Supplementary materials
Appendix 1 Microsatellites, their size and fluorescent dyes used in the single multiplex for genotyping European pine martens in Bresse. Ma1, Ma2, Ma8, Ma9, Ma10, Ma19, Gg7 were devellopped by Davis and Strobeck (1998) and Ggu454 by Walker et al. (2001)
Microsatellite locus / Size (bp) / Fluorescent dyeGg7 / 160-175 / Pet
Ggu454 / 120-128 / Vic
Ma01 / 195-208 / Fam
Ma02 / 160-179 / Fam
Ma08 / 95-119 / Fam
Ma09 / 135-146 / Pet
Ma10 / 149-173 / Vic
Ma19 / 186-209 / Ned
Davis CS and Strobeck C (1998), Isolation, variability, and cross-species amplification of polymorphic microsatellite loci in the family Mustelidae. Mol Ecol, 7:1776–1778
Walker CW, Vilà C, Landa A, Lindén M and Ellegren H (2001), Genetic variation and population structure in Scandinavian wolverine (Gulo gulo) populations. Mol Ecol, 10: 53–63
Appendix 2 R CODE to compute pairwise Euclidean “ground” distance
library(raster)
## Load raster of elevation
r <- raster("elevation.adf")
## Load individuals’ coordinates
coord <- read.table("coord.txt",sep="\t",dec=".",header=T)
mat <- matrix(nrow=dim(coord)[1],ncol=dim(coord)[1])
for (m in 1:dim(coord)[1]) {
for(j in 1:m) {
A <- cbind(coord[m,1],coord[m,2])
B <- cbind(coord[j,1],coord[j,2])
lin1 <- Line(rbind(A,B))
lin2 <- Lines(list(lin1), ID="a")
lin3 <- SpatialLines(list(lin2))
eleval <- unlist(extract(r3, lin3))
distc <- c()
for (i in 1:length(eleval)-1) {
distc[i] <- sqrt((20^2)+(eleval[i+1]-eleval[i])^2)
}
mat[m,j] <- sum(distc)
} }
Appendix 3 Screeplots of significant partial Mantel r ordered from highest r to lowest r for IBRLS and IBRelev models in BRE and ARG. Characteristics of the ten best IBRLS and IBRelev models in BRE and ARG.
Figure A1 Screeplot of the significant partial Mantel r values used to select the IBRLS scenarios in BRE, 43 scenarios were significant and the 2nd was selected based on its RS value as the LCDLC/BRE*.
Table A1 Mantel’s r, p-value of the Mantel’s r, RS coefficient and resistance values attributed to the different landscape elements of the ten first significant IBRLS scenarios in BRE (43 scenarios were significant and the 2nd, in bold, was selected based on its RS value as the LCDLC/BRE*).
Rank / Mantel’s r / p-value / RS / Highways / Roads / Ponds / Buildings / Hedgerows / Shrubby land / Forest and copses / Open areas1 / 0.0877 / 0.039 / 0.1546 / 100 / 1 / 100 / 1 / 25 / 1 / 25 / 100
2 / 0.0874 / 0.043 / 0.1565 / 100 / 1 / 100 / 1 / 25 / 1 / 25 / 50
3 / 0.0873 / 0.039 / 0.1466 / 100 / 1 / 100 / 1 / 50 / 1 / 50 / 100
4 / 0.0873 / 0.042 / 0.1454 / 100 / 1 / 100 / 1 / 100 / 1 / 50 / 100
5 / 0.0870 / 0.041 / 0.1546 / 100 / 1 / 100 / 1 / 100 / 1 / 25 / 50
6 / 0.0868 / 0.040 / 0.1544 / 100 / 1 / 100 / 1 / 50 / 1 / 25 / 50
7 / 0.0865 / 0.041 / 0.1435 / 100 / 1 / 100 / 1 / 100 / 25 / 50 / 100
8 / 0.0864 / 0.038 / 0.1541 / 100 / 1 / 100 / 1 / 25 / 25 / 25 / 50
9 / 0.0863 / 0.040 / 0.1444 / 100 / 1 / 100 / 1 / 50 / 25 / 50 / 100
10 / 0.0863 / 0.039 / 0.1430 / 100 / 1 / 100 / 1 / 100 / 50 / 50 / 100
Figure A2 Screeplot of the significant partial Mantel r values used to select the IBRLS scenarios in ARG, 468 scenarios were significant and the 5th was selected based on its RS value as the LCDLC/ARG*.
Table A2 Mantel’s r, p-value of the Mantel’s r, RS coefficient and resistance values attributed to the different landscape elements of the ten first significant IBRLS scenarios in ARG (468 scenarios were significant and the 5th, in bold, was selected based on its RS value as the LCDLC/ARG*).
Rank / r / pval / RS / Highways / Roads / Ponds / Buildings / Hedgerows / Shrubby land / Forest and copses / Open areas1 / 0.1149 / 0.045 / 0.1927 / 100 / 25 / 100 / 1 / 1 / 100 / 1 / 100
2 / 0.1149 / 0.041 / 0.1927 / 100 / 25 / 100 / 1 / 25 / 100 / 1 / 100
3 / 0.1149 / 0.041 / 0.1927 / 100 / 25 / 100 / 1 / 50 / 100 / 1 / 100
4 / 0.1149 / 0.041 / 0.1927 / 100 / 25 / 100 / 1 / 100 / 100 / 1 / 100
5 / 0.1145 / 0.037 / 0.1963 / 100 / 25 / 100 / 1 / 1 / 25 / 1 / 50
6 / 0.1145 / 0.037 / 0.1963 / 100 / 25 / 100 / 1 / 25 / 100 / 1 / 50
7 / 0.1145 / 0.043 / 0.1963 / 100 / 25 / 100 / 1 / 50 / 100 / 1 / 50
8 / 0.1145 / 0.042 / 0.1963 / 100 / 25 / 100 / 1 / 100 / 100 / 1 / 50
9 / 0.1143 / 0.046 / 0.1959 / 100 / 25 / 100 / 1 / 1 / 50 / 1 / 50
10 / 0.1143 / 0.045 / 0.1959 / 100 / 25 / 100 / 1 / 25 / 50 / 1 / 50
Figure A3 Screeplot of the significant partial Mantel r values used to select the IBRelev scenario in ARG, 58 scenarios were significant and the 2nd was selected based on its RS value as the LCDelev/ARG*.
Table A2 Mantel’s r, p-value of the Mantel’s r, RS coefficient and parameters for the inverse Gaussian function resistance function optimum (Optimumalti) and its standard deviation (SDalt) of the ten first significant IBRelev scenarios in ARG (58 scenarios were significant and the 2nd, in bold, was selected based on its RS value as the LCDelev/ARG*).
Rank / r / p-value / RS / Optimumalti / SDalti1 / 0.1541 / 0.006 / 0.2730 / 1800 / 250
2 / 0.1536 / 0.007 / 0.2797 / 1700 / 250
3 / 0.1493 / 0.007 / 0.2759 / 1900 / 250
4 / 0.1487 / 0.011 / 0.2572 / 1600 / 250
5 / 0.1464 / 0.010 / 0.2626 / 2000 / 500
6 / 0.1452 / 0.011 / 0.2537 / 1900 / 500
7 / 0.1447 / 0.011 / 0.2560 / 2100 / 750
8 / 0.1446 / 0.011 / 0.2642 / 2100 / 500
9 / 0.1441 / 0.012 / 0.2595 / 2200 / 750
10 / 0.1439 / 0.013 / 0.2539 / 2200 / 1000
Appendix 4 Correlation between elevation and landscape elements. For each elevation value (from 132 to 1301 meters a.s.l), using the R software, we extracted the number of pixels corresponding to each landscape element (e.g. pixels of forest or shrubs) present at this elevation class. We then computed relative percentage for each landscape element.
Figure A4 Correlation between the percentage of shrubs cover and elevation (in meter). The shaded area corresponds to elevation associated with the lowest cost of the averaged IBRelev scenarios and the vertical green line corresponds to the optimal elevation (1700m).
Figure A5 Correlation between the percentage of forested cover and elevation (in meter). The shaded area corresponds to elevation associated with the lowest cost of the averaged IBRelev scenarios and the vertical green line corresponds to the optimal elevation (1700m).
Figure A6 Correlation between the percentage of open area and elevation (in meter). The shaded area corresponds to elevation associated with the lowest cost of the averaged IBRelev scenarios and the vertical green line corresponds to the optimal elevation (1700m).