Fish and Incardona, J.P., Collier, T.K., Scholz, N.L., 2004. Defects in cardiac function precede morphological abnormalities in fish embryos exposed to polycyclic aromatic hydro- Wildlife Service Deepwater Horizon Natural Resource Damage carbons. Toxicol. Appl. Pharmacol. 196 (2), 191–205. http://dx.doi.org/10.1016/j. Assessment (#F12PD01069). taap.2003.11.026. Incardona, J.P., Gardner, L.D., Linbo, T.L., Brown, T.L., Esbaugh, A.J., Stieglitz, J.D., French, B.L., Labenia, J.S., Laetz, C.A., Mager, E.M., 2014. Deepwater Horizon crude Appendix A. Supplementary material oil impacts the developing hearts of large predatory pelagic fish. Proc. Natl. Acad. Sci. 111 (15), E1510–E1518. http://dx.doi.org/10.1073/pnas.1320950111. Kopf, P.G., Walker, M.K., 2009. Overview of developmental heart defects by dioxins, PCBs Supplementary data associated with this article can be found in the and pesticides. J. Environ. Health Sci. C: Environ. Carcinog. Ecoloxicol. Rev. 27 (4), online version at http://dx.doi.org/10.1016/j.ecoenv.2017.04.010. 276–285. Mazet, J.A.K., Newman, S.H., Kirsten, V.K., Tseng, F.S., Holcomb, J.B., Jessup, D.A., References Ziccardi, M.H., 2002. Advances in oiled bird emergency medicine and management. J. Avian Med. Surg. 16 (2), 146–149. Mostrom, M.S., Campbell, C., 1994. Livestock Field Investigations of Two Ranches Associated with a Pipeline Break. Alberta Research Council; Alberta Agriculture, Batista, J.S., Câmara, A.C.L., Almeida, R.D., Olinda, R.G., Silva, T.M.F., 2013. Poisoning Food and Rural Development, Edmonton. by crude oil in sheep and goats. Rev. Med. Vet. 11, 217–520. Solomon, G.M., Janssen, S., 2010. Health effects of the Gulf Oil Spill. JAMA 304 (10), Bradley, J.M., Cryar, K.A., Hajj, M.C., Hajj, E.C., Gardner, J.D., 2013. Exposure to diesel 1118–1119. http://dx.doi.org/10.1001/jama.2010.1254. exhaust particulates induces cardiac dysfunction and remodeling. J. Appl. Physiol. Tissier, F., Dussauze, M., Lefloch, N., Theron, M., Lemaire, P., Le Floch, S., Pichavant- 115, 1099–1106. Rafini, K., 2015. Effect of dispersed crude oil on cardiac function in seabass Brette, F., Machado, B., Cros, C., Incardona, J.P., Scholz, N.L., Block, B.A., 2014. Crude oil Dicentrarchus labrax. Chemosphere 134, 192–198. http://dx.doi.org/10.1016/j. impairs excitation-contraction coupling in fish. Science 343 (February), 772–776. chemosphere.2015.04.026. Brook, R.D., Rajagopalan, S., Pope, C.A., Brook, J.R., Bhatnagar, A., Diez-Roux, A.V., Holguin, F., Hong, Y., Luepker, R.V., Mittleman, M.A., Peters, A., 2010. Particulate

82 Ecotoxicology and Environmental Safety 146 (2017) 83–90

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Toxic effects of orally ingested oil from the Deepwater Horizon spill on MARK laughing gulls ⁎ K.E. Horaka, , S.J. Bursianb, C.K. Ellisa, K.M. Deanc, J.E. Linkb, K.C. Hanson-Dorrd, F.L. Cunninghamd, K.E. Harre, C.A. Pritsosf, K.L. Pritsosf, K.A. Healyg, D. Cacelac, S.A. Shrinera a USDA APHIS National Wildlife Research Center, Fort Collins, CO, United States b Department of Animal Science, Michigan State University, East Lansing, MI, United States c Abt Associates, Boulder, CO, United States d USDA APHIS National Wildlife Research Center, Mississippi Field Station, Mississippi State University, Starkville, MS, United States e Urika Pathology, LLC, Mukilteo, WA, United States f Department of Agriculture, Nutrition, and Veterinary Science, University of Nevada, Reno, Reno, NV, United States g US Fish and Wildlife Service, Deepwater Horizon Natural Resource Damage Assessment and Restoration Office, Fairhope, AL, United States

ARTICLE INFO ABSTRACT

Keywords: The explosion of the Deepwater Horizon oil rig released millions of gallons of oil into the environment, sub- Oil toxicity sequently exposing wildlife, including numerous bird species. To determine the effects of MC252 oil to species Deepwater Horizon relevant to the Gulf of Mexico, studies were done examining multiple exposure scenarios and doses. In this study, MC252 oil laughing gulls (Leucophaeus atricilla, LAGU) were offered fish injected with MC252 oil at target doses of 5 or Laughing gull 10 mL/kg bw per day. Dosing continued for 27 days. Of the adult, mixed-sex LAGUs used in the present study, 10 Oxidative stress of 20 oil exposed LAGUs survived to the end of the study; a total of 10 of the oil exposed LAGUs died or were Gulf of Mexico euthanized within 20 days of initiation of the study. Endpoints associated with oxidative stress, hepatic total glutathione (tGSH), oxidized glutathione (GSSG) and reduced glutathione (rGSH) significantly increased as mean dose of oil increased, while the rGSH:GSSG ratio showed a non-significant negative trend with oil dose. A significant increase in 3-methyl histidine was found in oil exposed birds when compared to controls indicative of muscle wastage and may have been associated with the gross observation of diminished structural integrity in cardiac tissue. Consistent with previous oil dosing studies in birds, significant changes in liver, spleen, and kidney weight when normalized to body weight were observed. These studies indicate that mortality in response to oil dosing is relatively common and the mortality exhibited by the gulls is consistent with previous studies examining oil toxicity. Whether survival effects in the gull study were associated with weight loss, physiologic effects of oil toxicity, or a behavioral response that led the birds to reject the dosed fish is unknown.

1. Introduction fish, seasonal berries, and garbage (Burger, 1988). LAGUs are com- monly found on shores, parking lots and landfills, foraging on the To assess the specific adverse effects of MC252 oil on Gulf of Mexico ground, and in shallow waters. They are classified by the International relevant avian species, we designed a series of avian toxicity studies Union for Conservation of Nature (IUCN) as a species of least concern using different routes of exposure and dosages over time. Background following population increases and a halt to the collection of their eggs on the oil spill and an overview of the avian toxicity studies are outlined for food. Their abundance and flexible diet make the species a useful in Bursian et al. (2017). Laughing gulls (Leucophaeus atricilla; LAGUs) potential model for studying the effects of DWH oil across a broad range were chosen specifically for inclusion in these tests because they were of species. impacted by the Deepwater Horizon (DWH) spill, are easily caught and Prior to the study described herein, a pilot study was done to ex- managed in captivity, and are large enough to supply needed blood and amine multiple dosing methods and doses to determine the effects of tissues for analysis. The LAGU is a small black-headed gull that com- DWH oil on LAGU. Initial studies were based on previous work by monly nests in large groups of up to 50,000. As an omnivore and sca- Leighton (1986). These studies used oral gavage of oil:fish slurries as a venger, its diet consists of both terrestrial and aquatic invertebrates, method of dosing. This dosing method allowed for very accurate

⁎ Correspondence to: National Wildlife Research Center, 4101 LaPorte Avenue, Fort Collins, CO 80521, United States. E-mail address: [email protected] (K.E. Horak). http://dx.doi.org/10.1016/j.ecoenv.2017.07.018 Received 18 August 2016; Received in revised form 5 July 2017; Accepted 7 July 2017 Available online 16 August 2017 0147-6513/ Published by Elsevier Inc. K.E. Horak et al. Ecotoxicology and Environmental Safety 146 (2017) 83–90 measurements of dose. However, rapid transit time through the gastro- the golden shiner minnows were replaced by larger (8-10 cm) brooder intestinal (GI) tract may have limited absorption; therefore a dosing minnows that were injected with 400 µL oil (in the swim bladder) per method based on delivery of the DWH oil via food fish was developed to fish on day 5. Dosing continued for 27 days. more closely represent natural routes of oil ingestion encountered in the wild and provide a realistic exposure scenario. 2.3. Blood sampling

2. Methods Blood samples were collected during quarantine, on day 0, and twice weekly until the conclusion of the study. Blood smears were All experiments were approved by the Institutional Animal Care and prepared for white blood cell and red blood cell (RBC) assessment. A Use Committee of the United States Department of Agriculture (USDA), 20 µL aliquot of fresh whole blood was fixed for identification of re- Animal and Plant Health Inspection Service (APHIS), Wildlife Services, ticulocytes and Heinz bodies by transmission electron microscopy. All National Wildlife Research Center (NWRC, Approval 2108), Fort blood samples were collected early in the morning prior to providing Collins, CO, USA. morning food rations. Once a week during the oral dosing study, approximately 300 µL of 2.1. Animal collection and husbandry additional plasma was collected for determination of clinical chemis- tries. If additional blood was available, larger volumes of blood were Thirty-five resident LAGUs were captured in the Gulf of Mexico in centrifuged so that a 40 µL sample of heparinized plasma and a 300 µL April 2013 and transported to the USDA National Wildlife Research sample of packed RBCs could be collected for total antioxidant capacity Center animal care facility where they were quarantined and in- measurements. The targeted collection volumes were 1.0 mL and dividually housed in pens. Pens were one of two sizes, either 2.4 m × 0.40 mL for the first and second weekly draws, respectively, to allow for 1.2 m × 1.8 m (length × width × height) or 2.1 m × 2.1 m × 2.4 m. preparation of blood smears, determination of clinical chemistries, and Birds were randomly assigned to cage types so that study groups measurement of antioxidant endpoints. Clinical chemistries and white were evenly distributed in the two cage types. Cages were constructed blood cell types examined are outlined in Dean et al. (2017). of coated wire mesh and were set up to face each other to reduce iso- lation stress. 2.4. Necropsy Upon arrival, each bird was examined for any sign of injury or disease. All pens contained a food bowl and a shallow plastic tank filled Moribund birds (Toth, 2000) were euthanized and necropsied be- with water, which was changed daily. Dri-deck, carpet, or a grate was fore the end of the study to provide fresh tissues adequate for analysis; provided in a portion of each pen for foot relief from concrete floors. In half of the remaining birds were necropsied on day 27 and half were addition, each pen contained a tree stump for perching and loafing. necropsied on day 28 with half of the birds in each treatment group Birds were maintained on a 12L:12D light cycle with a mean room necropsied on each day. At necropsy, each bird was weighed and blood temperature of approximately 22 °C. Bird condition was monitored was collected as described above. Birds were euthanized by cervical daily for overt changes in body temperature, behavior and food intake. dislocation and additional blood was collected by cardiac puncture. Twice weekly, body weight (bw) and cloacal body temperature were Plasma was aliquoted to determine 3-methyl histidine concentration. measured manually. The brain, heart, kidneys, liver, GI tract, spleen, thyroid, and adrenal glands were collected and weighed to the nearest 0.1 mg. If a 2.2. Feeding and dosing bird appeared to be in breeding condition, the gonads were collected and weighed. In addition, all organs were assessed for gross abnorm- LAGUs were fed approximately 75-100 g fresh fish and 25 g Mazuri alities and if present, were documented with digital photographs. Fish Analog 50/10 Frozen 5T8L (Purina Mills, St. Louis, Missouri) per The brain, heart, lungs, spleen, bursa, thymus, gonads, one adrenal bird per day. Food fish were obtained from the Mississippi Field Station gland, and one thyroid gland were placed in 10% neutral buffered (pogies, Brevooritia patronus) or purchased from I.F. Anderson Farms formalin for histopathology. The liver was sectioned into six pieces. (Lonoke, Arkansas, golden shiner minnows, Notemigonus crysoleucaso). Sections for assessment of cytochrome P450 (CYP) activity, oxidative LAGUs were randomly assigned to one of three treatment groups: 1) a damage and PAH concentration were flash frozen in liquid nitrogen, control group (n = 10) that was provided untreated minnows, 2) a with the remaining portion placed in 10% neutral buffered formalin. group dosed daily with up to 5 mL oil/kg of the bird's body weight (bw) The kidney that was not preserved for histopathology was cut into five through provision of oil-containing fish as described below (n = 10), sections for assessment of oxidative damage. The GI tract was then and 3) a group dosed daily with up to 10 mL oil/kg bw (n = 10). sectioned into four pieces (esophagus to glandular stomach, section of Frozen minnows were thawed and each minnow was injected with duodenum containing pancreas, colon from the level of the ceca to the approximately 200 µL or 400 µL of artificially weathered MC252 oil cloaca, and a 2 g portion of the duodenum). The sections were slit using a 21- or 22-gauge needle and 1 mL-syringe. Dosing of LAGU was vertically and rinsed thoroughly in phosphate buffered saline. The first accomplished by providing multiple fish with a total dose of MC252 oil three sections were placed in neutral buffered formalin. (DWH7937, batch# B030112) equal to approximately 5 or 10 mL/kg bw per day. The oil-containing minnows (and unoiled minnows for 2.5. Histopathology control birds) were offered to the LAGUs in the morning and the rest of their daily ration (pogies, minnows, and Mazuri diet) was provided All tissue samples obtained at necropsy were preserved in 10% after all oil-containing fish had been consumed (control birds received neutral-buffered formalin solution prior to being routinely processed. additional rations at the same time the majority of dosed birds received Paraffin-embedded tissues were sectioned at approximately 5 µm, additional rations). The proportion of minnows, pogies, and Mazuri diet mounted on glass slides, and stained with hematoxylin and eosin. All offered was standardized between the three groups. The total food ra- slides were examined by a board certified veterinary pathologist (Zoo/ tion was approximately 100-125 g of food/day. Oil consumption was Exotic Pathology Service, Carmichael, CA). calculated based on the number of fish that were consumed. At the outset of the experiment, small golden shiner minnows (ap- 2.6. Antioxidant capacity proximately 5-6 cm long and weighing 3-4 g) were injected with 200 µL of oil and offered to the LAGUs. However, birds frequently rejected Samples were sent to Dr. Chris Pritsos at the University of Nevada these fish. Assuming that larger fish would make the oil less detectable, Agricultural Experiment Station for analysis of markers of antioxidant

84 K.E. Horak et al. Ecotoxicology and Environmental Safety 146 (2017) 83–90

Table 1 oil consumption and food consumption may have been a factor for Target doses, actual mean doses consumed, and the number of days each bird survived those birds. Finally, because many birds did not consistently consume the 28 day trial. Percent weight changes for each bird are across the number of days it all of the oil-injected fish offered, mean oil exposure did not con- survived. sistently meet our target doses of 5 or 10 mL/kg bw per day. Therefore, Bird Target dose Mean dose Days alive % Weight Proportion of days actual mean doses consumed by each individual were calculated per kg body change bird received full (Table 1) and were used in the analyses in place of the target doses. weight food rations Hematologic and plasma clinical chemistry endpoints were com- ff 526a 5 3.16 9 -25.6 0.33 pared using linear mixed e ects regression models, where the mean 530a 5 1.51 11 -12.7 0.18 dose and the sample day were modeled as a fixed effect and individual 518a 5 2.50 14 -33.8 0.36 birds were modeled as a random effect (modeled as a random slope) for a 514 5 2.80 18 -14.6 0.50 endpoints measured multiple times. Models included day, treatment 516a 5 3.97 20 -16.1 0.65 (mean oil dose as mL/kg bw/day), and a treatment*day interaction 517 5 3.64 28 -9.9 0.70 501 5 4.74 28 10.0 0.78 term. Sample day and treatment were modeled as continuous variables, 531 5 4.83 28 3.9 0.81 where treatment was defined as the average daily consumption de- 505 5 4.88 28 2.3 0.85 termined from daily observations of actual oil consumed by each in- 523 5 4.94 28 -2.3 0.78 dividual bird (Table 1). Oxidative stress endpoints, 3-methyl histidine, 519a 10 2.80 7 -28.1 0.29 532a 10 1.15 11 -10.2 0.09 and relative organ weights (i.e., organ weight as a percentage of body 520a 10 4.08 14 -8.5 0.21 weight) were compared using linear models where each endpoint was 529a 10 6.56 18 2.0 0.50 modeled as a function of the mean dose ingested. Statements of sig- 500 10 7.91 28 -2.1 0.56 nificance are on based individual models and p-values < 0.05. 513 10 8.61 28 -1.0 0.59 Because the reference laboratory noted issues with plasma sample 502 10 8.94 28 -3.4 0.81 522 10 9.35 28 -8.0 0.78 quality for some endpoints, they provided a quantitation of lipolysis 533 10 9.49 28 -0.9 0.67 and hemolysis in the samples. As random error has been demonstrated 527 10 9.52 28 -7.8 0.93 as a potential effect of these characteristics, plasma samples with he- 503 0 0.00 28 8.2 1.00 molysis scores of 2 or 3 (moderate or severe) were not included in the 504 0 0.00 28 3.0 1.00 507 0 0.00 28 6.5 1.00 statistical analysis and samples with lipemia scores of 2 or 3 (moderate 509 0 0.00 28 1.9 1.00 or severe) were excluded in the statistical analysis of selected analytes 511 0 0.00 28 -4.9 1.00 (CLSI [2012] has a discussion of interference resulting from hemolysis 512 0 0.00 28 -6.2 1.00 or lipemia). 524 0 0.00 28 -2.0 1.00 All analyses were conducted using R, version 3.2.2 (R Development 525 0 0.00 28 -11.1 1.00 528 0 0.00 28 -7.8 1.00 Core Team, 2008). Regressions were run using the lme4 package (Bates 534 0 0.00 28 -8.4 1.00 et al., 2015) and the survival plot was generated using the survminer package. a Individual did not survive through day 28. capacity (see Pristos et al., 2017, in press, for methods). 3. Results

3.1. Food consumption and oil intake 2.7. Statistical analyses Daily food consumption for oil-dosed birds was somewhat depressed Only birds that survived to the end of the study were used for re- during the first week of the study prior to the switch to larger minnows gression analyses because many endpoints were only measured at ne- (Fig. 1). Food consumption (g/kg bw) was similar among groups for the cropsy. Moreover, many of the birds that did not survive to the end of birds that survived to the end of the trial (day 28, Fig. 1) with regres- the trial had reduced food consumption such that confounding between sion analysis indicating no significant differences in consumption. For

Fig. 1. Daily cumulative food consumption for laughing gulls orally exposed to artificially weath- ered MC252 crude oil. Note: many of the birds that did not survive through the end of the trial experi- enced decreased food consumption prior or death or euthanasia.

85 K.E. Horak et al. Ecotoxicology and Environmental Safety 146 (2017) 83–90 birds that did not survive until the end of the trial, individuals often receive additional rations not containing oil (Table 1). For the 5 mL/kg refused fish and food consumption was often limited for several days bw birds, birds that died early only received full rations an average of prior to death or euthanasia (Fig. 1). 40.4% of the time compared with 78.4% of the time for birds that Birds were only offered their full ration after they consumed their survived for the full 28 days. As a consequence, the birds that died early oil-injected fish for their target dose of 5 or 10 mL/kg bw. Therefore, if lost an average of 20.6% of their body weight compared to a weight birds failed to eat all oiled fish they did not get a full ration of food that gain of 0.8% in the birds that survived until the end of the testing day. Animals that did not consistently get full rations of food had lower period. Similarly, for the 10 mL/kg bw birds, the 4 individuals that died survival (as determined by the number of days alive) compared with the early only got full food rations an average of 27.3% of the time and lost birds that often received full rations (Table 1). For the birds in the 11.2% of their body weight compared to birds that survived receiving 5 mL/kg bw group, 5 of 10 died or were euthanized prior to the end of full food rations 72.3% of the time and losing an average of 3.9% of the study (day 28). All of these birds had reduced food consumption their body weight. and lost weight (12.7–33.8%) prior to death or euthanasia (Table 1). The mean daily dose of oil consumed by these birds ranged from 1.51 to 3.3. Gross observation and histopathology 3.97 mL/kg bw whereas mean daily dose of oil consumed by birds that survived until the end of the trial ranged from 3.64 to 4.88 mL/kg bw Dose of oil consumed was only correlated with liver, spleen, and (Table 1). kidney organ weights when normalized to body weight (i.e., relative Four of the 10 birds in the 10 mL/kg bw group died prior to the end weight). Relative liver and relative kidney weights both significantly of the study. These birds consumed their full food ration for fewer days increased with oil dose (p < 0.001, p = 0.010 respectively, Fig. 3) (mean = 27.25% of the days alive) than the birds that survived to the while relative spleen weight decreased as oil dose increased (p = 0.050, end of the study (mean = 74.67%). The mean daily dose of oil con- Fig. 3). Although relative heart weight did not significantly change with sumed by the birds that died prior to the end of the study ranged from oil dose, the hearts of oil-dosed birds were noted to have a loss of 1.15 to 6.56 mL/kg bw whereas the mean daily dose of oil consumed by structural integrity. This endpoint was not quantified as it was not birds that survived until the end of the trial ranged from 7.91 to anticipated; however, the generally more flaccid cardiac tissue of oil- 9.52 mL/kg bw (Table 1). dosed birds resembled dilated cardiomyopathy. There was a high incidence of parasitism in all groups, including: nematodiasis in the esophagus and small intestine; trematodiasis in the ff 3.2. Adverse health e ects colon, kidney, and proventriculus; coccidiosis in the kidney; and probable Sarcocystis sp. in the heart. Periductal inflammation involving Of the 31 adult, mixed-sex LAGUs used in the study, 21 LAGUs large collecting ducts was a frequent renal lesion seen in all three survived the entire 28 days; a total of 10 LAGUs died or were eu- treatment groups, albeit more prominent in the oil-dosed groups which thanized within 20 days of initiation of the study. LAGUs were eu- may be a result of oil excretion acting as an irritant. thanized based on veterinary assessment of severe distress or when the Inflammatory lesions were common across all three dosing groups. animals were moribund to ensure that necropsies could be performed Hepatic inflammation (hepatitis, cholangitis, or cholangiohepatitis) was on fresh carcasses and that a complete suite of endpoints could be present in more than half of LAGUs necropsied. Respiratory in- sampled. All of the birds that died or were euthanized prior to the end flammation (pneumonia or air sacculitis) or cardiac inflammation of the trial exhibited clinical signs of lethargy and reduced food intake. (myocarditis associated with Sarcocystis) was present in nearly a third fi While cloacal temperature did not change signi cantly during the trial, of LAGUs. Tenosynovitis was confirmed in two LAGUs at necropsy al- two birds in the 5 mL oil/kg bw group showed a large drop in tem- though animal observation records indicated 21 of 30 LAGUs limping at perature prior to death. One control bird was euthanized on day 4 of the least once. Once limping was observed in multiple animals, additional study (due to presumed capture myopathy) and replaced with a bird flooring was added for foot relief and most limping resolved. captured and maintained in the same conditions. Survival across the 28 fi days of the trial was signi cantly lower for birds in the oil-dosed groups 3.4. Oxidative stress compared to birds in the control group (p = 0.051, Fig. 2). Mortality in birds that did not survive until the end of the testing The hepatic antioxidant endpoints measured at necropsy were af- period may have been associated with weight loss caused by dosed fected by oil in a dose-dependent manner. Hepatic total glutathione fi birds refusing to eat their full ration of oiled sh such that they did not (TGSH), oxidized glutathione (GSSG), and reduced glutathione (rGSH) all significantly increased as mean dose of oil increased (p < 0.001 for all; Fig. 4). The rGSH:GSSG ratio was not statistically significant, but showed a decreasing trend with increasing oil dose. Interestingly, the data for these parameters were heteroskedastic; i.e., the variance in- creased with the mean response such that the spread in the data for these endpoints was higher for higher oil doses compared to the con- trols. Superoxide dismutase activity showed a decreasing trend as mean dose of oil increased (p = 0.205; Fig. 4).

3.5. Hematology

Packed cell volume (PCV), and concentrations of white blood cells, heterophils, lymphocytes, monocytes, eosinophils, and basophils were not affected by oil dose. For the limited number of samples that were evaluated, Heinz bodies were rarely (< 1%) identified in RBCs sampled from control birds and frequently (10-40%) found in RBCs from oil- Fig. 2. Survival curves for laughing gulls orally exposed to target doses of 0 mL/kg bw, dosed birds as analyzed by transmission electron microscopy (TEM; 5 mL/kg bw, or 10 mL/kg bw of artificially weathered MC252 crude oil. Solid lines show Fig. 5). Anecdotally, no difference in numbers of Heinz bodies related to the survival probability over time for the three groups and the dashed lines are the increasing oil dose was noted; erythrocytes from both 5 mL oil/kg bw fi con dence intervals associated with the survival probabilities. and 10 mL oil/kg bw birds contained similar numbers of dense

86 K.E. Horak et al. Ecotoxicology and Environmental Safety 146 (2017) 83–90

Fig. 3. Relative organ weights (i.e., organ weight as a percentage of body weight) as a function of mean dose of oil ingested for laughing gulls orally exposed to artificially weathered MC252 crude oil. inclusions that lacked structure, consistent with Heinz bodies. Ery- structure and electron density with Heinz bodies. Structures consistent throcytes in oil-dosed birds tended to have smudged nuclei lacking with degenerate organelles were only identified in oil-dosed birds chromatin detail and contained many dark cytoplasmic inclusions with (Fig. 5). the same homogeneity and electron density as Heinz bodies. Although the dark cytoplasmic inclusions were not membrane-associated as Heinz bodies classically are in mammals, they were consistent in

Fig. 4. Activity levels of hepatic total glutathione, reduced glutathione, oxidized glutathione, and superoxide dismutase in laughing gulls orally exposed to artificially weathered MC252 crude oil.

87 K.E. Horak et al. Ecotoxicology and Environmental Safety 146 (2017) 83–90

Conversely, for oil-dosed birds that did not survive through the end of the study, food consumption was often depressed for a number of days prior to death or euthanasia. This refusal of food may have been a behavioral response to the oil-injected minnows or may have had an unidentified physiological basis. Previous studies have reported variable effects of oil exposure on avian food consumption. Herring gull (Larus argentatus) nestlings orally dosed with 10 mL of Prudhoe Bay crude oil/kg bw/day or more con- sumed less food beginning on the third day of dosing compared to controls and began to lose weight on the fifth day of dosing (Leighton, 1986). Conversely, American kestrels (Falco sparverius) administered feed containing 3.0% oil from the Mexican Ixtoc I well blowout avoided feed for the first week of the study, but then consumed significantly more feed than birds receiving feed containing 0.3% oil and an equivalent amount of feed as controls (Pattee and Franson, 1982). Despite the increase in food consumption, these birds lost weight (Pattee and Franson, 1982). Hyperphagia with no weight gain was re- ported for herring gull chicks administered a single oral dose of 0.3 mL Fig. 5. Dark cytoplasmic inclusions with the same homogeneity and electron density as Kuwait or South Louisiana crude oil/kg bw (Miller et al., 1978). The Heinz bodies and structures consistent with degenerate organelles (red arrows). The lack of weight gain was attributed to reduced transport of essential yellow arrow indicates a normal cell. amino acids and possibly glucose across the GI tract as indicated by in vitro assays and pathological changes in the GI tract (Miller et al., 3.6. Clinical chemistry/plasma proteins 1978). Oil-induced hyperphagia has also been reported in adult Pekin ducks (Anas platyrhynchos domesticus) fed diets that provided approxi- Of the endpoints examined from plasma samples, 3-methyl histidine mately 2.9 mL South Louisiana crude oil, 2.5 mL Kuwait crude oil, or was the only one affected by oil dose. 3-methyl histidine, a marker of 1.3 mL No. 2 fuel oil/kg bw/day (Holmes et al., 1978). There were no muscle wastage, was positively correlated with oil exposure (R2 = 0.41, changes in food consumption in the LAGU in this study that survived to p = 0.001; Fig. 6). There were no significant changes in any of the other the end of the study, but animals that succumbed prior to the end of the clinical chemistry/plasma proteins measured. study showed reduced food consumption.

4.2. Adverse health effects 4. Discussion Five of the 10 birds that died or were euthanized were in the 5 mL 4.1. Food consumption and oil intake oil/kg bw dose group and 4 of the 10 birds assigned to the 10 mL oil/kg bw group died or were euthanized compared to none in the control In the current study, much effort was made to ensure that test birds group, suggesting there was a dose-related response to oil exposure in were consuming both oil-injected fish and unoiled fish rations. While the present study. However, since many of these birds refused the oil- food consumption was depressed for the first few days of the study in injected fish (and in some cases did not consume additional rations oil-dosed birds, when larger oil-injected minnows were offered begin- when they were offered), the potential impact of oil on these birds is ning on day 5, consumption in all three treatment groups was similar confounded with food consumption, and the effects of oil and food for birds that survived to the end of the study, presumably because the consumption cannot be disentangled. Nonetheless, the mortalities in oil was less detectable by the LAGU compared to the smaller minnows. the 5 mL oil/kg bw group may indicate that even low doses of oil could

Fig. 6. Levels of 3-methylhistidine, a marker of muscle wastage, in laughing gulls orally exposed to artificially weathered MC252 crude oil.

88 K.E. Horak et al. Ecotoxicology and Environmental Safety 146 (2017) 83–90 induce mortality in wild populations of birds if oil-free food sources are (1986) reported that the most predominant lesion in the livers of her- not available and birds have limited opportunities to find un- ring gull chicks and Atlantic puffin nestlings dosed daily with 10 mL of contaminated resources. Prudhoe Bay crude oil/kg bw consisted of enlarged Kupffer cells that Previous studies using oil doses consistent with the present study were filled with gold-brown pigment indicative of hemosiderin and have shown mortality, loss of body weight, and lethargy in Herring gull phagocytized erythrocytes. Necrosis of individual hepatocytes and (Larus argentatus) chicks, Atlantic puffin(Fratercula arctica) nestlings, apoptosis were prevalent in the gulls. Hepatic hemosiderosis in oil-ex- and American kestrels (Falco sparverius). Some of these studies point to posed birds has been reported in numerous previous studies (Fry and cold stress as a possible cofactor increasing the toxic effects of oil. Lowenstine, 1985; Pattee and Franson, 1982; Yamato et al., 1996). In Although LAGU in this study did not experience cold stress, they were the present study, hemosiderosis was minimal regardless of the severity reluctant to consume oiled food and therefore were calorically re- of the anemia and no difference between oil-dosed birds and control stricted (Miller et al., 1978; Pattee and Franson, 1982; Leighton, 1986). birds with normal PCVs was identified upon standard hematoxylin and Lafferty and Holt (2003) have suggested that additional stressors eosin (H & E) staining or Prussian blue staining (Khan and Nag, 1993). caused by the presence of oil in the environment, such as increased Anemia with minimal hemosiderosis indicates that a lack of cellular competition for oil-free food resources or reduced mobility, may cause regeneration in the bone marrow may also be a component of the an- increased mortality at lower doses. While additional stressors such as emia. The damaging effects of components of crude oil on bone marrow temperature or competition were not addressed in the present study, have been documented in mammals (Linet et al., 1996). Therefore, it the parasitism and inflammation noted across study groups may in- follows that MC252 oil could have induced hypoplastic marrow con- dicate that additional stressors might affect the impact of oil exposure tributing to lack of regeneration of erythrocytes. in wild populations. Collectively, these studies indicate that mortality in response to oil dosing is relatively common and the mortality ex- 4.4. Hematology hibited by the gulls is consistent with previous studies examining oil toxicity. Whether these effects in the gull study were associated with The presence of Heinz bodies in oil exposed LAGUs is consistent weight loss, physiologic effects of oil toxicity, or a behavioral response with the finding of changes in hepatic glutathione corresponding to that led the birds to reject the dosed fish is unknown. systemic oxidative stress. While Heinz bodies are indicative of hemo- lytic anemia, no significant change in PCVs was detected. As many of 4.3. Gross observation and histopathology the oil exposed birds were inappetent, especially prior to death, it is possible that dehydration resulted in hemoconcentration and increased The increase in relative kidney weights in oil-dosed birds in the PCVs, masking hemolytic anemia due to oil exposure. The presence of present study may indicate renal dysfunction and is consistent with Heinz bodies and degenerate organelles in the RBCs in both groups of previous studies. Cassin's auklets (Ptychoramphus aleuticus) exposed to oil-dosed LAGUs (but not in the control group) indicates that a com- oil via external application and common murres (Uria aalge) recovered ponent of oil-induced anemia could be the result of oxide radical da- from a spill of bunker C fuel oil had renal tubular necrosis (Fry and mage of RBCs and intravascular hemolysis as previously documented Lowenstine, 1985). Mallard (Anas platyrhynchos) ducklings fed a diet (Lee et al., 1985; Leighton, 1985, 1993; Troisi et al., 2007). This re- containing 5.0% South Louisiana crude had tubular inflammation and sponse is consistent with that reported in other avian species dosed with degeneration in the kidney (Szaro et al., 1978). On the other hand, oil (Fry and Lowenstine, 1985; Hartung and Hunt, 1966; Lee et al., Leighton et al. (1986) did not define renal lesions in herring gull chicks 1985; Leighton, 1985, 1993; Pattee and Franson, 1982; Szaro et al., and Atlantic puffin nestlings dosed with Prudhoe Bay crude oil. Al- 1978) or exposed to oil as the result of a spill (Yamato et al., 1996). though renal dysfunction did not impact survivorship in the present study, it is an interesting finding that may have impacts on wild po- 4.5. Oxidative stress pulations. In this study, relative liver weights were significantly increased in There was evidence of an increase in hepatic tissue oxidative stress oil-dosed birds, but histological lesions were minimal and generally in the LAGUs based on increases in liver GSH and GSSG and a negative occurred in birds of all three groups. Increases in liver weight have been trend in the rGSH:GSSG ratio. Rodríguez-Estival et al. (2016) have re- reported in previous avian oil exposure studies. Holmes et al. (1978) ported the utility of assessing oxidative stress as an early indicator of reported that adult Pekin ducks consuming approximately 6 mL of diminished health. GSH is the predominant non-protein thiol in cells South Louisiana crude per day had increased relative liver weights and plays a key role in maintaining the redox homeostasis within the compared to controls, but relative liver weights of ducks consuming cell. Induction of de novo synthesis of GSH occurs as an adaptive re- 6 mL of Kuwait crude were comparable to control weights. American sponse to oxidative stress (Biswas and Rahman, 2009). GSSG is the kestrels fed a diet containing 0.3% or 3.0% crude from the Ixtoc I well oxidized form of GSH and increases during oxidative stress, primarily blowout for one, two, or four weeks did not show absolute or relative by the reactions of the antioxidant enzyme GSH peroxidase. The ratio of liver weights that were significantly different compared to controls GSH to GSSG can be used as a marker of oxidative stress (Zitka et al., (Pattee and Franson, 1982). Herring gull chicks administered a single 2012) which was demonstrated in this study by a decreasing trend with oral dose of 0.3 mL Kuwait or South Louisiana crude oil/kg bw had increasing oil dose. The process of inactivation of PAHs found in crude increased liver weights when necropsied nine days later (Miller et al., oil results in the formation of oxides and reactive oxygen radicals that 1978). An increase in liver weights of herring gull chicks receiving five are part of the oxidative stress insult to the birds. Consistent with the daily oral doses of 10 mL of Prudhoe Bay crude oil/kg bw was reported present study, Leighton et al. (1985) reported an increase in GSH/PCV by Peakall et al. (1989), and mallard ducklings fed diets containing in herring gull nestlings dosed with comparable amounts of crude oil 2.5% and 5.0% South Louisiana crude oil for eight weeks had sig- from the Prudhoe Bay oil spill. nificant increases in liver weights (Szaro et al., 1978). Both Miller et al. (1978) and Peakall et al. (1989) reported hepatic 4.6. Clinical chemistry/plasma proteins activity of mixed function oxidase enzymes were significantly increased in the absence of hepatic pathology, suggesting that the increase in liver 3-methyl histidine was the only parameter in the clinical chemistry weight could be attributed to a compensatory metabolic response. Szaro and plasma protein panels to have been affected by oil consumption. 3- et al. (1978) reported that liver pathology in ducklings fed oil-con- methyl histidine concentration increased as oil dose increased in- taining feed was subtle, consisting of generally minimal hypertrophy dicating the possibility of damage to the musculature and general and vacuolation of hepatocytes and bile duct proliferation. Leighton muscle wastage. This result may be correlated with the general

89 K.E. Horak et al. Ecotoxicology and Environmental Safety 146 (2017) 83–90 observation of potential dilated cardiomyopathy. This endpoint was exposed to oil. Arch. Environ. Contam. Toxicol. 14, 725–737. Harr, K.E., Rishniw, M., Rupp, R.L., Cacela, D., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C., more thoroughly examined in double crested cormorants that had been Healy, K., Horak, K., Link, J.E., Reavill, D., Bursian, S.J., 2017. Dermal exposure to externally oiled (Harr et al., 2017, in press). Although they may not be weathered MC252 crude oil results in echocardiographically identifiable systolic detectable in visual scans of oiled environments, these changes could myocardial dysfunction in double-crested cormorants (Phalacrocorax auritus). fl Ecotoxicol. Environ. Saf (in press). alter a bird's ability to successfully complete long ights, migrate, or Hartung, R., Hunt, G., 1966. Toxicity of some oils to waterfowl. J. Wildl. Manag. 30, flee from predators. While changes in 3-methyl histidine were noted in 564–570. the oil exposed birds, a corresponding change in creatine phosphoki- Holmes, W.N., Cavanaugh, K.P., Cronshaw, J., 1978. The effects of ingested petroleum on nase was not detected. oviposition and some aspects of reproduction in experimental colonies of mallard ducks (Anas platyrhynchos). J. Reprod. Fertil. 54, 335–347. Khan, R.A., Nag, K., 1993. Estimation of hemosiderosis in seabirds and fish exposed to 5. Conclusion petroleum. Bull. Environ. Contam. Toxicol. 50, 125–131. Lafferty, K.D., Holt, R.D., 2003. How should environmental stress affect the population dynamics of disease? Ecol. Lett. 6, 654–664. Overall, the birds tested in this study exhibited similar responses to Lee, Y.Z., Leighton, F.A., Peakall, D.B., Norstrom, R.J., O'Brian, P.J., Payne, J.F., oil exposure (e.g., change in relative organ weights, signs of potential Fahimtula, A.D., 1985. Effects of ingestion of hibernia and Prudhoe Bay crude oils on hemolytic anemia, and oxidative damage) compared to many other hepatic and renal mixed function oxidase in nestling herring gulls (Larus argentatus). Environ. Res. 36, 248–255. species studied previously. While the underlying causes of mortality Leighton, F.A., 1985. Morphological lesions in red blood cells from herring gulls and were not attributable to specific physiological or behavioral effects, the Atlantic puffins ingesting Prudhoe Bay crude oil. Vet. Pathol. 22, 393–402. significant mortality in dosed groups compared to the control group Leighton, F.A., 1986. Clinical, gross, and histological findings in herring gulls and Atlantic puffins that ingested Prudhoe Bay crude oil. Vet. Pathol. 23, 254–263. indicates that exposure to MC252 DWH oil had a negative impact on Leighton, F.A., 1993. The toxicity of petroleum oils to birds. Environ. Rev. 1, 92–103. laughing gulls. Leighton, F.A., Lee, Y.Z., Rahimutula, A.D., O'Brian, P.J., Piakall, D.B., 1985. Biochemical and functional disturbances in red blood cells of herring gulls ingesting Prudhoe Bay Funding crude oil. Toxicol. Appl. Pharmacol. 81, 25–31. Linet, M.S., Yin, S.N., Travis, L.B., Li, C.Y., Zhang, Z.N., Li, D.G., Rothman, N., Li, G.L., Chow, W.H., Donaldson, J., Dosemeci, M., Wacholder, S., Blot, W.J., Hayes, R.B., This work was supported by the US Fish and Wildlife Service as part 1996. Clinical features of hematopoietic malignancies and related disorders among of the Deepwater Horizon Natural Resource Damage Assessment. benzene-exposed workers in China. Benzene Study Group. Environ. Health Perspect. 104 (Suppl 6), S1353–S1364. Miller, D.S., Peakall, D.B., Kinter, W.B., 1978. Ingestion of crude oil: sublethal effects in Acknowledgements herring gull chicks. Science 199, 315–317. Pattee, O.H., Franson, J.C., 1982. Short-term effects of oil ingestion on American kestrels (Falco sparverius). J. Wildl. Dis. 18, 235–241. This study was supported by many and we are grateful for their Peakall, D.B., Norstrom, R.J., Jeffrey, D.A., Leighton, F.A., 1989. Induction of hepatic excellent assistance: Nicole Mooers, Jeremy Ellis, Rachael Moulton, mixed function oxidases in the herring gull (Larus argentatus) by prudhoe bay crude Benjamin Robey, Dr. Michael Scott, and Alicia Withrow. oil and its fractions. Comp. Biochem. Physiol. 94C, 461–463. Pritsos, K.L., Perez, C.R., Muthumalage, T., Dean, K.M., Cacela, D., Hanson-Dorr, K., Cunningham, F., Bursian, S.J., 2017. Dietary intake of Deepwater Horizon oil-in- References jected live food fish by double-crested cormorants resulted in oxidative stress. Ecotoxicol. Environ. Saf (in press). R Development Core Team, 2008. R: A Language and Environment for Statistical ff Bates, D., Maechler, M., Bolker, B., Walker, S., 2015. Fitting linear mixed-e ects models Computing. R Foundation for Statistical Computing, Vienna, Austria (ISBN 3-900051- – using lme4. J. Stats. Softw. 67, 1 48. 07-0). 〈http://www.R-project.org〉. Biswas, S.K., Rahman, I., 2009. Environmental toxicity, redox signaling and lung in- Rodríguez-Estival, J., García-De Blas, E., Smits, J.E.G., 2016. Oxidative stress biomarkers fl – ammation: the role of glutathione. Mol. Asp. Med. 30, 60 76. indicate sublethal health effects in a sentinel small mammal species, the deer mouse ff Burger, J., 1988. Foraging behavior in gulls: di erences in method, prey, and habitat. (Peromyscus maniculatus), on reclaimed oil sands areas. Ecol. Indic. 62, 66–75. – Colonia. Waterbirds 11, 9 23. Szaro, R.C., Dieter, M.P., Heinz, G.H., 1978. Effects of chronic ingestion of south Bursian, S.J., Alexander, C.R., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Ellis, Louisiana crude oil on mallard ducklings. Environ. Res. 17, 426–436. C.K., Godard-Codding, C.A., Guglielmo, C.G., Hanson-Dorr, K.C., Harr, K.E., Healy, Toth, L.A., 2000. Defining the moribund condition as an experimental endpoint for an- K.A., Hooper, M.J., Horak, K.E., Isanhart, J.P., Kennedy, L.V., Link, J.E., Maggini, I., imal research. ILAR J. 41, 72–79. Moye, J.K., Perez, C.R., Pritsos, C.A., Shriner, S.A., Trust, K.A., Tuttle, P.L., 2017. Troisi, G., Borjesson, L., Bexton, S., Robinson, I., 2007. Biomarkers of polycyclic aromatic Overview of avian toxicity studies for the Deepwater Horizon natural resource da- hydrocarbon (PAH)-associated hemolytic anemia in oiled wildlife. Environ. Res. 105, – mage assessment. Ecotoxicol. Environ. Saf. 142, 1 7. 324–329. Dean, K.M., Bursian, S.J., Cacela, D., Carney, M.W., Cunningham, F.L., Dorr, B.S., Yamato, O., Goto, I., Maede, Y., 1996. Hemolytic anemia in wild seaducks caused by Hanson-Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Lipton, I., McFadden, A.K., marine oil pollution. J. Wildl. Dis. 32, 381–384. McKernan, M.A., Harr, K.E., 2017. Changes in white cell estimates and plasma Zitka, O., Skalickova, S., Gumulec, J., Masarik, M., Adam, V., Hubalek, J., Trmkova, L., chemistry measurements following oral or external dosing of double-crested cor- Eckschlager, T., Kizek, R., 2012. Redox status expressed as GSH:GSSG ratio as a fi morants, Phalacrocorax auritus, with arti cially weathered MC252 oil. Ecotoxicol. marker for oxidative stress in paediatric tumour patients. Oncol. Lett. 4, 1247–1253. Environ. Saf (in press). Fry, M.D., Lowenstine, L.J., 1985. Pathology of common murres and Cassin's auklets

90 Ecotoxicology and Environmental Safety 146 (2017) 91–97

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Effect of oral exposure to artificially weathered Deepwater Horizon crude oil MARK on blood chemistries, hepatic antioxidant enzyme activities, organ weights and histopathology in western sandpipers (Calidris mauri) ⁎ Steven J. Bursiana, , Karen M. Deanb, Kendal E. Harrc, Lisa Kennedyd, Jane E. Linka, Ivan Magginid, Chris Pritsose, Karen L. Pritsose, R.E. Schmidtf, Christopher G. Guglielmod a Department of Animal Science, Michigan State University, 474 South Shaw Lane, East Lansing, MI 48824, United States b Abt Associates, 1881 Ninth St., Ste 201, Boulder, CO 80302-5148, United States c Urika Pathology, Mukilteo, WA 98275, United States d Department of Biology, Advanced Facility for Avian Research, University of Western Ontario, London, ON, Canada N6G 1G9 e University of Nevada-Reno, Max Fleischmann Agriculture Bldg. 210, Reno, NV 89557, USA f Zoo/Exotic Pathology Service, 6020 Rutland Drive #14, Carmichael, CA 95608, United States

ARTICLE INFO ABSTRACT

Keywords: Shorebirds were among birds exposed to Mississippi Canyon 252 (MC252) crude oil during the 2010 Deep Water Deepwater Horizon oil spill Horizon (DWH) oil spill in the Gulf of Mexico. The western sandpiper (Calidris mauri) was chosen as one of four Natural Resource Damage Assessment species for initial oral dosing studies conducted under Phase 2 of the avian toxicity studies for the DWH Natural Western sandpipers Resource Damage Assessment (NRDA). Thirty western sandpipers were assigned to one of three treatment Oral oil toxicity groups, 10 birds per group. The control group was sham gavaged and the treatment groups were gavaged with 1 − or 5 mL oil kg bw 1 daily for 20 days. Periodic blood samples for hemoglobin measurements were collected during the trial. A final blood sample used to determine hemoglobin concentration in addition to complete blood counts, plasma clinical chemistries, haptoglobin concentration and plasma electrophoresis was collected when birds were euthanized and necropsied on day 21. Tissues were removed, weighed and processed for subsequent histopathological evaluation. There were numerical decreases in hemoglobin concentrations in oil-dosed birds over the 21-day trial, but values were not significantly different compared to controls on day 21. There were no significant differences between controls and oiled birds in complete blood counts, plasma chemistries, haptoglobin concentration, and plasma electrophoresis endpoints. Of the hepatic oxidative stress endpoints assessed, the total antioxidant capacity assessment (Trolox equivalents) for the control group was lower − − compared to the 1 mL oil kg bw 1 group. Absolute liver weights in the 5 mL oil kg bw 1 group were significantly greater compared to controls. While not conclusive, the numerical decrease in hemoglobin concentration and significant increase in absolute liver weight are consistent with exposure to oil. Histological changes in the adrenal gland could be considered a non-specific indicator of stress resulting from exposure to oil. It is possible that the quantity of oil absorbed was not sufficient to induce clearly evident hemolytic anemia or that the western sandpiper is relatively insensitive to ingested oil.

1. Introduction and the DWH NRDA are given in Bursian et al. (2017). The western sandpiper (WESA) was chosen as a test species because it is often found Shorebirds were among birds exposed to Mississippi Canyon 252 in the Gulf of Mexico during the winter months and during migration (MC252) crude oil during the 2010 Deep Water Horizon (DWH) oil spill (Nebel et al., 2002) and thus is representative of other migratory in the Gulf of Mexico. The western sandpiper (Calidris mauri) was shorebirds that might have been exposed to MC252 crude oil, its small chosen as one of four species for initial oral dosing studies conducted size is applicable to methods used in metabolism and flight performance under Phase 2 of the avian toxicity studies for the DWH Natural studies that are included in the Phase 2 avian toxicity studies, and Resource Damage assessment (NRDA). Details about the DWH oil spill because researchers at the University of Western Ontario have success-

⁎ Corresponding author. E-mail addresses: [email protected] (S.J. Bursian), [email protected] (K.M. Dean), [email protected] (K.E. Harr), [email protected] (J.E. Link), [email protected] (I. Maggini), [email protected] (C. Pritsos), [email protected] (R.E. Schmidt), [email protected] (C.G. Guglielmo). http://dx.doi.org/10.1016/j.ecoenv.2017.03.045 Received 5 August 2016; Received in revised form 24 March 2017; Accepted 27 March 2017 Available online 14 April 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 91–97

− fully used this species for laboratory metabolism and flight studies in 10 birds that were gavaged daily with 5 mL oil kg bw 1 for 20 days. the past. Birds were weighed daily prior to dosing to standardize the dose across A preliminary oral toxicity study conducted with WESAs followed individuals. A mealworm homogenate was prepared at the ratio of six similar protocols described in the literature (Leighton, 1986) as well as mealworms per 2 mL water placed in a 12 mL polypropylene tube and recommendations from an expert panel. Birds were gavaged once with homogenized with an Omni 2000 (Omni International) variable speed a 1:1 mixture of artificially weathered MC252 oil and mealworm slurry tissue homogenizer. This mixture was centrifuged at 2000g in a Galaxy − − that provided a dose of 20 mL oil kg body weight 1 (bw 1) or on four Mini microcentrifuge (VWR International) to remove cuticle particles − consecutive days at daily doses of 10 or 20 mL oil kg bw 1. Results of that would clog the gavage needle. The dosed birds were gavaged with this initial oral toxicity trial indicated few effects characteristic of oil a 1:1 oil/mealworm slurry. For each bird, the appropriate volume of oil exposure that include anemia, decreased nutrient absorption, altered was combined with the appropriate volume of mealworm homogenate in a stress response, and decreased immune function (Szaro et al., 1978; microcentrifuge tube and the mixture thoroughly vortexed for 30 s. Leighton et al., 1985; Leighton, 1985, 1986, 1993; Peakall et al., 1989). Half the daily dose was administered initially, and the second half was More detailed information about the effects of oil exposure in birds can administered an hour later. The single dose of oil/mealworm slurry be found in Bursian et al. (2017). Rapid clearance of the oil adminis- administered to the birds was calculated as bw (kg)×0.5 − − tered by gavage (20–30 min) was assumed to have prevented sufficient (mL kg bw 1)×2 for the 1 mL kg bw 1 dose group and bw (kg)×2.5 − − polycyclic aromatic hydrocarbon (PAH) absorption to cause oil toxicity (mL kg bw 1)×2 in the 5 mL kg bw 1 dose group. The oil/mealworm as indicated by no significant decrease in packed cell volume (PCV). slurry was administered to a manually restrained bird through a 5.08- The present oral toxicity study incorporated modifications of the cm, 20-gauge stainless steel gavage needle attached to a 1-mL Luer lock dosing methods used in the initial oral toxicity study to better emulate glass syringe. Dosing of birds began on March 7, 2013. field conditions and potentially increase PAH absorption. The daily oil − doses were decreased from 10 and 20 mL kg bw 1 to 1 and 5 mL kg 2.4. Blood collection − bw 1. The number of consecutive days that dosing occurred increased from four to 20 to prolong the duration of exposure and extend the time Adult WESAs generally weigh from 22 to 35 g and blood samples over which endpoints indicative of exposure to oil were evaluated. The taken during the study were cumulatively limited to 10% of the bird's present study was undertaken to contribute to the definition of the total blood volume. Approximately 75 µl of blood was collected from all appropriate dose of oil to be administered in order to induce hemolytic birds two days prior to dosing, and on days 8 and 15 in a heparinized anemia in these birds. microhematocrit tube after pricking the brachial vein with a 27-gauge needle. This blood was used for hemoglobin measurements (VetScan 2. Materials and methods iStat analyzer; Abaxis Medical Diagnostics).

2.1. Study approval 2.5. Necropsy

This study was reviewed and approved by the University of Western Birds were necropsied on day 21 following collection of as much Ontario's Council on Animal Care (approval number 2012-027). blood as possible from the brachial vein as described above. Birds were then euthanized by cervical dislocation. Blood collected at necropsy 2.2. Western sandpiper capture, housing and diet was used to determine hemoglobin concentration, and complete blood count (CBC). In addition, plasma clinical chemistries (VetScan VS2 Thirty adult, mixed-sex WESAs that had previously been held at the analyzer; Abaxis Medical Diagnostics) were analyzed using an avian/ University of Western Ontario's Advanced Facility for Avian Research reptilian or liver profile rotor. Plasma haptoglobin concentration and (AFAR) were used in this study. These birds were captured in Delta, plasma electrophoresis were analyzed by the University of Miami's British Columbia, Canada using mist nets in July 2012 under the Avian and Wildlife Laboratory when sufficient blood was available. guidelines of the University of Western Ontario's Council on Animal Upon exposure of the body cavity, all organs were assessed grossly Care and according to permit CA-0256 from the Canadian Wildlife for abnormalities, which, if present, were documented by digital Service. Prior to the study, WESAs were maintained in one of the images. The liver and brain were removed and weighed to the nearest specialized 2.4×3.7 m shorebird rooms at AFAR under 12:12 light 0.1 g. The liver was sectioned into five 50 mg samples that were placed conditions at approximately 19 °C. Birds were fed ad libitum a diet of in individual cryovials and frozen in liquid nitrogen for subsequent 80% Mazuri Waterfowl Starter and 20% Purina Aquamax Fingerling determination of oxidative damage. A sixth liver sample was frozen in Starter 300. The diet was supplemented with approximately 50 meal- liquid nitrogen for subsequent determination of cytochrome P450 worms per 20 birds every other day. (CYP) activity (Alexander et al., 2017). The remaining portion of liver One week prior to study initiation (March 1, 2013), birds were was placed in 10% neutral buffered formalin for subsequent histo- transferred into a large holding room under the same photoperiod and pathology. The gastrointestinal (GI) tract, spleen, kidneys, heart, lung temperature conditions. Dividers were placed in the room to create six and adrenal glands were also placed in 10% neutral buffered formalin 0.9×1.8×1.8 m corrals. There were two corrals per treatment group for subsequent histopathology. with five birds in each corral. Birds remained in their respective corrals throughout the trial. Their diet remained unchanged. 2.6. Assessment of hepatic oxidative damage

2.3. Toxicant, treatments and dosing Oxidative damage in liver was assessed on liver homogenates prepared from the individual liver subsamples described in 2.5. Total, The toxicant was artificially weathered-MC252 oil collected on July oxidized and reduced glutathione (TGSH, GSSG, and RGSH, respec- 26, 2010 during the DWH spill and artificially weathered (batch #: tively), malondialdehyde +4-hydroxylalkenals (MDA), and total anti- B030112, TDI-Brooks International, College Station, TX) as described in oxidant power (Trolox), were assayed according to (Pritsos et al. 2017). Forth et al. (2016) prior to receipt for use in the study. Western sandpipers were randomly chosen for the oral toxicity study and 2.7. Histopathology assigned to one of three treatment groups: a control group of 10 birds that were sham gavaged twice daily for 20 days; a group of 10 birds that Histopathology was performed by a board certified veterinary − were gavaged daily with 1 mL oil kg bw 1 for 20 days; and a group of pathologist (Zoo/Exotic Pathology Service), using standard paraffin

92 S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 91–97 embedding, sectioning and hematoxylin and eosin (H & E) staining 40 techniques. Organs were reviewed by the pathologist prior to a segment being removed for paraffin embedding. Any lesions were included for analysis. ## 30 * ## ## ** 2.8. Statistical analysis

Body weights and hematologic and plasma clinical chemistry values 20 determined over multiple time points were modeled as a repeated Weight (g) measures analysis of covariance (ANCOVA) with interaction, where dose is a categorical explanatory variable and days is a continuous explanatory variable. Differences among treatment groups on day 0 10 Control were evaluated. Hepatic oxidative stress and organ weight endpoints 1 ml oil kg bw-1 were analyzed by one-way analyses of variance (ANOVA) utilizing the -1 5 ml oil kg bw PROC MIXED procedure of SAS (SAS Institute, v. 9.4). Means were 0 compared using the least squares means function and considered 0 5 10 15 20 25 significantly different if p < 0.05. Relative organ weights were ana- Day of study lyzed by one-way ANOVA utilizing the PROC MIXED procedure of SAS Fig. 1. Effect of daily oral dosing with artificially weathered MC252 oil on body weight of following arc sin transformation. ANOVA assumptions were tested by western sandpipers (Calidris mauri). *,** Body weight of the 5 mL oil/kg bw birds was visual inspection of diagnostic plots of the heteroscedasticty and greater than control birds on days 1 and 14 (p < 0.05). Body weight in the 5 mL oil kg − − symmetry of the residuals, and conformity was satisfactory for most bw 1 treatment was greater than the 1 mL oil kg bw 1 on day 1 (p < 0.05). #,##Body − analytes. Assumptions were poorly met for heterophils and monocytes. weight on day 1 in the 5 mL oil kg bw 1 group was greater than that on days 7, 14, and Statistical significance is reported for only one analyte, Trolox equiva- 21 (p < 0.05). lents concentration, for which all assumptions were met including formal hypothesis test for normality of the residuals. Non-parametric Kruskal-Wallis tests were conducted, but not reported, as a way to confirm the conclusions of the hypothesis tests using ANOVA, and results were in agreement with respect to p < 0.05 criteria for all analytes except glucose, where the non-parametric test indicated statistical significance that was not indicated by ANOVA. Proposed reference intervals were augmented and verified in accordance with the American Society for Veterinary Clinical Pathology (ASVCP) guidelines using MedCalc (Version 14.12.0 64 bit; MedCalc Software, Ostend, Belgium) and a more stringent setting of the Dixon Test using confidence levels of 0.1 or Tukey's Outlier Test (Geffré et al., 2011; Friedrichs et al., 2012).

3. Results

3.1. Clinical signs and mortality

No bird dosed with oil had clinical signs indicative of toxicity such Fig. 2. Effect of daily oral dosing with artificially weathered MC252 oil on hemoglobin as lethargy and reduced food consumption. All birds survived the 21- concentration in western sandpipers (Calidris mauri). Regression analysis showed no fi ff day trial. Excreta from dosed birds contained oil within 30 min after signi cant main e ect of treatment and no treatment*day interaction, however there was −1 dosing. While the amount of oil in the excreta could not be quantified, it a significant (p < 0.05) effect of day. In the 1 mL oil kg bw treatment hemoglobin on −1 was obvious that the excreta were oily. day 21 was lower than on days 1 and 14 (p < 0.05). In the 5 mL oil kg bw treatment, fi ff hemoglobin was lower on days 7 and 21 than on day 1, on days 14 and 21 than day 7, and There was a signi cant dose e ect (p < 0.05) for WESA bw during lower on day 21 than on day 14 (p < 0.05). Dotted lines indicate reference intervals for − the 21-day study (Fig. 1), presumably due to changes in the 5 mL oil kg untreated WESAs from this population (10.4–15.6 g dL 1). − bw 1 group. All groups lost weight from day 0 to day 7 (2.0%, 2.5%, −1 −1 and 8.5% in the control, 1 mL oil kg bw , and 5 mL oil kg bw phocyte and monocyte counts were not significantly altered by oral oil −1 groups, respectively), but bw in the control and 1 mL oil kg bw dosing (Table 1). groups returned to day 0 values by day 21 (1.3% and 0.66% greater − than bw on day 0). However, the 5 mL oil kg bw 1 group lost approximately 5.9% bw over the 21-day period. 3.3. Plasma clinical chemistry

3.2. Hematology Very few plasma chemistry endpoints changed as a result of oral exposure to MC252 oil. There were no significant differences in plasma There was a significant day effect (p < 0.05) for hemoglobin activities of alanine aminotransferase, alkaline phosphatase, creatine concentration (Fig. 2). Hemoglobin concentrations were 9%, 8%, and phosphokinase, or gamma glutamyl transferase at necropsy on day 21 − 16% lower on day 21 compared to day 0 for the control, 1 mL oil kg for control, 1 and 5 mL kg bw 1 treatments. There were no significant − − bw 1 and 5 mL oil kg bw 1 groups, respectively (Fig. 2). One (10%) differences in the plasma concentrations of bile acids, cholesterol, − − 1 mL oil kg bw 1and two (20%) 5 mL oil kg bw 1 birds had glucose, haptoglobin, total protein, urea, or uric acid at necropsy for − hemoglobin values below the reference interval (indicative of anemia; control, 1 and 5 mL kg bw 1 treatments. Neither plasma calcium at Fig. 2). necropsy or plasma sodium concentration during the 21-day trial White blood cell estimates from blood drawn at necropsy on day 21 changed significantly. There was a significant day effect (p < 0.05) were not significantly different. Basophil, eosinophil, heterophil, lym- for plasma chloride concentration (Fig. 3). Mean concentrations were

93 S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 91–97

Table 1 Effect of oral dosing with artificially weathered MC252 oil on hematology, CBCs, plasma chemistries, plasma minerals, plasma electrophoresis, hepatic antioxidant enzymes and organ weights in western sandpipers (Calidris mauri) after 20 days of oil treatment.

− − Control 1 mL oil kg bw 1 5 mL oil kg bw 1

Analyte n Mean SE n Mean SE n Mean SE p-value treatment

− Hemoglobin concentration (g dL 1) 7 12.8 0.3 6 12.9 0.3 7 12.0 0.6 0.730 − White blood cell count (x109 L 1) 10 5.57 1.41 10 4.59 0.45 10 5.84 0.89 0.651 − Heterophil count (x109 L 1) 10 2.36 0.97 10 1.41 0.15 10 1.99 0.67 0.619 − Lymphocyte count (x109 L 1) 10 2.12 0.27 10 2.24 0.23 10 2.80 0.40 0.273 − Monocyte count (x109 L 1) 10 0.39 0.20 10 0.24 0.08 10 0.32 0.09 0.722 − Eosinophil count (x109 L 1) 9 0.31 0.15 10 0.25 0.07 10 0.21 0.11 0.853 − Basophil count (x109 L 1) 9 0.47 0.08 10 0.45 0.12 10 0.52 0.12 0.903 − Alanine aminotransferase activity (U L 1) 7 98.0 12.6 6 114.2 16.9 7 113.7 17.2 0.705 − Alkaline phosphatase activity (U L 1) 7 112 13.8 6 105 13.8 7 119 14.9 0.794 − Creatine phosphokinase activity (U L 1) 7 1025 111.3 6 1078 245.5 7 1038 202.5 0.708 − Gamma glutamyl transferase activity (U L 1) 7 10.7 2.4 6 13.2 2.4 7 10.4 1.4 0.482 − Bile acid concentration (µmol L 1) 7 54.9 10.0 6 55.8 7.9 7 64.9 8.2 0.675 − Cholesterol concentration (mg dL 1) 7 293.1 22.6 6 344.8 22.4 7 322.1 34.0 0.435 − Glucose concentration (mg dL 1) 7 334.7 17.5 6 357.5 17.9 7 309.7 6.1 0.101 − Haptoglobin concentration (mg mL 1) 5 0.70 0.22 4 0.37 0.11 4 0.67 0.25 0.503 − Total protein concentration (g dL 1) 7 3.7 0.4 6 3.5 0.2 7 3.8 0.3 0.856 − Urea concentration (mg dL 1) 7 7.0 1.6 6 6.8 0.6 7 7.1 1.7 0.989 − Uric acid concentration (mg dL 1) 7 22.7 1.9 6 21.9 2.0 7 19.8 1.7 0.538 − Plasma calcium concentration (mg dL 1) 7 8.8 0.3 6 8.9 0.2 7 9.2 0.3 0.605 − Plasma chloride concentration (mg dL 1) 7 126.6 1.0 6 126.7 1.0 7 128 1.0 0.519 − Plasma sodium concentration (mmol L 1) 7 149.3 3.3 6 147.2 3.3 7 143.3 3.6 0.445 − Pre-albumin concentration (g dL 1) 7 0.001 0.001 5 0.002 0.002 7 0.003 0.002 0.831 − Plasma albumin concentration (g dL 1) 7 2.67 0.14 5 3.04 0.29 7 2.71 0.24 0.495 − Plasma α−1 globulin concentration (g dL 1) 7 0.10 0.03 5 0.12 0.01 7 0.13 0.03 0.982 − Plasma α−2 globulin concentration (g dL 1) 7 0.18 0.03 5 0.23 0.03 7 0.27 0.04 0.190 − Plasma β-globulin concentration (g dL 1) 7 0.60 0.13 5 0.58 0.05 7 0.64 0.11 0.910 − Plasma gamma-globulin concentration (g dL 1) 7 1.11 0.28 5 1.08 0.25 7 1.08 0.25 0.944 Plasma albumin: globulin ratio 7 1.58 0.19 5 1.59 0.06 7 1.46 0.16 0.808 − Hepatic total glutathione (nmol mg 1) 10 45.9 2.2 10 51.1 2.2 10 45.3 2.2 0.143 − Hepatic oxidized glutathione (nmol mg 1) 10 3.15 0.39 10 3.10 0.39 10 2.59 0.39 0.549 − Hepatic reduced glutathione (nmol mg 1) 10 39.63 2.26 10 44.90 2.26 10 40.11 2.26 0.209 − Malondialdehyde + 4-hydroxylalkenal concentration (nmol mg 1) 10 0.389 0.028 10 0.330 0.028 10 0.348 0.028 0.315 Brain weight (g) 10 0.60 0.01 10 0.60 0.01 10 0.58 0.01 0.215 Brain weight relative to bw (%) 10 2.39 0.06 10 2.32 0.05 10 2.17 0.08 0.0828 Liver weight relative to bw (%) 10 2.93 0.19 10 2.99 0.14 10 3.28 0.21 0.352

plasma pre-albumin, albumin, α−1 globulin, α−2 globulin, β-globulin or γ-globulin, and in the albumin: globulin ratio at necropsy on day 21 (Table 1).

3.4. Oxidative stress

The majority of hepatic oxidative stress endpoints were unchanged as a result of exposure to MC252 oil (Table 1). There were no significant differences in total glutathione, oxidized glutathione, reduced glutathione, or malondialdehyde +4-hydroxylalkenals (Table 1). Trolox equivalents concentration, which is used to represent the total antioxidant capacity of blood or organs, was significantly lower (p < 0.05) in the control group as compared to the 1.0 mL oil kg − bw 1 group (Fig. 4).

3.5. Relative and absolute organ weights and histopathology Fig. 3. Effect of daily oral dosing with artificially weathered MC252 oil on plasma chloride concentration in western sandpipers (Calidris mauri). Regression analysis showed fi ff fi ff There were no signi cant di erences in relative liver weights no signi cant main e ect of treatment and no treatment*day interaction, however there −1 fi was a significant (p < 0.05) effect of day on plasma chloride concentration. In the 5 mL (Table 1), but birds in the 5 mL oil kg bw group had signi cantly − oil kg 1 bw treatment, plasma chloride concentration was lower on day 1 than on days 14 greater (p < 0.05) absolute liver weights compared to controls (Fig. 5). and 21 and lower on day 7 than on days 14 and 21 (p < 0.05). Dotted lines indicate There were no significant differences in relative or absolute brain − − − reference intervals for untreated WESAs from this population (114–132 mmol L 1). weight of birds in the control, 1 mL oil kg bw 1 and 5 mL oil kg bw 1 groups (Table 1). − 1.6 mmol L 1 in the control group during the 21-day period, but A summary of histological lesions is presented in Table 2.Of − plasma chloride in the 1 and 5 mL kg bw 1 groups increased by 2.7% particular note are the following: 1) interrenal cell hypertrophy in the − and 2.4% respectively from day 0 to day 21. adrenal gland of birds in the 1 and 5 mL oil kg bw 1 treatment groups, − Plasma electrophoresis endpoints were not affected during the 21- increasing in severity in the 5 mL oil kg bw 1 dose group; 2) presence − day trial. There were no significant changes in the concentrations of of pulmonary hemorrhage in the 1 and 5 mL oil kg bw 1 dose groups

94 S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 91–97

Table 2 Summary of histopathological lesions in tissues of western sandpipers (Calidris mauri) following daily oral dosing with artificially weathered MC 252 oil for 20 days.

− − Organ/Diagnosis Control 1 mL oil kg bw 1 5 mL oil kg bw 1

Adrenal Gland Interrenal hypertrophy – 1.8 (3) 2.5 (2) Protozoal cysts –– 2.0 (1) Mineralization 1.5 (1) –– Brain Mineralization –– 2.0 (1) Heart Intimal papillary formation 1.7 (7) 2.3 (8) 2.5 (5) Myocarditis 3.0 (1) – 2.8 (2) Myocardial degeneration 1.5 (1) –– GASTROINTESTINAL TRACT Proventriculus Nematodiasis 2.6 (6) 2.6 (5) 2.4 (10 Proventriculitis 2.5 (5) – 2.5 (3) Fig. 4. Effect of daily oral dosing with artificially weathered MC252 oil on total Glandular cysts 2.0 (2) –– antioxidant capacity (mM Trolox equivalents) in livers of western sandpipers (Calidris Ventriculus mauri). The lower and upper boundaries of the boxes indicate the 25th and 75th Mucosal mineralization 1.7 (6) 1.5 (2) 3.5 (1) percentiles, respectively. The black line within the boxes is the median while the white Nematodiasis 2.0 (11) –– line is the mean. Lower and upper whiskers indicate the 10th and 90th percentiles, Bacterial colonization 2.5 (1) – 4.0 (2) ab –– respectively. Trolox equivalents concentration was significantly lower (p < 0.05) in the Submucosal mineralization 2.0 (2) − control group than in the 1.0 mL oil kg bw 1. Small intestine Protozoal cysts 1.3 (2)a – 2.5 (1) Enteritis 3.5 (1) –– Large intestine Protozoal cysts 3.2 (5) – 4.0 (2) Spirilliform bacteria 1.1 (3) – 1.8 (3) Colitis 3.0 (4) –– Kidney Mineralization – 1.0 (3) 1.0 (1) Interstitial nephritis 2.3 (2) 1.3 (2) – Periductal inflammation with 2.0 (1) –– parasites Protozoal cysts 1.5 (1) –– Liver Hepatocyte pigment –– 1.0 (2) Cholangitis –– 1.0 (3) Cholangiohepatitis 2.0 (2) 2.3 (3) – Protozoal cysts –– 2.0 (1) Subcapsular degeneration –– 3.5 (1) Lung Congestion 2.4 (9) 2.4 (10) 2.7 (9) Hemorrhage – 2.5 (1) 3.0 (2) Pneumonia – 3.0 (1) – Fig. 5. Effect of daily oral dosing with artificially weathered MC252 oil on liver weight of Parasitic pneumonia – 1.0 (1) – western sandpipers (Calidris mauri). The lower and upper boundaries of the boxes indicate Mineralization –– 2.5 (1) the 25th and 75th percentiles, respectively. The black line within the boxes is the median Pancreas –– while the white line is the mean. Lower and upper whiskers indicate the 10th and 90th Cystic ductal hyperplasia 3.5 (1) percentiles, respectively. abAbsolute liver weight was significantly greater (p < 0.05) in Spleen − – – the 5 mL oil kg bw 1 group than in the control group. Lymphoid follicle formation 2.0 (1)

a Numbers in the table represent mean severity score (range of 1–minimal to 4–severe) but absent in the control birds; 3) absence of hemosiderosis in the liver of the particular lesion while numbers in parentheses refer to the number of birds of all birds; 4) a general lack of inflammatory response in treated birds affected. such as enteritis found in the small intestine in controls but not treated birds; colitis found in control but not treated birds; cholangiohepatitis − found in the control and 1 mL oil kg bw 1 treatment but not in the of dosing increased from four to 20 consecutive days. As was the case − 5 mL oil kg bw 1 dose group; interstitial nephritis found in the control with the preliminary study, there were few statistically significant − − and 1 mL oil kg bw 1 treatment but not in the 5 mL oil kg bw 1 dose changes in the endpoints evaluated in the present study. group; There were two factors that may have contributed to the apparent lack of significant changes. First, even with smaller volumes of oil gavaged twice daily into the birds, gut transit time did not change 4. Discussion and conclusions appreciably, which may have resulted in insufficient absorption of PAHs to induce hemolytic anemia. Second, it was difficult to obtain 4.1. Hematology blood samples of sufficient volume and quality for analyses with repeated sampling in small birds such as the WESA. Despite these A preliminary oral dosing study with WESAs failed to demonstrate constraints, there was a 16% decrease in hemoglobin concentration in fi − hemolytic anemia with short durations of oral dosing with arti cially the 5 mL oil kg bw 1 group suggesting that oral exposure of WESAs to weathered MC252 oil. The absence of hemolytic anemia was presum- MC252 oil resulted in a mild, uncompensated anemia in two 5 mL oil kg ff − ably due to a gastroprotective e ect resulting in hyperperistalsis and bw 1 birds. This response is consistent with that reported in other diarrhea. In the current study, daily doses were decreased from 10 and − − − − avian species dosed with oil (Hartung and Hunt, 1966; Szaro et al., 20 mL kg bw 1 day 1 to 1 and 5 mL kg bw 1 day 1 and the duration

95 S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 91–97

1978; Pattee and Franson, 1982; Leighton et al., 1983, 1985; Leighton, to reports on herring gulls, common murres (Uria aalge) and Cassins 1985; Fry and Lowenstine, 1985) or exposed to oil as the result of a spill auklets (Ptychoramphus aleuticus) that developed anemia when exposed (Fry and Lowenstine, 1985; Yamato et al., 1996). It is unfortunate that to Prudhoe Bay crude oil (Fry and Lowenstine, 1985; Leighton, 1986). oxidative damage in RBCs could not be confirmed with Heinz body Upon exposure to weathered MC252, double-crested cormorants not identification, however, oxidative endpoints in hepatic tissue support only exhibited coagulopathy but also hematochezia (loss of blood in the that the anemia had a hemolytic component. Decreased bone marrow feces) (Cunningham et al., 2017; Harr et al., 2017). External blood loss production due to oxidative damage may also contribute to the would result in anemia without hemosiderin accumulation in the liver decrease in erythron mass. Total antioxidant capacity is a measurement and it is hypothesized that the same toxicant was inducing similar of a system's total chemical and enzymatic capacity to protect against conditions in both species. an oxidative insult. These measurements are typically standardized by While not conclusive, the numerical decrease in hemoglobin con- comparing the system's antioxidant capacity to an equivalent antiox- centration and significant increase in absolute liver weight are con- idant concentration of Trolox, a water-soluble Vitamin E analog. The sistent with exposure to oil. It is possible that the quantity of oil advantage of these measurements is that they can take into account absorbed was not sufficient to induce clearly evident hemolytic anemia synergistic interactions and minor components of the oxidative defense or that the WESA is relatively insensitive to ingested oil. While there that may not be captured when assaying for a specific antioxidant were no histological changes that were characteristic of exposure to oil, component. The results presented in this study suggest that in response interrenal cell hypertrophy, which was observed to a greater extent in to the exposure to oil, WESA are increasing their overall antioxidant the adrenal glands in treated birds, could be considered a non-specific capacity in the liver. This suggests that WESA are experiencing an indicator of stress resulting from exposure to oil. As discussed pre- oxidative insult due to the ingestion of oil. viously, it is also possible that the oil-induced changes in blood-related endpoints could not be detected because of the difficulty collecting 4.2. Organs blood samples that were not hemolyzed.

− Birds in the 5 mL oil kg bw 1 group had increased absolute liver Acknowledgements weights compared to controls. An increase in liver weight was also reported in other avian oil studies. Holmes et al. (1978) reported that The studies appearing in this special issue were funded by the U.S. adult mallards (Anas platyrhynchos) consuming approximately 6 mL of Fish and Wildlife Service as part of the Deepwater Horizon Natural South Louisiana crude oil daily had increased relative liver weights Resource Damage Assessment. compared to controls, but relative liver weights of ducks consuming 6 mL of Kuwait crude were comparable to control weights. American References kestrels (Falco sparverius) fed feed containing 0.3% or 3.0% crude oil from the Ixtoc I well blowout did not have absolute or relative liver Alexander, C., Hooper, M.J., Cacela, D., Smelker, K.D., Calvin, C.S., Dean, K.M., Bursian, weights that were significantly different compared to controls (Pattee S.J., Cunningham, F.L., Hanson-Dorr, K.C., Horak, K.E., Isanhart, J.P., Link, J.E., Shriner, S.A., Godard-Codding, C.A., 2017. CYP1A protein expression and catalytic and Franson, 1982). Herring gull (Larus argentatus) chicks administered activity in double-crested cormorants experimentally exposed to Deepwater Horizon a single oral dose of 0.3 mL Kuwait or South Louisiana crude oil kg Mississippi Canyon 252 oil. Ecotox. Environ. Safe in press. − bw 1 had increased liver weights when necropsied nine days later Bursian, S.J., Alexander, C.R., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Ellis, C.K., Godard-Codding, C.A., Guglielmo, C.G., Hanson-Dorr, K.C., Harr, K.E., Healy, (Miller et al., 1978). An increase in liver weight of herring gull chicks K.A., Hooper, M.J., Horak, K.E., Isanhart, J.P., Kennedy, L.V., Link, J.E., Maggini, I., −1 receiving five daily oral doses of 10 mL Prudhoe Bay crude oil kg bw Moye, J.K., Perez, C.R., Pritsos, C.A., Shriner, S.A., Trust, K.A., Tuttle, P.L., 2017. was reported by Peakall et al. (1989), and mallard ducklings fed diets Overview of avian toxicity studies for the Deepwater Horizon Natural Resource containing 2.5% or 5.0% South Louisiana crude oil for eight weeks had Damage Assessment. Ecotox. Environ. Safe in press. Cunningham, F.L., Dean, K.M., Hanson-Dorr, K.C., Harr, K.E., Healy, K., Horak, K., Link, significant increases in liver weight (Szaro et al., 1978). In both the J.E., Shriner, S., Bursian, S.J., Dorr, B.S., 2017. Development of methods for avian oil Miller et al. (1978) and Peakall et al. (1989) studies, hepatic activity of toxicity studies using the double-crested cormorant (Phalacrocorax auritus). Ecotox. mixed function oxidase enzymes was significantly increased in the Environ. Safe in press. Forth, H.P., Mitchelmore, C.L., Morris, J.M., Lay, C.R., Suttles, S.E., Lipton, J., 2016. absence of hepatic pathology, suggesting that the increase in liver Characterization of dissolved and particulate phases of water accommodated weight was a compensatory metabolic response. fractions used to conduct aquatic toxicity testing in support of the Deepwater Horizon natural resource damage assessment. Envir. Toxicol. Chem. . 〈http://onlinelibrary. wiley.com/doi/10.1002/etc.3672/full〉. 4.3. Histopathology Freeman, B.M., 1970. The effect of adrenocorticotrophic hormone on adrenal weight and adrenal ascorbic acid in normal and bursectomised fowl. Comp. Biochem. Physiol. Interrenal cell hypertrophy in the adrenal glands was increased in 32, 755–761. Friedrichs, K.R., Harr, K.E., Freeman, K.P., Szladovits, B., Walton, R.M., Barnhart, K.F., oil-dosed birds compared to control birds, which has been previously Blanco-Chavez, J., 2012. ASVCP reference interval guidelines: determination of de reported (Mazet et al., 2002). Interrenal hypertrophy in birds has been novo reference intervals in veterinary species and other related topics. Vet. Clin. associated with a variety of causes (Garren and Barber, 1955; Siegel, Pathol. 41, 441–453. fi Fry, D.M., Lowenstine, L.J., 1985. Pathology of common murres and Cassin's auklets 1959; Freeman, 1970). Although not a speci c marker for crude oil exposed to oil. Arch. Environ. Contam. Toxicol. 14, 725–737. exposure, the results indicate a definite systemic stress in exposed birds. Garren, H.W., Barber, C.W., 1955. Endocrine and lymphatic gland changes occurring in The weathered MC252 oil used in the present study theoretically young chickens with fowl typhoid. Poult. Sci. 34, 1250–1258. ff had low levels of volatile petroleum components (Forth et al., 2016), Ge ré, A., Concordet, D., Braun, J.P., Trumel, C., 2011. Reference value advisor: a new freeware set of macroinstructions to calculate reference intervals with Microsoft and therefore, should cause minimal direct damage to the respiratory Excel. Vet. Clin. Pathol. 40, 107–112. tract. Yet, there was an increase in pulmonary hemorrhage, indicating Harr, K.E., Bursian, S.J., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Hanson- damage to the lungs and a decreased ability to exchange oxygen, in the Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Reavill, D.R., Shriner, S.A., Schmidt, R.E., 2017. Comparison of organ weights and histopathology between double-crested treated birds that was not present in control birds. Coagulopathy was cormorants (Phalacrocorax auritus) dosed orally or externally with artificially documented using activated clotting time in double-crested cormorants weathered Mississippi Canyon 252 crude oil. Ecotox. Environ. Safe in press. (Phalacrocorax auritus) exposed to the same batch of weathered MC252 Hartung, R., Hunt, G.S., 1966. Toxicity of some oils to waterfowl. J. Wildl. Man. 30, 564–570. oil (Cunningham et al., 2017). Therefore, we postulate that the Holmes, W.N., Cavanaugh, K.P., Cronshaw, J., 1978. The effects of ingested petroleum on pulmonary hemorrhage in the WESAs may also have resulted from oviposition and some aspects of reproduction in experimental colonies of mallard coagulopathy similar to that reported for DCCOs. ducks (Anas platyrhynchos). J. Reprod. Fertil. 54, 335–347. Leighton, F.A., 1985. Morphological lesions in red blood cells from herring gulls and The incidence of hemosiderosis in the liver was not increased in Atlantic puffins ingesting Prudhoe Bay crude oil. Vet. Pathol. 22, 393–402. treated birds compared to controls in the present study. This contrasts

96 S.J. Bursian et al. Ecotoxicology and Environmental Safety 146 (2017) 91–97

Leighton, F.A., 1986. Clinical, gross, and histological findings in herring gulls and Atlantic Auk 119, 922–928. puffins that ingested Prudhoe Bay crude oil. Vet. Pathol. 23, 254–263. Pattee, O.H., Franson, J.C., 1982. Short-term effects of oil ingestion on American kestrels Leighton, F.A., 1993. The toxicity of petroleum oils to birds. Environ. Rev. 1, 92–103. (Falco sparverius). J. Wildl. Dis. 18, 235–241. Leighton, F.A., Peakall, D.B., Butler, R.G., 1983. Heinz body hemolytic anemia from the Peakall, D.B., Norstrom, R.J., Jeffrey, D.A., Leighton, F.A., 1989. Induction of hepatic ingestion of crude oil: a primary toxic effect in marine birds. Science 220, 871–873. mixed function oxidases in the herring gull (Larus aregentatus) by Prudhoe Bay crude Leighton, F.A., Lee, Y.Z., Rahimtula, A.D., O’Brien, P.J., Peakall, D.B., 1985. Biochemical oil and its fractions. Comp. Biochem. Physiol. 94C, 461–463. and functional disturbances in red blood cells of herring gulls ingesting Prudhoe Bay Pritsos, K.L., Perez, C.R., Muthumalage, T., Dean, K.M., Cacela, D., Hanson-Dorr, K., crude oil. Toxicol. Appl. Pharmacol. 81, 25–31. Cunningham, F.L., Bursian, S.J., Link, J.E., Shriner, S.A., Horak, K.E., Pritsos, C.A., Mazet, J.A.K., Newman, S.C., Gilardi, K.V.K., Tseng, F.S., Holcomb, J.B., Jessup, D.A., 2017. Dietary intake of deepwater horizon oil-injected live food fish by double-crested Ziccardi, M.H., 2002. Advances in oiled bird emergency medicine and management. cormorants resulted in oxidative stress. Ecotox. Environ. Safe in press. J. Avian Med. Surg. 16, 146–149. Siegel, H.S., 1959. The relation between crowding and weight of adrenal glands in Miller, D.S., Peakall, D.B., Kinter, W.B., 1978. Ingestion of crude oil: sublethal effects in chickens. Ecology 40, 495–498. herring gull chicks. Science 199, 315–317. Szaro, R.C., Albers, P.H., Coon, N.C., 1978. Petroleum: Effects on mallard egg Nebel, S., Lank, D.B., O’Hara, P.D., Fernandez, G., Haase, B., Delgado, F., Estela, F.A., hatchability. J. Wildl. Man. 42, 404–406. Ogden, L.J.E., Harrington, B., Kus, B.E., Lyons, J.E., Mercier, F., Ortego, B., Yamato, O., Goto, I., Maede, Y., 1996. Hemolytic anemia in wild seaducks caused by Takekawa, J.Y., Warnock, N., Warnock, S.E., 2002. Western sandpipers (Calidris marine oil pollution. J. Wildl. Dis. 32, 381–384. mauri) during the nonbreeding season: spatial segregation on a hemispheric scale.

97 Ecotoxicology and Environmental Safety 146 (2017) 98–103

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Low level exposure to crude oil impacts avian flight performance: The MARK Deepwater Horizon oil spill effect on migratory birds ⁎ Cristina R. Pereza, John K. Moyea, Dave Cacelab, Karen M. Deanb, Chris A. Pritsosa, a Department of Agriculture, Nutrition, and Veterinary Sciences, University of Nevada, Reno, United States b Abt Environmental Research, United States

ARTICLE INFO ABSTRACT

Keywords: In 2010, the Deepwater Horizon oil spill released 134 million gallons of crude oil into the Gulf of Mexico making Deepwater Horizon it the largest oil spill in US history. The three month oil spill left tens of thousands of birds dead; however, the Flight performance fate of tens of thousands of other migratory birds that were affected but did not immediately die is unknown. We Homing pigeon used the homing pigeon as a surrogate species for migratory birds to investigate the effects of a single external Migration oiling event on the flight performance of birds. Data from GPS data loggers revealed that lightly oiled pigeons Oil spill took significantly longer to return home and spent more time stopped en route than unoiled birds. This suggests that migratory birds affected by the oil spill could have experienced long term flight impairment and delayed arrival to breeding, wintering, or crucial stopover sites and subsequently suffered reductions in survival and reproductive success.

1. Introduction effect of oil on birds (Leighton, 1993). Feathers saturated by oil become matted and lose the critical functions of water repellency, insulation, The acutely lethal effects of heavy oiling to birds are well known and flight (Leighton, 1993). Birds rely on feathers for flight, insulation, from many oil spills around the world (e.g., Exxon Valdez, Nestucca, and buoyancy. Feather fouling due to oil exposure causes the feathers to Apex Houston and various spills in the North Sea) (Burger, 1993). As lose their critical properties, and has thus been attributed as the pri- expected, following the Deepwater Horizon (DWH) spill, many oiled mary cause of mortality of seabirds following oil pollution (Leighton, bird carcasses were recovered from shoreline areas along the Gulf of 1993; O’Hara and Morandin, 2010). Crude oil causes disruption in the Mexico (GOM). However, in addition, there were many birds observed microstructure of feathers, causing a collapse in the hooks, barbs, and during and after the spill that were oiled and still alive (Haney et al., barbules that are critical in maintaining the integrity and functionality 2014). This large number of live oiled birds stands in contrast to many of the feathers, such as their ability to repel water (O’Hara and other large oil spills, where oiled birds died relatively quickly through a Morandin, 2010). Alterations in feather microstructure can occur even combination of loss of feather buoyancy, thermoregulatory function, with thin oil sheens of less than 3 µm of crude oil. Feather fouling from and toxicity. These other large spills with high acute bird mortality quantities of oil as small as 10 ml results in lethally reduced thermo- have typically occurred in colder climates where the birds that became regulation in seabirds (O’Hara and Morandin, 2010; Hartung, 1967). oiled were rafting seabirds for which the loss of feather function and However, direct mortality may not be the only injury resulting from effects on metabolism had immediate and direct consequences. In feather fouling due to oil exposure. There is the potential for other non- contrast, the birds exposed to oil by the DWH spill were predominantly acutely lethal effects to occur, such as the impact of feather fouling on shorebirds that spend a much shorter amount of time on the water than flight aerodynamics and performance. do rafting sea birds (USFWS, 2011) and thus are not faced with the Besides the body contour feathers responsible for thermoregulation, same set of thermoregulatory and metabolic challenges, particularly the flight feathers of the wings and tail and are also subject to feather compared to rafting seabirds in northern climates. damage upon oil exposure. Worn feathers have been associated with Regardless of environmental temperature, the exposure of crude oil higher costs of flight and increased risk of predation (Hendenstrom, to the plumage of avian species is a notable threat. External oil con- 2003). Preening of oil from feathers and eventual ingestion can then tamination of feathers is the most common form of exposure, and the lead to a host of other effects including gastrointestinal irritation and effect of oils on feathers has been considered the single most harmful hemorrhaging, anemia, reproductive impairment, depressed growth,

⁎ Corresponding author. E-mail address: [email protected] (C.A. Pritsos). http://dx.doi.org/10.1016/j.ecoenv.2017.05.028 Received 11 July 2016; Received in revised form 16 May 2017; Accepted 17 May 2017 Available online 05 June 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 98–103 and osmoregulatory dysfunction (Friend and Franson, 1999). The in- the birds to return home in a timely manner. The morning of the flight, formation in the current literature helps to identify some of the effects birds were hand trapped, weighed to the nearest gram, dummy weights of oil on individual systems or responses in birds. However, currently removed and GPS dataloggers (accuracy: 4.2 m R95) attached, and available information from the literature and the field is not sufficient loaded into a crate for transport to the release site. Time and date, bird to fully characterize the nature and extent of the injuries to the tens of band number, colored band, GPS datalogger number, and pre-flight thousands of oiled birds exposed from the DWH oil spill, or to quantify weight were recorded. Upon arrival at the release site, GPS dataloggers those injuries in terms of effects on bird viability. were turned on, and birds were given a 20 min resting period prior to There is a great deal of literature on the husbandry of homing pi- release. The second of the two flocks was released 10 min after the first. geons (Columba livia), and they are a species that can be used and A shorter wait period was given if the previously released flock had maintained with ease in a captive environment. They are also one of the vanished before the end of the 10 min wait period. Release times were few, if not the only, free-flying species that can be easily trained to noted and recorded for each of the two flocks at each release. return to their home loft after being released from distant, unfamiliar After the five baseline flights, the Rim Fire in Yosemite Park filled locations. Many studies have investigated this homing behavior (re- the Washoe Valley with heavy smoke for a considerable amount of viewed in Guilford and Biro, 2014). For many decades, the homing time. This complicated the fight schedule, and the birds did not fly pigeon has been used as a paradigmatic model for studying navigation during this time. The first experimental flight occurred once the smoke behavior in birds (Wallraff, 2005). Homing pigeons are considered to cleared (Sept. 6, 2013), approximately two weeks after the last baseline possess the same navigational mechanisms as migratory birds flight. Birds from both groups had difficulties during this flight resulting (Wiltschko and Wiltschko, 2003), and therefore can be used as a sur- in the unexpected loss of birds, which we believe was a result of time off rogate species for migratory birds as their flight paths form release site and issues due to residual smoke in the air. Ten experimental flights to their home loft can be tracked and studied. These unique char- (EF) were conducted. However, due to the issues occurring on EF1, data acteristics make the homing pigeon an appropriate model to assess how from EF1 were removed from analyses (n=9 EF). Six control birds and exposures to environmental contaminants affect flight behavior. Pre- seven oiled birds were left and participated in the remaining experi- vious studies from our lab have investigated the effects of methylmer- mental flights. Day-of and day-prior pre-flight procedures were similar cury (Moye et al., 2016), carbofuran and diazinon (Brasel et al., 2007), to those conducted during baseline flights. Five experimental flights and chlorpyrifos and aldicarb (Moye and Pritsos, 2010) on avian flight occurred every other day followed by 5 flights twice a week during ability. September and October of 2013. Release times were noted and recorded The current study was designed to investigate effects on flight after for each of the two flocks at each release. a single low level application of crude oil. Birds were trained to return home from a distance of 81 km and multiple flights were conducted. We 2.2. Oil application hypothesized that after oiling, birds would perform more poorly in returning home. This was tested by examining a number of flight per- External oiling of the birds occurred the morning of the first ex- formance parameters including flight duration, flight speed, and stop perimental flight (EF1), September 6, 2013, at the release site. duration collected using Global-Positioning-System loggers Artificially weathered crude oil collected from the Deepwater Horizon (TechnoSmart, Rome, Italy; Steiner et al., 2000). MC252 spill was transported to the release site in a Koolatron© (Brantford, Ontario) and maintained at 21 °C. Pre-flight procedures 2. Materials and methods were similar to those of baseline flights. Once arriving at the release site, a bird was randomly selected, and unoiled photos were taken of The methods used in this study were carried out in accordance with that bird. Oil was then applied on the primary wing and tail feathers approved guidelines. These guidelines were approved by the following a standardized oiling regime covering 20% of the total body Institutional Animal Care and Use Committee at the University of surface area (Perez et al., 2014). Based on the different oiling rate ca- Nevada, Reno (protocol # 00541). Birds were provided with Nutriblend tegories established by the U.S. Fish and Wildlife Service (DWH Trus- Green Pigeon Chow (Purina Mills) and water ad libitum while being tees, 2016), 20% surface area coverage corresponded to the lightly housed. oiled category. For each wing, a length of 5 cm was measured along the dorsal side of the first primary, and from the end of that length a line 2.1. Training pigeons and experimental flights was drawn to the tip of the last primary. The area created was then painted dorsally with a 2.5 cm bristle paintbrush. For the tail, a 2 cm The two groups of birds used in this study were housed at two band of oil was applied along the tips of the primary tail feathers. Oiling different sites to avoid any cross contamination. The oiled group of procedures were repeated with each randomly selected bird until all birds was housed in pigeon loft located at the Agricultural Experiment birds within the group were oiled. During oiling, the bristle paintbrush Station at the University of Nevada, Reno. The control group of birds was saturated with oil such that the bristles did not come into contact was housed in a pigeon loft located at the University of Nevada's with the feathers. We were concerned that running the bristles of a Mainstation Farm in Sparks, Nevada (approx. 1.6 air kilometers apart). clean paintbrush on the feathers of the control birds would create an Both groups of birds were initially trap trained to learn how to use the artificial disruption of feather structure, thus, no sham treatment was mechanisms needed to enter the lofts as well as to become familiar with applied to the control birds. All of the birds in this study were ac- the local surroundings around the loft. The birds were then trained at customed to being regularly handled and the additional handling time gradually increasing distances from the loft until a final distance of was not expected to bias the results. The additional handling time of 80.5 km was reached. At the 32.2 km training distance, a dummy each bird in the oiled group before the first experimental flight was weight was attached to each bird in order to mimic the weight of the approx. 5–7 min. Handling time on all other experimental flights was GPS data loggers during flight (approx. 10 g). the same for the two groups. After birds were trained to return home from a distance of 80.5 km, five baseline flights (BF) were conducted to establish reference flight 2.3. Flight duration times for each of the birds to be used during the experimental flights. Eleven potential controls and ten potential oiled birds remained after Time of flight was determined from the release time recorded at the training and were available for baseline flights. Flights occurred in release site and arrival time recorded by the GPS dataloggers within an August 2013, every other day if weather permitted. Feed was removed 80–100 m radius around the home loft using mapping software (Fugawi at 1 pm PST at each loft the day prior to the flight in order to encourage Global Navigator Moving Map Software, Toronto, Canada). Logger data

99 C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 98–103

Table 1 Effect sizes and 95% confidence intervals between groups for flight duration, stop duration, and flight speed. Estimates were derived from the full linear mixed effects models including treatment group (oil and ctrl), flight period (baseline “BF” and experimental “EF”), and treatment group*flight period and bird ID (n=13) as a random effect. Confidence intervals that do not overlap zero are shown in bold.

Flight duration Flight speed Stop duration

Effect 95% CI Effect 95% CI Effect 95% CI

BF oil - BF ctrl 49.1 (−35.2, 133.4) 2.2 (−1.79, 6.16) 5.3 (−67.6, 78.2) BF ctrl - EF ctrl −18.7 (−88.3, 50.9) 2 (−1.68, 5.72) −22.9 (−76.3, 30.4) BF oil - EF oil −119.5 (−185.1,−54) −2(−5.69, 1.77) −113.7 (−167.8, −59.6) EF oil - EF ctrl 150 (80.2, 219.8) 6.2 (2.73, 9.61) 96 (27.2, 164.8) were filtered to remove any information collected by the GPS devices confidence intervals did not overlap zero. Data from birds in the oil before the official release time and after the official arrival time. If a group (n=7) and control group (n=6) that participated in both base- bird's time of flight was abnormally long, due to limited battery life of line (n=3) experimental flights (n=9) were included in the analyses. the dataloggers (approx. 12 h), arrival time was based on the time they Only birds with complete flight track data were included in the analyses were observed at the loft the next day. Pigeons are not active in the dark for stop time and flight speed (n=131 total observations), flight and were assumed to not be flying during dark hours. Thus, we did not duration data included all observations (n=177). One oiled bird did include these hours in their flight duration values. For birds that arrived not return from the fifth experimental flight (day 38, 11 days after the next day, flight times were calculated based on release time to time oiling). All available data from this bird until the day it was lost are of sunset plus one hour, plus sunrise to observed arrival time. included in the analyses. Linear mixed models were also used to assess the influence of stop 2.4. Flight speed duration and flight speed (continuous variables) on predicting flight duration. Data from complete flight tracks were included in these Instantaneous ground speeds (meters covered per second of flight) analyses (n=131). Additionally, to assess the effect of subsequent were obtained from the data collected by the GPS dataloggers. The flights during experimental flights on flight duration values, linear average speed during sustained flight was calculated for the entire mixed effects regression models with a repeated measures structure flight of each bird with a complete flight track data file. We included were compared. The full model included flight number, treatment, and flight speeds between 30 and 110 km/h. We applied upper and lower flight number*treatment interaction as fixed effects with bird ID as a limits to remove periods of stopping or periods of excessive tail wind random effect. All mixed effects models were fit by maximum like- assistance. Once these periods were removed, a mean flight speed was lihood and were analyzed using packages lme4 (Bates et al., 2013) and calculated for each bird at each flight. Birds that arrived the next day lmerTest (Kuznetsova et al., 2014) in R version 3.2 (R Core Team, did not have complete flight tracks and their flight speed was not 2016). Degrees of freedom were obtained using Satterthwaite approx- considered in analysis. imation in lmerTest.

2.5. Stop duration 3. Results

Each data file downloaded from the GPS dataloggers was examined 3.1. Flight duration for periods of stopped flight. Stops were defined when flight speed re- corded by the dataloggers decreased to less than 10 km/h for at least a The best model for the flight duration data included main effects of five minute period. The duration of individual stops were recorded for flight period, treatment group, and the interaction between flight each bird with a complete flight track during each flight. Individual period and treatment group. Analyzing the effect sizes of the estimates stop durations were added to get the total time spent stopped for each obtained from this model showed no difference in flight duration be- bird during each flight. Birds that arrived the next day did not have tween the two treatment groups during baseline flights (Table 1), ver- complete flight tracks and their overnight stop duration was not con- ifying the two groups were similar in flight performance prior to sidered in analysis. treatment. After oiling, the oiled birds took 1.6 times longer to return from the release site compared to their baseline flights, while there was fl 2.6. Statistical methods no corresponding change between ight periods of the control birds (Fig. 1A). Further, flight duration was significantly greater among oiled fl fl To test for effects of oiling on flight, we assessed differences in flight birds than control birds during experimental ights and ight durations fi fl fl duration, flight speed, and stop duration by treatment group (oil or increased from the rst ight to the ninth ight in both groups (Fig. 2). control) and flight period (baseline “BF” or experimental “EF”) using linear mixed effects models (LMM). Linear mixed models were created 3.2. Flight speed and stopped duration for each response variable (flight duration, flight speed, and stop duration) with explanatory variables added sequentially and individual The best fit LMM for flight speed included only the treatment group bird ID (n=13) added as a random effect to control for repeated parameter. This model had a lower AIC score than the full model, in- measures (random intercept). We analyzed the importance of ex- dicating that the addition of flight period and the interaction of treat- planatory variables on the response variable by the comparing the AIC ment group and flight period did not further explain the variation in the scores of the full models to the reduced models. The explanatory vari- speed data. However, examining the effect sizes of the point estimates ables included in the best fit model were interpreted as being sub- from the full model found a significant difference between the control stantial in predicting the response variable. Full models included main and oil group during experimental flights, with the oiled birds having a effects of flight period, treatment group, and the interaction of flight greater flight speed by 6.2 km/h (Table 1). There were no significant period and treatment group. Effect sizes were determined based on changes occurring in either group between flight periods, validating the differences of least square means between groups (BF ctrl, BF oil, EF stronger support for the treatment only model, such that the oiled birds ctrl, EF oil). Effect sizes were considered to be significant if their 95% were slightly faster flyers regardless of treatment (Fig. 1B). Conversely,

100 C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 98–103 A

300

200

Flight Duration (min) Flight Duration 100 Baseline Experimental Control Oil Control Oil Fig. 2. Individual flight duration values for each experimental flight for control birds (n=6, unfilled symbols) and oil birds (n=7, filled symbols). Both groups flew 9 experi- B mental (post-oiling) flights. Regression lines fit to a 2 parameter LMM including flight 75.0 number and treat group as fixed effects and bird ID as random effect. Control (dashed line): y=13.1x +114; Oil (solid line): y=13.1x +267.

72.5 AIC comparison found that the full model was the best fit model for the stop duration data. There was no significant difference between groups during baseline flights (Table 1). The oil group significantly increased 70.0 their stop duration from baseline to experimental flights by 114 min (Fig. 1C). The stop duration for the oil group was also significantly greater than the control group during experimental flights, and the 67.5 control group did not significantly change their stop duration between flight phases. On average, individuals from both groups spent some fl 65.0 time stopped during both the baseline and experimental ights, how- fl

Flight Speed (km/hr) ever after being oiled, none of the birds in the oiled group ew con- tinuously home without stopping while some of the unoiled birds did. Baseline Experimental 62.5 The pattern of increased stop duration was similar to the pattern Control Oil Control Oil seen for increased stop duration. In attempts to determine whether flight speed or stop duration better predicted flight duration, linear mixed models were compared (Table 2). The best fit model included 3 C parameters including the interaction between stop duration and flight speed. However, the difference in AIC scores between the flight speed 200 only model and the stop duration only model showed that stop duration is more important in predicting flight duration than flight speed. There was not strong support for a treatment group effect. Thus, regardless of 150 treatment group, birds that spent more time stopped en route had longer flight times (Fig. 3). 100 4. Discussion

50 This study demonstrates that a single oil application on the flight feathers of homing pigeons inhibited their ability to sustain flight and Stop Duration (min) Stop Duration substantially increased the overall time they required to complete 0 Baseline Experimental flights. We observed that the oiled birds in this study significantly in- creased their flight duration compared to their unoiled flights and Control Oil Control Oil Table 2 Fig. 1. Point estimates and 95% CI by treatment group and flight period. Estimates de- fl rived from full models including treatment group, flight period, and treatment group*- Linear mixed models for response variable ight duration with predictive variables of fl ff flight period and bird ID as a random effect (n=13). Control group consisted of 6 birds, stop duration, ight speed, and treatment group with bird ID as a random e ect (n=13) including metrics for model support. the oil group consisted of 7 birds and both groups flew 5 baseline flights (BF) and 9 fl fl experimental ights (EF) from an 81 km ight distance. A. Flight duration- BF control: Models df deviance deltaAIC logLik 161 min (99, 223); BF oil: 210 min (153,267); EF control: 180 min (129, 230); EF oil: 330 min (281,378) B. Flight speed- BF control: 68.64 km/h (65.8, 71.5); BF oil: Null (random effect only) 3 1620.9 526.7 −810.44 70.83 km/h (68.1, 73.6); EF control: 66.6 km/h (64.3, 69.0); EF oil: 72.79 km/h (70.3, Speed 4 1135.6 524.8 −809.98 75.3) C. Stop duration- BF control: 57 min (4, 110); BF oil: 62 min (12, 112); EF control: Stop 4 1620.0 40.4 −567.78 80 min (32, 128); EF oil: 176 min (127, 225). Stop + speed 5 1107.6 13.8 −553.78 Stop + speed + stop*speed 6 1091.2 0.0 −545.62 Stop + speed + stop*speed + trtGroup 7 1090.9 1.7 −545.43

Top-performing model listed in bold.

101 C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 98–103

indicates that non-oiled birds are able to fly repeated, uninterrupted 81 km flights without stopping; however, this was not the case for birds that were lightly oiled. On average, the birds in the treated group spent almost 50% of their total flight duration not flying. It could be argued that perhaps the oiled birds were solely stopping to preen oil from their feathers and may not be stopping to rest at all. However, stopped time increased after oiling over a 35 day period and suggests that this was not the case because they not only continued to make stops, but in- creased their stop durations long after oil application. The impact of feather fouling on flight aerodynamics due to oil contamination of the flight feathers is likely the main factor driving reduced flight ability. Photographs of all the birds were taken throughout the study to assess the changes in oil spreading and feather fouling with time and repeated flights. We observed that oiled feathers immediately became matted and clumped and over time and appeared frayed and brittle. The peak of feather damage was observed 20 days post-oiling, after 7 experimental flights. We believe the weakened in- Fig. 3. Relationship of total stop duration and corresponding flight duration for all ob- tegrity of individual feathers was a causative factor in flight impairment servations with complete flight tracks (n=131). Regression line (solid line) and con- fidence intervals (dashed lines) fit to 1 parameter LMM flight duration ~stop duration found in the oiled birds of this study. All birds have feathers and the including bird ID as a random effect (n=13). Regression line: y=1.03x +99.2. majority rely on them for flight used in daily activities and long dis- tance migration. Thus, any bird subjected to external exposure of crude compared to unoiled birds flying on the same days given the same oil could be susceptible to oil-induced flight impacts, which likely environmental conditions. Additionally, the increased flight durations translates into inhibited activities, such as an inability to escape pre- became greater as the experimental flights continued. This suggests that dators (Maggini et al., 2017b), forage, or failure to complete migration. despite only a single external oiling exposure, the effects of that oiling were carried over to subsequent flights up to 35 days following the 4.1. Implications for migrating birds exposure. Thus, the 59% increase in flight duration observed in the oiled group for the 9 experimental flights is not just due to an initial The coast of the Gulf of Mexico (GOM) is an important area for effect on one or two flights. This implies that birds exposed to even a migrating birds, specifically Neotropical migrants that travel to and single light oiling event could experience long term flight impairment. from their northern breeding grounds in the Northern US and Canada to We also observed considerable individual variability in the response of their southern wintering grounds in Central and South America. The the oiled birds despite identical treatments (Fig. 2), suggesting that geographic location of the GOM coast makes it an important stopover some individuals may be subject to greater oil-induced flight effects habitat because it is the last and first suitable area before and after than others. crossing the Gulf of Mexico water body. The Gulf of Mexico is also home We hypothesized two possible explanations for the observed in- to a large number of resident bird species, making it an important area creased flight durations of the oiled birds 1) slower flying speed and/or for birds all year long. The timing (April 20 to July 15) of the spill and 2) increased time spent stopped during flight. On average, the oiled extended persistence of the DWH oil allowed for the possibility of both birds had a greater flight speed than the control birds regardless of spring and fall migrants to be impacted (Henkel et al., 2012). In the flight period, as no changes between flight periods were observed in current study, light oil exposure was shown to inhibit a bird's flight either group. This pattern seen in flight speeds does not correspond to ability during 81 km flights, which could be analogous to a migratory the patterns seen in flight duration. On the other hand, the oiled birds bout (Ellegreen, 1993). Increasing the amount of time to complete a increased stop duration after oiling and stopped for longer periods of flight, ultimately decreasing migration speed, would dramatically time than the control birds during experimental flights. The relation- contribute to late arrival dates to desired destinations such as wintering ship between stop duration and flight duration had a slope approx- grounds, breeding grounds, or stopover sites. imating a 1:1 relationship (Fig. 3). However, including flight speed in For many spring migrants, breeding success and physical condition the model better explained the variation observed in flight duration decline with later arrival date to breeding grounds due to limited than stop duration alone, suggesting that although stop duration is the nesting site availability, poor mate opportunities, and limited food main effect predicting flight duration, flight speed and the interaction availability (Marra, 1998). The impacts of late arrival dates to breeding between the two are also important. The interaction of the two may be grounds have been shown in a large number of avian species to have needed to explain the variation observed in the flight duration values of negative consequences on reproductive success (Lozano et al., 1996; birds that had stop durations of zero. We believe that the time spent Bêty et al., 2003; Smith and Moore, 2005; Cooper et al., 2011). Several stopped during flight was required by the oiled birds to rest and alle- studies have looked at the direct relationship between the percent re- viate the stress of flying at higher costs. A recent study has found that duction in reproductive success and the number of days delayed. Gordo oiled western sandpipers (Calidris mauri) spend as much as 45% addi- et al. (2013) observed a 25% reduction in reproductive success from tional energy in endurance flights lasting 2 h when compared to unoiled only a two week delay in arrival time in white storks (Ciconia ciconia). birds (Maggini et al., 2017a). This study was conducted in a wind Bêty et al. (2003) found that an 11 day delay in arrival resulted in a tunnel which required the birds to fly continuously even if they were 25% reduction in eggs being laid by snow geese (Chen caerulescens). In a inclined to stop. Our data suggests that oiled free-flying birds in the study of nesting common eiders (Somateria mollissima), a 33% reduction wild may decide to stop flight for prolonged durations rather than in clutch size resulted from a 26 day delay in laying date (Descamps continue flight at higher energy costs. et al., 2011). Also, early nesting migrants lay more eggs and have In the current study, several control birds also had periods of heavier nestlings than individuals that are delayed (Lozano et al., stopped flight during both the baseline and experimental phases of the 1996). This is important when considering possible population level study. However, during the 9 experimental flights, multiple control effects because larger clutch sizes and heavier offspring increase the birds did not stop during flight and flew continuously home, whereas probability of recruitment into the breeding population. Reductions none of the oiled birds flew directly home without stopping. This seen in reproductive success due to delayed arrival to breeding grounds is variable across species and life history traits, however, it is clear that

102 C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 98–103 a delay of even a few days has the potential for negative consequences References on reproduction. Arriving at stopover sites during times of favorable conditions is Baker, A.J., et al., 2004. Rapid population decline in red knots: fitness consequences of necessary for successful migrations. Birds need adequate time and decreased refuelling rates and late arrival in Delaware Bay. Proc. R. Soc. Lond. B 271, 875–882. adequate resources to optimally refuel. Bates, D., Maechler, M., Bolker, B., Walker, S., 2013. lme4: linear mixed-effects models For the red knot (Calidris canutus), a long distance migrant, their last using Eigen and S4. R Package Version 1.0-5. 〈http://CRAN.R-project.org/package= lme4〉. stop in Delaware Bay is crucial for accumulating the nutrients needed Bêty, J., Gauthier, G., Giroux, J.-F., 2003. Body condition, migration, and timing of re- for the final flight to the breeding sites in the Canadian Artic (Baker production in snow geese: a test of the condition-dependent model of optimal clutch et al., 2004). Due to overharvesting of the red knots’ food source in the size. Am. Nat. 162, 110–121. Brasel, J.M., Collier, A.C., Pritsos, C.A., 2007. Differential toxic effects of carbofuran and area, they require more time at the stopover site to reach the fat stores diazinon on time of flight in pigeons (Columba livia): potential for pesticide effects on needed to complete the last leg of migration. Baker et al. (2004) found migration. Toxicol. Appl. Pharmacol. 219, 241–246. that individuals that arrived early to the stopover site in Delaware Bay Cooper, N.W., Murphy, M.T., Redmond, L.J., Dolan, A.C., 2011. Reproductive correlates of spring arrival date in the Eastern Kingbird Tyrannus tyrannus. J. Ornithol. 152, had a higher probability of surviving and returning in subsequent years 143–152. due to the advantage of having more time to refuel and reach optimal Descamps, S., Bêty, J., Love, O.P., Gilchrist, H.G., 2011. Individual optimization of re- production in a long-lived migratory bird: a test of the condition-dependent model of mass before departure for the breeding grounds. Birds that arrive in laying date and clutch size. Funct. Ecol. 25, 671–681. poor quality areas that are inadequate for proper refueling may be Dierschke, V., 2003. Predation hazard during migratory stopover: are light and heavy – forced to depart with less than optimal fuel stores and are at a higher birds under risk? J. Avian Biol. 34, 24 29. Ellegreen, H., 1993. Speed of migration and migratory flight lengths of passerine birds risk of unsuccessful migration and mortality (Dierschke, 2003). The during autumn migration in Sweden. Ornis Scand. 24, 220–228. probability of stopping in areas of poor quality becomes greater if birds Friend, M., Franson, J.C., 1999. Oil. In: Field Manual of Wildlife Diseases: General Field fl Procedures and Diseases of Birds. US Geological Survey, pp. 309–315. are required to make more frequent stops along their migratory ights, Gordo, O., Tryjanowski, P., Kosicki, J.Z., Fulín, M., 2013. Complex phonological changes which would be predicted based on the oiled bird flight data shown in and their consequences in the breeding success of a migratory bird, the white stork this study. This is especially important in areas of high disturbance and Ciconia ciconia. J. Anim. Ecol. 82, 1072–1086. Guilford, T., Biro, D., 2014. Route following and the pigeon's familiar area map. J. Exp. habitat loss, where required food resources may be limited. Further- Biol. 217, 169–179. more, the more frequent and prolonged stops of oiled birds along their Haney, J.C., Geiger, H.J., Short, J.W., 2014. Bird mortality from the Deepwater Horizon journey would also increase their risk of mortality as they would be a oil spill. II. Carcass sampling and exposure probability in the coastal Gulf of Mexico. Mar. Ecol. Prog. Ser. 513, 239–252. greater target for predation. Additionally, it has been observed in var- Harrison, X.A., Blount, J.D., Inger, R., Norris, D.R., Bearhop, S., 2011. Carry-over effects ious species of migratory birds (Bêty et al., 2003; Descamps et al., 2011; as drivers of fitness differences in animals. J. Anim. Ecol. 80, 4–18. Hartung, R., 1967. Energy metabolism in oil-covered ducks. J. Wildl. Manag. 31, Sandberg and Moore, 1996) that arrival at breeding grounds in poor 798–804. body condition results in reduced reproductive success. Hendenstrom, A., 2003. Flying with holey wings. J. Avian Biol. 34, 324–327. Fall migrants that arrive late to wintering grounds are also at a great Henkel, J.R., Sigel, B.J., Taylor, C.M., 2012. Large-scale impacts of the Deepwater Horizon oil spill: can local disturbance affect distant ecosystems through migratory risk. For example, Neotropical migrants that winter in the tropics face shorebirds? Biosci. 62, 677–685. habitat loss and fragmentation which has been considered a major Kuznetsova, A., Brockhoff, P.B., Christensen, R.H.B., 2014. lmerTest: tests for random and fi ff ff potential cause of songbird population declines (Stutchbury, 1994). It xed e ects for linear mixed e ect models (lmer objects of lme4 package). R Package Version 2.0-6. 〈http://CRAN.R-project.org/package=lmerTest〉. has been suggested that birds may be limited primarily by habitat Leighton, F.A., 1993. The toxicity of petroleum oils to birds. Environ. Rev. 1, 92–103. availability on wintering grounds, which leads to intense competition Lozano, G.A., Perreault, S., Lemon, R.E., 1996. Age, arrival date and reproductive success of male American Redstarts Setophaga ruticilla. J. Avian Biol. 27, 164–170. for winter territories and advantages to arriving early on the wintering Maggini, I., Kennedy, L.V., Macmillan, A., Elliott, K.H., Dean, K.M., Guglielmo, C.G., grounds (Stutchbury, 1994; Norris et al., 2003). Based on the flight 2017a. Light oiling of feathers increases flight energy expenditure in a migratory duration data from this study, birds impacted by the DWH oil spill shorebird. J. Exp. Biol (in press). Maggini, I., Kennedy, L.V., Macmillan, A., Elliott, K.H., Dean, K.M., Guglielmo, C.G., would likely have arrived late to their wintering grounds and would 2017b. Trouble on takeoff: crude oil on feathers reduces escape performance of have been forced to occupy poor quality habitat that may likely have shorebirds. Ecotoxicol. Environ. Saf. 141, 171–177. Marra, P.P., Hobson, K.A., Holmes, R.T., 1998. Linking winter and summer events in a resulted in overwinter mortality. In addition, research has shown that migratory bird by using stable-carbon isotopes. Science 282, 1884–1886. birds wintering in poor quality habitats arrive later to their breeding Moye, J.K., Pritsos, C.A., 2010. Effects of Chlorpyrifos and Aldicarb on flight activity and grounds the next year in poorer physical condition (Marra et al., 1998; related cholinesterase inhibition in homing pigeons, Columba livia: potential for migration effects. Bull. Environ. Contam. Toxicol. 84, 677–681. Harrison et al., 2011). Thus, oil exposed birds that arrived late to their Moye, J.K., Perez, C.R., Pritsos, C.A., 2016. Effects of parental and direct methylmercury wintering habitat after the DWH oil spill most likely exhibited oil-in- exposure on flight activity on young homing pigeons (Columba livia). Environ. Pollut. – duced carry over effects to the next breeding season. 5, 23 30. Norris, D.R., Marra, P.P., Kyser, T.K., Sherry, T.W., Ratcliffe, L.M., 2003. Tropical winter habitat limits reproductive success on the temperate breeding grounds in a migratory Author contributions bird. Proc. R. Soc. Lond. B 271, 59–64. O’Hara, P.D., Morandin, L.A., 2010. Effects of sheens associated with offshore oil and gas development on the feather microstructure of birds. Mar. Pollut. Bull. 60, 672–678. J.K.M., C.R.P, K.M.D. and C.A.P. designed experiments, J.K.M. and Perez, C.R., Moye, J.K., Pritsos, C.A., 2014. Estimating the surface area of birds: using the C.R.P performed experiments, D.C. analyzed data, and J.K.M., C.R.P, homing pigeon (Columba livia) as a model. Biol. Open 3 (6), 486–488. R Core Team, 2016. R: A language and environment for statistical computing. R and C.A.P wrote the paper. Foundation for Statistical Computing, Vienna, Austria. URL 〈https://www.R-project. org/〉. Competing financial interests Sandberg, R., Moore, F.R., 1996. Fat stores and arrival on the breeding grounds: re- productive consequences for passerine migrants. Oikos 77, 577–581. Smith, R.J., Moore, F.R., 2005. Arrival timing and seasonal reproductive performance in a The authors declare that they have no competing financial, profes- long distance migratory landbird. Behav. Ecol. Sociobiol. 57, 231–239. fl Steiner, I., et al., 2000. A GPS logger and software for analysis of homing in pigeons and sional or personal interests that might have in uenced the performance small mammals. Physiol. Behav. 71, 589–596. or presentation of the work described in this manuscript. Stutchbury, B.J., 1994. Competition for winter territories in a neotropical migrant: the role of age, sex, and color. Auk 111, 63–69. USFWS, 2011. Deepwater Horizon bird impact data from the DOI‐ERDC NRDA database Acknowledgements 12 May 2011. Wallraff, H.G., 2005. Avian Navigation. In: Pigeon homing as a Paradigm. Springer, Berlin. This work was funded in part by the U.S. Fish and Wildlife Service Wiltschko, R., Wiltschko, W., 2003. Avian navigation: from historical to modern concepts. as part of the Deepwater Horizon Natural Resource Damage Assessment Anim. Behav. 65, 257–272. and the Nevada Agricultural Experiment Station. The authors would like to thank Katie McGlamery and Breanna Sage for help in performing the experiments.

103 Ecotoxicology and Environmental Safety 146 (2017) 104–110

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Body mass change in flying homing pigeons externally exposed to Deepwater MARK Horizon crude oil ⁎ Cristina R. Pereza, John K. Moyea, Dave Cacelab, Karen M. Deanb, Chris A. Pritsosa, a Department of Agriculture, Nutrition, and Veterinary Sciences, University of Nevada, Reno, USA b Abt. Environmental Research, USA

ARTICLE INFO ABSTRACT

Keywords: The Deepwater Horizon oil spill contaminated thousands of miles of habitat valuable to hundreds of species of Deepwater Horizon migratory and resident birds of the Gulf of Mexico. Many birds died as a direct result of the oil spill; however, the Flight performance indirect effects of oil exposure on the flight ability and body condition of birds are difficult to assess in situ. This Homing pigeon study utilizes the homing pigeon as a surrogate species for migratory birds to investigate the effect of multiple Migration external oil exposures on the flight performance and body mass change of birds over a series of repeated flights Weight change from 136.8 km flight distance. Oiled pigeons took significantly longer to return home, lost more weight during flight, and were unable to recover their weight, resulting in reduction of body weight overtime. Based on our data, migratory birds that were oiled, even partially, by the Deepwater Horizon oil spill likely took longer to complete migration and were likely in poor body condition, increasing their risk of mortality and reproductive failure.

1. Introduction time and fuel deposition time (Alerstam, 2003), which are a direct result of flight ability and rate of weight gain during stopovers. Thus, it The Deepwater Horizon (DWH) oil spill released approximately 134 is important to understand how oil exposure impacts weight loss and million gallons of crude oil into regions ecologically important for birds flight ability concurrently over time, because perturbations in one or (DWH Trustees, 2016). Hundreds of bird species use the northern Gulf both of these phases of migration will affect overall migration speed, of Mexico all year long either as resident species or for various periods determining arrival time to destination sites. of the year for breeding, stop-over, or wintering as migratory species. Ingestion of oil through stimulated preening causes toxicological The location of the Gulf of Mexico intersects three of the four major effects such as gastrointestinal irritation (Friend and Franson, 1999). migratory flyways in North America and the extended persistence of oil Ingestion of oil as well as oil-induced alterations in behavior can result released by the spill overlapped with migration seasons in 2010 and in changes in the body condition of oiled birds. The subacute effects of 2011. Thus, spring and fall migrants as well as resident birds that utilize oil exposure on the body condition of avian species have been examined the area were potentially impacted by the oil spill. in a number of studies (Anderson et al., 2000; Burger and Tsipoural, The acutely lethal effect of external oiling on birds has been well 1998; Goldsworthy et al., 2000; Golet et al., 2002; Harris et al., 2011; documented. The single most devastating effect of oil on birds has been Leighton, 1986). Oil ingestion has caused a reduction in growth rate of considered to be its effect on feathers, affecting both thermoregulation young birds in a variety of species (Peakall, 1983; Szaro et al., 1978) and buoyancy and often resulting in mortality (Leighton, 1993). and can lead to retardation of feather growth (Crocker et al., 1975; However, we have little understanding of how the presence of small Hoffman, 1979; Ellenton, 1982). In addition, data from rehabilitation quantities of oil on the feathers of birds affects their ability to fly. Flight, efforts report that there is a strong negative correlation between body such as long distance migration, is energetically expensive and ade- weight and mortality of oiled birds (Frink and Miller, 1995; Jenssen, quate body condition is imperative to its successful completion. Birds in 1994). However, the effect of oil on weight loss has been hard to good condition (high body mass, high fat stores) are more likely to determine in wild birds. Birds in the wild show great variability in body complete migration early and increase breeding success compared to mass between sexes and ages and body mass changes markedly with birds in poor condition (low body mass, low fat stores) (Bêty et al., season (Harris et al., 2011). Additionally, the stress of bringing wild 2003; Kokko, 1999). The total duration of migration is the sum of flight birds into captivity leads to substantial weight loss in even unoiled,

⁎ Corresponding author. E-mail address: [email protected] (C.A. Pritsos). http://dx.doi.org/10.1016/j.ecoenv.2017.05.012 Received 25 July 2016; Received in revised form 5 May 2017; Accepted 8 May 2017 Available online 16 May 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 104–110 presumed healthy birds (Anderson et al., 2000). These factors make during April of 2014. Experimental flight 1 was conducted at site 9, comparisons of body mass between oiled and unoiled wild birds experimental flight 2 at site 2, experimental flight 3 at site 3, difficult. experimental flight 4 at site 2, and experimental flight 5 at site 3 In this study, we used captive homing pigeons (Columba livia)to (Fig. 1). Release times were noted and recorded for each of the two examine the effect of multiple external oil (MC 252 crude oil) exposures flocks at each release. on flight performance and body weight change during repeated flights from a 136.8 km flight distance. Birds in the wild near an oil spill are likely to be oiled on multiple occasions, thus the oil exposures of this 2.1. Oiling application study were designed to mimic a multiple oiling scenario. We predicted that the presence of oil on the flight feathers would cause flight to be External oiling of the birds occurred at the release site before release more energetically expensive and as a result be reflected in greater for each experimental flight. Artificially weathered crude oil collected weight loss during flight compared to unoiled bird flights and cause a from the Deepwater Horizon MC252 spill was transported to the release change in body mass over time. site in a Koolatron© (Brantford, Ontario) and maintained at 70°F (21 °C). Once arriving at the release site, a bird was randomly selected 2. Materials and methods and oil was applied on the primary wing and tail feathers following a standardized oiling regime covering 20% of the total body surface area The methods used in this study were carried out in accordance with (Perez et al., 2014). Based on the different oiling rate categories set in approved guidelines. These guidelines were approved by the place by the U.S. Fish and Wildlife Service (DWH Trustees, 2016), 20% Institutional Animal Care and Use Committee at the University of surface area coverage corresponded to the lightly oiled category. For Nevada, Reno (protocol # 00541). Birds were provided with Nutriblend each wing, a length of 5.08 cm was measured along the dorsal side of Green Pigeon Chow (Purina Mills) and water ad libitum while being the first primary, and from the end of that length a line was drawn to housed. the tip of the last primary. The area created was then painted dorsally The two groups of birds used in this study were housed at two with a 2.5 cm bristle paintbrush. For the tail, a 1.9 cm band of oil was different sites to avoid any cross contamination. The oiled birds were applied along the tips of the primary tail feathers. Oiling procedures housed in a pigeon loft located at the Agricultural Experiment Station at were repeated with each randomly selected bird until all birds within the University of Nevada, Reno. The control birds were housed in a the group were oiled. During oiling, the bristle paintbrush was pigeon loft located at the University's Mainstation Farm in Sparks, saturated with oil such that the bristles did not come into contact with Nevada (approx. 1.6 air kilometers apart). Both groups of birds were the feathers. There was concern that running the bristles of a clean initially trap trained to learn how to use the doors needed to enter the paintbrush on the feathers of the control birds would create an artificial lofts as well as to become familiar with the local surroundings around disruption of feather structure, thus, no sham treatment was applied to the loft. The birds were then trained at gradually increasing distances the control birds. All of the birds in this study were accustomed to being from the loft until a final distance of 136.8 km was reached. At the regularly handled and the additional act of painting was not expected to 32.2 km training distance, a dummy weight (18–20 g) was attached to bias the results. The additional handling time of each bird in the oiled each bird in order to mimic the weight of the GPS data loggers during group before each experimental flight was approx. 5–7 min. flight (approx. 10 g). After the birds were trained to return home from a distance of 136.8 km, three baseline flights (BF) were conducted to establish 2.2. Flight duration reference flight times for each of the birds to be used during the experimental flights. Sixteen potential controls and fifteen potential Flight duration was determined from the release time recorded at oiled birds remained after training and were available for baseline the release site and the arrival time of each bird recorded by the GPS flights. Baseline flights occurred in March 2014, every third day if dataloggers within an 80–100 m radius around the home loft using weather permitted. Feed was removed at 1 pm PST the day prior to the mapping software (Fugawi Global Navigator Moving Map Software, flight to increase motivation for the birds to return home in a timely Toronto, Canada). Logger data were filtered to remove any information manner. The morning of the flight, birds were hand trapped and collected by the GPS devices before the official release time and after weighed to the nearest gram, dummy weights were removed and GPS the official arrival time. If a bird's arrival was after the datalogger dataloggers (Steiner et al., 2000) were attached, and then the birds stopped recording because of limited battery life (approx. 12 h), arrival were loaded into a crate for transport to the release site. Time and date, at the home loft was not recorded by the GPS data loggers and flight bird band number, colored band, GPS datalogger number, and pre- time was estimated as follows. For birds that arrived the next day and flight weight were recorded. Upon arrival at the release site, GPS were found at the loft the following morning, flight times were dataloggers were turned on, and birds were given a 20 min resting estimated as the time between release and the time of sunset plus one period prior to release. The second of the two flocks was released hour, plus the time between sunrise the next morning and the observed 10 min after the first. A shorter wait period was given if the prior flock arrival time. This calculation assumed that the birds would not have released had vanished before the end of the 10 min wait period. In flown between sunset plus one hour and sunrise because pigeons are attempts to keep the birds from becoming exceedingly comfortable with not active during dark hours. We recognize that this calculation adds the same release site, the birds were released from different sites during error to the flight duration data for birds that did not arrive the day of both baseline and experimental flights. A series of release sites was release. This error would be an hour or two when the birds could have established along a 1.6 km road running east and west at 136.8 km from arrived and no one was present at the loft to observe actual arrival time. the home lofts. Baseline flight 1 was conducted at site 9, baseline 2 at However, this estimation method only had to be used 5 times out of 147 site 2, and baseline 3 at site 6 (Fig. 1). times total. Thus, we believe this is a reasonable estimation of arrival After the three baselines flights, four experimental flights (EF) were time in order to include this data for birds that did not arrive the same conducted. Eleven control birds and eleven oiled birds remained after day. On experimental flight 1 the control birds chose not to leave the baseline flights and participated in experimental flights. Day of and day area of release, causing half of the control birds to return the next day. prior pre-flight procedures for experimental flights were similar to From a flight distance of 138.6 km, extended flight times of this nature those conducted during baseline flights. Before each of the experi- are uncharacteristic; and for this reason, data from experimental flight mental flights, oil was applied to each bird in the oiled group at the 1 for both groups were removed from all analyses. release site. A total of five experimental flights occurred every third day

105 C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 104–110

Fig. 1. Geographic extent of release site (north) and home lofts (south). Map inset includes release sites over the course of the study. The birds were randomly released from 4 different release sites within a 1.6 km distance in attempts to avoid a loss in motivation to leave the release site due to familiarity. Baseline flight (BF) 1 and experimental flight (EF) 1 were released from site 9; BF2, EF2, and EF4 were released from site 2; BF3 was released from site 6; and EF3 and EF5 were released from site 3.

2.3. Bird weight mixed effects models (LMM) testing for the effect of added explanatory variables on the response variable. Our full model included flight During all flights (baseline and experimental), birds were weighed duration as the response variable, and main effects of flight period on flight days when crated for transport to the release site (pre-flight (baseline, pre-oiled or experimental, post-oiled), treatment group (oil or weight) and immediately after returning from a flight (post-flight control), and the interaction of flight period and treatment group. weight) before consumption of feed or water. Birds that arrived the Individual birds (n=22) were included as a random effect to control for next day were weighed when they were found at the loft, also before repeated measures. To test for effects of oiling on weight lost during having access to feed or water at the loft. A final weight for all of the flight, we created mixed effects models with weight loss as the response remaining birds was measured the day of necropsy (three days after the variable and the same fixed and random effects as those described for last flight). Mass loss during flight was calculated as the difference flight duration. Additionally, to determine if multiple oil exposures between the mass before a flight (pre-flight weight) and the mass after a caused weight change over time we looked at how pre-flight body flight (post-flight weight). weight changed with subsequent experimental flights. We ran a linear mixed effects model with a repeated measures structure with experi- 2.4. Feed consumption mental flight number modeled as a continuous variable and treatment group as a categorical variable. The full regression model included Feed consumption was measured on a group basis every day starting flight number, treatment group, and the interaction between treatment the day of the first baseline flight and ending the day after the last group and flight number. Individual birds were included as random experimental flight. Measurements were taken between 8am and 10am effects allowing for random intercept. Since EF1 was removed from PST each day. Daily individual feed consumption was calculated as analyses, only data from EF2 to EF5 were included (n=4 experimental grams of feed consumed per day divided by the number of birds in the flights). loft on that day. Best fit models were selected based on AIC comparison and the explanatory parameters included in the final model were interpreted as fi ff ff 2.5. Statistical methods having a signi cant e ect on the response variable. E ect sizes were determined based on differences of least square means between groups ff We tested for effects of oiling on flight performance by analyzing (BF ctrl, BF oil, EF ctrl, EF oil). E ect sizes were considered to be fi fi differences in flight duration for unoiled birds and oiled birds during signi cant if their 95% con dence intervals did not overlap zero. Data fl experimental flights and the change in flight duration from baseline to from birds that participated in both baseline (n=3) experimental ights experimental flights within treatment groups. We compared linear (n=4) are included in the analyses, 11 birds in each group. One control

106 C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 104–110

Table 1 Table 2 Effect sizes and 95% confidence intervals between treatment groups and flight periods for Effect sizes and 95% confidence intervals between treatment groups and flight periods for flight duration. Estimates were derived from the best fit linear mixed effects model weight loss during flight. Estimates were derived from the best fit linear mixed effects including treatment group, flight period, and treatment group*flight period and bird ID as model including treatment group, flight period, and treatment group*flight period and a random effect (n=22). Confidence intervals that do not overlap zero are shown in bold. bird ID as a random effect (n=22). Confidence intervals that do not overlap zero are shown in bold. Effect 95% CI Effect 95% CI BF oil - BF ctrl −119 (−230.30, −7.7) BF ctrl - EF ctrl 87.3 (−4.84, 179.5) BF oil - BF ctrl −3.2 (−9.92, 3.50) BF oil - EF oil −142.7 (−237.71, −47.6) BF ctrl - EF ctrl 0.3 (−4.96, 5.56) EF oil - EF ctrl 111 (7.10, 214.9) BF oil - EF oil −11.4 (−16.84, −5.91) EF oil - EF ctrl 8.5 (2.11, 14.81) bird and two oiled birds were lost during experimental flights. All available data from these birds up until day lost were included in 3.2. Weight change analyses. All mixed effects models were fit by maximum likelihood and were analyzed using packages lme4 (Bates et al., 2013) and lmerTest Weight loss during flight was found to be significantly affected by (Kuznetsova et al., 2014) in R version 3.2 (R Core Team, 2016). Degrees the main effects of treatment group, flight period, and the interaction of freedom were obtained using Satterthwaite approximation in lmerT- between treatment group and flight period. There was no difference est. The outputs of only the final models are discussed. between the two groups in weight loss during baseline flights (Table 2). To determine differences in feed consumption by treatment group, During experimental flights the oiled birds lost significantly more feed consumption data were assessed using t-tests to compare flight day weight during flight than the control birds. Furthermore, the oiled and non-flight day averages. We interpreted a p-value of less than 0.05 birds lost an additional 11.4 g during their experimental flights than as statistically significant. during their unoiled flights, while there was no corresponding change in weight loss between flight periods of the control birds (Fig. 3). The birds in both groups were able to maintain their pre-flight 3. Results weights throughout the baseline flights (Fig. 4). However, there was a significant time effect on the pre-flight weights between groups during 3.1. Flight duration experimental flights. The best fit model for pre-flight weight during experimental flights included treatment group, flight number, and the The final model for our flight duration data included main effects of interaction between treatment group and flight number. The pre-flight flight period, treatment group and the interaction between flight period weights of the oiled birds during experimental flights decreased over and treatment group. During baseline flights, the control birds per- time (β=−1.55, SE=1.11), whereas the pre-flight weights of the formed unusually poorly, resulting in a large difference between the control birds increased over time (β=7.54, SE=1.52). The slopes two groups (Table 1). During experimental flights, the control birds obtained from this model reflected weight change from EF2 to EF5 decreased their average flight duration approaching the level of the because this analysis did not include EF1. Both groups decreased weight oiled birds baseline flights (Fig. 2). The oiled birds were significantly from EF1 to EF2 (Fig. 4, flights 4 and 5). Thus, the regression lines fit slower than the control birds during experimental flights by 111 min. from the LMM model (Fig. 5) captured the difference in weight recovery More importantly, there was a significant difference in flight duration that occurred over time after both groups had initially experienced a between flight periods of the oiled birds, as they took almost twice as reduction in their body weight from EF 1 and EF2. long to complete their experimental flights compared to their baseline fl ights. 3.3. Feed consumption

There were noticeable trends in daily feed consumption in both

Fig. 2. Point estimates and 95% CI of flight duration by treatment group and flight Fig. 3. Point estimates and 95% CI of weight loss during flight by treatment group and period. Estimates from a 3 parameter LMM, including treatment group, flight period, and flight period. Estimates from a 3 parameter LMM, including treatment group, flight treatment group*flight period and bird ID as a random effect. Both treatment groups period, and treatment group*flight period and bird ID as a random effect. Both treatment consisted of 11 birds and flew a total of 3 baseline flights and 4 experimental flights from groups consisted of 11 birds and flew a total of 3 baseline flights and 4 experimental a distance of 136.8 km. BF control: 294 min (209, 379); BF oil: 175 min (90,260); EF flights from a distance of 136.8 km. BF control: 37.8 g (33.1, 42.5); BF oil: 34.6 g (29.8, control: 207 min (127, 286); EF oil: 318 min (242, 393). 39.3); EF control: 37.5 g (33.1, 41.8); EF oil: 46.0 (41.3, 50.6).

107 C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 104–110

a

b

Fig. 4. Group average pre-flight body mass over the three phases of the experiment; baseline flights, experimental flights, and necropsy (mean ± SEM). Data from experi- mental flight 1 shown for visual purposes and were not included in analysis.

Fig. 6. A) Average feed consumption on individual bird basis for each day of the flight study. Arrows indicate days that flights were conducted. Baseline flights were conducted Fig. 5. Individual preflight weights for the control (unfilled symbols, solid line) and oiled from 03/30/2014 to 04/06/2014. Experimental flights were conducted 04/09/2014 to (filled symbols, dashed line) birds during experimental flights 2–5 (EF1 removed from 04/21/2014. Experimental flight 1 was conducted on 04/09/2014(not included in analysis). Regression lines fit to a 3 parameter mixed effects model including flight analysis). Feed was removed at 1:00 pm PST on the days before flight. B) Comparison number, treatment group, and flight number*treatment group as fixed effects and of average feed consumption by group for baseline (BF) flight days, experimental (EF) individual birds as random effects. Control: y=7.54x+407; Oil: y=−1.55x+422. flight days, and non-flight days (mean ± SEM). On non-flight fasting days (days before flights), feed was removed at 1:00 pm PST. *p=0.045 versus control BF non-flight day 1, fl fl groups (Fig. 6A). During the period of the three baseline flights, both **p=0.006 versus control EF ight days, ***p < 0.001 versus control EF non- ight day 1, t-tests. groups showed similar patterns; the day before flight had the lowest amount of consumption and a large increase in consumption occurred feathers. The oiled birds in our study likely also experienced higher the day of flight, and the highest consumption levels occurred the day energy costs of flights and consequently took longer to complete oiled after flight. As shown in Fig. 6A, the control birds consumed more feed flights. The suggested increased energy expenditure of flight is likely that the oiled birds during the pre-oiled period. During the post-oiling reflected in the increased mass lost during experimental flights. period, this difference was greater, and the control birds consumed Although the observed weight loss cannot be precisely partitioned into significantly more food than oiled birds on flight days (t(6)=4.12, fat, water, and protein loss, this data is suggestive of increased energy p=0.006) and the days immediately following flights (t(6)=6.16, expenditure during flight as oiled birds would have used more fuel p < 0.001) (Fig. 6B). Throughout the experiment, feed consumption during flights. on the days before flight (feed removed at 1 PST) remained similar in The birds in both groups were in good body condition at the both groups. beginning of the study and were able to maintain their starting weight throughout the baseline flights (Fig. 4). Both groups then experienced a 4. Discussion drop in weight from EF1 to EF2. However, the two treatment groups recovered from this reduction in weight differently over the course of In this study, we found that oiled pigeons almost doubled the the experimental flights. The significant difference in the slopes of the amount of time it took to complete 136.8 km flights compared to their lines fit to pre-flight body weight data suggests that the control birds pretreatment flights. In association with reduction in flight perfor- were able to recover their weight after experiencing a decrease mance, oiled-treated birds had reduced body mass. The oiled birds not (positive slope), whereas the oiled birds were not able to recover their only had greater body mass loss during flight, but also did not recover weight after experiencing a decrease (negative slope) (Fig. 5). Instead, the weight lost, leading to an overall decrease in body mass over the negative trend continued even after the last experimental flight, as experimental flights. We suggest that the reduction in flight perfor- the oiled birds continued to lose weight until the day of necropsy (3 mance and associated increase in weight loss was partially a result of days after the last experimental flight; Fig. 4). The inability of the oiled higher costs of flight due to oiled flight feathers. Results from a recent birds to recover their weight resulted in an 11% reduction in body study has showed that oiled Western Sandpipers (Calidris mauri) flying weight compared to their starting weight. in a wind tunnel spent more energy during 2 h endurance flights than In the time between flights, food and water were provided ad non-oiled birds manifested by increased loss of fat mass and body mass libitum and thus full recovery of weight to the next flight would be during flight (Maggini et al., 2017). This was attributed to the expected. The ability of the birds in both groups to maintain their impairment of flight dynamics caused by the negative effect the weight during baseline flights and the ability of the control birds to physical presence of oil has on the aerodynamic properties of flight recover their weight during experimental flights supports that a two

108 C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 104–110 day period between 136.8 km flights is sufficient time for the birds to to leave without adequate energy stores and continuing migration at rest and refuel. It should be noted that our weight loss analysis included suboptimal body weight reduces the probability of successfully com- all birds. Given the free-flying aspect of this experimental design, it pleting migration as well as reduces the probability of successfully cannot be verified whether birds eat or drink while on route. However, reproducing once they arrive at the breeding grounds (Kokko, 1999). the mass loss data showed that there was only one instance in which a Additionally, late arrivals to breeding grounds have been shown to bird arrived at the loft with a heavier weight than when it was released. suffer reproductive impacts such as reduced clutch size and lower This suggests that although it is possible for our birds to consume hatching rates (Smith and Moore, 2005; Cooper et al., 2011; Gordo resources during flight, it did not occur regularly. et al., 2013; Descamps et al., 2011, Baker et al., 2004). Our data The repeated application of oil to the birds in the oil group did not indicate that oiled birds alter their flight performance and lose weight allow for the oil to dry and instead the oil remained in the liquid state immediately after a single oil exposure. Additional oil exposures further throughout the duration of the experimental flights, facilitating inges- contribute to these effects. Given that birds living in the Gulf of Mexico tion of oil through preening. The ingestion of oil likely contributed to area during the DWH oil spill were likely oiled multiple times, it is the decreased body mass. Studies have shown that ingestion of oil can possible that oil-impacted birds experienced similar effects after result in irritation of the gastrointestinal tract manifested by lesions and exposure, slowing down migration and subsequently affecting survival bleeding (Hartung and Hunt, 1966; Fry and Lowenstine, 1985) and GI and reproductive success. edema (Bursian et al., 2015). Damage to the gastrointestinal tract can prevent efficient nutrient absorption and likely cause animals to lose Author contributions weight and become weak (Frink and Miller, 1995). It has been estimated that as much as 50% of the external oil burden may be J.K.M., C.R.P, K.M.D. and C.A.P. designed experiments, J.K.M. and ingested through preening within the first week after exposure C.R.P performed experiments, D.C., C.R.P., and J.M. analyzed data, and (Hartung, 1963). Oil was applied to the pigeons in the oil group 5 J.K.M., C.R.P., and C.A.P. wrote the paper. different times at a dose of approximately 2 ml per application. The ingestion of oil through preening would likely be additive after every Competing financial interests exposure, and based on the estimation made by Hartung (1963),itis possible that the oiled pigeons in this study consumed as much as 5 ml The authors declare that they have no competing financial, profes- over a 13 day period. Thus, the combined stress of energy expensive sional or personal interests that might have influenced the performance flight and oil ingestion from multiple oil exposures likely enhanced the or presentation of the work described in this manuscript. weight loss of the oiled pigeons in this study. Additionally, the cumulative amount of oil applied to the body of the birds could have Acknowledgements induced thermoregulatory effects by the end of the study. It is well known that oiled birds increase their metabolic rate to maintain body This work was funded in part by the U.S. Fish and Wildlife Service temperature (reviewed in Leighton, 1993 and Jenssen, 1994) and may as part of the Deepwater Horizon Natural Resource Damage Assessment have contributed to the reduction of body weight observed in the oiled and the Nevada Agricultural Experiment Station. The authors would birds even after the flights had stopped. like to thank Katie McGlamery and Breanna Sage for help in performing An experimental study on the oiling of the plumage of sanderlings the experiments. The authors would also like to thank James Sedinger (Calidris alba) found that oiled birds declined in body mass immediately for helpful advice in data analysis. after a 20% oiling and continued to lose weight 14 days post oiling (Burger and Tsipoural, 1998). The weight loss observed in the sander- References lings study was attributed in part to changes in behavior, such as a reduction in feeding and resting and an increase in the amount of time Akkeson, S., Hedenstrom, A., 2007. How migrants get there: migratory performance and spent preening and bathing. Although this behavior was not quantified orientation. Bioscience 57, 123–133. http://dx.doi.org/10.1641/B570207. Alerstam, T., 2003. Bird migration speed. Avian Migration. Springer Berlin Heidelbergpp. in our study, the oiled birds were observed to increase preening and 253–267. there is evidence of reduced feeding after oiling particularly on days Alerstam, T., Lindström, Å., 1990. Optimal bird migration: the relative importance of after flight (Fig. 6A and B). This suggests that a combination of time, energy, and safety. Bird Migration. Springer Berlin Heidelbergpp. 331–351. Anderson, D.W., Newman, S.H., Kelly, P.R., Herzog, S.K., Lewis, K.P., 2000. An behavioral changes as well as physiological impacts likely contributed experimental soft-release of oil-spill rehabilitated American coots (Fulica americana): to the overall weight loss observed in the oiled pigeons of this study. I. Lingering effects on survival, condition, and behavior. Environ. Poll. 107, 285–294. http://dx.doi.org/10.1016/S0269-7491(99)00180-3. fi 4.1. Implications for migration Baker, A.J., et al.2004. Rapid population decline in red knots: tness consequences of decreased refuelling rates and late arrival in Delaware Bay. Proceedings R. Soc. Lond. B. 271, pp. 875–882. DOI: http://dx.doi.org/10.1098/rspb.2003.2663. In this study, we found that oiled birds took longer to complete Bates, D., Maechler, M., Bolker, B., Walker, S., 2013. lme4: linear mixed-effects models – 〈 flights, lost more weight during flight, and were unable to recover using Eigen and S4. R. Package Version 1, 0 5. http://CRAN.R-project.org/ package=〉 (lme4). weight like the unoiled birds resulting in reduced body weight. These Bêty, J., Gauthier, G., Giroux, J.-F., 2003. Body condition, migration, and timing of data have important implications for migrating birds exposed to oil. reproduction in snow geese: a test of the condition-dependent model of optimal – The impact of oil on flight duration and body condition could have clutch size. Am. Nat. 162, 110 121. ff Burger, J., Tsipoural, N., 1998. Experimental oiling of sanderlings (Calidris alba): behavior direct e ects on the overall speed of migration. Migration speed is and weight changes. Environ. Tox. Chem. 17, 1154–1158. http://dx.doi.org/10. determined by three variables: speed of flight, rate of energy accumula- 1002/etc.5620170623. tion, and rate of energy consumption (Akkeson and Hedenstrom, 2007). Bursian, S., Harr, K., Cacela, D., Cunningham, F., Dean, K., Hanson-Dorr, K., Horak, K., Link, J., Pritsos, C., 2015. Deepwater Horizon Avian Toxicity Phase 2: Double Crested However, migration speed is mostly dependent on refueling rate Cormorant (Phalacrocorax auritus) Oral Dosing Study (M22). DWH Birds NRDA (Alerstam and Lindstrom, 1990), as approximately 90% of the migra- Technical Working Group Report. tory period is spent at stopover sites (Hedenstrom and Alerstam, 1998). Cooper, N.W., Murphy, M.T., Redmond, L.J., Dolan, A.C., 2011. Reproductive correlates fl of spring arrival date in the Eastern Kingbird Tyrannus tyrannus. J. Ornithol. 152, Based on our data, oiled birds would have a reduced speed of ight, 143–152. http://dx.doi.org/10.1007/s10336-010-0559-z. decreased rate of energy accumulation, and increased rate of energy Core Team, R., 2016. R: a language and environment for statistical computing. R. Found. consumption during flight. Together, this would result in an overall Stat. Comput., Vienna, Austria (URL). 〈https://www.R-project.org/〉. ff slower migration speed as birds would need more time to complete Crocker, A.D., Cronshaw, J., Holmes, W.N., 1975. The e ect of several crude oils and some petroleum distillation fractions and intestinal absorption in ducklings (Anas flights and spend more time accumulating resources at stopover sites. If platyrhychos). Environ. Physiol. Biochem 5, 92–106. oiled birds cannot efficiently accumulate resources, they may be forced Deepwater Horizon Natural Resource Damage Assessment Trustees, 2016. Deepwater

109 C.R. Perez et al. Ecotoxicology and Environmental Safety 146 (2017) 104–110

Horizon oil spill: Final Programmatic Damage Assessment and Restoration Plan and 2000.9674235. Final Programmatic Environmental Impact Statement (Retrieved from). 〈http:// Hartung, R., 1963. Ingestion of oil by waterfowl. Mich. Acad. Sci. 48, 49–55. www.gulfspillrestoration.noaa.gov/restoration-planning/gulf-plan〉. Hartung, R., Hunt, G.S., 1966. Toxicity of some oils to waterfowl. J. Wildl. Manag. 30, Descamps, S., Bêty, J., Love, O.P., Gilchrist, H.G., 2011. Individual optimization of 564–570. reproduction in a long-lived migratory bird: a test of the condition-dependent model Hedenström, A., Alerstam, T., 1998. How fast can birds migrate? J. Avian Biol. 29, of laying date and clutch size. Funct. Ecol. 25, 671–681. http://dx.doi.org/10.1111/j. 424–432. 1365-2435.2010.01824.x. Hoffman, D.J., 1979. Embryotoxic and teratogenic effects of crude oil on mallard embryos Ellenton, J.A., 1982. Teratogenic activity of aliphatic and aromatic fractions of Prudhoe on day one of development. Bull. Environ. Contam. Toxicol. 22, 632–637. Bay crude and fuel oil no. in the chicken embryo. Toxicol. Appl. Pharmacol. 63, Jenssen, B.M., 1994. Review article: effects of oil pollution, chemically treated oil, and 209–215. http://dx.doi.org/10.1016/0041-008X(82)90043-6. cleaning on the thermal balance of birds. Environ. Poll. 86, 207–215. Friend, M., Franson, J.C., 1999. Oil. In: Field Manual of Wildlife Diseases: General Field Kokko, H., 1999. Competition for early arrival in migratory birds. J. Anim. Ecol. 68, Procedures and Diseases of Birds. US Geological Survey. pp. 309–315. 940–950. Frink, L., Miller, E.A., 1995. In: Frink, L. (Ed.), Principles of oiled bird rehabilitation. Kuznetsova, A., Brockhoff, P.B., Christensen, R.H.B., 2014. lmerTest: tests for random and Wildlife and Oil Spills: Response, Research, and Contingency Planning, (ed.). Tri- fixed effects for linear mixed effect models (lmer objects of lme4 package). R. State Bird Rescue and Research, Inc., Newark, DE, pp. 61–68. Package Version (2.0-6). 〈http://CRAN.R-project.org/package=lmerTest〉. Fry, D.M., Lowenstine, L.J., 1985. Pathology of common murres and Cassin's auklets Leighton, F.A., 1986. Clinical, gross, and histological finding in herring gulls and Atlantic exposed to oil. Arch. Environ. Contam. Toxicol. 14, 725–737. puffins that ingested Prudhoe Bay crude oil. Vet. Pathol. 23, 254–263. Goldsworthy, S.D., Giese, M., Gales, R.P., Brothers, N., Hammill, J., 2000. Effects of iron Leighton, F.A., 1993. The toxicity of petroleum oils to birds. Environ. Rev. 1, 92–103. Baron oil spill on little penguins (Eudytula minor). II. post-release survival of Maggini, I., Kennedy, L.V., Macmillan, A., Elliott, K.H., Dean, K.M., Guglielmo, C.G., rehabilitated oiled birds. Wildl. Res. 27, 573–582. http://dx.doi.org/10.1071/ 2017. Light oiling of feathers increases flight energy expenditure in a migratory WR99076. shorebird. J. Exp. Biol (in press). Golet, G.H., et al., 2002. Long-term direct and indirect effects of the ‘Exxon Valdez’ oil Peakall, D.B., 1983. Toxicity of crude oils and their fractions to nestling herring gulls 1. spill on pigeon guillemots in Prince William Sound, Alaska. Mar. Ecol. Prog. Ser. 241, physiological and biochemical effects. Mar. Environ. Res. 8, 63–71. 287–304. http://dx.doi.org/10.3354/meps241287. Perez, C.R., Moye, J.K., Pritsos, C.A., 2014. Estimating the surface area of birds: using the Gordo, O., Tryjanowski, P., Kosicki, J.Z., Fulín, M., 2013. Complex phonological changes homing pigeon (Columba livia) as a model. Biol. Open 3 (6), 486–488. and their consequences in the breeding success of a migratory bird, the white stork Smith, R.J., Moore, F.R., 2005. Arrival timing and seasonal reproductive performance in a Ciconia ciconia. J. Anim. Ecol. 82, 1072–1086. http://dx.doi.org/10.1111/1365- long-distance migratory landbird. Behav. Ecol. Sociobiol. 57, 231–239. 2656.12084. Steiner, I., et al., 2000. A GPS logger and software for analysis of homing in pigeons and Harris, M.P., Wanless, S., Webb, A., 2011. Changes in body mass of common guillemots small mammals. Physiol. Behav. 71, 589–596. Uria aalge in southeast Scotland throughout the year: implication for the release of Szaro, R.C., Dieter, M.P., Heinz, G.H., 1978. Effects of chronic ingestion of South cleaned birds. Ringing Migr. 20, 134–142. http://dx.doi.org/10.1080/03078698. Louisiana crude oil on mallard ducklings. Environ. Res 17, 426–436.

110 Ecotoxicology and Environmental Safety 146 (2017) 111–117

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Reprint of: Trouble on takeoff: Crude oil on feathers reduces escape ☆ MARK performance of shorebirds ⁎ Ivan Magginia, ,1, Lisa V. Kennedya, Kyle H. Elliotta,2, Karen M. Deanb, Robert MacCurdyc,3, Alexander Macmillana, Chris A. Pritsosd, Christopher G. Guglielmoa a Department of Biology, Advanced Facility for Avian Research, University of Western Ontario, 1151 Richmond St., London, Ontario, Canada N6A 5B7 b Abt Associates, 1881 Ninth Street, Boulder, CO 80302, USA c Cornell University, Department of Mechanical & Aerospace Engineering, 105 Upson Hall, Ithaca, NY 14853, USA d Department of Agriculture, Nutrition and Veterinary Sciences, University of Nevada, 1664 N. Virginia St., Reno, NV 89557, USA

ARTICLE INFO ABSTRACT

Keywords: The ability to takeoff quickly and accelerate away from predators is crucial to bird survival. Crude oil can disrupt Oil spill the fine structure and function of feathers, and here we tested for the first time how small amounts of oil on the Calidris mauri trailing edges of the wings and tail of Western sandpipers (Calidris mauri)affected takeoffflight performance. In ff Takeo speed oiled birds, the distance travelled during the first 0.4 s after takeoff was reduced by 29%, and takeoff angle was High-speed video decreased by 10° compared to unoiled birds. Three-axis accelerometry indicated that oiled sandpipers produced Accelerometry less mechanical power output per wingbeat during the initial phase of flight. Slower and lower takeoff would ODBA make oiled birds more likely to be targeted and captured by predators, reducing survival and facilitating the exposure of predators to oil. Whereas the direct mortality of heavily-oiled birds is often obvious and can be quantified, our results show that there are significant sub-lethal effects of small amounts crude oil on feathers, which must be considered in natural resource injury assessments for birds.

1. Introduction when wing loading (the weight of the bird relative to its wing area) is high (Burns and Ydenberg, 2002; Ortega-Jiménez et al., 2010). External Escaping predators is one of the main survival tasks for animals. factors such as natural feather abrasion, breakage, or sun damage may Like most birds, migratory shorebirds have evolved behavioural tactics also reduce feather quality and takeoff performance. Feathers can be- to minimize predation risk. Shorebirds can time migration in order to come contaminated with crude oil during oil spills, and whereas the avoid the peak of migratory raptors on their journey (Ydenberg et al., inability of heavily-oiled birds to fly is often obvious, the potential for 2004), and they travel in flocks using dilution or the confusion effect to small amounts of oil to impair flight performance has not been studied. reduce an individual's chance of being killed (Cresswell, 1994). To be During the 2010 Deepwater Horizon (DWH) oil spill in the Gulf of effective, these behavioural tactics must be accompanied by the ap- Mexico, about 3.2 million barrels of crude oil were discharged in the sea propriate ability to fly and manoeuver. In particular, when an attack over an uninterrupted period of about three months (NOAA, 2015). The occurs, individuals that are slow or become separated from the flock are spill affected at least 25,000 km2 of marine habitat and over 2100 km of most vulnerable. Takeoff performance is therefore one of the major coastal habitat (NOAA, 2015) in the Gulf of Mexico region. Both re- aspects of predation avoidance for migratory shorebirds and other sident and transient birds, such as migratory shorebirds, were affected flocking birds. by the spill and their exposure to crude oil persisted long after the Difficulties during takeoff can occur when individual birds moult discharge from the compromised well was stopped (NOAA, 2015). flight feathers (Swaddle and Witter, 1997; Swaddle et al., 1999), or Previous studies have typically only considered the acute effects of oil

DOI of original article: http://dx.doi.org/10.1016/j.ecoenv.2017.03.026 ☆ A publisher's error resulted in this article appearing in the wrong issue. The article is reprinted here for the reader's convenience and for the continuity of the special issue. For citation purposes, please use; Ecotoxicology and Environmental Safety Volume 141 pp. 171-177 ⁎ Correspondence to: Konrad-Lorenz Institute of Ethology, University of Veterinary Medicine, Savoyenstrasse 1a, 1160 Vienna, Austria. E-mail address: [email protected] (I. Maggini). 1 Current address: Konrad Lorenz Institute of Ethology, University of Veterinary Medicine Vienna, Savoyenstrasse 1a, A-1160 Wien, Austria. 2 Current address: Department of Natural Resource Sciences, McGill University, 21111 Lakeshore Road, Ste Anne de Bellevue, Quebec, Canada H9X 3V9. 3 Current address: Massachusetts Institute of Technology, Computer Science and Artificial Intelligence Lab, 32 Vassar St, Cambridge, MA, 02139, USA. http://dx.doi.org/10.1016/j.ecoenv.2017.05.018 Received 11 August 2016; Received in revised form 18 March 2017; Accepted 20 March 2017 Available online 07 June 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 111–117 leading to rapid death, such as toxicity after ingestion and the reduced baseline flight without accelerometer (video only), baseline flight with insulation of oiled feathers (Peakall et al., 1982; Fry and Lowenstine, accelerometer, oiled (or sham) flight without accelerometer, and oiled 1985). During the DWH spill, tens of thousands of birds were estimated (or sham) flight with accelerometer. to have been directly killed, and several thousand live oiled birds were Supplementary material related to this article can be found online at also observed (NOAA, 2015). The majority of these birds were assigned http://dx.doi.org/10.1016/j.ecoenv.2017.03.026. to “trace” or “light” oiled categories (less than 5% and 5–20% of body Due to an unrecoverable data storage drive failure, the September surface, respectively). 2013 videos were lost before they could be analyzed. We analyzed the We quantified for the first time the effects of crude oil on takeoff accelerometer data from September 2013, but we waited until ability of birds. We hypothesized that birds with lightly-oiled wing and November 2014 to repeat the video recordings to allow the birds to tail feathers, as are commonly observed during oil spills, would have replace their feathers, and to measure them while they were in a similar reduced takeoff performance (slower speed and lower takeoff angle). migratory state. We repeated the time-matched control experiment (see We studied the effects on wings and tail because these are the major below) with the same individual birds that were studied in September surfaces involved in creating lift during flight (Thomas, 1997; 2013, except that in this case birds were measured over two days with Pennycuick, 2008), and we expect takeoff to be impacted when these the baseline followed by the experimental flight. surfaces are not fully functional, as in the case of oil contamination. We We followed a four-day protocol: all birds flew baseline flights (BF) used high-speed video and three-axis accelerometers to quantify the on day one (video only) and two (video and accelerometer), and then effects of feather oiling on takeoff of western sandpipers. High-speed were oiled or sham-treated on day three for their experimental flights video is a standard method used to measure takeoff speed and angle (EF, video only). On day four they flew an additional flight carrying (Lind et al., 2010). Accelerometers are used to measure parameters that accelerometers. Between day three and four the sandpipers were held are relevant to takeoff, such as overall dynamic body acceleration without access to bathing pools so their feathers remained oiled until (ODBA), which has been shown to indicate mechanical power output in tested on day four. In June 2013 all birds were oiled after their baseline a variety of animal species, including birds (Wilson et al., 2006; Halsey flights. In September 2013 and November 2014 we added a time-mat- et al., 2009; Elliott et al., 2013; Duriez et al., 2014). Measuring ODBA ched control group to exclude the possible effect of habituation to the allowed us to deepen our understanding of the energy requirements of experimental schedule. Accelerometers were only deployed in June takeoff in birds with flight feathers contaminated by crude oil. 2013 and September 2013.

2. Materials and methods 2.3. Application of crude oil to feathers

2.1. Study birds The oil applied to the birds from the oiled group was MC 252 oil collected during the 2010 DWH Gulf of Mexico oil spill and artificially Western sandpipers (family Scolopacidae) winter in the Gulf of weathered (TDI-Brooks International, College Station, TX) prior to re- Mexico in large numbers (Morrison et al., 1993; Nebel et al., 2002), and ceipt for use in the studies. Birds from the oiled group were oiled on were one of the species exposed to MC252 oil from the DWH spill 25% of the total surface of wings and tail. Oil covered the tip of the (NOAA, 2015). They are representative of other birds of similar size and primary feathers and tail feathers (Fig. 1). This level of oiling re- habitat requirements. presented approximately 20% of the total body surface (light oiling) as We captured western sandpipers near Roberts Bank and Boundary determined from study skins in advance of the study, however, in a ’ ’ Bay in Delta, British Columbia, Canada (49°04 N; 122°58 W) in July standing bird, this represented less than 5% of the visible body surface. 2012 and July 2013. Upon capture they were held for up to one week in animal facilities at Simon Fraser University (Burnaby, BC, Canada) ff before same-day shipment to Toronto, Ontario, Canada. They were then 2.4. Takeo experimental procedure transported by vehicle to the Advanced Facility for Avian Research fffl (AFAR) at the University of Western Ontario, London, Ontario, Canada We conducted the takeo ights in a large, brightly lit animal room and maintained in captivity until the experiments. that was sub-divided by temporary walls and white curtains into a test The birds were housed in specialized 2.4 m×3.7 m shorebird rooms arena (length 500 cm, width 310 cm, height 290 cm). At a release point under 16L:8D (16 h of light, 8 h of darkness) light conditions at ap- near a corner of the arena, each bird was placed in an opaque box 20 cm proximately 22 °C. They were fed an ad libitum diet of 80% Mazuri above the ground surface and approximately 30 cm from a wall to the Waterfowl Starter (Purina, Agribrands Purina Canada, Woodstock, ON, bird's left side. A high-speed video camera (Motion Pro X4 plus, Canada) and 20% trout chow (Aquamax Fingerling Starter 300, Grey Integrated Design Tools, Inc.) was positioned perpendicular to the re- ff Summit, MO, USA) supplemented with ~50 mealworms/20 birds every lease point and recorded the takeo s at 200 frames per second (fps). other day. During winter 2013 the light cycle was switched to 12L:12D The researcher waited until the bird positioned itself facing the long to simulate conditions on the winter range. In mid-April 2013 the light dimension of the arena (perpendicular to a side-view video camera and cycle was changed to 14L:10D to photostimulate the birds into a mi- gratory condition. The test in June 2013 was performed under these photoperiodic conditions. The birds captured in July 2013 were tested in September 2013 and the tests were performed when they were ex- periencing 16L:8D. During the winter 2013–2014 they went through the same photoperiodic changes described above, and additional tests were performed while the birds were experiencing 14L:10D.

2.2. Study design and schedule

The study was performed in three sessions: the first in June 2013 Fig. 1. Patterns of oiling for the experiments on Western sandpipers. Crude oil was ap- using birds caught in July 2012 (N=10 oiled), the second in September plied to the trailing edge of the wing beginning 2.3 cm from the tip of the outermost 2013 using birds caught in July 2013 (N=7 oiled, N=7 controls), and primary feather to the tip of the 10th primary feather, and along a 0.7 cm margin of the the third in November 2014 (N=7 oiled, N=6 controls). In June 2013 tail. Sham treated birds were brushed in the same locations for the same duration with a and September 2013, the birds were tested sequentially over four days: dry paint brush. Illustration kindly provided by D.R. Smith.

112 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 111–117 away from the researcher). At this point, the box was removed and an acceleration due to gravity) in the x (forward), y (sideways), and z external stimulus (clicking sound produced by a dog-training clicker, (vertical) axes at 0.005-second intervals. one to three clicks) was given to induce takeoff. The observer behind To assess the energy required for takeoff from accelerometer data, the bird used angle markers on the ground to estimate the angle of we calculated overall dynamic body acceleration (ODBA, measured in deviation from the straight line perpendicular to the camera to correct g, with g being the Earth's acceleration due to gravity), a comprehen- for perspective. An example for such a video can be found in the sup- sive metric that includes both wingbeat amplitude and frequency. plementary materials. In addition to the video recordings, we deployed ODBA was calculated using two different equations: tri-axial acceleration loggers, which were custom-designed to minimize ODBA(L1norm)=−+−+− A A A A A A mass (Shafer et al., 2015). The accelerometers were fitted into a plastic dx sx dy sy dz sz (1) bird-mount logger carrier and attached to the bird with a leg-loop 2 2 2 harness, sized according to the bird's body weight following (Naef- ODBA(L2norm)=−+−+− (Adx A sx ) (Ady A sy ) (Adz A sz ) (2) Daenzer, 2007). where dynamic acceleration in direction i,Adi, was the acceleration measured at that data point and static acceleration in direction i,Asi, 2.5. Accelerometers was the average of acceleration over 0.5 s before and after the data point. To select whether to use the L1 or the L2 norm for calculating fl We used tri-axial custom-made acceleration loggers. The logger ODBA, we chose 5 representative ights and calculated the regression ffi board weighed 440 mg, and we used a plastic holder and harness that coe cient between the two norms. The correlation between the L1 and 2 added 250 mg, for a total added mass of 690 mg, averaging roughly the L2 norm for ODBA was very high (R =0.996), therefore we used 2.3% of a bird's body mass. The acceleration logger employed an the L1 norm (Eq. (1)) for analyses. fl MSP430F2274 microcontroller from Texas Instruments (Dallas, TX, We determined ight duration as the number of seconds that the fl USA) and a BMA150 accelerometer from Bosch Sensortec (Reutlingen, bird was actively ying by visually inspecting the accelerometer traces. fi Germany). The system was powered by a small rechargeable battery Flight start and end were de ned by the beginning and end of a (Panasonic ML614). The microcontroller scheduled sensor readings and rhythmic change in the z-axis of acceleration. Even though duration per ff stored the data to local Flash memory, where it was read later. The se was not important if considering takeo (for which the most im- fi fl BMA150 sensed three orthogonal axes of acceleration, had configurable portant phase is within the rst 0.5 s of ight), assessing duration was sensitivity and sample rates, and used a 10 bit analog to digital con- important to be able to exclude from our analysis any measurement verter. For this experiment, the BMA150 was configured to output 8 bit taken after the bird landed. data, with a maximum range of +/−8g at 200 samples per second. This configuration yielded a minimum sensitivity of 62.5 milli-g (with g 2.6. Data analysis being the Earth's acceleration due to gravity). Data capture was in- itiated manually via a toggle switch before the start of each flight and 2.6.1. Video data (takeoff speed and angle) continued until the onboard 30208 byte onboard memory was filled, The video evaluation included the estimation of distance flown and which took 50.35 s. When the memory was filled, the bird was re- takeoff angle at intervals of 0.1 s (20 frames) after the feet of the bird captured and the tag was read. left the ground. Most birds were off-frame by 0.5 s, and we therefore We obtained acceleration data from 7 out of 10 birds in June 2013 have data on distance and angle at 0.1, 0.2, 0.3, 0.4 and 0.5 s after (3 accelerometers failed) and all 14 birds in September 2013. In total, takeoff. The position of the bird at the moment when the feet first left we had acceleration data for 21 birds: 7 oiled birds in June 2013, and 7 the ground and every 20 frames was determined using software Image J controls and 7 oiled birds in September 2013. The data obtained from 1.47 (National Institutes of Health, http://rsb.info.nih.gov/ij/). Using the accelerometers were accelerations in g (with g being the Earth's the software, we could determine the distance travelled and angle

Fig. 2. Example of an accelerometer trace. The y-axis of the graph is arbitrary to ease viewing and the fluctuations in the lines represent wingbeats. The black upper line represents vertical acceleration (on the z-axis), the dark grey central line forward ac- celeration (on the x-axis), and the light grey lower line sideways acceleration (on the y-axis). The flight duration was determined to be during the period where oscillations in all three axes were large and rhythmic.

113 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 111–117

Fig. 3. Distance flown by Western sandpipers as a function of time after takeoff. (A) in the June 2013 experiment (N =10:

ANCOVA intercept F1,63 =395.543, p < 0.001; flight type

(baseline or oiled) F1,63 =65.158, p < 0.001; time F1,63

=1015.894, p < 0.001; flight type*time F1,63 =30.393, p < 0.001); (B) in the November 2014 experiment (N =14: in-

tercept F1,37 =6.747, p=0.013; treatment (oiled or un-oiled)

F1,11 =12.770, p=0.004; time F1,37 =25.922, p < 0.001; treat-

ment*time F1,37 =33.934, p < 0.001). Oiled treatment (solid lines and squares): grey = baseline flight, black = oiled flight. Control treatment (dashed lines and triangles): grey = baseline flight, black = experimental flight. Error bars represent standard errors.

relative to the starting position at every time point. Distance was cor- covariate in the model for difference in distance. As in the previous rected for perspective if the bird was not flying perpendicular to the analysis, non-significant terms were removed from the model, and only camera using the formula: the output of the final models is shown. We only considered the first 0.4 s of flight (rather than the first 0.5 s) because most of the birds were Distance(corrected)= Distance(measured)/cos (angle of deviation) out of frame after that time and the sample sizes at time =0.5 s were (3) considerably smaller and highly unbalanced. where the angle of deviation (estimated by the observer behind the bird) was in radians. 2.6.2. Accelerometer data (ODBA) We tested for effects of oiling by analyzing how distance (or angle) In a first step of the analysis we integrated ODBA over each wing- changed with time in oiled and un-oiled birds. For the June 2013 data beat. In addition, as ODBA varies with time since takeoff, but is only a we created a linear mixed model with distance (or angle) as the re- useful variable when averaged over several wingbeats, we averaged sponse variable (y), flight type (baseline, non-oiled flight ‘BF’ or ex- ODBA over 0.25-second intervals, representing roughly four wingbeats. perimental, oiled flight ‘EF’), time (0.1–0.5 s), the flight type*time in- We considered the takeoff to begin once the wingbeat amplitude started teraction, and body mass as fixed factors, and individual as a random to increase from baseline by visually inspecting the accelerometer factor. Since the values at every time point are correlated with the output. See Fig. 2 as an example. Wingbeats were separated by sub- previous time point, we added an autocorrelation structure (compound sequent minima in the z-axis of acceleration. symmetry) to the model. In the model for distance, angle was also in- For the sake of investigating patterns, in a first step we analyzed cluded as a fixed factor. The models were simplified by removing non- each treatment group and season separately (i.e. the June 2013 oiled significant terms and the AICs of each model were compared in order to birds, the September 2013 controls, and the September 2013 oiled choose the best final model. One bird was excluded from the analysis birds). We ran linear mixed models with ODBA as the response variable since it was a clear outlier in both its BF and EF flights (it flew off at a (y), flight type (BF or EF) and body mass as fixed factors, and individual very steep angle). as a random factor. We ran a model for each of the first 20 wingbeats, To avoid the complicated interpretation of three-way interaction respectively the first eight 0.25 s-intervals (i.e. each interval for the first terms (treatment*flight type*time), but maintain a similar statistical 2s of flight). The models yielded no significant differences between approach for analysis of the November 2014 data, we modeled the flights, but tendencies were noted. To investigate these patterns more difference in distance (or angle) between BF and EF for every individual closely we modeled the difference in ODBA between BF and EF at each using a linear mixed model. This allowed removal of flight type from wingbeat/interval. We included treatment, wingbeat/interval, the dif- the predictors, and a more straightforward interpretation of the results. ference in body mass between BF and EF and the treatment*wingbeat/ The difference in body weight between BF and EF was added as a interval interaction as fixed factors, and individual as a random factor. covariate in the models, and the difference in angle was added as a As described above, we modeled differences instead of absolute values

Fig. 4. Flight angle of Western sandpipers at different time points after takeoff. (A) in the June 2013 experiment (N =10: ANCOVA

intercept F1,62 < 0.001, p=0.998; flight type F1,62 =27.501,

p < 0.001; time F1,62 =16.565, p < 0.001; body mass F1,62

=0.434, p=0.512; flight type*time F1,62 =0.685, p=0.411); (B)

in the November 2014 experiment (N =14: intercept F1,37

=0.581, p=0.451; treatment F1,11 =0.159, p=0.698; time F1,37

=11.001, p=0.002; treatment*time F1,37 =12.577, p=0.001). Oiled treatment (solid lines and squares): grey = baseline flight, black = oiled flight. Control treatment (dashed lines and trian- gles): grey = baseline flight, black = experimental flight. Error bars represent standard errors.

114 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 111–117 in order to avoid the difficult interpretation of the three-way interaction of interest (treatment*flight*wingbeat/interval) by removing flight as a fixed factor. We merged wingbeats/intervals with similar differences in ODBA using contrasts (Crawley, 2007) to simplify interpretation. For every simplification step we tested for changes in the model deviance and stopped simplification when the change was significant (Crawley, 2007). All statistical analyses were performed using the software R 3.0.2 (R Core Team, 2012).

3. Results

Video recordings of takeoffflights were made with two different groups of sandpipers in June 2013 (N =10) and in November 2014 (N Fig. 6. Difference in ODBA between baseline and experimental takeoffflights of Western =14). See Section 2 for details of the experimental schedule, however, sandpipers for the first 2 s of flight. Time after takeoff was divided in intervals of 0.25 s. in brief, in June 2013 we measured all birds with oiled feathers fol- Groups of intervals were merged during analysis and grouped as “initial” (0.0–0.25 s after lowing a baseline flight test, whereas in November 2014 we included takeoff), “median” (0.25–0.75 s), and “late” (0.75–2.0 s). Grey triangles: control group; ff fi time-matched controls which received no oil on their second flight. black squares: oiled group. Di erences were not signi cant (intercept F1,65 =8.116, Birds that were oiled flew a significantly shorter distance (y) per unit p=0.006; Treatment F1,18 =0.061, p=0.808; Interval F2,65 =2.652, p=0.078; treat- time than in their baseline flight (Fig. 3). Similarly, oiled birds flew at a ment*interval F2,65 =1.227, p=0.300). lower angle relative to their baseline flight (Fig. 4). The oiled group had 4. Discussion higher takeoff angles than the control group already in their baseline flight (Fig. 4B). This was the result of randomly selecting birds with generally higher takeoff angles. However, while controls repeated their Western sandpipers with crude oil on the trailing edges of their ff fl performance in their second flight, the oiled group significantly de- wings and tail took o more slowly and ew at a lower angle than ff fl creased their takeoff angle after oiling. In the November 2014 experi- control birds. Feather damage is likely to a ect ight performance by ffi ment the time-matched controls maintained the same distance and decreasing lift and thrust, increasing drag, imbalance, and di culties to ff ff angle in the sham-oiled treatment flight as in their baseline flight take o (Beaufrère, 2009). The e ects of feather damage caused by fl (Figs. 3b and 4b). Within 0.4 s after takeoff, oiled birds covered 29% external oil contamination on ight, however, had not previously been less distance than on the baseline flight (data of both experiments examined. Though we have no indications about what properties of the ff combined: average distance covered on baseline flight =73.7 cm; oiled feathers were a ected by the oiling, we were able to demonstrate a ff ff flight =52.1 cm), and they flew at about a 10° lower angle than when strong e ect on takeo performance. fl unoiled. Migrating sandpipers feed in large ocks, and their stopover site The reduction of takeoff speed in oiled birds tended to be coupled selection is determined by both food availability and predation risk with a decrease in ODBA averaged over groups of wingbeats, beginning (Ydenberg et al., 2002). Above a certain body mass (about 26 g), ff from the third wingbeat after takeoff (wingbeats were grouped as fol- western sandpipers had lower take-o ability and selected safer but less fi lows: 1–2: “early”,3–20: “late”, Fig. 5). When averaging ODBA over pro table stopover sites (Ydenberg et al., 2002). Refueling at a stopover time intervals rather than wingbeats, the opposite pattern was ob- site in sandpipers is optimized to reach a body mass high enough to served: ODBA tended to be less during the initial part of the takeoff and resume migration within a short time (Alerstam and Lindström, 1990; greater later on (intervals were grouped as follows: 0.00–0.25 s after Hedenström and Alerstam, 1997). Switching to less rewarding stopover takeoff: “initial”, 0.25–0.75 s: “median”, 0.75–2.0 s: “late”, Fig. 6). sites, therefore, can cause delays to the overall migration (Lindström, 2003). A study of western and least sandpipers (Calidris minutilla) documented a 20% reduction in takeoff speed over the range of natural wing loadings studied (Burns and Ydenberg, 2002). Our observation of a 29% reduction in distance covered in oiled birds (which is equivalent to the relative reduction in speed during the given time range of 0.4 s) is even larger than the natural variation in speed caused by variation in body mass. Thus, lightly oiled birds may limit their refueling activity to safer sites, reducing fuel deposition rate and delaying migration. Oiled western sandpipers tended to produce less mechanical power per wingbeat than controls, as indicated by their lower ODBA values over the first two wingbeats. On the other hand, ODBA averaged over time intervals was slightly greater for oiled birds during the initial part of the flight. These differences were not statistically significant due to the large variation in the data. However, this suggests that oiled sandpipers may compensate for reduced power output by increasing their effort via greater wingbeat frequency. Despite the apparent added effort, oiled birds were unable to achieve the same performance (as indicated by flight distance) as controls. Additional studies are required to elucidate the mechanisms by which crude oil changes feather prop- Fig. 5. Difference in ODBA between baseline and experimental takeoffflights of Western erties and the relationship between oiling and the power requirement/ sandpipers for the first 20 wingbeats. Groups of wingbeats were merged during analysis output during takeoff in oiled birds. “ ” – “ ” – and grouped into early (wingbeats 1 2) and late (wingbeats 3 20). Grey triangles: The significance of a 29% speed reduction at takeoff is reflected in control group; black squares: oiled group. Differences were not significant (intercept F 1356 several different potential impacts due to predation by raptors (the =6.952, p=0.009; treatment F1,18 =0.660, p=0.427; wingbeat F1356 =5.132, p=0.024; main predators of shorebirds on migration). Raptor predation success is treatment*wingbeat F1356 =3.631, p=0.058).

115 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 111–117 increased if the prey is isolated from the flock (Buchanan et al., 1988), Michael Hooper (U.S. Geological Survey), and Clare Cragan (DOI Office remains at the periphery of the flock (Jennings and Evans, 1980; Inglis of the Solicitor) provided support and feedback on the manuscript. We and Lazarus, 1981), takes off slower (Whitfield, 1985), or needs a thank Yolanda Morbey for thoughtful criticism, editing and advice. longer time to reach cover (Bednekoff, 1996). It takes 0.25–0.7 s for an entire flock of shorebirds to initiate and complete a takeoff (Hilton References et al., 1999), indicating that our study addressed the biologically re- levant time frame. This also means that after about one second, the Alerstam, T., Lindström, Å., 1990. Optimal bird migration: the relative importance of delay due to oiling of feathers would be similar to that associated with time, energy, and safety. In: Gwinner, E. (Ed.), Bird Migration. Springer Verlag, – fl fl ff Berlin, Heidelberg, Germany, pp. 331 351. being the last bird to take ight, assuming an average ock takeo time Beaufrère, H., 2009. A review of biomechanic and aerodynamic considerations of the of 0.5 s. Oiled birds might thus stand out and become preferred targets avian thoracic limb. J. Avian Med. Surg. 23, 173–185. for predators. As a result, during an oil spill, the reduced escape per- Bednekoff, P.A., 1996. Translating mass dependent flight performance into predation risk: an extension of Metcalfe & Ure. Proc. R. Soc. Lond. B 263, 887–889. formance of oiled birds could disproportionately impact higher trophic Buchanan, J.B., Schick, C.T., Brennan, L.A., Herman, S.G., 1988. Merlin predation on levels if raptors and other predators selectively capture and consume wintering dunlins: hunting success and dunlin escape tactics. Wils. Bull. 100, them. 108–118. ff Some different scenarios could arise as an effect of decreased escape Burns, J.G., Ydenberg, R.C., 2002. The e ects of wing loading and gender on the escape flights of least sandpipers (Calidris minutilla) and western sandpipers (Calidris mauri). performance of shorebirds in a flock. One possibility is that the overall Behav. Ecol. Sociobiol. 52, 128–136. decrease in the population would remain constant, in the case where Crawley, M.J., 2007. The R Book. John Wiley and Sons, Ltd, Chichester, UK. ff Cresswell, W., 1994. Flocking is an effective anti-predation strategy in redshanks,Tringa the number of raptor attacks remains the same, but their e orts would – “ totanus. Anim. Behav. 47, 433 442. be directed preferably towards oiled birds (a version of the doomed Duriez, O., Kato, A., Tromp, C., Dell'Omo, G., Vyssotski, A.L., Sarrazin, F., Ropert- surplus” hypothesis, Errington, 1945). This would not impact the Coudert, Y., 2014. How cheap is soaring flight in raptors? A preliminary investigation fl overall sandpiper population, but would possibly cause toxicological in freely- ying vultures. PLoS One 9, e84887. ff Elliott, K.H., Le Vaillant, M., Kato, A., Speakman, J.R., Ropert-Coudert, Y., 2013. e ects on the predators (Zuberogoitia et al., 2006). Alternatively, the Accelerometry predicts daily energy expenditure in a bird with high activity levels. presence of escape-impaired individuals in the flock might attract ad- Biol. Lett. 9, 20120919. ditional predators that usually prey on shorebirds only in rare occa- Errington, P.L., 1945. Some contributions of a fifteen-year local study of the northern bobwhite to a knowledge of population phenomena. Ecol. Monogr. 15, 1–34. sions, such as harriers or caracaras (Page and Whitacre, 1975; Nebel Fry, D.M., Lowenstine, L.J., 1985. Pathology of common murres and Cassin's auklets and Ydenberg, 2005). Lastly, oiled sandpipers might choose safer, less exposed to oil. Arch. Environ. Contam. Toxicol. 14, 725–737. profitable stopover sites where predation is reduced (Ydenberg et al., Halsey, L.G., Green, J.A., Wilson, R.P., Frappell, P.B., 2009. Accelerometry to estimate fi energy expenditure during activity: best practice with data loggers. Physiol. Biochem. 2002). Due to the lower pro tability of these sites, and to the time take Zool. 82, 396–404. to remove the oil, this might result in delays of migration and late ar- Hedenström, A., Alerstam, T., 1997. Optimum fuel loads in migratory birds: distin- rival at the breeding sites, with consequences on reproductive success guishing between time and energy minimization. J. Theor. Biol. 189, 227–234. fl (Sandercock et al., 1999). Hilton, G.M., Cresswell, W., Ruxton, G.D., 1999. Intra ock variation in the speed of es- cape-flight response on attack by an avian predator. Behav. Ecol. 10, 391–395. A great number of oiled birds observed in the aftermath of spills are Inglis, I.R., Lazarus, J., 1981. Vigilance and flock size in brent geese: the edge effect. Z. assigned to “trace” and “light” oiled categories, but damages to them Tierpsychol. 57, 193–200. fi Jennings, T., Evans, S.M., 1980. Influence of position in the flock and flock size in the are often overlooked or poorly quanti ed (Timoney and Ronconi, – fi vigilance of the starling Sturnus vulgaris. Anim. Behav. 28, 634 635. 2010). Our ndings show that even small amounts of oil can negatively Lind, J., Jakobsson, S., Kullberg, C., 2010. Impaired predator evasion in the life history of impact escape flight performance of birds. The negative effects of oil on birds: behavioral and physiological adaptations to reduced flight ability. Curr. – sprint locomotion in birds contribute to putting them at greater risk of Ornithol. 17, 1 30. Lindström, Å., 2003. Fuel deposition rates in migrating birds: causes, constraints and mortality, and should be considered in resource injury and environ- consequences. In: Berthold, P., Gwinner, E., Sonnenschein, E. (Eds.), Avian Migration. mental impact assessments during oil spills. Springer Verlag, Berlin, Heidelberg, Germany, pp. 307–320. Morrison, R.I.G., Ross, R.K., Guzman, P.J., Estrada, A., 1993. Aerial surveys of Nearctic Ethical standards shorebirds wintering in Mexico: Preliminary results of surveys on the Gulf of Mexico and Caribbean coasts. Canadian Wildlife Service Progress Notes, no. 206. Canadian Wildlife Service, Ottawa, Ontario. This research is in compliance with regulations governing animal Naef-Daenzer, B., 2007. An allometric function to fit leg-loop harnesses to terrestrial birds. J. Avian Biol. 38, 404–407. research. Bird capture, housing, and experiments were performed under Nebel, S., Ydenberg, R.C., 2005. Differential predator escape performance contributes to a the guidelines of the University of Western Ontario Animal Use Sub- latitudinal sex ratio cline in a migratory shorebird. Behav. Ecol. Sociobiol. 59, 44–50. Committee (protocol 2012-027) and according to permit CA-0256 from Nebel, S., Lank, D.B., O'Hara, P.D., Fernández, G., Haase, B., Delgado, F., Estela, F.A., the Canadian Wildlife Service. Evans Ogden, L.J., Harrington, B., Kus, B.E., Lyons, J.E., Mercier, F., Ortego, B., Takekawa, J.Y., Warnock, N., Warnock, S.E., 2002. Western sandpipers (Calidris mauri) during the nonbreeding season: spatial segregation on a hemispheric scale. Author contributions Auk 119, 922–928. NOAA, 2015. Deepwater Horizon oil spill: Draft Programmatic Damage Assessment and Restoration Plan and Draft Programmatic Environmental Impact Statement. I.M. and C.G.G. conceived the idea and designed the study, K.D. and Downloadable from 〈http://www.gulfspillrestoration.noaa.gov/restoration- C.A.P. participated in designing the study, I.M., L.V.K., A.M., C.G.G and planning/gulf-plan/〉. K.H.E. performed the experiments, I.M. and C.G.G. wrote the manu- Ortega-Jiménez, V.M., Álvarez-Borrego, S., Arriaga-Ramírez, S., Renner, M., Bridge, E.S., 2010. Takeoffflight performance and plumage wettability in Cassin's Auklet script, I.M. and C.G.G. developed the methods, I.M. and K.H.E. analyzed Ptychoramphus aleuticus, Xantus's Murrelet Synthliboramphus hypoleucus and Leach's the data, R.M. provided the accelerometers, advised on data analysis, Storm-petrel Oceanodroma leucorhoa. J. Ornithol. 151, 169–177. and wrote sections of the manuscript. All authors have read and ap- Page, G., Whitacre, D.F., 1975. Raptor predation on wintering shorebirds. Condor 77, – fi 73 83. proved the nal version of the manuscript. Peakall, D.B., Hallett, D.J., Bend, J.R., Foureman, G.L., Miller, D.S., 1982. Toxicity of Prudhoe Bay crude oil and its aromatic fractions to nestling herring gulls. Environ. Acknowledgments Res. 27, 206–215. Pennycuick, C.J., 2008. Modelling the Flying Bird. Elsevier, Academic Press, London, UK. Core Team, R., 2012. R: A Language and Environment For Statistical Computing. R The studies appearing in this special issue were funded by the US. Foundation for Statistical Computing, Vienna, Austria (ISBN 3-900051-07-0, URL). Fish and Wildlife Service as part of the Deepwater Horizon Natural 〈http://www.R-project.org/〉. Resource Damage Assessment. We thank A.R. Gerson, D. Lank, B. Sandercock, B.K., Lank, D.B., Cooke, F., 1999. Seasonal declines in the fecundity of arctic- breeding sandpipers: different tactics in two species with an invariant clutch size. J. Schwarz, G. Anderson, I. Jong, M. Dick, and M. Rebuli for help during Avian Biol. 30, 460–468. fieldwork and bird care. D. Romney Smith provided the illustration. Shafer, M.W., MacCurdy, R., Shipley, J.R., Winkler, D., Guglielmo, C.G., Garcia, E., 2015. fl Katherine Healy and Peter Tuttle (U.S. Fish and Wildlife Service), The case for energy harvesting on wildlife in ight. Smart Mat. Struct. 24, 025031.

116 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 111–117

Swaddle, J.P., Witter, M.S., 1997. The effects of molt on the flight performance, body P.J., 2006. Moving towards acceleration for estimates of activity‐specific metabolic mass, and behavior of European starlings (Sturnus vulgaris): an experimental ap- rate in free‐living animals: the case of the cormorant. J. Anim. Ecol. 75, 1081–1090. proach. Can. J. Zool. 75, 1135–1146. Ydenberg, R.C., Butler, R.W., Lank, D.B., Guglielmo, C.G., Lemon, M., Wolf, N., 2002. Swaddle, J.P., Williams, E.V., Rayner, J.M., 1999. The effect of simulated flight feather Trade-offs, condition dependence and stopover site selection by migrating sandpi- moult on escape take-off performance in starlings. J. Avian Biol. 30, 351–358. pers. J. Avian Biol. 33, 47–55. Thomas, A.L.R., 1997. On the tails of birds. BioScience 47, 215–225. Ydenberg, R.C., Butler, R.W., Lank, D.B., Smith, B.D., Ireland, J., 2004. Western sand- Timoney, K.P., Ronconi, R.A., 2010. Annual bird mortality in the bitumen tailings ponds pipers have altered migration tactics as peregrine falcon populations have recovered. in Northeastern Alberta, Canada. Wils. J. Ornithol. 122, 569–576. Proc. R. Soc. Lond. B 271, 1263–1269. Whitfield, D.P., 1985. Raptor predation on wintering waders in southeast Scotland. Ibis Zuberogoitia, I., Martínez, J.A., Iraeta, A., Azkona, A., Zabala, J., Jiménez, B., Merino, R., 127, 544–548. Gómez, G., 2006. Short-term effects of the Prestige oil spill on the peregrine falcon Wilson, R.P., White, C.R., Quintana, F., Halsey, L.G., Liebsch, N., Martin, G.R., Butler, (Falco peregrinus). Mar. Poll. Bull. 52, 1176–1181.

117 Ecotoxicology and Environmental Safety 146 (2017) 118–128

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Toxicological and thermoregulatory effects of feather contamination with MARK artificially weathered MC 252 oil in western sandpipers (Calidris mauri) ⁎ Ivan Magginia,b, , Lisa V. Kennedya, Steven J. Bursianc, Karen M. Deand, Alexander R. Gersona,1, Kendal E. Harre, Jane E. Linkc, Chris A. Pritsosf, Karen L. Pritsosf, Christopher G. Guglielmoa a Advanced Facility for Avian Research, University of Western Ontario, London, ON, Canada N6G 1G9 b Konrad-Lorenz Institute of Ethology, University of Veterinary Medicine, Savoyenstrasse 1a, 1160 Vienna, Austria c Department of Animal Science, Michigan State University, 474 South Shaw Lane, East Lansing, MI 48824, United States d Abt Associates, 1881 Ninth St., Ste 201, Boulder, CO 80302-5148, United States e Urika, LLC, Mukilteo, WA 98275, United States f University of Nevada-Reno, Max Fleischmann Agriculture Bldg. 210, Reno, NV 89557, United States

ARTICLE INFO ABSTRACT

Keywords: The external contamination of bird feathers with crude oil might have effects on feather structure and thus on Respirometry thermoregulation. We tested the thermoregulatory ability of western sandpipers (Calidris mauri)ina Thermal conductance respirometry chamber with oil applied either immediately prior, or three days before the experiment. The Oxidative stress birds were then exposed to a sliding cold temperature challenge between 27 °C and −3 °C to calculate thermal Renal damage conductance. After the experiment, a large blood sample was taken and the liver extracted to measure a range of Anemia parameters linked to toxicology and oxidative stress. No differences in thermal conductance were observed among groups, but birds exposed to oil for three days had reduced body temperatures and lost more body mass during that period. At necropsy, oiled birds showed a decrease in plasma albumin and sodium, and an increase in urea. This is reflective of dysfunction in the kidney at the loop of Henle. Birds, especially when exposed to the oil for three days, showed signs of oxidative stress and oxidative damage. These results show that the ingestion of externally applied oil through preening or drinking can cause toxic effects even in low doses, while we did not detect a direct effect of the external oil on thermoregulation over the temperature range tested.

1. Introduction toxic effects of oil ingestion. However, currently available information from the literature and the field is not sufficient to characterize the Oil spills affect ecosystems and wildlife in numerous ways, but nature and extent of the injuries to oiled birds, or to quantify those perhaps the most obvious and dramatic impacts are on aquatic- injuries in terms of effects on bird viability. associated birds. Heavy oiling of feathers makes it impossible to fly, Studies conducted following past oil spills have focused on the reduces buoyancy, and greatly increases heat losses (Lambert et al., physiological effects of oil ingestion; however, there are far fewer 1982; Jenssen and Ekker, 1991; O’Hara and Morandin, 2010). Internal studies that focus on the physiological impacts of oil on feathers. exposure to oil by ingestion during feather preening or absorption Feathers are a unique feature of birds and through evolution they have through the skin leads to toxic pathologies (Hartung and Hunt, 1966; adapted to optimize both insulation and mechanical function for flight Eastin and Rattner, 1982; Pattee and Franson, 1982; Leighton et al., (Parkes, 1966; Regal, 1975; Prum, 1999). In the event of contamination 1985; Lee et al., 1986; Leighton, 1986; Hughes et al., 1990; Yamato with oil, these functions are disrupted (O’Hara and Morandin, 2010). et al., 1996; Walton et al., 1997; Newman et al., 2000; Seiser et al., The overall impact and the behavioral and ecological consequences of 2000; Troisi et al., 2007). Most heavily oiled birds die quickly. oiling include increased preening, reductions in food consumption, However, large numbers of birds might experience lower, non-lethal weight loss and even separation from unoiled members of the flock oil coverage. Such birds may experience delayed mortality or reduced (Larsen and Richardson, 1990; Sharp et al., 1996; Andres, 1997; Burger, fitness (Henkel et al., 2012; Montevecchi et al., 2012) because of the 1997; Burger and Tsipoura, 1998). These studies reported low direct

⁎ Corresponding author at: Konrad-Lorenz Institute of Ethology, University of Veterinary Medicine, Savoyenstrasse 1a, 1160 Vienna, Austria. E-mail addresses: [email protected] (I. Maggini), [email protected] (L.V. Kennedy), [email protected] (S.J. Bursian), [email protected] (K.M. Dean), [email protected] (A.R. Gerson), [email protected] (K.E. Harr), [email protected] (J.E. Link), [email protected] (C.A. Pritsos), [email protected] (C.G. Guglielmo). 1 Current address: Department of Biology, University of Massachussetts, Amherst, MA 01003, United States. http://dx.doi.org/10.1016/j.ecoenv.2017.04.025 Received 11 August 2016; Received in revised form 7 April 2017; Accepted 11 April 2017 Available online 28 April 2017 0147-6513/ © 2017 Elsevier Inc. All rights reserved. I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 118–128 mortality, but the low quantity of oil ingested through preening is likely electrolyte imbalance, impaired organ function, oxidative damage, to have long-lasting sub-lethal effects, as demonstrated indirectly by the immune system activation) could have important negative effects on immediate loss of weight in oiled sanderlings (Burger and Tsipoura, thermoregulation if oxygen limitation and other factors alter resting 1998). To our knowledge, there are no published studies on the metabolic rate and/or reduce thermogenic capacity during shivering. physiological effects of ingested and externally applied oil on shorebird Previous studies have reported increased metabolic rates in birds that thermoregulation. ingested oil (Butler et al., 1986; Jenssen, 1994). Under normal conditions, birds maintain their body temperature In this study, the patterns of heat production in western sandpipers and vital functions by producing heat metabolically. The rate of (Calidris mauri) externally contaminated with oil in two scenarios were metabolic heat production is constant within a range of ambient investigated: birds that had the opportunity to preen and bathe (thereby temperatures, the thermoneutral zone (TNZ), because of regulatory possibly drinking contaminated water), thus eliminating some oil from changes in posture and feather insulation (Scholander et al., 1950). This their feathers but ingesting it in low quantities; and birds that did not constant metabolic heat production is defined as the basal metabolic have the opportunity to preen, thus having potentially reduced feather rate (BMR) when the bird is in a resting, postabsorptive state within the insulation but no internal effects of oil. We hypothesized that external TNZ (McNab, 1997). Below a lower critical ambient temperature, oil would have an effect on feather insulation resulting in an increase in feather insulation and other passive changes are no longer enough to thermal conductance. This should not affect BMR unless the lower maintain body temperature, and additional heat has to be produced, critical temperature is increased so much that measurements would fall mainly through shivering (Hohtola, 1981, 2004; Barré et al., 1985). outside the TNZ. Ingestion of oil, on the other hand, may induce Thus, at low ambient temperatures metabolic heat production must hypothermia or hyperthermia, thus affecting metabolic rates (including increase, thereby increasing metabolic rate as indicated by oxygen BMR) by lowering or raising the whole curve, but not affecting the slope consumption and carbon dioxide production. Shivering metabolic rates of thermal conductance. Since these birds might still have oil on their have been shown to increase linearly as ambient temperature decreases feathers, however, a combination of both scenarios is possible in birds (Scholander et al., 1950). Body size is the main factor affecting BMR in that were allowed to preen. We measured this using respirometry, a birds (Lasiewski and Dawson, 1967; Bennett and Harvey, 1987; well-established technique that is commonly used to measure the

McKechnie et al., 2006), but differences in feather insulation are metabolic rates of animals by accurately measuring O2 consumption ̇ ̇ ̇ responsible for differences in the slope of shivering thermogenesis, (VO2) and CO2 production (VCO2), and evaporative water loss (VH2O) e.g. among species from different environments (Scholander et al., (Lighton, 2008). These parameters allow the determination of meta- 1950; Hudson and Kimzey, 1966). The steeper the slope, the higher the bolic rate, rates of water loss, the relative contribution to energy from thermal conductance of a bird will be. High thermal conductance fat, and carbohydrate oxidation, and they allow the calculation of indicates reduced insulation, increased heat loss, and thus higher thermal conductance. In addition to the values directly related to energetic requirements to maintain a constant body temperature and thermoregulation, several endpoints that indicate toxic effects of low physiological homeostasis. amounts of oil ingested when preening or drinking contaminated water External exposure to oil decreases a bird's thermoregulatory ability, were measured from blood and liver samples taken at necropsy. which can be critical for survival, especially in cold climates (Perry et al., 1978). Oiled feather barbules become matted, resulting in a 2. Materials and methods decreased insulative function of the feathers (Lambert et al., 1982; O’Hara and Morandin, 2010). For waterbirds living in cold climates, the 2.1. Study species associated effects of reduced buoyancy (which increases the surface area exposed to cold water) and reduced feather insulation can result in Western sandpipers are long-distance migrating shorebirds. Western rapid hypothermia and death. There is not a large body of literature, sandpipers were captured near Roberts Bank and Boundary Bay in however, investigating the effects of external oil coverage on water- Delta, BC, Canada in July 2012 (80 individuals) under the guidelines of birds in more temperate regions, or on shorebirds in any climate. the University of Western Ontario Animal Use Sub-Committee (protocol External oil exposure is reported to cause increased heat production 2012-027) and according to permit CA-0256 from the Canadian (McEwan and Koelink, 1973; Erasmus et al., 1981) or hypothermia Wildlife Service. They were held for up to one week at Simon Fraser (Jenssen and Ekker, 1991), as well as behavioral modifications such as University (Burnaby, BC, Canada) before same-day shipment by air increased preening, increased aggressiveness, altered feeding behavior, cargo to Toronto, ON, Canada. They were transported by vehicle to the and reduced resting time (Jenssen, 1994; Walton et al., 1997; Burger Advanced Facility for Avian Research (AFAR) at the University of and Tsipoura, 1998). One potential effect of a reduction of feather Western Ontario, London, ON Canada and maintained in captivity until insulation due to exposure to crude oil is an increase in thermal experiments. These birds were initially maintained in one of the conductance, and the shift of the lower critical temperature (and thus specialized 2.4 m×3.7 m shorebird rooms at AFAR under 16 L:8D the TNZ) to higher values. (16 h of light, 8 h of darkness) light conditions at approximately External oiling also has additive toxicological effects via ingestion 22 °C. In 2012, during the winter, the light cycle was switched to through preening. When ingested by birds at concentrations that are 12 L:12D to simulate conditions on their winter range. In mid-April not acutely lethal, oil can cause a wide range of adverse effects, 2013, the light cycle was changed to 14 L:10D. The experiments were including hemolytic anemia (Leighton et al., 1985; Leighton, 1986; run between October 6 and October 23, 2013, corresponding to the Yamato et al., 1996; Newman et al., 2000; Seiser et al., 2000; Troisi beginning of wintering in free-ranging birds. Birds were maintained on et al., 2007). This damage results from oxidative damage to the an ad libitum diet of 80% Mazuri Waterfowl Starter and 20% trout structure of hemoglobin causing precipitates to form that can coalesce chow (Aquamax Fingerling Starter 300). The diet was supplemented into Heinz bodies in red blood cells (Leighton et al., 1983). Addition- with ~50 mealworms/20 birds every other day. ally, when cormorants were orally and dermally exposed to weathered MC252 crude oil external blood loss through the gastrointestinal tract 2.2. External dosing due to coagulopathy was documented (Bursian et al., 2017a, in press) Animals can only compensate for this damage by producing more red The toxicant was Mississippi Canyon 252 (MC252) oil collected blood cells to account for the loss of hemoglobin. If this compensation is during the 2010 Deepwater Horizon Gulf of Mexico oil spill and incomplete, the animal suffers from a loss of the oxygen-carrying artificially weathered (Forth et al., 2017). MC252 oil was applied capacity of blood. This loss of oxygen-carrying capacity, along with a externally with a brush to cover 10 cm2 of the back (2 cm2) and belly variety of other effects of oil on the physiological state of the bird (e.g. (8 cm2) of the experimental birds. The body cover obtained was

119 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 118–128 approximately 20% of the total body surface of the non-flying bird. This Ambient air was pumped into the system using a vacuum pump oiling pattern was chosen to simulate light levels of oiling in a natural (model DOA-P704-AA, Gast Manufacturing), was dried through two situation. The average weight of oil applied was 0.30 g (range: 0.21 – Drierite™ cylinders and a PC-4 Peltier Effect Dryer (Sable Systems Int.) 0.39 g). There was no difference between the weight of oil applied to before entering the chamber with the bird. The four chambers were birds in the 0.5 d and the 3.5 d oiling groups (Student's t-test: located in a Sanyo Incubator (MIR-154) that allowed for controlled t=−0.824, df=25.7, p=0.42, see below for a description of oiling temperature. A baseline measurement of 20 min was taken every two groups). Sham treatment followed the same procedure but a dry brush hours to control for drift in the analyzers. Baseline air bypassed the was used. chambers and entered directly the Flow Multiplexer (Sable Systems Int.). Baseline air and outcoming air from the chambers were sub- 2.3. Treatment groups and schedule sampled using a subsampling pump and passed through a water vapour

analyzer (RH-300, Sable Systems Int.), a CO2 analyzer (CA-2A, Sable A total of 44 birds was divided into three experimental treatments: Systems Int.) and an O2 analyzer (FC-1B, Sable Systems Int.) serially, before being expelled into the ambient. The system was calibrated 1. Controls (N=14): Received a sham application on Day 1 and on Day every day two hours before the beginning of the experiment. Pure 3 immediately prior to the respirometry measurement; nitrogen gas was used to zero all analyzers, a certified standard gas mix

2. 3.5 d external oiling (N=15): Received an application of oil on Day (20.9% O2, 1.00% CO2, balanced with nitrogen) was used to span the 1 and a sham application on Day 3 immediately prior to the CO2 and the O2 analyzers, and water-saturated air at a dewpoint of respirometry measurement; 17 °C was used to span the water vapour analyzer. 3. 0.5 d external oiling (N=15): Received a sham application on Day 1 During measurement of BMR, birds were in a post-absorptive state and an external oil application on Day 3 immediately prior to the (food was removed about six hours prior to entering the experiment). respirometry measurement. Measurements within the TNZ were obtained at 27 °C (based on comparisons with similar-sized shorebirds, e.g. Piersma, 1995, All birds were moved into plastic holding corrals (1.5×1.5 m, Ruthrauff et al., 2013) during the first six hours of the respirometry maximum five birds/corral) three days before the respirometry mea- night. surements. At this time, the birds were weighed ( ± 0.01 g), scanned in To assess the metabolic response to low temperature, birds were a quantitative magnetic resonance (QMR, EchoMRI) analyzer for body exposed to a sliding cold challenge similar to those used by Vaillancourt composition measurement (Guglielmo et al., 2011), and birds from the et al. (2005) and Vézina et al. (2011) immediately after the six hours of 3.5 d oiling group were oiled while control birds and birds from the 0.5 BMR measurements at 27 °C. During the cold challenge, temperature d oiling group were sham treated. Even though all birds lost weight in was automatically decreased in a stepwise manner, by 10 °C per step, to the corrals (see results), the weights and fat loads at the time of a minimum temperature of −3 °C. At each step, once a stable ambient respirometry were comparable to values of wild birds (Guglielmo and temperature was reached, shivering metabolic rate was recorded. Each Williams, 2003). Birds often gain extra weight and fat under captive, ad temperature step lasted two hours. libitum feeding conditions. Immediately before the respirometry mea- In total, the birds were in the respirometry chambers for 12 h, surement, birds were weighed and scanned again in the QMR. At this starting at 8:00 PM and ending at 8:00 AM the following morning. point we used a thermocouple inserted under the wing to measure body temperature (Tb) of each bird. This method provides an accurate and 2.5. Metabolic rate calculations non-invasive measurement of superficial body temperature, which correlates well with internal (core) Tb (G.A Mitchell and C.G. Gugliel- Respirometry data were recorded using the ExpeData software mo, unpublished data from hermit thrushes Catharus guttatus,r2 (version 1.1.21, Sable Systems Int.) every second during the respiro-

=0.416, p < 0.001). After 12 h in the respirometry chambers Tb was metry run. Fractional concentrations of O2 and CO2 (the raw output measured again, and birds were weighed and scanned in the QMR for a from the respirometry) were corrected for the presence of water (CO2 ̇ ̇ third time. A correction factor for the values of fat (0.943) and lean and O2) and for the presence of CO2 (O2 only), and VO2 and VCO2 were (1.021) was applied in line with the previously published validation of calculated using the following equations (Lighton, 2008): the QMR (Guglielmo et al., 2011). The respirometry setup allowed F H O=(HO/0.803)/1000 (1) testing of four birds each night. Every night one cohort of birds e 2 2 consisting of at least one bird from each experimental group was tested. The fourth bird was chosen alternatively from either the 0.5 d or the 3.5 FeO2 =(O2/100)/(1-FeH2O) (2) d external oiling group.

FeCO2 = (CO2/100)/(1-FeH2O) (3) 2.4. Respirometry setup ̇ VO2 = Flow*((0.2095-FeO2) - FeO2*(FeCO2-0.0004))/(1-FeO2) (4) Flow-through respirometry was used to quantify both nighttime resting metabolic rate (or basal metabolic rate, BMR) and shivering V̇CO = Flow*((F CO -0.0004) - F CO *(0.2095-F O ))/(1-F CO ) (5) metabolic rate using a sliding cold challenge. BMR and shivering 2 e 2 e 2 e 2 e 2 metabolic rate were measured in four birds (one cohort) each night where H2O (Eq. (4)), O2 (Eq. (5)) and CO2 (Eq. (6)) were the raw data using a multiplexing respirometry setup. Multiplexing uses computer from the respirometry equipment and Flow was the measured flow rate − controlled mass flow switches, which allow the metabolic measurement (ml min 1). of multiple animals sequentially throughout a respirometry session. Birds were measured 30 times in total over the 12 h, at intervals of Using this system, only one channel can be measured at a time, and five minutes each. For each bird, 15 measurements were obtained at switching between channels occurs automatically. The frequency of 27 °C (basal metabolic rate), and five measurements each at 17 °C, 7 °C, channel switching depends on the response time of the system in use. and −3 °C. Since the lowest possible values as reflective of the values The response time depends on the sub-sampling flow rate, the length of for a resting bird were of interest, the values with the lowest 5-min ̇ tubing in the sub-sampling loop, and the inherent response time of the average for VO2 at each temperature were chosen for further calcula- analyzers in use (Lighton, 2008). To allow rapid multiplexing, these tions. parameters were determined and validated. In this study, channels were Metabolic rates in watts (W) at each temperature were obtained switched every five minutes. using the equation:

120 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 118–128

̇ W = VO2 *(16 + 5.164*RQ)/60 (6) GPOX assay kit #703102 (Cayman Chemical Co.). For lipid peroxida- tion (LPO) assessment, 1g of liver tissue was homogenized in 5–10 ml of where RQ (the respiratory quotient) was calculated using the equation: ice-cold 20 mM Tris buffer, pH 7.4, containing 5 mM BHT (to prevent ̇ ̇ RQ = VVCO2 / O2 (7) sample oxidation). The homogenate was centrifuged at 3000×g for 10 min at 4 °C and the resulting supernatant was diluted appropriately In some cases, RQ was lower than the normal range between 0.7 and and used in assay kit #FR22 (Oxford Biomedical Research) according to 1.0. In these cases, we applied a default value of 0.7 to RQ. Using this manufacturer's directions. To determine total antioxidant power ff value instead of the calculated value did not a ect the overall results of (Trolox), liver tissue was homogenized in 5–10 ml of ice-cold phosphate the study. buffered saline (PBS)/g tissue. The homogenate was centrifuged at − We calculated thermal conductance (W/°C) at 3 °C using the 3000×g for 12 min at 4 °C and the supernatant was collected for the formula: assay. Trolox was determined using assay kit #TA02 (Oxford Biomedical Research) according to manufacturer's directions (Pritsos C = MR/(Tb -Ta) (8) et al., 2017, in press). where MR was the metabolic rate at −3 °C, Tb was the body temperature as measured after the experiment, and Ta was constant 2.8. Statistics at −3 °C. 2.8.1. Energetics 2.6. Necropsy Differences among treatments in the initial body, fat, and lean mass were tested using one-way ANOVA. Differences among treatments in After the respirometry run and QMR analysis, a large blood sample body, fat, and lean mass change during the three-day period in the was taken from all experimental birds. Blood was collected from the corral and during the night of the respirometry run were then tested. ulnar vein extravascularly after the vein was lanced with the needle Each difference was compared as both absolute and relative values bevel using lithium-heparinized capillary tubes. Subsequently, addi- using ANCOVAs to test for treatment and initial body mass effects. An tional blood was extracted from the jugular vein using a heparin coated interaction term was included in the model but was removed when not insulin syringe from which all excess heparin was expelled before blood significant. Where treatment effect was significant, pairwise compar- draw. All blood was pooled and aliquoted appropriately for different isons using Bonferroni corrections were also performed. In the parti- analyses. Immediately after blood sampling, birds were euthanized by cular case of the change in weight during the three days in the corral, a cervical dislocation and necropsied. The liver was removed and divided p-value of 0.10 was accepted as significant, since there was the one- into six approximately equal sections and flash-frozen in liquid nitro- tailed prediction that the birds from the 3.5 d oiling group would lose gen. more weight than the other two groups that did not have oil. In all other Whole blood was analyzed immediately using VetScan VS2 and iStat tests, a P-value lower than 0.05 was accepted as significant. Similarly, fi analyzers (Abaxis Global Diagnostics) using Avian/Reptilian Pro le treatment differences in Tb before the respirometry run were tested fi Plus and Mammalian Liver Pro le rotors, and the iStat-E3+ cartridge, using a one-way ANOVA, and differences in the Tb change between the according to manufacturer's instructions. If an analyte was measured on beginning and the end of the run were tested using a multiple fi fi both the Avian/Reptilian Pro le Plus and Mammalian Liver Pro le regression model with treatment, initial Tb and initial body mass as rotors, only values from the Avian/Reptilian Profile Plus rotor were covariates. The model was simplified removing terms and the model used for data analysis. Blood smears were standardly prepared using the with the lowest Akaike Information Criterion (AIC) value was selected standard push method (Hoppe and Lassen, 1978) for complete blood as the best model. count (CBC) estimates. Manual differential counts of either 100 or 200 cells, dependent upon cellularity of the sample, was performed 2.8.2. Respirometry (Campbell and Ellis, 2013). Plasma for measurement of plasma protein We compared metabolic rates at 27 °C (BMR) using an ANCOVA electrophoresis was prepared by centrifuging whole, heparinized blood that included metabolic rate in W as a response variable, and treatment for two minutes at > 9000 rpm. Plasma protein electrophoresis was and initial body mass as explanatory variables. The interaction term performed using split beta gels on the SPIFE 3000 (Helena Labora- was included in the model initially but was removed as it was not tories). significant. We tested for pairwise differences between treatments using Bonferroni corrections. We compared resting metabolic rates (RMR)at 2.7. Assessment of oxidative damage ambient temperatures between 17 and −3 °C using a linear mixed effects model with metabolic rate in W as a response variable, body Oxidative damage in the liver was assessed on liver homogenates mass, ambient temperature, treatment, and the treatment*temperature prepared from the individual liver subsamples described above. For interaction term (to identify differences in the slope of the relationship) total, oxidized and reduced glutathione (TGSH, GSSG, and RGSH, as fixed factors, and bird ID as a random factor to test for differences respectively) liver tissue was homogenized in ice-cold 50 mM 2-(N between treatments in RMR. We compared thermal conductance of the morpholino) ethanesulphonic acid (MES), 1 mM EDTA buffer (pH 6)/g three experimental groups measured at −3 °C using a one-way ANOVA. of tissue. The homogenate was centrifuged at 10,000×g for 15 min at 4 °C and the supernatant was collected and kept on ice. The supernatant 2.8.3. Necropsy endpoints was deproteinated by addition of an equal volume of 0.1% metapho- Given the low sample sizes and the usually non-normal distribution sphoric acid and then vortexed. After allowing the mixture to stand at of the variable values, nonparametric Kruskal-Wallis tests were used to room temperature for 5 min, it was centrifuged at 5000×g for 5 min at assess differences among groups in blood and liver endpoints. If room temperature. The supernatant was collected and used in the assay significant, Nemenyi posthoc tests with Tukey corrections were used for TGSH and GSSG (kit #70300, Cayman Chemical). Reduced glu- for pairwise comparisons. tathione was calculated as the difference between TGSH and GSSG. To Proposed reference intervals were augmented and verified in determine glutathione peroxidase (GPOX) activity, liver tissue was accordance with the American Society for Veterinary Clinical homogenized in ice-cold 50 mM Tris-HCl, 5 mM EDTA, 1 mM DTT Pathology (ASVCP) guidelines using MedCalc (Version 14.12.0 64 bit; buffer (pH 7.5)/g tissue. The homogenate was centrifuged at 10,000×g MedCalc Software, Ostend, Belgium) and a more stringent setting of the for 15 min at 4 °C and the supernatant was collected and kept on ice. Dixon Test using confidence levels of 0.1 or Tukey's Outlier Test (Geffré The assay was performed on appropriately diluted supernatant with et al., 2011; Friedrichs et al., 2012). These were calculated from control

121 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 118–128

Table 1 Mean body weights ± SD of western sandpipers at three time points during the thermoregulation experiment. In parentheses are the mean ± SD relative change (in relation to the previous measurement) given as a percentage. Superscript letters indicate significant differences between treatments.

Controls [N=14] 0.5 d oiling [N=15] 3.5 d oiling [N=15]

Body weight [g] three days before respirometry 33.21 ± 3.83a 33.55 ± 2.70a 33.58 ± 3.87a

Body weight [g] immediately before respirometry 29.16 ± 3.86 29.30 ± 2.63 28.64 ± 3.74 (−12.3 ± 3.6%)a (−12.7 ± 3.3%)ab (−14.8 ± 2.4%)b

Body weight [g] immediately after respirometry 27.44 ± 3.80 27.58 ± 2.54 26.76 ± 3.74 (−6.0 ± 1.2%)a (−5.9 ± 0.5%)a (−6.6 ± 1.1%)a birds used in other western sandpiper studies (Bursian et al., 2017b; lost 1.20 ± 0.55 g of lean mass. Initial lean mass also had no significant

Maggini et al., 2017, this issue; Maggini et al., in press: hemoglobin: effect on rate of lean mass loss (ANCOVA: F1,40=2.33, P=0.135). N=26; albumin: N=13; sodium: N=27; chloride: N=24; urea: N=22). During the respirometry experimental night all birds lost additional All other statistical analyses were performed using the software R weight (Table 1). On average, controls lost 1.72 ± 0.24 g, birds from 3.0.2 (R Core Team, 2012). the 0.5 d oiling group lost 1.72 ± 0.17 g, and birds from the 3.5 d oiling group lost 1.88 ± 0.25 g. The difference in weight loss among groups

3. Results was not significant (ANCOVA: F2,40 =2.47, P=0.098). Initial body mass did not affect body mass change significantly during the respiro-

3.1. Energetics metry experiment (ANCOVA: F1,40=2.20, P=0.146). The loss of fat mass during the respirometry experimental night

Birds from the three experimental groups entered the experiment differed significantly among treatments (ANCOVA: F2,40=8.06, with similar body weights at the time they were put into the corrals P=0.001; Table 2). On average, controls lost 0.78 ± 0.10 g, birds from

(one-way ANOVA: F2,41=0.05, P=0.95), three days before the respiro- the 0.5 d oiling group lost 0.82 ± 0.11 g, and birds from the 3.5 d oiling metry run (Table 1). Birds from all three experimental groups lost group lost 0.70 ± 0.09 g of fat mass. Birds from the 3.5 d oiling group weight during the three days in the corrals. The difference in weight lost less fat mass than the two other treatments. This difference was loss between groups was significant at the 0.10 P-level (ANCOVA: significant between the 0.5 d and the 3.5 d oiling groups (P=0.001),

F2,40=3.18, P=0.053). Post-hoc tests with Bonferroni corrections but not between the 3.5 d oiling group and the controls (P=0.067). The showed that the difference was significant between birds from the 3.5 0.5 d oiling group birds did not differ significantly from controls d oiling group and controls (P=0.074), but not between the 0.5 d and (P > 0.5). Initial fat mass had a significant effect on fat mass loss during the 3.5 d group (P=0.212) or between controls and the 0.5 d oiling respirometry. Birds with high initial fat mass lost less fat than birds with group (P > 0.5). Initial body weight had no significant effect on the low initial fat mass (ANCOVA: F1,40=15.79, P < 0.001). change in body mass over the three days in the corral (ANCOVA: The loss of lean mass during the respirometry experimental night ff fi F1,40=1.42, P=0.240). On average, controls lost 4.04 ± 1.10 g, birds di ered signi cantly among treatments (ANCOVA: F2,40=6.85, from the 0.5 d oiling group lost 4.26 ± 1.17 g, and birds from the 3.5 d P=0.003). On average, controls lost 1.05 ± 0.32 g, birds from the 0.5 oiling group lost 4.95 ± 0.75 g. d oiling group lost 1.05 ± 0.18 g, and birds from the 3.5 d oiling group Initial body fat and lean mass did not differ among treatment groups lost 1.37 ± 0.30 g of lean mass. Birds from the 3.5 d oiling group lost fi when entering the corrals (one-way ANOVA, for fat: F2,41=0.37, signi cantly more lean mass than birds from the 0.5 d oiling group P=0.694; for lean: F2,41 =0.35, P=0.709; Table 2). There was a (P=0.005) and controls (P=0.006). Controls and birds from the 0.5 d significant effect of treatment on the loss of fat after three days in the oiling group did not differ from each other (P > 0.5). Initial lean mass ff corrals (ANCOVA: F2,40=5.38, P=0.009). On average, controls lost at the start of the respirometry night had no e ect on the change in lean 2.58 ± 0.98 g, birds from the 0.5 d oiling group lost 3.07 ± 0.74 g, and body mass over the night (ANCOVA: F1,40=2.01, P=0.155). These birds from the 3.5 d oiling group lost 3.46 ± 0.31 g of fat mass. The 3.5 results indicate that birds from the 3.5 d oil exposure group lost less fat d oiling birds lost significantly more fat mass than controls (P=0.011), (although not significantly), but more lean mass than the other two but did not differ significantly from the 0.5 d oiling birds (P=0.459). treatments during the respirometry experimental night, while birds The 0.5 d oiling birds did not differ significantly from controls from the 0.5 d oil exposure group and controls did not differ from each (P=0.304). Initial fat mass had no significant effect on the difference other. fi in fat mass (ANCOVA: F1,40=1.588, P=0.215). At the time of beginning the respirometry run, there were signi - There was no significant difference among groups in the change in cant differences in Tb among groups (F2,41=4.82, P=0.013, Table 3). fi lean mass (ANCOVA: F2,40=1.51, P=0.233) over the three days in the Birds from the 3.5 d oiling group had signi cantly lower Tb than corrals. On average, controls lost 1.20 ± 0.49 g, birds from the 0.5 d controls (P=0.018), and tended to have lower Tb than birds from the oiling group lost 0.91 ± 0.53 g, and birds from the 3.5 d oiling group 0.5 d oiling group (P=0.066). The Tb difference between the beginning

Table 2 Fat and lean mass of western sandpipers at different time points during the thermoregulation experiment. Mean absolute and relative differences from the previous value (in parentheses) are given ± SD. Superscript letters (for fat) and symbols (for lean) indicate significant differences between treatments.

Controls [N=14] 0.5 d oiling[N=15] 3.5 d oiling [N=15]

Fat [g] Lean [g] Fat [g] Lean [g] Fat [g] Lean [g]

Three days before respirometry 12.06 ± 2.48 17.26 ± 1.63 12.69 ± 1.85 16.92 ± 0.99 12.86 ± 3.33 16.93 ± 1.06

Immediately before respirometry 9.47 ± 2.59 16.07 ± 1.53 9.62 ± 1.77 16.01 ± 1.09 9.40 ± 3.20 15.73 ± 1.05 ◊ ◊ ◊ (−22.2 ± 9.6%)a (−6.9 ± 2.6%) (−24.5 ± 6.4%)ab (−5.4 ± 3.1%) (−28.0 ± 4.6%)b (−7.1 ± 3.2%)

Immediately after respirometry 8.70 ± 2.53 15.02 ± 1.45 8.80 ± 1.72 14.96 ± 1.02 8.70 ± 3.15 14.36 ± 1.12 □ □ (−8.6 ± 2.0%)ab (−6.5 ± 2.1%) (−8.8 ± 1.8%)b (−6.5 ± 1.0%) (−7.9 ± 1.7%)a (−8.7 ± 2.1%)°

122 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 118–128

Table 3 3.3. Toxicology and oxidative stress Body temperature of western sandpipers immediately before and immediately after the respirometry run, measured when birds were at room temperature (22 °C). Mean ± SD fi ff The VetScan VS2 analysis of plasma clinical chemistries provided a are given. Superscript letters indicate signi cant di erences between treatments (where – tested). level of sample hemolysis on a 0 3 scale, with 0 indicating no hemolysis and 3 indicating marked hemolysis. Three individuals were removed Controls 0.5 d oiling 3.5 d oiling from analysis because hemolysis index was ≥3 (one control, two from [N=14] [N=15] [N=15] the 3.5 d oiling group). fi ff Temperature [°C] 40.9 ± 1.0b 40.7 ± 0.8b 39.9 ± 1.0a There were signi cant di erences in eight of 35 analytes that were before evaluated in blood and liver of western sandpipers collected at necropsy Temperature [°C] 39.4 ± 0.8 39.3 ± 0.7 38.9 ± 0.8 (Table 4, Fig. 2). Albumin concentration was significantly lower in the after 3.5 d oiling group compared to the 0.5 d group (Nemenyi posthoc Difference −1.5 ± 1.1a −1.4 ± 1.2a −1.1 ± 0.9a comparisons, p=0.030) but did not differ from controls (P=0.160). Similarly, the albumin: globulin ratio differed significantly between the and the end of the respirometry run did not differ among treatments two oiling groups (P=0.019) but not between the 3.5 d oiling group and controls (P=0.312). Plasma urea concentrations were significantly (linear model: F2,37=1.508, P=0.235). The difference in Tb was significantly correlated with initial temperature (linear model: greater in both groups of oiled birds compared to controls (0.5 d vs. F =45.48, P < 0.001) and with initial body mass (F =4.53, control: P=0.006; 3.5 d vs. control: P=0.003). Plasma sodium 1,37 1,37 fi P=0.040). Heavier birds and birds with a higher initial temperature concentrations were slightly, but signi cantly lower in both groups of oiled birds compared to controls (0.5 d vs. control: P=0.043; 3.5 d vs. had a greater drop in Tb after the respirometry experimental night. control: P=0.010). Oxidative stress was assessed in all three groups by a cadre of fl 3.2. Respirometry biomarkers that re ect the oxidative status of an organism (Table 4, Fig. 3). Increases in reduced (0.5 d vs. control: P=0.166; 3.5 d vs. BMR measured at 27 °C differed significantly among treatments control: P=0.012) and total glutathione (0.5 d vs. control: P=0.044; 3.5 d vs. control: P=0.003) was observed in the two treated groups (linear model: F2,39=3.29, P=0.048) when controlling for the signifi- cant positive effect of body mass (F =4.48, P=0.041). Pairwise compared to the controls (although for reduced glutathione the 1,39 ff fi comparisons with Bonferroni corrections revealed no significant differ- di erence was not signi cant between the 0.5 d oiling group and controls). While oxidized glutathione was elevated in the two treatment ences among treatments, but consistent with the lower Tb of the 3.5 d oiling birds the BMR for the 3.5 d oiling birds was almost significantly groups compared to controls, these increases were not statistically fi fi lower than controls (P=0.055). The difference was not significant signi cant. Superoxide dismutase activity was signi cantly decreased between the 0.5 d and the 3.5 d oiling groups (P > 0.5) or between in the 3.5 d oiling group compared to the controls (P=0.007) but not in ff controls and 0.5 d oiling birds (P > 0.5). the 0.5 d oiling group (P=0.155). No di erences were observed between groups for glutathione peroxidase or for total antioxidant RMR increased with decreasing Ta in all three treatments (Fig. 1). A linear mixed effects model indicated significant effects of ambient power in liver tissues. Lipid peroxidation as determined by measuring malondialdehyde +4-hydroxylalkenals was increased in the 3.5 d temperature (F1,72=975.66, P < 0.001) and body mass (measured oiling group compared to the controls and the 0.5 d external oiling before the experiment: F1,39 =4.77, P=0.035), but no effects of group, although these differences were not significant after posthoc treatment (F2,39=2.70, P=0.080), nor of the treatment*temperature interaction term (not present in the best model, P=0.524 before model tests (0.5 d vs. control: P > 0.5; 3.5 d vs. control: P=0.068). simplification). Metabolic rate was higher at lower temperatures and 4. Discussion was positively correlated with body mass. The average values of thermal conductance calculated at −3°C − − − − ff were 2.56×10 2 W.°C 1 for controls, 2.56×10 2 W.°C 1 for the 0.5 d There were no detectable e ects of external oiling on basal (BMR) − − oiling group, and 2.48×10 2 W.°C 1 for the 3.5 d oiling group, and nor resting metabolic rate (RMR), or thermal conductance. However, ff the difference among treatments was not significant (one-way ANOVA: this experiment provided some insights into the e ects of preening and possible oil ingestion for a three-day period. Birds in the 3.5 d oiling F2,33=0.47, P=0.63). pre-treatment group had a decrease in Tb and an increased body mass loss (mostly fat). These aspects are likely not due to a loss of insulation from the feathers, since in that case we would have expected these birds to show increased thermal conductance during the respirometry run. Most likely, hypothermia and weight loss are results of the toxic effects of the ingested low doses of oil. Crude oil is known to cause such effects in birds (Hughes et al., 1990; Jenssen and Ekker, 1991; Culik et al., 1991; Jenssen, 1994; Balseiro et al., 2005). Oil is known to impair intestinal absorption and cause liver failure (Crocker et al., 1974; Miller et al., 1978; Szaro et al., 1978; Fry and Lowenstine, 1985; Leighton, 1986; Khan and Ryan, 1991; Balseiro et al., 2005), which in turn results in lower food absorption and weight loss. Hemolytic anemia, a common injury of oiled birds (Leighton et al., 1985; Leighton, 1986; Yamato et al., 1996; Newman et al., 2000; Seiser et al., 2000; Troisi et al., 2007) causes a decrease in oxygen carrying capacity that can affect thermo- regulation, and thus cause hypothermia (Wood, 1991). In our experi- ment, we did not observe a significant change in hemoglobin with Fig. 1. Metabolic rates of western sandpipers measured at different ambient tempera- oiling, but there were some birds with values below the reference tures. Error bars show standard errors. The graph shows raw values for metabolic rates (without controlling for body mass). The gray area represents the thermoneutral zone interval, suggesting that this might be a mechanism. In mammals, slight (TNZ) in which BMR was measured. Since we have no precise data on the range of the hypothermia is often showed as a response to the ingestion of toxicants, TNZ in western sandpipers, we added a shaded area representing the uncertainty. though in some cases hyperthermia/fever can be observed (Gordon,

123 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 118–128

Table 4 Mean ± SEM of 35 analytes measured after necropsy of western sandpipers following the respirometry experiment. Sample sizes are given in square brackets. χ2 and P-values of Kruskal- Wallis tests are also given.

Analyte Control 0.5 d oil 3.5 d oil χ2P

Hemoglobin (g/dl) 12.5 ± 0.3 [8] 12.1 ± 0.4 [14] 11.7 ± 0.4 [12] 1.94 0.380 Albumin (g/dl) 2.63 ± 0.29 [4] 1.06 ± 0.40 [7] 2.00 ± 0.15 [6] 6.97 0.031 Alpha 1 globulins (g/dl) 0.09 ± 0.01 [4] 0.12 ± 0.02 [7] 0.15 ± 0.04 [6] 1.34 0.511 Alpha 2 globulins (g/dl) 0.17 ± 0.02 [4] 0.19 ± 0.02 [7] 0.21 ± 0.03 [6] 0.74 0.691 Beta globulins (g/dl) 0.34 ± 0.03 [4] 0.39 ± 0.06 [7] 0.32 ± 0.04 [6] 0.92 0.633 Gamma globulins (g/dl) 0.76 ± 0.09 [4] 0.81 ± 0.13 [7] 0.85 ± 0.11 [6] 0.24 0.888 Albumin: globulin ratio 1.96 ± 0.22 [4] 2.05 ± 0.09 [7] 1.43 ± 0.18 [6] 7.39 0.025 Alanine aminotransferase (U/l) 181.8 ± 24.5 [8] 227.1 ± 25.6 [14] 232.7 ± 32.9 [12] 0.73 0.695 Alkaline phosphatase (U/l) 96.5 ± 5.7 [8] 81.1 ± 7.9 [14] 74.6 ± 4.7 [12] 4.34 0.114 Aspartate aminotransferase (U/l) 1024 ± 73 [6] 704 [1] 1012 ± 9 [2] 2.76 0.252 Creatine phosphokinase (U/l) 4119 ± 555 [5] 4725 ± 495 [11] 3124 ± 375 [11] 5.31 0.070 Gamma-glutamyl transferase (U/l) 13.25 ± 2.53 [8] 7.57 ± 0.66 [14] 8.50 ± 1.60 [12] 5.14 0.077 Bile Acids (µmol/l) 36.4 ± 1.0 [9] 36.6 ± 1.3 [15] 41.2 ± 3.5 [12] 0.35 0.838 Urea (mg/dl) 4.5 ± 0.4 [8] 7.1 ± 0.4 [14] 8.1 ± 0.8 [12] 12.71 0.002 Uric acid (mg/dl) 4.99 ± 0.85 [9] 5.07 ± 0.63 [15] 5.03 ± 0.59 [12] 0.01 0.993 Cholesterol (mg/dl) 408.3 ± 16.5 [8] 423.9 ± 10.7 [14] 397.2 ± 13.9 [12] 2.06 0.358 Glucose (mg/dl) 325.9 ± 8.7 [9] 316.2 ± 8.2 [15] 311.0 ± 8.9 [12] 1.10 0.576 Total protein (g/dl) 3.29 ± 0.12 [9] 3.13 ± 0.09 [15] 2.98 ± 0.11 [12] 3.48 0.176 White blood cells (109/l) 4.55 ± 0.40 [10] 4.61 ± 0.33 [15] 4.11 ± 0.37 [15] 2.28 0.319 Eosinophils (109/l) 0.01 ± 0.01 [10] 0.01 ± 0.01 [15] 0.02 ± 0.01 [14] 0.53 0.766 Lymphocytes (109/l) 626 ± 263 [10] 360 ± 218 [15] 376 ± 176 [15] 1.82 0.403 Monocytes (109/l) 0.22 ± 0.04 [10] 0.14 ± 0.02 [15] 0.30 ± 0.08 [15] 4.25 0.120 Neutrophils (109/l) 719 ± 558 [10] 767 ± 446 [15] 902 ± 368 [15] 4.88 0.087 Glutathione disulfide actual (nmol/mg) 1.01 ± 0.26 [8] 1.50 ± 0.15 [15] 1.51 ± 0.15 [15] 2.63 0.268 Glutathione peroxidase (nmol/min/mg protein) 127.7 ± 10.7 [10] 121.2 ± 7.2 [15] 110.9 ± 5.0 [15] 2.56 0.278 Glutathione reduced (nmol/mg) 26.4 ± 3.8 [8] 29.6 ± 1.0 [15] 41.4 ± 4.3 [15] 8.09 0.017 Glutathione total (nmol/mg) 26.9 ± 3.5 [10] 32.6 ± 1.0 [15] 44.4 ± 4.5 [15] 11.18 0.004 Malondialdehyde ± 4-Hydroxylalkenals (nmol/mg) 0.71 ± 0.09 [10] 0.70 ± 0.09 [14] 0.91 ± 0.06 [14] 6.01 0.050 Superoxide dismutase (U/mg) 1.58 ± 0.10 [10] 1.31 ± 0.05 [15] 1.22 ± 0.03 [15] 9.11 0.011 Trolox equivalents (µmol/mg) 0.13 ± 0.04 [10] 0.09 ± 0.00 [15] 0.09 ± 0.00 [15] 0.63 0.731 Calcium (mg/dl) 8.28 ± 0.15 [9] 8.09 ± 0.08 [15] 8.31 ± 0.14 [12] 1.78 0.411 Phosphorus (mg/dl) 5.00 [1] 6.55 ± 1.25 [2] 5.97 ± 0.81 [6] 1.70 0.427 Potassium (mmol/l) 6.90 [1] 3.90 ± 0.30 [2] 4.97 ± 0.56 [6] 3.20 0.202 Chloride (mmol/l) 124.5 ± 1.0 [8] 122.8 ± 1.1 [14] 121.1 ± 1.6 [12] 5.15 0.076 Sodium (mmol/l) 141.8 ± 0.7 [8] 138.1 ± 0.8 [14] 138.2 ± 2.1 [12] 9.26 0.010

2010). A similar response could be expressed by the small birds in our regulatory costs would arise if oiled birds came in contact with water, study. Possibly, lowering Tb could be an adaptive response promoting as has been shown to be the case in penguins (Erasmus et al., 1981). energy conservation in oiled birds (Reinertsen, 1989). In the 3.5 d oiling birds that were exposed to oil for 72 h and could The lack of effects of oil on thermal conductance were surprising: be expected to have acute effects due to oral exposure to weathered previous studies found effects of external oil application on metabolic MC252 crude oil, albumin concentration and therefore the albumin: rates of birds exposed to low temperatures (Hartung, 1967; McEwan globulin ratio decreased, sodium and chloride concentrations also and Koelink, 1973; Lambert et al., 1982; Butler et al., 1986). However, decreased, while urea concentration increased. Urea may increase due there were different results depending on the method used for measur- to prerenal (dehydration), renal, or postrenal pathology. There was no ing metabolic rate. Metabolic rates of oiled birds were significantly evidence of obstruction in any bird upon necropsy indicating that higher than controls when measured with the doubly-labelled water postrenal disease was not present. In dehydration, sodium, chloride, method, but no difference was noted from the results of respirometry and albumin concentration would be expected to increase due to measurements (Butler et al., 1986). We have no satisfactory explanation hemoconcentration, yet the 3.5 d oiling group mean concentrations as to why these results differed. The results of our study indicate that a decreased. Hence, there must be a component of loss of sodium, moderate level of oiling does not induce substantial increases in chloride, and albumin. Urea is filtered at the glomeruli, reabsorbed in thermal conductance in western sandpipers. This could be due to the the proximal tubule and collecting ducts and actively secreted in the excellent insulation provided by feathers in these arctic breeding loop of Henle. There was no significant change in the uric acid shorebirds, which may ensure that they have some resistance to some concentration indicating that active transport in the proximal tubule degree of feather fouling and damage. It is also possible that with a was undamaged. Active sodium reabsorption and passive chloride reduced Tb metabolic rate was consistently lower in the 3.5 d oiling pre- reabsorption occurs in the distal tubule, the collecting ducts and the treatment group, making differences in thermal conductance more ascending loop of Henle. Combining the uremia, hyponatremia, and difficult to detect. Extending the low temperature test using a helox hypochloridemia, the induced damage is likely in the loop of Henle in atmosphere to measure summit metabolism (Msum) may have revealed the kidney that would result in increased urea due to lack of secretion effects of oil treatment and should be considered in future studies. and decreased sodium and chloride due to lack of reabsorption. Loss of Shorebirds are semi-aquatic birds that rarely submerge themselves albumin and electrolytes could also occur in the gastrointestinal tract completely in water. However, they might come into contact with water but this would not cause the uremia unless there was active bleeding while feeding. When oiled, feathers become matted and more perme- into the gastrointestinal tract. A component of gastrointestinal loss is able (O’Hara and Morandin, 2010). It is possible that feather insulation also possible concurrent with loop of Henle dysfunction. would be affected in a wet environment. We would expect thermal Typically, anemia is assessed using packed cell volume (PCV), conductance of oiled birds to increase when birds are wet. Although we however, in this case we are using the iStat measurement for hemoglo- did not collect data on this aspect, we suggest that higher thermo- bin. PCV was difficult to obtain accurately for these birds because a

124 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 118–128

Fig. 2. Measured values for five relevant metabolites measured in western sandpipers at necropsy following the respirometry measurement. Dashed horizontal lines represent the reference intervals calculated from a set of control birds from several western sandpiper studies (see text for details). number of blood samples had a significant degree of lysis. The iStat hydrophilic species via conjugation to glutathione, sulfates or glucur- calculates a value for hematocrit, which should correlate to PCV, from onides, though likely not before some oxidation has occurred. Bio- the hemoglobin concentration. As such, hemoglobin was used as a chemical processes to mitigate oxidative damage are upregulated, but substitute for PCV in this study. Anemia, defined as < 11.2 g hemoglo- in the face of repeated dosing the oxidative activities manifest as bin/dl, occurred in six treated birds (three from the 3.5 d oiling group damage to red blood cells and organs, including liver, kidney and and three from the 0.5 d oiling group). Given the short time between spleen. dosing and final hemoglobin measurement, the low number of birds Endotherms face thermal stress on a seasonal as well as daily basis that exhibited anemia was not unexpected. In the double-crested and need to regulate heat production and dissipation in order to cormorant (Harr et al., 2017) using the same artificially weather maintain a functional Tb. Environmental stressors such as exposure to DWH oil that was applied to feathers every three days, a low number crude oil either externally or internally or both can exacerbate the need of birds was anemic on day 6, the first sampling day. Anemia to regulate heat. The processes for thermoregulation involves the progressed in the cormorant study until day 21 when all dosed birds modulation of metabolism (Blagojevic, 2007) that can result in presented with anemia. Since the western sandpipers only received a increased production of reactive oxygen species (Lin et al., 2008). single dose equivalent to approximately half of that used (per kg) for The imbalance between reactive oxygen species production and anti- the cormorants, it is unsurprising that only a subset of western oxidant capacities creates oxidative stress that can lead to oxidative sandpipers showed a downward trend in hemoglobin concentration damage. This represents a life-history trade-off for the endotherm of by 3.5 days post-oiling. The fact that a single application of a small reactive oxygen species generation for maintenance of constant Tb. amount of oil can cause anemia shows the potential damage that oil can Organisms have developed elaborate antioxidant defense mechanisms do to a migratory bird species, especially if oiling occurs just prior to to protect against oxidative stress induced by increased reactive oxygen the metabolically expensive and stressful migration. species generation (Halliwell and Gutteridge, 1989; Blagojevic, 2007). Toxicological effects of crude oil polycyclic aromatic hydrocarbons This antioxidant response has been shown in endotherms in response to (PAHs) result from metabolism and detoxification processes that acute cold stress conditions (Davidovic, 1999; 2004; Beamonte- generate oxidatively damaging oxides. PAHs, like other xenobiotics, Barrientos and Verhulst, 2013; Petit and Vézina, 2014). While the can activate nuclear receptors, such as orphan nuclear receptors and the results in this study did show differences in Tb only in the 3.5 d oiling aryl hydrocarbon receptor that mediate the activation of detoxification group, we did observe increases in oxidative stress in the two oil treated processes (for review see Xu et al., 2005). Phase 1 detoxification of groups. Increased reactive oxygen species generation as a result of PAHs involves hydroxylation by microsomal cytochrome P450 depen- thermoregulation with a concomitant upregulation of glutathione dent mixed-function oxygenases. These enzymes are responsible for antioxidant defenses is consistent with these observations (see also converting the lipophilic parent compounds into more polar hydroxides Petit and Vézina, 2014). Lipid peroxidation was increased in the 3.5 d and epoxides. Phase II metabolic processes will then inactivate these external oiling group suggesting that the longer exposure time and

125 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 118–128

Fig. 3. Measurements of oxidative stress markers in western sandpipers at necropsy following the respirometry measurements. Superscript letters indicate grouping following Nemenyi posthoc comparisons. No significant differences among groups were found for malondialdehyde +4-hydroxyalkenals after the posthoc tests. longer period of oxidative stress could result in oxidative damage. This cortical hypertrophy and lipid pneumonia by 8 days post-dosing does not seem to be linked with the cold exposure treatment, but should (Hartung, 1967). These ducks defecated oil after approximately 6 h, be related to the toxicity of oil ingested during the 3 days before the as opposed to the western sandpipers in our study which began to clear experiment. oil in feces after as little as 10 min (pers. obs.). It is possible that the As weathered MC 252 oil has decreased volatile components irritant effects of the DWH artificially weather oil were greater than oils compared to crude oil, respiratory exposure would be decreased in tested by Hartung (1967); however, it is also likely that western comparison to other petroleum products. Respiratory exposure can sandpipers are more sensitive to the irritant effects resulting in a rapid cause acute changes within one hour while oral or dermal exposure excretion rate and less PAH metabolism. Alternatively, faster gut takes additional time for ingestion, absorption, metabolism, and effect. clearance times in the sandpipers might be simply due to their smaller Hence, the likelihood of the 0.5 d oiled birds having toxicant induced size as compared to ducks. Although the western sandpipers were pathologic change is significantly less than for those oiled for 3.5 days, potentially consuming similar volumes per kilogram to the ducks, they as there was limited time for organismal exposure. Black ducks were showing fewer adverse effects of the oil, which could be due to (Hartung, 1963) to which 4 ml or 20 ml of “moderate” density oil, lack of exposure time to PAHs or insufficient time between exposure which is likely to be of similar density to the artificially weathered oil and necropsy to fully assess physiological effects. Upregulation of liver used here, were able to preen approximately 15% or 25% of the total oil oxidative stress pathways and changes in plasma markers of kidney and volume respectively for the three-day period. Unfortunately, there was gastrointestinal function such as sodium, albumin and urea are only one bird per oil volume test so these are only rough estimates. indicators that sufficient oil volumes are being preened to cause a However, if we assume these values are correct, then this would be toxicologic response, but there may not have been sufficient time from equivalent to approximately 0.6 ml/kg or 5 ml/kg. The approximate exposure to measurement of the analytes to fully evaluate the toxico- volume of oil applied to the western sandpipers was 0.3g per bird, with logical effects. an average initial body weight of approximately 33g. The specific In conclusion, western sandpipers exposed to external doses of gravity of the artificially weathered oil was 0.9 (Forth et al., 2017), MC252 crude oil did not show effects on thermoregulation after a therefore resulting in application of approximately 10 ml/kg, which is maximum of 3.5 d exposure. However, there were toxic effects due to twice the maximum oil application of the Hartung (1963) study. At the oil ingestion through preening or drinking contaminated water. The end of the experiment, western sandpipers that had been exposed to oil effects of oil on Tb might be related with oxidative stress, which was for three days still appear oiled, suggesting that the estimates made by confirmed by liver analyses, or could be the result of an adaptive Hartung (1963) are applicable to this species. As such, the potential oil energy-saving mechanism as a response to the high cost of being oiled. consumption could be as high as 2.5 ml/kg in a three-day period. Ducks Mild anemia was produced in some individuals and there were given a single 2g/kg bolus dose via a stomach tube of various oils (most indications of renal and/or gastrointestinal dysfunction and, therefore, of them heavy oils) developed gastrointestinal irritations, fatty liver, lack of electrolyte homeostasis. Longer duration of oil exposure would

126 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 118–128 probably result in more marked effects. This external exposure study, in Davidovic, V., Dokic, I., Petrovic, N., Durasevic, S., Cvijic, G., 1999. Activity of antioxidant enzymes in rat skeletal muscle and brown fat: effect of cold and which small doses of oil were slowly ingested over time, resulted in propranolol. J. Therm. Biol. 24, 385–389. more significant effects than a single large volume gavage (Bursian Eastin, W.C., Rattner, B.A., 1982. Effects of dispersant and crude oil ingestion on mallard et al., 2017b, in press), which suggests that gavage might not be the ducklings (Anas platyrhynchos). Bull. Environ. Contam. Toxicol. 29, 273–278. ff Erasmus, T., Randall, R.M., Randall, B.M., 1981. Oil pollution, insulation and body appropriate method for assessment of toxic e ects. temperatures in the Jackass penguin Spheniscus demersus. Comp. Biochem. Physiol. 69A, 169–171. Ethical standards Forth, H.P., Mitchelmore, C.L., Morris, J.M., Lay, C.R., Suttles, S.E., Lipton, J., 2017. Characterization of dissolved and particulate phases of water accommodated fractions used to conduct aquatic toxicity testing in support of the Deepwater Horizon This research is in compliance with regulations governing animal natural resource damage assessment. Environ. Toxicol. Chem.. http://dx.doi.org/10. research. Bird capture, housing, and experiments were performed under 1002/etc.3803. (submitted for publication). the guidelines of the University of Western Ontario Animal Use Sub- Friedrichs, K.R., Harr, K.E., Freeman, K.P., Szladovits, B., Walton, R.M., Barnhart, K.F., Blanco-Chavez, J., 2012. ASVCP reference interval guidelines: determination of de Committee (protocol 2012-027) and according to permit CA-0256 from novo reference intervals in veterinary species and other related topics. Vet. Clin. the Canadian Wildlife Service. Pathol. 41, 441–453. Fry, D.M., Lowenstine, L.J., 1985. Pathology of common murres and Cassin's auklets – Author contributions exposed to oil. Arch. Environ. Contam. Toxicol. 14, 725 737. Geffré, A., Concordet, D., Braun, J.P., Trumel, C., 2011. Reference Value Advisor: a new freeware set of macroinstructions to calculate reference intervals with Microsoft I.M., A.R.G. and C.G.G. conceived the idea and designed the study Excel. Vet. Clin. Pathol. 40, 107–112. Gordon, C.J., 2010. Response of the thermoregulatory system to toxic insults. Front. and developed the methods, K.M.D. participated in designing the study, Biosci. (Elite) E2, 293–311. I.M. and L.V.K. performed the experiments, S.J.B., K.E.H., J.E.L., C.A.P. Guglielmo, C.G., Williams, T.D., 2003. Phenotypic flexibility of body composition in and K.L.P. performed analyses in the lab, I.M. analyzed the data, I.M., relation to migratory state, age and sex in the western sandpiper. Physiol. Biochem. Zool. 76, 84–98. S.J.B., K.E.H., C.A.P. and C.G.G. wrote the manuscript. All authors have Guglielmo, C.G., McGuire, L.P., Gerson, A.R., Seewagen, C.L., 2011. Simple, rapid, and read and approved the final version of the manuscript. non-invasive measurement of fat, lean, and total water masses of live birds using quantitative magnetic resonance. J. Ornithol. 152, S75–S85. Acknowledgements Halliwell, B.H., Gutteridge, J.M.C., 1989. Free Radicals in Biology and Medicine. Oxford University Press, Oxford. Harr, K.E., Rishniw, M., Rupp, T., Cacela, D., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C., The studies appearing in this special issue were funded by the US. Healy, K., Horak, K., Link, J.E., Reavill, D., Bursian, S.J., Cunningham, F., 2017. Fish and Wildlife Service as part of the Deepwater Horizon Natural Echocardiographic assessment of double crested cormorants (Phalacrocorax auritus) dermally exposed to weathered MC252 crude oil. (in press) Ecotoxicol. Environ. Saf Resource Damage Assessment. We thank A. Macmillan, M. Dick, K.H. (submitted for publication). Elliott and M. Rebuli for help with bird care and during the experi- Hartung, R., 1967. Energy metabolism in oil-covered ducks. J. Wildl. Manag. 31, ments. I. Lipton and A. McFadden provided quality control of the data. 798–804. Hartung, R., Hunt, G.S., 1966. Toxicity of some oils to waterfowl. J. Wildl. Manag. 30, 564–570. References Hartung, R., 1963. Ingestion of oil by waterfowl. Pap. Mich. Acad. Sci. Arts Lett. 48, pp. 49–55. Henkel, J.R., Sigel, B.J., Taylor, C.M., 2012. Large-scale impacts of the Deepwater Andres, B.A., 1997. The Exxon Valdez oil spill disrupted the breeding of black Horizon oil spill: can local disturbance affect distant ecosystems through migratory – oystercatchers. J. Wildl. Manag. 61, 1322 1328. shorebirds? BioScience 62, 676–685. Balseiro, A., Espi, A., Marquez, I., Perez, V., Ferreras, M.C., Marín, J.G., Prieto, J.M., Hohtola, E., 1981. Tonic immobility and shivering in birds: evolutionary implications. ff 2005. Pathological features in marine birds a ected by the prestige's oil spill in the Physiol. Behav. 27, 475–480. – north of Spain. J. Wildl. Dis. 41, 371 378. Hohtola, E, 2004. Shivering thermogenesis in birds and mammals, In: Barnes, B.M., Barré, H., Geloen, A., Chatonnet, J., Dittmar, A., Rouanet, J.-L., 1985. Potentiated Carey, (Eds.), Life in the Cold: Evolution, Mechanisms, Adaptation, and Application. muscular thermogenesis in cold-acclimated muscovy duckling. Am. J. Physiol. 249, In: Proceedings of the 12th International Hybernation Symposium. Biological Papers – R533 R538. of the Univ. of Alaska, number 27. Fairbanks, AK, USA, Inst. of Arctic Biology, Univ. Beamonte-Barrientos, R., Verhulst, S., 2013. Plasma reactive oxygen metabolites and non- of Alaska, pp. 241–252. ff enzymatic antioxidant capacity are not a ected by an acute increase of metabolic Hoppe, B.R., Lassen, E.D., 1978. Blood Smears and the Use of Wright's Stain 40. Iowa fi – rate in zebra nches. J. Comp. Physiol. B 183, 675 683. State University Veterinarianpp. 113–116. Bennett, P.M., Harvey, P.H., 1987. Active and resting metabolism in birds: allometry, Hudson, J.W., Kimzey, S.L., 1966. Temperature regulation and metabolic rhythms in – phylogeny and ecology. J. Zool. 213, 327 344. populations of the house sparrow, Passer domesticus. Comp. Biochem. Physiol. 17, Blagojevic, D.P., 2007. Antioxidant systems in supporting environmental and 203–217. – programmed adaptations to low temperatures. CryoLetters 28, 137 150. Hughes, M.R., Kasserra, C., Thomas, B.R., 1990. Effect of externally applied bunker fuel ff Burger, J., 1997. E ects of oiling on feeding behavior of sanderlings and semipalmated on body mass and temperature, plasma concentration, and water flux of glaucous- – plovers in New Jersey. Condor 99, 290 298. winged gulls, Larus glaucescens. Can. J. Zool. 68, 716–721. ff Burger, J., Tsipoura, N., 1998. E ects of oiling of sanderlings (Calidris alba): behavior and Jenssen, B.M., 1994. Effects of oil pollution, chemically treated oil, and cleaning on the – weight changes. Environ. Toxicol. Chem. 17, 1154 1158. thermal balance of birds. Environ. Pollut. 86, 207–215. Bursian, S.J., Cacela, D., Cunningham, F.L., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C., Jenssen, B.M., Ekker, M., 1991. Dose dependent effects of plumage-oiling on Healy, K., Horak, K., Link, J.E., Reavill, D.R., Schmidt, R.E., Shriner, S., Harr, K.E., thermoregulation of common eiders Somateria mollissima residing in water. Polar Res. 2017a. Comparison of organ weights and histopathology between double-crested 10, 579–584. fi cormorants (Phalacrocorax auritus) dosed orally or externally with arti cially Khan, R.A., Ryan, P., 1991. Long term effects of crude oil on common murres (Uria aalge) weathered Mississippi canyon 252 crude oil. Ecotoxicol. Environ. Saf (in press). following rehabilitation. Bull. Environ. Contam. Toxicol. 46, 216–222. Bursian, S.J., Dean, K.M., Harr, K.E., Kennedy, L.V., Link, J.E., Maggini, I., Pritsos, C., Lambert, G., Peakall, D.B., Philogène, B.J.R., Engelhardt, F.R., 1982. Effect of oil ff Pritsos, K.L., Schmidt, R.E., Guglielmo, C.G., 2017b. E ect of oral exposure to dispersant mixtures on the basal metabolic rate of ducks. Bull. Environ. Contam. fi arti cially weathered Deepwater Horizon crude oil on hematology and plasma Toxicol. 29, 520–524. chemistries, hepatic antioxidant enzyme activity, organ weights and histopathology Larsen, E.M., Richardson, S.A., 1990. Some effects of a major oil spill on wintering in western sandpipers (Calidris mauri). (this issue) Ecotoxicol. Environ. Saf (in press). shorebirds at Grays Harbor, Washington. Northwest. Nat. 71, pp. 88–92. ff Butler, R.G., Peakall, D.B., Leighton, F.A., Borthwick, J., Harmon, R.S., 1986. E ects of Lasiewski, R.C., Dawson, W.R., 1967. A re-examination of the relation between standard crude oil exposure on standard metabolic rate of Leach's storm-petrel. Condor 88, metabolic rate and body weight in birds. Condor 69, 13–23. – 248 249. Lee, Y.Z., O’Brien, P.J., Payne, J.F., Rahimtula, A.D., 1986. Toxicity of petroleum crude Campbell, T.W., Ellis, C.K., 2013. Avian and Exotic Animal Hematology and Cytology. oils and their effec on xenobiotic metabolizing enzyme activities in the chicken John Wiley & Sons. embryo in ovo. Environ. Res. 39, 153–163. ff Crocker, A.D., Cronshaw, J.T., Holmes, W.N., 1974. The e ect of a crude oil on intestinal Leighton, F.A., 1986. Clinical, gross, and histological findings in herring gulls and Atlantic – absorption in ducklings (Anas platyrhynchos). Environ. Pollut. 7, 165 177. puffins that ingested Prudhoe Bay crude oil. Vet. Pathol. 23, 254–263. Culik, B.M., Wilson, R.P., Woakes, A.T., Sanudo, F.W., 1991. Oil pollution of Antarctic Leighton, F.A., Peakall, D.B., Butler, R.G., 1983. Heinz-body hemolytic anemia from the ff penguins: e ects on energy metabolism and physiology. Mar. Pollut. Bull. 22, ingestion of crude oil: a primary toxic effect in marine birds. Science 220, 871–873. – 388 391. Leighton, F.A., Lee, Y.Z., Rahimtula, A.D., O’Brien, P.J., Peakall, D.B., 1985. Biochemical ff Davidovic, V., Petrovic, N., Dordevic, J., Durasevic, S., Cvijic, G., 2004. Acute e ect of and functional disturbances in red blood cells of herring gulls ingesting Prudhoe Bay cold on the antioxidant enzymes activities and uncoupling protein-1 content in the crude oil. Toxicol. Appl. Pharmacol. 81, 25–31. – brown fat of 6-hydroxydopamine-treated rats. J. Therm. Biol. 24, 385 389. Lighton, J.R.B., 2008. Measuring Metabolic Rates. Oxford University Press, New York.

127 I. Maggini et al. Ecotoxicology and Environmental Safety 146 (2017) 118–128

Lin, H., De Vos, D., Decuypere, E., Buyse, J., 2008. Dynamic changes in parameters of resulted in oxidative stress. Ecotoxicol. Environ. Saf. (in press). redox balance after mild heat stress in aged laying hens (Gallus gallus domesticus). Prum, R.O., 1999. Development and evolutionary origin of feathers. J. Exp. Zool. 285, Comp. Biochem. Physiol. 147C, 30–35. 291–306. Maggini, I., Kennedy, L.V., Macmillan, A., Elliott, K.H., Dean, K.M., Guglielmo, C.G., R Core Team, 2012. R: A language and environment for statistical computing. R submitted for publication. light oiling of feathers increases flight energy expenditure Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0 URL. in a migratory shorebird. J. Exp. Biol. (in press). 〈http://www.R-project.org/〉. Maggini, I., Kennedy, L.V., Elliott, K.H., Dean, K.M., MacCurdy, R., Macmillan, A., Pritsos, Regal, P.J., 1975. The evolutionary origin of feathers. Q. Rev. Biol. 50, 35–66. C.A., Guglielmo, C.G., 2017. Trouble on takeoff: crude oil on feathers reduces escape Reinertsen, R.E., 1989. The regular use of nocturnal hypothermia and torpor during performance of shorebirds. Ecotoxicol. Environ. Safe. 141, 171–177. energetically-demanding periods in the annual cycles of birds. Living in the cold II. McEwan, E.H., Koelink, F.C., 1973. The heat production of oiled mallards and scaup. Can. Elsevier, New York, pp. 107–116. J. Zool. 51, 27–31. Ruthrauff, D.R., Dekinga, A., Gill, R.E., Piersma, T., 2013. Identical metabolic rate and McKechnie, A.E., Freckleton, R.P., Jetz, W., 2006. Phenotypic plasticity in the scaling of thermal conductance in rock sandpiper (Calidris ptilocnemis) subspecies with avian basal metabolic rate. Proc. R. Soc. Lond. B 273, 931–937. contrasting nonbreeding life histories. Auk 130, 60–68. McNab, B.K., 1997. On the utility of uniformity in the definition of basal rates of Scholander, P.F., Hock, R., Walters, V., Johnson, F., Irving, L., 1950. Heat regulation in metabolism. Physiol. Zool. 70, 718–720. some arctic and tropical mammals and birds. Biol. Bull. 99, 237–258. Miller, D.S., Peakall, D.B., Kinter, W.B., 1978. Ingestion of crude oil: sublethal effects in Seiser, P.E., Duffy, L.K., McGuire, A.D., Roby, D.D., Golet, G.H., Litzow, M.A., 2000. herring gull chicks. Science 199, 315–317. Comparison of pigeon guillemot, Cepphus columba, blood parameters from oiled and Montevecchi, W., Fifield, D., Burke, C., Garthe, S., Hedd, A., Rail, J.F., Robertson, G., unoiled areas of Alaska eight years after the Exxon Valdez oil spill. Mar. Pollut. Bull. 2012. Tracking long-distance migration to assess marine pollution impact. Biol. Lett. 40, 152–164. 8, 218–221. Sharp, B.E., Cody, M., Turner, R., 1996. Effects of the Exxon Valdez oil spill on the black Newman, S.H., Anderson, D.W., Ziccardi, M.H., Trupkiewicz, J.G., Tseng, F.S., oystercatcher. Am. Fish. Soc. Symp. 18, 748–758. Christopher, M.M., Zinkl, J.G., 2000. An experimental soft-release of oil-spill Szaro, R.C., Dieter, M.P., Heinz, G.H., Ferrell, J.F., 1978. Effects of chronic ingestion of rehabilitated American coots (Fulica americana): ii. Effects on health and blood South Louisiana crude oil on mallard ducklings. Environ. Res. 17, 426–436. parameters. Envrion. Pollut. 107, 295–304. Troisi, G., Borjesson, L., Bexton, S., Robinson, I., 2007. Biomarkers of polycyclic aromatic O’Hara, P., Morandin, L.A., 2010. Effects of sheens associated with offshore oil and gas hydrocarbon (PAH)-associated hemolytic anemia in oiled wildlife. Environ. Res. 105, development on the feather microstructure of pelagic seabirds. Mar. Pollut. Bull. 60, 324–329. 672–678. Vaillancourt, E., Prud’Homme, S., Haman, F., Guglielmo, C.G., Weber, J.-M., 2005. Parkes, K., 1966. Speculations on the origin of feathers. Living Bird. 5, 77–86. Energetics of a long-distance migrant shorebird (Philomachus pugnax) during cold Pattee, O.H., Franson, J.C., 1982. Short-term effects of oil ingestion on American kestrels exposure and running. J. Exp. Biol. 208, 317–325. (Falco sparverius). J. Wildl. Dis. 18, 235–241. Vézina, F., Dekinga, A., Piersma, T., 2011. Shorebirds' seasonal adjustments in Perry, M.C., Ferrigno, F., Settle, F.H., 1978. Rehabilitation of birds oiled on two mid- thermogenic capacity are reflected by changes in body mass: how preprogrammed Atlantic estuaries. In: Proceedings of the Annual Conference of South Eastern and instantaneous acclimation work together. Int. Comp. Biol. 51, 394–408. Fish & Wildlife Agencies 32, 318–325. Walton, P., Turner, C.M.R., Austin, G., Burns, M.D., Monaghan, P., 1997. Sub-lethal Petit, M., Vézina, F., 2014. Elevating antioxidant defence is part of cold acclimatization in effects of an oil pollution incident on breeding kittiwakes Rissa tridactyla. Mar. Ecol. a small forest bird wintering in eastern Canada. Integr. Comp. Biol. 54 (suppl), e164. Prog. Ser. 155, 261–268. Piersma, T., 1995. Seasonality in basal metabolic rate and thermal conductance in a long- Wood, S.C., 1991. Interactions between hypoxia and hypothermia. Annu. Rev. Physiol. distance migrant shorebird, the knot (Calidris canutus). J. Comp. Physiol. B 165, 53, 71–85. 37–45. Xu, C., Li, C.Y., Kong, A.N., 2005. Induction of phase I, II and II drug metabolism/ Pritsos, K.L., Perez, C.R., Muthumalage, T., Dean, K.M., Cacela, D., Hanson-Dorr, K.C., transport by xenobiotics. Arch. Pharmacol. Res. 28, 249–268. Cunningham, F., Shriner, S., Horak, K., Pritsos, C.A., 2017. this issue. Dietary intake Yamato, O., Goto, I., Maede, Y., 1996. Hemolytic anemia in wild seaducks caused by of Deepwater Horizon Oil injected live food fish by double-crested cormorants marine oil pollution. J. Wildl. Dis. 32, 381–384.

128 Ecotoxicology and Environmental Safety 146 (2017) 129–133

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Reprint of: Assay Validation of the Cardiac Isoform of Troponin I in Double MARK Crested Cormorant (Phalacrocorax auritus) Plasma for Diagnosis of Cardiac ☆ Damage ⁎ Melissa Daigneaulta, Kendal E. Harrb, , Karen M. Deanc, Steven J. Bursiand a Center for Bird and Exotic Animal Medicine. 11401 NE 195th Street. Bothell, WA 98011 USA b Urika, LLC. 8712 53rd Place West, Mukilteo, WA 98275 USA c Abt Associates, 1881 Ninth St., Suite 201, Boulder, CO 80302 USA d Department of Animal Science, Michigan State University, East Lansing, MI 48824 USA

ARTICLE INFO ABSTRACT

Keywords: Cardiac abnormalities, initially found in Deepwater Horizon weathered MC252 crude oil exposed Double Crested Biomarker Cormorants (DCCOs) upon gross necropsy, were further investigated using echocardiography. Clinical and Deepwater Horizon statistically significant changes including decreased ventricular myocardial contractility and arrhythmia were Heart elucidated by echocardiography and interpreted by boarded cardiologists as potentially life threatening. The Oil-induced cardiomyopathy objective of this investigation was to initiate development of an antemortem, sensitive blood screening test for Polycyclic aromatic hydrocarbons (PAH) cardiac damage due to oil exposure of avian species. An assay for the cardiac isoform of troponin I (cTnI) which is known to be highly cross-reactive across mammalian species was chosen and analytically validated in DCCO. This is the first time this test has been analytically validated in avian species. All plasma samples from birds assessed as healthy had trace concentrations (< 0.016 ng/ml). The assays was precise and accurate revealing a coefficient of variation < 3% and an R2 > 0.99. Diagnostic investigation revealed that the test appears to have diagnostic potential for the diagnosis of cardiomyocyte damage. Diagnostic sensitivity and specificity were 91% and 73% in this laboratory population. Due to an equivocal sample population in which health could not be proven, further investigation is needed to diagnostically validate troponin I in the assessment of oil exposure in DCCO.

1. Introduction Diagnostic tools for cardiac assessment in standard diagnostic practice are either lacking or under used, due to anatomic and biochemical Necropsies performed as part of a weathered MC252 Deepwater differences between birds and mammals. Thus, the aim of this study Horizon oil exposure study (Cunningham et al., 2017) uncovered gross was to improve avian cardiac assessment capability, specifically post-oil morphological cardiac changes in double-crested cormorants (DCCO) exposure, by evaluating a troponin I ELISA immunoassay for use as a that indicated further evaluation of the heart was needed. In additional cardiac biomarker in DCCO. studies, echocardiographic analysis of DCCO dermally exposed to Cardiac troponin is a calcium protein complex responsible for the weathered MC252 revealed cardiac dysfunction, including dilative actin-myosin cross bridging that allows myocardial contraction. change in the ventricles and arrhythmia. (Harr et al., 2017). Thus, a Damage to the myocardiocyte membranes releases troponin I from simpler antemortem test for cardiac damage is warranted. Anemia, cardiac tissue and results in increased concentrations in the peripheral disrupted feather function, hypothermia, respiratory distress, seizures, blood. The cardiac isoform of troponin I (cTnI) is the standard of care in diarrhea, hepatic disease and renal disease (Mazet et al., 2002) have all human medicine for noninvasively diagnosing acute myocardial infarc- been reported in avian species secondary to exposure to petroleum tions (Antman, 2002; Sarko and Pollack, 2002; Newby et al., 2012; products. To the authors’ knowledge, cardiac disease due to petroleum Apple and Saenger, 2013). Cardiac troponin I also has been associated exposure has never been previously documented in adult birds. with a range of cardiac diseases in canine and feline patients as well as

DOI of original article: http://dx.doi.org/10.1016/j.ecoenv.2017.03.006 ☆ A publisher's error resulted in this article appearing in the wrong issue. The article is reprinted here for the reader's convenience and for the continuity of the special issue. For citation purposes, please use; Ecotoxicology and Environmental Safety Volume 141 pp. 52-56. ⁎ Corresponding author. E-mail address: [email protected] (K.E. Harr). http://dx.doi.org/10.1016/j.ecoenv.2017.05.016 Received 21 September 2016; Received in revised form 3 March 2017; Accepted 6 March 2017 Available online 30 May 2017 0147-6513/ © 2017 Published by Elsevier Inc. M. Daigneault et al. Ecotoxicology and Environmental Safety 146 (2017) 129–133 diseases that have secondary cardiac damage (Langhorn et al., 2013; Medcalc, Co., Ostend, Belgium). Diagnostic sensitivity and specificity Winter et al., 2014). CTnI's structure and function is highly conserved were manually calculated using the formulas among mammalian and avian species (Hastings et al., 1991) and has diagnostic sensitivity = true positive/(true positives + false negatives) been evaluated in marmosets, swine, cattle, guinea pigs, rats and mice (O’Brien et al., 2006). Therefore, this potential biomarker of myocar- diagnostic specificity = true negatives/(true negatives + false positives) diocyte damage was chosen for further evaluation and potential assessment. 3. Results 2. Materials and methods All pretreatment samples (n=26) contained trace concentrations of Samples for this study were collected from 26 DCCO captured in cTnI (< 0.016 ng/ml). Post- treatment samples from control birds had a Mississippi and Alabama in accordance with National Wildlife Research mean; median ± SD of 11,672; 2896 ± 16,030 and a range of – Center standard operating procedures under IACUC approval number 2 45,569 ng/ml. Post-treatment samples from treated birds had a QA2326. All birds were captured and retained in captivity under the mean; median ± SD of 7339; 2021 ± 11,342.34 and a range of – authority of Federal Permit # MB019065-3, and Mississippi and 0 35,043 ng/ml. Comparison using a two-tailed Student's t-test of equal fi ff Alabama state (#8017) scientific collection permits. Cormorants were variance revealed no signi cant di erence (P=0.47) between post individually housed as described by Cunningham et al. (2017). Briefly, treatment control and treated birds. Comparison of samples from birds were allowed to acclimate to captivity for a minimum of 21 days posttreatment birds to their pretreatment samples in a paired t-test fi ff prior to initiation of the study. A total of 25 DCCOs allocated to a revealed a signi cant di erence (P < 0.001). control group (n=12, 5 male, 7 female) and an exposed group (n=13, 6 males, 7 females) were used in this trial. During the course of the trial, 3.1. Accuracy one bird from the control group and two birds from the treatment group died and were not replaced. Therefore, the final number of birds in the Results of the serial dilution are presented in Table 1 and linear control and exposed group was 11 birds each to total 22 in the study. regression is presented in Fig. 1. Visual inspection of the linear Birds were exposed to petroleum via topical application over the breast regression graph indicates the data are linear and symmetrically and back at a moderate or 16–40% coverage. Oil on exposed birds distributed. Linear regression analysis reveals an intercept of 2 (13 g) and water on control birds (13 g) was applied every three days −0.29 ng/ml, a slope of 1.01 and R > 0.99. P values for the intercept through day 15 of the trial (on days 0, 3, 6, 9, 12, and 15). Detailed and slope are less than 0.02 and 0.01, respectively. The 95% Confidence description of application is available in Cunningham et al. (2017). Interval of the slope and y-intercept were calculated to be 0.98 to 1.05 Blood samples for cTnI analysis were collected at the end of and −0.52 to −0.053, respectively, indicating excellent agreement. acclimation on day −3 and just prior to euthanasia and necropsy on days 23 and 24 via a heparinized syringe from the jugular, tarsal, or 3.2. Precision basilic vein. Immediately after collection, samples were placed in lithium heparin, centrifuged, and archived in an ultra-cold freezer at Standard deviation (SD) and coefficient of variation (CV) across NWRC-MSFS for later analysis. All blood samples were processed to clinically relevant concentrations produced in this experiment were storage archive within 4 h of collection. Each sample underwent one calculated (Table 2). SD was highest at the 1:2 dilution at 0.09 ng/ml freeze-thaw cycle. and smallest at 0.01 ng/ml for 1:8, 1:16, and 1:32 dilutions. CV ranged Forty-eight cormorant samples (26 baseline and 22 post-study) were from 0.7% to 2.5% with the most precise measurement at the 1:8 analyzed concurrently in accordance with manufacturer directions dilution and the least precise measurement at the 1:32 dilution. using the Centaur Cardiac Troponin I TnI-Ultra assay (Siemens Comparison of the manufacturer's precision curves compared to DCCO Healthcare Diagnostics Inc., Tarrytown, NY) on the ADVIA Centaur plasma were produced by plotting the mean troponin concentration CP System (Siemens Healthcare Diagnostics Inc., Tarrytown, NY). This versus total CV (Fig. 2). system automatically performs the sandwich immunoassay via the following steps. One hundred µL of sample, 100 µL of polyclonal goat 3.3. Analytic sensitivity anti-troponin I antibody labeled with acridinium ester and 2 biotiny- lated mouse monoclonal anti-troponin I antibodies, and 50 µL of To evaluate analytic sensitivity, the lower limit of detection (LLOD) ancillary reagent are combined and incubated for 2.75 min at 37 °C. was evaluated by a serial blank measurement. A blank sample After the incubation period, 150 µL of magnetic latex particles con- composed of reagent and sterile water was analyzed 20 times in jugated with streptavidin is added and incubated for 5 min at 37 °C. The duplicate over 10 days and produced 0.016 ng/ml as the highest result. system then separates, aspirates, and washes the cuvettes with ADVIA This is above the vendor claim of 0.006 ng/ml. The LLOD can also be Centaur Wash 1 (phosphate buffered saline with sodium azide and calculated by the analyte value where the CV is unacceptable, in this surfactant). The chemiluminescent reaction is then initiated by adding case a CV that exceeds 20%. However, none of the CVs approached this 300 µL of Acid Reagent and 300 µL Base Reagent. Relative light units value. (RLUs) are detected by the system and reported as a numerical value. The four samples with the highest values were combined to form a Table 1 single pooled sample of adequate volume. The pooled sample was Calculated (Expected) and Measured Diluted Samples. assayed in triplicate and the average was reported. The remaining SAMPLE Calculated diluted Average of measured diluted % accuracy volume of the high value pool was diluted with ADVIA Centaur Multi- sample (ng/ml) sample (ng/ml) Diluent 11 in 1:2 serial dilutions to create a 1:2, 1:4, 1:8, 1:16, and 1:32 set of dilutions. Five replicates of each dilution were analyzed. These 1:1 13.43 13.43 100.00 results were used to establish accuracy, precision, and analytic sensi- 1:2 6.72 6.36 94.64 1:4 3.36 3.02 89.88 tivity (lower limit of detection) of the cTnl-Ultra assay using DCCO 1:8 1.68 1.34 79.76 heparinized plasma. To investigate accuracy, serial dilutions averages 1:16 0.84 0.62 73.81 were then compared to the calculated (expected) dilution results. 1:32 0.42 0.31 73.81 Statistical analysis including linear regression, correlation, and descriptive statistics were performed using Medcalc (version 16.4.3

130 M. Daigneault et al. Ecotoxicology and Environmental Safety 146 (2017) 129–133

Table 3A Samples I included in sensitivity and specificity calculations.

A Oil treated birds with troponin levels > 0.016 ng/ml at day 23, 24 B Control birds with troponin levels > 0.016 ng/ml at day 23, 24 C Oil treated birds with troponin levels < 0.016 ng/ml at day 23, 24 D All baseline (day −3) Samples E Control birds with troponin levels < 0.016 ng/ml ad day 23, 24

Table 3B Sensitivity and specificity chart with all samples included.

Sensitivity and specificity with all samples Sensitivity and specificity with all Fig. 1. Comparison of calculated and measured troponin concentrations. The solid line is included samples included the line of best fit. The dashed lines indicate 1 SD. True True True True Table 2 positive negative positive negative Average, SD, and CV of diluted samples. Test positive 10 10 Test AB SAMPLE Average (ng/ml) SD (ng/ml) CV (%) positive Test negative 1 27 Test CD,E 1:2 6.36 0.09 1.40 negative 1:4 3.02 0.05 1.60 1:8 1.34 0.01 0.68 1:16 0.62 0.01 1.4 1:32 0.31 0.01 2.50 Table 3C Sensitivity and specificity chart without post-study samples from control birds.

Sensitivity and specificity without Sensitivity and specificity without necropsy samples from control birds necropsy samples from control birds

True True True True positive negative positive negative

Test positive A B Test 10 None positive Test negative C D, E Test 126 negative

3.5. Reference interval

All samples collected on Day −3 (n=26) were below the limit of detection (< 0.016 ng/ml). These samples were collected before the study was initiated when birds were quarantined on ponds, i.e. in a natural environment before application of petroleum and housing and Fig. 2. Precision curve of cormorant troponin levels compared to the manufacturer handling stress that caused trauma. Therefore, this is the time at which reported precision curve. birds are assessed as healthy by the authors. The normal reference interval is proposed as less than or equal to 0.016 ng/L.

3.4. Diagnostic sensitivity and specificity 4. Discussion

To calculate diagnostic sensitivity and specificity each sample was In the present study, the Centaur CP TnI-ultra immunoassay was classified as true positive, true negative, false positive or false negative. analytically validated and found to be likely diagnostically useful for All baseline samples (collected at day −3) were classified as negatives. the detection of cardiomyocyte damage in avian species, though The necropsy sample from control birds not exposed to oil were differentiation of oil-induced damage was not possible in this study. classified as negative although the population was considered to be Analytical evaluation of this assay revealed good to excellent accuracy, equivocal as animals were housed in laboratory conditions with precision, and analytic sensitivity. These values were similar to or concrete flooring. Oil-treated birds were classified as positive. The better than studies evaluating cTnI in canine species Table 4 (Winter treated birds with troponin levels greater than 0.016 ng/ml at day 23 or et al., 2014; Langhorn et al., 2013). The American Society of Veterinary 24 were classified as true positives. The control birds with troponin Clinical Pathologists (ASVCP) published guidelines recommend no levels greater than 0.016 ng/ml at day 23 or 24 were classified as false more than a 20% CV for troponin measurements (Harr et al., 2013). positives. Twenty-seven samples were classified as true negatives; ten The CV for this validation study was well under the ASVCP recommen- samples as true positives. One sample was classified as false negative; dation at no more than 2.5%. ten samples were classified as false positives. Diagnostic sensitivity was Although diagnostic sensitivity was excellent, the diagnostic speci- 91% and diagnostic specificity was 73%. If the control birds, housed on ficity was found to be lower than desired due to a high number of false concrete, with troponin levels greater than 0.016 ng/ml are excluded, positives. Reasons for the high occurrence of false positives, in necropsy the diagnostic specificity becomes 100%. Table 3A, Table 3E, Table 3C. samples from post-treatment control birds, include physical damage due to laboratory housing leading to cardiac damage. The cormorants

131 M. Daigneault et al. Ecotoxicology and Environmental Safety 146 (2017) 129–133

Table 4 oped over the past two decades for the human market which have Comparison of accuracy and precision to Winter et al. (2014) and Langhorn et al. (2013). variable quality assessment and diagnostic utility due to antibody and assay differences (Apple and Saenger, 2013). The assay chosen in this Accuracy (slope) Accuracy (Y Precision (CV) ng/ml intercept) ng/ml cormorant study uses an epitope in a conserved region as it is cross- reactive with every mammal tested. Concentration of cTnI also Winter et al. n/a n/a < 5.7% determined by iSTAT (Heska Corp., Loveland, CO) was reported as a (canine) − potential diagnostic aid in a captive Red-tailed Hawk (Buteo jamaicen- Langhorn et al. 1.00 0.2 < 6.4% fi (canine) sis) later de nitively diagnosed with right sided cardiomyopathy and Langhorn et al. 1.05 −0.05 < 7.8% congestive heart failure (Knafo et al., 2011). This study did not find (feline) cTnI concentrations to be useful in the cardiomyopathy diagnosis. Present study 1.01 −0.29 < 2.5% While sample number is low, these results indicate that the antibody (DCCO) used in the iSTAT assay now marketed by Abaxis, Inc., Union City, CA is likely not crossreactive in avian species. The cTnI iSTAT assay is validated for dogs and is not supported by Abaxis, Inc. in avian species. were housed on concrete floors and would often crash land on the hard With further investigation, cTnI can be useful in identifying avian surface of the concrete when trying to move between the feeding tank patients with cardiac disease. CTnI, is a valuable, sensitive diagnostic and their secondary perch. This accidental self-damaging behavior, screening tool to establish cardiac damage in humans and canines and which also resulted in increased creatine kinase activities (Dean et al., has been proposed as a biomarker for assessing cardiac damage in pre- 2017), could artificially increase the troponin concentrations. The high clinical toxicity testing (Reagan, 2010). However, cTnI, it is not used as creatine kinase reference value for cormorants housed at this facility a replacement for, but rather in conjunction with, echocardiogram and was greater than 2000 IU. Typically, domestic bird species have high electrocardiogram (Thygesen et al., 2012; Langhorn et al., 2013; Newby creatine kinase reference values of 500 IU with that value doubling to et al., 2012) in diagnosis of definitive cardiac disease. Mildly increased 1000 IU in wild-caught or rehabilitated animals due to increased cTnI concentrations have been associated with renal disease in humans capture and handling trauma typically expected. In waterfowl, such and canines. The mechanism behind this noncardiogenic increase has as coots’ rehabilitation values double from a mean of 56 to 140 IU in not been elucidated but does not appear to be secondary to azotemia comparison to controls and again are much lower than what we found (Winter et al., 2014). Further investigation is needed to understand the in the cormorants (Newman et al., 2000) If baseline values were limitations of cTnI in cormorants and to validate cTnI in other avian considered true negatives and all samples from birds at necropsy were species. considered true positives, diagnostic specificity improved to 100% and Increased cTnI levels have been correlated with increased ischemia diagnostic sensitivity remained the same at 91%. This indicates that and a poorer prognosis in humans (Newby et al., 2012) and in dogs cTnI has potential to be a biomarker for cardiac myocyte damage in free (Fonfara et al., 2010). A correlation between increased cTnI and more ranging cormorants, including in oil exposure. Further study, where severe cardiovascular lesions on necropsy was not elucidated in this cormorants are housed in a more natural habitat, such as on ponds and study. Further information is needed prior to correlating severity of dirt, is warranted to better assess the diagnostic specificity of this assay. increased cTnI levels with severity of disease and prognosis. In dogs The proposed reference interval is therefore based on day −3 cTnI levels begin to rise after 2 h and peak in 12–24 h after cardiac samples when birds were housed on ponds and minimally handled for ischemia (Fonfara et al., 2010). There were only two sampling time experimentation. All samples collected during the quarantine period points in this study designed to analytically validate this assay in (n=26) prior to cardiac insult were below the limit of detection cormorants. cTnI levels are not available before day 23 or 24 post oil (0.016 ng/ml). This is similar to the mean canine reference values of exposure. Further studies to determine the peak of cTnI after myocyte 0.017 ng/ml, 0.02 ng/ml or 0.015 ng/ml found by Winter et al. (2014), damage are still needed. Langhorn et al. (2013), and Serra et al. (2010) respectively. In human Also reported in this issue, cardiac damage documented via studies the 99th percentile of apparently healthy humans ranging from echocardiogram and necropsy was associated with topical application 17 to 91 years of age was demonstrated to be 0.04 ng/ml. Diagnostic of weathered MC252 (Harr et al., 2017). Besides this study which cut off was recommended by European Society of Cardiology and the included the same population of subjects, oil has not been reported to American College of Cardiology as a result exceeding the 99th cause cardiac disease in adult avian species to the author's knowledge. percentile (Siemens Centaur insert). As stated previously, the cormor- Oil exposure has been established to cause cardiac damage in a broad ant reference interval will need further study to attain the ASVCP range of piscine species including sol, zebrafish, seabass and tuna recommended number of 80 samples for complete validation. (Claireaux and Davoodi, 2010; Tissier et al., 2015; Brette et al., 2014). Experimentally testing analytic specificity and reportable range of The cardiac damage is theorized to occur by an impairment of the test results were beyond the scope of this project. According to the cardiac excitation-contraction coupling mechanism via a decrease in manufacturer, possible interferents will impact the assay as follows: calcium current and calcium cycling (Brette et al., 2014). In the same 500 mg/dL of hemoglobin, 1000 mg/dL of triglycerides, 20 mg/dL of study, oil exposure was also shown to prolong the action potential in conjugated or unconjugated bilirubin and 10 ng/ml of biotin will tuna. Many of these piscine studies suggested that humans may also produce less than a 10% change in results. The reportable range is have cardiac change secondary to crude oil exposure; however, this has 0.006–50 ng/ml according to Siemens package insert. We have mod- not been documented in human studies (Goldstein et al., 2011; Soloman ified the reportable range due to blank assessment to 0.016 ng/ml to and Janssen, 2010; D ’andrea and Reddy, 2014). 50 ng/ml. Currently PCV and serum chemistry, including electrolytes and To the authors’ knowledge, cTnI has not yet been validated as a glucose, are recommended as part of an assessment protocol for oil cardiac biomarker in cormorants or any other avian species. A cTnI exposed birds (Mazet et al., 2002). Due to the cardiac damage pilot study was performed in eight Black Spanish Poults (Meleagris secondary to MC252 documented by Harr et al. and this study, we gallopavo) using an iSTAT cTnI (Abbott Point of Care Inc., Princeton, propose that troponin I should be further investigated as a screening NJ), but the authors did not find a significant difference in poults tool in birds exposed to oil. Similar to troponin's use in humans and treated with doxorubicin, a known myocyte toxin, and the control canines, troponin does not replace the diagnostic value of echocardio- group (McCleery et al., 2015). This failure of differentiation may be due gram and ECG in defining cardiac disease diagnosis. Rather cTnI is a to lack of crossreactivity of the different antibody that is used in the noninvasive measurement of cardiomyocyte damage and can be used as iSTAT analyzer. Numerous cardiac troponin I assays have been devel- a sensitive screening biomarker.

132 M. Daigneault et al. Ecotoxicology and Environmental Safety 146 (2017) 129–133

Acknowledgements Healy, K., Horak, K., Link, J.E., Reavill, D., Bursian, S.J., Cunningham, F.L., 2017. Dermal exposure to weathered MC252 crude oil results in echocardigraphically identifiable systolic myocardial dysfunction in double crested cormorants The studies appearing in this special issue were funded by the U.S. (Phalacrocorax auritus). Ecotoxicol. Environ. Saf. Accept. Fish and Wildlife Service as part of the Deepwater Horizon Natural Hastings, K.E., Koppe, R.I., Marmor, E., Bader, D., Shimada, Y., Toyota, N., 1991. Structure and developmental expression of troponin i isoforms. J. Biol. Chem. 266 Resource Damage Assessment. Special thanks to Katie Hanson-Dorr, Dr. (29), 19659–19665. Dr. Brian Dorr and the staff at the National Wildlife Research Center, Knafo, S.E., Rapoport, G., Williams, J., Brainard, B., Driskell, E., Uhl, E., Crochik, S., Mississippi for their technical assistance with animal care and sampling Divers, S.J., 2011. Cardiomyopathy and right-sided congestive heart failure in a red- – on this project. Also, thanks to Dr. Francis Moore and Shelley van tailed hawk (Buteo jamaicensis). J. Avian Med. Surg. 25 (1), 32 39. Langhorn, R., Willesen, J.L., Tarnow, I., Kjelgaard-Hansen, M., 2013. Evaluation of a Proosdy for all of their communication and excellent work on sample high-sensitivity assay for measurement of canine and feline serum cardiac troponin I. analysis. Vet. Clin. Pathol. 42 (4), 490–498. Mazet, J.A., Newman, S.H., Gilardi, K.V., Tseng, F.S., Holcomb, J.B., Jessup, D.A., References Ziccardi, M.H., 2002. Advances in oiled bird emergency medicine and management. J. Avian Med. Surg. 16 (2), 146–149. McCleery, B., Johns, S., Jesty, S., Jones, M., 2015. Cardiac troponin I as a biomarker of Antman, E.M., 2002. Decision making with cardiac troponin tests. N. Engl. J. Med. 346, myocyte injury in turkeys. Session #345. Paper presented at Association of Avian 2079–2082. Veterinarians Conference. San Antonio, Texas. Apple, F.S., Saenger, A.K., 2013. The state of cardiac troponin assays: looking bright and Newby, L.K., Jesse, R.L., Babb, J.D., Christenson, R.H., De Fer, T.M., Diamond, G.A., moving in the right direction. Clin. Chem. 59 (7), 1014–1016. http://dx.doi.org/10. Fesmire, F.M., Geraci, S.A., Gersh, B.J., Larsen, G.C., Kaul, S., 2012. ACCF 2012 1373/clinchem.2013.203307. expert consensus document on practical clinical considerations in the interpretation – Brette, F., Machado, B., Cros, C., Incardona, J.P., Scholz, N.L., Block, B.A., 2014. Crude oil of troponin elevations. J. Am. Coll. Cardiol. 60, 2427 2463. impairs cardiac excitation-contraction coupling in fish. Science 343, 772–776. Newman, S.H., Anderson, D.W., Ziccardi, M.H., Trupkiewicz, J.G., Tseng, F.S., Claireaux, G., Davoodi, F., 2010. Effect of exposure to petroleum hydrocarbons upon Christopher, M.M., Zinkl, J.G., 2000. An experimental soft-release of oil-spill ff cardio-respiratory function in the common sole (Solea solea). Aquat. Toxicol. 98, rehabilitated American coots (Fulica americana): II. E ects on health and blood – 113–119. parameters. Environ. Pollut. 107 (3), 295 304. ’ Cunningham, F.L., Dean, K.M., Hanson-Dorr, K.C., Harr, K.E., Healy, K., Horak, K.E., Link, O Brien, P.J., Smith, D.E., Knechtel, T.J., Marchak, M.A., Pruimboom-Brees, I., Brees, J.E., Shriner, S., Rupp, T.L., Bursian, S.J., Dorr, B.S., 2017. Development of methods D.J., Spratt, D.P., Archer, F.J., Butler, P., Potter, A.N., Provost, J.P., Snyder, P.A., fi for avian oil toxicity studies using the double crested cormorant (Phalacrocorax Reagan, W.J., 2006. Cardiac troponin I is a sensitive, speci c biomarker of cardiac – auritus). Ecotoxicol. Environ. Saf this issue (submitted for publication). injury in laboratory animals. Lab. Anim. 40, 153 171. D ’andrea, M.A., Reddy, G.K., 2014. Hematological and hepatic alterations in nonsmoking Reagan, W.J., 2010. Troponin as a biomarker of cardiac toxicity: Past, present, and future. – residents exposed to benzene following a flaring incident at the British petroleum Toxicol. Pathol. 38, 1134 1137. – plant in Texas City. Environ. Health 13, 115–132. http://dx.doi.org/10.1186/1476- Sarko, J., Pollack, C.V., 2002. Cardiac troponins. J. Emerg. Med. 23, 57 65. ’ 069X-13-115. Serra, M., Papkonstantinou, S., Adamcova, M., O Brien, P.J., 2010. Veterinary and Dean, K.M., Bursian, S.J., Cacela, D., Carney, M.W., Cunningham, F.L., Dorr, B.S., toxicological applications for the detection of cardiac injury using cardiac troponin. – Hanson-Dorr, K.C., Healy, K.A., Horak, K.E., Link, J.E., Lipton, I., McFadden, A.K., Vet. J. 185, 50 57. ff McKernan, M.A., Harr, K.E., 2017. Changes in white cell estimates and plasma Soloman, G.M., Janssen, S., 2010. Health e ects of the Gulf oil spill. J. Am. Med. Assoc. – chemistry measurements following oral or external dosing of double-crested 304 (10), 1118 1119. ff cormorants, Phalacrocorax auritus, with artificially weathered MC252 Oil. Ecotoxicol. Thygesen, K., Alpert, J.S., Ja e, A.S., Simoons, M.L., Chaitman, B.R., White, H.D., 2012. fi – Environ. Saf. (Accepted, this issue). Third universal de nition of myocardial infarction. Circulation 60, 1581 1598. fl Fonfara, S., Loureiro, J., Swift, S., James, R., Cripps, P., Dukes-McEwan, J., 2010. Cardiac Tissier, F., Dussauze, M., Le och, N., Theron, M., Lemaire, P., Floch, S.L., Pichavant- fi ff troponin I as a marker for severity and prognosis of cardiac disease in dogs. Vet. J. Ra ni, K., 2015. E ect of dispersed crude oil on cardiac function in seabass – 184, 334–339. (Dicentrarchus labrax). Chemosphere 134, 192 198. Goldstein, B.D., Osofsky, H.J., Lichtveld, M.Y., 2011. The Gulf Oil Spill. N. Engl. J. Med. Winter, R.L., Saunders, A.B., Gordon, S.G., Miller, M.W., Sykes, K.T., Suchodolski, J.S., 364 (14), 1334–1348. http://dx.doi.org/10.1056/NEJMra1007197. Steiner, J.M., 2014. Analytical validation and clinical evaluation of a commercially Harr, K.E., Flatland, B., Nabity, M., Freeman, K.P., 2013. ASVCP guidelines: allowable available high-sensitivity immunoassay for the measurement of troponin I in humans – total error guidelines for biochemistry. Vet. Clin. Pathol. 42 (4), 424–436. for use in dogs. J. Vet. Cardiol. 16, 81 89. Harr, K.E., Rishniw, M., Rupp, T.L., Cacela, D., Dean, K.M., Dorr, B.S., Hanson-Dorr, K.C.,

133 Ecotoxicology and Environmental Safety 146 (2017) 134

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Acknowledgements for this issue

The studies appearing in this special issue were funded by the U.S. as well as all of the responders on the scene whose hard work in doc- Fish and Wildlife Service as part of the Deepwater Horizon Natural umenting the impact of the spill on bird life contributed to improving Resource Damage Assessment. All raw data collected, as well as draft our understanding of oil toxicity to birds. reports submitted as part of the Final Programmatic Damage Specifically, we would like to thank Josh Lipton (Abt Associates), Assessment and Restoration Plan (PDARP) and Final Programmatic Jeff Gleason (Bureau of Ocean Energy Managenment, Regulation and Environmental Impact Statement (PEIS) are available using the fol- Enforcement), John Isanhart and Kevin Reynolds (US Department of lowing links: Interior), Michael Fry, Jon Hemming, Roger Helm, Chuck Hunter, http://www.gulfspillrestoration.noaa.gov/restoration-planning/ Carolyn Marn, Moira McKernan, Anne Secord, Kim Trust, Veronica gulf-plan/ Varela and Sara Ward (US Fish and Wildlife Service), Nelson Beyer, Dan https://dwhdiver.orr.noaa.gov/explore-the-data Esler, John French and Mike Hooper (US Geological Survey), and The manuscripts prepared for this issue would not have been pos- Douglas Beltman, Jess Fallon (Virginia Tech University) and Kirk sible without the assistance of a broad group of researchers, veter- Klasing (University of California Davis) for all of their valuable input in inarians, biologists and rehabilitation specialists whose input guided planning the scope of the DWH avian toxicity testing program. and informed the direction of the research. The enormity of the We would also like to thank Carolyn Cray and her entire staff at the Deepwater Horizon oil spill and its impact on bird life represented an University of Miami Comparative Pathology Lab for their invaluable unprecedented challenge for understanding the broad effects of oil technical assistance on a wide range of clinical measurements, Mike toxicity on birds. Several panels of experts were convened to discuss the Scott and Alicia Withrow at Michigan State University for electron direction that the DWH avian toxicity test program should take and it microscopy, Gina Saizan (Louisiana Oil Spill Coordinator’sOffice) and was these discussions that helped to shape the excellent research and John Wiebe (Louisiana Department of Wildlife and Fisheries) for cri- novel findings discussed in this issue. We would like to thank all of the tique and discussion, and finally to Jamie Holmes (Abt Associates) and following people who were involved in those early planning discussions Caitlyn Quinn for keeping all of the projects on track.

http://dx.doi.org/10.1016/j.ecoenv.2017.08.069

Available online 13 September 2017 0147-6513/