Computational protocol: A Rapid, Strong, and Convergent Genetic Response to Urban Habitat Fragmentation in Four Divergent and Widespread Vertebrates

Similar protocols

Protocol publication

[…] We used the program ARLEQUIN to estimate pair-wise FST values between patches using the infinite-allele model and 1000 permutations for significance , . We also calculated pair-wise FST between arrays within large and core patches with ARLEQUIN to show genetic divergence between sampling sites that were located within a patch of continuous habitat. For this calculation we also included some sampling sites from core areas of continuous habitat that were outside of the Simi Hills (our study area), but within SMMNRA, with an average of 4.28 km (range 1.8–6.6 km) separating these sites.To examine patterns of sample clustering based on genetic similarity, we used the program STRUCTURE v. 2.3.1 . We chose the LOCPRIOR model , assumed populations were not admixed and that allele frequencies were correlated between populations, and ran 100,000 MCMC chains with a 10,000 burn-in. We ran seven runs each of K = 1 to K = number of sample sites () for each species. We compiled results from our STRUCTURE runs with the program STRUCTURE HARVESTER (Dent Earl, http://taylor0.biology.ucla.edu/struct_harvest/). To determine the most likely K, we calculated the posterior probabilities of the mean of seven runs at each K (; ).Isolation by distance, as revealed by a correlation between pairwise genetic and geographic (Euclidean) distances using a Mantel test, was performed using IBDWS 3.14 . IBDWS uses a Reduced Major Axis (RMA) regression to estimate the slope and intercept of the isolation by distance relationship.To test for the effect of major roads, highways, and patch age on genetic divergence, we performed partial Mantel tests in IBDWS 3.14. Partial Mantel tests determined correlations of roads presence (RDS), highway presence (HWY), and patch age of isolation (AGE) on a genetic divergence matrix, while holding geographic distance constant. Tests were performed separately, one for each of these three variables, and all animals that were captured within a patch were used to calculate a patch average genetic divergence (FST; as calculated in ARLEQUIN, see above). The presence of major roads and the presence of Highway 23 were used separately in the analysis because the highway in our study area is larger and has more traffic than other roads. Also, several habitat fragments are only separated by major roads. Age of isolation was chosen because this measure incorporates not only when roads and freeways were built, but also when residential and commercial developments were erected.We mapped genetic distance on the landscape using Alleles in Space (AIS) and the landscape shape interpolation . We used a Delaunay triangulation-based connectivity network to identify midpoints between our sample sites, then the raw genetic distance (Dij) at each midpoint was calculated . This genetic distance measure is similar to Nei's standard genetic distance (Ds; ), where Dij is 0 if individuals are completely genetically identical, and Dij is 1 if individuals are completely genetically dissimilar. We did not calculate the residual genetic distance, because we did not find a significant isolation by distance effect in the Simi Hills samples for any species (see ). By this method, a landscape of genetic distances between sampling sites are expressed as “surface heights” and are displayed as a 3-dimensional graph. To better visualize the AIS height output, we imported the output file into ArcGIS 9.3 (ESRI Corporation, Redlands, CA) and created a 2-dimensional color hot-spot map overlaid on the geographic study area. Colors correspond to “heights” of genetic distance between points (e. g. ). [...] We used the program GENALEX to calculate the genetic diversity indices of within-patch expected heterozygosity (He), observed heterozygosity (Ho), number of effective alleles (NA), and relatedness (RLR) . We used the Lynch & Ritland (1999) estimator of relatedness because it has been shown to perform well in simulations for a wide range of marker data and population structure . We performed a rarefaction analysis using the web-based program RERAT which uses multiple simulations to determine the change in relatedness values as additional microsatellite loci are added. In RERAT, we performed 100 simulations and used the Lynch and Ritland (1999) relatedness analysis for each of the four species. For lizards, cores and large patches had three pitfall trap arrays while small patches had one (). To reduce bias because of array clustering, we calculated pair wise relatedness of all individuals caught in the same array, and then used the mean of those within-array measures to calculate within patch relatedness.We used the program STATA 9 (StataCorp, College Station, TX) to transform variables until they approached normal distributions and then to examine the relationship between the indices of genetic diversity and the size, degree of isolation, and age of the habitat patches. We used unpaired t-tests (with unequal variance when necessary) and Bonferroni corrections to compare genetic diversity measures between small and large/core habitat patches. Degrees of freedom for t-tests were calculated using the Satterthwaite (1946) method . We lumped large patches and core areas for this analysis because, for these small species, population size is likely equivalently large in the large patches and the core areas, and because the numbers of sites were relatively small for core areas (n = 2) and large patches (n = 3). To test for a relationship between patch isolation and genetic diversity, we used linear regression to examine the relationship of the genetic diversity indices with the size, pair wise age of isolation, and proximity (PROX) of the habitat patches, where the degree of isolation of a patch is the inverse of proximity. Spearman's rank correlations were used to test for significant associations between patch age and genetic diversity. […]

Pipeline specifications

Software tools Arlequin, IBDWS, AIS, GenAlEx
Application Population genetic analysis
Organisms Homo sapiens