Supplemental Material
Evaluating connectivity between Natura 2000 sites within the montado agroforestry system: a case study using landscape genetics of the wood mouse (Apodemus sylvaticus)
Jacinta Mullins () · Fernando Ascensão · Luciana Simões · Leonardo Andrade · Margarida Santos-Reis · Carlos Fernandes
Centre for Environmental Biology, University of Lisbon, Campo Grande, 1749-016 Lisbon, Portugal
S1: Microsatellite PCR conditions
PCRs were conducted in 10 μL volumes containing 2 µL of genomic DNA extract, 0.8 mM dNTP (Bioline), 6.25 µg Bovine Serum Albumin (New England Biolabs), 2.5 mM MgCl2, 1X PCR buffer, 0.5 U SURF HotTaq DNA polymerase (StabVida) and 0.5 µM of a primer mix containing locus-specific forward and reverse primers and a fluorescently labelled M13 primer tail (Schuelke 2000).
Two different PCR protocols were used for genotyping, a two-step PCR and a touchdown PCR. The two-step protocol consisted of an initial denaturation step of 95 °C for 15 mins followed by 10 cycles of 94 °C for 30 s, 65 °C for 30 s and 72 °C for 45 s, then 30 cycles of 94 °C for 30 s, 60 °C (As11, As20, SFM2, TNF, SCFM2) or 55 °C (SCFM6), locus annealing temperature for 30 s and 72 °C for 45 s, with a final extension of 20 mins at 72 °C. The touchdown procedure also included the 15 min initial denaturation, followed by 94 °C for 30 s, 65 °C for 30 s (dropping 1 °C every cycle for 12 cycles), 72 °C for 45 s, finishing with another 25 cycles of 94 °C for 30 s, 53 °C (As34 and SCFM9) for 30 s, 72 °C for 45 s, and a final extension for 20 mins at 72°C. Amplifications were performed in a GeneAmp 2720 thermal cycler (Applied Biosystems) and loci were combined in post-PCR multiplexes (1 μL each) for fragment analysis. Fragments were either analysed in-house using an ABI310 Genetic Analyzer or sent to Macrogen (Seoul, Korea) in 96-well plates. PCR products from three samples in every plate sent to Macrogen were also analysed in-house to merge the results. GENEMAPPER 3.7 (Applied Biosystems) was used to score the fragments against a GS500 ROX-labelled size standard. Alleles were assigned using bins created in TANDEM (Matschiner and Salzburger 2009) from sized alleles of a subset of 205 individuals. As the observed heterozygosity was high the samples were amplified once for each locus. Ambiguous genotypes were repeated until the peak profile could be scored or the sample was discarded. All extractions and amplifications were carried out in a UV cabinet with dedicated equipment and negative controls were used to monitor for contamination.
S2: Genetic diversity of wood mice in the study area. N = number of individuals genotyped; A = number of alleles; Ar = allelic richness; PAr = private allelic richness; Ho = observed heterozygosity; He = expected heterozygosity; Fis = inbreeding coefficient. Values were averaged across loci within each area
N / 71 / 60 / 262
A / 22.67±6.40 / 21.56±5.41 / 26.00±9.12
Ara / 22.13±6.21 / 21.56±5.41 / 21.99±6.45
PAra / 1.31±0.95 / 1.16±0.97 / 0.88±0.82
Ho / 0.82±0.04 / 0.82±0.07 / 0.80±0.05
He / 0.93±0.02 / 0.93±0.02 / 0.93±0.02
Fis / 0.13±0.02 / 0.12±0.02 / 0.14±0.01
aAllelic richness and private allelic richness were based on a sample of 120 genes (Sudoeste, the smallest ‘population’)
S3: Exact test p-values for the Hardy-Weinberg Equilibrium test in GENEPOP. Loci that remained significant after Bonferroni correction are indicated in bold (9 tests, adjusted α = 0.00556)
As20 / 0.1124 / 0.0034 / 0.1097
SFM2 / 0.0002 / 0.0346 / 0.2996
As7 / 0.0001 / 0.0244 / 0.2459
As34 / 0.0000 / 0.0071 / 0.0074
As11 / 0.0000 / 0.0000 / 0.0014
TNF / 0.0114 / 0.7599 / 0.2868
SCFM6 / 0.0000 / 0.1768 / 0.0001
SCFM2 / 0.0000 / 0.0059 / 0.0775
SCFM9 / 0.0000 / 0.0033 / 0.0000
S4: Null allele frequencies per locus in GENEPOP. Loci with null allele frequencies greater than 0.1 are indicated in bold for each area
As20 / 0.0345 / 0.0635 / 0.0412
SFM2 / 0.0462 / 0.0509 / 0.0525
As7 / 0.0538 / 0.0572 / 0.0314
As34 / 0.0936 / 0.0692 / 0.1087
As11 / 0.1067 / 0.1088 / 0.0934
TNF(CA) / 0.0353 / 0.0222 / 0.0283
SCFM6 / 0.0930 / 0.0485 / 0.1076
SCFM2 / 0.0675 / 0.0774 / 0.0416
SCFM9 / 0.0848 / 0.1059 / 0.1139
S5: Putative parent-offspring or full-sibling pairs identified with ML-RELATE. Pairs were sampled within the same trapping site, with corresponding capture date for each pair. The thirteen individuals that were dropped from the dataset are highlighted in bold text
(g) / Capture Date / Individual 2 / Weight
(g) / Capture Date / Relationship
As027f / 21.0 / 06/01/2009 / As028m / 21.0 / 06/01/2009 / FS
As173f / 19.5 / 21/01/2010 / As175f / 21.5 / 21/01/2010 / FS
As188f / 31.0 / 20/02/2010 / As189m / 12.5 / 20/02/2010 / FS
As189m / 12.5 / 20/02/2010 / As197f / 8.5 / 21/02/2010 / FS, PO
As193f / NA / 22/02/2010 / As202f / 20.5 / 22/02/2010 / PO
As193f / NA / 21/02/2010 / As195m / 21.5 / 21/02/2010 / FS, PO
As195m / 21.5 / 21/02/2010 / As196m / 22.5 / 21/02/2010 / FS
As221f / 8.0 / 17/03/2010 / As224m / 17.0 / 19/03/2010 / FS
As500m / 19.0 / 06/05/2010 / As501f / 27.5 / 06/05/2010 / FS
As506f / 29.5 / 12/05/2010 / As507f / 19.0 / 13/05/2010 / FS, PO
As574f / 20.0 / 02/02/2011 / As578f / 34.0 / 02/02/2011 / FS,PO
As644f / 28.5 / 21/01/2010 / As645m / 36.0 / 21/01/2010 / FS
As070f / 34.0 / 19/11/2009 / As085u / 10.0 / 19/11/2009 / FS, PO
As280f / 34.5 / 20/04/2010 / As285m / 33.5 / 21/04/2010 / PO
As709f / 24.0 / 22/05/2011 / As710f / 22.0 / 22/05/2011 / FS
S6: Bayesian clustering analysis with structure. The clustering solution with the highest likelihood and lowest variation across replicates was K = 1. The bar-plot of individual membership to each cluster for K = 2 (below) and higher values (not shown) further supported the lack of clear population subdivision between Cabrela and Sudoeste
S7: Bayesian clustering analysis with GENELAND. Black dots represent individual samples. The colors represent genetic clusters assigned by GENELAND using the correlated allele frequencies model
S8: Spatial autocorrelation analysis. Correlogram plots of the average autocorrelation coefficient, r, as a function of distance for classes of 1 km (a) and 3 km (b) in width.
S9: Optimization of resistance surface cost values by exhaustive permutation of cost values assigned to land cover categories. Boxplots representing 625 Mantel r correlation coefficients for each of the land cover categories ‘High’, ‘Medium’, ‘Low’ and ‘Other’.
S10: Reciprocal causal modeling Mantel test results for the top 20 resistance surfaces (RS) of the land cover model. The table includes the test sequence number (of 625 hypotheses) and the costs assigned to the high (H), medium (M), low (L) and other (O) land cover categories as detailed in the manuscript. The top 20 resistance surfaces ranked by Mantel r for the causal modeling approach of Cushman et al. (2006) are also shown for comparison.
ID / High / Medium / Low / Other / ID / High / Medium / Low / Other
H392 / 1 / 16 / 1 / 8 / H459 / 4 / 16 / 2 / 4
H197 / 1 / 16 / 1 / 4 / H284 / 4 / 16 / 4 / 4
H470 / 2 / 16 / 1 / 8 / H297 / 4 / 16 / 2 / 2
H200 / 1 / 16 / 1 / 2 / H294 / 4 / 16 / 1 / 4
H374 / 1 / 16 / 1 / 16 / H329 / 4 / 16 / 4 / 2
H474 / 2 / 16 / 1 / 4 / H407 / 4 / 16 / 1 / 2
H199 / 1 / 16 / 1 / 1 / H437 / 4 / 16 / 2 / 1
H480 / 2 / 16 / 1 / 2 / H579 / 2 / 8 / 1 / 2
H477 / 2 / 16 / 1 / 1 / H331 / 4 / 16 / 4 / 1
H542 / 2 / 16 / 1 / 16 / H335 / 4 / 16 / 4 / 8
H005 / 1 / 16 / 2 / 8 / H409 / 4 / 16 / 1 / 1
H003 / 1 / 16 / 2 / 4 / H457 / 4 / 16 / 2 / 8
H433 / 2 / 16 / 2 / 1 / H474 / 2 / 16 / 1 / 4
H002 / 1 / 16 / 2 / 1 / H480 / 2 / 16 / 1 / 2
H001 / 1 / 16 / 2 / 2 / H578 / 2 / 8 / 1 / 1
H050 / 1 / 8 / 1 / 2 / H606 / 2 / 8 / 2 / 2
H434 / 2 / 16 / 2 / 4 / H108 / 8 / 16 / 4 / 4
H432 / 2 / 16 / 2 / 2 / H248 / 4 / 8 / 2 / 2
H259 / 1 / 8 / 1 / 1 / H286 / 4 / 16 / 1 / 8
H048 / 1 / 8 / 1 / 4 / H432 / 2 / 16 / 2 / 2
* Top 20 of 625 surfaces ranked by support (GD~Focal RS|Alternative RS - GD~Alternative RS|Focal RS). GD = genetic distance, RS = resistance surface.
** Ranked by partial Mantel r value (GD~RS|Euclidean)
S11: Representative map of cumulative current flow for the optimal landscape resistance surface parameterized with genetic data. Blue represents low current flow (high resistance) and red represents high current flow (low resistance). The polygons indicate the borders of the two Natura 2000 sites defining our study area.
References
Matschiner M, Salzburger W (2009) TANDEM: integrating automated allele binning into genetics and genomics workflows. Bioinformatics 25:1982–1983
Schuelke M (2000) An economic method for the fluorescent labeling of PCR fragments. Heredity 18:233–234
13
Landscape genetics of wood mice in S Portugal, Supplemental Material