The Detection and Quantification of Aquatic Reptilian Environmental DNA Clare Isabel Ming-Ch'eng Adams Iowa State University

Iowa State University Capstones, Theses and Graduate Theses and Dissertations Dissertations

2017 The detection and quantification of aquatic reptilian environmental DNA Clare Isabel Ming-Ch'eng Adams Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/etd Part of the Biodiversity Commons, Ecology and Evolutionary Biology Commons, Genetics Commons, Natural Resources and Conservation Commons, and the Natural Resources Management and Policy Commons

Recommended Citation Adams, Clare Isabel Ming-Ch'eng, "The detection and quantification of aquatic reptilian environmental DNA" (2017). Graduate Theses and Dissertations. 15481. https://lib.dr.iastate.edu/etd/15481

This Thesis is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected].

The detection and quantification of aquatic reptilian environmental DNA

Clare Isabel Ming-ch’eng Adams

A thesis submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

MASTER OF SCIENCE

Major: Ecology and Evolutionary Biology

Program of Study Committee: Fredric Janzen, Major Professor Julie Blanchong John Nason Kevin Roe Michelle Soupir

Iowa State University Ames, Iowa 2017

TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ...... iv

ABSTRACT ...... vi

CHAPTER 1. INTRODUCTION ...... 1

CHAPTER 2. COMPARISON OF PAINTED TURTLE (CHRYSEMYS PICTA) ENVIRONMENTAL DNA EXTRACTION TECHNIQUES IN A LENTIC POND SYSTEM ...... 17 Introduction ...... 18 Materials and Methods ...... 21 Results ...... 25 Discussion ...... 27

CHAPTER 3. ESTIMATING AQUATIC REPTILE DENSITY UNDER FIELD CONDITIONS USING ENVIRONMENTAL DNA...... 49 Introduction ...... 50 Materials and Methods ...... 52 Results ...... 57 Discussion ...... 60

CHAPTER 4. SUMMARY AND CONCLUSIONS ...... 88

iii

ACKNOWLEDGMENTS

I would like to thank my major professor, Fredric Janzen, and my committee members, Julie Blanchong, John Nason, Kevin Roe, and Michelle Soupir for their guidance and support throughout the course of this research. Fred, I thank you for giving me the chance and support to become a scientist. I hope to one day view science as you do and develop the skills to see the “big picture” in turtle research, ecology, evolutionary biology, and life. Secondly, I want to thank Julie for being my Preparing Future Faculty mentor and taking the time to sit down and painstakingly work on writing with me. Without your generosity I wouldn’t be where I am now. Lastly, I want to thank Dr. Reg for being my diversity mentor. I arrived here not knowing how to navigate this space and you are the reason I am still here and able to complete my degree. I will miss our chats and your one- liners that helped me get through the day, week, and years.

In addition, I would also like to thank the Janzen Lab. Luke, you are way too kind to me and overly generous with your time. These projects would not be where they are currently without your input, guidance, and help. Rebecca, thank you for looking out for me in the lab.

I appreciate that you took me under your wing and helped me navigate the Janzone. Morgan, you are the best undergrad I have ever had. I cannot wait to see how you grow. I would be nowhere without my friends, especially Andrew and Jermaine. Andrew, thanks for putting up with me for a whole entire year and being an ever-supportive friend. I miss our late-night chats about life overlooking the “skyline” of Ames. Jermaine, words cannot describe how much you mean to me. Thanks for always having my back, yo. You are my emotional lighthouse in the dark. Tori, Monica, Amy G, Hilary – It has been fantastic getting to know iv you and I wish you the best. I wish our time together had been longer, but on to the next great adventure.

Thank you Jack Gallup, the department faculty, and staff for making time and space for me at Iowa State University. Jack, I wish we had met sooner. Additionally, I want to thank my mom and dad for supporting me all the way, especially when times have been tough. Your ability to adapt to my needs is amazing and I am super thankful I got you as my parents. Zach, Kiran, Bradley, thank you for being my oasis whenever I went back to the East

Coast. I miss you so much. I also want to offer my appreciation to turtles, without whom, this thesis would not have been possible. Too often we forget that how we treat nature, in your case – turtles, is a reflection upon humanity itself. Turtles are the best!

ABSTRACT

Monitoring individuals of a population is aided by sampling techniques for determining organism presence. However, invasive sampling methods may harm target and non-target individuals, not capture a fully representative population demographic, or may be difficult to use due to secretive and seasonally active species. Recent developments in non- invasive technology propose environmental DNA (eDNA) as a solution to mitigate some of these challenges. Environmental DNA is DNA captured from target organisms that is extracted from environmental samples such as water, soil, or air. Although this technique has been widely explored for fish and amphibian species, it is used less often for aquatic reptiles.

This thesis attempts to create an eDNA methodology for an imperiled reptilian taxa, turtles, for future monitoring use. We set up four experimental ponds with varying numbers of turtles

(0, 11, 23, 38) and sampled once every three days throughout the spring field season to determine effects of painted turtle (Chrysemys picta) density and time on eDNA technique utility. The first chapter compares two common eDNA methodologies, filtration and sodium acetate precipitation in a Midwestern lentic semi-natural environment. The second chapter uses the more efficient filtration to quantify eDNA based upon turtle abundance in the same environment.

We conclude that eDNA may not currently be an effective monitoring method for aquatic turtles. Overall, filtration was a more effective approach to capturing turtle eDNA than sodium acetate precipitation but visual surveys of turtles in our experimental setup led to an even higher rate of detection. Despite developing a sensitive qPCR protocol, we were unable to amplify turtle eDNA sufficiently to distinguish it from the negative control. We vi nonetheless identified a rank-order trend positively correlated with turtle density despite not obtaining large amounts of species-specific turtle eDNA. Furthermore, we found that total eDNA concentration did not follow a linear pattern throughout the field season. While we cannot recommend eDNA for aquatic turtles at this time, we believe this method could be refined with further technological advances such as better inhibition removers. 1

CHAPTER 1

INTRODUCTION

Population ecology addresses the dynamics of how a group of individuals in the same species interacts with other biotic and environmental factors (Wells and Richmond, 1995). Part of population ecology attempts to describe how groups of individuals in a certain species vary demographically – such as in abundance or density (Hutchinson, 1991; Turchin, 1999). The factors that shape population numbers may be environmental or density dependent (MacArthur,

1958). For example, extreme environmental events such as a heat wave may cause population decline, whereas rate of growth depends on previous population abundance (Kendall, 1949;

Garrabou et al., 2009). The need to predict changes drives the importance in studying these factors; too much growth or decline in numbers of individuals within a population can disrupt an ecosystem or forewarn extinction (Brown et al., 1995; Brooks et al., 2002). Yet, to make the most accurate predictions, ecologists must have some knowledge of a species’ presence and abundance.

One factor in population growth and decline is population abundance itself. Besides density-independent environmental events, populations depend upon reproductively mature adults to prosper (Frankham, 1995). Temporary decline in population abundance can result in demographic changes causing Allee effects or population fragmentation (Courchamp et al.,

1999; Marchand and Litvaitis, 2004; Epps et al., 2005). Differing densities may also lead to the inability to find a mate (Stephens and Sutherland, 1999). Difficulty in finding mates can increase inbreeding whereas genetic drift may additionally reduce allelic diversity (Maruyama and

Fuerstt, 1985; Allendorf, 1986; Charlesworth and Charlesworth, 1987). Both inbreeding and drift 2 may result in less adaptive evolutionary potential overall for the population through fixation of deleterious alleles, reducing population fitness (Charlesworth and Charlesworth, 1987; Caughley,

1994). Conversely, sufficiently large populations with plentiful migration or resources can buffer against these effects, increasing population density and abundance (Kimmel et al., 1998; Sakai et al., 2001).

A growing population’s high abundance may have ecological consequences as well. For example, besides invasive species’ cost to the human economy, they fundamentally change local community biodiversity because their presence, by definition, has impactful biological consequences (Mack et al., 2000; Pimentel et al., 2005; Vilà et al., 2011). Invasive organisms often start as a small population or a few individuals, thus it is imperative to know current abundance to predict future population growth (Mack et al., 2000). A need for sensitive abundance detection is therefore necessary for proper management.

For ecologists determining presence and abundance, methodology must be chosen very carefully. In natural environments, rarely do ecologists have the privilege to obtain a truly random sample of the entire population, especially if using sampling equipment or if the species is rare (Vaughan and Ormerod, 2003; Börger et al., 2006). Furthermore, some species have secretive life histories that help them elude capture. For example, hellbenders (Cryptobranchus alleganiensis) may hide under large river rocks that can make hand capture difficult (Krysko and

Nickerson, 2003). In addition to secretive life histories, specific methodologies themselves may have limitations. For instance, baited hoop traps may yield few juvenile painted turtles

(Chrysemys picta) whereas basking traps favor adult females (Ream and Ream, 1966). Although 3 we may not currently be able to obtain a truly random sample of some populations due to these restrictions, new detection technology inches towards this goal.

To obtain a multi-faceted view of a population, biologists have traditionally used demographic and genetic tools. For example, program MARK estimates population abundance with capture-recapture sampling data (White and Burnham, 1999). Capture methods used to obtain such information for aquatic vertebrates, such as fyke nets, gill nets, transects, hoop nets, and minnow traps, yield population demographics and natural history characteristics such as sex ratio, age class structures, and survival rates (Ryan et al., 2002; Willson and Dorcas, 2004;

Hardie et al., 2006). Built over time, repeated surveys increase the ability to predict metapopulation dynamics using site occupancy modeling (Mackenzie et al., 2003).

Unfortunately, traditional sampling techniques may also harm individuals, even if used properly (Jerde et al., 2011; Evans et al., 2017). Electrofishing may unintentionally harm smaller non-target fish species or even sensitive target species themselves (Snyder 2003). Nets not set up high enough to allow for airflow during a flooded stream may drown air-breathing individuals.

Lethal or invasive sampling (such as drawing blood) may induce stress in an animal (Snyder,

2003; Evans et al., 2017). Furthermore, elusive species may be difficult to sample depending on their natural history. Taken together, the need for more effective, non-invasive sampling tools is clear.

Assessing genetic material increases our understanding of allelic diversity that may work synergistically with demographic data to predict population dynamics. Many programs and 4 statistical methods (BAPS, K-means clustering, MEGA) have been developed for analyzing such data and identifying patterns of genetic structure (Corander et al., 2004; Tamura et al., 2007).

Thus, once changes in allelic diversity, inbreeding, deleterious mutations, and genetic drift are assessed, managers and conservationists can use non-invasive environmental DNA (eDNA) tools to predict population dynamics.

Environmental DNA (eDNA) encompasses genetics techniques developed and used by microbiologists, recently emerging as a population monitoring method for macroecology (Ogram et al., 1987; Steffan and Atlas, 1991). This novel tool promises to minimize the need for contact with individuals of interest (Lodge et al., 2012). It also offers the ability to survey over a wide area. Furthermore, all life stages present in a sample can be detected as a species’ DNA does not change from larval to adult stage (Ardura et al., 2015; Zaiko et al., 2015). Though eDNA is still in its infancy, a few lines of research have emerged from its rapid growth (Wilcox et al., 2013;

Klymus et al., 2015; Goldberg et al., 2016; Valentini et al., 2016). To further refine this technique, some studies focus on the physicality of eDNA and how the physical properties of eDNA interact in water (Barnes et al., 2014). This work includes assessing accumulation and degradation rates, output of eDNA by different life stages of the same species, and how water chemistry interacts with fragmented DNA in the water column (Maruyama et al., 2014; Strickler et al., 2015; Barnes and Turner, 2016). Though eDNA faces methodological challenges yet to be fully resolved, once refined, it has great potential to complement traditional sampling techniques and provide useful insight into population dynamics, particularly aquatic species that are difficult to detect using more traditional techniques.

The current body of eDNA literature has focused mainly on aquatic organisms and can be divided into multiple-species and single-species assay designs. Multi-species approaches usually address one or several taxa and focus on eDNA barcoding – utilizing universal primers and sequencing amplicons to determine species or genus presence (Kelly et al., 2014; Zaiko et al.,

2015; Valentini et al., 2016). Single-species assay designs focus more on invasive or imperiled taxa due to eDNA’s sensitivity and the need to monitor these species (Dejean et al., 2012;

Goldberg et al., 2013). Because of the potential aforementioned sampling and monitoring benefits, eDNA techniques recently have been widely explored in aquatic taxa such as fish and amphibians. Despite the recent boom in eDNA studies, one imperiled taxon has been inadequately addressed– freshwater turtles.

Freshwater turtles (Testudines) are an easily recognized group of shelled reptiles.

Unfortunately, over 60% of turtle species have declined substantially or gone extinct primarily because of habitat loss, poaching, the pet trade, and disease (Gibbons et al., 2000; Böhm et al.,

2013; van Dijk et al., 2014). Due to the high incidence of imperilment, easily discerning population presence and abundance may aid management and conservation efforts for a wide variety of species (Beaudry et al., 2008; Markle and Chow-Fraser, 2014; van Dijk et al., 2014).

Furthermore, some species are challenging to study due to their secretive natural histories, intricate habitats to sample, or complex seasonal activity patterns (Ream and Ream, 1966;

Christiansen et al., 1985). Thus, turtles are a clade that could benefit from the development of eDNA techniques.

The research presented in this thesis addresses ways to identify and quantify painted turtle (Chrysemys picta) eDNA as a model for endangered freshwater turtles. This wide-ranging

North American species overlaps in habitat and distribution with other, more imperiled species such as Blanding’s turtles (Emys blandingii) or spotted turtles (Clemmys guttata) (Ernst and

Lovich, 2009). Furthermore, its mitochondrial genome has been sequenced to allow for the design of primers (Jiang et al., 2016). If eDNA turtle techniques are both successful and sensitive, then the nuclear genome could be explored as well for population genetic eDNA

(Shaffer et al., 2013). Because this species is common and can survive in high densities, it is ideal for facilitating eDNA research. Historically, C. picta has been observed in densities from

7.2 kg/ha to 106kg/ha (Iverson, 1982; Congdon et al., 1986). Thus, the research presented here attempts to create an eDNA system via two major undertakings: creating an eDNA assay to detect turtle presence and then extending the eDNA assay for an abundance-biomass model in freshwater turtles.

Thesis Organization - Abstracts

Chapter 1: Comparison of painted turtle (Chrysemys picta) environmental DNA extraction techniques in a lentic pond system.

Ecological studies rely on species detection, thus it is imperative to identify efficient and effective ways to monitor individuals, especially for conservation. Environmental DNA (eDNA) provides scientists with non-invasive techniques for observing target populations especially when traditional approaches are challenging or inefficient. We tested two eDNA methods, glass microfiber (GMF) and sodium acetate (SA), for extraction efficiency of total eDNA and painted 7 turtle (Chrysemys picta) specific eDNA in a lentic pond system. We seeded isolated, semi- natural ponds with adult turtles and regularly collected water samples from the ponds from 1

April through 30 June 2015. We detected no difference between the two methods in total eDNA extracted (t = -0.94, p = 0.35), although GMF was more likely than SA to detect species-specific eDNA (χ2 = 12.023, p < 0.001). Our results indicate GMF filtration may be better for C. picta detection. Even so, further methodological development is needed before eDNA can be used for population management of aquatic reptiles.

Chapter 2: Estimating aquatic reptile density under field conditions using environmental

DNA.

Abundance monitoring is imperative to understand population fluctuations. One emerging ecological tool for monitoring species abundance is environmental DNA (eDNA), the technique of obtaining target organism DNA from environmental samples such as water, soil, or air. We attempt to determine if eDNA can be used for monitoring aquatic reptile abundance in a semi-natural lentic pond environment. Using four outdoor, experimental ponds with varying painted turtle (Chrysemys picta) densities, we quantified both total eDNA and turtle-specific eDNA for comparison across ponds between 1 April and 30 June 2016. We found no linear effect of date, but detected a non-linear trend, on total eDNA and noted a difference in total eDNA between ponds. We were mostly unable to amplify sufficient turtle-specific eDNA from the water samples, however. Thus, we cannot presently conclude that eDNA is an effective monitoring method for all aquatic reptiles in the wild.

References

Allendorf, F. W. 1986. Genetic drift and the loss of alleles versus heterozygosity. Zoo Biology

5:181–190. doi: 10.1002/zoo.1430050212

Ardura, A., A. Zaiko, J. L. Martinez, A. Samuiloviene, Y. Borrell, and E. Garcia-Vazquez. 2015.

Environmental DNA evidence of transfer of North Sea molluscs across tropical waters

through ballast water. Journal of Molluscan Studies 81:495–501. doi:

10.1093/mollus/eyv022

Barnes, M. A., and C. R. Turner. 2016. The ecology of environmental DNA and implications for

conservation genetics. Conservation Genetics. 17:1-17. doi: 10.1007/s10592-015-0775-4

Barnes, M. A., C. R. Turner, C. L. Jerde, M. A. Renshaw, W. L. Chadderton, and D. M. Lodge.

2014. Environmental conditions influence eDNA persistence in aquatic systems.

Environmental Science and Technology 48:1819–1827. doi: 10.1021/es404734p

Beaudry, F., P. G. deMaynadier, and M. L. Hunter. 2008. Identifying road mortality threat at

multiple spatial scales for semi-aquatic turtles. Biological Conservation 141:2550–2563.

doi: 10.1016/j.biocon.2008.07.016

Böhm, M., B. Collen, J. E. M. Baillie, P. Bowles, J. Chanson, N. Cox, G. Hammerson, M.

Hoffmann, S. R. Livingstone, M. Ram, A. G. J. Rhodin, et al. 2013. The conservation

status of the world’s reptiles. Biological Conservation 157:372–385. doi:

10.1016/j.biocon.2012.07.015

Börger, L., N. Franconi, G. De Michele, A. Gantz, F. Meschi, A. Manica, S. Lovari, and T.

Coulson. 2006. Effects of sampling regime on the mean and variance of home range size

estimates. Journal of Animal Ecology 75:1393–1405. doi: 10.1111/j.1365-

2656.2006.01164.x 9

Brooks, T. M., R. A. Mittermeier, C. G. Mittermeier, G. A. B. Da Fonseca, A. B. Rylands, W. R.

Konstant, P. Flick, J. Pilgrim, S. Oldfield, G. Magin, and C. Hilton-Taylor. 2002. Habitat

loss and extinction in the hotspots of biodiversity. Conservation Biology 16:909–923.

doi: 10.1046/j.1523-1739.2002.00530.x

Brown, J. H., D. W. Mehlman, and G. C. Stevens. 1995. Spatial variation in abundance. Ecology

76:2028. doi: 10.2307/1941678

Caughley, G. 1994. Directions in conservation biology. Journal of Animal Ecology 63:215–244.

doi: 10.2307/5542

Charlesworth, D., and B. Charlesworth. 1987. Inbreeding depression and its evolutionary

consequences. Annual Review of Ecology and Systematics 18:237–268. doi:

10.1146/annurev.es.18.110187.001321

Christiansen, J. L., J. A. Cooper, J. W. Bickham, B. J. Gallaway, and M. D. Springer. 1985.

Aspects of the natural history of the yellow mud turtle Kinosternon flavescens

(Kinosternidae) in Iowa: a proposed endangered species. Southwestern Naturalist

30:413–425. doi: 10.2307/3671274

Congdon, J. D., J. L. Greene, and J. W. Gibbons. 1986. Biomass of freshwater turtles: A

geographic comparison. American Midland Naturalist 115:165–173. doi:

10.2307/2425846

Corander, J., P. Waldmann, P. Marttinen, and M. J. Sillanpaa. 2004. BAPS 2: enhanced

possibilities for the analysis of genetic population structure. Bioinformatics 20:2363–9.

doi: 10.1093/bioinformatics/bth250 10

Courchamp, F., T. Clutton-Brock, and B. Grenfell. 1999. Inverse density dependence and the

Allee effect. Trends in Ecology & Evolution 14:405–410. doi: 10.1016/S0169-

5347(99)01683-3

Dejean, T., A. Valentini, C. Miquel, P. Taberlet, E. Bellemain, and C. Miaud. 2012. Improved

detection of an alien invasive species through environmental DNA barcoding: The

example of the American bullfrog Lithobates catesbeianus. Journal of Applied Ecology

49:953–959. doi: 10.1111/j.1365-2664.2012.02171.x van Dijk, P. P., J. Iverson, A. Rhodin, B. Shaffer, and R. Bour. 2014. Turtles of the World, 7th

Edition: Annotated checklist of taxonomy, synonymy, distribution with maps, and

conservation status. Chelonian Research Monographs 5:329–479. doi:

10.3854/crm.5.000.checklist.v7.2014

Epps, C. W., P. J. Palsbøll, J. D. Wehausen, G. K. Roderick, R. R. Ramey, and D. R.

McCullough. 2005. Highways block gene flow and cause a rapid decline in genetic

diversity of desert bighorn sheep. Ecology Letters 8:1029–1038. doi: 10.1111/j.1461-

0248.2005.00804.x

Ernst, C. H., and J. E. Lovich. 2009. Turtles of the United States and Canada. In.: 2nd edition.

JHU Press.

Evans, N. T., P. D. Shirey, J. G. Wieringa, A. R. Mahon, and G. A. Lamberti. 2017. Comparative

cost and effort of fish distribution detection via environmental DNA Analysis and

Electrofishing. Fisheries 42:90–99. doi: 10.1080/03632415.2017.1276329

Frankham, R. 1995. Effective population size/adult population size ratios in wildlife: a review.

Genetics Research 66:95–107. 11

Garrabou, J., R. Coma, N. Bensoussan, M. Bally, P. Chevaldonne, M. Cigliano, D. Diaz, J. G.

Harmelin, M. C. Gambi, D. K. Kersting, J. B. Ledoux, et al. 2009. Mass mortality in

Northwestern Mediterranean rocky benthic communities: effects of the 2003 heat wave.

Global Change Biology 15:1090–1103. doi: 10.1111/j.1365-2486.2008.01823.x

Gibbons, J. W., D. E. Scott, T. J. Ryan, K. a. Buhlmann, T. D. Tuberville, B. S. Metts, J. L.

Greene, T. Mills, Y. Leiden, S. Poppy, and C. T. Winne. 2000. The global decline of

reptiles, déjà vu amphibians. BioScience 50:653–666. doi: 10.1641/0006-

3568(2000)050[0653:TGDORD]2.0.CO;2

Goldberg, C. S., A. Sepulveda, A. Ray, J. Baumgardt, and L. P. Waits. 2013. Environmental

DNA as a new method for early detection of New Zealand mudsnails (Potamopyrgus

antipodarum). Freshwater Science 32:792–800. doi: 10.1899/13-046.1

Goldberg, C. S., C. R. Turner, K. Deiner, K. E. Klymus, P. F. Thomsen, M. A. Murphy, S. F.

Spear, A. McKee, S. J. Oyler-McCance, R. S. Cornman, M. B. Laramie, et al. 2016.

Critical considerations for the application of environmental DNA methods to detect

aquatic species. Methods in Ecology and Evolution 7:1299–1307. doi: 10.1111/2041-

210X.12595

Hardie, S. A., L. A. Barmuta, and R. W. G. White. 2006. Comparison of day and night fyke

netting, electrofishing and snorkelling for monitoring a population of the threatened

golden galaxias (Galaxias auratus). Hydrobiologia 560:145–158. doi: 10.1007/s10750-

005-9509-9

Hutchinson, G. E. 1991. Population studies: Animal ecology and demography. Bulletin of

Mathematical Biology 53:193–213. 12

Iverson, J. B. 1982. Biomass in turtle populations: A neglected subject. Oecologia 55:69–76. doi:

10.1007/BF00386720

Jerde, C. L., A. R. Mahon, W. L. Chadderton, and D. M. Lodge. 2011. “Sight-unseen” detection

of rare aquatic species using environmental DNA. Conservation Letters 4:150–157. doi:

10.1111/j.1755-263X.2010.00158.x

Jiang, J.-J., E.-H. Xia, C.-W. Gao, and L.-Z. Gao. 2016. The complete mitochondrial genome of

western painted turtle, Chrysemys picta bellii (Chrysemys, Emydidae). Mitochondrial

DNA Part A 27:787–788. doi: 10.3109/19401736.2013.873900

Kelly, R. P., J. A. Port, K. M. Yamahara, and L. B. Crowder. 2014. Using environmental DNA

to census marine fishes in a large mesocosm. PLoS ONE 9:e86175. doi:

10.1371/journal.pone.0086175

Kendall, D. G. 1949. Stochastic processes and population growth. Journal of the Royal Statistical

Society. Series B (Methodological) 11:230–282.

Kimmel, M., R. Chakraborty, J. P. King, M. Bamshad, W. S. Watkins, and L. B. Jorde. 1998.

Signatures of population expansion in microsatellite repeat data. Genetics 148:1921–

1930.

Klymus, K. E., C. A. Richter, D. C. Chapman, and C. Paukert. 2015. Quantification of eDNA

shedding rates from invasive bighead carp Hypophthalmichthys nobilis and silver carp

Hypophthalmichthys molitrix. Biological Conservation 183:77–84. doi:

10.1016/j.biocon.2014.11.020

Krysko, K., and M. Nickerson. 2003. Surveying for hellbender salamanders, Cryptobranchus

alleganiensis (Daudin): A review and critique. Applied Herpetology 1:37–44. doi:

10.1163/157075403766451216 13

Lodge, D. M., C. R. Turner, C. L. Jerde, M. A. Barnes, L. Chadderton, S. P. Egan, J. L. Feder, A.

R. Mahon, and M. E. Pfrender. 2012. Conservation in a cup of water: Estimating

biodiversity and population abundance from environmental DNA. Molecular Ecology

21:2555–2558. doi: 10.1111/j.1365-294X.2012.05600.x

MacArthur, R. H. 1958. Population ecology of some warblers of northeastern coniferous forests.

Ecology 39:599–619. doi: 10.2307/1931600

Mack, R. N., D. Simberloff, W. M. Lonsdale, H. Evans, M. Clout, and F. A. Bazzaz. 2000.

Biotic invasions: Causes, epidemiology, global consequences, and control. Ecological

Applications 10:689–710. doi: 10.2307/2641039

Mackenzie, D. I., J. D. Nichols, J. E. Hines, M. G. Knutson, A. B. Franklin, M. G. Knu, and A.

B. Franklin. 2003. Estimating site occupancy, colonization, and local extinction when a

species is detected imperfectly. Ecology 84:2200–2207.

Marchand, M. N., and J. A. Litvaitis. 2004. Effects of landscape composition, habitat features,

and nest distribution on predation rates of simulated turtle nests. Biological Conservation

117:243–251. doi: 10.1016/j.biocon.2003.07.003

Markle, C. E., and P. Chow-Fraser. 2014. Habitat selection by the Blanding’s turtle (Emydoidea

blandingii) on a protected island in Georgian Bay, Lake Huron. Chelonian Conservation

and Biology 13:216–226. doi: http://dx.doi.org/10.2744/CCB-1075.1

Maruyama, A., K. Nakamura, H. Yamanaka, M. Kondoh, and T. Minamoto. 2014. The release

rate of environmental DNA from juvenile and adult fish. PLoS ONE 9:e114639. doi:

10.1371/journal.pone.0114639 14

Maruyama, T., and P. A. Fuerstt. 1985. Population bottlenecks and nonequilibrium models in

population genetics. II. Number of alleles in a small population that was formed by a

recent bottleneck. Genetics 111:675–689.

Ogram, A., G. S. Sayler, and T. Barkay. 1987. The extraction and purification of microbial DNA

from sediments. Journal of Microbiological Methods 7:57–66. doi: 10.1016/0167-

7012(87)90025-X

Pimentel, D., R. Zuniga, and D. Morrison. 2005. Update on the environmental and economic

costs associated with alien-invasive species in the United States. Ecological Economics

52:273–288. doi: 10.1016/j.ecolecon.2004.10.002

Ream, C., and R. Ream. 1966. The influence of sampling methods on the estimation of

population structure in painted turtles. American Midland Naturalist 75:325–338. doi:

10.2307/2423395

Ryan, T. J., T. Philippi, Y. A. Leiden, M. E. Dorcas, T. B. Wigley, and J. W. Gibbons. 2002.

Monitoring herpetofauna in a managed forest landscape: effects of habitat types and

census techniques. Forest Ecology and Management 167:83–90. doi: 10.1016/S0378-

1127(01)00692-2

Sakai, A. K., F. W. Allendorf, J. S. Holt, D. M. Lodge, J. Molofsky, K. A. With, S. Baughman,

R. J. Cabin, J. E. Cohen, N. C. Ellstrand, D. E. McCauley, et al. 2001. The population

biology of invasive species. Annual Review of Ecology and Systematics 32:305–332.

doi: 10.1146/annurev.ecolsys.32.081501.114037

Shaffer, H. B., P. Minx, D. E. Warren, A. M. Shedlock, R. C. Thomson, N. Valenzuela, J.

Abramyan, C. T. Amemiya, D. Badenhorst, K. K. Biggar, G. M. Borchert, et al. 2013.

The western painted turtle genome, a model for the evolution of extreme physiological 15

adaptations in a slowly evolving lineage. Genome Biology 14:R28. doi: 10.1186/gb-

2013-14-3-r28

Snyder, D. E. 2003. Invited overview: Conclusions from a review of electrofishing and its

harmful effects on fish. Reviews in Fish Biology and Fisheries 13: 445-453.

doi:10.1007/s11160-004-1095-9

Steffan, R. J., and R. M. Atlas. 1991. Polymerase chain reaction: Applications in environmental

microbiology. Annual Reviews of Microbiology 45:137–161. doi: 10.1038/jid.2013.1

Stephens, P. A., and W. J. Sutherland. 1999. Consequences of the Allee effect for behavior,

ecology and conservation. Trends in Ecology & Evolution 14:401–405. doi:

10.1016/S0169-5347(99)01684-5

Strickler, K. M., A. K. Fremier, and C. S. Goldberg. 2015. Quantifying effects of UV-B,

temperature, and pH on eDNA degradation in aquatic microcosms. Biological

Conservation 183:85–92. doi: 10.1016/j.biocon.2014.11.038

Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: Molecular Evolutionary Genetics

Analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24:1596–1599.

doi: 10.1093/molbev/msm092

Turchin, P. 1999. Population regulation: A synthetic view. Oikos 84:153. doi: 10.2307/3546876

Valentini, A., P. Taberlet, C. Miaud, R. Civade, J. Herder, P. F. Thomsen, E. Bellemain, A.

Besnard, E. Coissac, F. Boyer, C. Gaboriaud, et al. 2016. Next-generation monitoring of

aquatic biodiversity using environmental DNA metabarcoding. Molecular Ecology

25:929–942. doi: 10.1111/mec.13428 16

Vaughan, I. P., and S. J. Ormerod. 2003. Improving the quality of distribution models for

conservation by addressing shortcomings in the field collection of training data.

Conservation Biology 17:1601–1611. doi: 10.1111/j.1523-1739.2003.00359.x

Vilà, M., J. L. Espinar, M. Hejda, P. E. Hulme, V. Jarošík, J. L. Maron, J. Pergl, U. Schaffner, Y.

Sun, and P. Pyšek. 2011. Ecological impacts of invasive alien plants: a meta-analysis of

their effects on species, communities and ecosystems. Ecology Letters 14:702–708. doi:

10.1111/j.1461-0248.2011.01628.x

Wells, J. V., and M. E. Richmond. 1995. Populations, metapopulations, and species populations:

What are they and who should care? Wildlife Society Bulletin 23:458–462.

White, G. C., and K. P. Burnham. 1999. Program MARK: survival estimation from populations

of marked animals. Bird Study 46:S120–S139. doi: 10.1080/00063659909477239

Wilcox, T. M., K. S. McKelvey, M. K. Young, S. F. Jane, W. H. Lowe, A. R. Whiteley, and M.

K. Schwartz. 2013. Robust detection of rare species using environmental DNA: The

importance of primer specificity. PLoS ONE. doi: 10.1371/journal.pone.0059520

Willson, J. D., and M. E. Dorcas. 2004. A comparison of aquatic drift fences with traditional

funnel trapping as a quantitative method for sampling amphibians. Herpetological

Review 35:148–150.

Zaiko, A., J. L. Martinez, J. Schmidt-Petersen, D. Ribicic, A. Samuiloviene, and E. Garcia-

Vazquez. 2015. Metabarcoding approach for the ballast water surveillance - An

advantageous solution or an awkward challenge? Marine Pollution Bulletin 92:25–34.

doi: 10.1016/j.marpolbul.2015.01.008 17

CHAPTER 2

COMPARISON OF PAINTED TURTLE (CHRYSEMYS PICTA) ENVIRONMENTAL

DNA EXTRACTION TECHNIQUES IN A LENTIC POND SYSTEM

This is a manuscript currently under review with the peer-reviewed journal Conservation

Genetics Resources

Abstract

Ecological field studies rely on species detection, thus it is imperative for sampling techniques to identify organism presence. Environmental DNA (eDNA) provides scientists with non-invasive techniques for observing target populations especially when traditional approaches are challenging or inefficient. We tested two eDNA methods for extraction efficiency of total eDNA and painted turtle (Chrysemys picta) specific eDNA in a lentic pond system: glass microfiber

(GMF) and sodium acetate (SA). We seeded isolated, semi-natural ponds with adult turtles and regularly collected water samples from 1 April through 30 June 2015. We detected no difference between methods in total eDNA extracted (t=-0.94, p=0.35), although GMF was more likely than

SA to collect painted turtle eDNA (χ2=12.023, p<0.001). Our results indicate GMF filtration may be better for collecting aquatic turtle eDNA. Even so, we cannot recommend either method due to an overall low detection rate (19-42%) of turtle presence. Despite detection, eDNA may not be the best approach for monitoring the presence of freshwater turtles. 18

Introduction

The ability to detect organisms in their environment is a necessary first step for most empirical ecological research. Finding individuals yields a core element of population information, including natural history knowledge about habitat use and behavioral patterns

(Mackenzie et al. 2003; Valeix et al. 2009). Presence data can lead to more targeted efforts and direct sampling to understand community structure, population demographics, and biological interactions. However, finding secretive organisms may be challenging. Target species may be elusive or only seasonally active (Rees et al. 2014; Sigsgaard et al. 2015), habitat may be difficult for researchers to access, and sampling may be labor-intensive (Swales 1987; Spear et al. 2015). Secretive species may not be detected because of small size or because sexes, juveniles, or phenotypes may vary among individuals (Ream and Ream 1966; Rees et al. 2014;

Sigsgaard et al. 2015; Spear et al. 2015; Zaiko et al. 2015b). In addition to detection issues, traditional sampling – often direct capture methods – may stress or harm target species and catch or kill non-target species (Edwards and Jones 2014). The totality of these factors point to the need to develop new, non-invasive sampling techniques (Taberlet et al. 1999).

Non-invasive methods have emerged to address problems associated with traditional sampling (Lodge et al. 2012). These techniques include: surveying urine or fecal samples, collecting host blood from parasites, installing camera traps, and collecting environmental DNA

(Taberlet and Luikart 1999; Taberlet et al. 1999; Valiere and Taberlet 2000; Bohmann et al.

2013; Schnell et al. 2015). Catch per unit effort may increase using non-invasive techniques rather than traditional methods, and furthermore alleviate stress and harm to both targets and non-targets (Lane 2006). Yet, some non-invasive methods cannot perfectly detect individuals, certain life stages, or low populations (Zaiko, Martinez, Schmidt-Petersen, et al. 2015). 19

The non-invasive methodology environmental DNA offers a solution to these issues while providing some of the same benefits as other non-invasive methodologies. Environmental

DNA (eDNA) is DNA obtained from samples collected from areas thought to be inhabited by the species of interest, such as DNA shed into a stream from the sloughed skin of an aquatic amphibian. An eDNA approach can detect the presence and abundance of target species and may even be more sensitive than traditional survey methodologies (Miya et al. 2015; Smart et al.

2015; Wilcox et al. 2016). Additionally, because eDNA is a genetic technique, adults and phenotypically different juvenile stages can be detected because they share the same conserved

DNA sequence, alleviating the problem of observing cryptic life stages when using traditional sampling methods (Zaiko et al. 2015a,b). Furthermore, eDNA can be less labor-intensive and cheaper than traditional methods (Lodge et al. 2012; Yamanaka and Minamoto 2016). Thus, there is incentive to invest in developing these methodologies.

Multiple species-specific and site-specific workflows have been developed for collecting eDNA in aquatic environments. Most general workflows include four steps: (1) a decision about how much water to capture, which usually depends on the concentration and extraction methods

(Goldberg et al. 2016), (2) concentration of the sample, which may be a filtration or centrifugation method (Piaggio et al. 2014), (3) eDNA extraction, usually through Qiagen or

PowerWater extraction kits, phenol-choloroform-isoamyl extraction, or sodium acetate precipitation (Renshaw et al. 2015), and (4) amplification, via one of multiple traditional molecular methods (Doi et al. 2015; Rees et al. 2015). Although this general workflow has been employed, it has yet to be fully validated in the field for many aquatic reptiles and few repeatedly measure the same location to confirm presence. 20

Few field workflows exist for reptile eDNA. Piaggio et al. (2014) published the first successful field protocol for reptiles, identifying eDNA of invasive Burmese pythons (Python bivittatus) in an aquatic Floridian habitat. Using universal vertebrate eDNA primers of the 12S ribosomal gene, Kelly et al. (2014) examined a 4.5 million liter aquarium mesocosm containing two sea turtles but were not able to collect turtle eDNA, perhaps due to low turtle density. Davy et al. (2015) developed primers targeting a CO-1 region of the mitochondrial genome for freshwater turtle species eDNA, but did not test this technique extensively in a field setting. Field water differs substantially from laboratory water because of the presence of PCR inhibitors, non- target DNA, and increased levels of DNA degradation (Strickler et al. 2015). Recent work on flattened musk turtles (Sternotherus depressus) used phenol-choroform-isoamyl extraction to inform a site occupancy model based on the presence or absence of eDNA within sample replicates (de Souza et al. 2016). These lotic sites had relatively clear, slow-flowing water.

Additionally, Lacoursire-Roussel et al. (2016) detected wood turtle (Glyptemys insculpta) presence with species-specific filtration methodology and found its eDNA more plentiful in river systems compared to lake environments. While methodology has been developed for detecting reptile presence in a few specific environments, reptile eDNA detection in lentic pond waters with higher incidence of inhibitor presence has not been successfully validated in a field setting

(Moyer et al. 2014; Jane et al. 2015).

In this study, we describe a method to collect and extract turtle eDNA from a semi- natural pond habitat. We target the painted turtle (Chrysemys picta) as a model organism because its nuclear and mitochondrial genomes have been sequenced (Shaffer et al. 2013; Jiang et al.

2016). Therefore, if one of our eDNA methods is successful, multiple genes could be targeted for population genetic eDNA usage. Additionally, as C. picta is broadly distributed across North 21

America (Ernst and Lovich 2009) our protocol could be used in similar habitats, resulting in further applicability. As this species exists in habitat with endangered species, C. picta may be used as a model for more imperiled species such as the Blanding’s turtle (Emys blandingii) (Reid and Peery 2014). Our goal was to identify a working methodology for turtles occupying a lentic pond environment. We compared the use of (1) a glass microfiber filtration method with (2) a sodium acetate extraction method to determine which technique yielded more total eDNA and turtle-specific eDNA in a semi-natural controlled system. We expected filtration and Qiagen kit extraction to be more effective than sodium acetate extraction because the former collected eDNA from higher volumes of water.

Materials and Methods

Field methodology

We drew water samples from four outdoor experimental ponds (19m L x 15m W x 1.5m

D each) at the Iowa State University (ISU) Horticulture Research Station. We filled the ponds in

October 2014 with water from a nearby lake that contains painted turtles (Table 1). We enclosed each pond with PVC pipe and solid plastic fencing along with an electric fence encircling the entirety of all ponds. We took three samples from experimental ponds during 20-31 March 2015 to provide negative controls before introducing turtles.

We seeded the ponds with adult painted turtles on 1 April 2015 (0, 11, 23, 38 turtles; hereafter Ponds A-D, respectively). Despite all efforts to keep turtles separate, one turtle was seen in Pond A within the first two weeks of the experiment. Thus, our negative control pond,

Pond A originally seeded with zero turtles, should be regarded as a pond with one turtle and 22 therefore a very low density pond (Table 1). Unfortunately, this turtle was never captured and therefore we do not know its biomass.

After introducing turtles, we collected water samples once every three days from 1 April to 30 June 2015, which corresponds with the height of post-hibernation activity and the nesting season (Ernst 1971). We wore disposable gloves to collect samples, changed gloves between ponds, and ensured our boots did not touch pond water to minimize contamination. We collected samples about 1m from the perimeter at a randomized location, differing for each pond and sample day. We obtained both a 1L sample for GMF methods and a 15mL sample for SA methods simultaneously, thus resulting in paired samples. We transported samples to the laboratory immediately after collection and stored them at 4°C until DNA extraction, usually no more than 72 h.

Sodium acetate methodology

We used a sodium acetate capture and extraction method modified from a protocol used by Piaggio et al. (2014). Briefly, we collected 15mL of water in a 50mL Corning plastic sterile tube from ponds A-D. We added 33mL of 95% ethanol and 1.5mL of 3M sodium acetate to each sample, then allowed each sample to precipitate overnight at -20°C. We centrifuged samples at

3220g for 45 min at 6°C and then discarded the supernatant. We added 10mL of 70% ethanol, vortexed the tube, and centrifuged again at 3220g for 10 min. We extracted DNA from the samples using Qiagen’s DNeasy Blood and Tissue Kit with manufacturer’s protocols but heated the elution buffer to 65°C prior to use (Davy et al. 2015).

Filtration methodology

Using a 10% bleach sterilized glass jar, we collected 1L of water from each pond on each sampling day. Filtration took place under a UV-sterilized hood within 72 h of collection, using one 110mm GMF/F Whatman Filter per sample. We placed filters onto a Buchner funnel attached to a flask for vacuum filtration; we stopped filtration if the filter clogged, but ran all samples for a minimum of 5h. Once 1L of water was filtered, we placed the filter into a 50mL tube and haphazardly added 5-10mL 95% ethanol. We stored the filter at -20°C until extraction.

We extracted DNA from filters using a protocol modified from Goldberg et al. (2011) and Piaggio et al. (2014). We cut filters in half, then into small strips to ease extraction. We added 230µL buffer ATL and 5µLproteinase K to the top of a Qiagen QIAshredder tube and incubated the samples at 65°C overnight (Goldberg et al. 2011). We then extracted eDNA from the elute with the DNeasy Blood and Tissue kit following manufacturer’s instructions with the exception of heating the elution buffer to 65°C before elution (Davy et al. 2015).

Quantification and amplification to confirm eDNA

We analyzed all samples by Nanodrop using elution buffer as a blank, quantifying total eDNA concentration. We used primers from the CO-1 region of the C. picta mitochondrial genome (Davy et al. 2015). We tested the primers against other sympatric turtle species

(Chelydra serpentina, Trachemys scripta, Apalone spinifera, Graptemys ouachitensis,

Graptemys geographica) to ensure species specificity. We utilized DNA samples that had already been extracted using Qiagen’s Blood and Tissue Kit following manufacturer’s protocols obtained from the Janzen Laboratory at Iowa State University. 24

We ran 30µL PCR reactions with 10 µL Environmental Taq Mastermix 2.0, 4 µL MgCl2,

11 µL ddH2O, 1 µL Primer R, 1 µL Primer F, and 3 µL of template. We used a Touchdown PCR procedure, beginning with 10 min at 95°C followed by annealing at 65°C, touching down to

59°C, repeating 34 more cycles of 95°C for 15s, 59°C for 15s, and 72°C for 30s. We employed a final 10 min extension step at 72°C and held final products at 4°C (see S1 for full procedure).

We ran samples on a 1.5% agarose gel with a 100 bp DNA ladder (Thermo Fisher Scientific) and examined the gel for a band of 230 bp in size, aligning with a positive control. We used C. picta blood and C. picta laboratory water extracted with the same methods to serve as positive controls; we ran all samples with PCR negative controls. For any PCR that amplified negative controls, we discarded the entire gel and assumed contamination.

Sequencing

We submitted a C. picta blood positive control, C. picta laboratory eDNA positive control, 13 May 2015 Pond D GMF method, and 13 May 2015 Pond D SA method, and a no template control for sequencing. A small sample of PCR product was run on a 1.5% agarose gel with a 100bp DNA ladder (Thermo Fisher Scientific) to confirm amplification prior to purifying the PCR products with a PureLink PCR Purification Kit (Invitrogen). We submitted samples to the ISU DNA Facility for Sanger sequencing and evaluated species specificity by submitting amplicon sequences to NCBI’s BLAST database (Altschul et al. 1990).

Statistics

The lowest concentration of total eDNA was -1.90 ng/µL, most likely due to machine error, thus we added 1.91 ng/µL to all samples. We omitted dates without a SA or GMF 25 counterpart (n=6) from the dataset. We then log-transformed the remaining total eDNA values to meet assumptions of normality and fit a linear model using the “lm” function of R to examine the effects of pond, date, and extraction method on total eDNA concentration (ng/µL). Because our model showed no effect of pond on total eDNA despite among-pond variance in turtle biomass, we treated ponds equally in this regard.

To examine turtle-specific eDNA amplification, we assigned samples either a 0 or 1 (no amplification or amplification, respectively) depending on whether the sample amplified using conventional PCR. A generalized linear model with a binomial logit-link function showed no effect of pond or date, thus we proceeded to compare extraction methods. We used a McNemar’s test on all extracted samples from all four ponds to determine if GMF and SA values differed in detection of turtle-specific eDNA presence (McNemar 1947; Marascuilo et al. 1988). This test determines whether marginal frequencies differ in matched paired data using a 2x2 contingency table (McNemar 1947). We ran a separate test using eDNA data from the three ponds intentionally seeded with turtles. Unfortunately, Pond A is not a true control as it had one rogue turtle. We used packages plyr and reshape2 for data formatting and analyzed these observations using the command “mcnmar.test” on a 2x2 contingency table comparing amplifications by methodology in R ver. 3.3.1 (Wickham 2007, 2011; R Core Team 2013).

Results

Total eDNA extracted from ponds

We obtained 290 water samples from the ponds in spring and early summer 2015 that yielded eDNA concentrations (ng/µL) for statistical analysis. For PCR, our negative controls 26 failed to amplify. Additionally, our March samples failed to amplify, suggesting little to no turtle eDNA presence prior to our introduction of turtles. We fit an initial linear model to the eDNA concentrations to test for the effect of date, pond, extraction method, and their interactions on eDNA concentration. Neither date nor pond (despite unequal variance in turtle numbers) nor any interactions between date, pond, and extraction method affected eDNA concentration (all p- values>0.05). After removing interactions, re-analysis detected an effect of date on total eDNA concentration (T278=4.72, p<0.001), but still no impact of pond or extraction method (p>0.1).

The effect of date corresponded to an increase of 1.78 ng/µL in eDNA concentration over the course of the 96-day study. The range of values obtained from our total eDNA samples was -1.91 ng/µL to 53.2 ng/µL.

Turtle-specific eDNA extracted from ponds

We tested the detectability of C. picta eDNA in water samples after extraction and amplification by examining gels from PCR for a band of the correct size. Compared to Pond A, which had only one turtle, Pond B (with the highest density of turtles) yielded five more positive detections with the GMF method and one more positive detection with the SA method (Table 1).

Ponds C and D also trended towards a higher abundance of positive detections with the GMF method, GMF detection did not vary statistically significant between ponds (F3,235=0.862, p=0.46). There was also no overwhelming support for an effect of date on turtle eDNA

(F1,235=3.168, p=0.076). Overall, we found C. picta specific eDNA in GMF extractions in 42% of samples (50/120), whereas we amplified it in SA extractions 19% of the time (23/120) (Table

1). 27

McNemar’s test indicated that the GMF filtration methodology extracted C. picta eDNA more reliably than did the SA approach (χ2=12.023, p<0.001) (Table 2). After excluding pond A,

McNemar’s test results comparing C. picta eDNA methodology were not demonstrably different from the McNemar’s test with pond A (χ2=7.500, p<0.01) (Table 3).

Sequencing

The five samples of PCR-amplified CO-1 amplicon were submitted for sequencing – C. picta blood positive control, C. picta laboratory eDNA positive control, 13 May 2015 Pond D

GMF method, and 13 May 2015 Pond D SA method – and then BLASTed using NCBI’s

Genbank database to the C. picta mitochondrial genome (Altschul et al. 1990; Boratyn et al.

2012). Blood extraction, eDNA positive control, and GMF filtered C. picta samples matched

100% with the C. picta mitochondrial genome while the second closest match was 95% to

Malaclemys terrapin, a brackish water turtle inhabiting the Atlantic and Gulf Coasts of North

America. The SA sample BLASTed 98% to the C. picta mitochondrial genome and 95% to

Pseudemys rubriventris, an east and southeastern North American river turtle. We concluded that we most likely amplified C. picta DNA. The no template control submitted failed to amplify any vertebrate DNA.

Discussion

Our experiment yielded two notable findings. First, we detected C. picta presence in outdoor ponds repeatedly over time in semi-natural conditions. Thus, our work supports the prospect that eDNA technology could be applied successfully to collect aquatic reptile eDNA in 28 lentic freshwater systems. Second, GMF filtration method had a higher incidence of finding C. picta -specific eDNA than did the SA approach, although total eDNA (ng/µL) concentration did not differ between the two methods. Interestingly, our experiment showed turtle density did not influence the detection of turtle eDNA with either method.

Environmental DNA used alone may not yet be the most effective method for C. picta detection in lentic systems due to our low collection rate of turtle eDNA (19-42%). Trapping freshwater turtles often has a high success rate, finding an average of 0.9 to 9.2 turtles (90%+ presence detection) per night in a fyke net across the Midwest (Vogt 1980; Bluett et al. 2011).

Hoop net traps routinely averaged more than one turtle per day in a central Illinois lake system

(Bluett et al. 2011). Thus, for eDNA techniques to achieve a similar detection probability, multiple environmental samples must be taken. We currently suggest the use of a multi-sample eDNA method to target areas for more labor-intensive trapping in pond-like conditions. Thus, eDNA techniques could provide useful information.

Total environmental DNA

In our experiment, total eDNA density (ng/µL) did not differ substantially between the two extraction methodologies. Thus, both methods may have extracted total eDNA at a similar efficiency. This could be because total eDNA levels were low in the ponds and both methods could extract this concentration effectively. However, considering the high productivity of our ponds, which at times induced filter clogging, a low total eDNA concentration is unlikely.

Alternatively, total eDNA concentration in water could be so diffuse that the difference in water volumes sampled (15mL – 1000 mL) was not large enough to reveal a different total eDNA concentration. Although GMF filtration collects more water than SA methods, SA extraction 29 may be more effective at obtaining total eDNA. GMF filtration may not catch all fragmented, non-bound eDNA as water is filtered (Barnes and Turner 2016). Thus, SA extraction may be efficient enough to make up for a water volume differential resulting in no substantial total eDNA difference between the two methods. It should be noted that interactions could occur between pond water and potential inhibitors, substrate, or degrading substances, and therefore we cannot rule these out as possible influences on total eDNA concentration (Barnes et al. 2014;

Barnes and Turner 2016; Shogren et al. 2016).

We found an effect of date on total eDNA after removing the interaction effects from our model containing date, pond, and extraction method. Despite having a statistically significant total effect, the overall mean increase of 1.78 ng/µL total eDNA over the course of our 96-day study seems marginal compared to the range of values (-1.9 ng/µL – 53.2 ng/µL) obtained in this study. It is possible this 3% increase was due to human error; as we extracted eDNA from samples our skill at doing so likely improved. Alternatively, as spring progressed into summer, ponds become more biologically active (Knutson et al. 2004). Algae, amphibian activity, and aquatic insect populations all naturally increase through spring, possibly heightening pond productivity and thus the production of eDNA (Buchanan 1905; Merritt and Cummins 1996;

Smith et al. 1999). However, this slight effect of increased total eDNA does not change our overall results of finding no support that either method performs better in collecting total eDNA.

Species-specific environmental DNA

Despite GMF and SA not yielding different total eDNA concentrations, GMF filtration was more effective than the SA method at obtaining turtle eDNA. This difference may be attributed to GMF filtering roughly 67 times more water than the SA method (1L vs 15mL, 30 respectively). Another possible explanation is that turtle-specific eDNA is not sufficiently concentrated in water samples. Sodium acetate extractions rely on salts to extract and bind DNA, thus turtle eDNA may be too sparsely distributed to bind effectively. Additionally, GMF filters are composed of woven glass fibers that form non-uniform pore sizes (Barnes and Turner 2016).

This design could more easily capture a wide variety of eDNA particles, including fragments of degraded target DNA. Overall, we found GMF rather than SA methodology more effective for capturing reptile eDNA in lentic systems.

Our turtle detection rates were similar to some eDNA pond studies. For example, Ficetola et al. (2008) amplified bullfrog (Lithobates catesbeianus) eDNA from ponds in 22% to 88% of

PCR replicates per site sampled, which is comparable to our 19% to 42% findings. Piaggio et al.

(2014) positively amplified species-specific eDNA in 11% to 77% of PCR replicates in samples with a known presence of Burmese pythons (Python bivittatus). For imperiled species, Rees et al.

(2015) detected crested newt (Triturus cristatus) presence in 11% to 75% of sample replicates.

Furthermore, Moyer et al. (2014) detected African jewelfish (Hemichromis lifalili) presence in about 12% of samples in a similar experimental pond setup. Interestingly, there is evidence that ponds sampled in different places and depths collect varying amounts of eDNA with each sample; some pond locations positively detect organism presence while other locations of the same pond were below detection limit (Eichmiller et al. 2014). However, it is important to note many other studies have found 90-100% presence detection across sample replicates in varied species of fish and amphibians (Thomsen et al. 2012; Goldberg et al. 2013; Laramie et al. 2015).

Thus, our detection rate of 19-42% for turtle presence is low but similar to the few available aquatic vertebrate studies.

Limitations of eDNA

Due to field limitations, our experiment does not lend itself well to discerning false positive rates. Since one turtle intruded into Pond A, we cannot distinguish between the detection of a single turtle and background false positive rate. If Pond A were a true negative control with no turtles, a false positive rate could have been obtained with any positive eDNA detection.

Mobile animals such as birds could have transferred turtle DNA to any pond through residual water on their body, resulting in a process type I error (Darling and Mahon 2011; Merkes et al.

2014). Moreover, the pond water was derived from a source that contains C. picta. Despite the fact negative controls taken before the experiment did not amplify turtle eDNA and that water was placed in the ponds months prior to the experiment, residual C. picta eDNA still could have persisted in low concentrations resulting in another process type I error.

It is also possible contamination occurred sometime during the extraction process.

However, water filtration took place under a hood and UV-sterilized bench in a separate room from DNA amplification to decrease the risk of contamination (Pedersen et al. 2015; Goldberg et al. 2016). Furthermore, our sequenced samples matched C. picta. We have no cause to believe

M. terrapin or P. rubriventris were in our ponds, therefore, we judge our sequenced amplicons to be species-specific to C. picta. Given the statistically significant differences between the two eDNA extraction methodologies, we conclude laboratory contamination was not an issue germane to our core findings.

As with other studies, PCR inhibitor presence may have affected turtle eDNA detection.

Previous studies have consistently cited an effect of inhibition, and we suspect our study is no different (McKee et al. 2015; Takahara et al. 2015; Adrian-Kalchhauser and Burkhardt-Holm

2016; Goldberg et al. 2016). Anecdotally, increases in turtle density were positively correlated 32 with increased turbidity, perhaps signaling the increased presence of algae, secondary plant compounds, or humic acids – all known PCR inhibitors (Wilson 1997; Schrader et al. 2012; Jane et al. 2015). Furthermore, as the field season progressed, pond water increasingly became more turbid. Environmental samples may be subject to PCR inhibitors, including plant polysaccharides and tannins, humic substances, and feces (Wilson 1997; Schrader et al. 2012; McKee et al.

2015). However, further experimentation is needed to confirm the exact relationship.

Implications for reptilian eDNA

Our pond system is representative of common habitat for freshwater turtles in North

America. Many freshwater turtles, including C. picta, live in slow-moving backwaters and clear to muddy areas of rivers, ponds, and lakes (Ernst and Lovich 2009). Although our experimental ponds exemplify these aquatic habitats, eDNA techniques should be validated for other aquatic turtle taxa. Painted turtles occur in the same habitat as more imperiled turtle species such as

Blanding’s turtles (Emys blandingii) or spotted turtles (Clemmys guttata) (Joyal et al. 2001;

Litzgus and Mousseau 2004; Ernst and Lovich 2009). Both species face threats from anthropogenic habitat destruction and fragmentation (Ernst and Lovich 2009; Reid and Peery

2014). By enabling conservation biologists to more quickly, non-invasively, and sensitively sample for the presence of imperiled freshwater turtle species in North America, eDNA methods may ease tracking distribution of animals, detecting movement over active seasons, and finding target areas for further sampling.

Although eDNA offers many benefits, some concerns about the technique remain. One criticism is that there is no actual capture of the study organism (Darling and Mahon 2011).

Without tangible organismal capture, it is difficult to ascertain demographic (age, class, size) 33 information other than presence. Additionally, the source of eDNA is unclear as DNA may come from feces of predators, contamination from other nearby environments, or even from carcasses that failed to quickly degrade (Merkes et al. 2014; Ficetola et al. 2016). Soil can retain eDNA for time periods exceeding target organism presence (Dejean et al. 2011; Barnes and Turner 2016;

Shogren et al. 2016). These issues have yet to be resolved by the eDNA community, but may be mitigated by using traditional trapping and invasive capture methodology once an eDNA technique has targeted specific areas for capture (Goldberg et al. 2016). Furthermore, this gap drives the need for controlled studies with known organism presence under natural or semi- natural conditions. Until additional environmental genetics techniques are developed, such as environmental RNA or the targeting of variable regions in the nuclear genome for SNP-based population genetics, eDNA may currently be limited to aiding in presence and abundance studies.

Once fully refined, eDNA may be added to the non-invasive methodology toolbox for ecologists and conservationists to use. In particular, imperiled taxa with difficult-to-study life histories, such as turtles, may benefit from improved non-invasive sampling methodology (Böhm et al. 2013; Spear et al. 2015). Less handling stress on endangered species via non-invasive methods prevents physical harm and reduces their vulnerability to predators. Cryptic life stages may become more easily studied. Because site-specific effort may decrease when using eDNA, larger areas can be targeted for presence/absence sampling (Jerde et al. 2011; Pierson et al.

2016). This expansion may lead to finding new populations of endangered species that would not have been targeted for study if missed by traditional sampling (Pierson et al. 2016). Overall, eDNA ultimately may prove to be a useful tool for managers and ecologists.

Summary

We detected painted turtle (Chrysemys picta) eDNA in semi-natural pond conditions.

From the two methodologies tested, glass microfiber filtration (GMF) and sodium acetate extraction (SA), we found the GMF method to be more successful in detecting C. picta eDNA.

While our experiment was successful in amplifying turtle eDNA from a subset of water samples, we cannot recommend eDNA technique for C. picta or other aquatic turtle species at this time due to lower detection rates compared to traditional sampling methodology. Our hope is that this method will be refined for later use in other species and for population ecology and conservation purposes.

REFERENCES

Adrian-Kalchhauser, I., and P. Burkhardt-Holm. 2016. An eDNA assay to monitor a globally

invasive fish species from flowing freshwater. PLoS One 11:e0147558.

Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic local alignment

search tool. J. Mol. Biol. 215:403–410.

Barnes, M. A., and C. R. Turner. 2016. The ecology of environmental DNA and implications for

conservation genetics. Conserv. Genet. 17:1-17.

Barnes, M. A., C. R. Turner, C. L. Jerde, M. A. Renshaw, W. L. Chadderton, and D. M. Lodge.

2014. Environmental conditions influence eDNA persistence in aquatic systems. Environ.

Sci. Technol. 48:1819–1827.

Bluett, R. D., E. M. Schauber, C. K. Bloomquist, and D. A. Brown. 2011. Sampling assemblages

of turtles in Central Illinois: A case study of capture efficiency and species coverage. Trans.

Illinois State Acad. Sci. 104:127–136.

Böhm, M., B. Collen, J. E. M. Baillie, P. Bowles, J. Chanson, N. Cox, G. Hammerson, M.

Hoffmann, S. R. Livingstone, M. Ram, A. G. J. Rhodin, S. N. Stuart, P. P. van Dijk, B. E.

Young, L. E. Afuang, A. Aghasyan, A. García, C. Aguilar, R. Ajtic, F. Akarsu, L. R. V

Alencar, A. Allison, N. Ananjeva, S. Anderson, C. Andrén, D. Ariano-Sánchez, J. C.

Arredondo, M. Auliya, C. C. Austin, A. Avci, P. J. Baker, A. F. Barreto-Lima, C. L. Barrio-

Amorós, D. Basu, M. F. Bates, A. Batistella, A. Bauer, D. Bennett, W. Böhme, D.

Broadley, R. Brown, J. Burgess, A. Captain, S. Carreira, M. del R. Castañeda, F. Castro, A.

Catenazzi, J. R. Cedeño-Vázquez, D. G. Chapple, M. Cheylan, D. F. Cisneros-Heredia, D.

Cogalniceanu, H. Cogger, C. Corti, G. C. Costa, P. J. Couper, T. Courtney, J. Crnobrnja- 36

Isailovic, P. A. Crochet, B. Crother, F. Cruz, J. C. Daltry, R. J. R. Daniels, I. Das, A. de

Silva, A. C. Diesmos, L. Dirksen, T. M. Doan, C. K. Dodd, J. S. Doody, M. E. Dorcas, J.

Duarte de Barros Filho, V. T. Egan, E. H. El Mouden, D. Embert, R. E. Espinoza, A.

Fallabrino, X. Feng, Z. J. Feng, L. Fitzgerald, O. Flores-Villela, F. G. R. França, D. Frost,

H. Gadsden, T. Gamble, S. R. Ganesh, M. A. Garcia, J. E. García-Pérez, J. Gatus, M.

Gaulke, P. Geniez, A. Georges, J. Gerlach, S. Goldberg, J. C. T. Gonzalez, D. J. Gower, T.

Grant, E. Greenbaum, C. Grieco, P. Guo, A. M. Hamilton, K. Hare, S. B. Hedges, N.

Heideman, C. Hilton-Taylor, R. Hitchmough, B. Hollingsworth, M. Hutchinson, I. Ineich, J.

Iverson, F. M. Jaksic, R. Jenkins, U. Joger, R. Jose, Y. Kaska, U. Kaya, J. S. Keogh, G.

Köhler, G. Kuchling, Y. Kumlutaş, A. Kwet, E. La Marca, W. Lamar, A. Lane, B. Lardner,

C. Latta, G. Latta, M. Lau, P. Lavin, D. Lawson, M. LeBreton, E. Lehr, D. Limpus, N.

Lipczynski, A. S. Lobo, M. A. López-Luna, L. Luiselli, V. Lukoschek, M. Lundberg, P.

Lymberakis, R. Macey, W. E. Magnusson, D. L. Mahler, A. Malhotra, J. Mariaux, B.

Maritz, O. A. V Marques, R. Márquez, M. Martins, G. Masterson, J. A. Mateo, R. Mathew,

N. Mathews, G. Mayer, J. R. McCranie, G. J. Measey, F. Mendoza-Quijano, M. Menegon,

S. Métrailler, D. A. Milton, C. Montgomery, S. A. A. Morato, T. Mott, A. Muñoz-Alonso,

J. Murphy, T. Q. Nguyen, G. Nilson, C. Nogueira, H. Núñez, N. Orlov, H. Ota, J.

Ottenwalder, T. Papenfuss, S. Pasachnik, P. Passos, O. S. G. Pauwels, N. Pérez-Buitrago,

V. Pérez-Mellado, E. R. Pianka, J. Pleguezuelos, C. Pollock, P. Ponce-Campos, R. Powell,

F. Pupin, G. E. Quintero Díaz, R. Radder, J. Ramer, A. R. Rasmussen, C. Raxworthy, R.

Reynolds, N. Richman, E. L. Rico, E. Riservato, G. Rivas, P. L. B. da Rocha, M. O. Rödel,

L. Rodríguez Schettino, W. M. Roosenburg, J. P. Ross, R. Sadek, K. Sanders, G. Santos-

Barrera, H. H. Schleich, B. R. Schmidt, A. Schmitz, M. Sharifi, G. Shea, H. T. Shi, R. 37

Shine, R. Sindaco, T. Slimani, R. Somaweera, S. Spawls, P. Stafford, R. Stuebing, S. Sweet,

E. Sy, H. J. Temple, M. F. Tognelli, K. Tolley, P. J. Tolson, B. Tuniyev, S. Tuniyev, N.

üzüm, G. van Buurt, M. Van Sluys, A. Velasco, M. Vences, M. Veselý, S. Vinke, T. Vinke,

G. Vogel, M. Vogrin, R. C. Vogt, O. R. Wearn, Y. L. Werner, M. J. Whiting, T. Wiewandt,

J. Wilkinson, B. Wilson, S. Wren, T. Zamin, K. Zhou, and G. Zug. 2013. The conservation

status of the world’s reptiles. Biol. Conserv. 157:372–385.

Bohmann, K., I. B. Schnell, and M. T. P. Gilbert. 2013. When bugs reveal biodiversity. Mol.

Ecol. 22:909-911.

Boratyn, G. M., A. A. Schäffer, R. Agarwala, S. F. Altschul, D. J. Lipman, and T. L. Madden.

2012. Domain enhanced lookup time accelerated BLAST. Biol. Direct 7:12.

Buchanan, R. 1905. Notes on the algae of Iowa. Proc. Iowa Acad. Sci. 14:47-84.

Darling, J. A., and A. R. Mahon. 2011. From molecules to management: Adopting DNA-based

methods for monitoring biological invasions in aquatic environments. Environ. Res.

111:978–988.

Davy, C. M., A. G. Kidd, and C. C. Wilson. 2015. Development and validation of environmental

DNA (eDNA) markers for detection of freshwater turtles. PLoS One 10:e0130965. de Souza, L. S., J. C. Godwin, M. A. Renshaw, and E. Larson. 2016. Environmental DNA

(eDNA) detection probability is influenced by seasonal activity of organisms. PLoS One

11:e0165273.

Dejean, T., A. Valentini, A. Duparc, S. Pellier-Cuit, F. Pompanon, P. Taberlet, and C. Miaud.

2011. Persistence of environmental DNA in freshwater ecosystems. PLoS One 6:e23398.

Doi, H., K. Uchii, T. Takahara, S. Matsuhashi, H. Yamanaka, and T. Minamoto. 2015. Use of

droplet digital PCR for estimation of fish abundance and biomass in environmental DNA

surveys. PLoS One 10:e0122763.

Edwards, K. E., and J. C. Jones. 2014. Trapping efficiency and associated mortality of

incidentally captured small mammals during herpetofaunal surveys of temporary wetlands.

Wildl. Soc. Bull. 38:530-535.

Eichmiller, J. J., P. G. Bajer, and P. W. Sorensen. 2014. The relationship between the distribution

of common carp and their environmental DNA in a small lake. PLoS One 9:e112611.

Ernst, C. H. 1971. Population dynamics and activity cycles of Chrysemys picta in southeastern

Pennsylvania. J. Herpetol. 5:151–160.

Ernst, C. H., and J. E. Lovich. 2009. Turtles of the United States and Canada. 2nd ed. JHU Press.

Ficetola, G. F., C. Miaud, F. O. Pompanon, and P. Taberlet. 2008. Species detection using

environmental DNA from water samples. Biol. Lett 4:423–425.

Ficetola, G. F., P. Taberlet, and E. Coissac. 2016. How to limit false positives in environmental

DNA and metabarcoding? Mol. Ecol. Resour. 16:604-607.

Goldberg, C. S., D. S. Pilliod, R. S. Arkle, and L. P. Waits. 2011. Molecular detection of

vertebrates in stream water: A demonstration using rocky mountain tailed frogs and Idaho

giant salamanders. PLoS One 6:e22746.

Goldberg, C. S., A. Sepulveda, A. Ray, J. Baumgardt, and L. P. Waits. 2013. Environmental

DNA as a new method for early detection of New Zealand mudsnails (Potamopyrgus

antipodarum). Freshw. Sci. 32:792–800.

Goldberg, C. S., C. R. Turner, K. Deiner, K. E. Klymus, P. F. Thomsen, M. A. Murphy, S. F.

Spear, A. McKee, S. J. Oyler-McCance, R. S. Cornman, M. B. Laramie, A. R. Mahon, R. F. 39

Lance, D. S. Pilliod, K. M. Strickler, L. P. Waits, A. K. Fremier, T. Takahara, J. E. Herder,

and P. Taberlet. 2016. Critical considerations for the application of environmental DNA

methods to detect aquatic species. Meth. Ecol. Evol. 7:1299–1307.

Jane, S. F., T. M. Wilcox, K. S. Mckelvey, M. K. Young, M. K. Schwartz, W. H. Lowe, B. H.

Letcher, and A. R. Whiteley. 2015. Distance, flow and PCR inhibition: eDNA dynamics in

two headwater streams. Mol. Ecol. Resour. 15:216–227.

Jerde, C. L., A. R. Mahon, W. L. Chadderton, and D. M. Lodge. 2011. “Sight-unseen” detection

of rare aquatic species using environmental DNA. Conserv. Lett. 4:150–157.

Jiang, J.J., E.H. Xia, C.W. Gao, and L.Z. Gao. 2016. The complete mitochondrial genome of

western painted turtle, Chrysemys picta bellii (Chrysemys, Emydidae). Mitochondrial DNA

Part A 27:787–788.

Joyal, L. A., M. McCollough, and M. L. Hunter. 2001. Landscape ecology approaches to wetland

species conservation: a case study of two turtle species in southern Maine. Conserv. Biol.

15:1755–1762.

Kelly, R. P., J. A. Port, K. M. Yamahara, and L. B. Crowder. 2014. Using environmental DNA

to census marine fishes in a large mesocosm. PLoS One 9:e86175.

Knutson, M. G., W. B. Richardson, D. M. Reineke, B. R. Gray, J. R. Parmelee, and S. E. Weick.

2004. Agricultural ponds support amphibian populations. Ecol. Appl. 14:669–684.

Lacoursière-Roussel, A., G. Côté, V. Leclerc, L. Bernatchez, and M. Cadotte. 2016. Quantifying

relative fish abundance with eDNA: a promising tool for fisheries management. J. Appl.

Ecol. 53:1148–1157.

Lane, J. 2006. Can non-invasive glucocorticoid measures be used as reliable indicators of stress

in animals? Anim Welf. 15:331-342. 40

Laramie, M. B., D. S. Pilliod, and C. S. Goldberg. 2015. Characterizing the distribution of an

endangered salmonid using environmental DNA analysis. Biol. Conserv. 183:29–37.

Litzgus, J. D., and T. A. Mousseau. 2004. Home range and seasonal activity of southern spotted

turtles (Clemmys guttata): Implications for management. Copeia 2004:804–817.

Lodge, D. M., C. R. Turner, C. L. Jerde, M. A. Barnes, L. Chadderton, S. P. Egan, J. L. Feder, A.

R. Mahon, and M. E. Pfrender. 2012. Conservation in a cup of water: Estimating

biodiversity and population abundance from environmental DNA. Mol. Ecol. 21:2555–

2558.

Mackenzie, D. I., J. D. Nichols, J. E. Hines, M. G. Knutson, A. B. Franklin, M. G. Knu, and A.

B. Franklin. 2003. Estimating site occupancy, colonization, and local extinction when a

species is detected imperfectly. Ecology 84:2200–2207.

Marascuilo, L. A., C. L. Omelich, and D. V Gokhale. 1988. Planned and post hoc methods for

multiple sample McNemar (1947) tests with missing data. Psychol. Bull. 103:238–245.

McKee, A. M., S. F. Spear, and T. W. Pierson. 2015. The effect of dilution and the use of a post-

extraction nucleic acid purification column on the accuracy, precision, and inhibition of

environmental DNA samples. Biol. Conserv. 183:70–76.

McNemar, Q. 1947. Note on the sampling error of the difference between correlated proportions

or percentages. Psychometrika 12:153–157.

Merkes, C. M., S. G. McCalla, N. R. Jensen, M. P. Gaikowski, and J. J. Amberg. 2014.

Persistence of DNA in carcasses, slime and avian feces may affect interpretation of

environmental DNA data. PLoS One 9:e113346.

Merritt, R. W., and K. W. Cummins. 1996. An introduction to the aquatic insects of North

America, 4th edition. Dubuque, IA: Kendall Hunt Publishing. 41

Miya, M., Y. Sato, T. Fukunaga, T. Sado, J. Y. Poulsen, K. Sato, T. Minamoto, S. Yamamoto, H.

Yamanaka, H. Araki, M. Kondoh, and W. Iwasaki. 2015. MiFish, a set of universal PCR

primers for metabarcoding environmental DNA from fishes: detection of more than 230

subtropical marine species. R. Soc. Open Sci. 2:150088.

Moyer, G. R., E. Díaz-Ferguson, J. E. Hill, C. Shea, and Z. Kawabata. 2014. Assessing

environmental DNA detection in controlled lentic systems. PLoS One 9:e103767.

Pedersen, M. W., S. Overballe-Petersen, L. Ermini, C. Der Sarkissian, J. Haile, M. Hellstrom, J.

Spens, P. F. Thomsen, K. Bohmann, E. Cappellini, I. B. Schnell, N. A. Wales, C. Carøe, P.

F. Campos, A. M. Z. Schmidt, M. T. P. Gilbert, A. J. Hansen, L. Orlando, and E. Willerslev.

2015. Ancient and modern environmental DNA. Philos. Trans. R. Soc. Lond. B. Biol. Sci.

370:20130383.

Piaggio, A. J., R. M. Engeman, M. W. Hopken, J. S. Humphrey, K. L. Keacher, W. E. Bruce,

and M. L. Avery. 2014. Detecting an elusive invasive species: A diagnostic PCR to detect

Burmese python in Florida waters and an assessment of persistence of environmental DNA.

Mol. Ecol. Resour. 14:374–380.

Pierson, T. W., A. M. McKee, S. F. Spear, J. C. Maerz, C. D. Camp, and T. C. Glenn. 2016.

Detection of an enigmatic plethodontid salamander using environmental DNA. Copeia

104:78–82.

R Core Team. 2013. R: A language and environment for statistical computing.

Ream, C., and R. Ream. 1966. The influence of sampling methods on the estimation of

population structure in painted turtles. Am. Midl. Nat. 75:325–338.

Rees, H. C., K. Bishop, D. J. Middleditch, J. R. M. Patmore, B. C. Maddison, and K. C. Gough.

2014. The application of eDNA for monitoring of the great crested newt in the UK. Ecol.

Evol. 4:4023–4032.

Rees, H. C., K. C. Gough, D. J. Middleditch, J. R. M. Patmore, and B. C. Maddison. 2015.

Applications and limitations of measuring environmental DNA as indicators of the presence

of aquatic animals. J. Appl. Ecol. 52:827-831.

Reid, B. N., and M. Z. Peery. 2014. Land use patterns skew sex ratios, decrease genetic diversity

and trump the effects of recent climate change in an endangered turtle. Divers. Distrib.

20:1425–1437.

Renshaw, M. A., B. P. Olds, C. L. Jerde, M. M. Mcveigh, and D. M. Lodge. 2015. The room

temperature preservation of filtered environmental DNA samples and assimilation into a

phenol-chloroform-isoamyl alcohol DNA extraction. Mol. Ecol. Resour. 15:168–176.

Schnell, I. B., R. Sollmann, S. Calvignac-Spencer, M. E. Siddall, D. W. Yu, A. Wilting, and M.

T. P. Gilbert. 2015. iDNA from terrestrial haematophagous leeches as a wildlife surveying

and monitoring tool - prospects, pitfalls and avenues to be developed. Front. Zool. 12:24.

Schrader, C., A. Schielke, L. Ellerbroek, and R. Johne. 2012. PCR inhibitors - occurrence,

properties and removal. J. Appl. Microbiol. 113:1014–1026.

Shaffer, H. B., P. Minx, D. E. Warren, A. M. Shedlock, R. C. Thomson, N. Valenzuela, J.

Abramyan, C. T. Amemiya, D. Badenhorst, K. K. Biggar, G. M. Borchert, C. W. Botka, R.

M. Bowden, E. L. Braun, A. M. Bronikowski, B. G. Bruneau, L. T. Buck, B. Capel, T. A.

Castoe, M. Czerwinski, K. D. Delehaunty, S. V Edwards, C. C. Fronick, M. K. Fujita, L.

Fulton, T. A. Graves, R. E. Green, W. Haerty, R. Hariharan, O. Hernandez, L. W. Hillier,

A. K. Holloway, D. Janes, F. J. Janzen, C. Kandoth, L. Kong, A. P. J. de Koning, Y. Li, R. 43

Literman, S. E. McGaugh, L. Mork, M. O’Laughlin, R. T. Paitz, D. D. Pollock, C. P.

Ponting, S. Radhakrishnan, B. J. Raney, J. M. Richman, J. St John, T. Schwartz, A.

Sethuraman, P. Q. Spinks, K. B. Storey, N. Thane, T. Vinar, L. M. Zimmerman, W. C.

Warren, E. R. Mardis, and R. K. Wilson. 2013. The western painted turtle genome, a model

for the evolution of extreme physiological adaptations in a slowly evolving lineage.

Genome Biol. 14:R28.

Shogren, A. J., J. L. Tank, E. A. Andruszkiewicz, B. Olds, C. Jerde, and D. Bolster. 2016.

Modelling the transport of environmental DNA through a porous substrate using continuous

flow-through column experiments. J. R. Soc. Interface 13:20160290.

Sigsgaard, E. E., H. Carl, P. R. Møller, and P. F. Thomsen. 2015. Monitoring the near-extinct

European weather loach in Denmark based on environmental DNA from water samples.

Biol. Conserv. 183:46–52.

Smart, A. S., R. Tingley, A. R. Weeks, A. R. Van Rooyen, and M. A. McCarthy. 2015.

Environmental DNA sampling is more sensitive than a traditional survey technique for

detecting an aquatic invader. Ecol. Appl. 25:1944–1952.

Smith, G. R., J. E. Rettig, G. G. Mittelbach, J. L. Valiulis, and S. R. Schaack. 1999. The effects

of fish on assemblages of amphibians in ponds: a field experiment. Freshw. Biol. 41:829–

837. Blackwell Science, Ltd.

Spear, S. F., J. D. Groves, L. A. Williams, and L. P. Waits. 2015. Using environmental DNA

methods to improve detectability in a hellbender (Cryptobranchus alleganiensis)

monitoring program. Biol. Conserv. 183:38–45.

Strickler, K. M., A. K. Fremier, and C. S. Goldberg. 2015. Quantifying effects of UV-B,

temperature, and pH on eDNA degradation in aquatic microcosms. Biol. Conserv. 183:85–

92.

Swales, S. 1987. The use of small wiremesh traps in sampling juvenile salmonids. Aquat. Res.

18:187–196.

Taberlet, P., and G. Luikart. 1999. Non-invasive genetic sampling and individual identification.

Biol. J. Linn. Soc. 68:41–55.

Taberlet, P., G. Luikart, and L. P. Waits. 1999. Noninvasive genetic sampling: Look before you

leap. Trends Ecol. Evol. 14:323-327.

Takahara, T., T. Minamoto, and H. Doi. 2015. Effects of sample processing on the detection rate

of environmental DNA from the common carp (Cyprinus carpio). Biol. Conserv. 183:64-

69.

Thomsen, P. F., J. Kielgast, L. L. Iversen, C. Wiuf, M. Rasmussen, M. T. P. Gilbert, L. Orlando,

and E. Willerslev. 2012. Monitoring endangered freshwater biodiversity using

environmental DNA. Mol. Ecol. 21:2565–2573.

Valeix, M., A. J. Loveridge, S. Chamaillé-Jammes, Z. Davidson, F. Murindagomo, H. Fritz, and

D. W. Macdonald. 2009. Behavioral adjustments of African herbivores to predation risk by

lions: Spatiotemporal variations influence habitat use. Ecology 90:23–30.

Valiere, N., and P. Taberlet. 2000. Urine collected in the field as a source of DNA for species

and individual identification. Mol. Ecol. 9:2150–2152.

Vogt, R. C. 1980. New methods for trapping aquatic turtles. Copeia 1980:368.

Wickham, H. 2007. Reshaping data with the reshape package. J. Stat. Softw. 21:1–20.

Wickham, H. 2011. The split-apply-combine strategy for data analysis. J. Stat. Softw. 40:1–29. 45

Wilcox, T. M., K. S. McKelvey, M. K. Young, A. J. Sepulveda, B. B. Shepard, S. F. Jane, A. R.

Whiteley, W. H. Lowe, and M. K. Schwartz. 2016. Understanding environmental DNA

detection probabilities: A case study using a stream-dwelling char Salvelinus fontinalis.

Biol. Conserv. 194:209–216.

Wilson, I. G. 1997. Inhibition and facilitation of nucleic acid amplification. Appl. Environ.

Microbiol. 63:3741–3751.

Yamanaka, H., and T. Minamoto. 2016. The use of environmental DNA of fishes as an efficient

method of determining habitat connectivity. Ecol. Indic. 62:147–153.

Zaiko, A., J. L. Martinez, A. Ardura, L. Clusa, Y. J. Borrell, A. Samuiloviene, A. Roca, and E.

Garcia-Vazquez. 2015a. Detecting nuisance species using NGST: Methodology

shortcomings and possible application in ballast water monitoring. Mar. Environ. Res.

112:64–72.

Zaiko, A., J. L. Martinez, J. Schmidt-Petersen, D. Ribicic, A. Samuiloviene, and E. Garcia-

Vazquez. 2015b. Metabarcoding approach for the ballast water surveillance - An

advantageous solution or an awkward challenge? Mar. Pollut. Bull. 92:25–34.

Table 1. Arrangement of adult painted turtles (Chrysemys picta) in experimental ponds and eDNA amplification results. Note that one turtle appeared in Pond A although none were released into that pond. Number of positive amplifications with each method — Glass Microfiber

Filteration (GMF) or Sodium Acetate (SA) — are out of 120 total each.

Pond # turtles Original # turtles # positive # positive stocked biomass (g) total eDNA eDNA amplifications amplifications with GMF with SA A 0 0 1 11 5 B 11 6068 11 15 6 C 23 9918 23 9 5 D 38 12990 38 12 7

Table 2. McNemar’s contingency table including Pond A results. Yes indicates a positive eDNA detection, no indicates no eDNA detection.

GMF Yes GMF No SA Yes 13 10 SA No 34 59

Table 3. McNemar’s contingency table not including Pond A results. Yes indicates a positive eDNA detection, no indicates no eDNA detection.

GMF Yes GMF No SA Yes 11 7 SA No 25 44

APPENDIX A

SUPPLEMENTAL INFORMATION

PCR Procedure

1. 95 °C for 10:00 2. 95 °C for 0:15 3. 65 °C for 0:15 4. 72 °C for 0:30 5. 95 °C for 0:15 6. 64 °C for 0:15 7. 72 °C for 0:30 8. 95 °C for 0:15 9. 63 °C for 0:15 10. 72 °C for 0:30 11. 95 °C for 0:15 12. 62 °C for 0:15 13. 72 °C for 0:30 14. 95 °C for 0:15 15. 61 °C for 0:15 16. 72 °C for 0:30 17. 95 °C for 0:15 18. 60 °C for 0:15 19. 72 °C for 0:30 20. 95 °C for 0:15 21. 59 °C for 0:15 22. 72 °C for 0:30 23. Repeat Steps 20-22, 34x 24. 72 °C for 10:00 25. Hold at 4 °C

Alternative eDNA Methods

It should be noted that a wide variety of other methodologies were explored for troubleshooting. We believe that it is to the benefit of the eDNA community to include the methodologies which yielded negative results to help guide further protocol refinement. Extraction techniques which did not work included: Whatman Glass Microfiber GF/D (pore size too big), FTA paper. Collection and Filtration techniques that did not work: Adding 567 Buffer ATL and 63 of Proteinase K of the Qiagen Blood and Tissue Kit (too much), Adding 400 µL Buffer AL and 400 µL of EtOH (too much), Spinning the AW1, AW2, and Elution step for twice as long (no effect), Elute 400 µL instead of 200 µL using Qiagen’s Blood and Tissue Kit (DNA too dilute), Overnight proteinase K incubation at 65C instead of 56°C (probably too high)(Goldberg et al. 2011), Q-tip extractions (USGS)(didn’t work for our system), QIAamp DNA Micro Kit (Not 48 sure why it didn’t work), QIAamp Fast DNA Stool Mini Kit (not sure why it didn’t work), PCI Extraction with CTAB buffer, PCI Extraction with Longmire’s buffer, PCI extraction with EtOH as a buffer, PCI extraction with no buffer (Renshaw et al. 2015). Steps that did not work for amplification included the incorporation of BSA, the use of Plant Taq, reducing the total reaction amount to 16µL, and varying concentration of gels from 1.5 to 3% (Adams, unpublished data). Ultimately, we chose a GMF filtration method and SA method because these had been used previously with Burmese Python studies, the only other reptile published on at the time of starting this study.

Autocorrelation Plots – These plots show whether our total environmental DNA samples (ng/µL) are independent. ACF on the Y axis is the autocorrelation factor, Lag on the X axis represents time. Each graph is split up by Pond. Because there is no pattern in the data and all but one spike is not statistically significant, we assume little to no evidence of day on total environmental DNA (ng/µL).

CHAPTER 3

ESTIMATING AQUATIC REPTILE DENSITY UNDER FIELD CONDITIONS USING

ENVIRONMENTAL DNA

This is a manuscript that will be submitted to the peer-reviewed journal Copeia

Abstract

Density monitoring is imperative to understand population fluctuations. One emerging ecological tool for monitoring species is environmental DNA (eDNA), the technique of obtaining target

DNA from environmental samples such as water, soil, or air. We evaluated if eDNA can be used for monitoring aquatic reptile density in a semi-natural lentic pond environment. Using four outdoor experimental ponds with varying Painted Turtle (Chrysemys picta) densities, we quantified both total eDNA and species-specific eDNA for comparison across ponds between 1

April and 30 June 2016. We found significant differences in total eDNA among ponds and a non-linear effect of time on total eDNA accumulation. We were largely unable to amplify turtle eDNA from the water samples despite developing a sensitive species-specific assay, highlighting the limitations of detecting this aquatic reptile under field conditions. Nonetheless, turtle eDNA increased in an expected rank-order pattern with increasing turtle density. Thus, eDNA retains potential to effectively measure density of aquatic reptiles in lentic systems. 50

Introduction

Monitoring population growth and decline is necessary to model future population trends and may illuminate an organism’s biology (Caswell, 1989). Changes in population density may have downstream demographic effects on range, metapopulation structure, and niche availability

(Wells and Richmond, 1995; Holt, 2009). Population stochasticity, whether due to environmental factors, anthropogenic pressures, or biotic interactions – e.g., disease, intrinsic growth and age class, fecundity, or predation – changes population density (Kendall, 1949; Sutcliffe et al., 1996;

Saether and Bakke, 2000; Frankham, 2008; Caswell, 2012). Increasing population density can indicate favorable environmental or biotic conditions for a species. For example, novel habitat can increase habitat and food resources thereby expanding a previously constrained niche

(Marzluff et al., 2001; Mills, 2001). Biotic host densities influence disease transmission; for instance, increased elk density enhances Brucellosis disease (Cross et al., 2007). Thus, monitoring current population levels may predict future population density.

Central to population monitoring is the need for a sensitive detection method. Recently, environmental DNA (eDNA) has received attention for being able to sensitively reveal the presence of target species, especially where traditional methods fall short. Because of the promise to identify target species despite low densities, several papers have addressed eDNA detection limits (Lodge et al., 2012; Dougherty et al., 2016; Goldberg et al., 2016; Secondi et al.,

2016) and even used eDNA for detecting population density (Dejean et al., 2012; Goldberg et al.,

2013). Previous studies have mainly focused on relating eDNA density to organism biomass in the laboratory or from semi-natural aquatic systems. For example, Klymus (2015) found a positive correlation between eDNA shedding rates of Bighead and Silver Carp

(Hypophthalmichthys nobilis and H. molitrix) and carp biomass under laboratory conditions. 51

Takahara (2012) detected a positive correlation between Common Carp (Cyprinus carpio) density and eDNA copy number, or number of target eDNA copies per sample, in a laboratory and field (i.e. lagoon) experiment. In addition, eDNA copy number correlated with patterns of microhabitat use by carp in a lake system (Eichmiller et al., 2014). Doi et al. (2016) noted a positive correlation between the eDNA copy number of a fish species (Plecoglossus altivelis) and its biomass within a Japanese riverine system. These biomass-eDNA copy number trends also held true for some tropical fish species (Robson et al., 2016).

However, not all aquatic field studies find a positive correlation between biomass and eDNA copy number. Spear et al. (2015) attempted to quantify Hellbender (Cryptobranchus alleganiensis) eDNA in a natural system but found eDNA copy number did not relate to density of salamanders in the population. In addition, Biggs et al. (2015) noted that eDNA copy number did not predict Great Crested Newt (Triturus cristatus) density in field ponds. Other studies have found a modest interaction between total biomass and eDNA copy number (Evans et al., 2016;

Kelly et al., 2017). Thus, eDNA may not always predict density in a natural aquatic setting.

Despite breakthroughs in detecting animal density in fish and some amphibian systems, there remains a dearth of studies trying to quantify aquatic reptilian populations with eDNA in the field (Piaggio et al., 2014; Hunter et al., 2015). This lacuna is notable because, for example, turtles are among the most at risk vertebrates, with over 60% of modern species listed as threatened, endangered, or extinct (Böhm et al., 2013; van Dijk et al., 2014). We used the

Painted Turtle (Chrysemys picta) because it exhibits a wide geographic range across North

America, existing in the same habitat as multiple endangered turtle species (Ernst and Lovich,

2009). Therefore, techniques applicable to determining Painted Turtle density may be adaptable 52 for co-occurring imperiled species. Additionally, in North American aquatic systems, Painted

Turtle densities range between 7.2 and 106 kg/ha (Iverson, 1982; Congdon et al., 1986).

The goal of our experiment was to determine the relationship between turtle eDNA and turtle density. We populated semi-natural ponds with varying numbers of turtles and correlated

Painted Turtle eDNA in water samples with adult Painted Turtle biomass in this lentic system.

Previous work under field conditions has assessed presence of freshwater turtle species in a lentic pond system (Davy et al. 2015; Adams et al., unpubl.) and a semi-aquatic turtle has been detected in riverine environments (Lacoursière-Roussel et al., 2016a). Additionally, one study has used site-occupancy in slow-flowing streams to quantify the minimum number of eDNA samples needed to determine presence of an endangered turtle (de Souza et al., 2016). However, we are unaware of any studies that relate density of individual turtles directly with eDNA copy number in a field setting. Furthermore, our sampling regime allowed us to examine potential changes in eDNA over a 3-month period. We hypothesized the amount of total eDNA and turtle eDNA would linearly increase with time and turtle density. A relationship between eDNA concentration and turtle density between ponds and throughout time could deliver an eDNA based management tool for Painted Turtle and other, more imperiled freshwater turtles.

Materials and Methods

Study sites and samples

We seeded four closed-system outdoor ponds with Painted Turtle at the Iowa State

University Horticulture Farm in 2016. These ponds were natural with respect to abiotic variables and water was not treated in any way. Although these ponds were the same dimensions (19m L x

15m W x 1.5m D each), they varied in number of adult turtles (0, 11, 23, 38) and initial biomass 53

(0g, 6088g, 9198g, and 12990g, respectively). We labeled these ponds as zero (0 turtles at a density of 0kg/ha), low (11 turtles and a density of 6kg/ha), medium (23 turtles at a density of

9kg/ha) and high (38 turtles at a density of 13kg/ha) density. We placed turtles in the ponds on 1

April 2016, which coincides with their post-hibernation activity (Ernst 1971).

We took randomized, 250mL water samples approximately 0.75m from the edge of each pond once every three days starting 1 April through 30 June 2016. We took samples in 10% bleach sterilized, autoclaved glass Nalgene jars. When sampling, we used gloves and did not touch the water’s edge with our feet to prevent pond-to-pond contamination. We immediately transported samples to Iowa State University, stored them in a 4°C refrigerator, and extracted

DNA within 48 hours.

Environmental DNA extraction protocol

We optimized our eDNA protocol by first testing multiple published eDNA methods and commercially available extraction kits before settling on the following methods (Adams et al., in review). We processed samples under a UV-sterilized hood to ensure sterility. We vacuum filtered water samples through a 0.8µm-pore cellulose nitrate filter. Once filtration was finished, we immediately folded the filter inward and put it into a QIAshredder with 350µL buffer ATL and 25µL proteinase K (Goldberg et al., 2013; Deiner et al., 2015; Dunker et al., 2016; Secondi et al., 2016; Tsuji et al., 2016). We then incubated the sample overnight at 65°C (Taberlet et al.,

1996; Olds et al., 2016). After the overnight incubation, we spun down the QIAshredder column for 2min at 14,000 rpm and added 200 µL buffer AL and 200 µL 95% ethanol to the elute. After vortexing, we put the solution into a DNeasy Blood and Tissue Kit spin column and spun the sample in a microcentrifuge for 2min at 14,000 rpm (Goldberg et al., 2013). We followed 54

Qiagen’s Manufacturer’s instructions starting with the addition of 500 µL Buffer AW1 (step 5) until elution (step 7). We eluted the samples with 200 µL EDTA (low TE) buffer heated to 65°C

(Davy et al., 2015). We also filtered and extracted three negative laboratory control samples using Culligan Nanopure water in this same way. We quantified samples immediately to get total eDNA (ng/µL) with a Nanodrop 2000 (Thermo Scientific) and stored them at -20°C.

Total eDNA

We quantified each sample against the EDTA low-TE elution buffer using a Nanodrop

2000 obtain total eDNA. Furthermore, we recorded 230/260 and 260/280 ratios to characterize the quality of eDNA obtained (Desjardins and Conklin, 2010). We used a linear model to evaluate the effect of local temperature (°C), cloud cover (%), and precipitation (mm) on total

Nanodrop concentration.

A priori linear hypothesis

We created a linear model examining the effect of time on eDNA amount. Our a priori hypothesis was that total eDNA would increase from 1 April to 30 June. To test this hypothesis, we fit a linear model to total eDNA concentration per pond after log transformation to fit assumptions of normality. We excluded one negative concentration value from the analyzed data set, assuming it arose from machine malfunction. We also assessed an interaction between pond and Julian date in this model.

We performed an ANOVA to test for differences in amounts of total eDNA among the ponds. Our a priori hypothesis was that ponds with a higher concentration of turtles would have 55 a higher total eDNA concentration due to increased organismal biomass. We used the packages ggplot2, dplyr, and lsmeans for program R 3.2.3 (R, ver. 3.2.3).

A posteriori non-linear hypothesis

Due to suspected non-linearity of our dataset, we ran an exploratory generalized additive model (GAM) (Hastie and Tibshirani, 1986; Zuur et al., 2009). A GAM is a likelihood-based regression that uses a smoothing, nonparametric function, similar to a spline (Hastie and

Tibshirani, 1986). We graphed total eDNA concentrations throughout the field season with an additive smoothing term for date. In this model, we averaged concentrations across all ponds per sampling day. Furthermore, we incorporated the daily weather variables into the GAM: mean air temperature (°C), precipitation (mm), and cloud cover (%). AIC model selection was used to compare the GAM and linear model. We used the mgcv package for R 3.2.3.

Turtle-specific eDNA

No lentic pond, species-specific qPCR protocol existed at the time of sampling for the

Painted Turtle, therefore we developed our own. Thermo Fisher Scientific designed a primer- probe combination from Painted Turtle mtDNA using GenBank Accession numbers

KF874616.1, NC_023890.1, NC_002073.3, and AF069423.1. Primer and probe sequence can be ordered using Taqman Assay APMFWY7_C_PICTA_V2 from Thermo Fisher Scientific. We ran a subset of our samples due to cost and time constraints. We used samples from all ponds from dates spaced at roughly 2-week intervals: 30 March, 16 April, 1 May, 16 May, 31 May, 15 June, and 30 June. 56

We performed a qPCR assay as follows: 20µL PerfeCTa qPCR ToughMix (Quanta

Biosciences, MD), 10µL nanopure water and 2 µL of the Taqman primer/probe reaction mix, and 8µL of 1:4 diluted template for a final reaction volume of 40µL. We diluted all samples due to the presence of inhibitors (McKee et al., 2015a). Reaction conditions were as follows: 10:00m initial denaturation at 95C, followed by 45 cycles of 95°C for 15s and 60°C for 45s. We ran qPCR reactions in triplicate and averaged the quantification cycle (Cq) values. We assumed samples that did not return a Cq value were below detection limit and excluded them from analyses. We tested species-specificity of the primer/probe set by attempting to amplify extracted blood samples from five sympatric turtle species (Chelydra serpentina, Trachemys scripta,

Apalone spinifera, Graptemys ouachitensis, Graptemys geographica). These turtle species all yielded Cq values ≥7 higher than Painted Turtle amplification, denoting species specificity (Sow et al., 2009; Bustin et al., 2009). We ran standard curves using DNA extracted from blood and

Painted Turtle eDNA from laboratory water in a 1:2 dilution series. Additionally, we ran one sample (31 May, high density pond) alongside these standard curves at the same dilutions. Due to non-linear eDNA amplification, we chose a 1:4 dilution for all samples (McKee et al., 2015).

Using more concentrated eDNA consistently failed to improve eDNA amplification, likely due to inhibition (McKee et al., 2015). All qPCR runs contained no template controls (NTC) in triplicate. We only considered values <33 Cq to ensure our samples were distinct from background amplification (Sow et al., 2009; Bustin et al., 2009).

In addition to assessing absolute Cq values, we examined the ordered trend of lowest Cq value to highest Cq value among ponds and controls, with abundance corresponding to 1/Cq.

Thus, we expected the highest turtle density per pond to have the lowest Cq value followed by ponds with medium, low, and zero densities of turtles. We also included positive controls (DNA 57 extracted from blood and turtle laboratory water) and a negative control (NTC), expecting extracts from blood to have the highest concentration of turtle eDNA, followed by turtle lab water, and the NTC. We evaluated the statistical significance of this ordering with Jonckheere’s trend test, which is similar to the Kruskal-Wallis test but used specifically to assess a priori ordering hypotheses (Jonckheere, 1954). Our null hypothesis was that there was no trend order, whereas our alternative dictated the strict trend: turtle blood, turtle laboratory water, high turtle density pond, medium turtle density pond, low turtle density pond, zero turtle density pond, then

NTC.

Results

We seeded four ponds with different numbers of adult Painted Turtle (0, 11, 23, 38) to examine the effect of turtle density on both total eDNA and turtle-specific eDNA in a closed- system field setting. We collected data from Julian day 90 to Julian day 181 in 2016, once every three days except for the first seven days in which data were collected daily. We collected 144 samples on 36 d over the 91 d study period. The average air temperature over the sampling period was 16.4°C. A linear model showed no effect of daily air temperature, precipitation, cloud

2 cover, or any interactions on total eDNA concentration (F7,127 = 1.008, all P > 0.15, adjusted R

< 0.001) (Supplementary Figures 4-7). Due to the small R2 value, weather appears to not explain variation in total eDNA.

A priori linear hypothesis

We found no statistically significant effects of pond, date, or their interactions on total eDNA concentration in a full linear model (P > 0.10 in all ponds, date, and resulting interactions except for the high turtle density pond: t135=1.9, p=0.06). Overall, there was no linear trend of date (t138=0.68, P = 0.50) or interaction between pond and date on total eDNA data (P > 0.1)

(Fig. 1). When we removed the interaction, the high turtle density pond differed from the pond that contained no turtles (t138 = 3.13, P < 0.01). Tukey’s post-hoc tests also confirmed statistically significant differences between the intercepts of the high turtle density pond and those of the other three turtle-containing ponds (all P ≤ 0.01). The high turtle density pond had more total eDNA compared to all other ponds; the other three ponds did not differ from each other (all P > 0.4).

A posteriori non-linear hypothesis

Given that the linear a priori hypothesis did not describe these data well, a GAM with smoothing function (day) was fit to the total-eDNA concentration to explore non-linear trends

(Fig. 2). The intercept was statistically different from zero (P < 0.001) and the smoothing term

2 (Julian Day) was statistically significant as well (F6.5,7.7 = 3.45, P < 0.01, adjusted R =0.25).

Between Julian days 135 and 160, the model detected a parabolic curve not captured in the previous linear models (Fig. 2). Additionally, AIC values indicated the GAM outperformed the linear models with and without an interaction between pond and date (Supplementary material).

While the GAM is purely for exploratory purposes, it highlights that not all variation for total eDNA may be explained via a linear model. 59

Including sampling temperature, precipitation, cloud cover, and interactions between all three in the GAM did not improve model fit (all P > 0.1). When interactions were removed to form a reduced model, there was still no substantive effect of any weather variable (P > 0.5).

However, in both models, date remained a significant smoothing factor (both P < 0.01).

Turtle-specific eDNA

From our qPCR dataset, we obtained 27 Cq values from seven sampling days. One sample, the zero turtle density pond on 15 June, was below our detection limit and did not return a Cq value. Our NTC samples amplified at an average Cq value of 40.06 (sd=1.98; sem=0.66) and our positive control Cq was 21.43 (sd=0.39; sem=0.11). The mean of all samples (excluding positive and negative controls) was 38.27 Cq (average sd = 0.86, average sem=0.48). The lowest mean value (i.e. highest eDNA abundance) for any sample was the high turtle density pond on 31

May, with 31.06 Cq (sd=0.39, sem=0.11). This reading is more than 7 Cq values away from our

NTC, rendering it able to be considered for analysis (Sow et al., 2009). The next highest eDNA abundance was for the medium turtle density pond on 15 June, with 33.92 Cq (sd=0.08, sem=0.04), which is not more than 7 Cq values away from the NTC and therefore not sufficiently distinguishable from background amplification. Thus, with only one sample meeting detection criteria, we could not statistically analyze individual Cq values (Fig. 3). That we detected background signal, however, indicates our assay was sensitive and that potential turtle- specific eDNA concentrations in our samples were simply too low.

Regardless of individual Cq values relative to background amplification, we could assess whether average Cq values followed an expected trend of turtle-specific eDNA concentrations.

The rank-order obtained for highest to lowest amplification of turtle-specific eDNA was thus: 60 turtle blood, turtle lab water, high turtle density pond, medium turtle density pond, low turtle density pond, zero turtle density pond, and our NTC (Supplementary material). This ranking of turtle-specific eDNA concentrations exactly matched our alternative hypothesis, and

Jonckheere’s test suggested a meaningful order to these samples (P < 0.001).

Discussion

We found no evidence of a linear increase in total eDNA over time but we did detect a non-linear trend, with a parabolic increase then decline between 15 May and 10 June 2016. The pond with the high turtle density also yielded a higher amount of total eDNA than all other ponds. One out of 28 samples amplified substantial turtle-specific eDNA (the high turtle density pond on 31 May). Thus, we must conclude our qPCR protocol for turtle-specific eDNA did not effectively detect turtles or quantify turtle density. Overall, although we find that total eDNA fluctuates in a field lentic pond system non-linearly, we cannot discern quantitative patterns regarding turtle-specific eDNA, indicating considerable detection limitations. Even so, our rank- order analysis supported the expected trend of increased turtle-specific eDNA with increased turtle density.

Total eDNA

The lack of total eDNA increase with time in the experimental ponds was unexpected.

Past laboratory eDNA studies have indicated that, after adding organisms to water, eDNA increases within a short amount of time (Takahara et al., 2012; Klymus et al., 2015). Indeed, this also happens for reptiles and amphibians (Dejean et al., 2011; Piaggio et al., 2014). Thus, the 61 lack of total eDNA increase over time may have been due to factors other than the presence of turtles and assumed presence of turtle eDNA.

Interestingly, total eDNA differed between the pond with the highest turtle density and all other ponds. Given that this pond had a turtle density of 13kg/ha, turtles likely shed more eDNA into this pond than into the other ponds. Behaviorally, more turtles also may have disturbed the pond substrate more often, increasing surface area of water particulates that could have bound eDNA. From previous studies, we know eDNA binds to substrate debris that can be swept into the water column (Barnes and Turner, 2016). These fine particulates may have been accidentally captured during sampling. Furthermore, we know ponds with turtles tend to increase ecosystem function (Lindsay et al., 2013). Indeed, the high turtle density pond was the only pond in which we consistently observed a Leopard Frog (Lithobates pipiens) along with Bullfrogs (Lithobates catesbeianus). It is impossible to ascertain exactly why total eDNA was higher in this pond, but a higher turtle density may have contributed to this higher productivity.

Perhaps the most interesting result was that of the sudden increase of total eDNA around

15 May and then the sudden decrease around 10 June. Grass mowing, which may have increased the presence of biological material in the water column, occurred on 25 May around the ponds but also occurred earlier in the experiment and did not produce the same effect. As these were ponds in the field, it is possible that algal blooms increased total eDNA. Algae grow in Iowa ponds and can exhibit boom and bust population cycles throughout warm months (Buchanan,

1907; Heisler et al., 2008). Interestingly, the low density turtle pond seemed to have the most distinct cycling pattern of total eDNA (Supplementary material). If our study had continued throughout the summer, this relationship could have been evaluated. 62

Weather did not detectably affect total eDNA amount in this study. Generally, weather variables are thought to influence degradation of eDNA (Darling and Mahon, 2011; Strickler et al., 2015; Barnes and Turner, 2016). Although cloud cover ranged from 95% to 0% across the sampling period (Supplementary material), the presence of solar UV radiation had little impact on total eDNA concentration in our experiment. That weather patterns had minimal effect on total eDNA perhaps suggests that eDNA can readily persist in the environment. Potentially, a threshold of eDNA saturation could exist (Dejean et al., 2011). For example, if abiotic factors degrade eDNA, there may be enough organisms to replenish that degraded eDNA. Nevertheless, this finding challenges prior assumptions about weather-related eDNA degradation in the field.

Turtle-specific eDNA

We developed a sensitive assay for detecting and quantifying turtle eDNA. We detected background Painted Turtle signal despite thorough use of UV-sterilizing equipment before qPCR amplification, isolation of qPCR preparation from DNA extraction, and much care to prevent contamination. Although the majority of our turtle-specific eDNA samples did not differ enough from the background noise to allow quantitative analysis, the raw abundances do qualitatively follow the expected rank-order pattern from highest-turtle density pond to lowest-turtle density pond. Thus, if there had been an increased concentration of eDNA in our samples, we might have obtained enough copies of eDNA for quantitative analysis. Because our negative control amplified, and similar turtle mtDNA has been amplified in our laboratory space before, perhaps amplifying another region of the Painted Turtle genome would aid in eliminating the DNA signal in the negative control (Shaffer et al., 2013; Jiang et al., 2016). Regardless, given that our 63 negative control amplified, we suspect we developed a sensitive assay whose detection limit for turtle eDNA lies below our detected background noise.

Biologically, turtles simply may not shed eDNA into the environment at the same rates as previously studied organisms. A metabarcoding study using universal primer sets and NGS sequencing found half as many eDNA reads for amphibians compared to fish (Valentini et al.,

2016). Besides potentially having lower biomass, many adult amphibians do not have gills and may spend some time out of the water, unlike fish (Lenfant and Johansen, 1967). Similarly, turtles lack gills and most integument is keratinized, thus they may not shed eDNA as readily as a mucus layer like amphibians and fish (Weldon et al., 1993; Ernst and Lovich, 2009).

Furthermore, turtles shed scutes and skin in pieces (rather than as rafts of cells), which, due to their mass, may sink into substrate and be unlikely to be detected as readily by eDNA methodology. Thus only excrement, tears, and saliva may be primary shedding mechanisms for turtle eDNA (Parmenter, 1981; Northmore and Granda, 1991; Ernst and Lovich, 2009).

It is also possible that turtle eDNA accounts for a miniscule fraction of total eDNA, such that any increase would have marginal effects at best. Metabarcoding studies find hundreds of organisms within their sequencing, thus, it is not unreasonable that although present, turtle eDNA would remain in very small concentrations (Zaiko et al., 2015; Valentini et al., 2016).

This scenario would align well with results of mesocosm studies specifically targeting fish and turtles in a semi-controlled environment (Kelly et al., 2014).

As with other eDNA studies, our experiment likely suffers from DNA inhibition in the environmental samples. When standard curves were run, sample dilutions always had a lower Cq value than the full sample itself, signaling the presence of inhibitors (McKee et al., 2015b). With

DNA extracted from Painted Turtle blood and Painted Turtle laboratory water, this was not the 64 case, indicating no inhibitor presence. Inhibition is common in eDNA field studies; indeed, it is a known issue and addressed through various protocols. Special buffers during extraction, using clean-up kits, using BSA in PCR reactions, and diluting template for PCR reactions are common ways of minimizing the effect of inhibitory compounds (Barnes et al., 2014; Davy et al., 2015;

McKee et al., 2015a; Renshaw et al., 2015; Wegleitner et al., 2015; Carim et al., 2016; de Souza et al., 2016; Goldberg et al., 2016; Pierson et al., 2016). Common environmental inhibitors include plant secondary compounds such as polysaccharides, pectin, xylan, phenols and tannins

(Wilson, 1997; Schrader et al., 2012). Soil also contains known PCR inhibitors including humic acids, minerals such as calcium, and inorganic compounds (Wilson, 1997; Schrader et al., 2012).

Other proteases, urea, and competing DNA may additionally inhibit reactions or decrease reaction efficiency (Schrader 2012). Despite the use of ToughMix (QuantaBiosciences), specifically made to reduce the effects of PCR inhibition, we were unable to amplify enough turtle eDNA to effectively relate to turtle density.

Conclusion

We presently cannot recommend eDNA for monitoring aquatic turtle density under field conditions. We obtained just one substantially amplifiable sample of turtle eDNA from pond water despite obtaining turtle-specific eDNA from lab water and developing a sensitive qPCR assay. On the other hand, we detected an expected positive relationship between turtle density and turtle-specific eDNA, hinting at a possible covariance between turtle density and eDNA extracted. Still, this study highlights current limitations of detecting aquatic reptile eDNA density under field conditions. We are hopeful that advances in both technology and technique will soon realize the full potential of eDNA for monitoring turtle populations. We recommend evaluating 65 techniques using droplet digital PCR because those may better quantify eDNA at low density and with more finesse than qPCR (Nathan et al., 2014; Doi et al., 2015). That fish and amphibians have well developed eDNA techniques lends optimism to the view that eDNA eventually can be used to monitor populations of aquatic turtles. 66

REFERENCES

Barnes, M. A., and C. R. Turner. 2016. The ecology of environmental DNA and implications for

conservation genetics. Conservation Genetics 17:1-17.

Barnes, M. A., C. R. Turner, C. L. Jerde, M. A. Renshaw, W. L. Chadderton, and D. M. Lodge.

2014. Environmental conditions influence eDNA persistence in aquatic systems.

Environmental Science and Technology 48:1819–1827.

Biggs, J., N. Ewald, A. Valentini, C. Gaboriaud, T. Dejean, R. A. Griffiths, J. Foster, J. W.

Wilkinson, A. Arnell, P. Brotherton, P. Williams, and F. Dunn. 2015. Using eDNA to

develop a national citizen science-based monitoring programme for the great crested

newt (Triturus cristatus). Biological Conservation 183:19–28.

Böhm, M., B. Collen, J. E. M. Baillie, P. Bowles, J. Chanson, N. Cox, G. Hammerson, M.

Hoffmann, S. R. Livingstone, M. Ram, A. G. J. Rhodin, et al. 2013. The conservation

status of the world’s reptiles. Biological Conservation 157:372–385.

Buchanan, R. 1907. Notes on the algae of Iowa. Proceedings of the Iowa Academy of Science

14:47–84.

Bustin, S. A., V. Benes, J. A. Garson, J. Hellemans, J. Huggett, M. Kubista, R. Mueller, T.

Nolan, M. W. Pfaffl, G. L. Shipley, J. Vandesompele, and C. T. Wittwer. 2009. The

MIQE guidelines: Minimum information for publication of quantitative real-time PCR

experiments. Clinical Chemistry 55:611–622.

Carim, K. J., K. S. McKelvey, M. K. Young, T. M. Wilcox, M. K. Schwartz. 2016. A protocol

for collecting environmental DNA samples from streams. Gen. Tech. Rep. RMRS-GTR- 67

355. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain

Research Station. 18 p.

Caswell, H. 1989. Matrix population models:construction, analysis, and interpretation. Sinauer

Associates Inc., Sunderland, MA.

Caswell, H. 2012. Matrix models and sensitivity analysis of populations classified by age and

stage: a vec-permutation matrix approach. Theoretical Ecology 5:403–417.

Christiansen, J. L., J. A. Cooper, J. W. Bickham, B. J. Gallaway, and M. D. Springer. 1985.

Aspects of the natural history of the yellow mud turtle Kinosternon flavescens

(Kinosternidae) in Iowa: a proposed endangered species. Southwestern Naturalist

30:413–425.

Congdon, J. D., J. L. Greene, and J. W. Gibbons. 1986. Biomass of freshwater turtles: A

geographic comparison. American Midland Naturalist 115:165–173. doi:

10.2307/2425846

Cross, P. C., W. H. Edwards, B. M. Scurlock, E. J. Maichak, and J. D. Rogerson. 2007. Effects

of management and climate on elk Brucellosis in the Greater Yellowstone Ecosystem.

Ecological Applications 17:957–964.

Darling, J. A., and A. R. Mahon. 2011. From molecules to management: Adopting DNA-based

methods for monitoring biological invasions in aquatic environments. Environmental

Research 111:978–988.

Davy, C. M., A. G. Kidd, and C. C. Wilson. 2015. Development and validation of environmental

DNA (eDNA) markers for detection of freshwater turtles. PLoS ONE 10:e0130965. 68

Deiner, K., J. C. Walser, E. Mächler, and F. Altermatt. 2015. Choice of capture and extraction

methods affect detection of freshwater biodiversity from environmental DNA. Biological

Conservation 183:53-63.

Dejean, T., A. Valentini, A. Duparc, S. Pellier-Cuit, F. Pompanon, P. Taberlet, and C. Miaud.

2011. Persistence of environmental DNA in freshwater ecosystems. PLoS ONE

6:e23398.

Dejean, T., A. Valentini, C. Miquel, P. Taberlet, E. Bellemain, and C. Miaud. 2012. Improved

detection of an alien invasive species through environmental DNA barcoding: The

example of the American bullfrog Lithobates catesbeianus. Journal of Applied Ecology

49:953–959.

Desjardins, P., and D. Conklin. 2010. NanoDrop microvolume quantitation of nucleic acids.

Journal of visualized experiments: Journal of Visualized Experiments 45:2565. van Dijk, P. P., J. Iverson, A. Rhodin, B. Shaffer, and R. Bour. 2014. Turtles of the world, 7th

edition: Annotated checklist of taxonomy, synonymy, distribution with maps, and

conservation status. Chelonian Research Monographs 5:329–479.

Doi, H., R. Inui, Y. Akamatsu, K. Kanno, H. Yamanaka, T. Takahara, and T. Minamoto. 2016.

Environmental DNA analysis for estimating the abundance and biomass of stream fish.

Freshwater Biology 62:30–39.

Doi, H., K. Uchii, T. Takahara, S. Matsuhashi, H. Yamanaka, and T. Minamoto. 2015. Use of

droplet digital PCR for estimation of fish abundance and biomass in environmental DNA

surveys. PLoS ONE 10:e0122763. 69

Dougherty, M. M., E. R. Larson, M. A. Renshaw, C. A. Gantz, S. P. Egan, D. M. Erickson, and

D. M. Lodge. 2016. Environmental DNA (eDNA) detects the invasive rusty crayfish

Orconectes rusticus at low abundances. Journal of Applied Ecology. 53:722-732.

Dunker, K. J., A. J. Sepulveda, R. L. Massengill, J. B. Olsen, O. L. Russ, J. K. Wenburg, and A.

Antonovich. 2016. Potential of environmental DNA to evaluate northern pike (Esox

lucius) eradication efforts: An experimental test and case study. PLoS ONE 11:e0162277.

Eichmiller, J. J., P. G. Bajer, and P. W. Sorensen. 2014. The relationship between the distribution

of common carp and their environmental DNA in a small lake. PLoS ONE 9:e112611.

Ernst, C. H., and J. E. Lovich. 2009. Turtles of the United States and Canada. 2nd edition. JHU

Press, Baltimore, MD.

Evans, N. T., B. P. Olds, M. A. Renshaw, C. R. Turner, Y. Li, C. L. Jerde, A. R. Mahon, M. E.

Pfrender, G. A. Lamberti, and D. M. Lodge. 2016. Quantification of mesocosm fish and

amphibian species diversity via environmental DNA metabarcoding. Molecular Ecology

Resources 16:29–41.

Frankham, R. 2008. Genetic adaptation to captivity in species conservation programs. Molecular

Ecology 17:325–333.

Gedalanga, P. B., and B. H. Olson. 2009. Development of a quantitative PCR method to

differentiate between viable and nonviable bacteria in environmental water samples.

Applied Microbiology and Biotechnology 82:587–596.

Goldberg, C. S., A. Sepulveda, A. Ray, J. Baumgardt, and L. P. Waits. 2013. Environmental

DNA as a new method for early detection of New Zealand mudsnails (Potamopyrgus

antipodarum). Freshwater Science 32:792–800. 70

Goldberg, C. S., C. R. Turner, K. Deiner, K. E. Klymus, P. F. Thomsen, M. A. Murphy, S. F.

Spear, A. McKee, S. J. Oyler-McCance, R. S. Cornman, M. B. Laramie, et al. 2016.

Critical considerations for the application of environmental DNA methods to detect

aquatic species. Methods in Ecology and Evolution 7:1299–1307.

Hastie, T., and R. Tibshirani. 1986. Generalized additive models. Statistical Science 1:297–310.

Heisler, J., P. M. Glibert, J. M. Burkholder, D. M. Anderson, W. Cochlan, W. C. Dennison, Q.

Dortch, C. J. Gobler, C. A. Heil, E. Humphries, A. Lewitus, et al. 2008. Eutrophication

and harmful algal blooms: A scientific consensus. Harmful Algae 8:3–13.

Holt, R. D. 2009. Bringing the Hutchinsonian niche into the 21st century: Ecological and

evolutionary perspectives. Proceedings of the National Academy of Sciences USA

106:19659–19665.

Hunter, M. E., S. J. Oyler-McCance, R. M. Dorazio, J. A. Fike, B. J. Smith, C. T. Hunter, R. N.

Reed, and K. M. Hart. 2015. Environmental DNA (eDNA) sampling improves

occurrence and detection estimates of invasive Burmese pythons. PLoS ONE

10:e0121655.

Iverson, J. B. 1982. Biomass in turtle populations: A neglected subject. Oecologia 55:69–76.

Jaeger, C. P., and V. A. Cobb. 2012. Comparative spatial ecologies of female Painted Turtles

(Chrysemys picta) and Red-Eared Sliders (Trachemys scripta) at Reelfoot Lake,

Tennessee. Chelonian Conservation and Biology 11:59–67.

Jiang, J.-J., E.-H. Xia, C.-W. Gao, and L.-Z. Gao. 2016. The complete mitochondrial genome of

western Painted Turtle, Chrysemys picta bellii (Chrysemys, Emydidae). Mitochondrial

DNA Part A 27:787–788. 71

Jonckheere, A. R. 1954. A distribution-free k-sample test against ordered alternatives.

Biometrika 41:133-145.

Kelly, R. P., C. J. Closek, J. L. O’Donnell, J. E. Kralj, A. O. Shelton, and J. F. Samhouri. 2017.

Genetic and manual survey methods yield different and complementary views of an

ecosystem. Frontiers in Marine Science 3:283.

Kelly, R. P., J. A. Port, K. M. Yamahara, and L. B. Crowder. 2014. Using environmental DNA

to census marine fishes in a large mesocosm. PLoS ONE 9:e86175.

Kendall, D. G. 1949. Stochastic processes and population growth. Journal of the Royal Statistical

Society. Series B (Methodological) 11:230–282.

Klymus, K. E., C. A. Richter, D. C. Chapman, and C. Paukert. 2015. Quantification of eDNA

shedding rates from invasive bighead carp Hypophthalmichthys nobilis and silver carp

Hypophthalmichthys molitrix. Biological Conservation 183:77-84.

Lacoursière-Roussel, A., Y. Dubois, E. Normandeau, and L. Bernatchez. 2016a. Improving

herpetological surveys in eastern North America using the environmental DNA method.

Genome 59:991-1007.

Lacoursière-Roussel, A., M. Rosabal, and L. Bernatchez. 2016b. Estimating fish abundance and

biomass from eDNA concentrations: Variability among capture methods and

environmental conditions. Molecular Ecology Resources 16:1401–1414.

Laramie, M. B., D. S. Pilliod, and C. S. Goldberg. 2015. Characterizing the distribution of an

endangered salmonid using environmental DNA analysis. Biological Conservation

183:29–37.

Lefevre, K., and R. J. Brooks. 1995. Effects of sex and body size on basking behavior in a

northern population of the Painted Turtle, Chrysemys picta. Herpetologica 51:217–224. 72

Lenfant, C., and K. Johansen. 1967. Respiratory adaptations in selected amphibians. Respiration

Physiology 2:247–260.

Lindsay, M. K., Y. Zhang, M. R. J. J. Forstner, and D. Hahn. 2013. Effects of the freshwater

turtle Trachemys scripta elegans on ecosystem functioning: an approach in experimental

ponds. Amphibia-Reptilia 34:75–84.

Lodge, D. M., C. R. Turner, C. L. Jerde, M. A. Barnes, L. Chadderton, S. P. Egan, J. L. Feder, A.

R. Mahon, and M. E. Pfrender. 2012. Conservation in a cup of water: Estimating

biodiversity and population abundance from environmental DNA. Molecular Ecology

21:2555–2558.

Maruyama, A., K. Nakamura, H. Yamanaka, M. Kondoh, and T. Minamoto. 2014. The release

rate of environmental DNA from juvenile and adult fish. PLoS ONE 9:e114639.

Marzluff, J. M., K. J. McGowan, R. Donnelly, and R. L. Knight. 2001. Causes and consequences

of expanding American Crow populations. In: Avian Ecology and Conservation in an

Urbanizing World. Springer US, Boston, MA, pp. 331–363.

McKee, A. M., D. L. Calhoun, W. J. Barichivich, S. F. Spear, C. S. Goldberg, and T. C. Glenn.

2015a. Assessment of environmental DNA for detecting presence of imperiled aquatic

amphibian species in isolated wetlands. Journal of Fish and Wildlife Management 6:498–

510.

McKee, A. M., S. F. Spear, and T. W. Pierson. 2015b. The effect of dilution and the use of a

post-extraction nucleic acid purification column on the accuracy, precision, and inhibition

of environmental DNA samples. Biological Conservation 183:70–76.

Mills, C. E. 2001. Jellyfish blooms: are populations increasing globally in response to changing

ocean conditions? Hydrobiologia 451:55-68. 73

Minamoto, T., M. Fukuda, K. R. Katsuhara, A. Fujiwara, S. Hidaka, S. Yamamoto, K.

Takahashi, and R. Masuda. 2017. Environmental DNA reflects spatial and temporal

jellyfish distribution. PLoS ONE 12:e0173073.

Moss, J. A., J. Gordy, and R. A. Snyder. 2014. Effective concentration and detection of

Cryptosporidium, Giardia, and the microsporidia from environmental matrices. Journal

of Pathogens 2014:408204.

Nathan, L. M., M. Simmons, B. J. Wegleitner, C. L. Jerde, and A. R. Mahon. 2014. Quantifying

environmental DNA signals for aquatic invasive species across multiple detection

platforms. Environmental Science and Technology 48:12800–12806.

Northmore, D., and A. Granda. 1991. Ocular dimensions and schematic eyes of freshwater and

sea turtles. Visual Neuroscience 7:627–635.

Olds, B. P., C. L. Jerde, M. A. Renshaw, Y. Li, N. T. Evans, C. R. Turner, K. Deiner, A. R.

Mahon, M. A. Brueseke, P. D. Shirey, M. E. Pfrender, et al. 2016. Estimating species

richness using environmental DNA. Ecology and Evolution 6:4214–4226.

Parmenter, R. R. 1981. Digestive turnover rates in freshwater turtles: The influence of

temperature and body size. Comparative Biochemistry and Physiology Part A:

Physiology 70:235–238.

Piaggio, A. J., R. M. Engeman, M. W. Hopken, J. S. Humphrey, K. L. Keacher, W. E. Bruce,

and M. L. Avery. 2014. Detecting an elusive invasive species: A diagnostic PCR to detect

Burmese python in Florida waters and an assessment of persistence of environmental

DNA. Molecular Ecology Resources 14:374–380. 74

Pierson, T. W., A. M. McKee, S. F. Spear, J. C. Maerz, C. D. Camp, and T. C. Glenn. 2016.

Detection of an enigmatic plethodontid salamander using environmental DNA. Copeia

104:78–82.

Renshaw, M. A., B. P. Olds, C. L. Jerde, M. M. Mcveigh, and D. M. Lodge. 2015. The room

temperature preservation of filtered environmental DNA samples and assimilation into a

phenol-chloroform-isoamyl alcohol DNA extraction. Molecular Ecology Resources

15:168–176.

Robson, H. L. A., T. H. Noble, R. J. Saunders, S. K. A. Robson, D. W. Burrows, and D. R. Jerry.

2016. Fine-tuning for the tropics: application of eDNA technology for invasive fish

detection in tropical freshwater ecosystems. Molecular Ecology Resources 16:922–932.

Rowe, J. W. 2003. Activity and movements of Midland Painted Turtles (Chrysemys picta

marginata) living in a small marsh system on Beaver Island, Michigan. Journal of

Herpetology 37:342–353.

Saether, B.-E., and O. Bakke. 2000. Avian life history variation and contribution of demographic

traits to the population growth rate. Ecology 81:642–653.

Schmelzle, M. C., and A. P. Kinziger. 2016. Using occupancy modelling to compare

environmental DNA to traditional field methods for regional-scale monitoring of an

endangered aquatic species. Molecular Ecology Resources 16:895–908.

Schrader, C., A. Schielke, L. Ellerbroek, and R. Johne. 2012. PCR inhibitors - occurrence,

properties and removal. Journal of Applied Microbiology 113:1014–1026.

Secondi, J., T. Dejean, A. Valentini, B. Audebaud, and C. Miaud. 2016. Detection of a global

aquatic invasive amphibian, Xenopus laevis, using environmental DNA. Amphibia-

Reptilia 37:131–136. 75

Shaffer, H. B., P. Minx, D. E. Warren, A. M. Shedlock, R. C. Thomson, N. Valenzuela, J.

Abramyan, C. T. Amemiya, D. Badenhorst, K. K. Biggar, G. M. Borchert, et al. 2013.

The western Painted Turtle genome, a model for the evolution of extreme physiological

adaptations in a slowly evolving lineage. Genome Biology 14:R28.

Smart, A. S., R. Tingley, A. R. Weeks, A. R. Van Rooyen, and M. A. McCarthy. 2015.

Environmental DNA sampling is more sensitive than a traditional survey technique for

detecting an aquatic invader. Ecological Applications 25:1944–1952. de Souza, L. S., J. C. Godwin, M. A. Renshaw, and E. Larson. 2016. Environmental DNA

(eDNA) detection probability is influenced by seasonal activity of organisms. PLoS ONE

11:e0165273.

Sow, F. B., J. M. Gallup, R. E. Sacco, and M. R. Ackermann. 2009. Laser capture

microdissection revisited as a tool for transcriptomic analysis: Application of an excel-

based qPCR preparation software (PREXCEL-Q). International Journal of Biomedical

Science 5:105–124.

Spear, S. F., J. D. Groves, L. A. Williams, and L. P. Waits. 2015. Using environmental DNA

methods to improve detectability in a hellbender (Cryptobranchus alleganiensis)

monitoring program. Biological Conservation 183:38–45.

Strickler, K. M., A. K. Fremier, and C. S. Goldberg. 2015. Quantifying effects of UV-B,

temperature, and pH on eDNA degradation in aquatic microcosms. Biological

Conservation 183:85–92.

Sutcliffe, O. L., C. D. Thomas, and D. Moss. 1996. Spatial synchrony and asynchrony in

butterfly population dynamics. Journal of Animal Ecology 65:85-95. 76

Taberlet, P., S. Griffin, B. Goossens, S. Questiau, V. Manceau, N. Escaravage, L. P. Waits, and

J. Bouvet. 1996. Reliable genotyping of samples with very low DNA quantities using

PCR. Nucleic Acids Research 24:3189–3194.

Takahara, T., T. Minamoto, H. Yamanaka, H. Doi, and Z. Kawabata. 2012. Estimation of fish

biomass using environmental DNA. PLoS ONE. 7:e35868.

Tsuji, S., H. Yamanaka, and T. Minamoto. 2016. Effects of water pH and proteinase K treatment

on the yield of environmental DNA from water samples. Limnology 18:1–7.

Valentini, A., P. Taberlet, C. Miaud, R. Civade, J. Herder, P. F. Thomsen, E. Bellemain, A.

Besnard, E. Coissac, F. Boyer, C. Gaboriaud, et al. 2016. Next-generation monitoring of

aquatic biodiversity using environmental DNA metabarcoding. Molecular Ecology

25:929–942.

Wegleitner, B. J., C. L. Jerde, A. Tucker, W. L. Chadderton, and A. R. Mahon. 2015. Long

duration, room temperature preservation of filtered eDNA samples. Conservation

Genetics Resources 7:789-791.

Weldon, P. J., B. J. Demeter, and R. Rosscoe. 1993. A survey of shed skin-eating

(dermatophagy) in amphibians and reptiles. Journal of Herpetology 27:219-228.

Wells, J. V., and M. E. Richmond. 1995. Populations, metapopulations, and species populations:

what are they and who should care? Wildlife Society Bulletin 23:458–462.

Wilson, I. G. 1997. Inhibition and facilitation of nucleic acid amplification. Applied and

Environmental Microbiology 63:3741–3751.

Zaiko, A., J. L. Martinez, J. Schmidt-Petersen, D. Ribicic, A. Samuiloviene, and E. Garcia-

Vazquez. 2015. Metabarcoding approach for the ballast water surveillance - An

advantageous solution or an awkward challenge? Marine Pollution Bulletin 92:25–34. 77

Zuur, A. F., E. N. Ieno, N. Walker, A. A. Saveliev, and G. M. Smith. 2009. Mixed effects models

and extensions in ecology with R. In.: Springer New York, NY.

Figure 1. Total eDNA concentration (ng/uL) by day of the year (days 90 – 181) of 2016. Values were log-transformed to fit assumptions of normality. Each color represents a different pond, each of which contained a different density of adult Painted Turtle (Chrysemys picta); grey bars indicate 95% confidence intervals per each pond. Data are graphed as points, and linear models are plotted for each pond to describe temporal trends. 79

Figure 2. Total eDNA (ng/uL) concentration averaged across all ponds analyzed using a general additive model. Day of the year is the smoothing term; dotted lines represent 95% confidence intervals. 80

Figure 3. Amplification (quantification cycle=Cq) of Painted Turtle (Chrysemys picta) eDNA per pond and date. Varying colors and symbols represent each different pond: the zero density pond had 0 turtles, the low density pond had 11 turtles, the medium density pond had 23 turtles, and the high density pond had 38 turtles. Bars indicate SEM, points indicate the average triplicate value of each sample. The positive control from extracted Painted Turtle blood (P) and the no template control (NTC) were plotted at day 75 to facilitate comparisons.

APPENDIX A SUPPLEMENTAL INFORMATION

25 y = -1.425ln(x) + 19.689 R² = 0.9981 20

Cq Cq value 15

0 0 0.5 1 1.5 2 2.5 3 3.5 Dilution Factor

Supplementary Figure 1. A scatterplot of DNA extracted from blood dilution standard curve. The best fit equation to the data is 1.425ln(x) + 19.689 with an R2=0.998.

eDNA standards 29.5

29 y = -1.519ln(x) + 25.888 28.5 R² = 0.9994

27.5 Cq Cq value 27

26.5

25.5 0 0.2 0.4 0.6 0.8 1 1.2 Dilution Factor

Supplementary Figure 2. A scatterplot of (+) eDNA painted turtle laboratory water standard curve. The best fit equation to the data is y=-1.519ln(x) + 25.888 with an R2=0.9994.

Supplementary Table 1. Comparisons of pond intercepts using the lsmeans package in program

R. Tukey-adjusted p-values indicate a significant difference between ponds.

Pond Comparison Intercept Standard df t-ratio p-value Significance Difference Error Estimate (ng/uL) No Turtles-High Turtles -11.98 3.72 136 -3.224 0.0085 ** No Turtles-Medium 3.50 3.72 136 0.943 0.7818 Turtles No Turtles-Low Turtles -3.41 3.72 136 -0.917 0.7960 High Turtles-Medium 15.48 3.72 136 4.167 0.0003 *** Turtles High Turtles-Low Turtles 8.57 3.72 136 2.307 0.1014 Medium Turtles-Low -6.91 3.72 136 -1.859 0.2505 Turtles

Supplementary Table 2. Table comparing AIC values of different models along with each model’s degrees of freedom (df).

Model df AIC

GAM Model 10.84155 1191.721

Linear Model without Pond*Date 6 1208.497

Interaction

Repeated Measures Model 10 1211.625

Linear Model with Pond*Date 9 1212.619

Interaction

Supplementary Figure 3. Daily mean air temperature data throughout the 2016 field season. 84

Supplementary Figure 4. Precipitation data throughout the 2016 field season.

30.00 40.00

35.00 25.00 30.00 20.00 25.00

15.00 20.00

15.00 Temperature (C) Temperature

10.00 (mm) Precipitation 10.00 5.00 5.00

0.00 0.00 85 105 125 145 165 185 Julian Day (2016)

Temperature (°C) Precipitation (mm)

Supplementary Figure 5. Daily mean temperature and precipitation data graphed together throughout the 2016 field season. 85

Supplementary Figure 6. Daily cloud cover (%) estimated at the time of sampling throughout the

2016 field season.

Supplementary Figure 7. Individual total eDNA concentrations (ng/uL) by pond and day of the year (Jdate) in 2016. Each pond contained a different density of adult Painted Turtle (Chrysemys picta).

Supplementary Table 3. Cq averages for Painted Turtle (Chrysemys picta) eDNA from water samples obtained from supplemental ponds throughout the 2016 field season.

Sample Cq Average 0 Turtles 39.95 38 Turtles 36.59 23 Turtles 38.15 11 Turtles 38.63 Positive control (P) 21.43 No template control (NTC) 40.07 (+) eDNA Turtle Lab Water 27.47

Supplementary Figure 8. Boxplot of Cq value averages for Painted Turtle (Chrysemys picta) eDNA from water samples obtained from experimental ponds throughout the 2016 field season.

CHAPTER 4

SUMMARY AND CONCLUSION

The goal of my thesis work was to elucidate a methodology for extracting C. picta specific environmental DNA (eDNA) and relating it to C. picta density for potential population monitoring and conservation applications. This process included finding the presence of C. picta eDNA in a lentic semi-natural pond system, repeated over time. We found GMF filtration to be more effective than SA extraction in detecting C. picta presence, though visual surveys were more effective at detecting the presence of turtles. Additionally, total eDNA concentration did not follow a linear trend across the 2016 field season. We were unable to amplify turtle-specific eDNA despite having developed a sensitive, species-specific qPCR reaction.

One area of exploration I hope to further research is the effect of inhibition on samples, especially in conjunction with describing eDNA abundance. For example, some samples benefitted from dilution to mitigate inhibition. Our semi-natural field samples likely contained plant secondary compounds that cause PCR inhibition (Schrader et al., 2012; McKee et al.,

2015b). Indeed, eDNA samples are notorious for their “dirtiness,” often containing secondary plant compounds, off-target eDNA, and humic acids (McKee et al., 2015a; Sassoubre et al.,

2016). Sample dilution, magnetic beads, specially designed Taq mastermixes, extraction using

QIAshredders, additional Proteinase K, heating elution buffer, and Zymo’s Inhibition removal kit have all been popular ways of mitigating inhibition in eDNA experiments (Goldberg et al., 2013,

2016; Davy et al., 2015; McKee et al., 2015b; Mckelvey et al., 2016; Olds et al., 2016; Tsuji et al., 2017). Inhibition probably confounds eDNA abundance studies as well. If eDNA is to truly 89 be used to estimate organism abundance, we must understand the “ecology” of eDNA – how

DNA interacts with its environment and what role inhibitors may play in this process (Barnes and Turner, 2016). Sampling an environment that is more productive biologically may encounter more inhibitors. Importantly, this is a consideration for eDNA sampling in a field setting

(Goldberg et al., 2016).

This thesis also highlights the limitations to eDNA technique. For example, eDNA can travel and may not be representative of its source (Wilcox et al., 2016), as Pond A may have shown. If eDNA methodology is sensitive enough, this technique may pick up eDNA from these alternative sources in the absence of the target organism (Rees et al., 2014; Thomsen and

Willerslev, 2015). One common example of natural contamination is when a predator releases the target organism’s eDNA in its fecal matter, or carries eDNA-positive water on its feathers, feet, or appendages from one pond to another (Merkes et al., 2014; Thomsen and Willerslev,

2015). Alternatively, a single carcass may continue to release eDNA (Merkes et al., 2014).

Additionally, absence of an organism’s DNA can never be proven – only presence (MacKenzie and Nichols, 2004). Thus to be, say, 95% confident of organism absence, scientists must use site occupancy modeling, which dictates that the amount of samples needed to support absence claims can vary with environmental condition (de Souza et al., 2016; Schmelzle and Kinziger,

2016).

Even though using eDNA has limitations, many technical hazards can be overcome with creative solutions. Synthetic positive controls allow for distinguishing between actual true positives and false positives, preventing contamination (Wilson et al., 2015). The use of Taqman probes in 90 qPCR may increase specificity and seems to be the direction of species-specific eDNA studies

(Goldberg et al., 2016). Having separate rooms for extraction and amplification with UV sterilization and bleach is quickly becoming the norm to prevent false positives and contamination (Goldberg et al., 2016). Negative controls at every step ensure every technical part of the experiment is sterile. From these precautions, scientists who employ eDNA can maintain confidence in their results.

To get around aforementioned limitations and the inability to obtain population demographic information, eDNA’s best use is currently a precursory sampling method before more intensive, targeted sampling. Because of the time it takes to find target organisms, using eDNA as a survey technique may be effective for finding new presence of species (Jerde et al., 2011; Pierson et al.,

2016). Several studies have used eDNA to search for novel sites of unknown occupancy, such as finding new populations of an endangered plethodontid salamander (Pierson et al., 2016) or detecting new invasive species presence (Darling and Mahon, 2011; Dejean et al., 2012; Newton et al., 2015). With new knowledge of where to target sampling efforts, scientists save time and money in finding and managing populations (Jerde et al., 2011; Bohmann et al., 2014).

Therefore, mere detection of presence with an eDNA assay is useful.

Beyond presence, it is our hope that eDNA abundance techniques pave the way to using non- invasive methods such as eDNA for population genetics. While we may never be able to tell population demographics with purely eDNA, population genetics may be within our reach in the next decade. The non-invasive methodology toolbox is growing rapidly and with the ability to sensitively detect genetic material from environmental samples, the ability to target sequences 91 for SNP detection is on the horizon. Scientists are already employing ddPCR for more sensitive assays to obtain copy number quantity at low abundances (Nathan et al., 2014; Cao et al., 2015;

Doi et al., 2015). With SNP data, it may be possible to get alleles from a population of a specific species (Livak, 1999). It is unlikely individual-level data can be elucidated, but the allelic diversity within a population may be inventoried via eDNA. Environmental DNA’s strength lies within its sensitivity and ability to “sample” organisms that may be secretive. Additionally, some eDNA may come in cellular form, an alternate route to obtaining genotypes (Turner et al., 2014).

A tagging system – potentially with antibodies and fluorescence – would have to be developed to sort these cells (Turner et al., 2014). Once individual genetic information is obtained, traditional population genetics analytical techniques may be employed to obtain levels of inbreeding, past bottlenecks, or population structure.

From adding an eDNA tool to the population biologist’s toolbox, this technique adds to organismal monitoring ability. Furthermore, eDNA is ideal as a non-invasive method because it does not harm the target organism (Taberlet et al., 1999). It is my hope that eDNA will be refined and implemented for identifying population presence and abundance. With eDNA abundance measures regularly taken, intensive traditional sampling may not need to happen, which could save valuable resources. Additionally, if eDNA for abundance assays is sensitive enough, population fluctuations could be tracked to alert managers should populations start to rapidly change (Dejean et al., 2012). From there, exploration of environmental and biotic factors may determine the cause of such a change and how it might influence the population. Although not addressed in this particular thesis, eDNA has the potential to survey species diversity

(Valentini et al., 2016). Surveys such as these taken over time can be used to address questions 92 of community composition that may become important in this time of rapid environmental change (IPCC 2007). Global change is increasingly ubiquitous and the demand for knowing how populations adjust to external and internal challenges is imperative to predicting future outcomes.

REFERENCES

Barnes, M. A., and C. R. Turner. 2016. The ecology of environmental DNA and implications for

conservation genetics. Conservation Genetics 17: 1-17. doi: 10.1007/s10592-015-0775-4

Bohmann, K., A. Evans, M. T. P. Gilbert, G. R. Carvalho, S. Creer, M. Knapp, D. W. Yu, and

M. de Bruyn. 2014. Environmental DNA for wildlife biology and biodiversity

monitoring. Trends in Ecology and Evolution 29:358–367. doi:

10.1016/j.tree.2014.04.003

Cao, Y., M. R. Raith, and J. F. Griffith. 2015. Droplet digital PCR for simultaneous

quantification of general and human-associated fecal indicators for water quality

assessment. Water Research 70:337–349. doi: 10.1016/j.watres.2014.12.008

Darling, J. A., and A. R. Mahon. 2011. From molecules to management: Adopting DNA-based

methods for monitoring biological invasions in aquatic environments. Environmental

Research 111:978–988. doi: 10.1016/j.envres.2011.02.001

Davy, C. M., A. G. Kidd, and C. C. Wilson. 2015. Development and validation of environmental

DNA (eDNA) markers for detection of freshwater turtles. PLoS ONE 10:e0130965. doi:

10.1371/journal.pone.0130965

Dejean, T., A. Valentini, C. Miquel, P. Taberlet, E. Bellemain, and C. Miaud. 2012. Improved

detection of an alien invasive species through environmental DNA barcoding: The

example of the American bullfrog Lithobates catesbeianus. Journal of Applied Ecology

49:953–959. doi: 10.1111/j.1365-2664.2012.02171.x

Doi, H., K. Uchii, T. Takahara, S. Matsuhashi, H. Yamanaka, and T. Minamoto. 2015. Use of

droplet digital PCR for estimation of fish abundance and biomass in environmental DNA

surveys. PLoS ONE 10:e0122763. doi: 10.1371/journal.pone.0122763 94

Goldberg, C. S., A. Sepulveda, A. Ray, J. Baumgardt, and L. P. Waits. 2013. Environmental

DNA as a new method for early detection of New Zealand mudsnails (Potamopyrgus

antipodarum). Freshwater Science 32:792–800. doi: 10.1899/13-046.1

Goldberg, C. S., C. R. Turner, K. Deiner, K. E. Klymus, P. F. Thomsen, M. A. Murphy, S. F.

Spear, A. McKee, S. J. Oyler-McCance, R. S. Cornman, M. B. Laramie, et al. 2016.

Critical considerations for the application of environmental DNA methods to detect

aquatic species. Methods in Ecology and Evolution 7:1299–1307. doi: 10.1111/2041-

210X.12595

Intergovernmental Panel On Climate Change (IPCC). 2007. IPCC Fourth Assessment Report:

Climate Change 2007. Intergovernmental Panel on Climate Change.

Jerde, C. L., A. R. Mahon, W. L. Chadderton, and D. M. Lodge. 2011. “Sight-unseen” detection

of rare aquatic species using environmental DNA. Conservation Letters 4:150–157. doi:

10.1111/j.1755-263X.2010.00158.x

Livak, K. J. 1999. Allelic discrimination using fluorogenic probes and the 5′ nuclease assay.

Genetic Analysis: Biomolecular Engineering 14:143–149. doi: 10.1016/S1050-

3862(98)00019-9

MacKenzie, D. I., and J. D. Nichols. 2004. Occupancy as a surrogate for abundance estimation.

Animal Biodiversity and Conservation 27:461–467.

McKee, A. M., D. L. Calhoun, W. J. Barichivich, S. F. Spear, C. S. Goldberg, and T. C. Glenn.

2015a. Assessment of environmental DNA for detecting presence of imperiled aquatic

amphibian species in isolated wetlands. Journal of Fish and Wildlife Management 6:498–

510. doi: 10.3996/042014-JFWM-034 95

McKee, A. M., S. F. Spear, and T. W. Pierson. 2015b. The effect of dilution and the use of a

post-extraction nucleic acid purification column on the accuracy, precision, and inhibition

of environmental DNA samples. Biological Conservation 183:70–76. doi:

10.1016/j.biocon.2014.11.031

Mckelvey, K. S., M. K. Young, W. L. Knotek, K. J. Carim, T. M. Wilcox, T. M. Padgett-

Stewart, and M. K. Schwartz. 2016. Sampling large geographic areas for rare species

using environmental DNA: A study of bull trout Salvelinus confluentus occupancy in

western Montana. Journal of Fish Biology 88:1215–1222. doi: 10.1111/jfb.12863

Merkes, C. M., S. G. McCalla, N. R. Jensen, M. P. Gaikowski, and J. J. Amberg. 2014.

Persistence of DNA in carcasses, slime and avian feces may affect interpretation of

environmental DNA data. PLoS ONE 9:e113346. doi: 10.1371/journal.pone.0113346

Nathan, L. M., M. Simmons, B. J. Wegleitner, C. L. Jerde, and A. R. Mahon. 2014. Quantifying

environmental DNA signals for aquatic invasive species across multiple detection

platforms. Environmental Science and Technology 48:12800–12806. doi:

10.1021/es5034052

Newton, J., A. Sepulveda, K. Sylvester, and R. Thum. 2015. Potential utility of environmental

DNA for early detection of Eurasian watermilfoil (Myriophyllum spicatum). Journal of

Aquatic Plant Management 54:46–49.

Olds, B. P., C. L. Jerde, M. A. Renshaw, Y. Li, N. T. Evans, C. R. Turner, K. Deiner, A. R.

Mahon, M. A. Brueseke, P. D. Shirey, M. E. Pfrender, et al. 2016. Estimating species

richness using environmental DNA. Ecology and Evolution 6:4214–4226. doi:

10.1002/ece3.2186 96

Pierson, T. W., A. M. McKee, S. F. Spear, J. C. Maerz, C. D. Camp, and T. C. Glenn. 2016.

Detection of an enigmatic plethodontid salamander using environmental DNA. Copeia

104:78–82. doi: 10.1643/CH-14-202

Rees, H. C., B. C. Maddison, D. J. Middleditch, J. R. M. Patmore, and K. C. Gough. 2014. The

detection of aquatic animal species using environmental DNA - a review of eDNA as a

survey tool in ecology. Journal of Applied Ecology. doi: 10.1111/1365-2664.12306

Sassoubre, L. M., K. M. Yamahara, L. D. Gardner, B. A. Block, and A. B. Boehm. 2016.

Quantification of environmental DNA (eDNA) shedding and decay rates for three marine

fish. Environmental Science and Technology 50:10456–10464. doi:

10.1021/acs.est.6b03114

Schmelzle, M. C., and A. P. Kinziger. 2016. Using occupancy modelling to compare

environmental DNA to traditional field methods for regional-scale monitoring of an

endangered aquatic species. Molecular Ecology Resources 16:895–908. doi:

10.1111/1755-0998.12501

Schrader, C., A. Schielke, L. Ellerbroek, and R. Johne. 2012. PCR inhibitors - occurrence,

properties and removal. Journal of Applied Microbiology 113:1014–1026. doi:

10.1111/j.1365-2672.2012.05384.x de Souza, L. S., J. C. Godwin, M. A. Renshaw, and E. Larson. 2016. Environmental DNA

(eDNA) detection probability is influenced by seasonal activity of organisms. PLoS ONE

11:e0165273. doi: 10.1371/journal.pone.0165273

Taberlet, P., G. Luikart, and L. P. Waits. 1999. Noninvasive genetic sampling: Look before you

leap. Trends in Ecology and Evolution 14:323–327. doi: 10.1016/S0169-5347(99)01637-

7 97

Thomsen, P. F., and E. Willerslev. 2015. Environmental DNA - An emerging tool in

conservation for monitoring past and present biodiversity. Biological Conservation

183:4–18. doi: 10.1016/j.biocon.2014.11.019

Tsuji, S., H. Yamanaka, and T. Minamoto. 2017. Effects of water pH and proteinase K treatment

on the yield of environmental DNA from water samples. Limnology 18:1–7. doi:

10.1007/s10201-016-0483-x

Turner, C. R., M. A. Barnes, C. C. Y. Xu, S. E. Jones, C. L. Jerde, and D. M. Lodge. 2014.

Particle size distribution and optimal capture of aqueous macrobial eDNA. Methods in

Ecology and Evolution 5:676–684. doi: 10.1111/2041-210X.12206

Valentini, A., P. Taberlet, C. Miaud, R. Civade, J. Herder, P. F. Thomsen, E. Bellemain, A.

Besnard, E. Coissac, F. Boyer, C. Gaboriaud, et al. 2016. Next-generation monitoring of

aquatic biodiversity using environmental DNA metabarcoding. Molecular Ecology

25:929–942. doi: 10.1111/mec.13428

Wilcox, T. M., K. S. McKelvey, M. K. Young, A. J. Sepulveda, B. B. Shepard, S. F. Jane, A. R.

Whiteley, W. H. Lowe, and M. K. Schwartz. 2016. Understanding environmental DNA

detection probabilities: A case study using a stream-dwelling char Salvelinus fontinalis.

Biological Conservation 194:209–216. doi: 10.1016/j.biocon.2015.12.023

Wilson, C. C., K. M. Wozney, and C. M. Smith. 2015. Recognizing false positives: Synthetic

oligonucleotide controls for environmental DNA surveillance. Methods in Ecology and

Evolution 7:23–29. doi: 10.1111/2041-210X.12452