Conservation Genetics of Eastern Massasauga Rattlesnakes
Sistrurus c. catenatus

H. Lisle Gibbs, Department of Biology, McMaster University

The Animal: The eastern massasauga rattlesnake, Sistrurus c. catenatus is in decline throughout its range in eastern North America (Greene & Campbell 1992), is classified as threatened in Canada (Weller & Parsons 1991), and unofficially regarded as such in the U.S. Among the largest extant populations are those found in two regions of Ontario: the eastern shore of Georgian Bay and the Bruce Peninsula (Fig. 1). Aside from moderately large populations that may persist on the lower peninsula of Michigan, relatively small habitat-isolated remnant populations are all that remain in the states of Illinois, Indiana, Iowa, Minnesota, Missouri, New York, Ohio, Pennsylvania, and Wisconsin (see Johnson & Menzies 1993). The current fragmented distribution of massasaugas appears to be a result of several interacting factors including narrow habitat use (Reinert & Kodrich 1982; Weatherhead & Prior 1992), climatic change and the natural succession of vegetation communities (Weller & Oldham 1993; Johnson 1995), and destruction of habitat by humans (Greene & Campbell 1992).

Figure 1. Distribution of known remnant populations of massasaugas in the Great Lakes region (stars), including those used in the present study: Killbear Provincial Park, Beausoleil Island, Bruce Peninsula National Park, Cicero Swamp, and Springfield. Wavy lines on the eastern shores of Georgian Bay, the Bruce Peninsula, and Manitoulin Island represent the approximate ranges of the largest extant regional populations of this species (from Gibbs et al. 1997).

To develop effective recovery strategies for this species it would be useful to know how genetically distinct isolated populations are from each other. Such information would provide an empirical basis for defining local populations and conservation management units (c.f. Moritz 1994) and for ranking populations vis a vis their conservation value. To address this issue a group of scientists including field biologists at Carleton University (Kent Prior, Chris Parent and Pat Weatherhead) and a molecular ecologist at McMaster University (H. Lisle Gibbs) have been funded by Parks Canada to conduct DNA-based genetic analyses on rattlesnake populations in eastern Canada and the United States. Below, I briefly describe the genetic techniques we have used, the major results obtained so far and the conservation implications of our findings. Much of this material is summarized from the paper by Gibbs et al. (1997).

Genetic Tools: For our genetic analyses, we used a recently-discovered type of DNA marker called microsatellite DNA markers (see Quellar et al 1993). These genetic markers are based on the fact that at specific positions in the DNA of all organisms (called loci) there exists non-coding (or junk) DNA which is made of small repeats 2-3 bases in size such as, for example the repeat unit CA, which is made up of the bases cytosine and adenine. The number of these repeats present in the two alleles at a particular locus in any individual is highly variable apparently due to high mutation rates at these loci. Some individuals will have, for example, alleles consisting of 5 and 10 repeat units while others will have alleles with 15 and 25 units respecitively. The advantage of these highly variable markers is that they can allow us to detect when populations become isolated from each other. This is because random genetic process such as genetic drift will cause genetic differences to develop between two populations which are no longer exchanging individuals through migration; this phenomenon can be more easily detected using highly variable marker systems.

As described by Gibbs et al. (1998) six microsatellite loci were isolated for massasauga rattlesnakes and then used for genetic analyses of samples collected from five eastern massasauga rattlesnake populations: two from the U.S. (Springfield, Ohio and Cicero Swamp near Syracuse, New York) and three from around Georgian Bay, Ontario (Bruce Peninsula National Park, Beausoleil Island which is part of Georgian Bay Islands National Park, and Killbear Provincial Park). DNA was obtained from small (50 ul) blood samples collected from the caudal vein of individual snakes (25 - 80 per population) which had been captured and then released unharmed. These samples were taken to the lab at McMaster where genetic analyses were performed by amplifying the microsatellite loci from the DNA of all individuals using the polymerase chain reaction (PCR) and then determining the sizes of the alleles present in different individuals by running the PCR products out on a polyacrylamide gel. Figure 2 shows an example of the type of variation detected among samples when using snake-specific microsatellite loci. The data was then analyzed using standard population genetic methods (cf. Hartl and Clark 1987).

Figure 2. Allelic variation at microsatellite locus Sc 07 among 13 individuals (A - M) from Killbear Provincial Park . Lanes marked with + are the products of amplifications from a plasmid clone containing a known size (176 bp) allele (from Gibbs et al. 1998).

Major Results: Genetic analyses of the type described above yielded three major results with respect to population genetic structure in these snakes. These are described in more detail in the paper by Gibbs et al. (1997).

1. All Sampled Populations Are Genetically Distinct. All of the populations that were compared were genetically different, even populations, such as Beausoleil and Killbear, that were < 50 km apart from each other. This was determined in three ways. First, tests which compared the frequency of alleles between all pair-wise combinations of populations showed statistically significant differences for 97% (58 of 60) possible population-by-locus comparisons. Thus, geographically separate massasauga populations differ significantly in their allele frequencies and this pattern is consistent amongst populations regardless of the geographical distance between them.

A second way to compare differentiation between populations is to calculate an Fst value (see Hartl and Clark 1987) which estimates the proportion of the overall genetic variation pooled across all loci which is, on average, unique to specific populations. In general, Fst values of 0.05 - 0.15 indicate moderate differentiation whereas values > 0.15 indicate substantial differentiation between populations. Fst values for massasauga populations ranged from a low of 0.085 calculated for geographically isolated populations occupying opposite sides of Georgian Bay (Bruce and Beausoleil) to a high of 0.261 for the comparison of Killbear and Cicero populations. The overall Fst value was 0.164 indicating that within-population variation averaged 16% of the overall variation detected.

Finally, each massasauga population we sampled also contained a substantial proportion of unique alleles. Between 14.8 - 32.7% (mean = 22.7%) of all alleles detected within population samples were found to be population-specific. Furthermore, between 2.0 - 7.2% (mean = 5.5%) of these unique alleles occurred at frequencies of 5%. The number of individuals sampled per population does not appear to have had a significant effect on the proportion of population-specific alleles detected, since the proportion of alleles classified as unique is roughly the same in populations from which the largest (Killbear) and smallest (Springfield) number of snakes were sampled. These results indicate that each of our study populations contain a distinct subset of the overall genetic variation resolved by the microsatellite markers we used.

Overall, our results suggest that geographically separate populations of massasaugas are genetically distinct and that such populations harbor a unique and substantial portion of the total genetic variation found in this sub-species. The only way in which high levels of differentiation could develop and persist through time is if migration of individuals between populations was extremely limited or non-exisitent. On the basis of these results we conclude that these snake populations are demographically isolated from each other.

2. Genetically-distinct Sub-Populations May Exist: One of our most surprising results is that additional genetic analyses suggest that genetically distinct sub-populations of massasaugas may exist on extremely small geographic scales of < 2 km. The best evidence for this comes from the Killbear population. Here, as described by Gibbs et al. (1998), although samples were collected from an extremely limited area (9 km2), most samples could be clustered into two groups (Twin Points; n = 27 and Blind Bay; n = 41) which had been collected at sites whose geographic centers were roughly 1.5 km apart. Comparisons of these samples showed highly significant differences in allele frequencies at five of six loci and a highly significant overall Fst value of 0.040. Similar results, albeit for samples collected from more widely separated populations, were obtained for the Bruce and Ohio populations. Overall, these results from three different sites support the interpretation that microgeographic genetic differentiation on a scale of < 5 km exists within at least some of the sampled populations. To our knowledge differentiation on this geographic scale has never before been reported for snake populations.

Biological and Conservation Implications: Our results have implications for understanding the biology of these snakes and for developing conservation plans to protect them. From a biological perspective, one major result was that all sampled populations are genetically distinct and that each contains population-specific alleles. For some populations (e.g. Cicero and Springfield) this result was not surprising, because these populations are isolated from other populations by large landscape barriers composed of unsuitable urban and agricultural habitats. However, we found similar levels of genetic distinctiveness for populations around Georgian Bay. These populations are geographically quite close to each other (Beausoleil and Killbear sites are approximately 50 km apart; Fig. 1), and high quality habitat occupied by other massasauga populations exists between them (Weller & Oldham 1993). Nonetheless, these populations differ significantly in allele frequencies at most loci, have positive values for Fst, and contain substantial numbers of unique alleles. These population-level results suggest that gene flow between populations is very restricted and that the populations we sampled have been genetically isolated from each other for some time. Whatever the actual time that these populations have been isolated, it certainly exceeds the several hundred years that encompasses the period of European settlement, with its accompanying large-scale alteration of habitat (e.g., logging). Thus, low levels of gene flow and genetic isolation appear to be the natural state for eastern massasauga populations rather than being recently induced by humans, although habitat destruction is certainly a major threat to extant populations (Prior and Weatherhead, in prepartion).

The fact that we detected genetic differences between snakes sampled only a few km apart suggests that natal dispersal is very limited, which in turn should increase the likelihood of inbreeding. A related possibility is that the fine-scale structure we detected may be a consequence of sampling sets of closely-related individuals. Such a phenomenon could result if very few females produce most of the recruits in a particular area. To assess the merit of these various biological explanations for the fine-scale genetic differentiation that we observed will require studies that document the reproductive success of individual males and females, as well as dispersal patterns and survival rates of juveniles and adults in an area where microgeographic differentiation has been detected. We are currently conducting such a study on individuals in the Killbear population (Parent et al., unpublished data).

From a conservation perspective, our data help define the spatial and ecological scale at which massasauga populations should be monitored and managed. At the very least, we should now consider the regional populations of the Bruce Peninsula and Georgian Bay to be composed of a series of partially independent local populations or management units (Moritz 1994). However, additional fine-scale genetic data (e.g. locality-based samples < 10 km apart) are required for a more precise delimitation of what constitutes a local population of these snakes

One practical implication of our results is that if local populations in regions such as Georgian Bay suffer catastrophic population reductions or extinctions (e.g. via fire or disease), natural re-population through immigration is likely to be very slow. If so, active managed translocations might be necessary to re-establish populations. Furthermore, given the genetic distinctiveness of individual populations, the source of donor individuals for such translocations would have to be carefully evaluated if maintaining the current genetic structure is deemed an important conservation goal.

Finally, that individual massasauga populations appear to contain unique portions of the total genetic variation suggests that each population is likely to be of particularly high conservation value. This argument relies on a long-standing assumption in conservation genetics that presumably neutral variation such as that detected using microsatellites is correlated with variation in traits which would preserve the evolutionary potential of this species (e.g. Avise 1994). We have no information on such a relationship in these animals, but such a correlation has been documented in other species (e.g. Vrijenhoek et al. 1985).

Acknowledgements. I am grateful to K. Hedgecock, T. Jaworski, C. Parent, D. Sweiger, and M. Villeneuve for assistance in obtaining blood samples from massasaugas, to L. Collins and L. DeSouza for careful preparation and analysis of DNA samples, and to my collaborators in this work, K. Prior, P. Weatherhead, C. Parent and G. Johnson. Primary funding for this work was provided by Parks Canada and we especially thank B. Hughson, as well as B. Hutchinson and B. Stephenson for their efforts in managing this process.

LITERATURE CITED

Avise JC (1994) Molecular markers, natural history, and evolution. Chapman and Hall, New York.

Gibbs, HL, Prior KA, Weatherhead PJ, Johnson G (1997) Genetic structure of populations of the threatened eastern massasauga rattlesnake (Sistrurus c. catenatus): Evidence from microsatellite DNA markers. Molecular Ecology (in press).

Gibbs HL, Prior KA, Parent C (1998) Characterization of DNA microsatellite loci from a threatened snake, the eastern massasauga rattlesnake (Sistrurus c. catenatus) and their use in population studies. Journal of Heredity (in press).

Greene HW, Campbell JA (1992) The future of the pitvipers. In Biology of the pitvipers (eds. J. A. Campbell & E. D. Brodie), pp. 421-427. Selva, Tyler, Texas.

Hartl D, Clark AG (1987) Principles of population genetics, 2nd. ed. Sinauer Associates, Sunderland, Massachusetts.

Johnson B, Menzies V (eds) (1993) International symposium and workshop on the conservation of the eastern massasauga rattlesnake. Metro Toronto Zoo, Toronto

Johnson G (1995) Spatial ecology, habitat preference, and habitat management of the eastern massasauga, Sistrurus c. catenatus in a New York weakly- minerotrophic peatland. Ph.D. dissertation, SUNY, Syracuse, New York.

Moritz C (1994) Applications of mitochondrial DNA analysis in conservation: a critical review. Molecular Ecology, 3, 401-411.

Prior KA, Weatherhead PJ (In preparation). National Recovery Plan for the Eastern Massasauga Rattlesnake. For Ontario Ministry of Natural Resources

Quellar DC, Strassman JE, Hughes CR (1993) Microsatellites and kinship. Trends in Ecology and Evolution, 8, 285-288.

Reinert HK, Kodrich WR (1982) Movements and habitat utilization by the massasauga, Sistrurus catenatus catenatus. Journal of Herpetology, 16, 162-171.

Vrijenhoek RC, Douglas ME, Meffe GK (1985) Conservation genetics of endangered fish populations in Arizona. Science, 229, 400-402.

Weatherhead PJ, Prior KA (1992) Preliminary observations of habitat use and movements of the eastern massasauga rattlesnake (Sistrurus c. catenatus). Journal of Herpetology, 26, 447-452.

Weller WF, Oldham MJ (1993) Historic and current distribution and status of the eastern massasauga rattlesnake, Sistrurus catenatus, in Ontario, Canada. In International symposium and workshop on the conservation of the eastern massasauga rattlesnake (eds Johnson B, Menzies V), pp. 35-39. Metro Toronto Zoo, Toronto.

Weller WF, Parsons HJ (1991) Status of the eastern massasauga rattlesnake, Sistrurus catenatus, in Canada. COSEWIC, Ottawa.