Understanding the biotic consequences of Pleistocene range shifts and fragmentation remains a fundamental goal in historical biogeography and evolutionary biology. Here, we combine species distribution models (SDM) from the present and two late Quaternary time periods with multilocus genetic data (mitochondrial DNA and microsatellites) to evaluate the effect of climate-induced habitat shifts on population genetic structure in the Large-blotched Ensatina (Ensatina eschscholtzii klauberi), a plethodontid salamander endemic to middle and high-elevation conifer forest in the Transverse and Peninsular Ranges of southern California and northern Baja California. A composite SDM representing the range through time predicts two disjunct refugia, one in southern California encompassing the core of the species range and the other in the Sierra San Pedro Mártir of northern Baja California at the southern limit of the species range. Based on our spatial model, we would expect a pattern of high connectivity among populations within the northern refugium and, conversely, a pattern of isolation due to long-term persistence of the Sierra San Pedro Mártir population. Our genetic results are consistent with these predictions based on the hypothetical refugia in that (i) historical measures of population connectivity among stable areas are correlated with gene flow estimates; and (ii) there is strong geographical structure between separate refugia. These results provide evidence for the role of recent climatic change in shaping patterns of population persistence and connectivity within the Transverse and Peninsular Ranges, an evolutionary hotspot.
Geneland coordinates
This is the spatial coordinate file (in UTM format) for Geneland analyses. There is one line per individual and two columns representing the x-axis and y-axis coordinate. Original latitude/longitude coordinates were collected in the field using a hand-held GPS (WGS84 datum) and subsequently converted to UTM format in ArcView.
Geneland genotypes
This is the microsatellite genotypes file for Geneland analyses. There is one line per individual with 20 columns (diploid, codominant data). Missing data are represented by 000.
Geneland individual labels
This is the file of individual labels for Geneland analyses in the same order as the genotypes and coordinates files.
Structure infile
This is the input file for Structure analyses. The second column contains sample group information used as a prior. Missing data are represented by -9.
structure.txt
Genepop infile
This is the input file for Genepop v4 analyses. Individual sample names have been replaced by x and y coordinates in meters. Microsatellite and mitochondrial data are included. Missing data are represented by 000.
Genepop.txt
MrBayesDir infile
Nexus file for modified version of MrBayes 3.1.2 with Dirichlet priors on branch lengths. Partitioning scheme and substitution models determined using PartitionFinder.
mbdir.nex
IBDWS infile for IBD regression
Matrixes of geographic distance (in meters; above diagonal) and genetic distance (below diagonal) for Isolation by Distance Web Service analysis. Genetic distances calculated as Fst/(1-Fst) in Genepop v4.2.
IBDWS euclidean.txt
IBDWS infile for IBR regression
Matrixes of resistance distance (above diagonal) and genetic distance (below diagonal) for Isolation by Distance Web Service analysis. Genetic distances calculated as Fst/(1-Fst) in Genepop v4.2.
IBDWS resistance.txt
Alignment phylip format
MtDNA alignment in phylip format for PartitionFinder analysis.
klauberi.phy
PartitionFinder infile
Input file for PartitionFinder analysis.
partition_finder.cfg
Infile for SGoF+ analysis
Input file for sequential goodness-of-fit test to correct for multiple comparisons implemented in SGoF+. The file consists of an integer indicating the total number of tests (102) and after it two columns with pairs of identifiers and p-values.
PvalSGoF.dat
Full data file
This is the complete data file.
data full.txt
All SDM points
This file contains latitude and longitude coordinates for the 91 unique localities used to construct species distribution models.
all point localities.csv
subset point localities
This file contains latitude and longitude coordinates for the 23 localities representing a subset of the total data used to construct species distribution models. California records were filtered to one locality record per 10 km2 grid cell, resulting in a total of N=23 records. We performed this analysis to investigate whether the greater number of records in California unduly influenced the model.