Skip to main content
Dryad logo

Data from: Low bottleneck detection in long-lived species despite lost genetic diversity: a case study of tuatara and eastern massasauga rattlesnakes

Citation

Bradke, Danielle R. et al. (2021), Data from: Low bottleneck detection in long-lived species despite lost genetic diversity: a case study of tuatara and eastern massasauga rattlesnakes, Dryad, Dataset, https://doi.org/10.5061/dryad.wpzgmsbn3

Abstract

Population bottlenecks can reduce genetic diversity and may lead to inbreeding depression. However, some studies have provided evidence that long lifespans buffer negative genetic effects of bottlenecks. Others have cautioned that longevity might merely mask the effects of genetic drift, which will still affect long-term population viability. We used microsatellite data from actual populations of tuatara (Sphenodon punctatus) and eastern massasaugas (Sistrurus catenatus) as a starting point for simulated population declines to evaluate the performance of bottleneck tests under a range of scenarios. We quantified losses in genetic diversity for each scenario and assessed the power of commonly used tests (i.e., M-ratio, heterozygosity excess, and mode-shift) to detect known bottlenecks in these moderate- to long-lived species. Declines in genetic diversity were greater in bottlenecks simulated for eastern massasaugas, the shorter-lived species, and mode-shift and heterozygosity excess tests were more sensitive to population declines in this species. Conversely, M-ratio tests were more sensitive to bottlenecks simulated in tuatara. Despite dramatic simulated population declines, heterozygosity excess and mode-shift tests often failed to detect bottlenecks in both species, even when large losses in genetic diversity had occurred (both allelic diversity and heterozygosity). While not eliminating type II error, M-ratio tests generally performed best and were most reliable when a critical value (Mc) of 0.68 was used. However, in tuatara simulations, M-ratio tests had high rates of type I error when Mc was calculated assuming θ = 10. Our results suggest that reliance on these tests could lead to misguided species management decisions.

Methods

The original/raw tuatarta genotypes were obtained at seven microsatellite loci collected from 272 individuals sampled from Takapourewa/Stephens Island, New Zealand between March 2003–March 2007. Sampling occurred at eight sites on the island and was distributed throughout the available remnant forest and pasture landcover types.

For simulated tuatara and eastern massasauga microsatellite genotypes we used program BOTTLESIM to to simulate multiple bottleneck scenarios for each species. For all simulations, we performed 1,000 iterations and set the reproductive system as dioecy with random mating, a 1:1 sex ratio, and 100% generational overlap.

For simulation dataset 1, simulations varied by 1) length of time (tB): 50, 100, and 200 years, 2) starting population size (N0): estimated census population size and one order of magnitude above and below, and 3) bottlenecked population size (NB): 20, 50, 100, 1000, and N0 (the lattermost NB = N0 scenario representing a constant population size with no bottleneck). For eastern massasauga simulations, we assumed a census population size of 6,300 individuals, age of sexual maturity of 5 years, and longevity of 12 years. For tuatara simulations, we assumed a census population size of 25,000 individuals, age of sexual maturity of 14 years, and longevity of 80 years.

For simulation dataset 2, we randomly selected seven loci from the eastern massasauga dataset. Using this reduced dataset, we conducted a set of simulations with the same values for tB and NB as described above and set N0 equal to estimated census population size.

For simulation dataset 3, we simulated bottlenecks using the values for tB and NB described above and set N0 equal to estimated census population size for each species. Age of sexual maturity and longevity were set at 10 years and 50 years, respectively, for both species.

Usage Notes

All datasets are in Genepop format and missing data are indicated with 000000.