KDFWR Annual Research Highlights 2008

Home , Blacktail redhorse, Black redhorse, Catostomidae, Felis (constellation), Highland Rim, Modoc sucker

Kentucky Department of Fish and Wildlife Resources

Annual Research Highlights 2008

Volume II, Sept. 2009 Kentucky Department of Fish and Wildlife Resources Annual Research Highlights 2008 Volume II, Sept. 2009

Our Mission: To conserve and enhance fish and wildlife resources and provide opportunity for hunting, fishing, trapping, boating and other wildlife related activities.

Annual Research Highlights 2008 Foreword

The five primary goals of the Strategic plan are: 1) To conserve and enhance fish and wildlife populations and their habitats; 2) To increase opportunity for, and safe participation in hunting, fishing, trapping, boating, and other wildlife-related activities; 3) To foster a more informed and involved public; 4) To expand and diversify our user base and 5) To create a more diverse, effective, and efficient organization.

These two documents are available to the public, and are intended for frequent revision and re-adjustment to incorporate ever changing agency and Wood duck banding / Brian Clark public needs and interests. The 2008 Kentucky Department of Fish and Long Term Conservation measures for these species and habitats. Wildlife Resources Research Summary Planning in Kentucky In contrast, the 2007-2012 Strategic represents our targeted efforts to fulfill The Kentucky Department of Fish Plan addresses both fish and wildlife the goals of our Comprehensive Wild- and Wildlife Resources (KDFWR) un- management issues and agency issues life Conservation Strategy as well as derstands the importance of long-term as a whole. goals of the 2008-2012 Strategic Plan. planning to protect and manage the natural resources of the Commonwealth and to effectively serve hunters and Figure 1. Federal programs that fund many KDFWR projects. fishermen in Kentucky. Over the past several years, KDFWR has collabo- Federal Funding Source Program Goal rated with multiple outside agencies, non-profit organizations, professionals, To develop and implement and biologists to complete two impor- programs that benefit wildlife and State Wildlife Grant Program tant planning documents: the Compre- their habitats, specifically species (SWG) hensive Wildlife Conservation Strategy and habitats of conservation (completed in 2007; (http://fw.ky.gov/ concern kfwis/stwg/) and the 2008-2012 Ken- To restore, conserve, manage and tucky Department of Fish and Wildlife Wildlife Restoration Act enhance wild birds and mammals Resources Strategic Plan (http://fw.ky. (Pittman-Robertson) gov/pdf/strategicplan2008-2012.pdf). and their habitats Both of these documents are designed To fund fishery management Sport Fish Restoration to guide agency decisions; however, projects, boating access, and aquatic Program (Dingell-Johnson) they serve two unique purposes. The education Comprehensive Wildlife Conservation To protect and restore habitats on Strategy (CWCS) is Kentucky’s road- Landowner Incentive Program private lands to benefit species of map for sustaining fish and wildlife di- (LIP) versity. The two primary goals of the conservation concern CWCS are to identify and prioritize im- Cooperative Endangered To fund conservation projects portant species and habitats of conser- Species Conservation Fund for candidate, proposed, or listed vation concern within Kentucky and to (Section 6) species. successfully implement conservation

Kentucky Department of Fish & Wildlife Resources FOREWORD

These project summaries serve as a tes- tament to KDFWR’s vigilance in the conservation of the fish and wildlife resources that we hold in trust for the public.

Funding Sources and Guidance to Federal Programs The Kentucky Department of Fish and Wildlife Resources receives no general fund taxpayer dollars. As a result, the Department relies on hunting and fishing license fees, boat registra- tion fees and federal programs. Proj- ects that are entirely funded by the state are labeled “Kentucky Department of Fish and Wildlife Resources;” however, most of the projects included in this document are partially or fully funded by federal programs such as the State and Tribal Wildlife Grant Program, the Wildlife Restoration Act (Pittman- Robertson), the Sport Fish Restoration Program (Dingell-Johnson), the Land- owner Incentive Program (LIP), and the Cooperative Endangered Species Con- servation Fund (Section 6). These federal programs serve a variety of purposes; however, each has an (UNTINGAND&ISHING,ICENSES underlying goal of fish, wildlife, and/or habitat conservation. Brief descriptions &EDERAL2EIMBURSEMENT of each of these programs are shown in "OAT2EGISTRATION&EES Figure 1. These federal programs provided /THEREG)NCOME4AX#HECK OFFAND)NTEREST)NCOME approximately 16.6 million dollars to KDFWR in 2008 (Figure 2). For reference, we have included the state and fed- Figure 2. Kentucky Department of Fish and Wildlife Resources Funding Sources eral funding sources for each project; 2008 however, these projects may be additionally supplemented by outside funding project updates. Citations for all pub- the table of contents, an expected date provided by non-profit organizations or lished research with Kentucky Depart- of completion is listed for each project universities. When possible, we listed ment of Fish and Wildlife involvement with a finite end-date. This will facili- these sources in addition to the state and are included in the Table of Contents. tate looking up detailed summaries of federal funding sources. For each project For projects that have been completed completed projects in later years. summary, we also identify the specific and not yet published, a detailed sum- goals of the strategic plan or CWCS ful- mary will be included in the first por- Please use the following filled, as well as the KDFWR contact re- tion of the document. For projects that citation when referencing this sponsible for each project. have just begun, a brief 1-page over- document: view of the project is included in the Kentucky Department of Fish and How to Use This Document second portion (“project highlights”) Wildlife Resources Annual Research This document is divided into four of the document. For select ongoing Highlights, 2008. Volume II. Publica- main sections: published research, com- projects, brief updates are included in tion of the Wildlife and Fisheries Divi- pleted projects, project highlights, and the last section of this document. In sions. September, 2009, 79 pp.

Annual Research Highlights 2008 Table of Contents

Effects of Orientation and Weatherproofing on the Published Research Detection of Echolocation Calls in the Eastern United Contact SWG Coordinator Danna Baxley States...... 34 ([email protected]) for reprints of these publications. Identifying and Protecting Hibernation Roosts for Hopkins, R.L., M.D. Burns, B. Burr, and L.J. Hopman. Endangered Bats in Kentucky...... 37 2008. Building a centralized database for Kentucky fishes: Progress and future applications. Journal of the Kentucky Academy of Science 69 (2): 164-169. Project Highlights Larkin, J.L., D.S. Maehr, J.J. Krupa, J.J. Cox, K. These projects began in 2008 Alexy, D.E. Unger, and C. Barton. 2008. Small mammal response to vegetation and spoil conditions Fishes on a reclaimed surface mine in eastern Kentucky. Southeastern Naturalist 7(3):401-112. Analysis of the Environmental Requirements for Etheostoma cinereum and Percina squamata in the Lynch, W.L., and C.N. Moreira. 2008. Nest arrival Rockcastle River vocalizations of the Turkey Vulture Cathartes aura Estimated Completion Date: December 31, 2010.... 41 (Cathartidae: Falconiformes). Vulture News 59:3-6. Captive Propagation and Reintroduction of the Vukovich, M. and G. Ritchison. 2008. Foraging behavior Cumberland Darter and Kentucky Arrow Darter in of Short-Eared Owls and Northern Harriers on a Southeastern Kentucky reclaimed surface mine in Kentucky. Southeastern Estimated Completion Date: December 31, 2012.... 42 Naturalist 1(1):1-10. A Survey of Fishes of Rock Creek, Kentucky, with Emphasis on the Impact of Stocking Rainbow Trout on Native Fishes Completed Projects Estimated Completion Date: December 31, 2009.... 43

Big Game and Small Game Status, Life History, and Phylogenetics of the Amblyopsid Cavefishes in Kentucky Programs Estimated Completion Date: December 31, 2010.... 44 Hunters’ Use of the Kentucky Department of Fish and Wildlife Resources’ Telecheck System...... 7 Status Survey of the Northern Madtom, Noturus stigmosus, in the Lower Ohio River Using FLIR (forward-looking infrared radiography) Estimated Completion Date: June 30, 2010...... 45 to Estimate Elk Density and Distribution in Eastern Kentucky...... 9 Mollusks

Maternal Antibody Transfer and Meningeal Worm Advances in the Propagation of Rare and Endangered Infection in Kentucky Elk...... 13 Mussel Species...... 46

Estimating Black Bear Populations in Kentucky...... 17 Successful Reintroduction of Two Endangered and Two Candidate Mussel Species to the Big South Fork Assessment of Habitat Value for Recovering Disturbed Cumberland River, Kentucky...... 47 Warm-Season Grass Using Multi-Cover Habitat Assessment Model for the Northern Bobwhite...... 25 Successful Augmentation of the Fatmucket, Lampsilis siliquoidea, in the Elkhorn Creek, Kentucky...... 48

Wildlife Diversity and Fisheries Freshwater Mollusk Monitoring in the South Fork Life History and Population Characteristics of Kentucky River System Moxostoma poecilurum, the Blacktail Redhorse, in Estimated Completion Date: June 30, 2010...... 49 Terrapin Creek, Graves County, Kentucky...... 27

Kentucky Department of Fish & Wildlife Resources TABLE of CONTENTS

Restoration of Bur Oak on the Clay Wildlife Reptiles and Amphibians Management Area by Means of Direct Seeding...... 65 Life History and Population Assessment of the Western Cottonmouth in Western Kentucky...... 50 Natural Grassland Survey of the Original Barrens-Prairie Region of Kentucky Status Assessment and Conservation of the Eastern Estimated Completion Date: June 30, 2011...... 67 Hellbender Estimated Completion Date: July 1, 2012...... 51

Inventory, Monitoring, and Management of Amphibians Project Updates and Reptiles in Kentucky...... 52 This section includes brief updates for selected projects that began prior to 2008.

Birds Conservation Status and Habitat of the Longhead Darter Reproductive Success of Interior Least Terns in Western ( Percina macrocephala) in Kinniconick Creek, Kentucky Kentucky...... 69 Estimated Completion Date: June 30, 2010...... 53 Assessing Avian Use of Land Enrolled in Conservation The Common Raven in Cliff Habitat: Detectability and Practice 33 (CP33), Conservation Reserve Program Occupancy ...... 70 Estimated Completion Date: June 30, 2011...... 54 Migratory Shorebird Updates An Evaluation Tool for Avian Monitoring Programs... 55 Avian Influenza Monitoring throughout Kentucky...... 71 Estimating Abundance of Species of Concern in the Central Hardwoods Region...... 56 Migratory Shorebird, Colonial Water Bird and Woodcock Investigations Golden-winged Warbler Monitoring...... 58 Estimated Completion Date: June 30, 2010...... 72

Grassland Songbird Survey...... 59 Monitoring and Management of Kentucky’s Waterfowl ...... 73

Mammals Monitoring Canada Goose Populations in Kentucky Black Bear Resource Selection, Demographics and ...... 74 Movement Patterns in Kentucky...... 60 Mourning Dove Banding in Kentucky...... 75 Genetic Diversity, Structuring and Recolonization Patterns of Black Bears in Eastern Kentucky...... 61 Proactive Wood Duck Management in Kentucky...... 76

Genetic Characteristics of Restored Elk Populations in Habitat Restoration Updates Kentucky Estimated Completion Date: June 30, 2010...... 62 Impacts of Herbicide Application Following a Late Summer Burn...... 77 Geographic Distribution and Prevalence Cytauxzoon felis in Wild Felids...... 63 Establishing Shrubs and/or Brambles on Reclaimed Minelands...... 78 Habitat Restoration Bottomland Hardwood and Riparian Restoration in Obion Creek/Bayou de Chien Watersheds Estimated Completion Date: June 30, 2010...... 64

Annual Research Highlights 2008 White-tailed deer / James Inman Completed Projects

Kentucky Department of Fish & Wildlife Resources Big Game and Small Game Programs / COMPLETED PROJECTS

Hunters’ Use of the Kentucky Department of Fish and Wildlife Resources’ Telecheck System

Mark Damian Duda, Responsive Management; Tina Brunjes, Kentucky Department of Fish and Wildlife Resources

Introduction The Kentucky Department of Fish and Wildlife Resources first implemented a toll-free, automated phone-in Telecheck system in 1998 for collecting hunter harvest information. All deer, elk and turkey harvested must be reported through the Telecheck System by midnight on the day of harvest, and entering false information is unlaw- ful. Reporting through the Telecheck system takes about five minutes and has increased the speed of collecting hunter harvest information. Hunters who report harvests through the Ken- tucky Department of Fish and Wildlife Resources’ Telecheck Service can also review their reports online for accuracy including the hunter’s name, species and gender harvested, type of weapon used, county code indicating harvest location, and whether there were visible antlers or a beard. “Telecheck Review” was implemented on the Department’s website, fw.ky.gov, Monday, November 5, 2007, just before the 2007 statewide modern firearms deer season opener on November 10. The purpose of this study is to better understand hunter compliance with the Kentucky Department of Fish and Wildlife Resources’ Telecheck Report- ing System. To assess total harvest Wild turkeys / Joe Lacefield numbers as well as Kentucky licensed hunters’ and license-exempt landown- exempt landowners across the state, for standards established by the Council of ers’ compliance levels with the Tele- a total of 1,860 completed interviews. American Survey Research Organiza- check Reporting System, Responsive tions. Methods of instruction included Management conducted a statewide Methodology lecture and role-playing. The Survey telephone survey of licensed hunters and To ensure the integrity of the Center Managers and other professional license-exempt landowners. Responsive telephone survey data, Responsive staff conducted project briefings with Management’s goal was to interview Management has interviewers who the interviewers prior to the administra- 1,500 licensed hunters and 360 license- have been trained according to the tion of this survey. Interviewers were

Annual Research Highlights 2008 COMPLETED PROJECTS / Big Game and Small Game Programs

instructed on type of study, study goals and objectives, handling of survey questions, interview length, termination points and qualifiers for participation, interviewer instructions within the survey instrument, reading of the survey instrument, skip patterns, and probing and clarifying techniques necessary for specific questions on the survey instrument. The Survey Center Manag- ers and statisticians monitored the data collection, including monitoring of the actual telephone interviews without the interviewers’ knowledge, to evaluate the performance of each interviewer Wild turkeys / Joe Lacefield and ensure the integrity of the data. Results After the surveys were obtained by the Responsive Management inter- confidence interval of the telephone interviewers, the Survey Center Manag- viewed 1,503 Kentucky licensed hunters survey estimate, between 951 and 2,886 ers and/or statisticians checked each and 374 landowners owning 10 or more bobcats harvested. completed survey to ensure clarity and acres for a total of 1,877 interviews. completeness. The reported Telecheck deer harvest Discussion/Management Interviews were conducted Mon- numbers and Responsive Management’s Implications day through Friday from 9:00 a.m. to telephone survey numbers are in close Responsive Management’s num- 9:00 p.m., Saturday noon to 5:00 p.m., agreement. The 2007 Telecheck system bers indicate that hunters who harvest and Sunday from 5:00 p.m. to 9:00 indicates that 113,436 deer were har- deer and bobcat adequately use the p.m., local time. A five-callback design vested, and this is within the confidence Telecheck system to report harvested was used to maintain the representa- interval of Responsive Management’s animals; however, hunters who harvest tiveness of the sample, to avoid bias to- estimate of 106,701 to 128,710 deer. The turkey do not appear to be reporting all ward people easy to reach by telephone, reported Telecheck turkey harvest num- of their turkey harvest, particularly those and to provide an equal opportunity for bers are not in agreement with Respon- hunting on public land. The Telecheck all to participate. When a respondent sive Management’s telephone survey turkey harvest total of 28,898 is lower could not be reached on the first call, numbers. Responsive Management’s than and does not fall within Respon- subsequent calls were placed on differ- turkey harvest estimates are higher than sive Management’s 95% confidence ent days of the week and at different the numbers reported under the Tele- interval of 38,419 to 51,470 turkeys, as times of the day. The licensed hunter check system. Because the sample size estimated from telephone interviews data were weighted so that the percent- of hunters who hunted otter was small, with hunters. The results of this research ages of sample among the license types confidence intervals were not created, indicate that the Telecheck system is an exactly matched the distribution of the and it was not possible to make an ac- effective means to track deer and bobcat license types statewide as shown in the curate harvest estimate. The Telecheck harvest and is properly used by most table below. figure of 2,385 bobcats harvested by hunters. Discrepancies between the Tele- hunting or trapping falls with the 95% check numbers and numbers estimated from interviews with turkey hunters are problematic. This finding will result in Number of Licenses in Percent of License Type increased diligence and monitoring of Dept. Provided Database Sample turkey harvest in the upcoming years by the Kentucky Department of Fish and 2007 Combo Residential 64,511 26% Wildlife Resources. 2007 Hunting 79,552 32% Funding Sources: Kentucky Department 2007 Sportsman 31,985 13% of Fish and Wildlife Resources 2007 Senior/Disabled 69,306 28% KDFWR Strategic Plan. Goal 1. TOTAL 245,354 100% Strategic Objective 5.

Kentucky Department of Fish & Wildlife Resources Big Game and Small Game Programs / COMPLETED PROJECTS

Using FLIR (forward-looking infrared radiography) to Estimate Elk Density and Distribution in Eastern Kentucky

Lauren Dahl , John J. Cox, David S. Maehr, and Will Bowling, University of Kentucky; Karen Alexy and Tina Brunjes, Kentucky Department of Fish and Wildlife Resources; Joe Duchamp and Jeffery Larkin, Indiana University of Pennsylvania; David Unger Alderson Broaddus College

Introduction Wildlife management decisions are often driven by a species’ population size (Lancia et al. 1996, Meffee and Carroll 1997). Kentucky’s free ranging elk population was reintroduced from 1997 – 2001 with a total of 1,547 elk (Cervus elaphus nelsoni) released at 8 sites within eastern Kentucky. Initial studies demonstrated that Kentucky elk had a low mortality rate (0.07) and a high natality rate (0.86) during the initial 3 years post-reintroduction (Larkin et al. 2003) suggesting a high potential for a rapidly growing population. The successful reintroduction and subsequent population growth of elk (Cervus elephus) in Kentucky during the past decade has created unique population monitoring challenges for wildlife managers. The Kentucky elk model (KEM) predicted a population size of 5,700 elk in 2006, but these were solely based on mortality, natality and harvest rates and did not consider the available habitat Kentucky elk / Lauren Dahl or landscape patterns that facilitate the distribution and local abundance of elk. was needed to verify the KEM and set raphy, double sampling, mark-resight, Regardless, the predicted population harvest levels to maintain the popula- line-transects, and aerial visual surveys size suggested that the population had tion within the desired goal. (Cochran 1977, Lancia et al. 1996). nearly quadrupled in less than 10 years. Techniques typically used to visu- Detection problems arising from the As the population began to reach the ally survey large mammal populations rugged forested landscape of eastern set goal of 10,000 elk, a comprehensive to gain abundance and distribution data Kentucky limit the efficacy of these and accurate population assessment include: drive counts, aerial photog- sampling techniques. Restricted ground

Annual Research Highlights 2008 9 COMPLETED PROJECTS / Big Game and Small Game Programs

accessibility, elk avoidance of humans, in the cabin of the plane displayed the et al. 2006). By using a grid system we and a predominately forested landscape images received from the sensor in real created 2 x 2 km (4 km2) blocks within make traditional ungulate population time allowing the sensor operator to the FLIR survey transects which en- estimation techniques impractical in manipulate the sensor and zoom in on abled us to identify landscape charac- Kentucky’s elk restoration zone. An al- suspected elk locations. FLIR images teristics within an area that an elk may ternative technique that addresses many were digitally recorded, georeferenced, use within a day (average Kentucky elk of the weaknesses of conventional and analyzed following the completion daily movement in December is 1,200 survey methods is aerial-based forward- of the entire survey. The sensor opera- m; W. Bowling unpublished data). We looking infrared radiography (FLIR), a tor used species-specific morphological also created a 6 x 6 km (36 km2) block method that utilizes an infrared scope features (e.g. body size and shape) to surrounding each 4 km2 block to iden- attached to the underside of an aircraft distinguish elk from other large mam- tify landscape characteristics within to detect thermal radiation of target mals (white-tailed deer, cows, horses) an area representative of an average species. FLIR has been used success- in the study area. For each FLIR-iden- Kentucky elk home range (31 km2; W. fully to survey elk (Dunn et al. 2002), tified individual or group of elk, we Bowling unpublished data). moose (Alces alces, Adams et al. 1997) recorded group size and the majority We calculated several landscape and several other species in other areas activity (bedded, standing, or walking), metrics at both of the spatial scales of North America. Our research tested and landcover type. based on previous literature and the feasibility of FLIR (forward-look- We conducted the elk detectabil- biological importance to Kentucky’s ing infrared radiography) to survey elk ity portion of the study in December elk (Cook 2002, Geist 2002, Skovlin in rugged southeastern Kentucky. 2006 and 2007. FLIR detection rate of et al. 2002, Schneider 2006, Sawyer elk was calculated by comparing elk 2007, Olsson et al. 2007, Telesco et Study Area locations identified by FLIR and those al. 2007). Landscape metrics included We determined FLIR detection rate observed by ground crews visually or the area of forest land cover (ForCA), of elk on a cluster of surface mines near using telemetry <_ 30 minutes of FLIR the area of herbaceous land cover Hazard, KY. This area was inhabited locations. We also located white-tailed (HerbCA), herbaceous edge density by 70 vhf radio-collared elk and thus deer in the same area using spotlights (HerbED), herbaceous mean patch size allowed us to take advantage of marked and recorded their locations to calculate (HerbMPS), forest mean patch size animals to compare with FLIR identi- a FLIR misclassification rate between (ForMPS), herbaceous mean nearest fied elk. FLIR transect surveys were elk and deer within the same temporal neighbor (HerbMNN), herbaceous area subsequently conducted along transects window. weighted mean patch fractal dimen- near or radiating from the 8 original In December 2006, we surveyed sions (Herb AWMPFD), and urban core elk release sites within the original 14- elk within a 2 x 16 km transect block area index (UrbCAI). Additionally, we county elk restoration zone. Total area that either radiated in the four cardinal calculated road density at both spatial surveyed using transects was 41,000 directions surrounding each of the 8 elk scales (4kmRdDen and 36kmRdDen), ha, or approximately 4% of the original release sites, or were in specific areas and site influence (SiteInflu) a variable 14-county elk restoration zone. where information about elk was de- that combined the distance from each sired. We used this systematic sampling site and the number of elk released at Methods design to obtain minimum counts of elk each of those sites. We conducted a We conducted the FLIR survey in at each release site while also modeling principle component analysis (PCA) to December 2006 during leaf off and at and extrapolating a population esti- transform correlated variables which nocturnal hours when infrared visibil- mate for the majority of the Kentucky were determined using a Pearson’s ity and temperature contrast between elk restoration zone (Caughley 1974). product moment correlation coefficient animals and background objects were Landscape characteristics at FLIR- (Dahl 2008). highest. The FLIR unit consisted of identified elk locations were gathered To estimate the uncertainty that a PolyTech, Kelvin 350 II infrared from Kentucky land cover type data. the undetected elk groups contribute to radiometer that housed a Thermovi- Because the distribution of animals the population modeling process, we sion 1000 sensor (PolyTech Airborne is a function of several biological adjusted the observed elk groups by the Remote Sensing, Stockholm, Ger- interactions that cannot be described detection rate and repeated this process many). The FLIR unit was attached to in any one specific spatial scale we 100 times to equalize the effects of the the underside of the wing of a Cessna used two scales to assess the relation- distribution of these additional missed 206 fixed wing aircraft (Cessna Aircraft ship between the density of elk and the groups (Efron 1979, Buckland and Company, Wichita, Kansas) by a 4-axis landscape characteristics (Levin 1992, Elston 1993). We used multiple linear gyro-stabilized system. A monitor with- Buckland and Elston 1993, Stubblefield regression models to identify the rela-

10 Kentucky Department of Fish & Wildlife Resources Big Game and Small Game Programs / COMPLETED PROJECTS

tionship between the density of elk in each 4km2 block and the landscape variables. By using Akaike’s Information Criterion corrected for small sample size (AICc: Burnham and Ander- son 2002) we selected the top models from each of the 100 iterations. We then used the relative model weights to select the stron- gest models for the final model averaging process (Burnham and Anderson 2002). The final derived elk population model was vali- dated on a subset of FLIR survey data with- held prior to the model building process. The final derived model was then extrapolated Figure 1: Elk density distribution estimated within 4 km2 cells in the core area of the restoration to each 4 km2 block zone in southeastern Kentucky, USA, December 2006. within the core area of the restoration zone (7,076 km2). Elk Density = to overcome many of the weaknesses 1.61+0.07x1-0.63x2-0.18x3+0.06x4+0.33x5 of other commonly used visual survey Results Where: techniques for elk. For example, FLIR The detection rate of FLIR for x1= Site Influence (the number of allowed us to observe elk in habitats elk in Kentucky was derived from 44 elk released at each site and the where their cryptic coloration and comparisons between ground located distance to that site) behavior would normally cause them to collared elk and FLIR observed elk x2= 4 km2 Road Density be undetected by visual methods. Foot- locations. The derived detection rate x3= 36 km2 Road Density age from FLIR surveys was recorded was 76% for groups of < 10 individuals x4= 36 km2 Urban Core Area Index and could be reviewed several times and 100% for groups of >_ 10 individu- (the amount of central urban area) to verify counts, whereas resolution als. FLIR misidentified 2 of 31 (6.5%) x5= PCA1 (The ratio of herbaceous with many video recording devices deer observed by the ground crew as area to forest area at both the 4 is often poor or sensitive to vibration elk during the detection rate portion of km2 and 36 km2 spatial scale) from moving aircraft. We found similar the study. detection rates as other FLIR surveys of A total of 1,981 elk were detected We extrapolated the population ungulates and had success distinguish- using FLIR within the survey tran- model to the core area of the Kentucky ing elk from deer within the survey sects. The final population model was elk restoration zone and calculated a area. ultimately derived from the averaging population estimate of 7,001 (SE=772, Areas of high elk density were of the overall top 4 competing models. CI=5,488-8,514) individuals for 2006 associated with areas of extensive her- The averaged model suggested that 5 (Figure 1). baceous cover, near release sites where variables were strong predictors of elk large numbers of elk were initially density in Kentucky demonstrated in Discussion translocated, and in areas with low road the following equation: The ability of FLIR to detect the density. The population estimate for the heat signature of animals allowed us core area of the Kentucky elk restora-

Annual Research Highlights 2008 COMPLETED PROJECTS / Big Game and Small Game Programs

tion zone was slightly above the 5,700 Burnham, K. P., and D. R. Anderson. D. S. Maehr, P. Widen and M. W. elk estimated by the current KEM. 2002. Model selection and inference: Wichrowski. 2007. Movement and The total cost of our study was a practical information theoretic activity patterns of translocated elk $105,632 and included both the detect- approach. 2nd edition. Springer-Ver- (Cervus elaphus nelson) on an active ability testing and actual elk survey. lag, New York, NY. coal mine in Kentucky. Wildlife Biol- While this cost is higher than several Caughley, G. 1974. Bias in aerial ogy in Practice 3:1-8. other visual survey techniques, FLIR survey. Journal of Wildlife Manage- Sawyer, H., R. M. Nielson, F. G. Lin- was able to provide a relatively fast ment. 38:921-933. dzey, L. Keith, J. H. Powell and A. A. survey of a large geographic area. Cochran, W. G. 1977. Sampling tech- Abraham. 2007. Habitat selection of The potential downfalls of this survey niques. John Wiley and Sons, New Rocky Mountain elk in a nonforested technique include the time delay for York, New York. environment. Journal of Wildlife analysis of the footage, the relatively Cook, J. G. 2002. Nutrition and Food. Management 71:868-874. higher expenses as compared to other Pages 259-350 in D. E. Toweill and Schneider, J., D. S. Maehr, K. J. survey techniques, and the coordination J. W. Thomas, editors. North Ameri- Alexy, J. J. Cox, J. L. Larkin, and of the logistically complicated detec- can elk: ecology and management. B. C. Reeder. 2006. Food habits tion rate portion of the study. Smithsonian Institute Press, Wash- of reintroduced elk in Southeastern ington, D. C. Kentucky. Southeastern Naturalist Management Implications Dunn, W.C., J.P. Donelley, and W.J. 5:535-546. The elk population estimate de- Drausmann. 2002. Using thermal Skovlin, J. M., P. Zaeger, B. K. John- rived from this study has allowed state infrared sensing to count elk in the son. 2002. Elk habitat selection and wildlife managers to identify elk abun- southwestern United States. Wildlife evaluation. Pages 531-556 in D. E. dance and distribution patterns in east- Society Bulletin 30:963-967. Toweill and J. W. Thomas, editors. ern Kentucky. These data have provided Efron, B. 1979. Bootstrap methods: North American elk: ecology and managers with an alternate method to another look at the jackknife. Annals management. Smithsonian Institute evaluate the existing elk population of Statistics 7:1-26. Press, Washington, D. C. model, and will be important in deter- Geist, V. 2002. Adaptive behavioral Stubblefield, C. H., K. T. Vierling, and mining harvest allocation and intensity strategies. Pages 389-434 in D. E. M. A. Rumble. 2006. Landscape- within the elk zone. The information Toweill and J. W. Thomas, editors. scale attributes of elk centers of derived from this elk population model North American elk: ecology and activity in the central Black Hills of may also be used as a reference in the management. Smithsonian Institution South Dakota. Journal of Wildlife creation of more fine scale habitat suit- Press, Washington, D.C. Management 70:1060-1069. ability models, which could lead to the Lancia, R. A., J. D. Nichols, and K. Telesco, R. L., F. T. Van Manen, J. identification of land important for elk H. Pollock. 1996. Estimating the D. Clark, M. E. Cartwright. 2007. related-habitat protection and herd con- numbers of animals in wildlife Identifying sites for elk restoration in nectivity. This research also assessed populations. Pages 215-253 in T. Arkansas. Journal of Wildlife Man- the general use of FLIR as a survey A. Bookhout, editors. Research and agement 71:1393-1403. technique for wildlife in Kentucky. Management Techniques of Wild- The detection rate of FLIR for elk, the life and Habitats. Allen Press, Inc., Funding Sources: Pittman Robertson ability of FLIR to identify species, and Lawrence, KS. (PR) the experiences gained throughout this Larkin, J. L., D. S. Maehr, J. J. Cox, project help managers understand the D. C. Bolin, and M. W. Wichrowski. KDFWR Strategic Plan. Goal 1. limitations and benefits of FLIR. 2003. Demographic characteristics Strategic Objective 5. of reintroduced elk population in Literature Cited Kentucky. Journal of Wildlife Man- Adams, K. P., P. J. Pekins, K. A. agement 67:467-476. Gustapson, and K. Bontaites. 1997. Levin, S. A. 1992. The problem of pat- Evaluation of infrared technology for tern and scale in ecology. Ecology. aerial moose surveys in New Hamp- 73:1943-1967. shire. Alces 33:129-139. Meffee, G. K., and C. R. Carrol. 1997. Buckland, S. T., and D. A. Elston. Principles of conservation biology. 1993. Empirical models for the spa- Sinauer Associates Inc, Sanderland, tial distribution of wildlife. Journal MA. of Applied Ecology 30:378-495. Olsson, P. M. O., J. J. Cox, J. L. Larkin,

12 Kentucky Department of Fish & Wildlife Resources Big Game and Small Game Programs / COMPLETED PROJECTS

Maternal Antibody Transfer and Meningeal Worm Infection in Kentucky Elk

Willie Bowling, Lauren Dahl, as the meningeal worm’s intermediate density of intermediate hosts (Raskev- host (Anderson 1963). White-tailed itz et al. 1991) due to differing levels John J. Cox, and David S. deer are generally not pathologically of cover, moisture and vegetation Maehr, University of Kentucky; affected by meningeal worm infections regimes (Suominen 1999). Because the Karen Alexy, Tina Brunjes, Dan (Forrester and Lankester 1998), but severity of meningeal worm infection Crank, and Charlie Logsdon, P. tenuis infection in abnormal defini- in abnormal hosts is positively cor- Kentucky Department of Fish tive hosts often causes severe neuro- related with the number of P. tenuis logical trauma that can compromise an larvae ingested (Samuel et al. 1992), and Wildlife Resources individual’s fitness (Anderson 1964). cervids that use habitat supporting Habitat influences the relative higher densities of infected gastropods Introduction Elk (Cervus elaphus) were widespread throughout the area now com- prising the continental United States prior to European settlement, with an- ecdotal and archeological evidence suggesting that this species was the most widely distributed North American cervid (O’Gara and Dundas 2002). Elk were historically abundant in Kentucky, but unregulated hunting and habitat loss led to its statewide extirpation by 1850 (Funkhouser 1925). The species was absent from the state until 1997, at which time the Kentucky Department of Fish and Wildlife Resources (KDF- WR) initiated elk reintroduction efforts using animals from source populations in the western United States (Maehr et al. 1999). Twentieth century elk restoration efforts in eastern states were often unsuccessful. Although these failed attempts often lacked definitive scientific data regarding reasons for decline, managers often implicated meningeal worm infection (Carpenter et al. 1973). Parelaphostrongylus tenuis, the meningeal worm, is a parasitic nematode that can infect a wide variety of cervid species, including elk (Anderson et al. 1966). P. tenuis, like all members of the Protostrongylidae family, requires both intermediate and definitive hosts (Adamson 1986). The normal definitive host for P. tenuis is white-tailed deer (Odocoileus virginanus), and a variety of terrestrial gastropods species serve Kentucky elk / Joe Lacefield

Annual Research Highlights 2008 COMPLETED PROJECTS / Big Game and Small Game Programs

may be at increased risk for meningeal face mines, 9% agricultural or cleared age. Upon recapture, we obtained a worm infection (Raskevitz et al. 1991). lands, and 1% urban (Cox 2003). Coal blood sample and equipped the elk with Although Kentucky elk have a high extraction through surface mining a permanent radio collar. Whole blood probability of escaping fatal meningeal created herbaceous patches that varied was separated into serum and red blood worm infections if they reach adult- in size, ranging from relatively small cells using a centrifuge, then serum was hood (Larkin et al. 2003, Alexy 2004), openings to 1,200-ha fields (Cox 2003). stored in cryogenic vials at -23º C for the effect of this parasite on population The study area was a contiguous subsequent analysis. Prairie Diagnos- growth is unknown. Juvenile survival block of land, but differences in land tic Services (Regina, Saskatchewan, can enhance cervid population growth management resulted in demarcations Canada) conducted all P. tenuis sero- and colonization potential (Taber et al. between property boundaries. The diagnostic tests using indirect ELISAs 1982). Potential detrimental effects of Laurel Fork area encompassed ap- after Ogunremi et al. (2002). We used P. tenuis infection on juvenile elk may proximately 1,200-ha, and primarily a Fisher’s exact test to determine if calf be reduced if immunologically naïve consisted of recently reclaimed surface sex affected acquisition of P. tenuis age classes obtain a degree of para- mines (< 15 years) and remnant forest antibodies from maternal transfer (P <_ site resistance that aids survival until patches, though active surface mining 0.05), and performed logistic regression a complete immune response can be occurred on small portions of the pe- using forward selection to determine if achieved. ripheral landscape throughout the study. predicted calf birth weight was associ- One scenario of acquired resis- The Starfire area was an approximately ated with P. tenuis antibody presence tance could come from acquisition of 7,400-ha site that consisted primarily of (P <_ 0.15). We calculated predicted P. tenuis antibodies through passive reclaimed surface mines and remnant birth weight of neonates by multiplying maternal transfer. This mechanism of forest patches; active surface mining estimated age (days) by 0.635 kg/100 antibody transmission could protect ju- affected a substantial portion of the kcal assumed daily growth rate (Seward venile elk from fatal parasite infection area during the course of the study. The 2003). until they reach immunological matu- Beech Fork area comprised approxi- rity (Hattel et al. 2007), but maternal mately 700-ha, and consisted primarily Results P. tenuis antibody transfer has not been of recently reclaimed surface mines (< Maternal transfer of P. tenuis documented. Despite this conceptual 5 years), though some remnant forest antibodies occurred in over half of elk link between habitat use and P. tenuis patches remained. neonates. P. tenuis antibodies were infection in cervids, previous studies detected in 10 of 19 (53%) neonates in were not able to evaluate this relation- Methods the 2004 cohort, 22 of 40 (55%) neo- ship due to the absence of a reliable We captured elk neonates using nates in the 2005 cohort, and 21 of 38 antemortem test for P. tenuis infection three different search methods: moni- (55%) neonates in the 2006 cohort. Sex (sensu Welch et al. 1991). This compli- toring cows for parturition behavior, did not influence antibody transfer in cation has been ameliorated following ground-based field searches in tra- 2004-06, so data were pooled across all the development of an enzyme-linked ditional calving locations, and aerial years (P = 0.148). Likewise, predicted immunoassay (ELISA) that identifiesP. searches using a helicopter equipped birth weight was not significant toP. tenuis antibodies within blood samples with an observer using an infrared tenuis antibody presence in 2004-06, (Ogunremi et al. 1999, Ogunremi et scope. Calf capture occurred during the and data were pooled across all years al. 2002). Our research objective was months of May and June from 2004-06. (P = 0.951). to determine if maternal transfer of P. Capture techniques followed practices tenuis antibodies occurred in elk. described by Seward (2003). Captured Discussion elk calves were fitted with expandable Passive immunity was previously Study Area vhf radio collars and plastic ear-tags described in cervid species (Grimstad The Kentucky elk restoration zone to facilitate recognition of individuals, et al. 1987, Gaydos et al. 2002), but lies along the Cumberland Plateau, an and we collected approximately 20 mL this study is the first to demonstrate area historically characterized by rug- of blood from the jugular vein of each maternal transfer of P. tenuis antibodies ged topography covered with continu- animal. We recorded sex, approximate in elk. We observed P. tenuis antibody ous second and third growth decidu- age based on umbilicus healing and presence in over half of elk neonates ous forest. Land use has considerably hoof epithelium wear, geographic co- in all three years of this study, which altered the eastern Kentucky landscape ordinates of the capture site, and body suggests that passive maternal transfer resulting in a mosaic that was approxi- weight. We attempted to recapture each of anti-meningeal worm immunoglo- mately 80% second and third growth study animal from the 2004 and 2005 bins commonly occurs in this popula- forest, 10% active and reclaimed sur- cohorts at approximately 6 months of tion. Sex did not influence the presence

14 Kentucky Department of Fish & Wildlife Resources Big Game and Small Game Programs / COMPLETED PROJECTS

of meningeal worm antibodies in elk ficient data to statistically determine sity, Clemson, South Carolina, USA. calves, which is not surprising given if acquisition of maternal antibodies 161 pp. that ruminants acquire maternal anti- was associated with decreased likeli- Anderson, R. C. 1963. The incidence, bodies through ingestion of colostrum hood of P. tenuis-related mortality, but development, and experimental trans- within hours of birth (Carpenter 1956). neonate acquisition of meningeal worm mission of Pneumostrongylus tenuis This suggests that any adaptive value antibodies may have conferred fitness Dougherty (Metastrongyloidea: inherent to antibody transfer operates advantages unapparent in our data. Protostrongylidae) of the meninges independent of sex. of the white-tailed deer (Odocoileus Roulin and Heeb (1999) proposed Management Implications virginianus borealis) in Ontario. that antibody production is positively In this study, we found that more Canadian Journal of Zoology 41: correlated with nutritional intake, while than half of elk neonates acquired P. 775-802. ungulate birth weight is directly affect- tenuis antibodies through maternal Anderson, R. C. 1964. Neurologic ed by maternal dietary success during transfer. Antibody transmission from disease in moose infected experimen- gestation (Keech et al. 2000). Accord- females to their progeny will not occur tally with Pneumostrongylus tenuis ingly, we expected higher predicted unless the dam has been exposed to the from white-tailed deer. Veterinary birth weights to be positively associated pathogen and mounted an immune re- Pathology 1: 289-322. with prevalence of P. tenuis antibodies sponse (Lemke et al. 2003). Grindstaff Anderson, R. C., M. W. Lankester, and in elk calves, but failed to detect any et al. (2006) suggested that mainte- U. R. Strelive. 1966. Further experi- association. Potential explanations in- nance of an elevated immune response mental studies of Pneumostrongylus clude potential error in predicted birth required for transfer of maternal anti- tenuis in cervids. Canadian Journal weights due to inaccurate calf aging, bodies may decrease overall reproduc- of Zoology 44: 851-861. adequate maternal nutrition intake dur- tive ability, but our findings illustrate Boulinier, T., and V. Staszewski. 2007. ing gestation to permit the production that a substantial percentage of cow elk Maternal transfer of antibodies: of large calves while maintaining opti- in Kentucky successfully birthed calves raising immuno-ecology issues. mal immune performance, and differing while maintaining P. tenuis antibody Trends in Ecology and Evolution 23: genetic propensities for addressing P. production. Given this information, it 282-288. tenuis infection within the population. seems unlikely that meningeal worm Buechler, K., P. S. Fitze, B. Gottstein, Exposure of elk calves to maternal infection diminished the reproduc- A. Jacot, and H. Richner. 2002. P. tenuis antibodies could favorably tive potential of Kentucky elk over the Parasite-induced maternal response influence individuals in several ways. course of this study. The Kentucky elk in a natural bird population. Journal Passive immunity to the parasite could population has increased from 1,541 of Animal Ecology 71: 247-252. prevent infection while the juvenile elk translocated animals to an estimated Carpenter, J. W., H. E. Jordan, and is allocating resources to growth and 8,500 individuals in 2008 despite P. B. C. Ward. 1973. Neurological development, potentially increasing fit- tenuis infection. Consequently, habitat disease in wapiti naturally infected ness later in life (Buechler et al. 2002). management for the sole purpose of with meningeal worms. Journal of Early exposure to meningeal worm decreasing P. tenuis infection is likely Wildlife Diseases 9: 148-153. antibodies could challenge the neo- unnecessary in the Kentucky elk popu- Carpenter, P. L. 1956. Immunology nate’s immune system, thus increasing lation. However, we recommend that and serology. W. B. Saunders and the potential for a stronger, secondary monitoring efforts continue to identify Company, Philadelphia, PA, USA. immune response should the individual potential changes in herd demographics 351 pp. come into contact with infective P. over time. Cox, J. J. 2003. Community dynamics tenuis larvae later in life (Boulinier and among reintroduced elk, white-tailed Staszewski 2007). Alexy (2004) noted Literature Cited deer, and coyote in southeastern Ken- that calves and yearlings accounted for Adamson, M. L. 1986. Modes of tucky. Ph.D. Dissertation. Universi- 80% of the P. tenuis-related deaths in transmission and evolution of life ty of Kentucky, Lexington, Kentucky, the Kentucky elk herd. However, in this histories in zooparasitic nematodes. USA. 292 pp. study only 1 of 61 (1.6%) elk calves Canadian Journal of Zoology 64: Forrester, S. G., and M. W. Lankester. died from probable meningeal worm 1375-1384. 1998. Over-winter survival of first- infection. The low observed mortality Alexy, K. J. 2004. Meningeal worm stage larvae of Parelaphostrongylus rate in this highly vulnerable age class (Parelaphostrongylus tenuis) and ec- tenuis (Nematoda: Protostrongyli- suggests P. tenuis infection poses little toparasite issues associated with elk dae). Canadian Journal of Zoology threat to growth and viability of elk restoration in southeastern Kentucky. 76: 704-710. in this region. We did not have suf- Ph.D. Dissertation, Clemson Univer- Funkhouser, W. D. 1925. Wild life

Annual Research Highlights 2008 COMPLETED PROJECTS / Big Game and Small Game Programs

in Kentucky. Kentucky Geological 2003. Non-genetic inheritable poten- tors. Elk of North America: ecology Survey. Frankfort, Kentucky, USA. tial of maternal antibodies. Vaccine and management. Stackpole Books, 385 pp. 21: 3428-3431. Harrisburg, PA, USA. Gaydos, J. K., D. E. Stallknecht, D. Maehr, D. S, R. Grimes, and J. L. Lar- Welch, D. A., M. J. Pybus, W. M. Kavanaugh, R. J. Olson, and E. R. kin. 1999. Initiating elk restoration: Samuel, and C. J. Wilke. 1991. Fuchs. 2002. Dynamics of maternal the Kentucky case study. Proceed- Reliability of fecal examination for antibodies to hemorrhagic disease ings of the Annual Conference of the detecting infections of meningeal viruses (Reoviridae: Orbivirus) in Southeastern Association of Fish and worm in elk. Wildlife Society Bul- white-tailed deer. Journal of Wildlife Wildlife Agencies 53: 350-363. letin 19: 326-331. Diseases 38: 253-257. O’ Gara, B. W., and R. G. Dundas. Grimstad, P. R., D. G. Williams, and 2002. Distribution: past and pres- Funding Sources: Kentucky S. M. Schmitt. 1987. Infection of ent. Pages 67-120 in D. E. Toweill Department of Fish and Wildlife white-tailed deer (Odocoileus virgin- and J. W. Thomas, editors. North Resources ianus) in Michigan with Jamestown American elk: ecology and manage- Canyon virus (California serogroup) ment. Smithsonian Institution Press, KDFWR Strategic Plan. Goal 1. and the importance of maternal anti- Washington D.C., USA. Strategic Objective 5. body in viral maintenance. Journal Ogunremi, O., M. Lankester, and A. of Wildlife Diseases 23: 12-22. Gajadhar. 2002. Immunodiagnosis Grindstaff, J. L., D. Hasselquist, J. A. of experimental Parelaphostrongylus Nilsson, M. Sandell, H. G. Smith, tenuis infection in elk. The Canadian and M. Stjernman. 2006. Trans- Journal of Veterinary Research 66: generational priming of immunity: 1-7. maternal exposure to a bacterial Ogunremi, O., M. Lankester, J. Kend- antigen enhances offspring humoral all, and A. Gajadhar. 1999. Serolog- immunity. Proceedings of the Royal ical diagnosis of Parelaphostrongylus Society B 273: 2551-2557. tenuis infection in white-tailed deer Hattel, A. L., D. P. Shaw, J. S. Fisher, and identification of a potentially J. W. Brooks, B. C. Love, T. R. unique parasite antigen. Journal of Drake, and D. C. Wagner. Mortality Parasitology 85: 122-127. in Pennsylvania captive elk (Cervus Raskevitz, R. F., A. A. Kocan, and J. H. elaphus): 1998 – 2006. 2007. Jour- Shaw. 1991. Gastropod availability nal of Veterinary Diagnostic Investi- and habitat utilization by wapiti and gation 19: 334-337. white-tailed deer sympatric on range Keech, M. A., R. T. Bowyer, J. M. Ver enzootic for meningeal worm. Jour- Hoef, R. D. Boertje, B. W. Dale and nal of Wildlife Diseases 27: 92-101. T. R. Stephenson. 2000. Life-his- Roulin, A., and P. Heeb. 1999. The tory consequences of maternal condi- immunological function of allosuck- tion in Alaskan moose. Journal of ling. Ecology Letters 2: 319-324. Wildlife Management 64: 450-462. Samuel, W. M., M. J. Pybus, D. A. Lankester, M. W. and R. C. Anderson. Welch, and C. J. Wilke. 1992. Elk 1968. Gastropods as intermediate as a potential host for meningeal hosts of Pneumostrongylus tenuis worm: implications for translocation. Dougherty of white-tailed deer. Journal of Wildlife Management 56: Canadian Journal of Zoology 46: 629-639. 373-383. Seward. N. W. 2003. Elk calf survival, Larkin, J. L., K. J. Alexy, D. C. Bo- mortality, and neonatal habitat use lin, D. S. Maehr, J. J. Cox, M. W. in eastern Kentucky. M. S. Thesis, Wichrowski, and N. W. Seward. University of Kentucky, Lexington, 2003. Meningeal worm in a reintro- Kentucky, USA. 68 pp. duced elk population in Kentucky. Taber, R. D., K. Raedeke, and D. A. Journal of Wildlife Diseases 39: McCaughran. 1982. Population 588-592. characteristics. Pages 279-299 in J. Lemke, H., H. Hansen, and H. Lange. W. Thomas and D. E. Toweill, edi-

16 Kentucky Department of Fish & Wildlife Resources Big Game and Small Game Programs / COMPLETED PROJECTS

Estimating Black Bear Populations in Kentucky

Vincent J. Frary, Joseph genetic diversity will be dependent on bear species, the loss and fragmenta- Duchamp, and Jeffery L. continued demographic ties with bear tion of its forested habitat in the eastern populations in the neighboring states of Larkin, Indiana University of U.S. is pronounced (Maehr 1984) and West Virginia, Virginia, and Tennessee. several populations are in jeopardy. Pennsylvania; David S. Maehr, In Kentucky, reproduction in the spe- University of Kentucky; Steven Introduction cies was documented in 2003 after an Dobey, Kentucky Department of Prior to European settlement, the absence of more than a century (Unger Fish and Wildlife Resources central Appalachians were composed of 2007). Individual black bears have relatively unbroken mesophytic forest been confirmed in 26 eastern Kentucky that provided habitat for many spe- counties with most (90%) records from Abstract cies, including the black bear (Ursus those bordering West Virginia, Tennes- Natural colonization of a new or americanus) (Barbour and Davis 1973, see, and Virginia (Unger 2007). The historic range by carnivores is rare. In Whitaker and Hamilton 1998). Un- return of the black bear to Kentucky order to further understand dynamics of regulated timber harvests led to major coincides well with increasing bear these unique populations, and to devel- losses of forested habitat in the region populations in neighboring regions of op appropriate management guidelines (Yarnell 1998), which, combined with Virginia and West Virginia (Virginia to encourage further expansion, estab- overhunting, led to the extirpation of Department of Game and Inland Fish- lishment of baseline demographic data the black bear and other forest carni- eries 2003, West Virginia Department is critical. We estimated the distribution vores in the late 19th and early 20th cen- of Natural Resources 2006), suggesting and abundance of a naturally-recolo- turies (Barbour and Davis 1974). known black bear populations in these nized black bear (Ursus americanus) The return of the black bear to va- states as the most likely sources driving population in eastern Kentucky during cant range in Kentucky is at odds with recolonization. Interstate movements 2007 using mark-recapture statisti- the pattern of distributional collapse of male black bears radio-collared in cal analyses. We used genotyping of among the world’s eight bear species Kentucky have been documented across remotely collected hair samples at 6 (Servheen 1990). Although the black the borders of Tennessee, Virginia, and microsatellite loci to identify individual bear is among the world’s most secure West Virginia, indicating potential de- black bears in Kentucky. Our results suggest that the current distribution of the black bear throughout a 7,575-km2 study area is not uniform, but localized to portions of extreme southeastern Kentucky. A pattern of population concentration and constrained expansion near protected public lands at high elevations may be influenced by human activity, or may be related to the population existing below saturation density. Despite a limited distribution and low estimated abundance (33 males, 56 females) genetic diversity at genotyped loci was high (mean HE = 0.80) and did not indicate non-random mating. To encourage further black bear colonization of Kentucky, we recommend the expansion of the public lands network, and the regulation of anthropogenic activities that may cause further habitat loss and fragmentation. Until the population grows further, maintenance of high Bear snare site / Vince Frary

Annual Research Highlights 2008 COMPLETED PROJECTS / Big Game and Small Game Programs

mographic ties with populations in all province, which is characterized by permission, and 3) logistical feasibility three states (Unger 2007). forested hills and deep, narrow valleys (i.e. rugged topography occasionally Our study was initiated because no (Thornbury 1965). Elevations in this made accessing certain areas time- empirical information has been attained area generally ranged from 300 m to prohibitive or dangerous). Snares were regarding distribution and abundance of 500 m above sea level (Homer et al. baited with two 4.25 oz cans of sar- the black outside of several public lands 2004). dines, and monitored for the presence in eastern Kentucky (Unger 2007). The study area was predominantly of black bear hair during five 8-day We used molecular genetic techniques (89%) composed of mesophytic for- sampling sessions between 15 May and to identify individual black bears in ests, which is characterized by nearly 29 June 2007. Kentucky throughout a much larger 30 dominant tree species (Ricketts et Samples collected at hair snares area than typically addressed through al. 1999). These include maples (Acer were analyzed by Wildlife Genetics telemetry studies (Woods et al. 1999, spp.), oaks (Quercus spp.), hickories International (WGI, Nelson, British Mowat and Strobeck 2000). Because of (Carya spp.), magnolias (Magnolia Columbia, Canada). WGI genotyped all the abundance of forests in the region app.), birches (Betula spp.), and others hair samples at 6 nuclear loci (G10H, and because the overall quality of the (Barbour and Davis 1973). Common G10M, G10L, G10C, G1A, G1D) us- forests in this landscape appeared simi- understory shrubs throughout the study ing methods described in Woods et lar, we hypothesized that the black bear area included mountain laurel (Kalmia al. (1999) to identify individual black would be evenly distributed throughout latifolia) and rhododendron (Rhododen- bears. WGI employed genotyping eastern Kentucky. We also hypothesized dron spp.) (Barbour and Davis 1973). quality assurance and error-checking that observed genetic heterozygosity Active and reclaimed surface mines recommendations described in Paetkau would be high due to links with larger accounted for approximately 6% of (2003) to ensure confident identifica- populations in neighboring states. The the land cover in the study area (Sayler tion of individuals. Genotyping results results of this research are important 2006). The remaining portions of the allowed us to develop mark-recapture for understanding the process of natural landscape (5%) are classified mainly as encounter histories for each individual colonization in a wide-ranging forest agricultural and developed land (Sayler bear identified. WGI used a single sam- carnivore, and in determining the most 2006). ple from each individual to assign gen- appropriate management scenarios to der based on size polymorphism at the facilitate a self-sustaining population in Methods amelogenin gene (Ennis and Gallagher Kentucky. Sampling Design 1994). WGI used a mitochondrial test We overlaid a grid composed of to confirm species (D. Paetkau, WGI, Study Area 313 contiguous cells across the study personal communication) in samples The study area covered approxi- area. Each grid cell was 25 km² in where microsatellite DNA was insuf- mately 7,575 km² in eastern Kentucky size, which at the time field work was ficient to identify individuality, which and included portions of Floyd, Pike, conducted approximated the small- allowed the sample to be used to docu- Letcher, Knott, Perry, Leslie, Harlan, est known black bear home range in ment black bear occupancy. Clay, Bell, Knox, and Whitley coun- Kentucky (Unger 2007, Woods et al. WGI provided measures of genetic ties. The highest elevations were found 1999). We excluded 10 cells within diversity including mean expected and in the southeastern extent of our study this grid because they contained >80% observed heterozygosity, as well as the area, which included the highest points non-forested habitat (n=4), or because mean number of unique alleles at each in Kentucky (Black Mountain and landowner permission for access could locus. We used genotyping results to Cumberland Mountain). Here, eleva- not be obtained (n=6). complete tests for a Hardy-Weinberg tions averaged 450 m above sea level, We installed a barbed-wire hair Equilibrium between genotypes and reaching to 1,262 m above sea level snare in each of the remaining 303 linkage equilibrium between gene loci (Homer et al. 2004). The southeastern cells using methods similar to those to identify evidence of non-random extent of the study area also included described by Woods et al. (1999). Each mating in Kentucky’s bear population Pine Mountain, a narrow, mostly for- snare consisted of 2 strands of barbed (Frankham et al. 2002). We performed ested ridge that extends from Elkhorn wire, wrapped parallel to the ground these tests using program Genepop 3.4 City, Kentucky to Jellico, Tennessee, around 3-4 trees, at 25 and 50 cm over 20 batches of 5000 iterations (Ray- reaching 975 m above mean sea level above ground level. We selected snare mond and Rousset 1995). Alpha levels at its highest point (Homer et al. 2004). locations within each cell based on 3 were adjusted for successive tests us- Our study also included a portion of criteria, ranked in order of importance: ing a progressive Bonferroni correction the northern Cumberland Plateau of 1) presence of contiguously forested (Legendre and Legendre 1998), with the Appalachian Plateau physiographic habitat and bear forage 2) landowner initial significance defined as p< .05.

18 Kentucky Department of Fish & Wildlife Resources Big Game and Small Game Programs / COMPLETED PROJECTS

Table 1. GIS habitat variables used to predict site-specific occupancy of the black bear in 1999). We tested temporal eastern Kentucky in 2007. and behavioral variation in detection probability Variable Resolution Data source X¯ SD by ranking these models in program MARK ac- National Landcover Percent Forest cording to their match to 30m Dataset (Homer et al., 0.82 % .087 (PERFOR) the data using Akaike’s 2004) Information Criterion National Landcover (AIC; Akaike 1977) cor- Elevation (ELEV) 30m Dataset (Homer et al., 481.93 m 114.30 rected for small sample 2004) size (AICc; Burnham and National Landcover Anderson 2002). Model Percent Slope 30m Dataset (Homer et al., 21.67 % 2.61 goodness-of-fit was mea- (SLOPE) 2004) sured using median c-hat tests in program MARK. Human Population 25.46 100m GeoLytics (2003) 16.44 Density (POP) people/ km2 Subsequently, occupancy predictor variables were National Landcover Percent Canopy selected using a stepwise 30m Dataset (Homer et al., 74.05 % 9.55 Cover (CANCOV) backward selection where 2004) all non-correlated habitat Distance from variables were first in- United States Census Nearest Road 30m 497.61 m 240.92 cluded as covariates in Bureau (2000) (ROAD) occupancy models. We re- Distance from National Landcover moved variables according Nearest Forest Edge 30m Dataset (Homer et al., 172.52 m 87.64 to the lowest ratio between (EDGE) 2004) the beta-coefficient and the coefficient’s standard Distance from Kentucky Department error. Weak covariates Nearest Active 30m 2397.13 m 1609.69 of Mine Permits (2007) Surface Mine (MINE) were removed from models in a stepwise fashion Distance Kentucky Department until removal of variables From Nearest of Fish and Wildlife 30m 7287.13 m 6040.72 failed to improve AICc Conservation Land Resources Information values. We considered (CLAND) System (2001) models that were ranked Distance from Kentucky Department within 4 AICc values Colonization Source 30m of Geographic 18,799.75 m 11,497.60 of the best model to be (COLO) Information (2005) competing (Burnham and Anderson 2002), and we Occupancy landscape for each habitat variable. A averaged detection probabilities from We identified 9 habitat variables, window with a radius of 2,447 m cor- these models using program MARK. represented in ArcGIS 9.x (Environ- responded to an area of 18.80 km², the We averaged beta-coefficients of the oc- mental Systems Research Institute, average home range of female black cupancy covariates and their associated Redlands, California) Geographic bears in Kentucky (Unger 2007). All standard errors (Burnham and Anderson Information System (GIS) that we habitat variables were tested for cor- 2002, Moore and Swihart 2005) to fit a considered important in predicting relation (r>0.80) using a Pearson pair- generalized linear model (GLM) using a the probability of occupancy of each wise correlation matrix in program R logit-link transformation. sample location (Table 1). We calcu- software version 2.8.0 (R Development We calculated Moran’s I statistic lated neighborhood statistics in ArcGIS Core Team, Vienna, Austria). values (Moran 1950) for model residu- 9.x for each habitat variable at each We calculated black bear detection als in Program R to check for spatial sampling site using a moving window and occupancy probabilities using en- autocorrelation in our occupancy routine (Carroll et al. 1999, Apps et al. counter histories entered into MacKen- estimates (Moore and Swihart 2005). 2004), with each sampling location as- zie et al.’s (2002) occupancy models in Moran’s I values were calculated at signed the mean value of a surrounding program MARK (White and Burnham nine 5 km distance classes (i.e., 0-5km,

Annual Research Highlights 2008 9 COMPLETED PROJECTS / Big Game and Small Game Programs

5-10km, 10-15 km…40- Table 2. Competing 2007 Kentucky black bear occupancy models after inclusion of a spatial 45km), where 45 km cor- autocovariate. responded to the average width of the study area, AICc MODEL AICc ∆AICc Parameters Deviance or the maximum distance Weight at which we would ex- p(t), (ELEV - PUB + pect autocorrelation to Ψ 350.56 0 0.26 9 331.95 be significant (Moore auto) and Swihart 2005). We p(t), (ELEV – PUB + adjusted alpha levels for Ψ 350.96 0.40 0.21 11 328.05 auto + EDGE – PERFOR) successive calculations using a progressive Bon- p(t), Ψ (ELEV – PUB + ferroni correction (Legen- auto + EDGE – PERFOR 351.08 0.51 0.20 12 326.00 dre and Legendre 1998), - ROAD) with initial significance p(t), (ELEV – PUB + Ψ 351.15 0.59 0.19 10 330.40 defined as p<.05. The auto + EDGE) greatest distance at which spatial autocorrelation p(t), Ψ (ELEV – PUB + was significant defined auto + EDGE – PERFOR 351.66 1.10 0.15 13 324.40 the neighborhood in the – ROAD - POP) calculation of a spatial autocovariate term for occupancy mod- fered significantly from 1:1 α( = 0.05). one pair of loci (G1A, G10M, p < .003). els (Augustin et al. 1996). We included the autocovariate term described above Results Occupancy as an additional covariate in occupancy Field Sampling and Genetic Analyses Eight of 9 habitat variables were models in program MARK. Following A total of 1,402 hair samples were considered for inclusion in the oc- inclusion of the autocovariate, model collected from 254 hair snares. Of cupancy models. CANCOV was cor- selection was repeated as described these, 328 samples collected at 66 snares related with PERFOR and EDGE (r originally, through the stepwise re- exhibited microscopic characteristics of >0.80). Because we considered PER- moval of the weakest predictors of oc- black bear hair, and were submitted to FOR and EDGE to adequately indicate cupancy. We used averaged regression WGI for genotyping. Of the 328 sam- the abundance and quality of forested coefficients from competing models ples, 134 were discarded because of lack habitat, CANCOV was eliminated from to fit a GLM that accounted for spatial of sufficient DNA for extraction n( =131) consideration. autocorrelation and allowed calculation or because they were from non-target Prior to accounting for spatial of black bear occupancy probabilities species (n=3). A total of 194 samples autocorrelation, model selection in pro- for each cell. were positively identified as black bear, gram MARK resulted in 4 competing and of these, 192 were matched to 54 models. Median c-hat tests suggested Abundance individuals, including 20 males and 34 adequate model fit for these data (c-hat We used encounter histories females. From these individuals, 38 < 0.9). All competing models allowed developed for each black bear identi- were snared during only one sampling detection probability (p) to vary with fied in laboratory analyses to estimate session, 10 were snared twice, five were time throughout all sessions (t). Ac- abundance of each gender using closed- snared three times, and one was snared cording to these initial models, black capture statistical models (Otis et al. four times. bears were more likely to occupy areas 1978) in program MARK. We ranked Genetic variability within hair of high elevation (ELEV β = 2.51 ± models that incorporated heterogeneous, samples was high, with mean observed 0.71 SE), that were closer to conserva- temporal, and behavioral variation in (HO) and expected (HE) heterozygosi- tion lands (CLAND β = -1.87 ± 0.89). detection probability in program MARK ties calculated as 0.81 (S.D. = 0.05) and Black bears were also more likely to according to AICc values. Models that 0.80 (S.D. = 0.06), respectively. Mean occupy areas further from the interior were ranked within 4 AICc values of the number of observed alleles at each locus of a forest (EDGE β = 0.84 ± 0.64), “best” model were averaged (Burnham was 7.5 (S.D. = 1.76). Deviations from closer to roads (ROAD β = -0.74 ± and Anderson 2002). We used a Chi- Hardy-Weinberg equilibrium were not 0.61), and with lower amounts of forest square test to determine whether sex detected at any loci (p > .08). Linkage (PERFOR β = -0.62 ± 0.53), although ratios estimated by program MARK dif- disequilibrium was detected between these relationships were only weakly

20 Kentucky Department of Fish & Wildlife Resources Big Game and Small Game Programs / COMPLETED PROJECTS

Figure 1: Probability of black bear occupancy at each mark-recapture sampling MARK resulted in two competing mod- cell in eastern Kentucky in 2007. Calculations were performed using models that els; the first of which allowed detection accounted for spatial autocorrelation. probabilities to be equivalent during sampling sessions one and three, and during sampling sessions two, four, and five (Model: Mt13; Table 3). A competing model allowed for the same temporal variation, but also accounted for heterogeneity in detection probabilities between two unspecified groups (Mod- el: Mht13; Table 3). Abundance was estimated as 33 males (95% C.I. 20-56, S.E. =12.18) and 56 females (95% C.I. 34-95, S.E. = 20.02) after model averaging, resulting in a total population estimate of 89 (95% C.I. 54-151). Aver- age detection probability throughout all sessions was 0.20 (S.E. = .09). Sex ratios favored females (56F:33M, χ² = 5.94, p= 0.015).

Discussion This study developed the first empirical estimate of the naturally recolonized black bear population in Ken- tucky. Forested areas in a three-county area appear to be the centers of black bear abundance in the study area, and occupancy probabilities suggest that black bears are less likely to be encountered as one travels away from extreme southeastern Kentucky (Fig. 1). It is important to note that our results ex- clude McCreary County and a disjunct population that was introduced into the Big South Fork National Recreation Area in 1996-1997 (Eastridge 2000). Nonetheless, the results presented here revealed the primary areas of unassisted occupation in the state. The black bear sex ratio in this related to black bear occupancy. (ELEV β = 1.54 ± 0.75) and areas study favored females, and is similar Spatial autocorrelation in these in closer proximity to conservation to the pattern of captures in an ongo- original occupancy models was signifi- lands (CLAND β = -1.54 ± 0.90). The ing telemetry study (Unger 2007). This cant up to 10 km (p = .01). Therefore effects of other variables were negli- female bias bodes well for population we repeated occupancy model selec- gible. Application of this model across increase in the region and suggests that tion with the inclusion of a spatial au- our study area suggests that high prob- parts of eastern Kentucky may become tocovariate (AUTO). AUTO was found abilities of black bear occupancy are a source population for further expan- to be a positive predictor of occupancy localized to extreme southeastern Ken- sion (Rogers 1987, Onorato and Hell- (β = 0.96 ± 0.65) in five competing tucky (Fig. 1). gren 2001, Unger 2007). Past studies models (Table 2). After accounting for of bear colonization have demonstrated spatial autocorrelation, bears were still Abundance that density (and, thus, the probability more likely to occupy high elevations Abundance estimation in program of occupancy) decreases with distance

Annual Research Highlights 2008 21 COMPLETED PROJECTS / Big Game and Small Game Programs

Table 3. Competing 2007 Kentucky black bear abundance models. diversity reported in other bear populations where many of the same loci were AICc examined (Paetkau et al. 1998, Belant MODEL AICc ∆ AICc Parameters Deviance Weight et al. 2005, Dixon et al. 2007). Linkage disequilibrium was detected at one pair {Mt13} 37.0626 0 0.85706 4 55.67912 of loci. Considered with other evidence it is likely not the result of non-random {Mht13} 40.6447 3.5821 0.14294 7 52.9847 mating, but a result of recent admix- ture of alleles within the population from core reproductive areas and males dispersing bears may be less abundant (Frankham et al. 2002). are often encountered at the periph- than predicted by Unger (2007), and, Small populations of bears typical- ery of occupied range (Swenson et thus, be a hindrance to colonization ly demonstrate lower genetic diversity al. 1998, Onorato and Hellgren 2001, (Sinclair 1992; Onorato et al. 2004). than we found in Kentucky (Paetkau Bales et al., 2005). Interestingly, with Black bear occupancy in Kentucky et al. 1998, Triant et al. 2004, Dixon the exception of a single male, we de- was positively correlated with elevation et al. 2007). High genetic diversity in tected no bears of either sex beyond 29 and negatively correlated with distance our results is likely due to demographic km from core areas. The reasons for from conservation lands. Although ties between Kentucky black bears and this are unclear, but it may reflect an- high elevations are not a requirement bears in neighboring states. In this case, thropomorphic and geographic barriers of black bear occupancy, these areas although the black bear population in to the west, and a population that is still in Kentucky are associated with large Kentucky is small compared to others well below saturation density (Sinclair tracts of forest that are relatively un- nearby in the Appalachian Mountains, 1992, Lidicker and Koenig 1996, On- fragmented. They also receive lower it possesses genetic characteristics of orato et al. 2004). levels of human use than do lower much larger, secure populations. High The pattern of distribution revealed elevations in the study area where genetic diversity in Kentucky’s popula- in this and other work (i.e. Unger 2007) many roads have been built and where tion may eventually permit a wider ar- suggests the geographic expansion of most people live. Conservation lands ray of management strategies (i.e. hunt- the population may be static. Records throughout our study area included ing) than are feasible for small, isolated of wandering individuals suggest that State and National Forests, National populations. dispersing subadult males do attempt Parks, State Nature Preserves, and State to colonize areas outside of our three- Wildlife Management Areas. The man- Management Implications county core area (Unger 2007), but agement of such areas mandates wild- The continued expansion of this due to their philopatric nature (Rogers life and forest stewardship that may be population in Kentucky may be affected 1987) females do not. A similar situ- absent on privately-owned lands. Such by the overall abundance of and con- ation prevails in south Florida where stewardship may have inadvertently nectivity to forested conservation lands male Florida panthers disperse from created a network of bear refuges that available to colonizing individuals. To core breeding habitat across a variety of has resulted in the uneven distribution encourage further black bear coloniza- landscape filters and barriers (Maehr et of the species in the most mountain- tion of Kentucky, the conservation land al. 2002). Despite more than a decade ous region of Kentucky. Based on the network in eastern Kentucky should be of such dispersal attempts the regional relative importance of elevation and expanded. Areas of particular concern pattern of panther distribution has not conservation land in our models, pub- for conservation are potential black changed. licly owned portions of Black Moun- bear habitat and colonization corridors Locations in our study area with a tain, Pine Mountain, and Cumberland identified previously by Unger (2007). high probability of occupancy generally Mountain in the extreme southeastern Throughout eastern Kentucky, practices correspond to the largest tracts of po- portions of the state appear to provide that result in large-scale fragmenta- tential black bear habitat identified by the best black bear habitat in Kentucky. tion and loss of forested habitat should Unger (2007). However, Unger (2007) Despite relatively small estimates be regulated. Future efforts aimed at identified bear habitat in our study area of abundance and geographic cluster- black bear population growth and the where our occupancy model predicted ing of the black bear in Kentucky, preservation of genetic diversity should the species to be absent. Eventually, genetic diversity in the population is concentrate on the identification and as currently occupied habitat reaches preservation of corridors connecting the high. Expected (HE) and observed (HO) saturation density, these areas may be heterozygosity, as well as mean number Kentucky black bear population with colonized by black bears. Alternatively, of alleles at genotyped loci compare fa- those in neighboring states. Periodic habitat suitable for reproducing and vorably to the highest levels of genetic hair snare surveys will likely be an im-

22 Kentucky Department of Fish & Wildlife Resources Big Game and Small Game Programs / COMPLETED PROJECTS

portant tool in understanding black bear York, USA. patriation of black bears to the Big dispersal and colonization behavior as Apps, C.D., B.N. McLellan, J.G. South Fork area of Kentucky and well as the factors that have, thus far, Woods, and M.F. Proctor. 2004. Tennessee. Thesis, University of Ten- limited a more uniform distribution Estimating grizzly bear distribution nessee, Knoxville, USA. across the state’s mountainous region. and abundance relative to habitat and Ennis, S., and T.F. Gallagher. 1994. human influence. Journal of Wildlife PCR based sex determination as- Acknowledgments Management 68:138-152. say in cattle based on the bovine We dedicate this manuscript to our Augustin, N.H., M.A. Mugglestone, amelogenin locus. Animal Genetics friend and co-author, Dr. David Maehr, and S.T. Buckland. 1996. An autolo- 25:425-427. who was killed on 20 June 2008 while gistic model for the spatial distribu- Frankham, R., J. Ballou, and D. Bris- conducting a black bear telemetry tion of wildlife. Journal of Applied coe. 2002. Introduction to conserva- flight in Florida. Dr. Maehr’s insight Ecology 33:339-347. tion genetics. Cambridge University and dedication to the conservation and Bales, S.L., E.C. Hellgren, D.M. Leslie Press, New York, New York, USA. restoration of wildlife and their habitats Jr., and J. Hemphill Jr. 2005. Dy- GeoLytics 2003. 2000 Population Den- will be missed. We thank Dr. Kerry namics of a recolonizing population sity by Block for the Conterminous Murphy and Steve Rosenstock for re- of black bears in the Ouachita Moun- United States [GIS Data]. GeoLytics views and comments. We thank Indiana tains of Oklahoma. Wildlife Society Incorporated, Brunswick, New Jer- University of Pennsylvania, University Bulletin 33:1342-1351. sey, USA. of Kentucky, Kentucky Department Barbour, R.W., and W.H. Davis. 1973. Homer, C., C. Huang, L. Yang, B. of Fish and Wildlife Resources, and Trees and shrubs of Kentucky. Uni- Wylle, and M. Coan. 2004. Develop- Shikar Safari Foundation International versity Press of Kentucky, Lexington, ment of a 2001 national landcover for financial and logistical support. We Kentucky, USA. database for the United States. Pho- also thank all those who granted access Barbour, R.W., and W.H. Davis. 1974. togrammetric Engineering and Re- to property including Cumberland Gap Mammals of Kentucky. University mote Sensing 70:829-840. National Historic Park, Big South Fork Press of Kentucky, Lexington, Ken- Kentucky Department of Geographic National Recreation Area, Kentucky tucky, USA. Information. 2005. County boundary State Nature Preserves Commission, Belant, J.L., J.F. Van Stappen, and D. polygons of Kentucky [GIS Data]. Kentucky Division of Forestry, Ken- Paetkau. 2005. American black bear Commonwealth Office of Technol- tucky Division of Parks, USDA Forest population size and genetic diversity ogy, Frankfort, Kentucky, USA. Service Daniel Boone District, Ken- at Apostle Islands National Lake- Kentucky Department of Mine Permits. tucky Department of Fish and Wildlife shore. Ursus 16:85-92. 2007. Mine permit locations [GIS Resources, Eastern Kentucky Univer- Burnham, K.P., and D.R. Anderson. data]. Kentucky Department for sity, The David School, Alice Lloyd 2002. Model selection and multi- Natural Resources, Frankfort, Ken- College, and hundreds of private land- model inference: a practical-theoretic tucky, USA. owners. Field assistance was provided approach. Second edition. Springer- Kentucky Fish and Wildlife Information by Dr. David Unger, Matt Nickols, An- Verlag, New York, New York, USA. System. 2001. Stewardship Lands drew Whittle, Kevin Kingsland, Nathan Carroll, C., W.J. Zielinski, and R.F. in Kentucky [GIS Data]. Kentucky Dryer, Tom Gieder, Geoff Grisdale, and Noss. 1999. Using presence-absence Department of Fish and Wildlife Re- Carson Lindbeck. data to build and test spatial habitat sources, Frankfort, Kentucky, USA. models for the fisher in the Klamath Legendre, P., and L. Legendre. 1998. Literature Cited region, U.S.A. Conservation Biology Numerical ecology. Second edition. Akaike, H. 1973. Information theory 13:1344-1359. Elseverier Science, Amsterdam. and an extension of the maximum Clark, J.D., D. Huber, and C. Servheen. Lidicker, W.Z., Jr., and W.D. Koenig. likelihood principle. Pages 267-281 2002. Bear reintroductions: lessons 1996. Responses of terrestrial verte- in B.N. Petrov and F. Csaki, editors. and challenges. Ursus:13:335-345. brates to habitat edges and corridors. Proceedings of the 2nd International Dixon, J.D., M.K. Oli, M.C. Wooten, Pages 85-109 in D.R. McCullough, Symposium on Information Theory T.H. Eason, J.W. McCown, and M.W. editor. Metapopulations and wildlife Akademiai Kiado, Budapest. (Repro- Cunningham. 2007. Genetic conse- conservation. Island Press, Washing- duced in pages 610-624 in S. Kot- quences of habitat fragmentation and ton, D.C., USA. zand and L.S. Johnson, editors. 1992. loss: the case of the Florida black MacKenzie, D.I., J.D. Nichols, G.B. Breakthroughs in Statistics, Volume bear (Ursus americanus floridanus). Lachman, S. Droege, J.A. Royle, and One, Foundations and Basic Theory. Conservation Genetics 8:455-464. C.A. Langtimm. 2002. Estimating Springer-Verlag, New York, New Eastridge, R. 2000. Experimental re- site occupancy rates when detection

Annual Research Highlights 2008 23 COMPLETED PROJECTS / Big Game and Small Game Programs

probabilities are less than one. Ecol- servation Biology 12:418-429. black bears (Ursus americanus luteo- ogy 83:2248-2255. Paetkau, D. 2003. An empirical explo- lus) as detected through noninvasive Maehr, D.S. 1984. Distribution of ration of data quality in DNA-based sampling. Conservation Genetics black bears in eastern North Ameri- population inventories. Molecular 5:647-659. ca. Proceedings of the Eastern Work- Ecology 12:1375-1387. United States Census Bureau. 2006. shop on Black Bear Management and Raymond, M., and F. Rousset. 1995. Ge- Second edition Tiger/line files-U.S. Research 7:74. nepop. Population genetics software roads [GIS Data]. .Accessed O.L. Bass, and T.S. Hoctor. 2002. nal of Heredity 86:248-249. 02-December-2007. Florida panther dispersal and con- Ricketts, T.H., E. Dinerstein, D.M. Ol- Unger, D. 2007. Population dynamics, servation. Biological Conservation son, C.J. Loucks, W. Eichbaum, D. resource selection, and landscape con- 106:187-197. DellaSalla, K. Kavanagh, P. Hedao, servation of a recolonizing black bear Moore, J.E., and R.K. Swihart. 2005. P. Hurley, K. Carney, R. Abell, and S. population. Dissertation, University of Modeling patch occupancy by forest Walters. 1999. Terrestrial ecoregions Kentucky, Lexington, USA. rodents: incorporating detectability of North America: a conservation as- Virginia Department of Game and In- and spatial autocorrelation with hier- sessment. Island Press, Washington, land Fisheries. 2002. Virginia black archically structured data. Journal of D.C., USA. bear management plan. . Accessed ous stochastic phenomena. Biometri- 84 in B.D. Chepko-Sade, and Z.T. 28-October-2008. ka 37:17-23. Halpin, editors. Mammalian disper- Whitaker, J.O., and W.J. Hamilton. Mowat, G., and C. Strobeck. 2000. sal patterns. University of Chicago 1998. Mammals of the eastern Estimating population size of grizzly Press, Chicago, Illinois, USA. United States. Comstock Publishing bears using hair capture, DNA pro- Sayler, K.L. 2006. Contemporary Associates, Ithaca, New York, USA. filing, and mark-recapture analysis. land cover change in the central Ap- West Virginia Department of Natural Journal of Wildlife Management palachians Ecoregion. United States Resources. 2006. West Virginia wild- 64:183-193. Geologic Survey. . Ac- www.wvdnr.gov/Wildlife/PDFFiles/ Den Bussche, and D.L. Doan-Crider. cessed 2 December 2006. wvwcap.pdf>. Accessed 28-October- 2004. Phylogeographic patterns with- Servheen, C. 1990. Status and con- 2008. in a metapopulation of black bears servation of the bears of the world. Woods, J.G., D. Paetkau, D. Lewis, (Ursus americanus) in the American International Conference on Bear Re- B.N. McLellan, M. Proctor, and C. southwest. Journal of Mammalogy search and Management. Monograph Strobeck. 1999. Genetic tagging of 85:140-147. No. 2. free-ranging black and brown bears. Onorato, D.P., and E.C. Hellgren. 2001. Sinclair, A.R.E. 1992. Do large mam- Wildlife Society Bulletin 27:616-627. Black bear at the border: natural mals disperse like small mammals? Yarnell, S. 1998. The southern Appa- recolonization of the Trans-Pecos. Pages 229-242 in N.C. Stenseth, W.Z. lachians: A history of the landscape. Pages 245-259 in D.S. Maehr, R.F. Lidicker, Jr., editors. Animal dispersal, General Technical Report SRS-18. Noss, and J.L. Larkin, editors. Large small mammals as a model. London: U.S. Department of Agriculture, U.S. mammal restoration: ecological and Chapman and Hall, London. Forest Service, Southern Research sociological challenges in the 21st Swenson, J.E., E. Sandegreen, and A. Station, Asheville, North Carolina, century. Island Press, Washington, Soderberg. 1998. Geographic expan- USA. D.C., USA. sion of an increasing brown bear Otis, D.L., K.P. Burnham, G.C. White, population: evidence for presatura- Funding Source: Pittman Robertson and D.R. Anderson. 1978. Statisti- tion dispersal. Journal of Animal (PR) cal inference from capture data on Ecology 67:819-826. closed animal populations. Wildlife Thornbury, W.D. 1965. Regional geo- KDFWR Strategic Plan. Goal 1. Monographs 62. morphology of the United States. Strategic Objective 5. Comprehen- Paetkau, D., L.P. Waits, P.L. Clarkson, John Wiley and Sons, New York, sive Wildlife Conservation Strategy: L. Craighead, E. Vyse, R. Ward, New York, USA. Appendix 3.9; Class Mammalia: and C. Strobeck. 1998. Variation in Triant, D.A., R.M. Pace III, and M. Taxa specific conservation project. genetic diversity across the range of Stine. 2004. Abundance, genetic di- North American brown bears. Con- versity and conservation of Louisiana

24 Kentucky Department of Fish & Wildlife Resources Big Game and Small Game Programs / COMPLETED PROJECTS

Assessment of Habitat Value for Recovering Disturbed Warm-Season Grass using a Multi-Cover Habitat Assessment Model for the Northern Bobwhite

Brittany Compton, Lindsey to determine if these disturbance re- goals of this study were to: 1) Evaluate Wilson College; Samantha gimes are effective in improving habitat effectiveness of several disturbance quality. The U.S. Fish and Wildlife practices (physical and chemical) on Fugate, Lindsey Wilson College; Service has recently-developed Habi- Green River Wildlife Management Cassie Niersel, Lindsey Wilson tat Evaluation Protocol (HEP) models Area using the HEP model for Northern College; Andrea Dykes, Lindsey for target species designed to assess Bobwhite Quail; 2) Thin warm season Wilson College; Paul Huffman, habitat quality for target species. These grasses to optimal densities for North- Lindsey Wilson College; Laura models use cover type and habitat qual- ern Bobwhite and promote appropriate ity indicators to assess habitat value stem densities and diversity of forbs Sidebottom, Lindsey Wilson for target species with known habitat and woody species; and 3) Increase College; Susan Monteleone, requirements. For Northern Bobwhite, amount of bare ground to improve Lindsey Wilson College, Brian HEP models characterize optimal habi- wildlife use of habitat area. Gray, Kentucky Department of tat as containing 50% up-right growth Fish and Wildlife Resources cover (forbs and woody) and 50% bare Methods and Materials ground. Beginning during the summer The 7- acre experimental tract on of 2006, a 7-acre experimental area was Green River Wildlife Management Area Introduction established on the Green River Wildlife was divided into five one-acre plots. Disturbance regimes such as disk- Management Area (GRWMA) to deter- Of these plots, four were mechanically ing, burning, and chemical treatment mine what type and rate of disturbance disturbed (disked 2x, 4x, 6x, 10x), one are often used to improve habitat qual- is most effective for improving North- plot was treated with 1 quart (41%) ity for wildlife. There are several ways ern Bobwhite habitat. Specifically, the Glyphosate, one plot was treated with

Figure 1.

Annual Research Highlights 2008 25 COMPLETED PROJECTS / Big Game and Small Game Programs

1.5 quarts (41%) Glyphosate, and the last plot was a control plot receiving no treatment. After implementation of treatments, habitat parameters were measured for each study plot (9 randomly sampled points per study plot). For each sample point, the following habitat parameters were collected: percent bare ground, percent grass cover, percent woody cover, pent forb cover, number of grass clumps, and number of woody stems. These specific Figure 2. habitat parameters were chosen because quality of cover grass clumps. plots. In terms of chemical versus me- for Northern Bobwhite is considered Bare ground cover ranged from chanical disturbance, data trends show the most important habitat characteris- 25% to 31% with the 6x plot having that the mechanical disturbance of disk- tic for the Habitat Evaluation Protocol the highest proportion of bare ground ing 6 times is similar to a 1.5 quart/acre Model designed by U.S. Fish and Wild- (31%). Percent cover of forbs ranged Glyphosate treatment. life Service. from 8% to 28% with the 10x plot hav- For each experimental plot, sample ing the highest forb component of all Management data were averaged and standard devia- treatments. Grass coverage ranged from Recommendations tions were calculated. Percent coverage 35% to 61%; the 2x and 4x plots had the Since the 6x and 10x treatments data were normalized using arcsine highest percent cover of grasses. Woody appear successful, but did not differ transformation. We compiled percent percent cover ranged from 6% to 24% enough from one another to warrant cover and stem density bar graphs to and the 6x plot had the highest percent- 10x disking, we recommend using assess patterns and trends in the col- age of woody cover (see Figure 2) either a 6x disking protocol or a 1.5 lected data. quart/acre (41%) Glyphosate applica- Discussion tion to optimize Northern Bobwhite Results In considering the HEP model of habitat and minimize labor and man- The number of grass clumps within the Northern Bobwhite and its pre- agement costs. treatment plots did not show a large ferred habitat, the 6x and 10x disking range between treatment types, while the improved habitat quality by providing Funding Source: Lindsey Wilson Col- number of woody stems varied from 8.4 adequate amounts of bare ground and lege, Kentucky Department of Fish and in the plot disked six times to less than cover. In these treatment areas, quail Wildlife Resources, Pittman Robertson 2 stems per sample in the plots disked would have enough room to move (PR) two and four times (see Figure 1). Plots through the grass without being detect- treated with either 1.0 quart or 1.5 quarts ed by predators. The 2x and 4x treat- KDFWR Strategic Plan. Goal 1. of Glyphosate exhibited woody growth ments did not appear successful in cre- Strategic Objective 5. Comprehen- intermediate to the 2x/4x and 6x/8x ating a suitable habitat for the Northern sive Wildlife Conservation Strategy: treatment plots. Glyphosate plots were Bobwhite since less bare ground and Appendix 3.2; Class Aves: Taxa spe- also characterized by greater numbers of less woody cover was present in these cific research project #2.

26 Kentucky Department of Fish & Wildlife Resources Wildlife Diversity and Fisheries / COMPLETED PROJECTS

Life History and Population Characteristics of Moxostoma poecilurum, the Blacktail Redhorse, in Terrapin Creek, Graves County, Kentucky

Brooks Burr and Robert Gerwig, the Blacktail Redhorse is restricted to known reproducing population prior to Southern Illinois University the Gulf Coastal Plain from eastern this study occurs in downstream North Texas to the Choctawhatchee system, KDFWR Contact: Ryan Oster Fork Obion River, Tennessee (Burr Florida and Alabama, except in the Mo- and Carney 1984). Terrapin Creek is a bile River Basin where it occurs above direct tributary to the North Fork Obion Introduction the Fall Line to the headwaters. The River. Therefore, an examination into The Blacktail Redhorse is a small Blacktail Redhorse reaches the north- the life history and population status, species of the genus Moxostoma that ernmost limit of its North American especially the possibility of Terrapin grows to a maximum total length of range in Terrapin Creek, Graves Coun- Creek supporting a reproducing popula- 508 mm (Carlander 1969). Habitats ty, Kentucky, where it was formerly tion of Blacktail Redhorse was essential for this species range from sluggish considered sporadic and rare (Burr and for future conservation efforts to main- Coastal Plain Streams with soft sand Warren 1986). The Blacktail Redhorse tain the species presence in the state of and silt substrates to more typically up- has a Kentucky Heritage status as en- Kentucky. land streams and rivers with high gradi- dangered and is considered critically This study is the first to examine ents and firm coarse substrates (Etnier imperiled (S1) in Kentucky (Nature- the life history of the Blacktail Red- and Starnes 1993). It can also be found Serve 2004). The species is one of 59 horse in any detail, and is intended to in the numerous impoundments of the fish species identified as a “species of reveal whether Terrapin Creek supports south-central United States. The Black- greatest conservation need” under the a reproducing population of the species. tail Redhorse tolerates (at least tempo- State’s comprehensive Wildlife Action Additional objectives were to determine rarily) brackish conditions based on a Plan. In Terrapin Creek, both adults and life history aspects (e.g., spawning pe- specimen collected from the Escambia juveniles have been collected from san- riod and habitat, age at first reproduc- River where salinities were above 5 ppt dy-bottomed pools: however the closest tion, diet analysis, recruitment trends, at the surface (Boschung and Mayden 2004). The species is a benthic feeder consistent with other members of the genus Moxostoma (Boschung and Mayden 2004). Spawning is known to occur in spring when water temperatures near 20 C in shoal areas that are 1-10 m wide (Kilken 1974). This too is consistent with other members of the genus Moxostoma which also utilize smaller streams for spawning (Meyer 1962; Bowman 1970; Curry and Spacie 1984). Overall however, the biology of the species has received little attention. The distribution of Terrapin Creek / Brooks Burr

Annual Research Highlights 2008 27 COMPLETED PROJECTS / Wildlife Diversity and Fisheries

critical habitat) crucial to an informed pect in obtaining nutrition from detritus rectilinear in shape (Pearson and Healy management plan for future conserva- and living plant material (Weisel 1962; 2003). All fish were examined to deter- tion efforts that will benefit this species Jenkins 1970; Jenkins and Burkhead mine overall condition such as, repro- and others found in the Terrapin Creek 1994). Contents of the “stomach” (the ductive condition (free flowing milt or watershed. straight gut from pharynx caudad to eggs, presence of tubercles, coloration), first descending loop--Bowman 1970; external parasites, the presence of any Methods and Materials Rupperecht and Jahn 1980; Moss et lesions or other abnormalities. Sampling efforts for Blacktail al. 1983; Peterson et al. 1999) and the Macrohabitat characteristics of Redhorse began on 14 January 2007 remaining anterior 2/3 of the intestinal each of 3 replicated sites were deter- and terminated on 05 October 2008. tract was suspended in water and agi- mined using a Rapid Bioassessment Sampling was usually conducted every tated gently to remove all food items Protocol designed by the U.S. Environ- 2-3 weeks.The following data were only when the anterior 1/3 was empty. mental Protection Agency (Barbour et compiled for a detailed examination The contents were identified as fol- al. 1999). Environmental parameters into the biology of the species: 1) age lows: invertebrates, algae, detritus, and such as bank stability, velocity/depth and growth; 2) reproductive timing and sediment. Invertebrates were identified regime, and riparian characteristics output; 3) food habits and diet analysis; to the lowest taxonomic level possible were scored using a range of values and 4) seasonal migrations. Age deter- (order or family). Food items were from 0-200 for 10 parameters. Sub- mination was made by counting annual expressed as frequencies of occur- strate type was assessed using a modi- rings on opercle bones. Lengths of fish rence and percent volume. In order to fied Wentworth scale (Cummins 1962) were recorded in millimeters. Three obtain percent volume of a food type as follows: sand (<4 mm), pebble (5-64 measurements were taken for each all material from each stomach was mm), cobble (64-256 mm), boulder fish: standard length (SL), fork length grouped according to taxon or category (>256 mm), silt, clay, and organic detri- (FL), and total length (TL). Reproduc- (detritus and sediment) and placed over tus following a Wolman pebble count. tive timing and potential output were a grid system (Windell 1971). The total Instream habitat features were noted estimated using the gonadosomatic number of squares that each group of and the total amounts of these features index (GSI). GSI = 100(GWW/WW), food items occupied was counted and within each reach were visually esti- where GW is the gonad wet weight and expressed as a proportion of the total mated. Water quality parameters were WW is total adjusted wet weight (de amount of squares that were occupied measured including pH, dissolved oxy- Vlaming et al. 1982; Crim and Glebe by all food material (White and Haag gen, and temperature. Land use activi- 1990). To estimate potential annual fe- 1977). The data were pooled accord- ties in the watershed were noted (and cundity, ovaries were removed and eggs ing to age class of fish (Timmons et al. photographed) to determine potential counted, measured (mm), and placed 1983; Peterson et al. 1999). To deter- threats to the watershed and ultimately in categories of developmental stage as mine any daily patterns of feeding ac- the Blacktail Redhorse population. outlined by Heins and Rabito (1986): tivity the fullness of the “stomach” and Microhabitat parameters were re- latent (LA), early maturing (EM), late time of capture were noted. corded for adult and juvenile redhorse maturing (LM), mature (MA) and ripe In order to ascertain any seasonal at their respective points of capture. (RE). Spawning behavior and habitat migrations within Terrapin Creek, 40 At each capture site the following were observed through observation of individuals were fitted with yellow T- procedures were used to determine mi- spawning fish from an elevated position bar Anchor Floy tags (Floy Tag Inc.) crohabitat use: 1) a series of 10 depths on the stream bank. Spawning habitat beginning on 08 June 2007 and ending were taken with a 1.5 m graduated staff parameters were recorded in a field on 18 July 2007. The tag was inserted in 0.05 m increments and averaged, 2) notebook including velocity, depth, between the first and second dorsal a series of 10 velocity measurements substrate characteristics, macrohabitat fin ray as to lock the T-bar in place. (m/sec) with a Marsh-McBirney flow type (riffle, run, or pool), and tempera- A total of 14 males and 26 females meter were recorded and averaged, and ture. were fitted with tags. All tagged indi- 3) instream location of capture was Food habits were determined for viduals were immediately released at determined by either stream margin or all retained adult specimens, and 10 their initial point of capture. All speci- mid-channel, macrohabitat type (riffle, age-0 individuals from each season. mens for mark-recapture data were run, or pool), or microhabitat type (un- The Blacktail Redhorse, similar to sexed, weighed to the nearest 0.1 g, dercut bank, large organic debris, root other members of the Catostomidae, and lengths in mm recorded (TL, SL, wad, or open channel). Large organic does not possess a true stomach but FL). Sexing of fish was accomplished debris was in the form of wood and was instead a large coiled gut that probably through examination of the anal fin; classified into three separate categories: increases digestion time, a critical as- males are fan shaped and females are clumps, jams, and scattered (Brooks

28 Kentucky Department of Fish & Wildlife Resources Wildlife Diversity and Fisheries / COMPLETED PROJECTS

et al. 2003). Microhabitat types were measured with a roll tape and percent total area was estimated for each habitat type.

Results Age Determination was made for a total of 101 sexed specimens by counting annual rings on the opercle bone (McConnell 1951; Scoppettone et al. 1986; Scoppettone 1988; Sum- merfelt and Hall 1987; Peterson et al. 1999). Males ranged in age from 1 to 7 years old with most (32 of 41,78.05%) between 1 and 3 years old. Those identified as females ranged in age from 1 to 7 years old and were more evenly distributed among age classes with the Male Blacktail Redhorse in breeding condition / Brooks Burr exception of only 1 female recorded as 1-year-old. The majority of individuals greater fertilization success (Burr and ryngeal apparatus. It should be noted captured in this study were young- Morris 1977). The trio performed this that young-of-the-year redhorse less of-the-year fish (63.96%), and were behavior 3 times within 25 minutes of than six months in age did not contain grouped in the age 0 category. These observation. Each individual spawning any discernible bits of vegetative or were determined to be age 0 by follow- act lasted approximately 5 seconds, other detrital stuffs; it is assumed that ing the development of young-of-the- though precise times were not recorded. the fine brown material found in young year individuals following the spawn- The female drifted downstream into stomachs was periphyton growing ating period in late May and early June. the adjacent pool and disappeared from tached to large pieces of detritus (leaf sight. The males moved upstream and pack) that accumulated in preferred Age and Growth were not seen again. Examination of habitat. Ontogenetic shifts in both Blacktail Redhorse were observed the substrate in the area that the spawn- categories of food stuffs and mean actively spawning only once during the ing behavior occurred did not reveal volumes (Table 3 and Figures 16 and course of this study on 4 June 2007 at any products (i.e., water-hardened 17) were evident as the fish approached approximately 1825 hrs. Behavior did eggs) of the spawning act. This spawn- maturity. These shifts are most likely not differ markedly from behaviors de- ing behavior occurred in a sandy run due to increases in mouth gape size as scribed for other species of Moxostoma with flows ranging between 0.08 to has been described for Spotted Suckers (Meyer 1962; Bowman 1970; Curry 0.17 m/sec (mean 0.11 m/sec). Depth (White and Haag 1977) and Blackfin and Spacie 1984; Jenkins and Burkhead ranged from 14 to 29 cm (mean 23.6 Suckers (Timmons et al. 1983) result- 1993). A single female was observed cm), the dominant substrate was sand ing in increasing exploitative benefits being followed by two males. The two (65%) and fine gravel (35%), and water for gathering food stuffs in increasing males lagged just behind the female on temperature was 21.5º C. It appeared quantities and compositions. In age 0 either side of her. In a sudden rush the that the spawning fish came from a pool specimens detritus/periphyton made female made quick, spastic vibrating that was immediately downstream (8.2 up 35.8% of the total volume of the motions while facing upstream. At this m) from the spawning site. During the diet with lesser amounts of Cladoc- same time the males quickly joined her course of the anticipated spawning pe- era (24.4%), Ostracoda (19.6%), and on both sides and commenced with a riod of 2008 several high water events Hydracarinae (12.3%). Chironomidae rapid vibrating action. It appeared that made observation impossible due to made up only 6.0% of the mean volume the caudal peduncles of these fish were increases in turbidity and depth. of the diet. The diet of juvenile fish scouring the substrate while swimming All samples of stomachs examined became more diverse in both respect upstream in the vibrating manner de- contained detritus and periphyton. No to volume and composition as they scribed. The large fan-shaped anal fins attempt was made to discern between approached age III. Detritus inputs be- of the males probably aided in direct- very fine detritus and periphyton due to came greater and ranged from 19.3% ing reproductive material towards the the difficult nature of separating such volume in age I fish to 20.7% in age II eggs that were being deposited assuring food stuffs after mastication by the pha- fish. These trends in increased inges-

Annual Research Highlights 2008 29 COMPLETED PROJECTS / Wildlife Diversity and Fisheries

tion of detrital inputs are consistent groundwater inputs must exist along in a variety of habitats but most often with observations on other species of most of Terrapin Creek. Flood events, near some kind of small aggregation catostomids ( Bowman 1970; Moyle however, were somewhat frequent and of woody debris (i.e., clumps sensu and Mariochi 1975; White and Haag were characterized by very rapid rises Brooks et al. 2003). They were general- 1977;Timmons et al. 1983; Pearson and falls in water levels (Figures 4 and ly located in fairly shallow (0.20 – 0.60 and Healey 2003). Insect items became 5). This is almost certainly a result m) runs with sandy bottoms and mod- more important during this time and of past channelization efforts and the erate flows relative to Terrapin Creek were dominated by the Chironomidae likely cause of overall low habitat as- (0.08 – 0.12 m/sec). These clumps were occurring in 88.9% of age I fish and sessment scores. either mid-channel or along the stream found in 100% of age II fish. The Chi- margin, most often along stream mar- ronomidae were the only insect group Microhabitat gins adjacent to current. encountered in juvenile fishes less than Microhabitat variables were re- two years old. corded for Blacktail Redhorse that were Discussion/Management Habitat characteristics were docu- captured in this study. Following early Reccomendations mented at each of the three replicate development (the first two months of The biology of the Blacktail Red- sites that were sampled for relative life) Blacktail Redhorse could be found horse is similar to that described for abundance data. Each site represented consistently along stream margins as other species of Moxostoma. As con- a typical reach of stream relative to the opposed to mid-channel locations. Be- sidered here, the species is a benthic site conditions. Sites varied between cause of this difference in habitat use trophic generalist, consistent with other 113 m to 144 m in length and contained the principal components analysis re- species of Redhorse that feed on detri- riffle, run, and pool habitat units. Aver- vealed no specific habitat requirements tus and the chironomid inhabitants typi- age depths ranged from 0.28 to 0.47 m for juvenile redhorse. Partial least cal of pool environments where organic between sites and had maximum depths squares regression and principal com- matter accumulates. The importance of of 1.3 m at the farthest downstream ponents analysis revealed strong habitat such fish in stream ecosystem function site to 0.56 m at the most upstream preferences of adults for pool environ- has been poorly studied, but that should site. Following protocols suggested by ments with large woody debris (e.g., not undermine the significance of these the Environmental Protection Agency log jams or clumps of large woody organisms. Redhorses, including the (Barbour et al. 1999) scores were based debris as defined by Brooks et al. 2003) Blacktail Redhorse, utilize organic al- on visual inspection and specific mea- and adequate depth (Figures 19 and lochthonous stream inputs readily and surements (depth, velocity, substrate, 20). Blacktail Redhorse, especially such material made up the majority of and instream cover). The results of adults, were particular in preference the diet of redhorse in this study; they these scores and observations were that of habitat. With the exception of adult also, however, ingest significant quanti- Terrapin Creek supports only marginal male Blacktail Redhorse, which occupy ties of aquatic invertebrates and there- to suboptimal habitat for Blacktail riffle and run areas in spring prior to the fore represent a unique trophic position. Redhorse. Habitat scores for pool spawning period, all adult redhorse in Their trophic behavior not only helps variability, epifaunal substrate avail- this study were captured in calm pool recycle coarse and fine particulate or- ability, sediment deposition, channel habitats with ample woody cover (Fig- ganic material as a primary consumer, flow status, bank stability, and chan- ure 21). Results of partial least squares but also as a secondary consumer nel alteration all had scores that were analysis (F = 14.99, p = 0.007) revealed ingesting significant quantities and between suboptimal and marginal as positively correlated habitat variables diversity of aquatic macroinvertebrates outlined in definitions and descriptions affecting abundance of Blacktail Red- (though chironomid larvae made up the in the protocol (Barbour et al. 1999). horse. The most highly ranked habitat largest quantity of insects ingested). Riparian vegetation scores were within variables included amount of detritus, This is especially important if one the optimal range. Discharge estimates depth, log jams, canopy cover, under- considers that suckers usually make up were conducted on 2 October 2007 cut bank, and location of habitat along the majority of fish biomass in aquatic and were calculated to be 49 m³/ sec. stream margins. Increases in velocity systems (Jenkins and Burkhead 1993). Water temperatures ranged from 5 to were negatively correlated with overall River Redhorse, Moxostoma carinatum, 23 º C during this study. We never wit- numbers of redhorse captured. Velocity may be an exception to this, and are nessed any sudden fluctuations in water explained 41.2% of the variance among specialized feeders on mollusks (Tatum temperature or cessation of flow, even predictors. These patterns were also ev- and Hackney 1969) and the species of during the drought conditions in the ident using principal components anal- western suckers (genus Catostomus) summer of 2007. We agree with Burr ysis (Figure 20). Juvenile and young- found in the large rivers of the south- and Warren (1986) that considerable of-the-year redhorse were encountered west are primarily invertebrate feeders

30 Kentucky Department of Fish & Wildlife Resources Wildlife Diversity and Fisheries / COMPLETED PROJECTS

(Dauble 1986). Power et al. (1996) Redhorse, but we note that no adult tions, one that is resident and one that defined a keystone species as “a species individuals were found after October is migratory. The Terrapin Creek popu- whose impact on its community or eco- and none before March. It is therefore lation could be a similar case; we only system is large, and disproportionately likely that adults use this system only recaptured a few individuals several large relative to its abundance.” Wheth- seasonally as a breeding area. Page and times, but did not recapture any other er or not the Blacktail Redhorse is a Johnston (1990) and Curry and Spacie individuals that were also captured with keystone species is not understood, but (1984) both summarized reproductive tagged individuals from the same loca- considering its overall large size, low behaviors of catostomids and described tion within a pool. The lack of Blacktail abundance in comparison to other spe- migrations into smaller stream systems Redhorse in winter samples could be cies of fish commonly encountered in where reproduction takes place. Burr the result of sampling bias because the Terrapin Creek, and the unique trophic and Morris (1977) described similar deepest pools were inaccessible be- guild it occupies, it could be speculated spawning migrations of Shorthead cause of depth (> 1.3 m), and the large that it is an important component to nu- Redhorse, Moxostoma macrolepidotum, and complex amounts of woody debris, trient cycling in this watershed. into a tributary of the Fox River in most notably log jams, which prevented Blacktail Redhorse are not a long- northeastern Illinois. Werner (1979) the use of effective use of seines and lived species in comparison to others hypothesized that White Suckers, Ca- electrofishing gear. Considering the in the family. Specimens examined tostomus commersonii, home to their overall amount of effort given to cap- during this study indicate a maximum natal spawning streams primarily on turing Blacktail Redhorse in Terrapin age of 7 years. Scoppettone (1988) the basis of olfactory cues. However, Creek through all seasons indicates found western species of suckers lived Bowman (1970) reported no site fidel- that the adults are, in reality, a transient as long as 43 years (e,g,, Deltistes ity to any specific spawning shoal. We population that uses Terrapin Creek for luxatus) and most lived between 15 to recaptured a tagged female a year later spawning and a summer refuge for fe- 30 years. Blacktail Redhorse longev- from the exact location where she was males following the spawn. ity is similar to other small suckers caught the previous year. This female The preferred habitat of adult such as Golden Redhorse, Moxostoma had recently completed spawning when Blacktail Redhorse is similar to that erythrurum, which lived a maximum of first caught and was found to have just described for other species of redhorse. 8 years (Meyer 1962) and the Blackfin completed spawning at the time of re- They are inhabitants of pools with Sucker, Thoburnia atripinnis (Timmons capture (through examination of the go- woody debris temporarily occupying et al. 1983), which did not exceed 5 nads). Perhaps there is some fidelity to shoals/riffles for spawning or mov- years of age. Growth rates of Black- spawning shoals in the Terrapin Creek ing upstream. The importance of large tail Redhorse in Terrapin Creek were population of Blacktail Redhorse. woody debris in streams cannot be somewhat slower than that calculated There were differences in seasonal understated. Debris dams are critically for other species of redhorse and was abundance of Blacktail Redhorse in important in the retention of organic much slower than what Kilken (1974) Terrapin Creek. Young-of-the-year fish materials in stream systems (Jones and reported in pond-reared individuals; he were nearly always present but adults Smock 1991) and this organic material found that young-of-the-year Black- appeared to be seasonal. Adult Black- is essential for secondary production tail Redhorse reached 186 mm TL in tail Redhorse first appeared in March of invertebrates and therefore the fish 6 months. Blacktail Redhorse only with an onset of males that occupied that feed upon both detritus and mac- reached 88 mm TL in their first year of large woody debris adjacent to flow- roinvertebrates. These debris dams, life in Terrapin Creek. A tagged female ing waters of runs and riffles. Female most notably log jams, not only are individual (tag 30) grew only 11 mm redhorse began to appear by the end of important in nutrient cycling through (from 283 mm SL to 294 mm SL) after March and occupied pool habitats with trophic levels but also increase habitat being recaptured nearly a year later. low flows and usually were found in heterogeneity (Smock et al. 1989) in Before the onset of this study the deepest water relative to available low gradient streams such as Terrapin it was not known if Terrapin Creek habitats. The mark – recapture efforts Creek. Nearly every pool that contained sustained a reproducing population revealed site fidelity (Table 5) with redhorse was immediately downstream of Blacktail Redhorse, and the down- the majority of recaptured specimens from a log jam or several large clumps stream North Fork Obion River, Ten- coming from the same habitats within of woody debris. These pools were nessee, was regarded as the closest re- the same pools from where they were formed from the scouring action of producing population (Burr and Carney initially encountered. Curry and Spacie water flowing through and over these 1984; Burr and Warren 1986). It is now (1984) suggested that Golden Redhorse obstructions. All undercut banks where clear that Terrapin Creek does have a found within the same stream could redhorse occurred were also immedi- reproducing population of Blacktail actually represent two distinct popula- ately downstream from large woody

Annual Research Highlights 2008 COMPLETED PROJECTS / Wildlife Diversity and Fisheries

debris. Woody debris also helps reduce the Black Redhorse, Moxostoma ecology of the Largescale Sucker the energy of flowing water, and this is duquesnei (Lesueur), in Missouri. (Catostomus macrocheilus) in the especially important in Terrapin Creek Transactions of the American Fisher- Columbia River. American Midland where channelization is quite evident. ies Society 3:546-559. Naturalist 116:356-367. DeVlaming, We consider the effects of channeliza- Brooks, K. N., P. F. Ffolliott, H. M. V. G., G. Grossman, and F. Chap- tion as still having a negative impact on Gregersen, and L. F. DeBano. 2003. man. 1982. On the use of gonosomat- the overall health of this system with Hydrology and the Management of ic index. Comparative Biochemistry continued increases in sedimentation, Watersheds. 3rd edition. Iowa State and Physiology 72A:31-39. stream bed instability, headcutting in University Press, Ames, Iowa. Etnier, D. A., and W. C. Starnes. 1993. the upstream reaches (especially above Burr, B. M., and M. A. Morris. 1977. The Fishes of Tennessee. University County Road 1485), downcutting, Spawning behavior of the Shorthead of Tennessee Press, Knoxville. and bank sloughing. Erosional stages Redhorse, Moxostoma macrolepi- Fuiman, L. A., and D. C. Witman. have been classified by Schumm et dotum, in Big Rock Creek, Illinois. 1979. Description and comparisons al. (1984) who developed a five stage Transactions of the American Fisher- of catostomid fish larvae:Catostomus classification for stream reaches that ies Society 106:80-82. catostomus and Moxostoma erythru- had undergone headward erosion. With Burr, B. M., and D. A. Carney. 1984. rum. Transactions of the American stage I being the natural stream chan- The Blacktail Redhorse, Moxostoma Fisheries Society 108:604-619. nel prior to headcutting and ending poecilurum (Catostomidae), in Ken- Greenwood, P. H., D. E. Rosen, S. H. with stage V where the stream is again tucky, with other additions to the Weitzman, and G. S. Myers. 1966. meandering on a new floodplain that State ichthyofauna. Transactions of Phyletic studies of teleostean fishes, may be significantly lower in elevation the Kentucky Academy of Science 45 with a provisional classification of than the old one that is now a terrace (1-2):73-81. living forms. Bulletin of the Ameri- (Ross 2001). We suggest that Terrapin Burr, B. M., and M. L. Warren, Jr. can Museum of Natural History 131: Creek is in stage III due to some bank 1986. A Distributional Atlas of Ken- 339-456. failures and loss of riparian vegetation tucky Fishes. Kentucky State Nature Guerrero, R. D., III, and W. L. Shel- into the stream as erosional processes Preserves Commission Scientific and ton. 1974. An aceto-carmine squash continue. In order to insure the contin- Technical Series No. 4. method for sexing juvenile fishes. ued existence of Blacktail Redhorse in Carlander, K. D. 1969. Handbook of Progressive Fish-Culturist 36:56. Terrapin Creek it is our judgement that Freshwater Fishery Biology, Volume Heins, D.C., and F. G. Rabito, Jr. 1986. large woody debris is essential in creat- 1. Iowa State University Press, Ames, Spawning performance in North ing the heterogeneity of habitat needed Iowa. Crabtree, R.E., C.W. Hamden, American minnows: direct evidence for this species, especially the presence D. Snodgrass, and C. Stevens. 1996. of the occurrence of multiple clutch- of shoal areas critical for spawning and Age, growth, and mortality of bone- es in the genus Notropis. Journal of pool habitats that allow the species to fish, Albula vulpes, from the waters Fish Biology 28:483-488. recover from the stressors of migration of the Florida Keys. Fishery Bulletin Jenkins, R. E. 1970. Systematic studies and spawning. Not only does woody 94:442-451. of the catostomid fish tribe Moxos- debris favor the Blacktail Redhorse but Crim, L.W., and B.D. Glebe. 1990. tomatini. Unpublished Doctoral Dis- also other species of fish found in Ter- Reproduction. Pages 529-553 in sertation. Cornell University, Ithaca, rapin Creek. C.B. Schreck and P.B. Moyle, New York. Jenkins, R. E., and N. M. editors. Methods for Fish Biology. Burkhead. 1993. Freshwater fishes of Literature Cited American Fisheries Society, Bethes- Virginia. American Fisheries Society, Barbour, Michael T., Jeroen Gerritsen, da, Maryland. Bethesda, Maryland. Blaine D. Snyder, and James B. Cummins, K.W. 1962. An evaluation Jones, J. B., and L. A. Smock. 1991. Stribling. 1999. Rapid Bioassessment of some techniques for the collec- Transport and retention of particulate Protocols for Use in Streams and tion and analysis of benthic samples organic matter in two low-gradient Wadeable Rivers: Periphyton, Ben- with special emphasis on lotic wa- headwater streams. Journal of the thic Macroinvertebrates, and Fish. ters. American Midland Naturalist. North American Benthological Soci- Second Edition. EPA 841/B-99/002. 67:477-504. ety 10:115-126. Office of Water, Washington, D. C. Curry, K. D., and A. Spacie. 1984. Dif- Kilken, R. H. 1974. Artificial spawning Boschung, H.T., and R.L. Mayden. ferential use of stream habitat by and hatching techniques for Blacktail 2004. Fishes of Alabama. Smithson- spawning catostomids. American Redhorse. Progressive Fish-Culturist ian Books, Washington, D.C. Midland Naturalist 111:267-279. 36:174. Bowman, M. L. 1970. Life history of Dauble, D. D. 1986. Life history and LeCren, E. D. 1947. The determination

32 Kentucky Department of Fish & Wildlife Resources Wildlife Diversity and Fisheries / COMPLETED PROJECTS

of the age and growth of the perch sippi, Jackson, Mississippi. from the Cahaba River, Alabama. (Perca fluviatilis) from the opercular Ross, S. T., M. T. O’Connell, D. M. Proceedings of the Annual Confer- bone. Jour. Animal Ecology 16:188- Patrick, C. A. Latorre, W. T. Slack, J. ence Southeastern Association of 204. G. Knight, and S. D. Wilkins. 2001. Game and Fish Commissioners McConnell, W.J. 1951. The opercular Stream erosion and densities of Ethe- 23(1969):255-261. bone as indicator of age and growth ostoma rubrum (Percidae) and asso- Timmons, T.J., and J. S. Ramsey. 1983. of the carp, Cyprinus carpio Lin- ciated riffle-inhabiting fishes: biotic Life history and habitat of the Black- naeus. Transactions of the American stability in a variable habitat. Copeia fin Sucker,Moxostoma atripinne Fisheries Society 81:138-149. Meyer, 2001:916-927. (Osteichthyes: Catostomidae). Co- W. H. 1962. Life history of three spe- Rupprecht, R.J., and L.A. Jahn. 1980. peia 2:538-541. cies of redhorse (Moxostoma) in the Biological notes on blue suckers in Weisel, G. F. 1962. Comparative study Des Moines River, Iowa. Transac- the Mississippi River. Transactions of the digestive tract of a sucker, tions of the American Fisheries Soci- of the American Fisheries Society Catostomus catostomus, and a preda- ety 91:412-419. 109:323-326. cious minnow, Ptychocheilus orego- Moss, R. E., F. Harders, and W.H. Schreck, C. B., and P. B. Moyle, edi- nense. American Midland Naturalist Tucker. 1983. Observations of the tors. 1990. Methods for fish biology. 68:334-346. natural history of the blue sucker American Fisheries Society, Bethes- Wener, R. G. 1979. Homing mechanism (Cycleptus elongatus Lesueur) in the da, Maryland. of spawning White Suckers in Wolf Neosho River. American Midland Schumm, S. A., M. D. Harvey, and C. Lake, New York. New York Fish and Naturalist 109:15-22. C. Watson. 1984. Incised channels: Game Journal 26:48-58. Moyle, P. B., and A. Marciochi. 1975. morphology, dynamics and control. White, D. S., and K. M. Haag. 1977. Biology of the Modoc Sucker, Catos- Water Resources Publications. Little- Food and feeding habits of the Spot- tomus microps, in northeastern Cali- ton, Colorado. ted Sucker, Minytrema melanops fornia. Copeia 1975:556-560. Scoppettone, G. G. 1988. Growth and (Rafinesque). American Midland NatureServe. 2004. NatureServe Ex- longevity of the cui-ui and longevity Naturalist 98:137-146. plorer: An online encyclopedia of life of other catostomids and cyprinids in Wilson, M. V. H. 1980. Oldest known [web application]. Version 4.7. western North America. Transactions Esox (Pisces: Esocidae), part of a NatureServe, Arlington, Virginia. Page, of the American Fisheries Society new Paleocene teleost fauna from L. M., and C. E. Johnston. 1990. 117:301-307. western Canada. Canadian Journal of Spawning in the creek chubsucker, Scoppettone, G. G., M. Coleman, and Earth Sciences 17:307-312. Windell, Erimyzon oblongus, with a review of G. A. Wedemeyer. 1986. Life history J. T. 1971. Food analysis and rate of spawning behavior in suckers (Catos- and status of the endangered cui-ui of digestion, p. 215-226. In: Methods tomidae). Environmental Biology of Pyramid Lake, Nevada. U.S. Fish and for assessment of fish production in Fishes 27:265-272. Wildlife Service, Fish and Wildlife fresh waters. W. E. Ricker (ed.). IBP Pearson, M. P., and M. C. Healey. 2003. Research 1. Handbook No. 3, 2nd edition. Black- Life-history characteristics of the en- Siebert, D. J. 1984. Cyprinoid interrela- well Scientific Publications, Oxford. dangered Salish Sucker (Catostomus tionships: gill arch structure and vas- sp.) and their implications for man- cularization. Page 192 in Abstracts Funding Source: State and Tribal agement. Copeia 2003:759-768. of the 64th Annual Meeting of the Wildlife Grant (SWG) Peterson, M.S., L. C. Nicholson, D. J. American Society of Ichthyologists Snyder, G. L. Fulling. 1999. Growth, and Herpetologists. KDFWR Strategic Plan. Goal 1. spawning preparedness and diet of Smock, L. A., G. M. Metzler, and J. Strategic Objective 5. Comprehen- Cycleptus meridionalis 80 (Catos- E. Gladden. 1989. Role of debris sive Wildlife Conservation Strategy: tomidae). Transactions of the Ameri- dams in the structure and function- Appendix 3.9; Class Actinopterygii can Fisheries Society 128:900-908. ing of low-gradient streams. Ecology and Cephalaspidomorphi: Taxa spe- Power, M. E., D. Tilman, J. A. Estes, B. 70:764-775. 81 cific conservation projects. A. Menge, W. J. Bond, L. S. Mills. Summerfelt, R. C., and G. E. Hall, edi- 1996. Challenges in the quest for tors. 1987. Age and Growth of Fish. keystones. BioScience 46:609-620. Iowa State University Press, Ames, Ross, S. T., W. M. Brenneman, W. T. Iowa. Slack, M. T. O’Connell, and T. L. Tatum, W. M., and P. A. Hackney. Peterson. 2001. Inland Fishes of Mis- 1970. Age and growth of river red- sissippi. University Press of Missis- horse, Moxostoma carinatum (Cope)

Annual Research Highlights 2008 COMPLETED PROJECTS / Wildlife Diversity and Fisheries

Effects of Orientation and Weatherproofing on the Detection of Echolocation Calls in the Eastern United States

Eric Britzke, U.S. Army Corps of Engineers; Brooke Slack, Kentucky Department of Fish and Wildlife Resources; Susan Loeb, U.S. Forest Service; Mike Armstrong, U.S. Fish and Wildlife Service

Introduction Acoustic monitoring of bat echolocation calls has allowed us to greatly expand our knowledge of bat ecology. Ultrasonic detectors permit non-obtru- sive sampling of the bat community and can be used to sample bats in habitats that can not be sampled using tradition- Deployed Anabat units / Brooke Slack al capture techniques (e.g., open fields, large rivers). Additionally, bat detectors the tube (O’Farrell 1998). The second ing endangered species. For example, can detect more species at a site than system is comprised of the microphone in Kentucky, the U.S. Fish & Wildlife capture techniques, although the use enclosed in a PVC tube pointed down Service Kentucky Field Office and of both techniques maximizes detec- at an acrylic-glass plate used for call the Kentucky Department of Fish & tion of all species (Murray et al. 1999, deflection (BatHat; Arnett et al. 2006). Wildlife Resources require Anabat II O’Farrell and Gannon 1999). Further, We are not aware of any published data systems in survey efforts for the feder- some ultrasonic detectors can be de- that has tested the effectiveness of the ally endangered Indiana bat (Myotis ployed to passively record the echolo- 2 weatherproofing options, despite the sodalis). Thus, it is critical to know the cation calls of bats without an observer importance of the potential impacts on effects of weatherproofing and detector present, thereby allowing a small crew the results and subsequent interpreta- orientation on the number and quality to sample multiple sites simultaneously tion of the data. of bat echolocation calls. The objective and for long periods of time (e.g., Gor- The ability to detect echolocation of this study was to determine how de- resen et al. 2008). calls of bats can also be affected by tector orientation and weatherproofing The Anabat II (Titley Electronics, the orientation angle of the detector affect the quantity and quality of bat Ballina, New South Wales, Australia) (Weller and Zabel 2002). Depending calls recorded. is an ultrasonic detection system that on the height of the detector, detector is widely used for the study of bats. orientation can significantly affect the Methods While the system allows for long number of bat detections. Although We used Anabat II detectors con- periods of automated recording, the the orientation is relatively fixed at 45° nected to a CF storage ZCAIM as well equipment is susceptible to damage from horizontal for the 2 weatherproof- as the SD1 units (Titley Scientific; from rain or high humidity. To protect ing methods, without weatherproofing www.titley.com.au). Before sampling, the equipment, researchers have devel- the researcher can use any orientation systems were calibrated following oped 2 types of weatherproofing. The between 0° and 90°. methods used by Larson and Hayes first protective measure was placement While the Anabat II system has (2000). At each sampling site, 5 An- of a detector in a polyvinyl chloride the potential to collect vast amounts abat II systems were deployed side by (PVC) tube in a waterproof box where of data, it is unknown to what extent side on tripods at 1.5 m. Each detector the echolocation calls of bats enter the detector orientation or weatherproof- was randomly assigned to an orienta- PVC tube and are transferred to the ing affects detectability of bat calls. tion or weatherproofing treatment. The microphone, while the water drains out Acoustic sampling is often used to sur- order of the treatments at each site was through small holes in the bottom of vey areas for species presence, includ- randomly determined. The detector

34 Kentucky Department of Fish & Wildlife Resources Wildlife Diversity and Fisheries / COMPLETED PROJECTS

orientations were 0° (horizontal), 45°, (F = 15.37; P < 0.001) with the BatHat orientation and weatherproofing. The and 90°(vertical) and the 2 weather- recording calls of lower quality than the total number of call sequences or files proofing types were the PVC tube and other treatments (Fig. 2c). is often used as a measure of overall the BatHat (EME Systems Inc., www. A total of 6 species were detected: activity (e.g., Hayes 1997). The num- emesystems.com). big brown bats (Eptesicus fuscus), red ber of pulses per sequence provides Detectors were set up before dark bats (Lasiurus borealis), hoary bats (L. a measure of the intensity of activity and calls were recorded for the entire cinereus), Indiana bats, little brown (Gorrensen et al. 2008). The percent- night. The following morning, units bats (M. lucifugus), and tri-colored age of calls surviving the identification were picked up and data were uploaded bats (Perimyotis subflavus). Red bats filter is a measure of the quality of the to a laptop computer using the CFCread and big browns were found at the most recordings. Factors affecting the detec- program (www.hoarybat.com). Data sites. Species richness varied signifi- tion, quality, or length of sequences from each unit were scanned with a cantly among treatments (P = 0.003) all have large impacts on the results of customized filter in Analook Version and the number of species recorded studies using Anabat detectors or other 4.9j to delete extraneous noise (Britzke with the BatHat was approximately acoustic sampling equipment. The 2003). We used the scanFiles option in ½ that of the other detectors (Fig. 3). consistent pattern of differences across Analook to determine the number of Detectors oriented in the horizontal all parameters suggests that the differ- files and the Countscan option to calcu- position also recorded fewer species ences among treatments are indeed real. late the total number of pulses for each than the PVC and 45°and 90° orienta- However, the only significant differenc- sequence. We then used a customized tions (Fig. 3). In general, the pattern es occurred between the BatHat and the filter (modified from Britzke and Mur- of higher detections by the PVC and other methods. Thus, use of the BatHat ray 2000) to extract parameters for spe- 45°and 90° orientations held for each should be avoided in future studies. cies identification. The parameters were of the species although northern long- Weller and Zabel (2002) found no identified using a mixed DFA model eared bats were only picked up by the difference between detectors oriented in the statistical program R (v. 2.2.1; vertical detectors and Indiana bats were at 30° or 45° if they were on 1.4 m http://www.r-project.org; Duchamp not as readily detected by the detectors high stands. However, we found that 2006). The total number of files, aver- with PVC or BatHats (Fig. 4). although not significant, the horizontal age number of pulses, and the percent- deployment tended to record lower age surviving the identification filter Discussion activity levels (number of files) and were compared among 5 weatherproof- Significant differences among the species richness than the other orienta- ing and orientation treatments using a treatment groups suggest that weather- tions. However, the appropriate orien- randomized block ANOVA. Treatments proofing and detector orientation may tation may depend on the question of were then compared using a Tukey test. have important impacts on the detec- interest. If a researcher is interested Average species richness was compared tion of the echolocation calls of bats. In in recording bats that forage near the using a Median test. particular, our data suggest that studies water surface such as gray bats (M. employing the BatHat system may de- grisescens), horizontal deployment is Results tect lower activity and species richness typically better than other orientations. A total of 17 sites were surveyed in than are present at a site. If researchers Therefore, researchers should try to use which sampling equipment functioned are simply interested in relative activity the orientation that maximizes detec- properly for all 5 treatments. The mean levels among sites, any weatherproof- tions of the species of interest. number of files recorded per night var- ing or orientation is acceptable as long A total of 6 species were detected ied by treatment (F = 4.02; P = 0.006). as detectors are deployed in a similar across the 5 treatments, but no treat- The mean number of files recorded per way among all sites. However, prob- ment detected all 6 species. The PVC, night by the BatHat was significantly lems may arise if researchers want 45°, and vertical (90°) treatments con- lower than all other treatments; the to conduct species identification or sistently recorded almost twice as many units with the PVC protection recorded compare their results to studies where species as the BatHat. Further, the the highest number of files per night another weatherproofing design or number of species recorded by detec- (Fig. 2a). The average number of pulses orientation was used. Future studies us- tors in the horizontal position tended to per file varied among treatment F( = ing passive sampling with the Anabat be lower than the detectors at 45°and 8.02; p < 0.001) and the BatHat had system should include discussion of the 90° even though the number of calls fewer pulses per file than all other treat- method of deployment and the potential recorded by detectors in the horizontal ments (Fig. 2b). The percentage of files impacts of those methods on the results. orientation was just slightly below the that passed through the ID filter were We measured a variety of param- number of calls recorded by detectors significantly different among treatments eters in this study to test the effects of in the vertical (90°) position, Thus, if

Annual Research Highlights 2008 COMPLETED PROJECTS / Wildlife Diversity and Fisheries

the BatHat or horizontal orientations obtained from these studies. Knowl- Duchamp, J. E., M. Yates, R.-M. Muzi- are used, researchers should deploy edge of differences among orientations ka, and R. K. Swihart. 2006. Estimat- additional equipment to gain a more ac- and weatherproofing designs can assist ing probabilities of detection for bat curate representation of species present management efforts by maximizing the echolocation calls: an application of (Duchamp et al. 2006). quality of data obtained through stud- the double-observer method. Wildlife When conducting surveys for ies involving acoustic surveys, thereby Society Bulletin 34:408–412. bats, such as the Indiana bat, multiple improving the impacts of management Gorresen, P.M., A. C. Miles, C. M. sampling sites may be required due to activities on bats. Todd, F. J. Bonaccorso, and T. J. project size. Many detectors are often Weller. 2008. Assessing bat detect- set throughout the project area to re- Acknowledgements ability and occupancy with multiple cord simultaneously. Similarly, studies This study was supported by the automated echolocation detectors. designed to test the effects of habitat Kentucky Department of Fish and Journal of Mammalogy 89:11-17. type or management activities on bat Wildlife Resources and the U.S. Fish Hayes, J. P. 1997. Temporal variation habitat use and activity often set detec- and Wildlife Service, Kentucky Ecolog- in activity of bats and the design of tors simultaneously in each habitat type ical Service Field Office. We thank Aric echolocation-monitoring studies. or treatment to control for the effects Payne and Corey Knight who helped Journal of Mammalogy 78:514-524. of temporal variation on activity (e.g., with data collection and the numerous Larson D.J., and J.P. Hayes. 2000. Vari- Loeb and Waldrop 2008). These situa- landowners that provided access to ability in sensitivity of Anabat II bat tions require the use of weatherproofing their lands. detectors and a method of calibration. equipment because detectors are widely Acta Chiropterologica 2:209-213. scattered across the landscape and sud- Literature Cited Loeb, S.C., and T. A. Waldrop. 2008. den storms are possible. Our results sug- Arnett, E. B., J. P. Hayes, and M. M. P. Bat activity in relation to fire and fire gest that detections obtained from detec- Huso. 2006. An evaluation of the use surrogate treatments in southern pine tors at two common orientations without of acoustic monitoring to predict bat stands. Forest Ecology and Manage- weatherproofing (45° and 90°) are simi- fatality at a proposed wind facility in ment 255: 3185-3192. lar to those from detectors with the PVC south-central Pennsylvania. An annual Murray, K. L., E. R. Britzke, B. M. weatherproofing. Thus, data can be eas- report submitted to the Bats and Wind Hadley, and L. W. Robbins. 1999. ily compared across studies that use any Energy Cooperative. Bat Conservation Surveying bat communities: a com- one of these methods. Our results also International, Austin, Texas. parison between mist nets and the suggest that if weatherproofing is used, Britzke, E. R. 2003. Use of ultrasonic Anabat II bat detector system. Acta that the PVC method be used instead of detectors for acoustic identification Chiropterologica 1:105-112. the BatHat. The detection cone for the and study of bat ecology in the east- O’Farrell, M. J. 1998. A passive moni- BatHat seems to be more directional ern United States. Ph.D. dissertation, toring system for Anabat II using a than the PVC weatherproofing and de- Tennessee Technological University, laptop computer. Bat Research News tectors without weatherproofing (person- Cookeville, TN. 39:147-150. al observation). Thus the performance Britzke, E. R., and K. L. Murray. 2000. O’Farrell, M. J., and W. L. Gannon. of the BatHats may be improved if the A quantitative method for selection 1999. A comparison of acoustic ver- detectors are deployed in the direction of identifiable search-phase calls us- sus capture techniques for the inven- of the most bat activity (e.g., parallel to ing the Anabat system. Bat Research tory of bats. Journal of Mammalogy stream flow). Whatever weatherproof- News 41:33-36. 80:24-30. ing method that is used (none, PVC, or Britzke, E. R., K. L. Murray, J. S. Weller, T. J., and C. J. Zabel. 2002. BatHat), the implications of that choice Heywood, and L. W. Robbins. 2002. Variation in bat detections due to de- on the results should be considered. Acoustic identification. Pp. 220-224 tector orientation in a forest. Wildlife in The Indiana bat: Biology and Society Bulletin 30:922-930. Management Implications Management of an Endangered Spe- Ultrasonic detectors are commonly cies (A. Kurta and J. Kennedy, eds.). Funding Source: State and Tribal used to study bat ecology. Results from Bat Conservational, Austin, TX. Wildlife Grant (SWG) these studies are often used to infer Duchamp, J. E. 2006. Modeling habitat use of bats and management de- bat community structure and spe- KDFWR Strategic Plan. Goal 1. cisions are then based on these conclu- cies distribution across fragmented Strategic Objective 5. Comprehen- sions. Our results suggest that micro- landscapes within the upper Wabash sive Wildlife Conservation Strategy: phone orientation and the presence of River basin. Dissertation, Purdue Appendix 3.9; Class Mammalia: weatherproofing may impact the results University, West Lafayette, IN. Taxa specific conservation project.

36 Kentucky Department of Fish & Wildlife Resources Wildlife Diversity and Fisheries / COMPLETED PROJECTS

Identifying and Protecting Hibernation Roosts for Endangered Bats in Kentucky

Jim Kennedy, Bat Conservation colonies and allow bats to re-establish Funding for this research was gener- historic roosts. There are numerous International ously provided through a State and examples of bat colonies increasing Tribal Wildlife Grant administered by KDFWR Contact: Brooke Slack when management is improved, and of the Kentucky Department of Fish and bats re-colonizing formerly abandoned Wildlife Resources (KDFWR). Introduction roosts (Humphrey 1978, Tuttle and Bat Conservation International Kennedy 2002). Given the potential for Methods and Materials (BCI) has long been a proponent for undiscovered, overlooked, and current- Objective 1: Identify and assess previ- habitat protection, especially of caves ly altered caves to provide increased ously overlooked hibernation roosts that are important to hibernating bats hibernacula and roosting sites for bats and develop management and restora- such as the critically Endangered In- of 11 species in KY, BCI conducted tion recommendations for the highest diana myotis. However, many caves this study in order to benefit natural priority sites. historically used by bats are no longer resource managers in Kentucky and the To accomplish our research goals, available due to poor management, bats and habitats that they manage. In we started with a pool of all caves in the including unrestricted visitation during order to explicitly address “Prioritized area, from which we derived a target list critical periods, physical alterations Research Projects #1 and #5 and Sur- of caves most likely to be hibernacula, from guano or saltpeter mining, com- vey Project #1” as found in Appendix based on the cave maps and/or descrip- mercialization for show caves, and 3.2, Class Mammalia of the Kentucky tions. Our selection criteria included restriction or closure of entrances due Wildlife Action Plan (Kentucky’s length and size of the cave; configura- to erosion, trash dumping, and purpose- Comprehensive Wildlife Conservation tion of passages and entrance(s); prox- ful filling to prevent entry (Humphrey Strategy 2005), BCI undertook the imity to other known M. sodalis sites; 1978). These caves are not currently project detailed herein under the leader- bats, guano, roost stains, or saltpeter considered “bat caves” and are not ship of Cave Resources Specialist and mining indicated on the map or descrip- managed as such, but if even a small Principle Investigator Jim Kennedy. tion; and input from other knowledge- number are identified and properly managed (including microclimate restoration, if necessary), it will increase the number of available roosts dramatically, allowing populations to increase (Tuttle and Kennedy 2002). One of our strategies in this regard is to locate these “overlooked” bat caves, prioritize their historic importance, and work with landowners and management agencies to correct problems and improve management to better protect current Indiana Bat cluster / Traci Hemberger

Annual Research Highlights 2008 COMPLETED PROJECTS / Wildlife Diversity and Fisheries

able cavers in the area. Caves were courtesy of Louisville Grotto (Kentucky tunistically during field crew searches then field-checked for suitability and to Speleofest guidebooks), Kentucky Spe- for target caves. The names of most of record important baseline data. Owner- leological Survey, and American Cave these opportunistic caves are currently ship was determined when possible, and Conservation Association. unknown, but none were significant bat permission secured. Contact informa- sites. Investigations were attempted at tion for thelandowner or their agent was Objective 2: Train cavers, biologists three additional caves, but permission obtained. At each site the field crew met and land managers in hibernation was denied for one, another was never with landowners to secure permission, cave recognition and documenta- located despite several attempts, and and provide initial education about the tion techniques, identifying threats to the third was not entered due to reports purpose of the study and the impor- roosts and developing solutions. from the owner of past flooding. During tance of hibernation caves. Location BCI used a formal presentation to the 2007–2008 field season, 5 new sites information was recorded for each site a large group of cavers and landowners were documented with Rafinesque’s visited, including time-averaged GPS and provided over 200 copies of the big-eared bat (Corynorhinus rafines- data. Information about each entrance Field Guide to Eastern Cave Bats to quii), a state Species of Concern. One, was documented, including size, shape, current and future natural resources and in Wayne County, is a significant winter aspect, vegetative cover, recent altera- land managers. More intensive cave roost for the species, with 195 bats tions (enlargements or restrictions), assessment training was provided by reported. Three new gray bat (Myotis amount of human disturbance, and any BCI to more than 15 other individuals grisescens) roosts were discovered, other important factors. Inside the cave who were then dispatched to the field to most likely summer bachelor colonies. the field crew documented locations, practice these skills (see Objective 1). Even more exciting was the discovery numbers, and species of any T&E or of a significant (Priority 2) Indiana bat other bat species, as well as locations Objective 3: Continue restoration work (Myotis sodalis) hibernacula in Wayne and size of bat sign, including roost on Saltpeter Cave in Carter County, a County. stains, guano deposits, and accumula- high priority site. tions of bat bones. Past bat populations BCI continued step-wise restora- Objective 2: Train cavers, biologists were estimated based on measurements tion efforts at Saltpeter Cave, Carter and land managers in hibernation of wall and ceiling staining left by County, KY through initiating efforts cave recognition and documenta- roosting bats, and/or guano deposits. to re-open a sinkhole above the north- tion techniques, identifying threats to Passage dimensions and environmental west passage of the cave. Using results roosts and developing solutions. parameters as temperature, humid- of a geo-probing study conducted last During the grant period we were ity, airflow, and presence of streams fall, BCI coordinated planning and as- able to provide a one-hour formal or pools were recorded. All roost sites sessment of the proposed project and presentation on bat cave assessment were photo-documented, as were other sought and received permission from and protection to more than 200 cav- important elements such as signs of the USFWS Kentucky Field Office to ers and landowners during the 2008 vandalism and other (non-bat) biologi- perform the restoration acts of remov- Kentucky Speleofest. A key component cal observations. No bats were handled ing recent fill material, installing a of that training was the distribution during these assessments. All work was culvert, replacing clean fill to hold the of more than 200 copies of the Field accomplished by Jim Kennedy, Cave culvert in place, and installing an ad- Guide to Eastern Cave Bats. We also Resources Specialist at BCI, or by cav- justable gate to modify airflow as may provided informal training to multiple ers and volunteers trained by him. A be necessary. landowners,volunteer cavers, and mem- large portion of the work was performed bers of our own field crews. by John Chenger and his team from Bat Results Conservation and Management (www. Objective 1: Identify and assess previ- Objective 3: Continue restoration work batmanagement.com). Chris Clark from ously overlooked hibernation roosts on Saltpeter Cave in Carter County, a the American Cave Conservation As- and develop management and restora- high priority site. sociation was instrumental in providing tion recommendations for the highest At Saltpeter Cave, in Carter Caves landowner contact information and as- priority sites. State Resort Park, BCI has been mak- sisting with data recording in the field. Forty caves were visited during ing stepwise efforts in restoring altera- Other personnel participating in this the field investigations. Thirty-four of tions to the cave’s microclimate caused study were Thor Bahrman, Kristi Lind- these were identified on our target list, by historic saltpeter mining and more berg, Kevin Rhome, Kyle Ryan, Anna approximately one third of all the caves recent commercialization. This past Scesny, Rose Sisler, and Bill Walden. on the list for central and eastern Ken- year we have been working with the Maps and other cave data were provided tucky. Five caves were added oppor- Environmental Services Division of

38 Kentucky Department of Fish & Wildlife Resources Wildlife Diversity and Fisheries / COMPLETED PROJECTS

Hinkle Engineering in Lexington, KY of the cave dictated that those areas requires physical modification. Such to develop a plan for excavating a filled should be the coldest in the winter. actions may involve a variety of actions sinkhole over the northwestern branch There are many dozens of caves still to including entrance restoration (enlarge- of the cave. This sink was formerly be checked that may be potential bat ment or restriction), air dams (in-cave open to the surface, providing moister caves, or have been bat caves histori- or at the surface), reopening closed air to that passage in the summer and cally. It is obvious that our investiga- entrances, closing artificial entrances, colder temperatures in the winter. Ex- tive approach – that of selecting target removal of obstructions (buildings over tensive roost staining indicates past use caves based on maps, descriptions, and entrances, walls, poorly-designed gates, by at least 10s of thousands of bats, recommendations from knowledgeable trash, rock, etc.). During the time frame but the passage is currently too dry and cavers – and then field checking those of this project, BCI has assessed the unusable in its current condition. Later caves, will bring to light more impor- past bat use and restoration potential this summer Hinkle Enginerring and tant cave habitats that require better of less than 1% of the known caves in Bat Conservation International will re- management. For most caves identified Kentucky. These data resulting from move recent fill and construction debris as important bat hibernacula, the next these efforts can now be used by state from the surface, hand-excavate the logical steps are to secure future man- and federal agencies to help guide fu- remaining fill connection into the cave, agement through a Cave Management ture efforts to increase the amount of install a culvert to prevent soil loss into Plan, Conservation Easement, or out- available habitat for hibernating bats the cave, backfill and grade around the right purchase, and restore altered mi- in Kentucky through site protection culvert, and secure access to the newly- croclimates when necessary or control and restoration work, ultimately help- opened entrance with an adjustable gate winter access through proper cave gates ing populations of endangered Indiana to modify airflow. (Kennedy 2004, 2006). Data collected myotis to recover. in this study were analyzed to identify Discussion and Management sites needing resto- Recommendations ration or increased Objective 1: Identify and assess previ- protection efforts. In ously overlooked hibernation roosts cases where caves and develop management and restora- still possess suitable tion recommendations for the highest temperatures but are priority sites. not currently used by Many of the best M. sodalis and bats, we recommend C. rafinesquiihibernation caves visited halting visitation of during this study occurred in or near the these caves during Eastern Coalfields province, a part of winter months to the Cumberland Plateau and Mountains encourage re-coloni- physiographic region. This is unsurpris- zation. Bat-friendly, ing, since relief is greater and caves can zero-air-flow-modifi- be deeper, potentially trapping more cation gates may be cold air. Counties with the greatest po- recommended at sites tential for discovering new M. sodalis with a strong history sites appear to be Clinton, Wayne, Pu- of local visitation and laski, Rockcastle, and possibly Jackson where simple signage and Estill. The more westward caves would be ineffective. visited appear to have greater use by Often, caves that M. grisescens. These caves are located have been known as at lower elevations in the Pennyroyal historical bat caves region of the Mississippian Plateaus have been modified physiographic province. These include by humans such that caves in Breckenridge, Hardin, Gray- they become unsuit- son, Hart, Warren, Barren, Metcalf, and able for hibernation. Adair counties. Several caves that were In these cases resto- visited were incompletely assessed ration to conditions due to the initial passages (entrance favorable for bat area) being too warm, when the layout re-colonization often Cave gate / Traci Hemberger

Annual Research Highlights 2008 9 Bobcat / Adam Jones Project Highlights

40 Kentucky Department of Fish & Wildlife Resources Fishes / PROJECT HIGHLIGHTS

Analysis of the Environmental Requirements for Etheostoma cinereum and Percina squamata in the Rockcastle River

Michael C. Compton and Christopher M. Taylor, Texas Tech University KDFWR Contact: Ryan Oster

he integrity of rivers and the per- Tsistence of aquatic life are under constant pressure from agricultural practices, urban sprawl, road development, deforestation, and mining activities. The effects of these disturbances upon the landscape have a direct and indirect impact on the aquatic biota and their environment. The Rockcastle River is no exception to these threats and is of immense concern given that it has an exceptionally high aquatic biodiversity and contains numerous unique species to Kentucky. Two species, the ashy darter and the olive darter, are of Rockcastle River / Michael Compton particular interest given their presence within the Rockcastle River watershed of the two species within 30 stream from two of the five stream reaches and their overall rarity in the common- reaches of the Rockcastle River. Data sampled. Currently no associations be- wealth. will be collected during the summer tween darter presence and habitat have Historically the ashy and olive months of 2008-2010 to determine been made. The importance of the data darters inhabited numerous stream what stream reaches within the Rock- is to ultimately identify and model the systems within the Cumberland and castle river watershed are inhabited by environmental conditions and habitat Tennessee River drainages among six the species, and within those stream preferences of the species, which will southeastern states, but their distribu- reaches, what microhabitats are used by provide KDFWR the needed information has become fragmented over time the species. In addition, fish community tion to ensure the species existence and their populations have declined. data will be collected to determine the within the Rockcastle River but also to Although various aspects of life his- overall health of the watershed and to enhance conservation efforts in other tory are known for both species, many determine if any species association ex- watersheds that contain or historically aspects are not fully understood, such ist between the ashy and olive darters have contained ashy and olive darters. as habitat preferences, tolerance to and any other species within the river. impacts, or a current conservation sta- Collecting efforts in 2008 yielded Funding Source: State and Tribal tus. In Kentucky, the Rockcastle River no ashy darter or olive darter indi- Wildlife Grants (SWG) contains the best populations of the viduals from eleven 3rd order reaches two species; therefore, the watershed surveyed within the Rockcastle water- KDFWR Strategic Plan. Goal 1. provides an excellent setting to model shed. However, three stream reaches Strategic Objective 5. Comprehen- habitat preferences and environmental within the 4th and 5th order portions of sive Wildlife Conservation Strategy: conditions for the target species. Mul- the Rockcastle River yielded nearly 30 Appendix 3.9; Class Actinopterygii tiple regression models will be devel- ashy darter individuals, in addition, 3 and Cephalaspidomorphi: Taxa spe- oped based on presence-absence data Olive darter individuals were collected cific research project #1.

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Fishes

Captive Propagation and Reintroduction of Cumberland Darter and Kentucky Arrow Darter in Southeastern Kentucky

Conservation Fisheries, Inc., in review to federally list the Cumber- for both species in a controlled setting Knoxville, TN land Darter as endangered, because of to produce offspring that can be used KDFWR Contact: Ryan Oster recent range curtailment and fragmen- to re-establish extirpated populations. tation resulting from habitat degrada- Brood stock for each species were col- tion. The Cumberland Darter is current- lected during winter 2008, and are cur- aptive propagation and reintroduc- ly limited to six streams in the Cumber- rently being held at the CFI facility in Ction are considered appropriate land River drainage above Cumberland Knoxville, where captive breeding will tools for the recovery and delisting of Falls. A recent status assessment of the occur in spring 2009. Methods for egg threatened and endangered fishes and Kentucky Arrow Darter in the upper and larval collections and rearing will to prevent declining species from being Kentucky River basin has shown that be applied and developed specifically added to the federal List of Endangered populations have declined considerably for each species, based on taxonomic or Threatened Wildlife. The Cumber- during the past two decades, having relationships and prior experience with land Darter (Etheostoma susanae) and become extirpated from many streams other species at CFI. Kentucky Arrow Darter (Etheostoma where they were known to occur his- Offspring produced in captivity sagitta spilotum) are two small native torically. Conservation Fisheries, Inc. will be released into carefully chosen fishes with naturally limited ranges in (CFI), in cooperation with KDFWR has streams where suitable habitat exists Kentucky. A proposed rule is currently begun developing spawning protocols and that lie within the historic range of each species. Reintroduction is anticipated either in late summer or fall 2009 or in 2010, depending on the number of offspring produced and rate of growth. Following reintroduction, survivability and movement patterns will be assessed through mark-recapture methods and through periodic monitoring using non-invasive methods, including visual census techniques such as snorkeling. Depending on age and time of release, stocked fish may be tagged with Visible Elastomer Implant (VIE) fluorescent marks. Use of different tag positions and colors permits future determination of age and potential dispersal information for recaptures. Kentucky Arrow Darter / J.R. Shute Funding Source: State and Tribal Wildlife Grants (SWG)

KDFWR Strategic Plan. Goal 1. Strategic Objective 5. Comprehen- sive Wildlife Conservation Strategy: Appendix 3.2; Class Actinopterygii and Cephalaspidomorphi: Taxa specific research project #8.

Cumberland Darter / Matt Thomas

42 Kentucky Department of Fish & Wildlife Resources Fishes / PROJECT HIGHLIGHTS

A Survey of Fishes of Rock Creek, Kentucky, with Emphasis on the Impact of Stocking Rainbow Trout on Native Fishes

Stephanie Brandt and Sherry and SGCN and whether they may be years in the upper section by other biol- Harrel, Department of influencing the distribution of SGCN in ogists. Sawfin shiners appear to be more Biological Sciences, Eastern Rock Creek are unknown. common in the upper portion of Rock The purpose of this study was to Creek and were not previously taken Kentucky University, Matt document present distributions, relative from below Hemlock Grove showing Thomas, Kentucky Department abundance and habitat use of fishes in a downstream range extension of dis- of Fish and Wildlife Resources Rock Creek, with emphasis on SGCN. tribution. Emerald darters appear to be In addition, we determined distribution uncommon with only seven individu- and habitat use of stocked rainbow trout als found in the lower section of Rock ock Creek, located in McCreary in relation to SGCN and examined their Creek. Rainbow trout, sawfin shiners RCounty, Kentucky, is a tributary gut contents for the presence/absence and emerald darters were collected to- to the Big South Fork Cumberland of SGCN. Sites on Rock Creek were gether at two sites. Gut contents of two River. The stream spans twenty-one sampled on a seasonal basis from sum- trout specimens included the remains of miles, with its origin in Pickett County, mer 2008 to spring 2009. All available telescope shiners and barcheek darters, Tennessee. The lower section from White Oak Creek to its confluence with the Big South Fork has been impacted by acid mine drainage, but reclamation efforts have been ongoing. Rock Creek supports a federally endangered fish species, blackside dace, as well as five species of greatest conservation need (SGCN) in Kentucky: sawfin shiner, emerald darter, ashy darter, Etheostoma sanguifluum / Matt Thomas bloodfin darter, and mountain brook lamprey. Since the early 1960’s, habitats at each site were thoroughly as well as aquatic insects. Preliminary the Kentucky Department of Fish and sampled using a combination of seining observations suggest that fish species Wildlife Resources has stocked Rock and backpack electrofishing. Any trout richness and abundance are comparable Creek annually, at multiple sites, be- collected were retained for gut contents between the unimpacted and the re- ginning at the junction of White Oak analysis. Presence of SGCN in gut con- claimed reaches of Rock Creek. Initial Creek upstream to the Kentucky/Ten- tents of trout would be interpreted as sampling has found no direct predation nessee state line with catchable-size (9- direct evidence of predation as a factor by trout on SGCN, but overlap in spe- 10 inch TL) rainbow trout (Oncorhyn- impacting these species. cies distributions is cause for concern. chus mykiss). Stockings occur once per Data analysis is ongoing, and month from March-June and from Sep- thirty-five fish species have been col- Funding Source: State and Tribal tember-December. Habitat destruction lected throughout Rock Creek samples Wildlife Grants (SWG) as well as predation may be negatively thus far. Two SGCN, sawfin shiner impacting SGCN. Research has shown and emerald darter, were present, as KDFWR Strategic Plan. Goal 1. that predation by stocked rainbow trout well as blackside dace. The ashy darter Strategic Objective 5. Comprehen- can influence intrastream (habitat) and and bloodfin darter were not collected. sive Wildlife Conservation Strategy: interstream (geographic) distributions Although not collected in the present Appendix 3.9; Class Actinopterygii of small native fishes. However, inter- study, the mountain brook lamprey and Cephalaspidomorphi: Taxa spe- actions between stocked rainbow trout has been collected within the last three cific research project #1.

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Fishes

Status, Life History, and Phylogenetics of the Amblyopsid Cavefishes in Kentucky

is known from just a single cave in Pulas- ki County. With the assistance of several cavers and caving- affiliated groups, we have documented this species in four additional caves and collected numerous invertebrate samples that are being sent to taxonomic experts for identification. Moreover, preliminary genetic analy- Cavefish / Dante Fenolio ses indicate that these populations Benjamin M. Fitzpatrick and chthys subterraneus). Although these are distinct from others associated with Matthew L. Niemiller, University species have been known to science the Cumberland Plateau in Tennessee since the early 1840s, little is known and from populations in the Mammoth of Tennessee about the demography and persistence Cave region of central Kentucky. This KDFWR Contact: Ryan Oster of local populations and the system- research will provide KDFWR with atic relationships among species and important data regarding the status, early 1800 species of subter- among populations within species. distribution, life history, and genetics of Nranean organisms are known Here we investigate the status, distri- these species. In addition, data acquired in North America and many species bution, ecology, and threats to popula- on other cave fauna can also be used remain undescribed. However, little is tions of these cavefishes. In particular when making conservation and man- known about the distributions and lim- we are conducting surveys and status agement decisions. iting factors of many of these species. assessments for each species within Over 95% of subterranean species in the state including both searches of Funding Sources: State and Tribal North America are considered vulner- historic and new localities, while ob- Wildlife Grant (SWG), University of able or imperiled, mainly because of taining life history data and acquiring Tennessee habitat degradation and restricted geo- tissue samples for genetic analyses. graphic ranges. Unfortunately, data on We also are using molecular tech- KDFWR Strategic Plan. Goal 1. the distribution and status of cave-obli- niques to investigate cryptic diversity, Strategic Objective 5. Comprehen- gate species are incomplete or lacking particularly in Typhlichthys, where sive Wildlife Conservation Strategy: entirely, making conservation and man- preliminary data suggest the existence Appendix 3.9; Class Actinopterygii agement decisions difficult. Therefore, of an undescribed species unique to and Cephalaspidomorphi: Taxa spe- a need exists to document subterranean Kentucky. Finally, we are conducting cific research project #1. diversity and identify threats that im- surveys and collecting specimens of pinge upon the continued survival of invertebrate cave organisms to deter- these species. mine species distributions and commu- Three species of Amblyopsid nity associations. cavefishes occur in Kentucky: Spring Surveys over the past year have Cavefish (Forbesichthys agassizii), focused in caves of Pulaski and Wayne Northern Cavefish (Amblyopsis spe- Counties in the Upper Cumberland laea), and Southern Cavefish (Typhli- watershed where the Southern Cavefish

44 Kentucky Department of Fish & Wildlife Resources Fishes / PROJECT HIGHLIGHTS

Status Survey of the Northern Madtom, Noturus stigmosus, in the Lower Ohio River

of the Northern Madtom in the lower Ohio River; 2) To determine why the recorded abundance of the Northern Madtom has been so diminutive in the Ohio River; 3) Identify habitats vital to the persistence of this species; 4) Identify potential threats to existing populations of the Northern Madtom; and 5) Recommend any ad- Northern madtom / Matt Thomas ditional effort required to ensure conservation Donovan Henry and Lennie have been made on the lower Ohio of the species in Kentucky. In addition to field sampling, we also conduct Pitcher, Three Rivers River where it borders Kentucky. This population has likely been overlooked habitat analysis, and collect field ob- Environmental Assessments LLC for many years due to difficulty sam- servations to address these goals. The KDFWR Contact: Ryan Oster pling large, dynamic systems such as information resulting form this project the Ohio River. More recent surveys (on habitat use and resource require- he Northern Madtom is a small, have been conducted on eastern Ken- ments) will be useful for determining Tsecretive inhabitant of large creeks tucky populations, but the Ohio River needs for conserving or enhancing this and rivers where there are moderate to population, which is on the western population, and will likely benefit oth- swift flows, and clean sand and gravel periphery of the species range, has not er species of conservation need such substrates. This species is listed as been investigated. This population may as benthic dwelling invertebrates (e.g. threatened, endangered, or of special be at risk due to further habitat deg- unionid mussels, gastropods, macroin- concern in every state in which it oc- radation by the construction of a new verterates, etc.). curs, and is disappearing from the lock and dam near Olmsted, Illinois margins of its range. In Kentucky, the at river mile (RM) 964.4. To facilitate Funding Sources: State and Tribal Northern Madtom is listed as special future management implications, a Wildlife Grant (SWG), Three Rivers En- concern by KDFWR’s Comprehensive focused and thorough survey, geared vironmental Assessments LLC, Illinois Wildlife Conservation Strategy, has a specifically toward this benthic dwell- Endangered Species Protection Board global rank of G3 (Vulnerable), and is ing species, is needed to determine the considered a Species in Greatest Need conservation status of the Northern Comprehensive Wildlife Conserva- of Conservation (SGNC). This species Madtom in the lower Ohio River along tion Strategy: Appendix 3.9 Class is sporadic and uncommon throughout Kentucky shores. Actinopterygiii, Priority monitoring its distribution, and occurs sporadically To address this information need, needs by taxonomic class (p. 1). Es- in the upper Kentucky, and Big Sandy, we established a sampling protocol tablish protocols, schedules, and sites with isolated but apparently stable pop- within five sections of the lower for long-term population monitoring ulations in the Licking and Salt Rivers Ohio River between the upper end of to assess status and trends for prior- of Kentucky. Furthermore, this species Smithland Pool and the confluence ity species. appears to be extirpated from the upper of the Ohio and Mississippi Rivers. Green River in Kentucky. In the Ohio Beginnning in 2008, we began exten- River there were only three records of sively sampling these sites to address this species before 1997. Since 1997, the following objectives: 1) Determine four collections of Northern Madtoms the present distribution and abundance

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Mollusks

Advances in the Propagation of Rare and Endangered Mussel Species

endangered or rare species. We devel- design. We have experimented with Center for Mollusk oped several small aquaculture systems several species, including Simponsias Conservation, Kentucky for use in research and propagation of ambigua, Alasmidonta viridis, Lampsi- Department of Fish and Wildlife juvenile mussels. Starter systems (nurs- lis siliquoidea, L. cardium, L. abrupta Resources Monte A. McGregor, ery stage I) include air lift downwellers (endangered), Epioblasma triquetra, E. Adam C. Shepard, Fritz Vorisek, for early stage juveniles, mini-riffle capsaeformis (endangered), and Villosa systems (nursery stage II) for the next trabilis (endangered). Growth for most J. Jacob Culp, Jim Hinkle, several months. The final nursery (stage species has been up to 15-20mm in less Andrew Wooldridge, Vaughn III) consists of variable sized upwelling than 9 months, regardless of time of Kauffman tanks that may be recirculating or open year. The rare animals are surrogates for other federally endangered species he Center for Mollusk Conserva- such as the Littlewing Pearlymussel Ttion made several advances on the and the Northern Riffleshell. We have propagation of mussels in 2008. Last improved both survival and growth year, we were able to propagate 5 new rates of mussels in captivity.

Funding Source: Endangered Species Act (Section 6) funds, U.S. Fish and Wildlife Service

KDFWR Strategic Plan. Goal 1. Strategic Objective 5. Comprehen- sive Wildlife Conservation Strategy: Appendix 3.2, Class Bivalvia. Prior- ity Research Project #1.

Juvenile mussels (2-10 months old) / Monte McGregor

Juvenile mussels (2-10 months old) / Monte McGregor Upwelling growout tank / Monte McGregor

46 Kentucky Department of Fish & Wildlife Resources Mollusks / PROJECT HIGHLIGHTS

Successful Reintroduction of Two Endangered and Two Candidate Mussel Species to the Big South Fork Cumberland River, Kentucky

Center for Mollusk Conservation, Kentucky Department of Fish and Wildlife Resources Monte A. McGregor, Adam C. Shepard, Fritz Vorisek, J. Jacob Culp, Jim Hinkle, Andrew Wooldridge, Vaughn Kauffman

n June 2008, The Kentucky Depart- Iment of Fish and Wildlife collabo- rated with The National Park Service (Big South Fork), United States Fish and Wildlife Service, Tennessee Wild- life Resources Agency, and the United States Geological Survey to reintroduce two federally endangered mussels and two mussels that are candidates for federal listing into the Big South Fork Cumberland River. The four species re- Big South Fork Cumberland River, Kentucky / Monte McGregor leased were the oystermussel (FE), the dromedary pearlymussel (FE), spec- m x 3 m) were selected and a known quantitative sampling, and on average taclecase (candidate species), and the number of mussels were released in the mussels did not move from their fluted kidneyshell (candidate species); each section. The spectaclecase, which original release area. We estimate that all are likely to be extirpated from the is generally found under large rocks, all except for 2 tagged mussels moved Big South Fork. Individuals from all were released outside the grid in habitat less than 3 meters. This site will con- four species were collected from the that was more suitable for the species. tinue to be monitored semi-annually to Clinch River, TN. After mussels were Three days after the release, mussels examine survival and reproduction of collected, they were tagged and held at could be seen at the surface siphoning all released species, as well as changes the Center for Mollusk Conservation and displaying fish host lures. in the entire mussel community at the in Frankfort, KY. On June 9 2008, we In September 2008, we went back release site. released approximately 300 individuals: to the site to monitor the reintroduced 97 oystermussel, 19 dromedary pearly- mussels. We completed another quan- Funding Source: Endangered Species mussel, 43 spectaclecase, and 142 titative survey that included the entire Act (Section 6) funds, U.S. Fish and fluted kidneyshell. Before the release, 6 m x 12 m release grid, with an ad- Wildlife Service quantitative sampling was done in the ditional several meters surrounding area where the mussels were to be re- the grid. This allowed us to encounter KDFWR Strategic Plan. Goal 1. leased. We found 12 species present in mussels that may have moved out- Strategic Objective 5. Comprehen- low densities. None of the species be- side of the release grid. Because we sive Wildlife Conservation Strategy: ing reintroduced were found. After the released all mussels in a known area, Appendix 3.2, Class Bivalvia. Prior- survey was completed, we selected a 6 and we recorded tag numbers in all 8 ity Research Project #1 and #2. m x 12 m grid to release the mussels. sections, we could track movement of To facilitate the release in known areas the released mussels. All species re- of exact densities, 8 sections (each 3 leased in the grid were located during

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Mollusks

Successful Augmentation of the Fatmucket, Lampsilis siliquoidea, in Elkhorn Creek, Kentucky

Center for Mollusk Conservation, Kentucky Department of Fish and Wildlife Resources Monte A. McGregor, Adam C. Shepard, Fritz Vorisek, J. Jacob Culp, Jim Hinkle, Andrew Wooldridge, Vaughn Kauffman

n August 2007 the Kentucky Depart- Iment of Fish and Wildlife released 100 individuals of the Fatmucket in Elkhorn Creek, KY. Released individuals were propagated at the Center for Mollusk Conservation for approximately 1.5 years. The mean length at release was 20.83 mm. Before the augmentation a quantitative survey was completed and 9 species were found in low densities, including the Fatmucket. Only two individuals of the Fatmucket were found in initial survey. A 5 m x Post-release tagged fatmuckets collected from 5 m release area was chosen based on Elkhorn Creek, Kentucky / Monte McGregor habitat stability and mussel density. All individuals were released in the area and allowed to bury into the substrate. Approximately one year later, in September of 2008, another quantitative survey was done to monitor the released individuals and evaluate the augmentation. We surveyed a concentrated area surrounding the release site and again found 9 species, this time with higher densities. We also found 12 of the released Fatmuckets. Their mean length after one year in the wild was 47.62 mm (an increase of 129% from the initial release size). No released Fatmuckets were found gravid (too young), but we will continue to monitor this site semi-annually to examine Pre-release tagged fatmuckets in 2007 / Monte McGregor growth, survival, and reproduction, as well as changes in the entire mussel Act (Section 6) funds, U.S. Fish and Strategic Objective 5. Comprehen- community at the release site. Wildlife Service) sive Wildlife Conservation Strategy: Appendix 3.2, Class Bivalvia. Prior- Funding Source: Endangered Species KDFWR Strategic Plan. Goal 1. ity Research Project #1.

48 Kentucky Department of Fish & Wildlife Resources Mollusks / PROJECT HIGHLIGHTS

Freshwater Mollusk Monitoring in the South Fork Kentucky River System

Redbird Ranger District of the Daniel Boone National Forest (DBNF) is one of the most intensively utilized for gas and mineral extraction. Other tributaries showed moderate to serious sedimentation likely attributable to runoff from coal mining and poor agricultural practices. KSNPC has been fortunate to have received intensive instream habitat mapping data collected by DBNF staff for the Redbird River. We hope to use this information to contrast against the mussel data in the system to help examine trends. Because the South Fork Kentucky River watershed is generally regarded as the best quality of the three forks of the Kentucky River, future projects to abate issues from sediment runoff and mining effluent should be Redbird River, quantitative sampling site / Ryan Evans considered.

Ryan Evans, Kentucky State have generated a sampling grid for the Funding Source: State and Tribal Nature Preserves Commission South Fork Kentucky River and have Wildlife Grants (SWG) been doing reconnaissance of potential KDFWR Contact: Danna Baxley field sites. We will be sampling approx- KDFWR Strategic Plan. Goal 1. imately 29 stations on the mainstem Strategic Objective 5. Comprehen- his project is focused on identify- South Fork Kentucky in the upcoming sive Wildlife Conservation Strategy: Ting the best remaining freshwater field season. Appendix 3.2, Class Bivalvia. Prior- mussel hotspots in the South Fork Preliminary observations indicate ity Survey Project #1. Kentucky River watershed. In addi- that the Redbird River is the best tion, basic inventory is being done for remaining tributary in the South freshwater snails as recent information Fork Kentucky River system on the group is generally lacking within for freshwater mollusks; how- the basin. ever, this does not infer that this In 2008, 28 sites were qualita- system is without threats. Strip tively sampled in the basin, with 1 site mining has obviously taken a toll quantitatively sampled. The project to on the upper sites in the Redbird date has located 16 species of freshwa- as no freshwater mussels were ter mussels, including records for the located despite intensive search round hickorynut (Obovaria subrotun- effort. Further, our quantita- da), a species of conservation need in tive sampling data at one of the Kentucky as well as the little spectacle- higher quality sites in the lower case, Villosa lienosa. KSNPC is still Redbird River identified during compiling information and identifying qualitative sampling indicates a freshwater snail specimens but field low density mussel population data indicates no KSNPC-listed spe- with little recent recruitment. Round hickorynut (Obovaria cies have been located to this point. We This information is troubling, as the subrotunda) / Ryan Evans

Annual Research Highlights 2008 9 PROJECT HIGHLIGHTS / Reptiles and Amphibians

Life History and Population Assessment of the Western Cottonmouth in Western Kentucky

Edward Zimmerer, Murray State University KDFWR Contact: John MacGregor

he Western Cottonmouth (Agkis- Ttrodon piscivorus leucostoma) is a large venomous snake that occurs in the Western Kentucky Coal Field and the Mississippi Embayment regions of Kentucky (West of Ohio and Butler counties). Usually found in sloughs, sluggish streams, bayous, and other slow-moving water habitats, the West- ern Cottonmouth prefers areas that are at least partially wooded that contain emergent or aquatic vegetation. Al- though rangewide population trends Western cottonmouth / Phil Peak for the cottonmouth are unknown, populations in Kentucky are thought Transponder (or PIT tag) containing a history information which will be nec- to be declining, primarily due to the unique electronic identification number, essary for implementing a long-term loss of high quality forested wetland so that these individuals can be identi- conservation plan for this snake. habitat as a result of surface mining, fied in subsequent years. By marking conversion to cropland, and urban/sub- these snakes with PIT tags, growth Funding Sources: Murray State Uni- urban development. As a result of the rates, survivorship, age-specific mortal- versity, State and Tribal Wildlife Grant perceived population decline of the ity rates, and other important life his- (SWG). Western Cottonmouth in Kentucky, this tory parameters will be assessed. species is listed as a Species of Great- We plan to continue monitoring Comprehensive Wildlife Conserva- est Conservation Need in Kentucky’s this important cottonmouth population tion Strategy: Appendix 3.3, Class Wildlife Action Plan. Until recently, in future years. Continued monitoring Reptilia. Priority Research Project major hibernacula and locations of will enable us to collect valuable life #1. high-density populations of this taxon were unknown in Kentucky. Beginning in 2008, Murray State University began monitoring the healthiest known cottonmouth population within Kentucky. At this site, drift fence arrays are used to capture and monitor cottonmouths moving from winter hibernacula to their spring and summer foraging areas. To date 1,619 cottonmouths have been captured at the site of interest, which is the only known area in Kentucky where these snakes are concentrated at such high densities. Of these 1,619 snakes, 439 have been marked with a Passive Integrated Western cottonmouth records for Kentucky

50 Kentucky Department of Fish & Wildlife Resources Reptiles and Amphibians / PROJECT HIGHLIGHTS

Status Assessment and Conservation of the Eastern Hellbender

Gregory Lipps and Mike Sisson, lations, and identify threats to hellbender These preliminary results are very simi- Gregory Lipps, LLC populations and habitats. Our goal is to lar to what is being reported elsewhere throughout the hellbender’s range -- a KDFWR Contact: John search for hellbenders in every historic location of occurrence, using records 75-80% decline in overall abundance MacGregor from the Natural Heritage Database and with declining populations comprised other sources. Searches consist of skin of old individuals. Excess siltation is orldwide declines of amphib- diving using a snorkel and mask while degrading many hellbender habitats, Wian populations have garnered lifting large rocks. Captured hellbend- caused by development and agriculture international attention over the past ers are measured and massed, and all along streams, ATV use, and in-stream two decades. The recent Global Am- injuries and deformities are carefully gravel mining. Going forward, we will phibian Assessment found that 32% documented. Prior to release, a micro- be examining the remaining 38 loca- of the known amphibian species are chip (identical to those used to identify tions of historical occurrence as well threatened with extinction or already pets) is placed under the skin to allow us as other areas of potential hellbender extinct, while 42% are declining. One to identify previously captured individu- habitat throughout the state. In addi- of these declining species is the hell- als. As larval hellbenders are thought to tion, we are working with our partners bender (Cryptobranchus alleganiensis), reside in the interstitial spaces of gravel to protect stream and riparian habitats the largest North American amphibian. and adults spend their lives in cavities where healthy hellbender populations Hellbenders are completely aquatic under large rocks, we are quantifying the still occur, so that this unique animal salamanders, living their lives in large substrate quality at each location using a will continue to contribute to the rich creeks and rivers, where they take ref- zig-zag pebble count procedure, provid- biodiversity of Kentucky. uge under large slab rocks and feed on ing one measurement of the health of the crayfish. While Kentucky makes up a habitat. Funding Sources: State and Tribal substantial part of the range of the East- To date, we have captured Eastern Wildlife Grant (SWG), Gregory Lipps, ern Hellbender, the current status and Hellbenders at only two of 24 historical LLC, and Columbus Zoo and Aquarium distribution of hellbenders in the state locations in the Licking and Kentucky is unknown. River watersheds. Furthermore, indi- Comprehensive Wildlife Conserva- In 2008, we began a survey to viduals of multiple size classes (indi- tion Strategy: Appendix 3.2, Class determine where hellbenders occur in cating recent successful reproduction) Amphibia. Priority Research Project Kentucky, collect information on popu- have been found in only one waterway. #1 and Priority Survey Project #2.

Eastern hellbender / Ralph Pfingsten

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Reptiles and Amphibians

Inventory, Monitoring and Management of Amphibians and Reptiles in Kentucky

Will Bird and Phil Peak, Kentucky Herpetological Society KDFWR Contact: John MacGregor

n the course of developing Ken- Itucky’s Comprehensive Wildlife Conservation Strategy (CWCS) it was determined by KDFWR that more baseline data needed to be collected in order to execute effective conservation action plans for our native reptile and amphibian species. The actual distributions for many reptile and amphibian species in Kentucky have not yet been entirely determined. Many of the records that we have in our current database are decades old, and the landscape has been altered to such a degree that the current distributions of many species remain unknown. Species for which Timber rattlesnake / Will Bird baseline data is most needed from all groups of reptiles and amphibians have where they could remain undetected. We were able to locate many of the been identified, as have the regions There are species of reptiles and am- species identified by the CWCS includ- within Kentucky where this information phibians for which AC has proven less ing some in areas where they had not should be gathered. effective. When targeting these species been recorded at all or for many years. Locating reptiles and amphibians we use box style funnel traps, along We were also able to secure new survey can be difficult. We begin the process with drift fence arrays, to assist in their locations where we deployed AC that by identifying locations where we location. we hope will lead to 2009 records. The believe targeted species can be found. Once animals are discovered, we surveys we conduct will help in identi- These locations are on state, federal, record their exact location and any fying key habitats for species of reptiles and even private lands. Once permis- other natural history observations such and amphibians in Kentucky. As more sion is granted to conduct surveys, as weather conditions, time of day, information is collected we believe we use different methods for locating etc. If the species is not common or that potential for the implementation reptiles and amphibians based on their if it is a species that could be difficult of effective conservation plans will be biological requirements. Because they to properly identify, we photograph it possible. are ectotherms we are able to utilize prior to release. All of our information Artificial Cover (AC) to locate many of is sent to KDFWR State Herpetologist Funding Sources: State and Tribal the animals we search for. Heavy metal John MacGregor for review prior to Wildlife Grant (SWG), Kentucky Herpe- objects that absorb heat from the sun’s being entered into the main database. tological Society rays and provide protection from the el- Should there be any question about the ements are set out at our study sites. We authenticity of entered data we are able Comprehensive Wildlife Conserva- also deploy large wooden boards which to provide photographic evidence. This tion Strategy: Appendix 3.4, Class retain moisture even during the drier process eliminates the need to preserve Amphibia. Priority Conservation months and provide refuge for many or collect the animals that we find. Action #2; Class Reptilia. Priority of the creatures that might otherwise 2008 proved to be a good year for Conservation Action # 2. stay far below the surface of the ground gathering reptile and amphibian data.

52 Kentucky Department of Fish & Wildlife Resources Birds / PROJECT HIGHLIGHTS

Reproductive Success of Interior Least Terns in Western Kentucky

Caitlin Borck and Robert B. degraded because of changes in the his- Swan Lake had 54 nests, but, likewise, Frederick, Eastern Kentucky toric flooding regime. In the absence of only one egg hatched. Arkema, our in- University, John Brunjes, natural islands, least terns have nested dustrial site had a much higher success. on islands created by US Army Corps This site had only 14 nests, but nest Kentucky Department of Fish of Engineers dredging operations or at success was 93% and hatch success was and Wildlife other alternative nesting sites. Due to 91%. Average clutch size for all sites prolonged flooding and high water from was 1.5. In 2009 we plan to continue istorically, interior least terns May to August 2008 least terns in west- to monitor these sites and others along H(Sterna antillarum athalassos) ern Kentucky were unable to nest on the Ohio and Mississippi Rivers in have nested along the Colorado, Red, islands along the Ohio and Mississippi Kentucky and record basic reproductive Arkansas, Missouri, Ohio, and Mis- rivers. With traditional sites unavail- data. If water levels allow, we also will sissippi Rivers. They were extremely able, least terns nested on agricultural evaluate and compare habitat suitability abundant in the 1800s but nearly extir- fields and industrial sites. of man-made versus natural islands. pated by the end of that century due to Nesting sites were discovered With the loss of so many suitable nest- demands of the millinery trade, hunting, through reports by the public and Ken- ing sites, management and protection of and egg collection. Their numbers de- tucky Wildlife Division employees. existing natural and alternative sites is clined again in the 1940s because of hu- Each site was systematically searched increasingly important. These data will man development along river banks and for nests until the entire area was cov- help us manage these sites to improve changes to the river channels. The loss ered. Nests were uniquely marked and tern reproductive success. of nesting habitat due to channelization, monitored every 2-4 days. Clutch size, dam and weir construction, irrigation, nest fate, and clues to fate were record- Funding Source: State and Tribal and reservoir construction is considered ed at each nest. The two agricultural Wildlife Grant (SWG) the greatest threat to least tern popula- sites, Swan Lake and Open Pond suf- tions. Human disturbance of nesting fered heavily from predation. Several Comprehensive Wildlife Conservation sites is considered the second greatest sites on Open Pond were accidentally Strategy: Appendix 3.4; Class Aves: threat to least tern populations while oth- plowed under. Open Pond had 89 nests, Taxa specific conservation action #1. er causes of reproductive failure include but only one egg hatched successfully. depredation and flooding. Because of the loss of their nesting habitat and population decline they were listed as federally endangered in 1985. Interior least terns are now only found in small, local populations in their former nesting range. Those nesting on islands on the Ohio and Mississippi rivers in Kentucky have lost natural islands or seen island quality Caitlin Borck checking fate of tern nest at the Swan Lake agricultural site / Bob Frederick

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Birds

The Common Raven in Cliff Habitat: Detectability and Occupancy

Josh Felch, John J. Cox, vens in an area. Should University of Kentucky; breeding locations of ravens be identified within Matthew Dzialak, Hayden-Wing the Commonwealth, Associates, WY; and Shawchyi additional benefits will Vorisek, Kentucky Department of include data on timing, Fish and Wildlife Resources productivity, and breeding season food habits. We will opportunisti- he cliff-nesting common raven cally gather similar data T(Corvus corax) is of conserva- on ravens discovered to tion interest in Kentucky and currently nest in non-cliff habitat listed as state threatened. The common in the state. raven was nearly extirpated in the east- Ultimately, we ex- ern U.S. by the mid 1900s as a result pect to gain a better idea of human persecution, loss of forest of the status of the com- habitat, and likely the absence of large mon raven population in mammal carrion. Today, portions of its Kentucky and to generate Raven hatchlings / Josh Felch range have been recolonized by rem- new information for de- nant populations including southeastern ter-spring 2010. tecting ravens at Kentucky’s cliffs and on Kentucky where a handful of sightings Detectability of ravens in cliff habitat features that might be important and nests have been observed during habitat will be estimated by conduct- in their occupancy of potential breeding the past 3 decades. The Commonwealth ing repeated visual/auditory surveys at sites. We hope to develop a monitoring appears to have extensive suitable 12-15 known raven breeding sites in protocol useful to wildlife managers and breeding habitat but ravens have re- the Southern Appalachians during the land stewards interested in long-term mained relatively rare and unstudied, 2009 breeding season (February – early monitoring, management, and conserva- thus little is known about the local May). These sites have been identi- tion of the species in cliff habitat. ecology or population status or this oft fied through coordination with biolo- reclusive corvid. gists, naturalists, birders, and others Funding Source: State and Tribal Our research will employ a detec- throughout the region, and paired with Wildlife Grant (SWG) tion-occupancy approach to assess an equal amount of unoccupied sites, distribution and abundance of ravens in will provide the basis for a site-attribute Comprehensive Wildlife Conserva- cliff habitat of eastern Kentucky. Our habitat model that will quantify breed- tion Strategy: Appendix 3.4; Class objectives are threefold. First, we plan ing habitat in the region. Key potential Aves: Taxa specific conservation ac- to quantify factors that affect our abil- breeding habitats in Kentucky will be tion #1. ity to detect ravens in cliff habitat by identified through communication with determining the intensity and duration state ornithologists, by historical obser- of survey visits at occupied raven nests vations, and by recent sightings. During required to detect these species with a the 2010 breeding season, we will con- level of confidence. Second, we will duct repeated visual/auditory surveys at quantify landscape attributes at known these potential sites as was done before nesting locations to create a landscape- at the known breeding sites. The detect- scale species predictive models in cliff ability estimate obtained during the first habitat of this region. Finally, using field season will be used to determine data from the first field season, we will the protocols for monitoring occupancy develop and initiate protocols for moni- at these sites, such as the necessary toring the occupancy of key potential allocation of survey effort needed to breeding habitats in Kentucky in win- reliably infer presence/absence of ra- Common raven / Josh Felch

54 Kentucky Department of Fish & Wildlife Resources Birds / PROJECT HIGHLIGHTS

An Evaluation Tool for Avian Monitoring Programs

Rua S. Mordecai, Shawchyi gling with evaluating their monitor- to evaluate bird monitoring programs, Vorisek, Kentucky Department of ing programs under the State Wildlife recommend improvements, and track Action Plans (SWAP). For instance, implemented improvements over time. Fish and Wildlife Resources; and many state agencies have supported and A secure online database was used in The Central Hardwoods Joint conducted various avian monitoring which biologists could evaluate moni- Venture projects throughout the years without toring programs, based on the devel- consideration of management priorities, oped questions, and to track improve- he need to coordinate bird moni- regional or continental conservation ments in monitoring over time. Ttoring efforts has been identified goals, statistical rigor, and data man- Kentucky served as the pilot by the U.S. North American Bird Con- agement. Thus, a regular assessment of state for this project as KDFWR staff servation Initiative’s (NABCI) as one existing and proposed projects is need- worked with the principal investigator of their highest priorities. As a result, ed by states so that current and future and the Central Hardwoods Joint Ven- the NABCI Monitoring Subcommit- population monitoring projects can be ture in developing the questions, testing tee identified a set of goals to improve altered to address current management the evaluations, and online applications. avian monitoring. These are: 1) Fully and bird conservation objectives. As a Additionally, three outside reviewers integrate monitoring into bird man- starting point in assessing states’ moni- were asked to evaluate the questions agement and conservation practices toring programs, we developed a set of for applicability. The monitoring evalu- and ensure that monitoring is aligned multiple choice questions based on the ation tool was also presented at the with management and conservation elements identified by the Monitoring Joint Southeast Quail Study Group and priorities, 2) Coordinate monitoring Subcommittee, which where then used Southeast Partners In Flight (SEPIF) programs among orga- Meeting in Columbia, SC in nizations and integrate March of 2009. The evaluation them across spatial tool was well received by the scales to solve conser- SEPIF working group and the vation or management group plans to bring this tool problems effectively, forward and integrate it into 3) Increase the value of future web-based systems and monitoring by improv- avian conservation planning ing statistical design, organizations. and 4) Maintain bird population monitoring Funding Sources: State and data in modern data Tribal Wildlife Grants (SWG) management systems. Although many Comprehensive Wildlife large-scale organiza- Conservation Strategy: Ap- tions (e.g., Joint Ven- pendix 3.4; Class Aves: Taxa tures and regional bird specific conservation action monitoring partner- #1. Appendix 3.9; Critical ships) are now attempt- Elements of Successful Moni- ing to align their bird toring in Kentucky. monitoring efforts with the four goals of the NABCI subcommittee, none of these groups have a standard way to measure and track progress toward these goals. Additionally, many states are strug-

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Birds

Estimating Abundance of Species of Concern in the Central Hardwoods Region

Shawchyi Vorisek and Kate Surveys were completed utilizing KD- produce population estimates and to Heyden, KDFWR; Frank R. FWR staff, as well as personnel from identify suitable forest habitat for target many other state and federal agencies. priority species, including Cerulean Thompson III, Jennifer Reidy, There were 580 road points and 107 Warblers, on public lands in the Central Kerri Cornell Duerr and Wes river points sampled. Road surveys Hardwoods Bird Conservation Region. Bailey, University of Missouri, sampled upland habitat; river surveys Analysis was completed by the Univer- Jane Fitzgerald, American Bird sampled bottomland habitat. There sity of Missouri in Columbia. Distance Conservancy. were 1437 individuals of target species and removal models were compared for detected, 44 of which were cerulean density estimation of forest birds, ac- warblers (Table 1). counting for variability in the probabil- n 2007 and 2008, KDFWR coordi- The goal of this survey effort was ity of detection of individuals during a Inated Cerulean Warbler surveys for to develop a sampling design, survey point count (Table 2). the Central Hardwoods Joint Venture. methods, and analysis methods to

Table 1. Detections of each target species from sampled areas (public forested land) within Kentucky, 2007–2008.

Worm- Acadian Cerulean Kentucky Louisiana Wood Site eating Total Flycatcher Warbler Warbler Waterthrush Thrush Warbler Bernheim 19 18 3 26 37 103 Research Forest Clay WMA 41 1 19 1 4 28 94 Clark’s River 18 7 7 6 1 1 40 NWR Dale Hollow 4 2 2 2 2 8 20 WMA Fort Knox 20 3 19 6 11 48 107 Green River 36 16 4 5 39 100 WMA Kleber WMA 3 1 Lake Cumberland 4 2 4 State Park Land Between the 115 6 41 19 52 85 318 Lakes NRA Mammoth Cave 102 18 101 19 77 141 458 NP Pennyrile SF 13 2 27 30 41 113 Taylorsville WMA 18 1 14 33 Tradewater WMA 3 2 3 2 10 Yellowbank WMA 14 5 7 5 10 41 Total 410 44 260 65 213 459 1437

56 Kentucky Department of Fish & Wildlife Resources Birds / PROJECT HIGHLIGHTS

Table 2. Population estimates (based on males) and 95% confidence intervals for each target species from top distance and top removal models from sampled areas (public forested land) within Kentucky, 2007–2008 (University of Missouri, unpubl. report).

Distance Removal Species Area (ha) Estimate 95% CL Estimate 95% CL

Acadian Flycatcher 14787 6912 (5285–9216) 5382 (4177–7441) Cerulean Warbler 14787 500 (235–1276) 1009 (423–3304) Kentucky Warbler 14787 3540 (2191–6075) 1697 (1514–2422) Louisiana Waterthrush 14787 1588 (521–8294) 390 (320–681) Worm-eating Warbler 14787 2746 (1839–4498) 1701 (1409–2474) Wood Thrush 14787 2037 (1476–2893) 5700 (3148–14020)

Funding Sources: State and Tribal Wildlife Grants (SWG)

KDFWR Strategic Plan. Goal 1. Strategic Objective 5. Comprehen- sive Wildlife Conservation Strategy: Appendix 3.4; Class Aves: Taxa specific conservation action #1.

Female and male cerulean warbler / Adam Smith

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Birds

Golden-winged Warbler Monitoring

Shawchyi Vorisek, Kentucky Department of Fish and Wildlife Resources and Sara Barker (Cornell Lab of Ornithology)

he Golden-winged TWarbler (Vermivora chrysoptera) is a Neotropi- cal migrant that is of high conservation concern in eastern North America. In Kentucky, they breed in the higher elevations of the Appalachian Mountain in the Southeast. They currently utilize the early sucessional habitats created on reclaimed mined lands. However, their populations are quickly declining. Fur- thermore, the species is hy- bridizing with Blue winged Warblers (Vermivora pinus) which have been expand- Golden-winged warbler / Patricia Hartman ing into the Golden-winged Warbler’s (GWWA) range in Southeast and a mobbing sequence throughout will enable us to determine population Kentucky, adding another threat to the the species range in the Appalachians. trends throughout the Appalachians and species. Similar threats face the species Kentucky participated in the project by portions of New York. rangewide. utilizing the protocol to survey points Thus, the Golden-winged Warbler where GWWAs were observed during a Funding Source: State and Tribal Working Group was established and 2003 atlas survey. Wildlife Grant (SWG), National Fish consists of ornithologists, conservation- Seven sites were surveyed in KY and Wildlife Foundation (NFWF) ists, and managers with the common from late May through early June in goal of ensuring the conservation of Bell, Harlan, McCreary, Pike, and Comprehensive Wildlife Conserva- GWWA populations through sound Whitley Counties. Twenty-seven points tion Strategy: Appendix 3.2; Class science, education, and management. were surveyed, with only 4 points hav- Aves: Taxa specific research project The group acknowledged that because ing GWWA only observations. All #6, taxa-specific survey project #3 GWWAs are patchily distributed and other points had either only BWWA, and #5. Appendix 3.9. New projects. poorly sampled by other surveys, such both species, or neither species ob- as the Breeding Bird Survey, long-term served, suggesting that isolated GWWA monitoring of the species is needed. As populations have declined. We plan to a result, the Cornell Lab of Ornithology continue monitoring the populations in (CLO) established a pilot monitoring 2009 following revised protocol from protocol in 2008 that incorporated the CLO that incorporates a spatially bal- use of playback, passive point counts, anced sampling design. Such a design

58 Kentucky Department of Fish & Wildlife Resources Birds / PROJECT HIGHLIGHTS

Grassland Songbird Survey

Shawchyi Vorisek (KDFWR; instance, the national Breeding Bird maintained at the county level. David Buehler, University of Survey is not sufficient in capturing The 2008 surveys yielded more Henslow Sparrows or grassland birds at Henslow Sparrows (n = 45) in KY Tennessee; Tiffany Beachy, smaller scales that would be useful to than any other state with pasture and University of Tennessee; Central land managers. hayfields being the most dominant Hardwoods Joint Venture As a result, this project was es- land use type. For all states com- tablished in an effort to develop an ad- bined, Henslow’s Sparrows occurred equate and readily repeatable method- in areas with a greater percentage of any grassland and early suc- ology for surveying the available early cool season grass fields and Northern Mcessional songbirds, such as successional habitat across a regional Bobwhites were found more often in Henslow’s Sparrows (Ammodramus bird conservation region (BCR), the a mixture of cool season grasses and henslowii) have been experiencing pre- Central Hardwoods BCR and attempts forbs. These results suggest that proper cipitous population declines in North to document priority grassland bird grassland management in KY (e.g. America. Causes include habitat altera- species’ distributions and habitat as- higher percent cover of native warm tions, succession, mowing regimes, sociations. Roadside and atlas surveys season grass stands) can favor priority fire suppression, and development. were conducted in priority bobwhite grassland species and that region-wide Additonally, the game species Northern focus areas within the states of IL, IN, surveys such as these can offer useful Bobwhite (Colinus virginianus) utilizes KY, and TN during May-July 2008. information on population trends and similar habitat and is also experienc- Roadside surveys consisted of 5-min- land use. Plans for 2009 include ex- ing major population declines. Thus, ute point counts that targeted several panding this protocol to cover counties management is clearly needed to ben- grassland species of conservation con- in other states within the Central Hard- efit these grassland species. However, cern. Habitat was also categorized into woods BCR and coordinating efforts few surveys are available to assess the different land use types. Surveys were so that states can conduct these surveys effects of management nor to track conducted at the county scale as agri- systematically to ultimately better man- population trends for these species. For culture census data from the USDA are age our grassland species.

Funding Sources: U.S. Fish and Wildlife Ser- vice, State and Tribal Wildlife Grant (SWG), Kentucky Department of Fish and Wildlife Re- sources, Central Hard- woods Joint Venture, University of Tennessee

Comprehensive Wild- life Conservation Strategy: Appendix 3.2; Class Aves: Taxa specific survey project #5.

Grasshopper sparrow / Don Martin

Annual Research Highlights 2008 9 PROJECT HIGHLIGHTS / Mammals

Black Bear Resource Selection, Demographics, and Movement Patterns in Kentucky

Ben Augustine, Rebekah Jensen, John T. Hast, and John J. Cox, dled and the 2009 den season resulted in the University of Kentucky; Steven documentation of 7 Dobey and Jayson Plaxico, litters and 13 cubs. As Kentucky Department of Fish demographic data ac- and Wildlife Resources cumulates, it appears initial fecundity estimates were overly op- fter being extirpated from Ken- timistic – 3.25 cubs/ Atucky in the late 19th century, the litter from 2002-2006. black bear (Ursus americanus) has Black bear productiv- expanded its range back into the state ity in eastern Ken- and the population may still be increas- tucky now appears to ing in both size and range. As forest be more consistent obligates with large space requirements with other black bear and seasonal variation in the location populations in the of forage resources, black bears in east- southern Appala- ern Kentucky must navigate through a chians. To date, east- matrix of natural and human-modified ern Kentucky black habitats to meet their life requisites. bears have averaged The interspersion of black bear habitat 2.7 cubs per litter. An- and areas of human activity often leads other trend is the in- to human-bear conflict, sometimes crease in the capture resulting in property damage and bear of females – 16 of the mortality. In order to manage black 33 females captured bears in a way that allows for popula- to date were captured tion persistence and minimizes human- in the past 2 years. bear conflict, an understanding of how While many demo- bears respond to landscape structure Bear tracks / John Cox graphic parameters and human activity is especially im- remain uncertain, the portant. Further, the black bear has be used to assess how black bears are black bear population in eastern Ken- recently become a game species in responding to landscape structure and tucky appears to be healthy and contin- Kentucky, necessitating an accurate es- human activity, ultimately leading to a ued research will ensure that wildlife timation of demographic parameters for spatially-explicit probability of occur- managers have the information needed robust population models and harvest rence map for the black bear in eastern to properly manage this species. regulations. Kentucky, which can inform manage- Research conducted by the Univer- ment decisions and land acquisition by Funding Source: Pittman Robertson sity of Kentucky has recently focused conservation organizations. While sta- (PR) on the collection of high-intensity GPS tistical analyses have not yet been con- data that allow for the assessment of ducted, it appears that females general- KDFWR Strategic Plan. Goal 1. fine-scale resource use and movement ly avoid areas of human activity, while Strategic Objective 5. Comprehen- patterns, while also allowing for the males are often attracted to human food sive Wildlife Conservation Strategy: continued collection of demographic sources, suggesting the distribution of Appendix 3.9; Class Mammalia: data. Since 2002, 48 bears have worn female black bears will ultimately be Taxa specific conservation project. GPS collars, with 24 being deployed more restricted than that of males. in the past year and at least 17 to be Since 2003, 31 litters have been deployed this summer. These data will documented with 69 cubs being han-

60 Kentucky Department of Fish & Wildlife Resources Mammals / PROJECT HIGHLIGHTS

Genetic Diversity, Structuring, and Recolonization Patterns of Black Bears in Eastern Kentucky

John Hast, John J. Cox, and Ben Augustine, University of Kentucky; Steven Dobey and Jayson Plaxico, Kentucky Department of Fish and Wildlife Resources

opulation genetics has become Pa popular and useful tool for addressing important management and conservation questions involving large mammals. The black bear (Ursus americanus) has successfully recolonized the extreme southeast portion of Kentucky following nearly a century of conjectured absence from the Com- monwealth. Recolonization of bears into Kentucky has spawned a myriad of questions involving the source population, travel corridors, and the genetic makeup of the resident bear population. It has been hypothesized that the bears Red dots represent the 63 hair samples taken for Kentucky black bears. currently present in Kentucky represent the outer fringes of a larger regional remnant population, many examples of Virginia black bear hair samples with metapopulation recently (if not cur- interchange between Virginia and Ten- intact follicles and geographical coor- rently) sourced by surrounding states. nessee have been documented using dinate data. University of Tennessee Although possible that bears residing radio tracking devices. has agreed to provide 30 hair samples within Kentucky are products of a small Our research will employ a popu- from the Smokey Mountains. In May lation genetics approach using micro- and June 2009, we will use hair snares satellite analysis of hair samples from and biopsy darting to collect additional black bears to examine the following hair samples in Virginia and Big South four project objectives: (1) evaluation Fork, Kentucky. Given that bears ap- of the genetic diversity of the Kentucky pear to be regionally increasing in num- black bear population, (2) evaluation of ber, and are perhaps transitioning from the genetic structure of the Kentucky an establishment to growth phase, our black bear population, (3) evaluation of findings should prove useful to adaptive possible source populations located in management of this socio-economically surrounding states, and (4) evaluation and ecologically important large game of recolonization patterns and corridors species. associated with the Kentucky black bear population. Hair samples will be Funding Source: Pittman Robertson (PR) collected from captured bears that are part of a complimentary ecological KDFWR Strategic Plan. Goal 1. study, as well as opportunistically from Strategic Objective 5. Comprehen- state employees dealing with road kill sive Wildlife Conservation Strategy: and poaching incidents. To date we Appendix 3.9; Class Mammalia: Trapped male bear / John Hast have obtained 63 Kentucky and 9 West Taxa specific conservation project.

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Mammals

Genetic Characteristics of Restored Elk Populations in Kentucky

Virginia Dunn, Steve Demarais and Bronson Strickland, Mississippi State University; Randy DeYoung, Texas A&M University – Kingsville; Tina Brunjes, Kentucky Department of Fish and Wildlife Resources

astern Kentucky currently has a Ethriving elk (Cervus elaphus) population, thanks to restoration efforts by the Kentucky Department of Fish and Wildlife Resources (KDFWR) beginning back in 1997. Retention of genetic diversity is important to the success of wildlife populations, including elk. Ge- netic diversity is important to individual and population survival, adaptiveness, growth and reproductive potential. Future management decisions, such as hunting season regulations, need to be made with the genetic structure of the Collared elk calf / Will Bowling population in mind. The KDFWR and Mississippi State University have teamed up to evaluate the genetic makeup of the eastern Ken- tucky elk herd. During fall and winter 2008-2009, biologists obtained tissue and hair samples from hunter harvested elk and will collect these samples again during the 2009-2010 hunting seasons. We will use DNA taken from these samples to evaluate the genetic makeup of the elk across the restoration area and compare this to their source populations in western states. Weight and antler size will also be evaluated. This information will allow the KDFWR to make future management decisions that will promote elk population health.

Funding Sources: Pittman Robertson (PR) and Mississippi State University

KDFWR Strategic Plan. Goal 1. Strategic Objective 5. Harvested elk / KDFWR

62 Kentucky Department of Fish & Wildlife Resources Mammals / PROJECT HIGHLIGHTS

Geographic distribution and prevalence Cytauxzoon felis in wild felids

Barbara C. Shock, University 1 serval [Leptailurus of Georgia; Staci M. Murphy, serval]), Missouri Southeastern Cooperative (39 bobcats), North Wildlife Disease Study; Laura Carolina (8 bobcats) and Oklahoma (20 L. Patton, Kentucky Department bobcats), were 27%, of Fish and Wildlife Resources; 55%, 33%, 79%, 63% Philip M. Shock, West Virginia and 60% respectively. Division of Natural Resources; The prevalence was Colleen Olfenbuttel, North lower in West Virginia (0%, 37 bobcats), Carolina Wildlife Resources Ohio (5%, 19 bob- Commission; Jeff Beringer, cats), Georgia (3%, Missouri Department of 69 bobcats) and North Conservation; Suzanne Prange, Dakota (3%, 114 Ohio Department of Natural bobcats). These data indicate that C. felis is Resources; Dorothy Fecske, widespread in bobcat North Dakota Game and Fish populations, but the Department; Matt Peek, Kansas spatial differences in Department of Fish and Wildlife; prevalence may relate Victor F. Nettles, Southeastern to differences in the distributions of the Cooperative Wildlife Disease two tick species. The Study; Holly Brown, University ultimate goal of this of Georgia; David S. Peterson, project is to inves- University of Georgia; and tigate intraspecific Michael J. Yabsley, University of variability of C. felis throughout the Eastern Georgia Bobcat / Adam Jones U.S. by comparison of ITS sequences pres- ytauxzoon felis, a tick-borne Louisiana, Missouri, North Carolina, ent in wild felids with those detected in Cprotozoal parasite of wild and do- North Dakota, Ohio, Oklahoma, and domestic cats and ticks. mestic felids, is the causative agent of West Virginia). The bobcat is believed cytauxzoonosis in domestic and exotic to be the primary reservoir for C. felis, Funding Sources: Southeastern Coop- felids. C. felis can be transmitted by but few studies have looked at the dis- erative Wildlife Disease Study, Various two tick species, Dermacentor varia- tribution and prevalence of the parasite State Agency Interests (including Ken- bilis and Amblyomma americanum. within wild felids. Blood and/or spleen tucky Department of Fish and Wildlife The distribution of these ticks overlap samples from hunter/trapper-killed fe- Resources). considerably throughout the Southern lids (n=420) were tested for C. felis by US, but D. variabilis ranges farther PCR, targeting the ribosomal internal KDFWR Strategic Plan. Goal 1. into northern states. The objective of transcribed spacer region 1 (ITS-1). Strategic Objective 5. the current project was to determine Prevalence was higher in southern the distribution and prevalence of C. states where both tick species are pres- felis in bobcats (Lynx rufus) and other ent. The prevalences in Kansas (41 bob- wild/exotic felids from ten eastern cats), Kentucky (74 bobcats), Louisiana states (Georgia, Kansas, Kentucky, (1 bobcat, 1 cougar [Felis concolor],

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Habitat Restoration

Bottomland Hardwood and Riparian Restoration in Obion Creek/Bayou de Chien Watersheds

Jeff Sole and Shelley Morris, The Nature Conservancy

he Obion Creek/ TBayou de Chien watersheds have been identified by multiple agencies and organizations as high- priority conservation areas. The majority of the bottomland hardwood forests within these two watersheds have been negatively impacted by incom- patible forestry or agricultural practices; consequently, most of the streams have very little riparian vegetation and are in need of restoration at- TNC biologist discussing habitat improvement practices in Obion Creek bottoms / Nathan Hicklin tention. Beginning in 2008, we sought to achieve the follow- burns to improve habitat for grassland Comprehensive Wildlife Conserva- ing conservation objectives on private songbirds within this project area. Ad- tion Strategy: Appendix 3.4, Pri- lands within these two watersheds: ditionally, we planted bottomland hard- oritized taxa-specific conservation restore four miles of riparian cover, wood trees on private property along actions, Class Mammalia; Appendix promote implementation of streambank Brush Creek, and will plant riparian 3.3, Conservation Action # 7, #14, stabilization projects such as grade sta- buffers on this same property during fall #32, #80, #97 #120, and #129. bilization, cedar revetments, and rock 2009. Also planned for fall 2009 is a veins, plant 150 acres of bottomland riparian restoration project along Cane hardwoods, conduct prescribed burns to Creek, a tributary to Bayou de Chien. improve habitat in target areas, and cre- We plan to continue this project ate ephemeral pools for pond-breeding through 2010 and hope to partner with amphibians. Restoring riparian cover federal, state, and local partners to and improving wildlife habitat in these restore habitat, improve water quality, watersheds will reduce sedimentation and abate threats to species of greatest and will help improve habitat for sev- conservation need within the Obion eral species of greatest conservation Creek/Bayou de Chien Watersheds. need including the relict darter and the federally listed Indiana Bat. Funding Sources: State and Tribal In 2008, we partnered with Quail Wildlife Grant (SWG), The Nature Unlimited to implement prescribed Conservancy

64 Kentucky Department of Fish & Wildlife Resources Habitat Restoration / PROJECT HIGHLIGHTS

Restoration of Bur Oak on the Clay Wildlife Management Area by Means of Direct Seeding

Wes Mattox, Nathan Gregory, Brian Wagoner, Scott Freidhof, and Stephen Bonney, Kentucky Department of Fish and

ur oak (Quercus macro- Bcarpa) is native to at least 103 Kentucky counties (Stein, Binion, and Acciavatti, 2003). It is an especially important tree in the Bluegrass Region of the state where it was historically a dominant component of the Bluegrass Ash-Oak Savannah community described by Braun (1950). Bur oak is commonly found growing on broad fertile ridges of the in- ner Bluegrass ecoregion and to a lesser extent in fertile bottom land within the rougher topography of Bur Oak seedlings in Blue-X direct seeding shelters / Wes Mattox the Hills of the Bluegrass. The Clay Wildlife Management Area (CWMA) er. Unfortunately, the many long, fertile past, leaving very little of their original lies within the Hills of the Bluegrass river bottoms located on the CWMA forest cover. Some of these bottoms ecoregion and borders the Licking Riv- have all been cleared and farmed in the have grown back to forest in recent

Table 1. Bur Oak Seed Collection Location Characteristics

# of Landscape Location County Ecoregion Soil* Elev. Trees Position Oxbow Road 2 Nicholas Hills of the Bluegrass NE mid-slope Eden flaggy silty clay 870’ Carpenter 1 Nicholas Outer Bluegrass NE slope Cythiana-Faywood 880’ Road #1 Carpenter 1 Nicholas Outer Bluegrass E road bank ridge Lowell silt loam 865’ Road #2 Cemetery 2 Nicholas Outer Bluegrass N lower slope Lowell silt loam 835’ Carlisle 1 Nicholas Hills of the Bluegrass Narrow ridge Faywood silty clay loam 920’ Bee Lick 1 Robertson Hills of the Bluegrass Creek Bottom Nolin silt loam 700’ Johnson Creek 2 Robertson Hills of the Bluegrass Creek Bottom Nolin silt loam 660’ Bridge HWY 324 1 Mason Outer Bluegrass Broad ridge Lowell silt loam 940’ Lewisburg Lowell-Faywood silt 3 Mason Outer Bluegrass W lower slope 780’ Baptist Church loam

Annual Research Highlights 2008 PROJECT HIGHLIGHTS / Habitat Restoration

years but these areas are dominated by deep approximately 8 feet apart in three and bare root seedlings. Preliminarily, light seeded species such as sycamore fields and one forested area. Blue-X rodent predation of seedlings has been (Platanus occidentalis) and box elder direct seeding shelters were placed over surprisingly low (<4%). However, this (Acer negundo). Black walnut (Juglans 750 of the planted acorns. Approxi- will be reassessed later this spring nigra) is typically the only hard mast mately 300 acorns were planted without (2009) after germination is complete producing species found in these bot- shelters but were marked with stakes. and seedlings have emerged. tomland forests. The purposes of this An additional 200 bare root bur project are to; 1) restore an important oak seedlings were planted March 4, Literature Cited native hard mast producing species to 2009 for comparison. Additional treat- Braun, E. L. 1950. Deciduous Forest the bottomland forests of the CWMA ments to the planting sites are planned of Eastern North America. Hafner and 2) to test the efficacy of direct for the spring and summer of 2009 and Press, New York. seeding of acorns compared to planting include treating fescue competition bare root seedlings. with clethodim herbicide and reducing Stein, J., Binion, D., and Acciavattii, Acorns were collected on the 2nd, the basal area within the forested plan- R. 2003. Field Guide to Native Oak 9th, and 17th of October 2008 from 14 tation to provide optimum light levels Species of Eastern North America. different bur oak trees in 9 locations in for seedling growth. USDA, Forest Service, FHTET- Nicholas, Mason, and Robertson Coun- These plantations will be moni- 2003-01, page 55. ties. The acorns were inspected and tored over the next several years to de- those that were desiccated, rotten, or termine rates of predation and browsing Funding Source: Kentucky Department excessively damaged by weevils were on acorns and seedlings, germination of Fish and Wildlife Resources discarded. The acorns were stored in rates of acorns, and to compare growth peat moss until planting on October 20th rates between seedlings with and with- KDFWR Strategic Plan Goal 1. Stra- and 23rd. Acorns were planted 1-2 inches out shelters and between direct seeded tegic Objectives 2 and 3.

Table 2. Plantation site descriptions

Size Location Cover Soil* Site Index* Elevation # of Acorns (acres) NRO = 80 Field 1 1.2 Mowed briars Elk silt loam 620’ 309 YP = 90 NRO = 80 Field 2 .7 Mowed fescue Allegheny loam 620’ 269 YP = 83 NRO = 80 Field 3 1.5 Mowed fescue Allegheny loam 620-630’ 300+ YP = 83 SG = 99 Field 4 .4 Woods Nolin silt loam 610’ 136 ECW = 72 NRO = 80 Field 5 .4 Fallow Field Allegheny loam 620-630’ 200 seedlings YP = 83

*Soil and Site Index data taken from Soil Survey of Nicholas and Bourbon Counties Kentucky. 1982. USDA, Soil Conserva- tion Service. (NRO = Northern Red Oak, YP = Yellow Poplar, SG = Sweetgum, and ECW = Eastern Cottonwood)

66 Kentucky Department of Fish & Wildlife Resources Habitat Restoration / PROJECT HIGHLIGHTS

Natural Grassland Survey of the Original Barrens-Prairie Region of Kentucky

Brian Yahn, Ecologist, Kentucky the Interior Low Plateau Karst Priority habitat remaining and provide informa- State Nature Preserves Conservation Area (ILPCA) over a 3 tion to refine and delineate regions of conservation focus. Commission year period (2008 – 2011). Identifying grassland habitats will take a 4-step ap- The project started in mid-August KDFWR Contact: Danna Baxley proach. First, existing data on grassland of 2008 and focused in Hardin, Larue sites will be collected and entered into and Grayson counties. The flight plan t the time of early settlement, a GIS database. Second, color aerial included 58 sites which were surveyed AKentucky had an estimated 2.5 to photography will be analyzed to select by air. After inspection, 36 of these 3 million acres of natural grasslands potential areas not previously identi- sites were visited on the ground. These (prairies) and open woodlands (barrens) fied. Third, sites selected in steps 1 and sites were then scored by evaluating that were common in the Pennyroyal/ 2 will be organized into a flight plan. 6 factors: habitat quality, species rar- Mitchell Plain and the Coastal Plain These sites will be flown-over (via heli- ity, invasive species abundance, size, regions and scattered throughout sur- copter) and inspected. Fourth, only se- landscape context and woody species rounding areas (including the Eastern lected sites from step 3 will be ground- encroachment. After scoring all the and Western Highland Rim and Shaw- truthed. Qualitative ground surveys sites visited, six scored high enough to nee Hills). These prairies and open will further identify the highest quality be considered of higher-quality prairie/ woodlands supported glade/barren habitat. So far, a wide diversity of these 6 were mostly small (< wildlife species. Many 50 acres) with thin, forested of these species are buffers. Outside of the thin, now rare or declin- forested buffers, most sites ing in Kentucky due were surrounded by crop ag- to the destruction of riculture (of low restoration the grassland habitat potential). that supported them. The project continues This includes species in 2009 with focus areas in such as the Henslow’s parts of Barren, Butler and sparrow, Lark spar- Grayson and all of Edmon- row, Short-eared owl, son and Hart counties. The Northern Harrier, East- same methodology is being ern corn snake, Eastern applied. slender glass lizard, and Six-lined racerun- Funding Sources: State and ner, as well as many Tribal Wildlife Grant (SWG), others. (These species Kentucky State Nature Pre- are listed by KDFWR serves Commission and The as species of great- Nature Conservancy est conservation need (SGCN)). Kentucky’s Comprehen- We propose to sive Wildlife Conservation identify remaining Strategy: Appendix 3.8, natural grassland and Terrestrial habitat guild woodland habitats that (Grassland/agriculture), harbor and sustain Objectives 1, 3, 4 and 5. these rare and declin- (vol. III - pgs. 63-65). ing wildlife species (SGCN). We will focus this inventory within Grassland / Brian Yahn

Annual Research Highlights 2008 Gadwall / USFWS Project Updates

68 Kentucky Department of Fish & Wildlife Resources / PROJECT UPDATES

Conservation Status and Habitat of the Longhead Darter (Percina macrocephala) in Kinniconick Creek, Kentucky

David J. Eisenhour, Joshua M. Schiering, and A. M. Richter, Morehead State

anagement of a rare Mfish requires knowledge of its past and present distribution and population size, in addition to basic biological information, such as habitat use. These are lacking for many nongame fishes, Longhead darter (Percina macrocephala) / David J. Eisenhour including the longhead darter (Percina macrocephala), a Kentucky level (1 m2) to examine differences in vulnerable to sedimentation, we recom- endangered species. This study exam- microhabitat usage. Data collection and mend that management strategies focus ines the distribution and population analysis of population trends have been on reducing sediment input. Currently, size and habitat use of longhead darters completed; analysis of habitat data is most sediment appears to be entering in Kinniconick Creek, Lewis County, nearly complete. from major tributaries, which suffer Kentucky. Data generated from this We found longhead darters in 15 from extensive (and presumably illegal) study will be used to develop manage- of 55 reaches sampled, extending the gravel mining and channelization. In ment plans that protect this species and known range to 50 stream km in Kin- addition, the upstream portion of the maintain biodiversity in Kinniconick niconick Creek. The highest densities population appears to exhibit source- Creek. occurred in the middle part of Kin- sink dynamics, which suggests migra- Prior to this survey, longhead dart- niconick Creek, which has the lowest tion to replace periodic, local extirpa- ers were known from Kinniconick Creek land disturbance. A total of 93 darters tions is necessary. We suggest that road from only eight specimens, all collected were encountered during the survey, crossings be built or modified so that prior to 1982; the present status of this including both young-of-the-year and they permit darter movements. endangered species was unknown in the adults, indicating successful reproduc- creek. In 2007 and 2008 we surveyed, tion and recruitment. We conservatively Funding Source: State and Tribal via canoe, 69 stream km of Kinniconick estimate the population in Kinniconick Wildlife Grants Program (STWG) Creek, dividing it into 198 sampling Creek to be 2000-5000 individuals. reaches, with each reach extending Although still an uncommon and local- KDFWR Strategic Plan. Goal 1. from the crest of one riffle to the crest ized fish within the system, it has an Strategic Objective 5. Comprehen- of the next riffle. At 55 randomly se- apparently stabile population. Long- sive Wildlife Conservation Strategy: lected reaches we sampled for presence head darters were typically found at the Appendix 3.9; Class Actinopterygii and abundance of darters and collected ends of long pools in areas of moderate and Cephalaspidomorphi: Taxa spe- habitat data. Darter sampling was done depth (0.3-0.8 m), little or no current, cific research project #1. primarily by snorkeling, supplemented and substrates of clean boulders and by seining and backpack electrofishing. cobbles. Preliminary analysis of habitat Fourteen habitat variables were collected data suggest that longhead darters are at the macro (reach) level to examine associated with reaches having long differences between reaches with and pools, and avoid areas with high silt without darters. Five additional habitat or sand or strong currents. Because variables were collected at the micro their habitat (low flow areas) is highly

Annual Research Highlights 2008 9 PROJECT UPDATES /

Assessing Avian Use of Land Enrolled in Conservation Practice 33 (CP-33), Conservation Reserve Program

Kate Heyden and Shawchyi sis tested for significant differences in densities between control and CP-33 Vorisek, Kentucky Department fields for the following species: Ameri- of Fish and Wildlife Resources; can robin, song sparrow, white-throated Wes Burger, Mark Smith, and sparrow, field sparrow Spizella( pusilla), Kristine Evans, Mississippi State swamp sparrow (Melospiza georgiana), University and other sparrows (American tree sparrow (Spizella arborea), savannah sparrow, white-crowned sparrow (Zonotrich- n 2004, the USDA’s Conservation ia leucophrys), fox sparrow (Passerella IReserve Program (CRP) introduced iliaca) and unknown sparrow). a new conservation practice called CP- In 2007-2008, field sparrow and 33. This conservation practice provides swamp sparrow had significantly habitat buffers for upland birds, and is greater densities at the CP-33 treatment designed to supply food and cover for transects than at the control transects. Song Sparrow / Kate Heyden bobwhite quail and other avian species Song sparrow, white-throated spar- in areas dominated by agricultural row row, other sparrow and American robin being twice as large, on average, than at crops. Specifically, CP-33 involves exhibited notably greater densities at control sites. planting native warm-season grasses, CP-33 treatment sites, but these values forbs, legumes, and a limited amount of were not significantly different from Funding Sources: State and Tribal shrubs around cropland field edges. To control site densities (Figure 1). Aver- Wildlife Grants (SWG) assess the actual value of CP-33 areas age cluster size (group size) was larger to wildlife, KDFWR coordinated moni- at CP-33 treatment transects than at KDFWR Strategic Plan. Goal 1. toring for winter use of CP-33 fields by control transects for all species in- Strategic Objective 5. Comprehen- avian species in 2007 and 2008. cluded in the analyses. Field sparrows sive Wildlife Conservation Strategy: In two field seasons, observers showed the most drastic comparison in Appendix 3.4; Class Aves: Taxa spe- completed 128 transect surveys (63 in cluster size with groups at CP-33 sites cific conservation action #1. CP33 fields and 65 in control fields). Surveys occurred in 15 counties during Figure 1. The density of wintering birds on CP-33 sites vs. control sites in January-March. Sixty-five species were Kentucky during 2007-2008. detected overall and nine Species of Greatest Conservation Need (SGCN), as listed in our State Wildlife Action Plan (SWAP), were observed: Ameri- can black duck (Anas platyrhynchos), American woodcock (Scolopax minor), loggerhead shrike (Lanius ludovi- cianus), northern bobwhite (Colinus virginianus), northern harrier (Circus cyaneus), red-breasted nuthatch (Sitta canadensis), rusty blackbird (Euphagus carolinus), savannah sparrow (Passer- culus sandwichensis) and short-eared owl (Asio flammeus). Data from 2007 and 2008 were pooled for analysis in 2008. The analy-

70 Kentucky Department of Fish & Wildlife Resources Migratory Shorebird Updates / PROJECT UPDATES

Avian Influenza Monitoring Throughout Kentucky

Rocky Pritchert, John Brunjes and Erin Harper, Kentucky Department of Fish and Wildlife Resources

vian influenza (AI), also known Aas “bird flu,” is caused by several different strains viruses commonly found in some species of water birds. AI viruses are labeled as either highly or lowly pathogenic depending on the response of poultry to the infection. Although many different subtypes of avian influenza exist, the highly pathogenic AI, H5N1 (HPAI H5N1) has been the cause of global concern as a potential pandemic threat. First detected in Southeast Asia in 1996, it has spread across the Asian continent and into Europe and Africa. Millions of poultry have been killed by this subtype throughout the regions of the world where the virus has been isolated. But to date, HPAI H5N1 has not been found in North America. While HPAI H5N1 is primarily an avian virus, there Avian influenza sampling / John Brunjes have been instances in which human infections have occurred. Most people primary objective of this program is to all of which have tested negative for have been infected through very close partner with states to embark on sam- the high-path H5N1 strain of avian in- contact with infected birds, usually pling and surveillance efforts sufficient fluenza. Represented in this sample are poultry. There is concern about the pos- to detect this pathogen should it exist 415 hunter-killed birds, 50 live birds. sibility of genetic exchange between within the Unites States. By collecting Primary species surveyed include mal- human and avian viruses. If this genetic cloacal and tracheal samples from wa- lards, wood ducks, green-winged teal, exchange were to occur, the potential terfowl, shorebirds, and other migratory Canada geese and 9 other waterfowl result would be a novel, possibly lethal, birds and submitting these samples to species. Under the current 2009 AI virulent influenza strain that is easily the National Animal Health Laboratory monitoring plan broad scale monitoring transmissible from human to human. Network (NAHLN) for testing, early in If deemed necessary by the USDA, Thus far, approximately 400 human detection of the high-path H5N1 strain we plan to continue these surveillance infections have been documented with is possible. States are separated into efforts in the Commonwealth in 2008. most infections resulting in death from one of three tiers (I, II or III), depend- HPAI H5N1. Disease-control centers ing upon the suspected vulnerability Funding Source: United States worldwide have prioritized the study to appearance of HPAI H5N1 strain. Department of Agriculture (USDA) of this pathogen. As a result of world- Kentucky is a tier III region, and thus wide concern regarding the presence labeled with a low probability of viral KDFWR Strategic Plan. Goal 1, and transmission of the HPAI H5N1, importation. In Kentucky, KDFWR Strategic Objective 1c. Comprehen- the U.S. Department of Agriculture collected approximately 465 samples sive Wildlife Conservation Strategy: (USDA) has launched a national avian during the 2008-09 period, from a Appendix 3.3; Priority Conservation influenza surveillance program. The broad geographic area within the state, Action #132.

Annual Research Highlights 2008 PROJECT UPDATES / Migratory Shorebird Updates

Migratory Shorebird, Colonial Water Bird and Woodcock Investigations

Erin Harper, John Brunjes to attach more transmitters to breeding February and 3 March when a total of and Rocky Pritchert, Kentucky adults. These efforts will help us man- 10,455 cranes were counted. Few to age for future problems created by the no cranes were observed after the first Department of Fish and Wildlife night-herons. week of March. Surveys are important Resources During migration, greater sandhill because they provide information to cranes (Grus Canadensis tabida) travel develop a chronology of migration n spring 2008, we began monitor- in large flocks through central regions through Kentucky and will be useful Iing a colony of black-crowned of Kentucky on there northward route in assessing changes in abundance and night-herons (Nycticorax nycticorax), to the Great Lakes and Canada. Sand- distribution. which have become a nuisance, nest- hill cranes have been using migratory These monitoring and survey ef- ing in suburban Louisville. They have stopover sites in Kentucky for many forts will provide a platform by which recently expanded their numbers and years. One of these major sites is east long-term trends may be assessed, and range within Kentucky. Their droppings of Cecilia in Hardin County where management actions that are consis- and waste pose health and aesthetic there are abundant grain fields with tent with long-term welfare of specific concerns for citizens living in this area. water-filled ditches or depressions populations may be implemented. Over 80 breeding pairs were nesting that provide food for the thousands of within a 0.2 mi2 area in Germantown. cranes that stop and rest. We conducted Funding Source: State Wildlife Grant In May 2008, nests were removed by 10 surveys between 5 December 2008 Program (SWG) USDA Wildlife Service personnel, with and 3 March 2009 at this site. We the hope that the night-herons would counted 15,262 sandhill cranes. The KDFWR Strategic Plan. Goal 1. renest elsewhere. About 75% of the peak migration occurred between 17 Strategic Objective 5. pairs renested, with some moving further south and expanding the colony among a 0.3 mi2 area. A large percentage of nests were successful. Young night-herons were seen later in the summer walking among tree branches waiting for parents to bring food. We captured and attached leg bands and transmitters to two adult night-herons in order to monitor behavior and movements. One transmitter was removed due to complications, while the other bird continued to roost and forage until it migrated south in October. With no nest or mate to tend to, it proved to be a non-breeding adult. In addition to continued monitoring efforts in 2009, we plan Black-crowned night heron / John Brunjes

72 Kentucky Department of Fish & Wildlife Resources Migratory Shorebird Updates / PROJECT UPDATES

Monitoring and Management of Kentucky’s Waterfowl

Rocky Pritchert, John Brunjes, each species observed along the routes. Western Kentucky Purchase Region are Thomas Young and Pat Hahs, In addition, observers record the number consistently greater than elsewhere in of coots, sandhill cranes and eagles en- the Commonwealth. These periodic wa- Kentucky Department of Fish countered along each route. Surveys are terfowl surveys are extremely because and Wildlife Resources conducted along established flight routes they provide much-needed information and ducks (dabblers and divers), geese pertaining to population sizes, species eriodic aerial surveys have been (Canada, snow, and blue), and coots are distribution and patterns of winter use on Pimplemented by KDFWR to moni- counted during each flight. Bald eagles Kentucky wetland habitats. Without this tor winter waterfowl populations across are also counted, when present, on these information, waterfowl management in the state. These wide-reaching surveys survey routes. Over the 2008/2009 win- Kentucky would be a daunting task. We are conducted about once every two ter period, we conducted six waterfowl anticipate that these waterfowl surveys weeks from late November through Feb- surveys between 29 November 2008 will continue for the 2009/10 winter ruary. All known major migration/win- and 15 February 2009. KDFWR observ- period and beyond, to ultimately provide tering waterfowl concentration areas are ers recorded 822,764 ducks, 135,722 land managers and biologists with data surveyed. These areas include the Mis- Canada geese, 112,099 snow geese, addressing long-term waterfowl trends sissippi River, Ohio River, major wet- and 177 eagles. The average number within Kentucky. lands in the Purchase and Green River of birds recorded per survey during the regions, as well as the major reservoirs 2008/09 wintering period was 137,127, Funding Source: Kentucky Depart- of central and northeastern Kentucky. 41,303 and 30, ducks, geese and eagles, ment of Fish and Wildlife Resources Survey observers identify birds to spe- respectively. Although waterfowl popu- cies ducks, geese, and swans whenever lations exist within each of these survey KDFWR Strategic Plan. Goal 1, Ob- possible and record the total number of routes, waterfowl populations within the jective 5.

Waterfowl / John Inman

Annual Research Highlights 2008 PROJECT UPDATES / Migratory Shorebird Updates

Monitoring Canada Goose Populations in Kentucky

Rocky Pritchert, John During that period Brunjes, and Erin Harper, Kentucky has observed Kentucky Department slow but steady growth in the number of tem- of Fish and Wildlife perate nesting Canada geese in the state. aterfowl managers use In 1994, Kentucky’s Wbanding data to assess spring population was survival and harvest rates, mi- estimated to be about gration routes, speed and tim- 18,000 total Canada ing of migration, population geese with the num- distribution and numerous basic ber of breeding pairs life history parameters. Ken- estimated to be about tucky Department of Fish and 3,000. The estimated Wildlife Resources have banded Canada goose roundup / Joe Lacefield population from the Kentucky nesting Canada geese 2008 survey was about since the early 1980’s. According to the birds captured in Kentucky). Ken- 32,500 in the surveyed area with about Ducks Unlimited, the Canada Goose tucky’s growing population of Canada 10,000 breeding pairs. The slow growth (Branta Canadensis) is second only to geese has afforded increased hunting of this population compared to that mallard ducks (Anas platyrhynchus) in opportunities across the state. Several observed in northern states can possibly the number of birds banded in North decades ago, hunting opportunities be attributed to two factors. The first, America with nearly 3.0 million birds were largely limited to the western por- Kentucky’s population of temperate banded through 2008. Although Canada tion of the state; today, giant Canada nesting Canada geese has a relatively Goose populations are globally secure, geese are present on water bodies state- low reproductive rate as observed by Kentucky, Virginia, Alabama, and Loui- wide. Through the continued collec- the low number of goslings banded siana list this some populations of this tion of data from capture and banding annually. The second factor may be species as vulnerable (S3 designation; efforts, KDFWR will be able to make related to harvest during the hunting Natureserve Explorer). Fortunately, informed management decisions re- season. KDFWR over the years has prudent management efforts in Ken- garding season dates and bag limits and slowly expanded Canada goose hunt- tucky as well as across the flyway have insure a sustainable and healthy popu- ing opportunities to be consistent with resulted in an increase in the number lation of these birds in Kentucky. the long-term welfare of the resource. of temperate nesting Canada geese in In addition to banding Canada Since the 1990’s the proportion of Ken- the Commonwealth, especially in the goose banding efforts, aerial spring tucky nesting temperate Canada geese central and northeastern portions of surveys are employed by KDFWR to harvested by area hunters has increased the state. The Kentucky Department of evaluate population status of nesting from 8% in 1990 to 52% of the total Fish and Wildlife Resources has been Canada geese. Since 1994, Mississippi harvest in 2007. Given the slow but a long-time participant (since the mid Flyway states have surveyed temperate steady growth of the population, this 1980s) in Canada Goose banding ef- nesting (resident) Canada goose popu- is consistent with KDFWR temperate forts. In 2008, a total of 1,359 Canada lations annually and have used the data nesting Canada goose management ef- geese were captured and banded in to assess the overall status of this im- forts. The information provided by this Kentucky and another 5,377 Canada portant goose population on both state- survey is critical for continued monitor- geese that were captured, banded and wide and flyway wide basis. In turn the ing and management of this population translocated to Kentucky from Michi- information is used to make harvest and KDFWR will again conduct spring gan were banded by KDFWR person- management decisions with focus on surveys in 2009. nel. Canada geese were banded from 25 temperate population with limited im- sites in Kentucky, representing 1,245 pacts on other migrant Canada goose Funding Source: Non Federal Aid (NFA) adults and 108 hatch year birds. The populations. KDFWR has conducted sex ratio for these birds approached spring temperate nesting Canada goose KDFWR Strategic Plan. Goal 1, Ob- 50:50 (we were able to sex all but 6 of surveys since their initiation in 1994. jective 5.

74 Kentucky Department of Fish & Wildlife Resources Migratory Shorebird Updates / PROJECT UPDATES

Mourning Dove Banding in Kentucky

John Brunjes, Kentucky Department of Fish and Wildlife Resources

lthough mourning doves are con- Asidered one of the most common bird species in eastern North America, long-term, nationwide monitoring (beginning in 1966) has revealed downward population trends. Given the economic and social importance of mourning doves, KDFWR, along with 32 other states nationwide, initiated a dove banding program in 2002 to better monitor and mange Kentucky’s and the nations mourning dove population. To date, more than 10 thousand bands have been applied to mourning doves in all five regions of Kentucky. In 2008, banding efforts occurred between July 1st and August 31st, and we successfully banded 1,837 birds from 61 sites statewide. Of these captured birds, 968 were adults and 869 were young birds, and we were able to determine the sex of 916 birds of which 66% were male and 33% were female. Interest- ingly, young birds were very difficult to catch until late in the trapping period with only 103 young birds banded by August 1. This may be explained by heavy early season rainfall which could have caused early season nest failures. Hunters reported the harvest of 70 banded in 2008, and have recovered Mourning dove banding / KDFWR over 600 bands since the onset of this program. One hunter harvested a bird representing 19 sites were obtained tory data, we will be able to make op- banded as an adult in 2004. Although from collection barrels, check stations timally informed harvest management most recoveries occurred in Kentucky, manned by KDFWR personnel. Ap- decisions. doves banded in Kentucky have also proximately 5000 additional wings been recovered in Illinois, Indiana, were obtained from mail surveys to Funding Source: Kentucky Department Arkansas, Louisiana, Missouri, Ohio, hunters. Valuable life history data were of Fish and Wildlife Resources West Virginia, North Carolina, South obtained from these wings including: Carolina, Georgia, Florida, Tennessee, age at harvest, sex ratios of harvested KDFWR Strategic Plan. Goal 1, Ob- Mississippi, Oklahoma, Texas, and birds and information on the timing of jective 5. North Dakota. nesting activities. We plan to continue In addition to banding, KDFWR banding and monitoring Kentucky’s also collects wings from harvested mourning dove population in 2009. By mourning doves. In 2008, 4,563 wings gathering regionally-specific life his-

Annual Research Highlights 2008 PROJECT UPDATES / Migratory Shorebird Updates

Proactive Wood Duck Monitoring and Management in Kentucky

Rocky Pritchert, John Brunjes, and Erin Harper, KDFWR

ood ducks are the most common Wbreeding duck in Kentucky, and 2008 marked yet another year of successful wood duck management by the Kentucky Department of Fish and Wildlife Resources. Historically, KD- FWR has proactively managed wood duck populations through banding, nest box programs, spring stream sur- Female green-winged teal and male wood duck / Joe Lacefield veys, and careful regulation of hunting. These management efforts have been average of 5.4. The 0.36 broods per harvest and survival rates, as well as har- important in restoring wood duck popu- mile and was similar to 0.34 broods vest distribution for local nesting wood lations, which approached nearly un- per mile observed in 2007, but well ducks. Since 1980, KDFWR has banded sustainable lows by the mid 1950’s due below the 0.47 reported in 2006 and between 1,700 - 3,000 birds annually, to over-hunting and habitat destruction. also lower than the long-term average after the primary nesting period and Fortunately, through prudent manage- of 0.4. However the lower brood counts prior to the September hunting season. ment efforts wood ducks are now com- are probably reflective of the drought In 2008, KDFWR captured 2,378 wood mon, permanent residents of Kentucky. experienced over the last two nesting ducks and placed leg bands on 2,035 of In addition to long-term banding pro- periods and these counts fall within them. Wood ducks were captured from grams, KDFWR has conducted annual the range of experience encountered numerous locations across the state. stream surveys to assess reproductive during the history of the survey. While Most wood ducks (688) were banded in effort across the state. Wood ducks are wood duck reproductive was somewhat the Green River region, but substantial secretive ducks which prefer to use depressed the last two years because of numbers (13%) were banded in eastern forested wetlands and mature wooded the drought, overall Kentucky’s wood regions of the state as well. Banding ef- stream/river riparian zones which are duck population is healthy and the fort was stratified by regions of harvest not easily surveyed using conventional trend remains stable. and this effort has been successful in aerial survey methods. Wood duck In addition to wood duck stream sur- identifying the southern populations as stream surveys are one of the longest veys, KDFWR also has a summer band- that segment Kentucky’s birds are most running surveys currently conducted by ing program for wood ducks. The impor- closely associated with. As a result of KDFWR, beginning in the mid 1960’s tance of nesting wood duck populations these banding efforts, KDFWR was it still continues today with some modi- was elevated when Kentucky and two granted operational status of the Septem- fication. In the early years observers other states were permitted to establish ber wood duck season in 2003. KDFWR only reported the number of broods an experimental September wood duck has committed to continued banding as and young per brood, regardless of the season. This early hunting opportunity one of the conditions for maintaining op- number of adults encountered. Today quickly gained favor with Kentucky’s erational status this season. the survey still monitors broods and waterfowl hunters. A major factor for KDFWR will continue to use young but also collects data on adults maintaining the September season was stream counts and band data to monitor such as sex, whether birds are paired that KDFWR would monitor local wood this important Kentucky resource. or whether females are present with duck numbers to evaluate the potential broods. The additional data provides harvest affects. The wood duck’s secre- Funding Source: Kentucky Department KDFWR with better insight into wood tive habits make this species somewhat of Fish and Wildlife Resources duck reproductive efforts. difficult to survey using standard tech- In 2008, the number of young per niques. Therefore, KDFWR established a KDFWR Strategic Plan. Goal 1, Ob- brood was 5.34 and near the long-term summer banding program to determine, jective 5.

76 Kentucky Department of Fish & Wildlife Resources Habitat Restoration Updates / PROJECT UPDATES

Impacts of Herbicide Application Following a Late Summer Burn

Brian Clark, Ben Robinson, a small field previously converted from were probably most affected by the fescue to native grasses and forbs. After herbicide due to timing of application. John Brunjes, Dave Frederick, re-growth of grasses half of the field Native warm-season grasses were more Chris Grasch, and Jim Barnard, was sprayed during early October with clumpy and forbs were more diverse Kentucky Department of Fish Select herbicide at a rate of 12 ounces and predominant 1 year after treatment. and Wildlife Resources per acre, while the eastern/downhill half was left untreated. Suppression Funding Source: Kentucky Depart- of the grasses was visibly evident by ment of Fish and Wildlife Resources t KDFWR headquarters we an- winter and persisted through the fol- Aecdotally gauged the impacts of lowing spring. In July 2008 following Comprehensive Wildlife Conserva- Select herbicide (grass killer, (chemical herbicide application, total grass cover tion Strategy: Appendix 3.3, Priority name Clethodim) on plants following was still noticeably reduced, especially Conservation Action #76. a late summer prescribed burn. In Sep- among cool season grasses such as tember 2007, a burn was conducted on tall fescue and brome species, which

Clethodim-treated (east/left side) vs. untreated side of field at KDFWR headquarters, 9 months following burning and subsequent spraying / Brian Clark

Annual Research Highlights 2008 PROJECT UPDATES / Habitat Restoration Updates

Establishing Shrubs and/or Brambles on Reclaimed Minelands

Scott Harp and Eric Williams, In an attempt to answer these No initial germination from the Kentucky Department of Fish questions, we created two test plots in chill-treated seed was evident in the the summer of 2007: a 0.76 acre plot summer/fall of 2008. However, subse- and Wildlife Resources and a 0.66 acre plot. These sites were quent research indicates that a “warm sprayed with Garlon 3A and disked to stratification” period is typically re- he successful establishment of bare ground in August & September quired before cold stratification in order Tshrubs and/or brambles on mine 2007. Seeds were ordered, soaked for for dormancy to break. The addition of reclamation sites is an important task 24-hours in water, and then placed the third test plot should indicate if that because these habitats provide struc- in cold storage substrate (43 degrees was the cause of the first 2 plots not ture, cover, and foraging areas for many farenheit) the first week of October. germinating. This is a long-term project types of wildlife. Additionally, these They remained in cold storage until which will continue until vegetation be- areas often act as movement corridors planting on June 8, 2008. Seven species comes established in each test plot such for wildlife and may serve as important were planted: Silky Dogwood, Gray that we may evaluate the success of our erosion control components of the land- Dogwood, Smooth Sumac, Elderberry, direct seeding efforts. If direct seeding scape. Unfortunately, it is often costly Coralberry, Blackberry, and Wild Plum. of shrubs/brambles proves to be effec- and difficult to establish and maintain A third 1.0 acre test plot was created in tive, we hope to have this practice valu- shrub/bramble habitats on reclaimed August 2008. This plot was prepared ated/approved by NRCS for potential mining sites. Competition and domina- identically to the previous 2 test plots, inclusion in Farm Bill programs. tion of reclamation sites by Sericea and was seeded with dormant seed on lespedeza, Lespedeza cuneata, an ag- August 15, 2008 to mimic the natural Funding Source: Pittman Robertson gressive invader of open areas, poses a distribution of seeds from fruit over (PR) considerable challenge when attempt- the summer months. Five species were ing to re-establish shrubs and brambles. planted: Silky Dogwood, Gray Dog- Comprehensive Wildlife Conserva- Previous efforts to plant seedlings have wood, Smooth Sumac, Wild Plum, and tion Strategy: Appendix 3.3, Priority not been successful; high mortality Serviceberry. Conservation Action #76. rates are usually present, making these efforts somewhat cost prohibitive. In 2007 we initiated a long-term study to assess success of direct seeding shrubs/brambles on reclaimed mine ground within Peabody Wildlife Management Area. Through this study, we hope to address the following questions: 1) Will direct seeding of shrubs/ brambles result in better survivability than planting seedlings? 2) Will developing shrubs out-complete sericea with only one year of control? 3) Will shrubs develop in a mixed species situation or will one species out-compete the others? 4) Is direct seeding efficient when considering time and money involved when compared to planting seedlings? 5) Will cold storage be sufficient to break seed dormancy and initiate germination, or will planting during natural summer conditions improve germination? Reclaimed mineland / Scott Harp

78 Kentucky Department of Fish & Wildlife Resources KDFWR Contacts

Sawfin shiner / Matt Thomas

ore information regarding Karen Alexy [email protected] Mthe project summaries within Jim Barnard [email protected] this publication can be attained by Danna Baxley [email protected] contacting the KDFWR authors or Stephen Bonney [email protected] contacts listed here. John Brunjes [email protected] Tina Brunjes [email protected] General questions can be directed to: Brian Clark [email protected] The Kentucky Department of Dan Crank [email protected] Fish and Wildlife Resources Steven Dobey [email protected] # 1 Sportsman’s Lane Dave Frederick [email protected] Frankfort, KY 40601 Scott Freidhof [email protected] 1-800-858-1549 Chris Grasch [email protected] [email protected] Brian Gray [email protected] Scott Harp [email protected] Erin Harper [email protected] Kate Heyden [email protected] Charlie Logsdon [email protected] William Lynch [email protected] John MacGregor [email protected] Monte McGregor [email protected] Wes Mattox [email protected] Ryan Oster [email protected] Laura Patton [email protected] Jayson Plaxico [email protected] Rocky Pritchert [email protected] Ben Robinson [email protected] Brooke Slack [email protected] Matt Thomas [email protected] Shawchyi Vorisek [email protected] Eric Williams [email protected]