HETEROZYGOSITY OF A WOOD TURTLE POPULATION IN CENTRAL NY Kyle Pursel Department of Biological Sciences This presentation will report on the work and results of a genetic study on the wood turtle, Glyptemys insculpta at two sites in Eastern Oswego County, NY for the Student/Faculty Challenge Grant. This study looked at tandemly repeating segments of nuclear DNA, referred to as microsatellites, to assess heterozygosit y and relatedness of marked turtles at both sites. This data is part of a larger project assessing the status of wood turtles at both sites as a part of my honors thesis. I. Introduction Wood turtles, Glyptemys insculpta are riparian species of turtle native to New York State and much of the Northeastern US, Great Lakes, and Southeastern Canada (Ernst et al., 1994). Declines have been documented in populations throughout their range in recent decades (Daigle & Jutras, 2005; Garber & Burger 1995; Harding, 1991; Saumure & Bider, 1998). As a result of the these declines, many states and Canadian provinces are providing protective status to wood turtles by listing th em as endangered, threatened, or a species of special concern (Bowen & Gillingham, 2004). In New York State, the wood turtle is a species of special concern (Al Breisch, personal communication). In 2005, Dr. Peter Rosenbaum and Kyle Pursel initiated surveys to find and study wood turtles in eastern Oswego County in New York State to determine the natural history, ecology, and status of this species in this pa rt of their range. In 2007, a genetic proponent was added to the study to assess the genetic vari ation of turtles found at the study sites and to aid in the assessment. A genetic component was added due to the va luable data genetics can provide about a population. It was determined that microsatellites, which measure tandemly repeating sequences of nuclear DNA and are generally considered to be good measures of heterozygosity, were chosen to calculate the genetic diversity of the population. Heterozygosity is the measure of heterozygotes in a population, and a heterozygote is an individual which has two separate alleles for a given gene. With microsatellites, an allele can be considered to be any variation in the number of tandem repeats (Frankham et al., 2004). For example, an individual that is heterozygous can have one allele for 30 repeats and another allele for 33 repeats, which a homozygous individual would have either two alleles with 30 or two alleles with 33 repeats. Aiding in the decision was also a study released in 2005 which calculated the heterozygosities of six populations
K. Pursel 2 in Quebec, Canada, which provided the data and names of primers proven to work with wood turtles for microsatellite studies (Tessier et al., 2005). II. Methods Blood samples were taken from each turtle using a vacuum syringe. Samples of 1/10 to cc were taken depending upon the size of the turtle and ease of blood taking. Samples were preserved in a clotting buffer and stored in a -40oF freezer until use. Thawed samples were extracted using a PureGene DNA extraction kit and the corresponding manufacturers protocol for clotted blood. The general pro cess consists of lysing the blood cells of the clotted blood were and protein denaturation for subsequent protein removal using proteinase K and centrifugation. For each extraction, roughly 50 L of clotted blood was added to 550 L of cell lysis solution and 3 L of proteinase K solution in a sterile 1.5mL labeled centrifuge tube. This was i nverted 25 times and incubated at 55oC overnight. The sample was then cooled to room temperature and 200 L of protein precipitation solution was added. The solution was then vortexed at high speed for 20 seconds and then placed on ice for 5 minutes before being centrifuged on the highest setting for approximately 5 minutes. The resulting supernatant containing the DNA was then poured into a new sterile 1.5mL labeled centrifuge tube containing 600 L of 100% isopropanol and was mixed by inversion 50 times. The sample was then centrifuged again on high for approximately 1.5 minutes. The new supernatant was poured out a nd the tube allowed to dry while inverted. Then, 600 L of 70% ethanol was added and then centrifuged for another 1.5 minutes. The supernatant was again drained and inverted for about 10 minutes to dry. Once dry, 20 L of DNA hydration solution was added to the DNA pellet and incubated at 65oC for approximately one hour. Once hydrated, the solution was stored at 4oC until needed. DNA was then quantified using a fluorometer. Once quantified, DNA extractions were diluted for PCR. Frozen PCR materials were allowed to thaw and then mixed to make a m aster solution that could be added to the DNA samples. Added materials incl ude 10X PCR buffer, dNTP, MgCl2, and Taq polymerase. During this time, the forward and reverse primer s for one of the five microsatellite primers was added to the master solution. The a ppropriate amounts of master mix and DNA solution were added to a sterile labeled 0.2m L centrifuge tube. This was centrifuged briefly before being placed in a thermocycler. Th e sequence of the program used for the thermocycler was adapted from that used by Tessier et al. (2005) and goes as follows: 2 minutes at 95oC for one cycle, 35 cycles of 94oC for 45 seconds, 54oC for 45 seconds, 72oC for one minute. Samples were then prepared and ran on a Beckman-Coulter CEQ 8000 Genetic Analyzer. Samples were placed in trays and ran in acrylamide gels through capillaries. Fragments would travel at different speeds thro ugh the gel in the capillaries and, when they reached a certain point, a laser would detect the fluorescent tag in the fragments and determine the base pair composition. Once comp lete, the data came out in graphs as a series of one or two peaks, depending on if that individual was heterozygous or homozygous for that loci. From these peaks, the amount of base pairs or nucleotide
Heterozygosity of a Wood Turtle Population in Central NY 3 compliments, was taken and analyzed using the freeware program PopGene to determine the heterozygosity, allele numbers and freque ncies, and Hardy-Weinberg Equilibrium for both sites. The freeware program Structur e was used to determine if both sites compromised one large population or separate populations. III. Results Three of the five loci tested yielded recordable results. The loci which worked were GmuB21, GmuD16, and GmuD93. Not all turtle blood samples worked, with twenty-three different individuals working w ith one to all of the three loci. Base pair data for loci GmuB21 was obtained for 14 individuals from Sloperville and 3 from Little Grindstone, 13 in Sloperville and 4 for Little Grindstone for loci GmuD16, and 11 for Sloperville and all samples for Little Grindstone with loci GmuD93. Analysis using the program Structure could no t differentiate individuals form either site into separate populations. Fig. 1 clearly s hows the close relatedness of turtles from each site to one another, with each turtle consis tently having approximat ely 50% likelihood of being assigned to one site or the other during an alysis. This data is consistent with both sites being part of a single larger population. Fig. 1 Triangle Plot of the likelihood of turtle being in separate populations. Note that all points lie roughly in the center (50%) ra nge of the plot. Red dots are individuals from Sloperville, green dots are i ndividuals from Little Grindstone
K. Pursel 4 Analysis using PopGene shows a total of 35 different alleles total from both sites together. Totals of 13, 9, and 13 different alleles were found for each loci (GmuB21, GmuD16, and GmuD93) respectively. Little Gri ndstone had fewer alleles per loci than Sloperville, but did contain 2 private alleles for GmuD93. Table 1 summarizes allele data for all loci and both site pooled together a nd separated. Sloperville consistently had larger numbers of alleles than Little Grindstone. Table 1. Observed Alle le Numbers per Loci =========================================================================== Locus Diploid # of alleles Observed allele total Sloperville LittleGrindstone =========================================================================== GmuB21 34 13 13 3 GmuD16 34 9 9 2 GmuD93 32 13 11 8 Mean 33 11.6667 St. Dev 2.3094 =========================================================================== In all three loci, the observed hetero zygosity was lower than the expected heterozygosity (Table 2). Tests of Hardy-We inberg equilibrium within both sites pooled together as one population show that loci Gm uB21 and GmuD93 were significantly out of equilibrium (B21 p=<0.001 Chi-Square=142.7; D93 p=0.004, Chi-Square=114.2), whereas loci GmuD16 was not significantly different fro m equilibrium (p=0.741, Chi-Square=30.2). Table 2. Observed and Expected Heterozygosity of Both Sites ============================================================ Locus Diploid Sample Size Observed Heterozygosity Expected Heterozygosity ============================================================ B21 34 0.5294 0.8663 D16 34 0.6471 0.7968 D93 32 0.6875 0.8851 Mean 33 0.6213 0.8494 St. Dev 0.0821 0.0465 ============================================================
Heterozygosity of a Wood Turtle Population in Central NY 5 Data for the individual loci and turtle were not obtained from the Quebec investigators (Tessier et al., 2005). A simple eye-ball comparison of data shows a similar trend amongst all populations (Table 3), in which all populati ons appear to have relatively high numbers of alleles and heterozygosity. However, thorough statistical analysis could not be conducted since only 3 of the 5 loci used in the Quebec studies yielded data for the NY population. Table 3. Heterozygosity and Allele Comparisons Between Quebec & NY Populations ============================================================================= Site Sample Size (turtles) Observed Allele # Observed Heterozygosity ============================================================================= Quebec FA* 46 50 0.561 Quebec SH* 39 46 0.673 Quebec SU* 8 40 0.886 Quebec MI* 19 45 0.804 Quebec TO* 6 36 0.767 Quebec DC* 18 54 0.837 Central NY ~23 35 0.621 ============================================================================ *See Tessier et al., 2005. for in-depth site data and name meanings. IV. Discussion Turtles from both appear to be or recently were part of a larger population that has and continues to be fragmented by human development of the land. Although 2 out of the 5 loci failed to yield reliable data, a relatively larg e number of alleles were found for Sloperville and the overall population. This data indicates that turtles from Sloperville and the overall population are possibly genetically stable. Howe ver, Little Grindstone had lower allele numbers for 2 of the 3 loci. This may be an indication that Little Grindstone, through recent isolation and human impacts, may be faci ng an ecological and genetic bottleneck, and might be losing some of the genetic diversity that is present at other sites. This is consistent with the turtle collection data, which only yi elded 5 turtles over 3 years. Since one loci, GmuD93, did yield a relatively decent number of alleles, it is possible that the addition of other loci may show that Little Grindstone turtles are more genetically variable than the current data suggests.
K. Pursel 6 The heterozygosity and Hardy-Weinberg data suggests that the populations have fewer heterozygotes than expected for the number of loci present. However, Hardy-Weinberg is based off of many assumptions of nonrandom mating, no selection pressures (for that gene), no mutation, no migration, no genetic drif t, and large populations. It is possible that sample sizes are not large enough to provide en ough data and that the population is within equilibrium. However, it is also possible th at there have been declines within the population, as appears to have happened at the Little Grindstone site, which could also lead to the elimination of some alleles and cha nges in allele frequency through genetic drift, since there should be little to no selection o ccurring for these genes, unless they are linked to selected genes. V. Acknowledgements I would like to thank the numerous volunteers w ho have helped me collect field data and turtles and the landowners who have graciously given me permission to conduct the field research and turtle surveys that led to this pr oject. Funding for this work was provided by a Student/Faculty Challenge Grant. I would especi ally like to thank my two advisors, Dr. Peter Rosenbaum and Dr. Amy Welsh, both of whom have provided much help and advice. VI. Literature Cited Bowen, K. D. & Gillingham, J. C. (2004). R9 Species conservation assessment for wood turtle Glyptemys insculpta (LeConte, 1830). U. S. Forest Service, Eastern Region, Milwaukee, WI. Daigle, C & Jutras, J. (2005). Quantitative evid ence of decline in a southern Quebec wood turtle ( Glyptemys insculpta ) population. Journal of Herpetology 39(1): 130-132. Ernst, C., Jeffrey, E., Lovich, E., and Barbour, R. W. (1994). Turtles of the United States and Canada Smithsonian Institute Press, Washington, D. C. Garber, S. & Burger, J. (1995). A 20-yr study documenting the relationship between turtle decline and human recreation. Ecological Applications 5(4): 1151-1162. Harding, J. H. (1991). A twenty year wood turtle study in Michigan: implications for conservation. First International Symposium on Turtles & Tortoises: Conservation and Captive Husbandry : 31-35. Saumure, R. A. & Bider, J. R. (1998). Impact of agricultural development on a population of wood turtles ( Clemmys insculpta ) in southern Quebec, Canada. Chelonian Conservation and Biology 3(1): 37-45. Tessier N, Paquette, S. R., & Lapointe, F. J. (2005). Conservation genetics of the wood turtle ( Glyptemys insculpta ) in Quebec, Canada. Canadian Journal of Zoology 83(6): 765-772.