Spatial and Temporal Patterns of Diversification in Leaf- toed Geckos (Phyllodactylidae: Phyllodactylus) throughout the Mexican Dry Forest
by
Christopher Blair
A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Ecology and Evolutionary Biology University of Toronto
© Copyright by Christopher Blair 2012 Spatial and Temporal Patterns of Diversification in Leaf-toed Geckos (Phyllodactylidae: Phyllodactylus) throughout the Mexican Dry Forest
Christopher Blair
Doctor of Philosophy
Department of Ecology and Evolutionary Biology University of Toronto
2012 Abstract
Understanding the ecological and evolutionary processes responsible for shaping patterns of genetic variation in natural populations is a long-standing goal in molecular ecology.
Although an extensive number of recent studies focus on patterns and processes throughout tropical rain forest ecosystems, substantially less effort has been placed on tropical dry forests (TDFs); a habitat known to harbour a large percentage of Earth’s diversity. In this thesis I use leaf-toed geckos of the genus Phyllodactylus to understand both the historical and contemporary processes influencing diversification throughout Mexico’s TDFs. In Chapter 2
I isolate and characterize microsatellite markers for the gecko P. tuberculosus. Chapter 3 uses these loci to conduct a landscape genetic analysis of the species near Alamos, Sonora. I find that the inclusion of landscape variables explains more genetic variance versus
Euclidean distance alone. Chapter 4 examines the evolutionary history of the P. tuberculosus group throughout western Mexico. Results suggest that habitat and climate shifts during the
Miocene and Pleistocene were important divers of diversification. Chapter 5 uses microsatellite and mtDNA markers to compare historical and contemporary demographic
ii parameters in P. tuberculosus. I find evidence for low historical gene flow and high female philopatry, recent reductions in population sizes, and higher correlations between landscape and contemporary gene flow versus historical gene flow and mtDNA divergence. In Chapter
6 I examine the biogeographic and taxonomic consequences of the dynamic history of Baja
California. My phylogenetic results provide evidence for a trans-peninsular seaway in the
Isthmus of La Paz region and suggests that P. xanti nocticolus warrants species status. My results also suggest the possibility of a seaway near the Loreto area that needs to be evaluated further. By combining multiple molecular marker-types and analytical methods, this thesis adds to our understanding of diversification processes throughout the threatened Mexican
TDF.
iii Acknowledgments
Taking on a graduate degree, especially a Ph.D. in a foreign country with limiting funding opportunities is no easy task. Throughout our graduate programs we must learn to troubleshoot potential research issues and partition the limited time we have in an effective manner. To this end, I feel that graduate school adds a second dimension of education beyond a researcher’s focus on a specific discipline. With that said, I would first like to thank my supervisor, Prof. Bob Murphy, for all his advice and guidance over the years. Bob accepted me into his lab knowing that I did not have my own research funding, in part, because I was an international student. However, he welcomed me into his lab with open arms and gave me the freedom to pursue my own research interests and was always available for advice if needed. I also thank the remainder of my thesis committee, Prof. Allan Baker and Prof. Marie-Josée-Fortin for their insightful discussions regarding methods of statistical analysis.
Special thanks also go to my fellow lab-mates Christina Davy, Andre Ngo, and Pedro Bernardo for all the help and advice they have given me. Graduate school, like life, is a continuous learning process and these individuals taught me much more than I ever thought possible. I would also like to thank graduate students from other labs at the Royal Ontario Museum (ROM) and the University of Toronto for their thoughtful discussions about research methods and philosophy. In particular, I thank Ida Conflitti, Patrick Schaefer, Ruben Cordero, Julio Rivera, Damon Dewsbury, Santiago Sanchez, Yessica Rico, Pasan Samarasin, Ilona Naujokaitis-Lewis, Oliver Haddrath, and Erika Tavares for all their help over the years.
As those of us in the field know, molecular genetic laboratory work can be a stressful, time-consuming process that results in many days without the generation of data. However, throughout my time at the ROM I was fortunate to have the help from a variety of talented individuals including Amy Lathrop, Kristen Choffe, and Oliver Haddrath. These knowledgeable researchers were always close-by to lend a helping hand when needed. I also thank Christopher Law and Pedro Bernardo for all their help in the laboratory and Christopher Wilson and Christopher Kyle for their assistance with 454 sequencing.
iv During my time as a Ph.D. student at the University of Toronto I had the opportunity to travel for my research. Because of Bob, I was fortunate to have the chance to conduct research on geckos throughout western Mexico. Conducting field work throughout the country over the course of three years enabled me to form new collaborations that I hope will last a lifetime. Working in Mexico also allowed me to grow as an individual and understand the lives and cultures of people from different backgrounds. I first thank my collaborator and friend Fausto Méndez de la Cruz for all of his help with logistics in Mexico including securing permits. Without Fausto, this thesis would not have been possible. Special thanks also go to his students Victor Jiménez-Arcos and Anibal Diaz de la Vega for all their help. Not only are these guys excellent herpers, but they are very knowledgeable about Mexico’s diverse ecosystems and about where and when to sample. I also thank Stephanie Meyer and Martin Gabriel Figueroa Martinez for their help with all of the Alamos work. Finally, I wish to thank Marie-Josée-Fortin and Helene Wagner for the opportunity to attend the landscape genetics conference in Santa Barbara. This was a tremendous opportunity and I hope to continue to collaborate on projects with the people I met.
As research is not possible without adequate funding, I must thank the appropriate funding sources for this research. The majority of this research was funded by a Natural Sciences and Engineering Research Council of Canada (NSERC) grant (A3148) issued to Bob. Funding for lab work was also provided by the Schad Foundation at the ROM. Field work in Mexico was partly funded through a Theodore Roosevelt Memorial Grant issued by the American Museum of Natural History.
Finally, and most importantly, I would like to thank my parents, Burt and Debbie Blair, and my brother, Michael Blair, for all their support over the years. Words cannot express how supportive they have been to me during my time as a struggling student. Throughout my time as a college student (over 12 years) my parents gave me the freedom to pursue my interests in life even at times when it was difficult financially. For this I will always be in your debt. Pursuing advanced degrees also takes us away from the people we care about, and although this is difficult on families, my parents understood that life is short and we must follow our hearts. You raised me to be a polite, caring, hard-working man and I always wanted to make you proud of me. I hope this thesis and degree adds to that endeavor.
v Table of Contents
Abstract……………………………………………………………………………………….ii
Acknowledgments…………………………………………………………………………....iv
Table of Contents…………………………………………………………………………….vi
List of Tables……………………………………………………………………………….xiv
List of Figures………………………………………………………………………..……xviii
List of Appendices……………………………………………………………………..….xxiii
CHAPTER 1—INTRODUCTION...... 1
1. Background………………………………………………………………………..1
2. Tropical Dry Forest……………………………………………………………….5
3. Leaf-toed Geckos………………………………………………………………….6
4. Objectives of Thesis…………………………………………………………….....7
References...... 9
Figure Legends...... 15
Figures...... 16
CHAPTER 2—USING NEXT-GENERATION DNA SEQUENCING FOR RAPID MICROSATELLITE
DISCOVERY IN MEXICAN LEAF-TOED GECKOS (PHYLLODACTYLUS
TUBERCULOSUS)...... 17
vi Abstract...... 17
Note...... 18
Acknowledgments...... 20
References...... 22
Tables...... 24
CHAPTER 3—LANDSCAPE GENETICS OF LEAF-TOED GECKOS IN THE SUB-TROPICAL DRY
FORESTS OF NORTHERN MEXICO...... 25
Abstract...... 25
1. Introduction...... 26
2. Materials and Methods...... 29
2.1 Sampling...... 29
2.2 DNA extraction and genotyping...... 30
2.3 Genetic diversity...... 30
2.4 Genetic differentiation...... 31
2.5 Population structure...... 32
2.6 Landscape genetic analysis...... 33
3. Results...... 40
3.1 Genetic diversity...... 40
3.2 Genetic differentiation...... 41
3.3 Population structure...... 41
vii 3.4 Landscape genetics...... 43
4. Discussion...... 44
4.1 Genetic diversity and population structure...... 45
4.2 Landscape genetics...... 47
4.3 Statistical methods in landscape genetics...... 51
4.4 Conservation implications for Mexican TDF...... 53
Acknowledgments...... 54
References...... 55
Tables...... 66
Figure Legends...... 71
Figures...... 72
Supplementary Material...... 76
CHAPTER 4—GENES, GEOGRAPHY, CLIMATE AND CRYPTIC DIVERSITY IN LEAF-TOED
GECKOS (PHYLLODACTYLIDAE: PHYLLODACTYLUS) ILLUMINATE
EVOLUTIONARY PROCESSES THROUGHOUT THE MEXICAN DRY FOREST...... 79
Abstract...... 79
1. Introduction...... 80
2. Materials and Methods...... 84
2.1 Study area and sampling...... 84
2.2 Laboratory methods...... 84
viii 2.3 Sequence alignment and haplotype estimation...... 85
2.4 Recombination...... 85
2.5 Gene-tree reconstruction...... 86
2.6 Species-tree reconstruction and divergence times...... 86
2.7 Diversification rates...... 88
2.8 Biogeography...... 89
2.9 Population genetic and historical demographic analysis...... 90
2.10 Migration rates and effective population sizes...... 91
2.11 Isolation-by-distance...... 92
3. Results...... 93
3.1 Sequence characteristics...... 93
3.2 Gene-tree analysis...... 94
3.3 Species-tree analysis and divergence times...... 94
3.4 Diversification rates...... 95
3.5 Biogeography...... 96
3.6 Population genetics and demographics...... 97
3.7 Migration rates and effective population sizes...... 97
3.8 Isolation-by-distance...... 98
4. Discussion...... 98
4.1 Phylogenetic analysis...... 99
ix 4.2 Divergence times and diversification rates...... 101
4.3 Biogeography...... 104
4.4 Population genetics and demographics...... 107
4.5 Migration...... 108
4.6 IBD and drainage systems...... 109
4.7 Gecko diversity and taxonomic implications...... 109
4.8 Conclusions...... 111
Acknowledgments...... 111
References...... 113
Tables...... 126
Figure Legends...... 129
Figures...... 131
Supplementary Material...... 137
CHAPTER 5—CONSIDERATION OF HISTORICAL DEMOGRAPHIC PROCESSES IN LANDSCAPE
GENETIC STUDIES OF GENE FLOW: A CASE STUDY WITH MEXICAN LEAF-TOED
GECKOS...... 151
Abstract...... 151
1. Introduction...... 152
2. Materials and Methods...... 156
2.1 Study site and tissue sampling...... 156
x 2.2 DNA extraction, amplification, and sequencing...... 157
2.3 Matrilineal genealogy...... 158
2.4 Population genetic analysis of mtDNA...... 160
2.5 Historical and contemporary migration rates...... 161
2.6 Historical and contemporary effective population sizes...... 163
2.7 Landscape genetics...... 164
3. Results...... 165
3.1 Genetic diversity and model selection...... 165
3.2 Matrilineal genealogy...... 166
3.3 Population genetic analysis of mtDNA...... 167
3.4 Historical and contemporary migration rates...... 168
3.5 Historical and contemporary effective population sizes...... 169
3.6 Landscape genetics...... 170
4. Discussion...... 171
4.1 Sequence divergence and mitochondrial genealogy...... 171
4.2 Historical and contemporary migration rates and effective
population sizes...... 173
4.3 Landscape genetics...... 175
4.4 Implications for Mexican biodiversity...... 178
4.5 Conclusions...... 179
xi Acknowledgments...... 181
References...... 182
Tables...... 196
Figure Legends...... 202
Figures...... 204
Supplementary Material...... 209
CHAPTER 6—MOLECULAR PHYLOGENETICS AND TAXONOMY OF LEAF-TOED GECKOS
(PHYLLODACTYLIDAE: PHYLLODACTYLUS) INHABITING THE PENINSULA
OF BAJA CALIFORNIA...... 218
Abstract...... 218
1. Introduction...... 219
2. Materials and Methods...... 220
2.1 Tissue collection...... 220
2.2 DNA extraction, amplification, and sequencing...... 221
2.3 Sequence alignment and phylogenetic analysis...... 223
3. Results...... 225
3.1 Genetic diversity...... 225
3.2 Phylogenetic analysis of concatenated data set...... 225
3.3 Phylogenetic analysis of nDNA...... 226
3.4 Phylogenetic analysis of mtDNA...... 226 xii 4. Discussion...... 227
4.1 Taxonomy of Phyllodactylus...... 227
4.2 Biogeography...... 229
Acknowledgments...... 233
References...... 234
Tables...... 240
Figure Legends...... 243
Figures...... 245
Supplementary Material...... 250
CHAPTER 7—CONCLUSIONS AND FUTURE DIRECTIONS...... 252
1. General Background...... 252
2. Mexican Tropical Dry Forest...... 252
3. Leaf-toed Geckos...... 254
4. General Results of Thesis...... 255
5. Future Directions...... 258
References...... 261
xiii List of Tables
CHAPTER 2
Table 1. Characterization of twelve microsatellite loci for the leaf-toed gecko Phyllodactylus
tuberculosus (n = 30) based on molecular cloning and 454 shotgun sequencing.
CHAPTER 3
Table 1. Characterization of genetic diversity of Phyllodactylus tuberculosus at each
sampling site included in this study based on data from 10 microsatellite loci.
Table 2. Pairwise genetic differentiation between populations estimated from 10
microsatellite loci. Values above diagonal represent Dest and values below diagonal
FST. Bold values of FST indicate significance (P < 0.05) after Bonferroni correction.
Table 3. Multiple regression on distance matrices (MRM) results showing the relationship
between pairwise genetic distance (linearized Dest) and least-cost path distances
incorporating landscape heterogeneity. Four different MRM analyses were
performed (models B–D) using a stepwise approach with a p-to-enter value <0.05.
Models in italics are those selected with only significant regression coefficients.
Optimal cost values used to parameterize resistance surfaces prior to calculating
each least-cost path were selected based on Mantel r correlation coefficients.
VIF=Variance Inflation Factor with values above 10 in boldface.
Table 4 Multiple regression on distance matrices (MRM) results showing the relationship
between pairwise genetic distance (linearized Dest) and resistance distances
xiv incorporating landscape heterogeneity. Optimal cost values used to parameterize
resistance surfaces prior to calculating resistance distances were selected based on
Mantel r correlation coefficients. For clarity, only models with significant support
are shown. VIF=Variance Inflation Factor.
CHAPTER 4
Table 1. Primer information for loci used in this study.
Table 2. Mitochondrial DNA sequence divergence between major lineages of leaf-toed
geckos (Phyllodactylus) recovered in the phylogenetic analysis. Values below
diagonal represent uncorrected p-distances and values above diagonal Tamura-Nei
corrected distances.
Table 3. Diversity statistics and tests for demographic changes for each major mitochondrial
lineage of leaf-toed geckos (Phyllodactylus) recovered in the phylogenetic analysis.
CHAPTER 5
Table 1. Nucleotide diversity statistics and their variances for all populations sampled for
mtDNA.
Table 2. DNA divergence between populations. Numbers above diagonal are the average
number of nucleotide differences and numbers below correspond to the average
number of substitutions per site (Dxy).
xv Table 3. Estimates of contemporary migration rates between populations. Values in
parentheses represent 95% confidence intervals. Bold values on diagonal represent
the proportion of individuals derived from the source population each generation.
Table 4. Estimates of contemporary and historical effective population sizes (Ne) and 95%
confidence intervals. Results in first three columns represent Ne estimates based on
the linkage disequilibrium method employed in the software LDNe. Results are
reported based on lowest allele frequency values used (0.05, 0.02, 0.01). Estimates
in last column represent historical Ne estimates using the coalescent model in
-4 MIGRATE and calculated assuming a mutation rate of 5 x 10 (Garza & Williamson
2001).
Table 5. Results of tests for population bottlenecks performed using the software
BOTTLENECK. Numbers represent P-values based on the corresponding mutation
model. Allelic distribution represents the test for a shift in the relative abundance of
alleles at different frequencies.
Table 6. Regression on distance matrix results showing the relationships between landscape
resistance and historical migration, contemporary migration, and mtDNA sequence
divergence (Dxy). Model M = MIGRATE; Model B = BAYESASS; Model MtDNA =
mtDNA divergence. Models in italics significant at an alpha = 0.05.
CHAPTER 6
Table 1. Taxon name, tissue number, sample ID, geographic location, GPS coordinates, and
Genbank accession numbers for all species of Phyllodactylus sequenced. ROM =
xvi Royal Ontario Museum; RWM = Collections of Robert W. Murphy housed at the
ROM; N/A = No data available.
Table 2. Percent sequence divergence (uncorrected p-distances from mtDNA) for Baja
Californian and mainland Mexican Phyllodactylus compared to Gekko gecko. N/A
= undefined distance between sequences.
xvii List of Figures
CHAPTER 3
Fig. 1. Sampling sites for all individuals and populations of Phyllodactylus tuberculosus
included in this study. Darker shades of gray represent tropical dry forest. Dark lines
represent streams and arroyos throughout the study area.
Fig. 2. Relationship between elevation and genetic diversity (allelic richness) for all
populations of Phyllodactylus tuberculosus included in this study.
Fig. 3. STRUCTURE results excluding prior information on sampling locality for
Phyllodactylus tuberculosus included in this study. A) Results for K=3 selected using
the ΔK method. B) Results for K=5 selected using Pr(X|K).
Fig. 4. Results from CIRCUITSCAPE depicting cumulative currents between populations of
Phyllodactylus tuberculosus based on the combined effect of slope and riparian
connectivity. ‘High’ in figure legend represents high levels of current whereas ‘Low’
represents low levels of current.
CHAPTER 4
Fig. 1. Map of collecting localities for all species and subspecies used in this study. Shaded
landscape represents the present distribution of Mexican dry forest. Solid lines
represent previously documented biogeographic barriers.
Fig. 2. Genealogical relationships among representatives of the P. tuberculosus species
group based on maximum likelihood analysis of mtDNA. Values adjacent to nodes
xviii represent bootstrap proportions estimated from 10000 replicates using the fast
bootstrapping algorithm in RAxML.
Fig. 3. Relaxed molecular clock analysis for the P. tuberculosus group based on mtDNA
sequences. Horizontal lines at nodes represent the 95% highest posterior density for
inferred nodal ages. Vertical line represents the hypothesized origin of the Mexican
dry forest (Becerra 2005). Horizontal scale is in million of years ago. Branch colors
represent rate of evolution (blue=slow; red=fast). Lineage Through Time plot
illustrates the accumulation of evolutionary lineages as a function of time.
Fig. 4. Species-tree analysis of the combined mitochondrial and nuclear data. Horizontal
lines at nodes represent the 95% highest posterior density for inferred nodal ages.
Horizontal scale is in million of years ago. Lineage Through Time plot illustrates the
accumulation of evolutionary lineages as a function of time.
Fig. 5. Historical biogeography of the P. tuberculosus species group based on the dated
mtDNA gene-tree. Colors represent the Mexican state where the animal was
captured. Letters at nodes represent ancestral areas with the highest likelihood from
Bayesian reconstructions. Green circles surrounding nodes represent postulated
vicariant events. Dotted lines represent hypothesized biogeographic barriers.
I=Sierra Madre Occidental; II=Mexican Volcanic Belt; III=Isthmus of Tehuantepec.
Fig. 6. Extended Bayesian skyline plots showing demographic trends through time for each
major mtDNA lineage recovered in the phylogenetic analysis. Dashed line represents
the median population size, while solid lines denote 95% confidence intervals. Scale
xix on the x-axis represents time in millions of years before present, while the y-axis
represents log population size (Ne) x generation time (τ).
CHAPTER 5
Fig. 1. Map of the Alamos region of Sonora showing sampling sites for leaf-toed geckos,
Phyllodactylus, included in the study. Site names in italics represent sites where
mtDNA sequences were obtained.
Fig. 2. A) Maximum parsimony (MP) strict consensus tree of eight most parsimonious trees
based on the concatenated mtDNA data for leaf-toed geckos, Phyllodactylus.
Numbers above nodes represent 10,000 nonparametric bootstrap proportions.
Numbers below nodes represent Bayesian posterior probabilities sampled from the
posterior distribution of trees. B) Bayesian majority rule phylogram resulting from
the optimal partitioning strategy of gene + codon position.
Fig. 3. Geographic distribution of major mitochondrial lineages for leaf-toed geckos,
Phyllodactylus, from throughout the Alamos region. Areas of sympatry are
represented by localities encircled by more than one color.
Fig. 4. Scatter plot with Reduced Major Axis (RMA) regression line illustrating isolation by
distance for leaf-toed geckos, Phyllodactylus, resulting from a Mantel test of 30,000
randomizations. PhiST/(1-PhiST) values used to calculate test statistic were based on
the uncorrected p-distances between haplotypes. Geographic distances shown in
kilometers.
xx Fig. 5. A) Frequency histogram of historical migration rates for leaf-toed geckos,
Phyllodactylus, estimated from MIGRATE. B) Frequency histogram of contemporary
migration rates estimated from BAYESASS.
CHAPTER 6
Fig. 1. Map of Baja California, the Gulf of California, southwestern USA, and northwestern
Mexico illustrating the geographic location of samples of Phyllodactylus included in
this study. Sample numbers correspond to those presented in Table 1. The break in
geographic ranges in the Isthmus of La Paz region is due to unsuitable natural habitat
(lack of rocky outcrops). Gray area represents the geographic range of P. nocticolus,
black area represents the geographic range of P. xanti, striped area represents the
range of P. unctus.
Fig. 2. Phylogenetic relationships of Mexican Phyllodactylus based on a branch-and-bound
maximum parsimony (MP) analysis of the concatenated dataset. Numbers above
nodes represent MP bootstrap proportions (BSP) resulting from 10,000
pseudoreplicates.
Fig. 3. Phylogenetic relationships of Mexican Phyllodactylus based on Bayesian inference
(BI) of the concatenated dataset. Numbers above and adjacent to nodes represent
Bayesian posterior probabilities (BPP; * = 1.0) sampled from the posterior
distribution of trees. Gekko gecko was removed from the tree due to its long branch
length.
xxi
Fig. 4. Phylogenetic relationships of Mexican Phyllodactylus based on maximum parsimony
(MP; A) and Bayesian Inference (BI; B) of nuclear DNA (nDNA). Numbers above nodes
on MP tree represent nonparametric bootstrap proportions (BSP). Numbers above nodes
on the Bayesian topology represent Bayesian posterior probabilities (BPP; * = 1.0). Refer
to text for a description of loci and total base pairs (bp) sequenced.
Fig. 5. Phylogenetic relationships of Mexican Phyllodactylus based on maximum parsimony
(MP; A) and Bayesian Inference (BI; B) of mitochondrial DNA (mtDNA). Numbers
above nodes on MP tree represent nonparametric bootstrap proportions (BSP). Numbers
above nodes on the Bayesian topology represent Bayesian posterior probabilities (BPP;
* = 1.0). Gekko gecko was removed from the BI tree due to its long branch length. Refer
to text for a description of loci and total base pairs (bp) sequenced.
xxii
List of Appendices
CHAPTER 3
Fig. S1. A) STRUCTURE results illustrating changes in Pr(X|K) under the aspatial model. B)
STRUCTURE results based on the second order rate of change (ΔK method) under the
aspatial model. For each K, 10 independent simulations were performed.
Fig. S2. A) STRUCTURE results illustrating changes in Pr(X|K) under the spatial model. B)
STRUCTURE results based on the second order rate of change (ΔK method) under the
spatial model. For each K, 10 independent simulations were performed.
CHAPTER 4
Appendix I. Locality information and GenBank accession numbers for all species and
populations used in this study.
Supplementary Table 1. MIGRATE-N estimates of mutation-scaled effective population sizes
and migration rates for Phyllodactylus tuberculosus and P. muralis along with
confidence intervals. Locus 'All' signifies that all three loci were used to calculate
values. 4Nm = number of migrants per generation (ӨM).
Supplementary Table 2. MIGRATE-N estimates of mutation-scaled effective population sizes
and migration rates for Phyllodactylus lanei along with confidence intervals. Locus 'All'
signifies that all three loci were used to calculate values. 4Nm = number of migrants per
generation (ӨM).
xxiii
Fig. S1. Relationships among representatives of the P. tuberculosus species group based on
maximum likelihood analysis of phased alpha-enolase haplotypes. For ease of
visualization only unique haplotypes are presented. Values adjacent to nodes represent
bootstrap proportions estimated from 10000 replicates using the fast bootstrapping
algorithm in RAxML. I=Sierra Madre Occidental; II=Mexican Volcanic Belt; III=
Isthmus of Tehuantepec.
Fig. S2. Relationships among representatives of the P. tuberculosus species group based on
maximum likelihood analysis of phased lamin-A haplotypes. For ease of visualization
only unique haplotypes are presented. Values adjacent to nodes represent bootstrap
proportions estimated from 10000 replicates using the fast bootstrapping algorithm in
RAxML. I=Sierra Madre Occidental; II=Mexican Volcanic Belt; III= Isthmus of
Tehuantepec.
Fig. S3. Relationships among representatives of the P. tuberculosus species group based on
maximum likelihood analysis of the concatenated data. For concatenation with mtDNA
a single phased nuclear haplotype was selected from both intron alignments for each
individual. Values adjacent to nodes represent bootstrap proportions estimated from
10000 replicates using the fast bootstrapping algorithm in RAxML.
CHAPTER 5
Appendix 1. Population and outgroup sampling with corresponding GenBank accession
numbers for all Phyllodactylus sequenced for mtDNA. Population ID for each
population follows that presented in all population genetic analyses. N/A = Not
applicable.
xxiv
Supplementary Table 1. Gene, number of sequenced bases, primer name and sequence, and
reference for all mtDNA loci used.
Supplementary Table 2. Total and number of potentially phylogenetically informative
characters for each data partition along with the chosen model of nucleotide substitution
based on AIC.
Supplementary Table 3. Alternate partitioning strategies tested using Bayes Factors.
Supplementary Table 4. MIGRATE estimates of historical asymmetric migration rates between
populations. Locus 'All' signifies that all 10 microsatellite loci were used for
calculations. ϴ = 4Neµ, M = m/µ, Nm = number of migrants per generation (ϴM/4). 1 =
Road to Navojoa, 2 = Alamos, 3 = Arroyo Tabelo, 4 = Aduana, 5 = Rio Cuchujaqui, 6 =
La Sierrita, 7 = Mocuzari, 8 = El Quintero, 9 = Choquincahui, 10 = San Antonio.
Supplementary Table 5. Multiple regression on distance matrices (MRM) results showing the
relationship between contemporary migration rates (estimates using BAYESASS) and
resistance distances incorporating landscape heterogeneity. Optimal cost values used to
parameterize resistance surfaces prior to calculating resistance distances were selected
based on Mantel r correlation coefficients. VIF = Variance Inflation Factor.
CHAPTER 6
Appendix I. Mitochondrial and nuclear loci sequenced for Phyllodactylus and their
corresponding primer sequences. Different combinations of primers were used to obtain
full reads of regions.
xxv
CHAPTER 1 Introduction
1. Background
Evolutionary and molecular ecological research often strives to understand spatial and
temporal patterns of biological diversification. Elucidating how and when species and
populations evolved allows us to draw robust conclusions regarding the ecological and
historical processes governing diversification. Recent advances in both DNA sequencing
technologies and analytical methods are providing empiricists with a toolbox of new
methods to test hypotheses of general interest to molecular ecology and evolution (Edwards
2008; Brito & Edwards 2009; Heled & Drummond 2008,2010; Storfer et al. 2010; Blair &
Murphy 2011). However, both careful selection of molecular markers and analytical
methods are required to identify accurately pattern-process relationships (Wang 2010,
2011).
The field of phylogeography seeks to identify the historical processes that shape the
spatial distribution of genetic lineages. Early animal studies relied on mitochondrial DNA
(mtDNA) sequences to test historical hypotheses including the effects of both major geologic
changes and climatic shifts on biological diversification (Hewitt 2000, 2004). Now, with the continual reduction in costs of DNA sequencing, these studies are beginning to incorporate variable nuclear DNA (nDNA) loci to provide a multilocus perspective regarding historical
processes shaping the evolutionary history of species and populations (Zhang & Hewitt 2003;
Brito & Edwards 2009). Analysis of multiple unlinked genetic loci is often beneficial as sole
1 CHAPTER 1—INTRODUCTION 2
reliance on mtDNA for evolutionary inference is fraught with potential issues including its maternal inheritance, which makes it generally unsuitable as a tool to measure gene flow
(Edwards & Bensch 2009). Further, phylogenetic analysis of a single gene provides but a single glimpse into history. Indeed, both theory and empirical data show that gene divergence will predate population and species divergence, producing a biased perspective regarding historical relationships and tempo of divergence (Edwards & Beerli 2000; McCormack et al.
2011). Because of stochasticity in the coalescent process (Kingman 1982), analysis of unlinked loci will often result in different histories and conclusions regarding the influence of historical processes on evolutionary patterns (Moore 1995; Maddison 1997; Degnan &
Rosenberg 2006, 2009). This is particularly the case in recently diverged groups separated by relatively short branch-lengths, where deep gene-coalescence is common (Fig. 1). The debate regarding the utility of mtDNA versus nDNA-based phylogeography and phylogenetics continues (Zink & Barrowclough 2008; Edwards & Bensch 2009) and has spawned a new era in molecular evolution and systematics that focuses efforts away from concatenation and gene-tree approaches (Edwards 2008; Knowles 2009). These so-called “species-tree” methods use multispecies coalescence to estimate phylogenetic relationships by explicitly modeling the stochastic nature of the coalescent process (Liu et al. 2008; Kubatko et al.
2009; Heled & Drummond 2010). Unlike concatenation, these species-tree methods statistically account for gene-tree/species-tree discordance in phylogenetic studies, most of which attribute the discordance to the incomplete sorting of ancestral alleles (e.g. Liu 2008).
New multilocus Bayesian methods are also being developed to reconstruct the demographic history of populations (Heled & Drummond 2008) and estimate historical migration rates and effective population sizes (Beerli & Felsenstein 2001; Beerli 2006).
CHAPTER 1—INTRODUCTION 3
Although the analysis of nDNA loci provides a more comprehensive perspective regarding patterns and processes of evolution, care must be taken when working with these data (Hare 2001; Zhang & Hewitt 2003). For example, unlike mtDNA that is inherited directly through the maternal line in most animals, nDNA is bi-parentally inherited and sequences can show signs of genetic recombination that can affect downstream phylogenetic and demographic inference (Arenas & Posada 2010; Martin et al. 2011). Determining haplotypic phase is also an issue when working with nDNA sequences that is traditionally overlooked in higher-level phylogenetic studies. Although molecular cloning procedures can be used to determine allelic phase, Bayesian methods have been developed to phase haplotypes in a statistical framework (Stephens et al. 2001). Subsequent analysis with phased nuclear data is still in its infancy and new models continue to be developed that incorporate genetic information from both alleles at a locus (Heled & Drummond 2008, 2010).
Regardless, recent evidence suggests that the arbitrary selection of one allele from each gene and concatenating it with mtDNA can lead to erroneous conclusions (Weisrock et al. 2012).
Thus, although phylogenetic and phylogeographic studies continue to adopt multilocus approaches, testing the performance of new and existing analytical methods is required if we are to select the appropriate methods to test our hypotheses.
Complementing the inferences made possible by molecular phylogenetics and phylogeography, the nascent field of landscape genetics seeks to understand the contemporary landscape and ecological factors influencing microevolutionary processes
(Manel et al. 2003; Holderegger & Wagner 2006, 2008). A key distinction between landscape genetics and phylogeography is the former’s focus on understanding contemporary ecological and evolutionary processes whereas the latter focuses on historical processes
CHAPTER 1—INTRODUCTION 4
(Manel et al. 2003; Storfer et al. 2010; Wang 2010). The focus of landscape genetics on
contemporary versus historical processes usually requires a certain class of molecular marker with enough variability to capture processes operating over the recent timescales of interest
(e.g. microsatellites). Thus, recent discussions have highlighted the importance of careful molecular marker selection in both phylogeography and landscape genetics (Storfer et al.
2010; Wang 2010, 2011). For example, sequence-based markers such as mtDNA and nDNA
evolve too slowly to capture ecological processes operating over recent timescales (see
Vandergast et al. 2007 for counter-argument). Conversely, microsatellites evolve too rapidly
to be useful for inferring historical processes (Zink 2010). Common questions addressed by
adopting a landscape genetics approach include the influence of environmental and landscape
variables on rates and patterns of gene flow, the influence of environmental and landscape
variables on the spatial distribution of adaptive genetic variation, and the influence of
environmental and landscape variables on certain ecological processes (Spear & Storfer
2008; Manel et al. 2010; Storfer et al. 2010; Balkenhol & Landguth 2011). Landscape
genetics can also be used for effective corridor design to maintain adequate rates of
functional connectivity between populations (Epps et al. 2007; Braunisch et al. 2010). The
enormous interest in landscape genetics comes from a combination of its interdisciplinary
nature—combining population genetics, landscape ecology, and spatial statistics—along with
the continual ease in which molecular genetic data can be gathered. This interest has
spawned numerous recent simulation studies that test the relative power of different
analytical methods for inferring landscape-genetic relationships (e.g. Balkenhol et al. 2009;
Landguth et al. 2010; Blair et al. 2012). Additional simulation and empirical studies are vital
CHAPTER 1—INTRODUCTION 5 if we are to gain an in-depth understanding of the processes governing the spatial distribution of genetic variation in natural populations.
2. Tropical Dry Forest
The tropical dry or deciduous forests (TDF) of western Mexico harbor an enormous diversity of species (Ceballos & García 1995; Gillespie et al. 2000; García 2006). In Mexico, these forests form a nearly continuous belt throughout much of the Pacific coast from southern
Sonora to Chiapas with isolated stands scattered throughout the south and east as well as in
Baja California Sur (Becerra 2005). Unlike tropical rainforests, the TDF is characterized by a distinct seasonality with up to 8 months of dry, arid-like conditions separated by 4 months of deluge (Murphy & Lugo 1986). These forests typically grow at low to moderate elevations where they gradually transition to pine-oak and pine forests (Robichaux & Yetman 2000).
Molecular phylogenetic evidence suggests that these forests originated approximately 30 million years ago in response to major geologic changes associated with the Sierra Madre
Occidental and the Mexican Volcanic Belt (Becerra 2005). Although the herpetofaunal diversity in these forests is substantial, effective conservation strategies throughout this threatened ecosystem have been lacking (García 2006) and we still know little about the historical processes important in shaping the evolutionary and biogeographic histories of organisms inhabiting this ecosystem (Becerra & Venable 2008). The few phylogenetic and phylogeographic studies focusing on genetic patterns of species in this region have found evidence for several cryptic mtDNA lineages that correspond to several traditional biogeographic barriers (e.g. Devitt 2006; Zarza et al. 2008). However, additional data from
CHAPTER 1—INTRODUCTION 6
diverse species are required to understand better the potential suite of historical processes
influencing cladogenesis throughout the TDF.
To date, no landscape genetic study has focused explicitly on TDF habitat. Indeed, at
present there is a general lack of landscape genetic studies focused on tropical ecosystems in
general (Storfer et al. 2010). This is problematic, as the majority of Earth’s species reside in these environments (Myers et al. 2000). Further, dry forests, particularly in Mexico, have been experiencing elevated rates of recent fragmentation due to continual anthropogenic influence (Trejo & Dirzo 2000). Thus, it remains unknown how species are responding to fragmentation. The adoption of a landscape genetic approach can help us understand the importance of landscape and environmental variables on functional connectivity and how populations are responding to recent habitat fragmentation. Thus, by combining multilocus phylogenetic analyses along with a landscape genetics we can begin to understand both the evolutionary history of species inhabiting TDF and if and how they are responding to recent changes in the landscape. The combination of molecular markers and analytical methods suitable for testing hypotheses operating over different temporal scales serves as a powerful approach to understand the evolutionary ecology of TDF dependent species.
3. Leaf-toed Geckos
Leaf-toed geckos, genus Phyllodactylus, encompass approximately 50 species distributed throughout the TDF of the Neotropics from northwestern Mexico and extending to southern
California and the peninsula of Baja California southward to north-central South America and the West Indies (Dixon 1962). Phyllodactylus are primarily nocturnal commonly found in a diverse array of microhabitat types depending on geographic region (Dixon 1962). In
CHAPTER 1—INTRODUCTION 7
Mexico, for example, most species are strictly saxicolous (found on rocky outcroppings),
although some species can also be found on trees or inside cacti (Dixon & Huey 1970).
Although these lizards are often quite common throughout their range, the phylogenetic
relationships, biogeographic history, population genetic structure, and taxonomy of this
diverse group of lizards remains uncertain (Dixon 1960, 1964). However, recent
investigations utilizing genetic data have documented cryptic diversity throughout the
Mexican dry forests (Blair et al. 2009). Further, their apparent reliance on TDF habitat makes
them an ideal choice to examine both historical and contemporary processes governing
genetic diversification throughout this ecosystem.
4. Objectives of Thesis
The goals of this thesis are to test multiple spatial and temporal hypotheses affecting genetic
diversification in leaf-toed geckos throughout the Mexican TDF. All fieldwork was in
collaboration with Prof. Robert W. Murphy (supervisor; University of Toronto), Prof. Fausto
Roberto Méndez de la Cruz (National Autonomous University of Mexico), and Victor Hugo
Jimenez Arcos (National Autonomous University of Mexico), with additional field assistance
provided by numerous other individuals (see Acknowledgments). In Chapter 2, I use a
combination of standard molecular cloning along with next-generation 454 DNA sequencing
to isolate 12 highly polymorphic microsatellite loci for the Mexican yellow-bellied gecko
Phyllodactylus tuberculosus saxatilis. In Chapter 3, I use these microsatellite loci to perform
a landscape genetic analysis of P. tuberculosus in the region surrounding Alamos, Sonora.
Specifically, I seek to determine if forest structure, stream connectivity, slope, anthropogenic disturbance, and minimum temperature influence rates and patterns of gene flow between
CHAPTER 1—INTRODUCTION 8 populations. In Chapter 4, I conduct a multilocus phylogenetic study of the P. tuberculosus species group (P. tuberculosus, P. lanei, and P. muralis) throughout the TDF of western
Mexico. Using new, powerful coalescent methods, I test hypotheses regarding the relative roles of climate change, geography, and geological formations on rates and patterns of cladogenesis in these geckos. Chapter 5 explores the use of the microsatellite data gathered in
Chapter 3 along with de novo mtDNA sequences by comparing and contrasting historical and contemporary demographic processes in P. tuberculosus throughout the region surrounding
Alamos. More specifically, I test for a significant difference in historical versus contemporary migration rates and effective population sizes, and how these metrics correlate with contemporary landscape structure. I then test the ability of mtDNA to capture contemporary landscape processes influencing genetic connectivity. Finally, in Chapter 6, I examine the taxonomic consequences of historical biogeographic processes operating throughout the peninsula of Baja California. This chapter is now published in the journal
Zootaxa (Blair et al. 2009). Chapter 7 briefly summarizes the main findings of the thesis. All chapters are written in collaboration with Robert W. Murphy. References follow the style of the journal Molecular Ecology.
CHAPTER 1—INTRODUCTION 9
References
Arenas M, Posada D (2010) The effect of recombination on the reconstruction of ancestral
sequences. Genetics, 184, 1133–1139.
Balkenhol N, Landguth EL (2011) Simulation modeling in landscape genetics: on the need to
go further. Molecular Ecology, 20, 667–670.
Balkenhol N, Waits LP, Dezzani RJ (2009) Statistical approaches in landscape genetics: an
evaluation of methods for linking landscape and genetic data. Ecography, 32, 818–830.
Becerra JX (2005) Timing the origin and expansion of the Mexican tropical dry forest.
Proceedings of the National Academy of Sciences USA, 102, 10919–10923.
Becerra JX, Venable E (2008) Sources and sinks of diversification and conservation
priorities for the Mexican tropical dry forest. PLoS One, 3, e3436.
doi:10.1371/journal.pone.0003436.
Beerli P (2006) Comparison of Bayesian and maximum likelihood inference of population
genetic parameters. Bioinformatics, 22, 341–345.
Beerli P, Felsenstein J (2001) Maximum likelihood estimation of a migration matrix and
effective population sizes in n subpopulations by using a coalescent approach.
Proceedings of the National Academy of Sciences of the USA, 98, 4563–4568.
Blair C, Murphy RW (2011) Recent trends in molecular phylogenetic analysis: where to
next? Journal of Heredity, 102, 130–138.
Blair C, F.R. Mendez de la Cruz, A. Ngo, J. Lindell, A. Lathrop, and R.W. Murphy (2009)
Molecular phylogenetics and taxonomy of leaf-toed geckos (Phyllodactylidae:
Phyllodactylus) inhabiting the peninsula of Baja California. Zootaxa, 2027, 28–42.
CHAPTER 1—INTRODUCTION 10
Blair C, Weigel D, Balazik M et al. (2012) A simulation-based evaluation of methods for
inferring linear barriers to gene flow. Molecular Ecology Resources.
Braunisch V, Segelbacher G, Hirzel AH (2010) Modelling functional landscape connectivity
from genetic population structure: a new spatially explicit approach. Molecular Ecology,
19, 3664–3678.
Brito PH, Edwards SV (2009) Multilocus phylogeography and phylogenetics using
sequence-based markers. Genetica, 135, 439–455.
Ceballos G, García A (1995) Conserving neotropical biodiversity: the role of dry forests in
western México. Conservation Biology, 9, 1349–1356.
Degnan JH, Rosenberg NA (2006) Discordance of species trees with their most likely gene
trees. PLoS Genetics, 2, 762–768.
Degnan JH, Rosenberg NA (2009) Gene tree discordance, phylogenetic inference, and the
multispecies coalescent. Trends in Ecology and Evolution, 24, 332–340.
Devitt TJ (2006) Phylogeography of the western lyresnake (Trimorphodon biscutatus):
testing aridland biogeographical hypotheses across the Nearctic-Neotropical transition.
Molecular Ecology, 15, 4387–4407.
Dixon JR (1960) Two new geckos, genus Phyllodactylus (Reptilia: Sauria), from Michoacan,
Mexico. The Southwestern Naturalist, 5, 37–42.
Dixon JR (1962) The leaf-toed geckos, genus Phyllodactylus, of northeastern South America.
The Southwestern Naturalist, 7, 211–226.
Dixon JR (1964) The systematics and distribution of lizards of the genus Phyllodactylus in
North and Central America. New Mexico State University Research Center, Scientific
Bulletin, 64–1, 1–139.
CHAPTER 1—INTRODUCTION 11
Dixon JR, Huey RB (1970) Systematics of the lizards of the gekkonid genus Phyllodactylus
of mainland South America. Los Angeles County Museum Contributions in Science, 192,
1–78.
Epps CW, Wehausen JD, Bleich VC et al. (2007) Optimizing dispersal and corridor models
using landscape genetics. Journal of Applied Ecology, 44, 714–724.
Edwards SV (2008) Is a new and general theory of molecular systematics emerging?
Evolution, 63, 1–19.
Edwards SV, Beerli P (2000) Perspective: gene divergence, population divergence, and the
variance in coalescence times in phylogeographic studies. Evolution, 54, 1839–1854.
Edwards SV, Bensch S (2009) Looking forwards or backwards in avian phylogeography? A
comment on Zink and Barrowclough 2008. Molecular Ecology, 18, 2930–2933.
García A (2006) Using ecological niche modeling to identify diversity hotspots for the
herpetofauna of Pacific lowlands and adjacent interior valleys of Mexico. Biological
Conservation, 130, 25–46.
Gillespie TW, Grijalva A, Farris CN (2000) Diversity, composition, and structure of tropical
dry forests in Central America. Plant Ecology, 147, 37–47.
Hare MP (2001) Prospects for nuclear gene phylogeography. Trends in Ecology & Evolution,
16, 700–706.
Heled J, Drummond AJ (2008) Bayesian inference of population size history from multiple
loci. BMC Evolutionary Biology, 8, 289.
Heled J, Drummond AJ (2010) Bayesian inference of species trees from multilocus data.
Molecular Biology and Evolution, 27, 570–580.
Hewitt G (2000) The genetic legacy of Quaternary ice ages. Nature, 405, 907–913.
CHAPTER 1—INTRODUCTION 12
Holderegger R, Wagner HH (2006) A brief guide to landscape genetics. Landscape Ecology,
21, 793–796.
Holderegger R, Wagner HH (2008) Landscape genetics. BioScience, 58, 199–207.
Kingman J (1982) The coalescent. Stochastic Processes and their Applications, 13, 235–248.
Knowles LL (2009) Estimating species trees: methods of phylogenetic analysis when there is
incongruence across genes. Systematic Biology, 58, 463–467.
Kubatko LS, Carstens BC, Knowles LL (2009) STEM: species tree estimation using
maximum likelihood for gene trees under coalescence. Bioinformatics, 25, 971–973.
Landguth EL, Cushman SA, Schwartz MK, McKelvey KS, Murphy M, Luikart G (2010)
Quantifying the lag time to detect barriers in landscape genetics. Molecular Ecology, 19,
4179–4191.
Liu L (2008) BEST: Bayesian estimation of species trees under the coalescent model.
Bioinformatics, 24, 2542–2543.
Liu L, Pearl DK, Brumfield RT, Edwards SV (2008) Estimating species trees using multiple-
allele DNA sequence data. Evolution, 62, 1–12.
Maddison WP (1997) Gene trees in species trees. Systematic Biology, 46, 523–536.
Manel S, Schwartz MK, Luikart G, Taberlet P (2003) Landscape genetics: combining
landscape ecology and population genetics. Trends in Ecology and Evolution, 18, 189–
197.
Manel S, Joost S, Epperson BK et al. (2010) Perspectives on the use of landscape genetics to
detect genetic adaptive variation in the field. Molecular Ecology, 19, 3760–3772.
Martin DP, Lemey P, Posada D (2011) Analysing recombination in nucleotide sequences.
Molecular Ecology Resources, 11, 943-955.
CHAPTER 1—INTRODUCTION 13
McCormack JE, Heled J, Delaney KS, Peterson AT, Knowles LL (2011) Calibrating
divergence times on species trees versus gene trees: implications for speciation history of
Aphelocoma jays. Evolution, 65, 184–202.
Moore WS (1995) Inferring phylogenies from mtDNA variation: mitochondrial gene trees
versus nuclear gene trees. Evolution, 49, 718–726.
Murphy PG, Lugo AE (1986) Ecology of tropical dry forest. Annual Review of Ecology and
Systematics, 17, 67–88.
Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity
hotspots for conservation priorities. Nature, 403, 853–858.
Robichaux RH, Yetman DA (2000) The tropical deciduous forest of Alamos. Biodiversity of
a threatened ecosystem in Mexico. The University of Arizona Press.
Spear SF, Storfer A (2008) Landscape genetic structure of coastal tailed frogs (Ascaphus
truei) in protected vs. managed forests. Molecular Ecology, 17, 4642–4656.
Stephens M, Smith N, Donnelly P (2001) A new statistical method for haplotype
reconstruction from population data. American Journal of Human Genetics, 68, 978–
989.
Storfer A, Murphy MA, Spear SP, Holderegger R, Waits LP (2010) Landscape genetics:
where are we now? Molecular Ecology, 19, 3496–3514.
Trejo RI, Dirzo R (2000) Deforestation of seasonally dry tropical forest towards its northern
distribution: a national and local analysis in México. Biological Conservation, 94, 133–
142.
Vandergast AG, Bohonak AJ, Weissman DB, Fisher RN (2007) Understanding the genetic
effects of recent habitat fragmentation in the context of evolutionary history:
CHAPTER 1—INTRODUCTION 14
phylogeography and landscape genetics of a southern California endemic Jerusalem
cricket (Orthoptera: Stenopelmatidae: Stenopelmatus). Molecular Ecology, 16, 977–992.
Wang IJ (2010) Recognizing the temporal distinctions between landscape genetics and
phylogeography. Molecular Ecology, 19, 2605–2608.
Wang IJ (2011) Choosing appropriate markers and analytical methods for testing landscape
genetic hypotheses. Molecular Ecology, 20, 2480–2482.
Weisrock DW, Smith SD, Chan LM et al. (2012). Concatenation and concordance in the
reconstruction of mouse lemur phylogeny: an empirical demonstration of the effect of
allele sampling in phylogenetics. Molecular Biology and Evolution, doi:
10.1093/molbev/mss008.
Zarza E, Reynoso VH, Emerson BC (2008) Diversification in the northern Neotropics:
mitochondrial and nuclear DNA phylogeography of the iguana Ctenosaura pectinata and
related species. Molecular Ecology, 17, 3259–3275.
Zhang DX, Hewitt GM (2003) Nuclear DNA analyses in genetic studies of populations:
practice, problems and prospects. Molecular Ecology, 12, 563–584.
Zink RM (2010) Drawbacks with the use of microsatellites in phylogeography: the song
sparrow Melospiza melodia as a case study. Journal of Avian Biology, 41, 1–7.
Zink RM, Barrowclough GF (2008) Mitochondrial DNA under siege in avian phylogeography. Molecular Ecology, 17, 2107–2121.
CHAPTER 1—INTRODUCTION 15
Figure Legends
Fig. 1. Example of a gene tree embedded within a species tree illustrating the influence of deep coalescence (black dots) on species tree inference. The figure shows that deep coalescence can result from both incomplete lineage sorting in recently diverged species (A and B) as well as from older speciation events shaped by short internal branches (Species E versus Species C and D). All allelic copies in Species C, D, and E share a most-recent common ancestor indicating that enough time has passed to reach reciprocal monophyly.
Different colour shades represent different ancestral species. Figure taken from Knowles
(2009).
CHAPTER 1—INTRODUCTION 16
CHAPTER 2
Using next-generation DNA sequencing for rapid microsatellite
discovery in Mexican leaf-toed geckos (Phyllodactylus
tuberculosus)
Abstract
I use a combination of 454 shotgun pyrosequencing and traditional molecular cloning to
isolate microsatellite loci for the Mexican yellow-bellied leaf-toed gecko (Phyllodactulus
tuberculosus). A pyrosequencing run on a 454 GS Junior yields 62 Mbp of data composed of
150k fragments with an average length of 412 bp. Among the fragments, 18 144 (12%)
contain a suitable microsatellite. Dinucleotides and tetranucleotides are most frequently
encountered. I then genotype 30 individuals of P. tuberculosus using 12 polymorphic loci. I find moderate allelic diversity and heterozygosity. No loci are in linkage disequilibrium and only two loci show deviations from Hardy-Weinberg expectations. These results further highlight the utility of next-generation sequencing for discovering molecular markers in non-model taxa and suggest that sequencing on a 454 GS Junior system is a rapid and cost-effective approach for microsatellite characterization.
17 CHAPTER 2—MARKER DEVELOPMENT 18
Leaf-toed geckos (Phyllodactylus) occur in arid to semi-arid regions of the Neotropics
(Dixon 1964). The yellow-bellied gecko, P. tuberculosus, ranges from southern Sonora,
Mexico to northern Costa Rica. In Mexico, species are commonly encountered throughout
tropical dry forests (TDF) habitat. TDF is considered one of Earth’s biodiversity hotspots
(Myers et al. 2000; García 2006) that are experiencing rapid rates of deforestation (Trejo &
Dirzo 2000). However, little is known about how fragmentation will influence population
connectivity for species inhabiting this ecosystem. Further, species within this genus pose a
conservation concern as local people actively kill species of the genus as they are assumed to
be venomous (pers. obs.). At present, the species has yet to be assessed by the IUCN. I
combine molecular cloning with NGS to isolate and characterize polymorphic microsatellite
loci for P. tuberculosus. The microsatellite loci obtained will be used for subsequent studies
that examine fine-scale population genetic structure of geckos within Mexico’s TDF.
In 2008 I sampled 30 individuals of P. tuberculosus from a single population adjacent
to the Río Cuchujaqui near Alamos, Sonora. Tail tips were collected for genetic material and
preserved in the field with 95% ethanol. DNA was extracted using standard phenol- chloroform procedures. Extracts were precipitated and resuspended in distilled water.
I first used standard molecular cloning methods to isolate microsatellite loci (Hamilton
et al. 1999). Genomic DNA was digested with the restriction enzymes NheI RsaI and XmnI.
Following digestion, SNX linkers were ligated to the digested fragments for amplification via
PCR. Following amplification, I enriched for the following microsatellite motifs: (CGG)4,
(AAAC)6, (AAT)12, (AAAT)8. Fragments containing the microsatellite motifs were captured
by streptavidin beads. The linker-ligated enriched DNA was then amplified, inserted into a
plasmid vector, and cloned into E. coli cells. Positive clones were then amplified using M13 CHAPTER 2—MARKER DEVELOPMENT 19
forward and reverse primers and directly sequenced using the BigDye Terminator Cycle
Sequencing Kit (Applied Biosystems). Sequences were then read on an ABI 3730 DNA
sequencer (Applied Biosystems) and edited using BIOEDIT version 7.0.9 (Hall 1999). A
single individual was also sent to Trent University's Natural Resources DNA Profiling and
Forensics Centre for 454 shotgun sequencing using a full run on a 454 GS Junior. For
genomic sequencing, I extracted DNA using a DNeasy Tissue Kit (QIAGEN) to obtain a
cleaner product. This extract was then sent to Trent for library construction, emulsion PCR,
and 454 sequencing.
MSATCOMMANDER (Faircloth 2008) was used to search for 454 fragments (~150k) containing microsatellites. When suitable clones and 454 fragments were found, I used
PRIMER3 (Rozen & Skaletsky 2000) to develop primers. The total data set consisted of 12
loci (5 from cloning and 7 from 454 sequencing). I chose loci from the 454 run containing
tetranucleotide repeats to minimize potential scoring errors due to stutter.
PCR amplification was performed using a 12.5 µL reaction volume containing 1.25
µL 10X PCR buffer containing 1.5 mM MgCl2, 0.1 µL BSA, 1 µL 10 mM of each primer,
0.4 µL 10 mM dNTPs and 0.1 µL taq DNA polymerase (New England Biolabs). Water and
MgCl2 concentrations were adjusted to obtain optimal PCR products (Table 1). Each forward
primer was fluorescently labelled with a 6-FAM M13 tag following Schuelke (2000). All
PCRs were performed on an Eppendorf AG 5345 thermal cycler using the following
conditions: initial denaturation at 95 °C for 2 min, followed by 35 cycles of 95 °C for 30 s,
annealing at the optimal temperature for 45 s (Table 1), and extension at 72 °C for 45 s, with a final extension temperature of 72 °C for 6 min. Genotyping was performed by mixing 2 µL of each PCR product with a mixture of Hi-Di and GENESCAN 500 LIZ size standard (Applied CHAPTER 2—MARKER DEVELOPMENT 20
Biosystems). Samples were read on an ABI 3730 DNA sequencer and alleles were scored
using the software GENEMARKER version 1.9.1 (SoftGenetics).
I used MICROSATELLITE ANALYSER version 4.05 (Dieringer & Schlötterer 2003) to
calculate the number of alleles per locus, allelic range, and observed and expected
heterozygosities (HO and HE, respectively). Hardy-Weinberg and linkage equilibrium (log
likelihood ratio statistic) were tested using GENEPOP version 4.0.10 (Rousset 2008) using
default Markov chain parameters. I used MICRO-CHECKER version 2.2.3 (van Oosterhout et al. 2004) to check for scoring errors due to stuttering, long allele dropout, or null alleles.
I was able to obtain reliable genotypes for 12 loci (three trinucleotide and nine
tetranucleotide). Diversity was moderate, with the number of alleles per locus ranging
between three and 14. Observed heterozygosity ranged between 0.6207 to 0.9000 and
expected heterozygosity ranged from 0.5668 to 0.9136 (Table 1). No loci showed signs of
linkage disequilibrium after Bonferroni correction (P > 0.05). Loci G2_37 and Pt1 showed
deviations from Hardy-Weinberg expectations with an excess of homozygotes (P < 0.01 and
P = 0.0224, respectively). No loci showed evidence of stutter or long allele dropout.
However, G2_37 showed evidence of null alleles (0.0835). This study adds to a small body
of literature that uses NGS to isolate polymorphic microsatellite markers in reptiles (e.g.
Castoe et al. 2009; Metzger et al. 2011; Smith et al. 2011) and suggests that the 454 GS
Junior is a rapid and cost-effective means to isolate microsatellite loci.
Acknowledgments
I kindly thank the generation sequencing facilities available at Trent University's Natural
Resources DNA Profiling and Forensics Centre for performing the 454 sequencing for this CHAPTER 2—MARKER DEVELOPMENT 21
study. I thank Martin Gabriel Figueroa for help in the field. Stephanie Meyer provided
logistic support for potential field sites. For laboratory assistance I thank Oliver Haddrath,
Amy Lathrop, and Kristen Choffe at the Royal Ontario Museum. I thank Brant Faircloth for assistance with MSATCOMMANDER. This work was supported by the Natural Sciences and
Engineering Research Council (NSERC) Discovery Grant A3148 (RWM) and the Theodore
Roosevelt Memorial Grant through the American Museum of Natural History (CB). CHAPTER 2—MARKER DEVELOPMENT 22
References
Castoe TA, Poole AW, Gu W, de Koning APJ, Daza JM, Smith EN, Pollock DD (2009)
Rapid identification of thousands of copperhead snake (Agkistrodon contortrix)
microsatellite loci from modest amounts of 454 shotgun genome sequence. Molecular
Ecology Resources, 10, 341–347.
Dieringer D, Schlötterer C (2003) MICROSATELLITE ANALYSER (MSA) a platform
independent analysis tool for large microsatellite data sets. Molecular Ecology Resources,
3, 167–169.
Dixon JR (1964) The systematics and distribution of lizards of the genus Phyllodactylus in
North and Central America. New Mexico State University Scientific Bulletin, 64, 1–139.
Faircloth BC (2008) MSATCOMMANDER: detection of microsatellite repeat arrays and
automated, locus-specific primer design. Molecular Ecology Resources, 8, 92–94.
García A (2006) Using ecological niche modelling to identify diversity hotspots for the
herpetofauna of Pacific lowlands and adjacent interior valleys of Mexico. Biological
Conservation, 130, 25–46.
Hall TA (1999) BIOEDIT: a user-friendly biological sequence alignment editor and analysis
program for Windows 95/98/NT. Nucleic Acids Symposium Series, 41, 95–98.
Hamilton MB, Pincus EL, Di Fiore A, Fleischer RC (1999) Universal linker and ligation
procedures for construction of genomic DNA libraries enriched for microsatellites.
BioTechniques, 27, 500–507.
Metzger C, Ferchaud A, Geiser C & Ursenbacher S (2011) New polymorphic microsatellite
markers of the endangered meadow viper (Vipera ursinii) identified by 454 high- CHAPTER 2—MARKER DEVELOPMENT 23
throughput sequencing: when innovation meets conservation. Conservation Genetics
Resources, 3, 589-592
Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity
hotspots for conservation priorities. Nature, 403, 853–858.
Rousset F (2008) GENEPOP’007: a complete re-implementation of the GENEPOP software
for Windows and Linux. Molecular Ecology Resources, 8, 103–106.
Rozen S, Skaletsky HJ (2000) Primer3 on the WWW for general users and for biologist
programmers. In: Krawetz S, Misener S, Krawetz S, editors. Miseners. Bioinformatics
methods and protocols: methods in molecular biology. Totowa (NJ): Humana Press. P.
365–386.
Schuelke M (2000) An economic method for the fluorescent labelling of PCR fragments.
Nature Biotechnology, 18, 233–234.
Smith A, Gardner M, Bull M & Driscoll D (2011) Primers for novel microsatellite markers in
“fire-specialist” lizards (Amphibolurus norrisi, Ctenotus atlas and Nephrurus stellatus)
and their performance across multiple populations. Conservation Genetics Resources, 3,
345-350.
Trejo I, Dirzo R (2000) Deforestation of seasonally dry tropical forest: a national and local
analysis in Mexico. Biological Conservation, 94, 133–142.
Van Oosterhout C, Hutchinson WF, Wills DPM, Shipley P (2004) MICRO-CHECKER:
software for identifying and correcting genotyping errors in microsatellite data.
Molecular Ecology Notes, 4, 535–538. CHAPTER 2—MARKER DEVELOPMENT 24
Table 1 Characterization of twelve microsatellite loci for the leaf-toed gecko Phyllodactylus tuberculosus (n = 30) based on molecular cloning and 454 shotgun sequencing.
GenBank Locus Method Repeat Unit Primer Sequence (5'–3') Label Ta (°C) MgCl2 (mM) Alleles Range (bp) HO HE Accession
G2_22 Cloning (GAA)19 F: TGCCAGAAGTGGAAAACTAACC FAM 60 1.50 14 231–282 0.9000 0.9136 R: ATTGTGGAACTGGCATACCC
G2_37 Cloning (CTT)12 F: GAGAGGAAGGGAAGGCAACT FAM 55 1.50 13 249–303 0.7000 0.8616 R: GCCGCCTGATACTCTTTTTG
G2_96 Cloning (CAAA)11 F: GATTTGTAAGCCGCCTCAGT FAM 55 1.50 8 197–229 0.7667 0.7373 R: TCCAAGGACCTTTGTTGTTTG
G2_85 Cloning (CAAA)6 F: TGTGTTCTGCAAGGGAGAAA FAM 55 2.00 5 243–259 0.7333 0.6531 R: GGAACCATGTGAAGCCATCT
G2_59 Cloning (CCG)3 F: CGACGAAGACTCCAGAGACC FAM 60 2.00 3 259–265 0.6207 0.5668 R: TGGATGTTGCTGCAATTTGT 454 Pt1 sequencing (AAAT)10 F: ATTTTCCCTGGCTTTCCTTT FAM 55 2.50 9 231–267 0.6818 0.8277 R: CTTCCATTGTCAACGGGACT 454 Pt2 sequencing (AAAT)10 F: CCCCCTCTGGTGGTTATATTT FAM 53 1.50 6 229–265 0.7667 0.7418 R: CATGTCTGGCCAGTGTTCTG 454 Pt6 sequencing (AATG)11 F: GCAGTTGTTTTACCGCATCA FAM 55 1.50 5 198–214 0.8333 0.6915 R: AAGGCAGGACAAACCCTACC 454 Pt7 sequencing (AATG)13 F: TTATATTGTATGCCGCCTTGG FAM 60 2.00 9 231–263 0.6897 0.8361 R: AGCTGGAACGTGGTTCAGAG 454 Pt12 sequencing (ATCC)10 F: TGGCACTGAAATGGATGCTA FAM 60 2.00 8 206–246 0.8148 0.8162 R: AGGGTGGGCTTTTAGATGGT 454 Pt15 sequencing (CATT)10 F: CCCTGCTTTCCAAAACATCT FAM 55 1.50 7 213–237 0.7931 0.8312 R: AATGTGAAGCGGGAGGGTAT 454 Pt19 sequencing (GGAT)15 F: AGCCATCCTGAGATGTTCGT FAM 54 2.00 7 183–219 0.8667 0.8395 R: GATTGTTGACATGCAAAGCTAGT
All forward primers were labelled with an M13 tail and amplified using the three primer method of Schuelke (2000).Ta, annealing temperature; MgCl2, concentration of MgCl2 in each reaction; HO, observed heterozygosity; HE, expected heterozygosity
CHAPTER 3
Landscape genetics of leaf-toed geckos in the sub-tropical dry
forests of northern Mexico
Abstract
Habitat fragmentation due to both natural and anthropogenic forces continues to threaten the evolution and maintenance of biological diversity. This is of particular concern in tropical regions that are experiencing elevated rates of habitat loss. Although less well-studied than tropical rain forests, tropical dry forests (TDF) contain an enormous diversity of species and continue to be threatened by anthropogenic activities including grazing and agriculture.
However, little is known about the processes that shape genetic connectivity in species
inhabiting TDF ecosystems. I adopt a landscape genetic approach to understanding functional
connectivity for leaf-toed geckos (Phyllodactylus tuberculosus) at multiple sites near the
northernmost limit of this ecosystem at Alamos, Sonora, Mexico. Traditional population
genetic analyses are combined with univariate and multivariate landscape analyses to test
hypotheses on the potential drivers of spatial genetic variation. Moderate levels of within-
population diversity and substantial levels of population differentiation are revealed by FST
and Dest. Analyses using STRUCTURE suggest the occurrence of from three to five genetic
clusters. Landscape genetic analysis suggests that forest cover, stream connectivity,
undisturbed habitat, slope, and minimum temperature of the coldest period explain more
genetic variation than do simple Euclidean distances. Additional landscape genetic studies
25 CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 26 throughout TDF habitat are required to understand species-specific responses to landscape and climate change and to identify common drivers. I also urge researchers interested in using multivariate distance methods to test for, and report, significant correlations among predictor matrices which can impact results, particularly when adopting least-cost path approaches.
1 Introduction
Dispersal is a fundamental process that can greatly influence ecological and demographic
trajectories within and between subpopulations (Clobert et al. 2001). For example, dispersal
often leads to gene flow, the transfer of genetic information from one population to another
(Slatkin 1987; Bohonak 1999). Maintaining adequate rates of gene flow is often beneficial
because populations experiencing low gene flow are susceptible to a loss of genetic diversity
due to inbreeding and drift (Slatkin 1987; Frankham et al. 2002; Vilá et al. 2002). A lack of
genetic diversity may also make it difficult for populations to adapt to changing
environmental conditions and eventually lead to local extinction (Lande 1993; Crispo et al.
2011).
Dispersal rates can result from a suite of different factors including species-specific
philopatry, intra- and interspecific interactions, predation pressures, physiological tolerances,
and simple geographic (Euclidean) distance (Spear et al. 2010). Landscape heterogeneity
often plays a substantial role in an organism’s choice to disperse or not (With et al. 1997;
Ricketts 2001; Bowne & Bowers 2004) and both natural and anthropogenic habitat- fragmentation can detrimentally affect the connectivity and persistence of populations CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 27
(Fahrig 1997, 1998, 2003, 2007). Landscape genetics seeks to explicitly quantify the influence of landscape and environmental variables on microevolutionary processes such as gene flow and natural selection (Manel et al. 2003; Holderegger & Wagner 2008). The approach extends traditional population genetic studies by explaining the spatial distribution of genetic variation by using components of the landscape. This particularly powerful approach to studying fine-scale population structure and its application is meeting with much success as Storfer et al. (2010) review. By combining rapidly evolving molecular markers such as microsatellites with novel approaches to statistical analysis, landscape genetics identifies a suite of environmental variables likely to influence population genetic structure
(Spear et al. 2005; Vignieri 2005; Wang et al. 2009). Identifying the landscape components facilitating or constraining gene flow can aid in delimiting areas for conservation (Neel
2008), for example, by designing corridors that maximize functional connectivity (Epps et al.
2007; Braunisch et al. 2010). Today, there is a dearth of landscape genetic studies focusing on tropical localities (Storfer et al. 2010), which harbour the majority of species (Myers et al.
2000). Thus, compared to temperate localities, relatively little is known about the processes influencing functional connectivity in species inhabiting this mega-diverse region.
Tropical dry forests (TDF), which also occur in the Neotropics, are a major biodiversity hotspot (Myers et al. 2000). They occur in a semi-continuous belt throughout the
New World from northern Mexico southward into northern South America. Both structural and functional differences differentiate these forests from tropical rainforests (Murphy &
Lugo 1986). The amount of annual rainfall is a primary distinction between these forests, with TDF experiencing up to eight months of arid-like conditions followed by four months of deluge. Further, canopy height is far lower in TDF than in rainforests and many plant species CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 28 in these forests possess thorny spines throughout their shoots. Although evidence suggests that these forests may be far more diverse than currently realized (Robichaux & Yetman
2000), habitat fragmentation due to both natural and anthropogenic factors is threatening the evolutionary potential of inhabiting species (Trejo & Dirzo 2000; Sigala-Rodríguez &
Greene 2009). Habitat fragmentation is of particular concern in rapidly developing countries such as Mexico, where dense continuous forest is being cleared for both livestock and agriculture (Trejo & Dirzo 2000). In Mexico, TDF reaches its northern limit near Alamos,
Sonora, whereas forest density is highest in the southwest states of Jalisco, Colima,
Michoacán, and Guerrero (Becerra 2005). Fragmentation of these forests is documented to have occurred for decades and continues to increase, yet we know relatively little about how fragmentation effects species and populations distributed throughout these ecosystems.
The diversity of Mexico’s herpetofauna is substantial, with approximately 1 000 described species and many more awaiting formal description (Robichaux & Yetman 2000;
Wilson et al. 2010). Flanking the Pacific coast, Mexico’s TDF also appears to be a centre of endemism for a variety of amphibian and reptilian species (García 2006). A variety of species of amphibians and reptiles are restricted to the TDF of western Mexico, including many species of leaf-toed geckos of the genus Phyllodactylus (Dixon 1964) These lizards inhabit arid to semi-arid areas from southern California southward through Middle America into northern South America and into the West Indies. Like many geckos, they are commonly found on vertical surfaces including bridges and buildings.
Herein I use the Mexican yellow-bellied gecko, Phyllodactylus tuberculosis, to test a variety of landscape genetic hypotheses operating in Mexico’s TDF. This species is an ideal choice to examine relationships between landscape and genetics for two reasons. First, the CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 29 gecko’s geographic distribution mirrors the distribution of TDF in Mexico. Second, this species, along with others in the genus, are at risk of local and area-wide extinctions due to habitat fragmentation and recent introductions of non-native, all female species, such as geckos of the genus Hemidactylus, which appears to be displacing leaf-toed geckos (pers. obs.). The study site lays in the northernmost limit of TDF near Alamos, Sonora. I test the following null hypotheses: 1) Genetic diversity not spatially structured, other than isolation- by-distance; 2) forest fragmentation does not influence functional connectivity and gene flow; and 3) univariate and multivariate landscape and environmental data do not correlate with the spatial distribution of genetic differentiation between populations. Rejection of these null hypotheses and drawing robust conclusions favouring alternative hypotheses requires principles and practices drawn from diverse research disciplines (Manel et al. 2003).
Therefore, I use recent advances in landscape genetic techniques (Storfer et al. 2007, 2010;
Balkenhol et al. 2009; Dyer et al. 2010) to test the null hypothesis (Cushman et al. 2006) and identify which landscape variables influence functional connectivity of leaf-toed geckos in this threatened ecosystem.
2 Materials and Methods
2.1 Sampling
My study area was the northern limits of TDF (Fig. 1), an area that today has a relatively high degree of forest cover compared to other locations throughout western Mexico due, in part, to the federal protection of land (Sierra de Alamos/Rio Cuchujaqui Reserve). The CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 30 landscape at lower elevations (generally under 500 m) consisted of tropical dry thornscrub that gradually transitioned into TDF closer to the Sierra de Alamos with increasing elevation.
From 2008 to 2010, I sampled 336 leaf-toed geckos from 12 different localities (mean = 28 individuals per locality) throughout the landscape surrounding the Alamos region. Sampling localities were chosen based on landscape characteristics to allow for testing hypotheses. I sampled on opposite banks of two relatively large rivers or arroyos (Cuchujaqui and Tabelo) to determine if these features served as barriers to gene flow. When possible, I obtained a minimum of 30 individuals per population. Following protocols approved by the Royal
Ontario Museum Animal Care Committee, tail tips were taken in the field and immediately preserved in 95% ethanol for subsequent genetic analysis.
2.2 DNA extraction and genotyping
DNA was extracted using standard phenol-chloroform procedures. I used the polymerase
chain reaction (PCR) to amplify 12 polymorphic microsatellite loci developed specifically
for Phyllodactylus tuberculosus (see Blair et al. in review for description of loci and PCR
conditions). Both negative and positive controls were run on each PCR plate. PCR products
were visualized on an ABI 3730 automated sequencer (Applied Biosystems Inc.) at the Royal
Ontario Museum. Genotyping was performed using the software GeneMarker v. 1.95
(SoftGenetics). I re-ran PCRs for approximately 10% of the samples to quantify any potential
errors in genotyping.
2.3 Genetic diversity
CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 31
I calculated diversity statistics for each site including number of alleles and allelic richness using MICROSATELLITE ANALYSER v. 4.05 (Dieringer & Schlötterer 2003). Observed (HO)
and expected (HE) heterozygosites were calculated in TFPGA v. 1.3 (Miller 1997). I tested
for site-specific deviations from Hardy-Weinberg (exact test) and linkage equilibrium using
GENEPOP on the web v. 4.0.10 (Raymond & Rousset 1995; Rousset 2008). Significance for
tests was assessed using the Markov chain method using 100 batches with 1 000 iterations
per batch. I used the false-discovery rate method to control for multiple comparisons
(Benjamini & Hochberg 1995).
2.4 Genetic differentiation
There has been considerable recent debate as to the best FST-like analogue for assessing
genetic differentiation between populations (Heller & Siegismund 2009; Jost 2009; Ryman &
Leimar 2009; Gerlach et al. 2010). The traditional FST metric (GST for multiple loci and alleles) has been shown to be highly sensitive to within-population heterozygosity, making it difficult to compare values between studies and markers (Hedrick 2005; Jost 2008). Further, because FST is dependent on the number of alleles and heterozygosity, two populations have shown low FST values despite sharing no alleles. For example, Gerlach et al. (2010) showed that in two hypothetical populations with a total of 16 alleles, FST was constrained to less than
0.1 even when the two populations shared no alleles. New metrics such as G’ST (Hedrick
2005) and D (Jost 2008) have been developed that circumvent some of the more common
issues found with FST (reviewed by Meirmans & Hedrick 2011). Although these new metrics have been criticised (Whitlock 2011), information about differentiation can be maximized by comparing multiple statistics (Meirmans & Hedrick 2011). Thus, I calculated both traditional CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 32
FST and Jost’s D between populations. Whereas FST dealt with fixed biallelic markers, D
measured the degree of allelic differentiation between populations and was particularly
useful for highly polymorphic markers such as microsatellites (Jost 2008). MICROSATELLITE
ANALYSER was used to calculate both global and pairwise multilocus FST following Weir &
Cockerham (1984). Significance of comparisons was assessed using 10 000 permutations while implementing a Bonferroni correction for multiple tests. I used SMOGD v. 1.2.5
(Crawford 2010) to calculate pairwise Dest based on harmonic means estimated over all 10
loci. I used the unbiased estimates of all metrics to account for artefacts of sample size (Jost
2008, 2009).
2.5 Population structure
I tested for the presence of genetic clusters using STRUCTURE v. 2.3.3 (Pritchard et al. 2000;
Falush et al. 2003) to classify population structure. STRUCTURE, a Bayesian program, used
Hardy-Weinberg and linkage disequilibrium to cluster individuals into populations (K) based
on data from multi-locus genotypes. I employed the admixture model with correlated allele
frequencies while specifying a range of K-values (1–12). For simplicity, I created a batch run
specifying the range of K-values to be evaluated and running 10 independent runs per K.
Each run was composed of a burnin of 30 000 followed by 100 000 iterations, which was
sufficient to reach convergence. I evaluated the most likely number of clusters using both the
Pr (X/K) method as described by the program’s authors and the ΔK method of Evanno et al.
(2005). STRUCTURE HARVESTER (Earl & vonHoldt 2011) was used to visualize outputs and calculate K based on both methods. To deal with the multimodality of utilizing multiple independent runs, I used CLUMPP v. 1.1.2 (Jakobsson & Rosenberg 2007) to permute the CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 33 admixture coefficients for the runs with the chosen K-value using the “Greedy” algorithm with 1 000 random input orders. DISTRUCT v.1.1 (Rosenberg 2004) was then used to
visualize the output from CLUMPP.
Geographic (spatial) distributions have provided valuable insights in population
genetic analyses (Guillot 2009). Thus, I compared the non-spatial STRUCTURE results to those
that incorporated information about sampling locations as prior information (Hubisz et al.
2009). I introduced an additional parameter (LOCPRIOR) into the clustering analysis by
specifying a different integer for each sampling location. Often this approach has increased
the ability of the algorithm to detect relatively weak population structure that may not be
found using the aspatial model (Hubisz et al. 2009). I did not perform fully spatial models
using programs such as TESS (Chen et al. 2007; Durand et al. 2009) and GENELAND (Guillot
et al. 2005a,b) because I had GPS data for populations and not individuals. Although
GENELAND can potentially accommodate this type of sampling by the introduced error term,
some subjectivity remained as to which values to use for this parameter.
2.6 Landscape genetic analysis
I calculated effective distances between populations using least-cost path modeling (Michels
et al. 2001; Adriaensen et al. 2003; Spear et al. 2005; Wang et al. 2009). This assessed the
influence of different landscape and environmental variables on population genetic structure
(Spear et al. 2010). Unlike using Euclidean distances (IBD) to correlate geographic and
genetic distances, effective distances between individuals or populations allowed for the
direct incorporation of landscape variables into the model (Cushman et al. 2006). These
effective distances were created in a GIS environment by parameterizing different resistance CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 34 surfaces that represented the hypothesized relationships between a specific habitat feature and gene flow (Spear et al. 2010). For example, if hypothesizing that urban development versus undisturbed forested habitat constrained the movement of individuals, I assigned a higher cost value to cells representing urban habitat. The difficulty in using resistance surfaces to model functional connectivity lied in the assignment of relative cost values for different landscape features (Rayfield et al. 2010). Although parameterization based on radio telemetry, mark-recapture studies or experimental approaches may be the most quantitative way to assess relative costs, often for many species these methods have not been feasible due to financial or logistical reasons (Spear et al. 2010). Parameterization based on expert opinion of assigning relative cost values provided an alternative approach to model gene flow and the values were tested in a model-selection framework.
I tested for the relative influence of several landscape variables on genetic differentiation based on pairwise Dest. Landscape variables were selected based on expert
knowledge of which habitat characteristics were most likely important in shaping patterns of
gene flow in the species (e.g. Spear et al. 2010). My first model was based on the traditional
isolation-by-distance (IBD) model (Wright 1943) that assumed genetic differentiation was a
by-product of simple Euclidean distance without regard to the landscape. Next, I tested a
variety of landscape genetic hypotheses that explicitly consider the intervening matrix
(Ricketts 2001). First, I tested for the influence of land cover-type (TDF vegetation) on
genetic connectivity by downloading a raster data set from the North American Land Change
Monitoring System (NALCMS). Nineteen different land cover-types were classified at a 250
m spatial resolution. I created resistance surfaces by re-classifying the data to assign higher
cost values to non-forested versus forested habitat. I tested several different cost ratios (1:2, CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 35
1:10, 1:100, 1:1000) to determine how parameterization might have influenced the results. I used the Mantel test function in the R package ECODIST (R Development Core Team 2008;
Goslee & Urban 2008) to test for both the presence of IBD and to select among the four
relative cost values chosen for parameterization. Optimal values were selected based on the
Mantel r correlation statistic using 10 000 randomizations.
Because P. tuberculosus occurred in lowland tropical environments only, my second set of analyses developed least-cost paths based on slope. These data were derived from a
GTOPO30 digital elevation model (DEM) with a 1 km2 spatial resolution produced from
Natural Resources Canada and the U.S. Geological Survey. This layer consisted of seven
elevation classes that were reclassified into slope using the Spatial Analyst extension of
ARCMAP 10 (Environmental Systems Research Institute) to test the prediction that gene flow
occurred primarily throughout lowland habitats. The study area encompassed an elevation range between 100 to 500 m above sea level. For computational efficiency and because I assumed a linear relationship between elevation and gene flow (e.g. Spear & Storfer 2008), I reclassified the data into values ranging between 1 and 9, with 9 representing the highest slope values in the study.
Because I often captured geckos adjacent to streams and arroyos, I tested if dispersal
occurs primarily through stream corridors. I first obtained a polyline file representing all of
Mexico’s streams and tributaries from the GISDataDepot, a site that compiled multiple data
layers based on ESRI’s Digital Chart of the World (DCW). I then converted this data-file into
a raster file with a cell size of 100 m2 and assigned different cost values to cells encompassing streams versus those that did not. I tested the same relative cost values as out CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 36 land cover analysis (1:2, 1:10, 1:100, 1:1000) and selected the best values based on Mantel correlations.
I tested for effects of minimum temperature of the coldest period of the year because these lizards are predominantly found in warm tropical lowland habitats. A significant correlation between gene flow and minimum temperatures was predicted. Temperature data were obtained from the WorldClim database at a resolution of 1 km2. Because these data
were continuous variables, I reclassified values to represent natural breaks in temperature. As
with slope, I assumed a linear relationship between temperature and gene flow and assigned cost values ranging from 1 to 9, with 9 representing the highest cost to gene flow
representing cells with the lowest temperature.
Finally, I utilized a multivariate resistance surface representing the combined effects
of anthropogenic land-use on gene flow. These data were obtained from the Wildlife
Conservation Society (WCS) and the Center for International Earth Science Information
Network (CIESIN) and represented the combined effects of population density, built-up
areas, roads, railroads, navigable rivers, coastlines, land-use, and nighttime lights. The data
were categorized based on the Human Influence Index with values ranging between 0 and 64,
with 0 representing no human influence and 64 representing maximum influence. To create
least-cost paths, I recategorized the continuous data into discrete classes and assigned cost values ranging from 1 to 9, with 9 representing the highest cost to gene flow representing cells with the highest anthropogenic influence.
For all least-cost path-analyses, I used the LANDSCAPE GENETICS TOOLBOX 1.2.3
(Etherington 2011) implemented in ARCMAP 10 to calculate effective distances between
sampling localities. This calculated both the cost-distance and the length of the least-cost CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 37 path between any two sampling points. Because both distances could have been sensitive to relative cost values (Rayfield et al. 2010), I tested several different relative values for categorical variables as described above. For all least-cost path-analyses, I used the cumulative cost-distance because this metric minimized the degree of multicollinearity in the predictors. For all least-cost path calculations, rasters were standardized to a cell size of 100 m2.
Functional connectivity and gene flow were modeled using circuit theory (McRae
2006; McRae & Beier 2007; McRae et al. 2008). Unlike least-cost path approaches to
modeling functional connectivity, a circuit theoretic approach was particularly powerful in that it modeled all possible pathways linking populations and accounted for wider habitat
swaths. This so-called isolation-by-resistance (IBR) approach has been shown to be powerful
in modeling gene flow and functional connectivity in both simulated and empirical data sets
(McRae 2006; McRae & Beier 2007). I calculated resistance distances between populations
using CIRCUITSCAPE 3.5.7 (Shah & McRae 2008). I first used the EXPORT TO CIRCUITSCAPE
tool in ARCMAP to modify the input rasters for use with CIRCUITSCAPE. I then calculated
resistance distances between each population for each landscape variable in CIRCUITSCAPE.
Each calculation used focal points in pairwise mode and a four-neighbors connection
scheme. All calculations were based on values of per-cell resistance.
Multiple regression analyses on distance matrices (MRM; Legendre et al. 1994;
Lichstein 2007) were used to evaluate landscape genetic relationships. Although a potentially
powerful method to elucidate landscape genetic relationships (Balkenhol et al. 2009), few
studies have incorporated MRM analyses (Dyer et al. 2010). Similar to the commonly used
partial Mantel test (Smouse et al. 1986), MRM was developed to test for significant CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 38
relationships between a dependent distance matrix (e.g. linearized Dest) and any number of
indicator matrices and identify the contribution of each explanatory variable to the overall fit
of the model. Further, MRM can model polynomial and nonlinear relationships (Lichstein
2007). Each distance matrix was unfolded into vectors representing pairwise distances. The
response vector was then regressed against each indicator vector and the significance of the
model was assessed by permuting the objects of the response matrix. Because several of the
landscape variables were highly correlated (forest cover, minimum temperature, slope,
anthropogenic disturbance), I created multivariate resistance surfaces using the Weighted
Sum function under the Spatial Analyst tool in ARCMAP. I created four different surfaces
representing a composite of the four correlated variables. To understand which of the four
variables were influencing genetic structure, I assigned a weight of 10 to a different input
raster for each of the four multivariate surfaces. For example, to test whether forest structure
was the primary feature influencing functional connectivity or not, I first multiplied cost
values in the forest raster by 10 before summing the values with the remaining three rasters. I
then re-calculated least-cost paths and resistances based on the new multivariate resistance
surfaces.
I performed all MRM using ECODIST. To select among competing models, I used a stepwise regression approach, although using a backwards elimination approach gave similar
results. First, I tested the predictive power of each distance metric to determine the variable
with the largest coefficient of determination. This variable was then added to the model and
subsequent variables were either added or removed depending whether their corresponding
p-value was lower or higher than the predetermined cut-off of p = 0.05, respectively.
Included variables were required to be statistically significant. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 39
Multivariate models using all variables were not created for the least-cost analyses due to a relatively high degree of multicollinearity among the predictors, which could result in coefficients with large variances, leading to erroneous conclusions regarding the direction and magnitude of slope. Kutner et al. (2004) suggested calculating Variance Inflation Factors
(VIF) for each predictor in a model to ascertain if collinearity might be a problem in parameter estimation. Like previous studies, I used VIF values > 10 as evidence for substantial multicollinearity (Dyer et al. 2010). Thus, I ran four separate models for regressions based on least-cost paths to minimize the potential error in estimated regression coefficients. All models contained both Euclidean distance and least-cost paths based on stream connectivity. Each of the four models contained one of the remaining variables (forest cover, slope, temperature, or anthropogenic disturbance). Because multicollinearity was not a problem in the resistance distances calculated from CIRCUITSCAPE I included all variables in
the stepwise regression.
To validate the stepwise procedure for selecting a model, I calculated values of
AIC—measures the amount of information lost in a given model i relative to the true model j
(Kullback & Leibler 1951)—for competing models. The best model minimized the amount
of information lost as represented by the combination of variables with the lowest AIC value.
Corrected values of AIC accounted for smaller sample sizes. Different combinations of
variables were compared to the null model of IBD to determine if the incorporation of
landscape variables explained more of the variation in Dest. I followed Burnham & Anderson
(2002) in assuming that models with ΔAIC < 2 were strongly supported, values between 3
and 7 represented models with considerably less support, and values above 10 were highly
unlikely. I used MUMIN (Barton 2009) to calculate AIC weights for each model. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 40
3 Results
3.1 Genetic diversity
The genotyping and scoring of microsatellite alleles had an error rate of less than 1%, and, thus, high reproducibility. After controlling for false discovery rates, some loci showed significant deviations from Hardy-Weinberg expectations within sites. For example, locus
G2_96 showed heterozygote deficits at six of the 12 sites, locus P7 at eight of 12, and locus
G2_59 at four of 12. However, only three alleles were present at locus G2_59 and, thus, there was a high probability that random chance resulted in significance. Because loci G2_96 and
P7 showed a significant heterozygote deficit at multiple sites, I ran preliminary analyses with and without these loci to see how results changed. Although results did not differ substantially, I adopted a conservative approach and chose to report results from subsequent analyses excluding these two loci.
After controlling for false discovery rates, a few loci showed signs of linkage disequilibrium. For example, locus G2_22 showed linkage to loci G2_96, P2, P7, P12, and
P19. Locus G2_85 showed signs of linkage with locus P7, and locus P2 with P15. However, linkage was suggested at only two of the 12 sites (Arroyo Tabelo B and Mocuzari). If linked,
I expected to find significant linkage across multiple sites. Therefore, I concluded that the loci were independent, unlinked markers.
Exploratory results of genetic diversity were shown in Table 1. In general, within-site diversity was moderate as shown by both expected heterozygosity and allelic richness.
Expected heterozygosity ranged from 0.7025 to 0.7812 and allelic richness from 54.2341 to
65.6058. Allelic richness was calculated based on the minimum sample size per locality (10 CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 41 individuals) over all 10 loci. Allelic diversity within loci over all populations ranged from 3 alleles at locus G2_59 to 26 alleles at locus G2_37 (mean number of alleles per locus = 13.1).
In general, diversity estimates were fairly similar among sites. There was a significantly positive relationship between elevation and genetic diversity (R2 = 0.5836; p = 0.004; Fig. 2).
3.2 Genetic differentiation
Global genetic differentiation was reasonably moderate based on both FST (FST = 0.0869; p =
0.0001) and D (Dest = 0.248). Pairwise measures of differentiation revealed moderate to high levels of population divergences (Table 2). Pairwise FST values ranged from zero between
Aduana and the road to Navojoa to 0.201 between La Sierrita and El Quintero. Further, the
majority of FST values were significant based on 10 000 random permutations of alleles after a Bonferroni correction. Similarly, values of Dest were moderate to high, showing similar
relative values between populations. However, estimated values of differentiation were
generally much higher based on Dest versus FST. El Quintero and Choquincahui showed the
greatest differentiation among all comparisons based on both statistics. I also detected a
significantly positive correlation between Euclidean and genetic distances (Mantel r =
0.4897, p < 0.001).
3.3 Population structure
I used two methods to assess the degree of genetic structure and to assign individuals to
populations. I first used the traditional aspatial model in STRUCTURE that did not incorporate
sampling locality information into the clustering. Results from the batch run suggested a K- CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 42 value of 5 using the Pr(X|K) method (Supplementary Fig. S1a). Conversely, inference based on the second order rate of change in likelihood values (ΔK method; Evanno et al. 2005) suggested a K = 3 (ΔK = 138.332; Fig. S1b). With a K of 3, the following localities were grouped together into clusters with varying degrees of admixture: Cluster 1= Road to
Navojoa, House/Alamos, Arroyo Aduana, Río Cuchujaqui A and B, La Sierrita Nature
Reserve, and San Antonio; Cluster 2 = Arroyo Tabelo A and B and Mocuzari; Cluster 3 = El
Quintero and Choquincahui (Fig. 2A). Similar clustering was found with K = 5. However, both sites Mocuzari and Cuchujaqui A,B formed an additional cluster with varying degrees of admixture from other clusters (Fig. 2B). Neither clustering analysis suggested that individuals on opposite banks of Rio Cuchujaqui or Arroyo Tabelo formed a distinct cluster.
However, individual-based Mantel tests found a significant barrier effect for both Rio
Cuchujaqui (Mantel r=0.1148; p <0.0001) and Arroyo Tabelo (Mantel r=0.1126; p=0.0006) after 10 000 permutations.
A second set of analyses utilized the model that explicitly incorporated prior information for sampling localities to aid in clustering. These results differed somewhat from the clustering results that did not utilize sampling localities. For example, the plot of K versus
Ln Pr(X) never reached a single peak (Fig. S2a) whereas the ΔK method suggested a K of 2
(ΔK = 80.915; Fig. S2b). However, the plot of K versus Ln Pr(X) showed that likelihood values began to stabilize at about K = 3, which was a value similar to that chosen using the
ΔK method in the aspatial analysis. At K = 3, cluster memberships and admixture coefficients were very similar to those of the aspatial analysis.
CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 43
3.4 Landscape genetics
Because the parameterization of resistance surfaces in landscape genetic studies of gene flow can be difficult, especially for categorical landscape variables such as forest cover, I first used Mantel r values to select among relative cost values. For forest cover, I tested four relative cost values (1:2, 1:10, 1:100, 1:1000), in each case assigning a cost value of 1 to
TDF and a higher cost to all other types of land-cover. The same relative cost values were tested on the stream data, with cells containing streams assigned a value of 1 and all other cells a higher cost value. For forest cover, Mantel r values were highest with a cost ratio of
1:2 (Mantel r = 0.4344, p = 0.002) whereas for the stream data the best cost surface was based on a 1:10 ratio (Mantel r = 0.4901, p < 0.001).
The MRM analysis based on least-cost path distances suggested that the incorporation of landscape variables explained significantly more variance in genetic differentiation than a simple IBD model (Table 3). Whereas Euclidean distance was able to explain approximately
44% of the variation in Dest values, the incorporation of landscape variables increased this value to approximately 57%. Because of substantial multicollinearity among some of the predictors, I ran four separate MRM analyses to test the influence of different variables. The most statistically supported model (using AIC weights) was based on a combination of
2 Euclidean distance and the degree of anthropogenic disturbance (R = 0.5725; wi = 0.974).
The least-cost paths based on temperature and slope explained slightly more variation in Dest than Euclidean distance (0.49 and 0.45 versus 0.44, respectively). A model containing
Euclidean distance and forest was also supported through stepwise regression. However, these two distances were highly correlated (Mantel r = 0.995; p<0.001) leading to very large
VIF values. After accounting for Euclidean distance in models, regression coefficients for CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 44 stream connectivity, forest, and undisturbed habitat (anthro model) were negative and, thus, associated with a lower D and higher gene flow.
Results of MRM based on resistance distances calculated in CIRCUITSCAPE were
similar to those based on least-cost path distances (Table 4). However, slope appeared to be an important variable influencing gene flow under a circuit-theoretic approach. The stepwise
MRM analysis suggested a model containing Euclidean distance, slope, and stream connectivity to best explain patterns of differentiation, with the latter variable showing a
negative relationship with Dest. This model explained approximately 20% more variance in
Dest than a pure IBD model (0.6192 vs. 0.4464). The most supported model based on AIC weights included Euclidean distance, slope, stream connectivity, and temperature (wi =
0.521), with stream being the only variable with a significantly negative coefficient. Visual examination of resistance distances based on a composite map of stream networks and slope was highly congruent with the genetic clusters inferred from the STRUCTURE analysis (Fig.
4). In most cases AIC weights were higher for models incorporating landscape variables
versus a model of simple IBD. VIF values were less than 10 for all models suggesting that
multicollinearity among predictors was not likely to be a problem.
4 Discussion
This study highlights the power of adopting a landscape genetics approach to understanding
genetic structure and functional connectivity for tropical organisms. Few landscape genetic
studies focus explicitly on tropical taxa (Storfer et al. 2010). This is problematic, because
tropical regions contain the majority of Earth’s biodiversity (Myers et al. 2000). To
exacerbate this issue, rates of deforestation and habitat fragmentation are increasing CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 45 exponentially throughout these regions and we still know relatively little about how organisms will respond to these continuing threats (Cascante et al. 2002; Burgos & Maass
2004; Miles et al. 2006; Becerra & Venable 2008). My results suggests that if we are to fully comprehend how habitat fragmentation influences functional connectivity in tropical organisms, additional landscape genetic studies are required to infer species-specific responses to continued habitat change.
4.1 Genetic diversity and population structure
Results suggest a moderate to high level of genetic diversity within populations of P.
tuberculosus surrounding the Alamos region. Diversity is not substantially higher in the
population near Alamos versus the other areas, as might be predicted given the tendency for
many gecko species to aggregate in human settlements. Further, a statistically significant
relationship occurs between elevation and allelic richness. These results corroborate field
observations that these lizards are predominantly encountered in TDF habitat. For example,
elevation of the sites ranges from 100m to 500m. This range in elevation spans two distinct
habitat types: tropical thornscrub and TDF (Robichaux & Yetman 2000). An abrupt change
in tropical vegetation occurs at approximately 400m as well as an apparent change in
abundance of geckos; more individuals are encountered per unit of time at El Quintero and
Choquincahui than the other sites.
Significant population structure occurs based on both pairwise differentiation
statistics and STRUCTURE runs. Most pairwise Fst values are high and statistically significant,
with the localities El Quintero and Choquincahui being most divergent. Similar results are
obtained based on values of Dest. Thus, leaf-toed geckos appear to exhibit substantial CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 46 population differentiation over relatively fine spatial scales. Blair et al. (unpublished data) examine the molecular phylogenetics of the P. tuberculosus species group; they report substantial genetic differentiation among populations and suggest that these lizards evolved in concert with Mexico’s TDF. However, few studies examine the population and/or landscape genetic structure of other lineages of geckos. Examining two species of gecko,
Hoehn et al. (2007) report significant differences in genetic diversity and structure. Their result suggests different dispersal abilities in sympatric species, with one species exhibiting a maximum dispersal distance of only 500 m. The results of my landscape genetic analysis of
P. tuberculosus also suggest that many gecko species may have limited dispersal abilities.
The STRUCTURE analyses reveal cryptic population structure. The number of genetic clusters, K, are inferred using two different methods both with and without incorporating prior information on sampling localities. Under the aspatial model, I detect three clusters using the ΔK method and five using Pr(X|K). Incorporating geographic locality information
as a prior changes the inference of K. The ΔK method suggests a K of 2, whereas the plot of
K versus Ln Pr(X) never reaches a peak. Populations at El Quintero and Choquincahui
always group into a single cluster. I never detect structure on opposite banks of the Río
Cuchujaqui or Arroyo Tabelo based on Bayesian clustering (i.e. individuals on opposite
banks were placed in the same cluster). However, Mantel tests did suggest a significant
barrier effect for these features. Although Bayesian clustering methods, and STRUCTURE in particular, can be a powerful tool for inferring recent linear barriers to gene flow (Safner et al. 2011; Blair et al. in press), their high type-1 error rates and difficulty in interpretation require caution when using these methods to test barrier hypotheses. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 47
Genetic clustering is a difficult task and many authors have sought to develop more sophisticated algorithms to achieve this endeavour. Although early algorithms are entirely aspatial in nature (e.g. Pritchard et al. 2000), recent algorithms and programs incorporate prior locality information into the clustering analysis (Guillot et al. 2005a; Chen et al. 2007;
Durand et al. 2009). These powerful spatial genetic models utilize GPS data and free or constrained Voronoi tessellation to cluster individuals into discrete units. In contrast, the new spatial method implemented in STRUCTURE incorporates geographic information by assigning
different codes to different populations (Hubisz et al. 2009). Although relatively
underutilized, this method is appealing in cases like this where I have geographic information
only by population and not by individual. As different results are often encountered when
using different Bayesian clustering programs (Safner et al. 2011; Blair et al. in press),
additional empirical and simulation studies are required to test the power of the spatial
approach in STRUCTURE in comparison to the fully spatial models implemented in other
software packages.
4.2 Landscape genetics
At present, few studies focus on understanding landscape-genetic relationships in
Neotropical vertebrates (Sork & Waits 2010; Storfer et al. 2010). My landscape genetic
analysis identifies several landscape variables important in shaping the genetic connectivity
of leaf-toed geckos. This suggests that the incorporation of landscape variables can explain
significantly more variation in genetic differentiation than a simple IBD model. These
conclusions make intuitive sense when the general biology and life-history of the study organism is considered. For example, Blair et al. (unpublished data) report that the evolution CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 48 of Mexican leaf-toed geckos is tied to the formation of the TDF throughout western Mexico.
Combining this information with qualitative observations of relative abundance, it seems reasonable to conclude that forest fragmentation will have detrimental effects on functional connectivity. Several landscape genetic studies also report a negative relationship between forest fragmentation and genetic connectivity in small vertebrate species (e.g. Spear &
Storfer 2008; Goldberg & Waits 2010), but again, most such studies focus on temperate systems. After controlling for Euclidean distance in my models, the least-cost path results show a negative relationship between forest connectivity and genetic differentiation, suggesting that gene flow is higher through forest patches (Table 3). However, I view these result with caution for two reasons. First, VIF values are exceptionally high for this model and this may be causing large variances in regression coefficients. Second, the study area contains a relatively large amount of undisturbed forest compared to localities in southern
Mexico. Thus, Euclidean distances and least-cost path distances based on forest are nearly identical. I suggest that additional landscape genetic studies are required in areas experiencing rapid loss of TDF in order to understand the effect of forest patch dynamics on functional connectivity.
The next supported model based on least-cost paths includes Euclidean distance and anthropogenic disturbance. Because I parameterize the multivariate anthropogenic resistance surface by assigning higher costs to disturbed areas, the significantly negative regression coefficient suggests that undisturbed habitat is associated with a lower Dest or higher gene
flow. These results are concordant with other recent studies that show a negative relationship
between anthropogenic disturbance and rates of gene flow (e.g. Wofford et al. 2005; Leclerc
et al. 2008; Raeymaekers et al. 2008). This concordance suggests that although geckos are CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 49 frequently encountered in close proximity to human settlements, these areas have a detrimental impact on genetic connectivity. Although geckos are common on abandoned houses in TDF habitat, individuals of P. tuberculosus are only present in and around houses in absence of introduced geckos of the genus Hemidactylus. Very few individuals of P. tuberculosus can be found syntopically with Hemidactylus and on one occasion I witnessed the head of a P. tuberculosus in the jaws of H. frenatus. Thus, in heavily anthropogenically influenced areas, it appears that the introduced Hemidactylus are directly competing against native species of Phyllodactylus. To exacerbate this issue, people often kill leaf-toed geckos on site as they believe the darker colour of these geckos indicates that they are venomous.
Conservation efforts should focus on educating local people on differences between native and non-native flora and fauna to aid in the maintenance and protection of the native species.
This is especially important for species commonly found close to human settlements.
The next most-significant variable in the least-cost models is the minimum temperature of the coldest period, which explains slightly more of the variance in Dest than
Euclidean distance. Neotropical dry forest is a seasonal forest with approximately eight months of warm, wet conditions and four months of dry, cooler conditions (Murphy & Lugo
1986). Because these lizards are commonly encountered in hot tropical lowland environments, I predict that localities experiencing colder temperatures will impede gene
flow. To parameterize the temperature resistance surface I assign higher cost values to cells
experiencing colder temperatures. The MRM results suggest that these geckos are avoiding
areas experiencing colder temperatures. Seasonality and climate are important variables
shaping connectivity for other species (Geffen et al. 2004; Hirao & Kudo 2004, 2008; Banks
et al. 2007; Dionne et al. 2008; Garroway et al. 2008). CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 50
My sampling includes geckos from near bridges and on rocky outcroppings adjacent to streams so I test the hypothesis that stream connectivity is an important component shaping patterns of gene flow. None of the results from MRM based on least-cost distances suggest that gene flow occurs primarily through riparian corridors. Indeed stream networks are often the first variable removed through the stepwise procedure except in the anthropogenic models. However, when stream connectivity is included in a highly supported model, its coefficient is negative; geckos may be following stream corridors. Riparian networks may or may not shape functional connectivity in small terrestrial vertebrates. For example, streams facilitate gene flow among populations of blotched tiger salamander
(Ambystoma tigrinum melanosticum) (Spear et al. 2005) as they do in Rocky Mountain tailed frogs (Ascaphus montanus; Spear & Storfer 2010) and the Pacific jumping mouse (Zapus trinotatus; Vignieri 2005). Conversely, gene flow occurs terrestrially in coastal tailed frogs
(Ascaphus truei) and does not follow riparian corridors (Spear & Storfer 2008). Combined, these results illustrate the utility of a GIS-based landscape genetic approach to understanding the influence of stream networks on genetic connectivity of small terrestrial vertebrates and reaffirm the necessity for examining species-specific processes in landscape genetic studies
(Storfer et al. 2010).
My results based on resistance-distances derived from a circuit theoretic approach are similar to those based on least-cost paths. However, slope becomes a more important predictor of genetic variation in the former models, where populations are separated by higher slopes experiencing lower rates of gene flow (Table 4). These results are similar to numerous other studies that show a direct relationship between topological relief, elevation, and slope on rates and patterns of gene flow in terrestrial vertebrates (e.g. Funk et al. 2005; CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 51
Spear et al. 2005; Pérez-Espona et al. 2008; Murphy et al. 2010). Temperature, stream connectivity, and forest appear as important components based on resistance-distances.
Unlike MRM, resistance-distances show a strong effect of stream connectivity on gene flow.
Circuit theoretic approaches complement least-cost path modeling and often explain more of the variance in genetic differentiation than more traditional methods (McRae & Beier 2007;
McRae et al. 2008). The model with the largest AIC weight based on resistance distances explains more variation in Dest than the least-cost models (0.6462 vs. 0.6161). However,
unlike the least-cost analysis, I am able to combine all resistance distances into the model due
to the lack of multicollinearity among predictors. My results corroborate previous findings in
suggest that a combination of circuit theoretic and least-cost models provides a powerful tool
for investigating functional connectivity in dynamic landscapes.
4.3 Statistical methods in landscape genetics
Landscape genetics is still a relatively new discipline (Manel et al. 2003) and a large number
of recent studies focus on testing the power of various analytical techniques for
understanding the influence of landscape variables on microevolutionary processes
(Cushman et al. 2006; Balkenhol et al. 2009; Cushman & Landguth 2010; Landguth et al.
2010; Safner et al. 2011; Blair et al. in press). Although the Mantel and partial Mantel tests
continue to be the most widely used methods to link landscape and genetic data (Storfer et al.
2010), recent research suggests that these methods suffer from low power and high type-1
errors (Balkenhol et al. 2009; Legendre & Fortin 2010). Recognizing these limitations,
Cushman et al. (2006) and Cushman & Landguth (2010) conclude that partial Mantel tests
implemented in a causal modeling framework can be a powerful tool. However, landscape- CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 52 genetic relationships are often multivariate and are best represented in models that simultaneously consider multiple landscape and environmental variables (Spear et al. 2010).
Thus, a MRM approach can serve as a powerful method to understanding the complex suite of factors important in shaping the spatial distribution of genetic variation (Balkenhol et al.
2009). My study highlights the value of applying MRM analyses to both least-cost and resistance distances. Surprisingly, few studies use this analytical method (Dyer et al. 2010).
Although potentially powerful, MRM approaches have limitations that need to be addressed (Legendre et al. 1994; Lichstein 2007). Multicollinearity often occurs among independent variables because they are in the form of distances and it will often manifest itself in least-cost analyses, with the severity depending on landscape-structure and the distance metric chosen. Although multicollinearity will not affect predictions of the variance of the dependent variable, it may have a consequence on individual regression coefficients because standard errors will be higher. Regression coefficients for predictors can change drastically depending on what other predictors are included in the model. Researchers interested in using MRM should examine the influence of multicollinearity on model results and how regression coefficients change with different models. Adopting rigorous model- selection criteria such as AIC or stepwise regression methods can serve as a means to understanding the best combination of explanatory variables. Although multicollinearity occurs in my data, coefficients change little with different explanatory models (Table 3,4).
Thus, I am confident in my conclusions regarding landscape-genetic relationships. In cases where coefficients change drastically between models, examination of VIF values may be necessary to understand the degree of correlation among the predictors. I recommend that future landscape genetic studies using methods such as MRM and Mantel approaches report CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 53
VIF for each model examined. The application of ridge regression techniques to landscape genetic studies is also an avenue that could be explored.
4.4 Conservation implications for Mexican TDF
Habitat fragmentation and extirpation continue to threaten tropical ecosystems throughout the
globe (Vitousek et al. 1997; Miles et al. 2006). Fragmentation of Mexico’s TDF is of
particular concern as these forests are the country’s predominant vegetation-type and are
known to be a biodiversity hotspot (Ceballos 1995; Robichaux & Yetman 2000; Becerra
2005). Trejo & Dirzo (2000) perform a time-series analysis of Mexico’s TDF and report that
by 1990 only 27% of intact forest remained due to continual anthropogenic conversion for
agriculture and pastureland. At present, the TDF near the Alamos region represents the
country’s most undisturbed tract of continuous forest due, in part, to federally protected
reserves (Robichaux & Yetman 2000). However, even in areas of relatively high forest cover,
slight fragmentation may have detrimental impacts to functional connectivity. Thus, this
study highlights the need for additional landscape genetic studies focusing on TDF
ecosystems to better understand how habitat fragmentation and climate change will influence
ecological and evolutionary processes. To this end, researchers should focus on developing
GIS data sets at finer spatial resolutions that will provide a more comprehensive examination
of the effect of landscape-level processes on the spatial distribution of genetic variation. The
combination of the high resolution landscape layers with highly polymorphic genetic markers
and sophisticated analytical techniques is required if we are to design movement corridors to
fully maximize functional connectivity for species inhabiting this threatened ecosystem.
CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 54
Acknowledgments
I thank Anibal Díaz de la Vega and Martín Gabriel Figueroa Martinez and Stephanie Meyer
for help with field work. For help with GIS and statistical analysis I thank Marie Josée
Fortin, Ilona Naujokaitis-Lewis, and Erika Crispo. Oliver Haddrath, Kristen Choffe, and
Amy Lathrop provided assistance with laboratory work. Funding was provided by the
Natural Sciences and Engineering Research Council of Canada (A3148), the Theodore
Roosevelt Memorial Fund, and the Schad Foundation at the Royal Ontario Museum. All
research was conducted using approved Animal Use Protocols. Collection (FLOR0030,
SGPA/DGFS/047097/09, SGPA/DGVS/02320/10, SGPA/DGVS/02372/11) and export
permits (028, 059) were obtained from SEMARNAT through the Universidad Nacional
Autónoma de México (UNAM).
CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 55
References
Adriaensen F, Chardon JP, De Blust G et al. (2003) The application of ‘least-cost’ modelling
as a functional landscape model. Landscape and Urban Planning, 64, 233–247.
Balkenhol N, Waits LP, Dezzani RJ (2009) Statistical approaches in landscape genetics: an
evaluation of methods for linking landscape and genetic data. Ecography, 32, 818–830.
Banks SC, Piggott MP, Williamson JE, Bove U, Holbrook NJ, Beheregaray LB (2007)
Oceanic variability and coastal topography shape genetic structure in a long-dispersing
sea urchin. Ecology, 88, 3055–3064.
Bartoń K (2009) MUMIN: Multi-Model Inference. R package version 0.12.2/r18, URL
http://R-Forge.R-project.org/projects/mumin/.
Becerra JX (2005) Timing the origin and expansion of the Mexican tropical dry forest.
Proceedings of the National Academy of Sciences, USA, 31, 10919–10923.
Becerra JX, Venable LD (2008) Sources and sinks of diversification and conservation
priorities for the Mexican tropical dry forest PLoS One, 3, e3436.
Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a practical and
powerful approach to multiple testing. Journal of the Royal Statistical Society B, 57, 289–
300.
Blair C, Weigel DE, Balazik M, et al. (in press) A simulation-based evaluation of methods
for inferring linear barriers to gene flow. Molecular Ecology Resources.
Bohonak AJ (1999) Dispersal, gene flow, and population structure. The Quarterly Review of
Biology, 74, 21–45. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 56
Bowne DR, Bowers MA (2004) Interpatch movements in spatially structured populations: a
literature review. Landscape Ecology, 19, 1–20.
Braunisch V, Segelbacher G, Hirzel AH (2010) Modelling functional landscape connectivity
from genetic population structure: a new spatially explicit approach. Molecular Ecology,
19, 3664–3678.
Burgos A, Maass MJ (2004) Vegetation change associated with land-use in tropical dry
forest areas of Western Mexico. Agriculture, Ecosystems, and Environment, 104, 475–
481.
Burnham KP, Anderson DR (2002) Model Selection and Multimodel Inference: A Practical
Information-Theoretic Approach, 2nd ed. Springer-Verlag.
Cascante A, Quesada M, Lobo JJ, Fuchs EA (2002) Effects of dry tropical forest
fragmentation on the reproductive success and genetic structure of the tree Samanea
saman. Conservation Biology, 16, 137–147.
Ceballos G, García A (1995) Conserving neotropical biodiversity: the role of dry forests in
western Mexico. Conservation Biology, 9, 1349–1356.
Chen C, Durand E, Forbes F, François O (2007) Bayesian clustering algorithms ascertaining
spatial population structure: a new computer program and a comparison study. Molecular
Ecology Notes, 7, 747–756.
Clobert J, Danchin E, Dhondt AA, Nichols JD (2001) Dispersal. Oxford University Press,
New York. Pp. 452.
Crawford NG (2010) SMOGD: software for the measurement of genetic diversity. Molecular
Ecology Resources, 10, 556–557. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 57
Crispo E, Moore JS, Lee-Yaw JA, Gray SM, Haller BC (2011) Broken barriers: human-
induced changes to gene flow and introgression in animals. BioEssays, 33, 508–518.
Cushman SA, McKelvey KS, Hayden J, Schwartz MK (2006) Gene flow in complex
landscapes: testing multiple hypotheses with causal modeling. American Naturalist, 168,
486–499.
Cushman SA, Landguth EL (2010) Spurious correlations and inference in landscape genetics.
Molecular Ecology, 19, 3592–3602.
Dieringer D, Schlötterer C (2003) (MSA): a platform independent analysis tool for large
microsatellite data sets. Molecular Ecology Resources, 3, 167–169.
Dionne M, Caron F, Dodson JJ, Bernatchez L (2008) Landscape genetics and hierarchal
genetic structure in Atlantic salmon: the interaction of gene flow and local adaptation.
Molecular Ecology, 17, 2382–2396.
Dixon JR (1964) The systematics and distribution of lizards of the genus Phyllodactylus in
North and Central America. New Mexico State University Research Center, University
Park, New Mexico, Scientific Bulletin, 64–1, 1–139.
Durand E, Jay F, Gaggiotti OE, François O (2009) Spatial inference of admixture proportions
and secondary contact zones. Molecular Biology and Evolution, 26, 1963–1973.
Dyer RJ, Nason, JD, Garrick RC (2010) Landscape modelling of gene flow: improved power
using conditional genetic distance derived from the topology of population networks.
Molecular Ecology, 19, 3746–3759.
Earl DA, vonHoldt BM (2011) STRUCTURE HARVESTER: a website and program for
visualizing STRUCTURE output and implementing the Evanno method. Conservation
Genetics Resources, DOI: 10.1007/s12686-011-9548-7 CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 58
Epps CW, Wehausen JD, Bleich VC et al. (2007) Optimizing dispersal and corridor models
using landscape genetics. Journal of Applied Ecology, 44, 714–724.
Etherington T (2011) Python based GIS tools for landscape genetics: visualising genetic
relatedness and measuring landscape connectivity. Methods in Ecology and Evolution, 2,
52–55.
Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using
the software STRUCTURE: a simulation study. Molecular Ecology, 14, 2611–2620.
Fahrig L (1997) Relative effects of habitat loss and fragmentation on population extinction.
Journal of Wildlife Management, 61, 603–610.
Fahrig L (1998) When does fragmentation of breeding habitat affect population survival?
Ecological Modeling, 105, 273–292.
Fahrig L (2003) Effects of habitat fragmentation on biodiversity. Annual Review of Ecology,
Evolution, and Systematics, 34, 487–515.
Fahrig L (2007) Non-optimal animal movement in human-altered landscapes. Functional
Ecology, 21, 1003–1015.
Falush D, Stephens M, Pritchard JK (2003) Inference of population structure using
multilocus genotype data: linked loci and correlated allele frequencies. Genetics, 164,
1567–1587.
Frankham R, Ballou JD, Briscoe DA (2002) Introduction to Conservation Genetics. New
York, Cambridge University Press, 617 p.
Funk WC, Blouin MS, Corn PS et al. (2005) Population structure of Columbia spotted frogs
(Rana luteiventris) is strongly affected by the landscape. Molecular Ecology, 14, 483–
496. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 59
García A (2006) Using ecological niche modelling to identify diversity hotspots for the
herpetofauna of Pacific lowlands and adjacent interior valleys of Mexico. Biological
Conservation, 130, 25–46.
Garroway CJ, Bowman J, Carr D, Wilson PJ (2008) Application of graph theory to landscape
genetics. Evolutionary Applications, 1, 620–630.
Geffen E, Anderson MJ, Wayne RK (2004) Climate and habitat barriers to dispersal in the
highly mobile grey wolf. Molecular Ecology, 13, 2481–2490.
Gerlach G, Jueterbock A, Kraemer P, Deppermann J, Harmand P (2010) Calculations of
population differentiation based on GST and D: forget GST but not all of statistics!
Molecular Ecology, 19, 3845–3852.
Goldberg CS, Waits LP (2010) Comparative landscape genetics of two pond-breeding
amphibian species in a highly modified agricultural landscape. Molecular Ecology, 19,
3650–3663.
Goslee S, Urban D (2008) Dissimilarity-Based Functions for Ecological Analysis. CRAN
repository. Available from http://cran.r-project.org/web/packages/ecodist/
Guillot G, Estoup A, Mortier F et al. (2005a) A spatial statistical model for landscape
genetics. Genetics, 170, 1261–1280.
Guillot G, Mortier F, Estoup A (2005b) GENELAND: a program for landscape genetics.
Molecular Ecology Notes, 5, 712–715.
Guillot G, Leblois R, Coulon A, Frantz A (2009) Statistical methods in spatial genetics.
Molecular Ecology, 18, 4734–4756.
Hedrick PW (2005) A standardized genetic differentiation measure. Evolution, 59, 1633–
1638. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 60
Heller R, Siegismund HR (2009) Relationship between three measures of genetic
differentiation GST, DEST and G’ST: how wrong have we been? Molecular Ecology, 18,
2080–2083.
Hirao AS, Kudo G (2004) Landscape genetics of alpine snowbed plants: comparisons along
geographic and snowmelt gradients. Heredity, 93, 290–298.
Hirao AS, Kudo G (2008) The effect of segregation of flowering time on fine-scale spatial
genetic structure in an alpine-snowbed herb Primula cuneifolia. Heredity, 100, 424–430.
Hoehn M, Sarre SD, Henle K (2007) The tales of two geckos: does dispersal prevent
extinction in recently fragmented populations? Molecular Ecology, 16, 3299–3312.
Holderegger R, Wagner HH (2008) Landscape genetics. BioScience, 58, 199–207.
Hubisz MJ, Falush D, Stephens M, Pritchard JK (2009) Inferring weak population structure
with this assistance of sample group information. Molecular Ecology Resources, 9, 1322–
1332.
Jakobsson M, Rosenberg NA (2007) CLUMPP: a cluster matching and permutation program
for dealing with label switching and multimodality in analysis of population structure.
Bioinformatics, 23, 1801–1806.
Jost L (2008) GST and its relatives do not measure differentiation. Molecular Ecology, 17,
4015–4026.
Jost L (2009) D vs. GST: response to Heller and Siegismund (2009) and Ryman and Leimar
(2009). Molecular Ecology, 18, 2088–2091.
Kullback S, Leibler RA (1951) On information and sufficiency. Annals of Mathematical
Statistics, 22, 79–86. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 61
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes. The American Naturalist, 142, 911–927.
Landguth EL, Cushman SA, Schwartz MK, McKelvey KS, Murphy M, Luikart G (2010)
Quantifying the lag time to detect barriers in landscape genetics. Molecular Ecology, 19,
4179–4191.
Leclerc E, Mailhot Y, Mingelbier M, Bernatchez L (2008) The landscape genetics of yellow
perch (Perca flaviscens) in a large fluvial ecosystem. Molecular Ecology, 17, 1702–1717.
Legendre P, Fortin MJ (2010) Comparison of the Mantel test and alternative approaches for
detecting complex multivariate relationships in the spatial analysis of genetic data.
Molecular Ecology Resources, 10, 831–844.
Legendre P, Lapointe FJ, Casgrain P (1994) Modeling brain evolution from behaviour: a
permutational regression approach. Evolution, 48, 1487–1499.
Lichstein JW (2007) Multiple regression on distance matrices: a multivariate spatial analysis
tool. Plant Ecology, 188, 117–131.
Manel S, Schwartz MK, Luikart G, Taberlet P (2003) Landscape genetics: combining
landscape ecology and population genetics. Trends in Ecology and Evolution, 18, 189–
197.
McRae BH (2006) Isolation by resistance. Evolution, 60, 1551–1561.
McRae BH, Beier P (2007) Circuit theory predicts gene flow in plant and animal populations.
Proceedings of the National Academy of Sciences, USA, 104, 19885–19890.
McRae BH, Dickson BG, Keitt TH, Shah VB (2008) Using circuit theory to model
connectivity in ecology and conservation. Ecology, 10, 2712–2724. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 62
Meirmans PG, Hedrick PW (2011) Assessing population structure: FST and related measures.
Molecular Ecology Resources, 11, 5–18.
Michels E, Cottenie K, Neys L, De Gelas K, Coppin P, De Meester L (2001) Geographical
and genetic distances among zooplankton populations in a set of interconnected ponds: a
plea for using GIS modelling of the effective geographical distance. Molecular Ecology,
10, 1929–1938.
Miles L, Newton AC, DeFries RS et al. (2006) A global overview of the conservation status
of tropical dry forests. Journal of Biogeography, 33, 491–505.
Miller MP (1997) Tools for Population Genetic Analyses (TFPGA). A Windows program for
the analysis of allozyme and molecular population genetic data. Distributed by author.
Murphy PG, Lugo AE (1986) Ecology of tropical dry forest. Annual Review of Ecology and
Systematics, 17, 67–88.
Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity
hotspots for conservation priorities. Nature, 409, 853–858.
Neel MC (2008) Patch connectivity and genetic diversity conservation in the federally
endangered and narrowly endemic plant species Astragalus albens (Fabaceae). Biological
Conservation, 141, 938–955.
Pérez-Espona S, Perez-Barberia FJ, McLeod JE et al. (2008) Landscape features affect gene
flow of Scottish Highland red deer (Cervus elaphus). Molecular Ecology, 17, 981–996.
Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using
multilocus genotype data. Genetics, 155, 945–959. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 63
R Development Core Team (2008) R: a language and environment for statistical computing.
R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL
http://www.R-project.org.
Raeymaekers JAM, Maes GE, Geldof S et al. (2008) Modeling genetic connectivity in
sticklebacks as a guideline for river restoration. Evolutionary Applications, 1, 475–488.
Rayfield B, Fortin MJ, Fall A (2010) The sensitivity of least-cost habitat graphs to relative
cost surface values. Landscape Ecology, 25, 519–532.
Raymond M, Rousset F (1995) GENEPOP (version 1.2): population genetics software for
exact tests and ecumenicism. Journal of Heredity, 86, 248–249.
Ricketts TH (2001) The matrix matters: effective isolation in fragmented landscapes. The
American Naturalist, 158, 87–99.
Robichaux RH, Yetman DA (2000) The tropical deciduous forest of Alamos. Biodiversity of
a threatened ecosystem in Mexico. The University of Arizona Press.
Rosenberg NA (2007) DISTRUCT: a program for the graphical display of population structure.
Molecular Ecology Notes, 4, 137–138.
Rousset F (2008) GENEPOP ‘007: a complete reimplementation of the GENEPOP software for
Windows and Linux. Molecular Ecology Resources, 8, 103–106.
Ryman N, Leimar O (2009) GST is still a useful measure of genetic differentiation—a
comment on Jost’s D. Molecular Ecology, 18, 2084–2087.
Safner T, Miller MP, McRae BH, Fortin MJ, Manel S (2011) Comparison of Bayesian
clustering and edge detection methods for inferring boundaries in landscape genetics.
International Journal of Molecular Sciences, 12, 865–889. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 64
Shah VB, McRae BH (2008) CIRCUITSCAPE: a tool for landscape ecology. In: Varoquaux G,
Vaught T, Millman J (eds). Proceedings of the 7th Python in Science Conference (SciPy
2008), pp. 62–66.
Sigala-Rodríguez JJ, Greene HW (2009) Landscape change and conservation priorities:
Mexican herpetofaunal perspectives at local and regional scales. Revista Mexicana de
Biodiversidad, 80, 231–240.
Slatkin M (1987) Gene flow and the geographic structure of natural populations. Science,
236, 787–792.
Smouse PE, Long JC, Sokal RR (1986) Multiple regression and correlation extensions of the
Mantel test of matrix correspondence. Systematic Zoology, 35, 627–632.
Sork VL, Waits LP (2010) Contributions of landscape genetics—approaches, insights, and
future potential. Molecular Ecology, 19, 3489–3495.
Spear SF, Storfer A (2008) Landscape genetic structure of coastal tailed frogs (Ascaphus
truei) in protected vs. managed forests. Molecular Ecology, 17, 4642–4656.
Spear SF, Storfer A (2010) Anthropogenic and natural disturbance lead to differing patterns
of gene flow in the Rocky Mountain tailed frog, Ascaphus montanus. Biological
Conservation, 143, 778–786.
Spear SF, Peterson CR, Matocq M, Storfer A (2005) Landscape genetics of the blotched tiger
salamander (Ambystoma tigrinum melanostictum). Molecular Ecology, 14, 25553–2564.
Storfer A, Murphy MA, Evans JS et al. (2007) Putting the ‘landscape’ in landscape genetics.
Heredity, 98, 128–142.
Storfer A, Murphy MA, Spear SP, Holderegger R, Waits LP (2010) Landscape genetics:
where are we now? Molecular Ecology, 19, 3496–3514. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 65
Trejo I, Dirzo R (2000) Deforestation of seasonally dry tropical forest: a national and local
analysis in Mexico. Biological Conservation, 94, 133–142.
Vignieri SN (2005) Streams over mountains: influence of riparian connectivity on gene flow
in the Pacific jumping mouse (Zapus trinotatus). Molecular Ecology, 14, 1925–1937.
Vilá C, Sundqvist AK, Flagstad O et al. (2002) Rescue of a severely bottlenecked wolf
(Canis lupus) population by a single immigrant. Proceedings of the Royal Society of
London B, 270, 91–97.
Vitousek PM, Mooney HA, Lubchenco J, Melillo JM (1997) Human domination of Earth’s
ecosystems. Science, 277, 494–499.
Wang IJ, Savage WK, Shaffer BH (2009) Landscape genetics and least-cost path analysis
reveal unexpected dispersal routes in the California tiger salamander (Ambystoma
californiense). Molecular Ecology, 18, 1365–1374.
Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population
structure. Evolution, 38, 1358–1370.
Whitlock MC (2011) G’ST and D do not replace FST. Molecular Ecology, 20, 1083–1091.
Wilson LD, Townsend JH, Johnson JD (2010) Conservation of Mesoamerican Amphibians
and Reptiles. Eagle Mountain Publishing, LC, Eagle Mountain, Utah, 812pp.
With KA, Garder RH, Turner MG (1997) Landscape connectivity and population
distributions in heterogeneous environments. Oikos, 78, 151–169.
Wofford JEB, Gresswell RE, Banks MA (2005) Influence of barriers to movement on within-
watershed genetic variation of coastal cutthroat trout. Ecological Applications, 15, 628–
637.
Wright S (1943) Isolation by distance. Genetics, 28, 139–156. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 66
Table 1. Characterization of genetic diversity of Phyllodactylus tuberculosus at each sampling site included in this study based on data from 10 microsatellite loci.
Population Elevation (m) n n(msat) HO HE # alleles allelic richness
Road to Navojoa 455 10 10 0.717 0.7547 65 65.0000 Alamos 375 36 35.2 0.7615 0.7725 88 64.8873 Tabelo A 167 27 26.3 0.7007 0.7143 70 56.5553 Tabelo B 199 30 29 0.7213 0.734 69 54.2341 Aduana 497 17 16.5 0.7562 0.7799 75 65.6058 Rio Cuchujaqui A 358 30 28.9 0.7487 0.7619 86 65.4514 Rio Cuchujaqui B 261 42 41.3 0.7717 0.7812 83 64.5302 Sierrita 483 38 36.4 0.6976 0.7073 89 63.8813 Mocuzari 124 30 29.6 0.7005 0.7125 70 55.2195 El Quintero 361 30 29 0.6903 0.7025 83 61.7606 Choquincahui (El Cobre) 433 31 29.2 0.7089 0.7214 81 59.3397 San Antonio 388 15 13.6 0.7242 0.7518 69 62.7309 n= number of individuals, n(msat) = number of individuals accounting for missing data, HO = observed heterozygosity, HE = expected heterozygosity; allelic richness calculated based on population with smallest sample size (n = 10). CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 67
Table 2 Pairwise genetic differentiation between populations estimated from 10 microsatellite loci. Values above diagonal represent Dest and values below diagonal FST. Bold values of FST indicate significance (P < 0.05) after Bonferroni correction.
Road to Navojoa Alamos Tabelo A Tabelo B Aduana Cuch A Cuch B La Sierrita Mocuzari El Quintero Choquincahui San Antonio Rd to Navojoa -- 0.106 0.181 0.174 0.000 0.119 0.075 0.048 0.094 0.325 0.311 0.113 Alamos 0.032 -- 0.197 0.166 0.072 0.086 0.095 0.139 0.151 0.217 0.309 0.032 Tabelo A 0.080 0.080 -- 0.050 0.125 0.196 0.223 0.257 0.133 0.260 0.309 0.248 Tabelo B 0.065 0.059 0.026 -- 0.143 0.224 0.232 0.252 0.100 0.293 0.377 0.197 Aduana ‐0.003 0.021 0.058 0.059 -- 0.074 0.084 0.069 0.123 0.207 0.269 0.083 Cuch A 0.039 0.028 0.082 0.076 0.024 -- 0.059 0.131 0.150 0.170 0.195 0.134 Cuch B 0.033 0.029 0.087 0.075 0.031 0.021 -- 0.137 0.197 0.236 0.319 0.101 La Sierrita 0.042 0.056 0.154 0.123 0.047 0.078 0.069 -- 0.176 0.345 0.378 0.155 Mocuzari 0.045 0.063 0.068 0.049 0.056 0.055 0.070 0.097 -- 0.267 0.318 0.172 El Quintero 0.149 0.118 0.147 0.151 0.128 0.096 0.111 0.201 0.150 -- 0.009 0.217 Choquincahui 0.140 0.122 0.142 0.150 0.125 0.092 0.114 0.199 0.142 0.008 -- 0.336 San Antonio 0.040 0.019 0.099 0.078 0.027 0.040 0.042 0.069 0.083 0.135 0.144 -- CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 68
Table 3 Multiple regression on distance matrices (MRM) results showing the relationship between pairwise genetic distance (linearized Dest) and least-cost path distances incorporating landscape heterogeneity. Four different MRM analyses were performed (models B–D) using a stepwise approach with a p-to-enter value <0.05. Models in italics are those selected with only significant regression coefficients. Optimal cost values used to parameterize resistance surfaces prior to calculating each least-cost path were selected based on Mantel r correlation coefficients. VIF=Variance Inflation Factor with values above 10 in boldface.
2 Model Variables β P Model R P VIF Model AICc ∆AIC Akaike Weight (wi) A1 Euclidean 8.49E-06 0.0001 0.444 0.0001 -97.46 20.34 0.0000453 A2 Stream 1.08E-06 0.01 0.2397 0.01 -76.8 40.48 1.48E-09
B1 Anthro 1.92E-07 0.0037 0.2631 0.0037 -78.86 38.42 4.15E-09 B2 Euclidean 9.20E-06 0.0063 0.4461 0.0015 2.53 -95.43 0.0000164 Stream -1.60E-07 0.7654 B3 Euclidean 2.17E-05 0.0017 0.5725 0.0002 9.39 -112.53 4.75 0.0849 Anthro -4.12E-07 0.0422 B4 Euclidean 2.87E-05 0.0005 0.6161 0.0001 16.35 -117.28 0 0.913 Anthro -5.18E-07 0.0055 11.24 Stream -8.03E-07 0.1447 3.03
C1 Temp 2.59E-07 0.0001 0.4889 0.0001 -103.01 14.27 0.000727 C2 Temp 3.48E-07 0.1257 0.4922 0.0001 18.3 -101.17 16.11 0.00029 Euclidean -3.14E-06 0.6983 C3 Temp 3.15E-07 0.0006 0.5018 0.0006 2.76 -102.43 14.85 0.000544 Stream -4.17E-07 0.4297
D1 Slope 3.33E-07 0.0002 0.4541 0.0002 -98.66 18.62 0.0000826 D2 Slope 2.66E-07 0.4414 0.4548 0.0004 27 -96.47 20.81 0.0000276 Euclidean 1.76E-06 0.8493 D3 Slope 3.96E-07 0.0038 0.4626 0.0015 2.87 -97.43 19.85 0.0000447 CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 69
Stream -3.46E-07 0.5382
E1 Forest 4.41E-07 0.0001 0.4093 0.0001 -93.46 23.82 0.00000614 E2 Euclidean 3.85E-05 0.014 0.4993 0.0004 101.76 -102.09 15.19 0.000459 Forest -1.63E-06 0.0398 E3 Euclidean 3.87E-05 0.0161 0.5 0.0001 102 -99.85 17.43 0.00015 Forest -1.62E-06 0.0501 102.33 Stream -9.91E-08 0.8517 2.54 CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 70
Table 4 Multiple regression on distance matrices (MRM) results showing the relationship between pairwise genetic distance (linearized Dest) and resistance distances incorporating landscape heterogeneity. Optimal cost values used to parameterize resistance surfaces prior to calculating resistance distances were selected based on Mantel r correlation coefficients. For clarity, only models with significant support are shown. VIF=Variance Inflation Factor.
2 Model Variables β P Model R P VIF Model AICc ∆AIC Akaike Weight (wi) A1 Euclidean 8.84E-06 0.0003 0.4464 0.0003 -74.68 17.39 0.0000873
C4 Euclidean 1.09E-05 0.001 0.6192 0.001 3.22 -90.5 1.57 0.238 Slope 5.31E-03 0.0147 1.78 Stream -2.87E-02 0.0352 4.15
D2 Euclidean 1.20E-05 0.0022 0.6462 0.0011 3.48 -92.07 0 0.521 Slope 2.17E-03 0.5584 5.62 Stream -3.74E-02 0.0204 5.12 Temp 4.84E-03 0.2765 6.75
D3 Euclidean 1.11E-05 0.0016 0.6287 0.0006 3.23 -89.36 2.71 0.135 Slope 5.54E-03 0.0147 1.83 Stream -4.09E-02 0.1127 9.61 Forest 5.21E-03 0.5212 5.44 CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 71
Figure Legends
Fig. 1. Sampling sites for all individuals and populations of Phyllodactylus tuberculosus included in this study. Darker shades of gray represent tropical dry forest. Dark lines represent streams and arroyos throughout the study area.
Fig. 2. Relationship between elevation and genetic diversity (allelic richness) for all populations of Phyllodactylus tuberculosus included in this study.
Fig. 3. STRUCTURE results excluding prior information on sampling locality for
Phyllodactylus tuberculosus included in this study. A) Results for K=3 selected using the ΔK
method. B) Results for K=5 selected using Pr(X|K)
Fig. 4. Results from CIRCUITSCAPE depicting cumulative currents between populations of
Phyllodactylus tuberculosus based on the combined effect of slope and riparian connectivity.
‘High’ in figure legend represents high levels of current whereas ‘Low’ represents low levels of current. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 72
Fig. 1
# Mocuzari
## Arroyo Tabelo A,B
# Arroyo# Aduana Road to Navojoa #Alamos # La Sierrita Choquincahui# (El Cobre) # Rio Cuchujaqui A,B # El Quintero # San Antonio
010205 Kilometers CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 73 Fig. 2
70.0000
65.0000
60.0000
55.0000 Allelic Richness
50.0000 y = 0.0255x + 52.884 R2 = 0.5836 45.0000 0 100 200 300 400 500 600 Elevation (m) CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 74
Fig. 3
A K=3 ∆K
i i s a a o o N A B A B t r u i o n i i i a r R o o e h n l l a u u r z t m e e q q r a o a u e u n c t l b b d a a i c i a a j j n n A S o u i / T T A u u A h h M Q u e n s c c q u u o a u S o C C h H C
B K=5 Pr(X|K)
i i s a a o o N A B A B t r u i o n i i i a r R o o e h n l l a u u r z t m e e q q r a o a u e u n c t l b b d a a i c i a a j j n n A S o u i / T T A u u A h h M Q u e n s c c q u u o a u S o C C h H C
CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 75
Fig. 4
# Mocuzari
# Arroyo Tabelo A,B
Arroyo Aduana # # Road to Navojoa # Alamos # Choquincahui (El Cobre) La Sierrita # ## # Rio Cuchujaqui A,B El Quintero # San Antonio
Legend Value High : 12.3473 010205 Kilometers
Low : 9e-006 CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 76
Supplementary Material
Fig. S1. A) STRUCTURE results illustrating changes in Pr(X|K) under the aspatial model. B)
STRUCTURE results based on the second order rate of change (ΔK method) under the aspatial model. For each K, 10 independent simulations were performed.
Fig. S2. A) STRUCTURE results illustrating changes in Pr(X|K) under the spatial model. B)
STRUCTURE results based on the second order rate of change (ΔK method) under the spatial model. For each K, 10 independent simulations were performed. CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 77
Fig. S1a
Fig. S1b CHAPTER 3—LANDSCAPE GENETICS OF P. TUBERCULOSUS 78
Fig. S2a
Fig. S2b
CHAPTER 4
Genes, geography, climate and cryptic diversity in leaf-toed
geckos (Phyllodactylidae: Phyllodactylus) illuminate
evolutionary processes throughout the Mexican dry forest
Abstract
Neotropical dry forests (TDF) are among the planet’s most diverse ecosystems. However, in comparison to moist tropical forests, little is known about the mode and tempo of biotic evolution throughout this threatened biome. Mitochondrial and nuclear DNA sequence data are used to investigate the drivers of diversification in leaf-toed geckos (Phyllodactylus) of
Mexico’s dry forests. Speciation and substantial, cryptic molecular diversity seems to have originated as the forest was forming due to orogenesis of the Sierra Madre Occidental and
Mexican Volcanic Belt in western Mexico. Multilocus species-tree and diversification analyses suggest that lineage accumulation was low and consistent until ~2.6 million years ago when diversification rates increased at an exponential rate. Key vicariant events play important roles in driving the biogeographic history of these lineages and southwestern
Mexico is suggested to be an area for future diversification. Estimates of historical migration rates suggest that this extraordinary diversity may also be driven by limited dispersal abilities. Multilocus tests of demographic expansion indicate weak evidence for increases in population size since the Last Glacial Maximum. These results suggest that the formation of
79 CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 80
TDF vegetation in the Miocene combined with Pleistocene climate shifts affected the evolution of leaf-toed geckos throughout western Mexico.
1 Introduction
The emergence of next generation DNA sequencing techniques and novel statistical methods are revolutionizing molecular phylogenetics and genealogical inference (Brito & Edwards
2009; Edwards 2009; Blair & Murphy 2011). New methods of phylogenetic inference account for gene-tree discordance (Liu 2008, Kubatko et al. 2009; Heled & Drummond
2010) and use more realistic models to estimate the timing of lineage divergence (Drummond et al. 2006). Multilocus algorithms have also been recently developed to estimate demographic history (Heled & Drummond 2008), reconstruct ancestral distributions (Yu et al. 2010), and estimate effective population sizes and asymmetric migration rates (Beerli &
Felsenstein 2001; Beerli 2006). These new statistical methods formulate and test hypotheses pertaining to the historical forces shaping the evolutionary histories of species and populations.
The tropical dry forests (TDF) of the Neotropics are home to an extraordinary diversity of species (Ceballos & García 1995; Gillespie et al. 2000; García 2006). Unlike tropical rainforests, the TDF is characterized by a distinct seasonality with up to 8 months of dry, arid-like conditions separated by 4 months of deluge (Murphy & Lugo 1986). Because of this seasonality, many plants lose their leaves in a similar fashion to the vegetation of temperate deciduous forests. It is hypothesized that the TDF of Mexico originated sometime between 30–20 million years ago (Ma) due to the further orogenesis of two of México’s major mountain systems: the Sierra Madre Occidental and the Mexican Volcanic Belt (MVB; CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 81
Becerra 2005). These mountain chains effectively block colder northern temperatures allowing the TDF to expand throughout much of western Mexico and Middle America.
Although Neotropical dry forests are a diversity hotspot (Myers et al. 2000), relatively little is known about the historical forces that drive and shape patterns of genetic diversity throughout this ecosystem (Werneck et al. 2011). Recent empirical analyses conflict as to the role Pleistocene climatic shifts played in driving lineage divergence in lowland Neotropical species (Klicka & Zink 1997; Weir & Schluter 2004; Zink et al. 2004;
Weir 2006; Zarza et al. 2008). Some discrepancy may be due to the different statistical methods employed, yet these results also suggest that additional data are required to test for the influence of palaeoclimatic shifts on rates of cladogenesis. Alternatively, the regions dynamic orogenesis and diverse landscape also may have played a vital role in speciation and lineage divergence (Devitt 2006; Mulcahy et al. 2008; Zarza et al. 2008).
How TDFs responded to the climatic fluctuations of the Pleistocene remains uncertain. Early studies suggest that during the Pleistocene TDF encompassed a much broader distribution that was subsequently fragmented as temperature and precipitation levels increased during the Pleistocene-Holocene transition (Pennington et al. 2000). This hypothesis partly explains the disjunct distributions of several plant and animal taxa restricted to these forests. Conversely, recent evidence from palaeodistribution modeling suggests that the distribution of TDF was much more disjunct during the Last Glacial
Maximum than present-day (Werneck et al. 2011). Regardless of the temporal dynamics of
TDF vegetation during the Pleistocene, it is likely that genetic signatures dating to this time period are present in species largely restricted to this ecosystem. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 82
Recent molecular phylogenetic and phylogeographic studies in Mexico focus on patterns and processes of species inhabiting highland pine and pine-oak habitats and suggest a dual effect of Neogene vicariance and Pleistocene climate change (Bryson et al.
2011a,b,c,d). The few empirical studies that focus on genetic patterns of species throughout the lowlands of western Mexico report several cryptic lineages (Hasbún et al. 2005; Devitt
2006; Zarza et al. 2008), many of which correspond to major biogeographical provinces within Mexico. Additional studies are warranted to adequately assess regional patterns of biodiversity, especially at a time when the TDF is severely threatened due to burnings, agriculture, urbanization, and selective logging (Trejo & Dirzo 2000; Burgos & Maass 2004).
Analyses of spatial distribution of cryptic diversity and hypothesis testing can illuminate palaeoclimatic, geologic, and geographic histories within Mexico’s TDF (sensu Che et al.
2010).
Mitochondrial DNA (mtDNA) continues to be the marker of choice for examining spatial patterns of lineage divergence in closely related groups. However, the validity of the approach for extrapolating population history is questioned (Barrowclough & Zink 2009;
Edwards & Bensch 2009). Some evidence suggests that a relatively large number of nuclear loci are required to obtain the same level of signal found with mtDNA (Lee & Edwards
2008). Further, analyses based on mtDNA resolve matrilineal histories only, whereas nuclear
DNA (nDNA) provides evidence of gene flow including paternal contributions to evolutionary history. It is also well-known that the slower rate of evolution and larger effective population sizes of nuclear versus mitochondrial markers result in longer coalescence times with the former (Degnan & Rosenberg 2006,2009). This realization has spawned a rapid growth of theory and computational methods aimed at estimating a species- CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 83 or population-tree versus a gene-tree (Edwards & Beerli 2000; Edwards 2009; Knowles
2009). Although these methods are particularly powerful at estimating phylogenetic relationships among recently diverged groups, few studies have utilized these techniques in a hypothesis-testing framework in Neotropical lowland taxa.
Leaf-toed geckos, genus Phyllodactylus, have a largely Neotropical distribution, ranging from southern California, the peninsula of Baja California, and the Sea of Cortes, to western lowlands of Mexico, Middle and South America, northeastern South America, and the West Indies (Dixon 1964). They occur in arid and semi-arid regions and are commonly found in tropical thornscrub and TDF; they avoid rainforests. The species are ideal for testing hypotheses pertaining to historical diversification within the Mexican TDF. Like many other geckos, Phyllodactylus appear to be microhabitat specialists, and are commonly found on rocky outcroppings adjacent to drainages. Little molecular work exists for these lizards (Blair et al. 2009). Thus, we still know relatively little about the evolutionary history of this widespread group.
Recent research demonstrates the utility of employing molecular phylogenetic analyses to understand historical processes and biotic evolution in highly diverse regions
(e.g. Che et al. 2010; Fujita et al. 2010). Thus, my primary goal is to use multilocus sequence data to elucidate the drivers of cladogenesis in leaf-toed geckos distributed throughout the
Mexican TDF. I test hypotheses pertaining to the influence of palaeoclimate, geology, geography, and landscape configuration on rates and patterns of cladogenesis for the P. tuberculosus species group using phylogenetic, biogeographic, demographic, molecular dating, and multilocus coalescent techniques. I am particularly interested in understanding CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 84 the relative roles of Pleistocene climate change and the origin of the Mexican TDF on rates and patterns of lineage diversification.
2 Materials and Methods
2.1 Study area and sampling
From 2008 to 2010 I sampled approximately 150 individuals of the P. tuberculosus group (P.
tuberculosus, P. lanei, P. muralis) from distant sites covering most of the Mexican ranges for
each species (Fig. 1). When possible, I obtained sample sizes large enough to capture
intrapopulation diversity. Tissue samples (tail tips) were collected in the field and directly preserved in 95% ethanol. Detailed sampling locality information was provided in
Supplementary Appendix I. I used Phyllodactylus xanti and Coleonyx variegatus as outgroup taxa.
2.2 Laboratory methods
Total genomic DNA was extracted using standard phenol-chloroform procedures following
initial digestion with proteinase K. I then amplified two mitochondrial genes (NADH
dehydrogenase subunit 4 (ND4) and cytochrome b (cyt b) following the polymerase chain
reaction (PCR) protocols described previously (Blair et al. 2009). Because I was also
interested in employing new multilocus analytical methods, I amplified and sequenced two
rapidly evolving nuclear introns (alpha-enolase and lamin-A) previously used for other
vertebrates (Friesen et al. 1997, 1999; Zarza et al. 2008). Primer sequences and annealing CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 85 temperatures used for each of the four loci were summarized in Table 1. Following amplification, each PCR product was separated on a 1% agarose gel stained with ethidium bromide. Bands were then excised and directly sequenced in both directions using the
BigDye Terminator Cycle Sequencing Kit (Applied Biosystems). Sequences were then read on an ABI 3730 automated sequencer (Applied Biosystems) and protocols followed the manufacturer’s recommendations.
2.3 Sequence alignment and haplotype estimation
Sequences were edited and assembled using BIOEDIT v7.05 (Hall 1999). Sequence identity
was confirmed using a BLASTN search. Mitochondrial genes were translated into their
corresponding amino acid sequence to check for premature stop codons which would suggest
a nuclear pseudogene. Next, all sequences were aligned manually in BIOEDIT. Haplotypes
from heterozygous individuals in the intron data were identified using PHASE V2.1 (Stephens
et al. 2001; Stephens & Donnelly 2003) with the default parameters and a threshold value of
0.90. Input files for PHASE were generated using SEQPHASE (Flot 2010). All unique
sequences were deposited in GenBank (accession numbers here upon acceptance).
2.4 Recombination
Although the inclusion of nDNA sequences provided a more comprehensive analysis of
evolutionary history, recombinant sequences have been shown to be a serious concern for subsequent phylogenetic and demographic analyses (Posada 2001, 2002; Arenas & Posada
2010). I used RDP3 (Martin et al. 2010) to test for the presence of recombination in the two CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 86 nuclear intron data sets. Because the data set was relatively large, I only implemented the RDP
(Martin & Rybicki 2000), GENECONV (Padidam et al. 1999), and MAXCHI (Maynard Smith
1992) methods as per recommendation of the program’s authors using the default parameters.
2.5 Gene-tree reconstruction
I used RAxML v7.2.6 (Stamatakis 2006) run through RAxMLGUI 0.93 to estimate
genealogical (microevolutionary) and phylogenetic (macroevolutionary) relationships for the
species group under a maximum likelihood (ML) framework. I performed several different
analyses on different data sets—mtDNA only, alpha-enolase only, lamin-A only, and
concatenated data—specifying an independent GTRGAMMA model for each data set. For
ease of visualization I only included unique sequences from the phased intron data sets. A
single phased haplotype from each individual of the intron data was concatenated with the
corresponding mtDNA sequence. Each analysis started from a maximum parsimony tree and
implemented a full ML search with 10 000 bootstrap replicates (BS) using the rapid bootstrap
algorithm (Stamatakis et al. 2008). Phyllodactylus xanti was used as the outgroup to root
each topology. I accepted BS values greater than 70 as strong support for a given node (Hillis
and Bull 1993).
2.6 Species-tree reconstruction and divergence times
Divergence dating in geckos can be difficult due to both the poor fossil record and the taxonomic uncertainty surrounding the fossils that are present (Grimaldi 1995; Kluge 1995;
Fujita & Papenfuss 2011). However, I was interested in gaining a broad understanding of the CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 87 tempo of lineage divergence in leaf-toed geckos and thus, I relied on previous estimates of mtDNA substitution rate in geckos (see below). Calibrating phylogenetic trees based on independently-derived mutation rates can be robust, particularly if a parametric distribution prior is applied to the rate (Ho & Phillips 2009). Because incomplete lineage sorting can be ubiquitous at and below the species level, I performed a species-tree analysis for the P. tuberculosus group using the coalescent model in the program *BEAST V.1.7.1 (Heled &
Drummond 2010; Drummond et al. 2012). Of the rapidly growing number of species-tree
methods available, *BEAST is one of the few algorithms that can simultaneously estimate individual gene-trees, the species-tree, population sizes, and divergence times under a
Bayesian framework. For the analysis, I defined a ‘species’ based on previous species and
subspecies designations, sampling locality, and results from the calculated mtDNA gene-tree
topology; an approach similar to Fujita et al. (2010). For computational efficiency I pruned
data for each individual gene to 3 individuals per species which has been shown to be
adequate for estimating species-trees (Heled & Drummond 2010; McCormack et al. 2011). I
created a text file defining traits for each species and used BEAUTI v.1.7.1 to create the input file for *BEAST. I used BEAUTI to define two species sets; one constraining the monophyly of
Phyllodactylus and the other based on the root. I specified a GTR+I+G model for the mtDNA
and an HKY model for both nDNA loci. Both phased nuclear haplotypes were used for the
analysis. I used an uncorrelated lognormal relaxed clock, Yule tree prior, and a random
starting tree. I specified a substitution rate for the mtDNA gene-tree of 0.0057 substitutions
per site per million years as estimated from previous studies on geckos and other small
vertebrates (Macey et al. 1999; Fujita & Papenfuss 2011). To incorporate uncertainty in this
value (Ho & Phillips 2009), I implemented a normal prior for the mtDNA rate with a mean CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 88 and initial value of 0.0057 and a standard deviation of 0.001. Further, because recent research suggested the utility of calibrating species-trees versus gene-trees in molecular dating analyses (McCormack et al. 2011), I placed a prior on the root node of the species-tree separating Phyllodactylus from Coleonyx. I specified a normal distribution with a mean of
140 and standard deviation of 20 (Gamble et al. 2008). Although using so-called “secondary calibrations” for molecular dating can be less than ideal, results are often robust, particularly if parametric prior distributions are used and subsequent conclusions based on 95% confidence intervals versus simple mean estimates (Sauquet et al. 2012). I ran five separate runs of 200 million generations each, sampling every 20 000 generations. Runs were combined in LOGCOMBINER and trees annotated in TREEANNOTATOR. Adequate sampling for
all parameters was assessed with ESS values of at least 200.
2.7 Diversification rates
I tested hypotheses on whether or not formation of the Mexican TDF as well as Pleistocene
climatic shifts substantially influenced rates of cladogenesis in leaf-toed geckos. Rejection of
the null hypotheses—no influence—would be concluded if diversification rates were
correlated with the timing of these events. I used LASER v. 2.3 (Rabosky 2006a,b) to test for
significant departures from diversification-rate constancy over evolutionary history. LASER
marked an improvement over previous methods in testing for temporal increases in
diversification rate versus decreases as measured by the gamma statistic (γ) based on a
constant rate test (Pybus & Harvey 2000). I calculated the test statistic ΔAICRC by comparing
the likelihood of multiple rate-constant and rate-variables models. Model 1 assumed a pure
birth or Yule model of diversification; Model 2 assumed a constant birth-death model; CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 89
Models 3 and 4 assumed density-dependent diversification rates under an exponential (DDX) or logistic distribution (DDL); Model 5 assumed a single change in diversification rate over time (yule2rate). A significantly positive value of the ΔAICRC statistic signified that a rate-
variable model was a better fit to the data than a rate-constant model using the equation
ΔAICRC = AICRC – AICRV
To visualize trends in lineage accumulation through time, I generated Lineage Through Time
(LTT) plots in LASER. Finally, I calculated γ to test for a decrease in diversification through
time as expected if the number of available niches was being depleted. Significantly negative
values of γ would indicate a decrease in diversification through time with a large number of internal nodes distributed towards the root of the phylogeny (Pybus & Harvey 2000). All
LASER analyses were based on both the dated mtDNA gene-tree topology and the dated
species-tree calculated from the *BEAST analysis.
2.8 Biogeography
I used RASP V2.0 (Yu et al. 2010, 2011) to estimate ancestral distributions and elucidate the
biogeographic history of the P. tuberculosus species group throughout western Mexico.
RASP, the most recent version of S-DIVA (Statistical Dispersal-Vicariance Analysis),
incorporated both phylogenetic uncertainty and multiple equally optimal ancestral states into
consideration when calculating ranges at ancestral nodes given a single or a set of
phylogenetic trees (Nylander et al. 2008; Harris & Xiang 2009). I implemented the Bayesian
Binary MCMC analysis using the default settings (2 runs, 50 000 cycles, 10 chains, sampling
frequency of 100) under an F81+Γ model. I estimated ancestral ranges based on the dated
mtDNA gene-tree obtained in *BEAST versus the species-tree or concatenated topology due CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 90 to the higher support values in the former. To estimate ancestral ranges I first combined tree files for each of the five independent runs of *BEAST implementing a burnin of 10%. I then
performed a Bayesian Binary analysis using the parameters above specifying the mtDNA
gene-trees file as input to RASP. Distribution was classified based on the Mexican state where the animal was captured and by the mtDNA gene-tree topology calculated from RAxML.
2.9 Population genetic and historical demographic analysis
I used DNASP v5 (Librado & Rozas 2009) to calculate nucleotide diversity statistics for each
major lineage recovered in the phylogenetic analysis of mtDNA sequences. These statistics
were not calculated for the intron data because of the relatively low variation within
populations. I calculated the number of haplotypes (H), haplotype diversity (h), average number of nucleotide differences assuming no recombination (k), number of segregating sites
(S), and nucleotide diversity (π). Sequence divergence between mitochondrial lineages
(uncorrected p-distances and Tamura-Nei corrected distances) was calculated in MEGA 5
(Tamura et al. 2011).
To investigate signs of demographic expansion with the mtDNA data I calculated the
Ramos-Onsins and Rozas’ R2 statistic (Ramos-Onsins & Rozas 2002), Tajima’s D (Tajima
1989) and Fu’s Fs (Fu 1997). Results from simulated data have suggested that Fu’s Fs test is
particularly powerful when testing for expansion in large data sets, whereas the Ramos-
Onsins and Rozas’ R2 is more sensitive for smaller populations (Ramos-Onsins & Rozas
2002). Significance of D followed the critical values of Tajima (1989). Significance of both
Fs and R2 was determined from 1000 coalescent simulations assuming no recombination.
Statistics were calculated for populations/lineages with at least four individuals. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 91
I also applied a multilocus Bayesian approach to infer historical changes in effective population sizes through time (Ho & Shapiro 2011). I used an Extended Bayesian Skyline model (Heled & Drummond 2008) implemented using BEAST to infer demographic history,
as simulation evidence suggested that doubling the number of independent loci in
demographic analyses reduces the error rate by a factor of √2 leading to more robust
parameter estimates (Heled & Drummond 2008). Extended Bayesian Skyline Plots were
obtained from each lineage recovered in the mtDNA gene-tree analysis. I used JMODELTEST
to calculate best-fit models for each gene within each lineage using BIC; model parameters
were estimated separately for each gene using BEAST. Bayesian skyline models required
either a known substitution rate or internal calibrations to date times of demographic
changes. I specified the same mtDNA substitution rate and priors as for the species-tree
analysis. I used starting rates of 0.005 substitutions per site per million years for the nDNA
data with a uniform prior between 0 and 100. All phased haplotypes were included. I then
specified an Extended Bayesian Skyline tree prior under a strict clock model and designated
the loci as either mitochondrial or autosomal. Operators were adjusted according to the
recommendations of the program’s authors. Each analysis was run for a 30 million
generations, logging parameters every 3 000 generations to obtain a final sample of 10 000
states. Adequate sampling for all parameters was assessed with ESS values of at least 200.
2.10 Migration rates and effective population sizes
I used MIGRATE-N V3.2.16 (Beerli & Felsenstein 2001; Beerli 2006) to estimate demographic parameters including effective population sizes and asymmetric migration rates between lineages recovered in the phylogenetic analysis. MIGRATE-N used the coalescent in a CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 92
Bayesian or maximum likelihood framework to calculate two parameters from the data, Ө and M, where Ө represented the mutation-scaled effective population size (4Neµ for a nuclear
locus, Neµ for mtDNA) and M represented the mutation-scaled immigration rate (m/µ). The
effective number of immigrants per generation was estimated by multiplying Ө and M. I
performed separate analyses on lineages of P. tuberculosus/muralis and P. lanei while specifying collecting locality and major mitochondrial lineages as separate populations. Only populations/lineages with at least five individuals were included to increase computation efficiency. I used default values for all starting parameters. To infer Ө and M, I specified two independent runs, static heating with four chains using recommended temperatures, a sampling increment of 100, 100 000 recorded steps, and a burnin of 10 000. To account for the difference in effective population sizes of the loci (mtDNA vs. nDNA), I introduced inheritance scalars (4 1 1) for the mtDNA, alpha-enolase, and lamin-A data, respectively. All phased nuclear haplotypes were included in the analysis. Convergence was assessed by examination of ESS values with a target of at least 1 000.
2.11 Isolation-by-distance
I used IBDWS V3.2.1 (Jensen et al. 2005) to test for significant correlations between
geographic (Euclidean) and genetic distance for both P. tuberculosus and P. lanei. As an
estimate of genetic distance, I used Rousset’s measure of differentiation based on sequence
data ((PhiST/(1 – PhiST)). Significance was assessed by Mantel tests (Mantel 1967) using 30
000 randomizations. Because I tested the hypothesis that dispersal was not influenced by
drainage configuration, I created a third binary matrix specifying a value of 0 for populations
adjacent to the same river body and a value of 1 for different systems. I then used partial CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 93
Mantel tests (Smouse et al. 1986) to test for a significant correlation between genetic distance and river configuration; a significant correlation was used to reject the null hypothesis.
Partial Mantel tests controlled for the influence of Euclidean distance when estimating a riverine effect. Data for Mexico’s major rivers and tributaries were obtained from the
Instituto Nacional de Estadistica y Geografia.
3 Results
3.1 Sequence characteristics
The concatenated mitochondrial and nuclear data set contained a total of 2334 characters.
The combined mtDNA data contained 1301 characters, 776 of which were variable and 670
potentially parsimony-informative. The phased alpha-enolase intron alignment including P. xanti contained a total of 441 characters, 134 of which were variable and 116 potentially parsimony-informative. The phased lamin-A alignment contained 592 characters, 113 of which were variable and 105 potentially parsimony-informative. No premature stop codons were found for either mitochondrial gene suggesting that authentic mitochondrial sequences were obtained. No recombinant breakpoints were detected for the nuclear loci with any of the three methods used. JMODELTEST selected the GTR+I+Γ model for both the concatenated mtDNA and alpha-enolase partitions and the HKY+Γ model for lamin-A. Within the mtDNA data the GTR+I+Γ model was selected for both ND4 and cytb, and HKY+ I+Γ for the tRNA.
No indels were present in the alpha-enolase alignment. However, the lamin-A data set contained a total of nine indels ranging from 1–20 bp in length.
CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 94
3.2 Gene-tree analysis
ML analysis of the mtDNA data resulted in a well-supported topology illustrating highly divergent maternal lineages (Fig. 2). Two deep divergences largely corresponded to species
(P. tuberculosus and P. lanei). However, P. tuberculosus magnus, P. muralis muralis, and P. muralis isthmus were not monophyletic. Two highly divergent and sympatric lineages (A1 and B1) occurred at El Charco, Jalisco. Bootstrap support for each lineage varied from moderate to high. Genetic structure was much less pronounced for both intron data sets versus the mtDNA and showed evidence for substantial incomplete lineage sorting. ML analysis of the phased alpha-enolase data resulted in three lineages that largely corresponded to species and sub-species (Supplementary Fig. S1). Conversely, ML analysis of the phased lamin-A data showed less structure, with several haplotypes of some species nested within haplotypes of other species (Supplementary Fig. S2). The concatenated ML analysis was largely congruent with the mtDNA gene-tree (Supplementary Fig. S3). However, lineage A1 from the mtDNA analysis shifted from the P. tuberculosus/muralis lineage to P. lanei.
Further, haplotypes from Cosalá, Sinaloa clustered more closely with haplotypes from
Alamos, Sonora versus haplotypes from Villa Unión, Sinaloa. Lastly, P. muralis isthmus was monophyletic in the concatenated tree and haplotypes for P. lanei rupinus from Michoacán were removed from lineage B2.
3.3 Species-tree analysis and divergence times
The five independent *BEAST runs all showed adequate mixing and convergence as estimated by ESS values greater than 200. The 95% highest posterior density (HPD) for ucld.stdev and CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 95 coefficient of variation did not include zero, indicating that the data did not conform to a strict clock model. The mtDNA gene-tree indicated that the most recent common ancestor of
Phyllodactylus was present during the Cretaceous ~75 Ma (HPD: 40–127; Fig. 3).
Cladogenesis progressed relatively slowly until about 2.67 Ma when the number of lineages increased rapidly up to the present. The *BEAST mtDNA gene-tree was highly congruent with the ML topology, with most of the maternal lineages sharing a most recent common ancestor in the Early Pleistocene. There was evidence for an increase in evolutionary rate for many lineages approximately 25 Ma; a time when speciation was occurring throughout the
species group.
The species-tree topology was generally less well-supported than the mtDNA gene-
tree and suggested a slightly different topology (Fig. 4). However, unlike the mtDNA tree, the species-tree analysis recovered strong evidence for a monophyletic P. lanei rupinus, P.
lanei lanei, and P. tuberculosus magnus. In contrast, P. tuberculosus saxatlis, P. muralis
isthmus, and P. muralis muralis were not monophyletic. Divergence times were more recent
than the mtDNA tree, with speciation and population divergence occurring in the Miocene.
3.4 Diversification rates
I used the maximum likelihood model in LASER to statistically quantify shifts in rate of cladogenesis throughout the tree and, thus, test for significant departures from a rate- constancy model of diversification. No decrease in diversification rate was detected through time over the evolutionary history of these geckos based on either the mtDNA gene-tree (γ =
6.364; P = 1) or the species-tree topology (γ = 1.485; P = 0.9312). The simultaneous comparison of multiple rate-constant and rate-variable models favoured the rate-variable CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 96 yule2rate model for the mtDNA tree (AIC = 22.85) indicating that rates of diversification increased through time (ΔAICRC = 16.0045). Cladogenesis progressed at a rate of
0.0242/million years until 2.67 Ma when the rate shifted to 0.3261/million years.
Significance of the calculated ΔAICRC statistic based on 5 000 simulated trees under a Yule
process (constant birth) indicated a significant difference (P = 0.0002). Diversification
through time progressed gradually up until the Early Pleistocene when cladogenesis followed
a more exponential distribution versus linear and never reached an asymptote (Fig. 3).
Diversification analysis based on the dated species-tree also suggested that a rate-variable
model explained the data better than a rate-constant model (yule2rate AIC = 81.91; bd AIC =
83.17; ΔAICRC = 1.2512). An increased in diversification rate from 0.0100 to 0.0406
occurred approximately 28 Ma (Fig. 4). However, this shift in diversification was not significantly different from a constant birth model (P = 0.2006).
3.5 Biogeography
The distance between independent runs of the Bayesian MCMC analysis was less than 0.01,
indicating congruence in ancestral range estimates. Different biogeographic histories were
recovered depending on which tree or sets of trees were used an input (results not shown).
However, all three data sets (mtDNA, concatenated, species-tree) suggested an origin of
Phyllodactylus in the state of Jalisco. I focus additional discussion on the mtDNA topology
because of the relatively high support for relationships (Fig. 5). Biogeographic histories were
presented for both the P. tuberculosus/muralis and P. lanei lineages. For P. tuberculosus, the
results suggested an ancestral distribution originating in Jalisco, with subsequent dispersal
and divergence to the north (P. tuberculosus saxatlis) and to the south (P. tuberculosus CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 97 magnus, P. muralis isthmus, P. muralis muralis). Three distinct vicariant events were hypothesized for the tuberculosus lineage. Ancestral distributions for P. lanei were also concentrated in the Jalisco region with a single vicariant event separating P. lanei lanei from
P. lanei rupinus.
3.6 Population genetics and demographics
Sequence divergence between major matrilines was exceptionally high, with uncorrected p-
distances and Tamura-Nei distances approaching 32% (Table 2). Genetic diversity within
lineages was low-to-moderate as revealed by nucleotide and haplotype diversities (Table 3).
None of the neutrality tests indicated signs of recent demographic expansion. However, significantly positive values of Fu’s Fs test indicated recent population subdivision for lineages A2 and B2. The extended Bayesian skyline plots also showed weak evidence for recent demographic expansion (Fig. 6). ESS values for all parameters were >100 and most were >200 indicating adequate sampling from the posterior distribution. I was unable to get
adequate sampling from the posterior distribution for lineage B1. Although the demographic.population size-changes parameter often suggested one, two, and three changes, the 95% confidence intervals were large for many lineages and no plot showed an obvious trend.
3.7 Migration rates and effective population sizes
Results from the MIGRATE-N analyses suggested that population sizes and migration rates
between populations were low. For the P. tuberculosus/muralis lineage, the number of CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 98
migrants per generation (4Nm) ranged from 0.0057 between lineages A3 and A10 to 6.7195
from lineage A5 to A9 (Supplementary Table 1). For P. lanei, the number of migrants per
generation (4Nm) was one or less for all comparisons of populations (Supplementary Table
2). Estimated effective population sizes (4Neu) for P. tuberculosus were smallest for lineages
A7 (0.0008) and A10 (0.00003). Effective population size for P. lanei at Chamela, Jalisco
(0.0006) was much smaller than sizes for the remaining populations.
3.8 Isolation-by-distance
The Mantel test for P. tuberculosus found a significant correlation between log geographic
distance and genetic distance (r = 0.2892, P = 0.0101). Partial Mantel tests showed no
significant correlation between drainage structure and genetic distance with (r = -0.0620 P =
0.2596) and without (r = 0.1502, P = 0.2596) controlling for the effects of Euclidean
distance. There was no significant correlation between log geographic distance and genetic
distance for P. lanei (r = -0.2625, P = 0.8128)
4 Discussion
Neotropical dry forests are one of Earth’s most diverse ecosystems based on both species
richness and number of endemics (Myers et al. 2000). However, these forests occur primarily in developing countries where the threat of extirpation is drastically increasing due to a suite
of anthropogenic factors (Trejo & Dirzo 2000). To focus conservation efforts in this highly
threatened system in Mexico, recent efforts seek to determine both specific areas of high
richness and endemism as well as source-sink areas for future diversification (García 2006; CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 99
Becerra & Venable 2008). However, comprehensive analyses from diverse scientific disciplines are required to fully understand the mode and tempo of biotic evolution in this ecosystem. By combining phylogenetic, biogeographic, molecular clock and demographic analyses from multiple loci, I seek to determine the patterns and the drivers of diversification in western Mexico’s TDF. Results suggest that these forests may be much more diverse than previously realized, and point to several potential mechanisms of diversification including
Pleistocene climatic cycling.
4.1 Phylogenetic analysis
My phylogenetic analysis suggests that the degree of cryptic molecular diversity for leaf-toed
geckos throughout western Mexico is substantial, with sequence divergence values
approaching 32%. Phyllodactylus tuberculosus and P. muralis are composed of at least 11
highly differentiated lineages whereas P. lanei contains five cryptic lineages. Phylogenetic
analysis of the nuclear haplotypes also reveals genetic structure, albeit less pronounced than
the mtDNA analysis. Analysis of the phased alpha-enolase haplotypes suggests that P.
tuberculosus saxatilis is not exchanging genetic information with other populations
throughout southern Mexico. These patterns are not resolved in analyses of the lamin-A.
The discrepancy between gene-trees from independent loci at and below the species
level often suggests gene flow or incomplete lineage sorting as causal factors (Degnan &
Rosenberg 2006,2009). Although sex-biased dispersal can explain these patterns, no data to
suggest this occurs. Interspecific hybridization could also be a factor. However, my results
suggest that haplotypes are shared between individuals over relatively broad spatial scales.
This would seem counter to a hypothesis of limited dispersal abilities in geckos. More likely, CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 100 the differences between the mtDNA and nDNA gene trees is due to the incomplete sorting of ancestral haplotypes due to the larger effective population sizes of the nuclear markers. One potentially powerful approach to deal with rampant incomplete lineage sorting is to employ powerful species-tree methods that incorporate the stochasticity of the coalescent process into genealogic and phylogenetic inference (Liu et al. 2008,2009; Heled & Drummond
2010). Indeed the multilocus *BEAST analysis resulted in a different hypothesis versus both
the mtDNA and concatenated analyses, albeit with lower statistical support. However,
several species and subspecies that were not monophyletic in the gene-tree analysis (e.g. P.
tuberculosus magnus) were monophyletic in the species-tree analysis. These results add to
the growing body of literature and suggest that combining novel species-tree methods with
traditional gene-tree based approaches can be a powerful means to elucidate historical
relationships at and below the species level.
These results highlight the utility of incorporating both haplotypes from heterozygous
individuals in genealogical and phylogenetic inference. Many researchers choose to construct
a genealogy choosing one of the two haplotypes from heterozygous individuals. The arbitrary
(or statistical) choice of including one haplotype over the other can result in biased
conclusions regarding genetic structure and evolutionary relationships (Weisrock et al.
2012). This is of particular concern in studies at and below the species level where
incomplete lineage sorting is likely to be common. As multilocus studies are now
commonplace, I urge researchers to present genealogies and perform analyses with both
nuclear haplotypes so correct genetic structure can be visualized. The continual development
of more sophisticated phylogenetic and demographic algorithms (e.g. Heled & Drummond
2008,2010) is now making this a feasible endeavour. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 101
A growing body of literature focuses on phylogenetic patterns and processes throughout the diverse ecosystems of Mexico. Some of these studies test climatic and geologic hypotheses influencing the evolution of montane species (Bryson et al. 2011a,b,c), whereas others focus on lowland tropical taxa (Mulcahy & Mendleson 2000; Mulcahy et al.
2006). My study adds to the limited number of vertebrate studies focusing on the TDF of western Mexico that find extensive cryptic molecular diversity throughout the region
(Hasbún et al. 2005; Devitt 2006; Zarza et al. 2008). However, to my knowledge this study is only the second to incorporate variable nuclear markers to infer historical processes.
Although not as variable as the mtDNA, the intron data allowed for a multilocus perspective of processes governing the evolutionary history of the P. tuberculosus group. I urge researchers to continue to scan the genome for variable nuclear markers to test historical processes effecting organisms distributed throughout the TDF of western Mexico. The combination of markers with both different mutation rates and inheritance modes can provide a much more comprehensive investigation into patterns and processes of molecular evolution.
4.2 Divergence times and diversification rates
The relaxed molecular clock analysis of the mtDNA gene-tree dates the most recent common
ancestor of Phyllodactylus to the Cretaceous approximately 75 Ma, corroborating other
recent molecular evidence for an ancient origin of geckos (Gamble et al. 2008). The
diversification analysis in LASER suggests that cladogenesis progressed gradually up until
about 2.67 Ma when the rate of lineage accumulation increased up to the present. I also find
evidence for increased rates of evolution and speciation approximately 25 Ma. Interestingly, CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 102 these results are highly congruent to a time-calibrated phylogeny and diversification analysis of Bursera, a conspicuous member of the Mexican TDF (Becerra 2005). For example, it is hypothesized that Bursera also has an origin dating back to the Cretaceous approximately
70–80 Ma. Further, the tree genus is hypothesized to have experienced rapid diversification at around 20–30 Ma, a time when extensive tectonic activity was altering Mexico’s Sierra
Madre Occidental and MVB, and creating the TDF. Most of the species identified in my mtDNA analysis coalesce within this time frame. Taken together, this suggests that
Phyllodactylus evolved in concert with this genus of trees, and that the formation of the
Mexican TDF may be responsible for driving cladogenesis in these geckos. The dated species-tree also suggests an influence of climate change during the Miocene as a driver of cladogenesis in these geckos. Although many divergence dates in the species-tree are younger than those of the mtDNA as theory predicts, my diversification analysis did suggest a moderate shift in diversification rate approximately 28 Ma.
The importance of Pleistocene climatic shifts in shaping the evolutionary history of
Neotropical taxa is a hotly debated topic (e.g. Klicka & Zink 1997; Hewitt 2000). My results suggest a substantial influence of both Miocene/Pliocene and Pleistocene effects in shaping the evolutionary history of these geckos. For example, most of the mtDNA lineages date back to the Early Pleistocene approximately 2.5 Ma, making the shifts in TDF habitat during the Quaternary (Pennington et al. 2000; Werneck et al. 2011) a likely model to explain patterns of intraspecific divergence in these geckos. My results are also concordant with previous studies that suggest an older timeframe of speciation in Neotropical lowland taxa
(Weir & Schluter 2004; Weir 2006) as many of the species and populations diverged during the Miocene. However, my results cannot refute the hypothesis of Pleistocene effects on CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 103 diversification patterns within the TDF of western Mexico. This generally agrees with other studies that find evidence for both Pleistocene and older climatic and geologic events important in shaping the evolution of TDF biota (Pennington et al. 2004).
The continual reduction of next-generation sequencing costs is resulting in studies that incorporate information from multiple unlinked genetic loci to test evolutionary hypotheses. These multilocus data sets have, in turn, spawned the development of several new phylogenetic algorithms that explicitly model the stochastic nature of gene coalescence
(Heled & Drummond 2008;2010; Liu 2008). The species-tree method implemented in
*BEAST is particularly powerful given its flexibility regarding model specification and priors
for a variety of parameters. An additional attractive feature of *BEAST is its ability to
simultaneously estimate gene- and species-trees. Previously, molecular clock analyses
focused on obtaining divergence dates for a given gene-tree topology. However, in analyses
such as these it can be difficult to select the appropriate tree prior for the data (Ho 2005). For
example, a Yule prior is generally appropriate for data at and above the species level whereas
coalescent priors are suitable for intrapopulation sequences. However, many data sets, such
as this, consist of sequences where neither a Yule nor a coalescent prior would be appropriate
to estimate a gene-tree. In this case, a species-tree model would be an appropriate alternative
as ‘species’ and ‘populations’ better fit the Yule species-tree prior (A. Drummond, pers.
comm.). Further, although these species-tree methods were explicitly designed to deal with
multilocus sequence data, the can also be applied to single locus data to provide a rough
estimation of the species- or population-tree (Heled & Drummond 2010). As inappropriate
selection of tree priors coupled with incorrect fossil information can severely bias divergence dating analyses (Ho 2005,2007; Ho & Phillips 2009), employing a *BEAST model may be CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 104 more accurate for particular sampling schemes. My study also adds to a recent study in demonstrating the utility of calibrating species-trees versus gene-trees (McCormack et al.
2011). I predict that this will become more common in the near future due to both the increase in multilocus data sets and the ease in which this can now be performed (Drummond et al. 2012)
4.3 Biogeography
My biogeographic analysis suggests several key vicariant events shaping the history of both the P. tuberculosus/P. muralis lineage and the P. lanei lineage. Interestingly, the proposed
events correspond to several previously hypothesized biogeographic barriers. First, my
analysis suggests the clade has its origin in the state of Jalisco, southwestern Mexico with
subsequent dispersal to the north and south. Previous studies also suggest that this region
may explain both current patterns of richness and endemicity as well as potential source areas
for future diversification (García 2006; Becerra & Venable 2008). Becerra and Venable
(2008) suggest that the high levels of species accumulation in the area are due to the diverse
set of environmental conditions generated by the confluence of several of Mexico’s major mountain chains. Becerra (2005) postulates an ancestral distribution of the tree genus
Bursera in the area. Thus, it appears that Jalisco’s TDF is an important area for species accumulation and diversification across broad taxonomic groups.
My biogeographic analysis for the P. tuberculosus/P. muralis lineage suggests three
vicariant events that coincide with previously hypothesized biogeographic boundaries. Event number one partitions individuals into groups north and south of the MVB. Numerous studies suggest that this mountain chain serves as a biogeographic barrier between Nearctic and CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 105
Neotropical biotas (Marshall & Liebherr 2000; Mulcahy & Mendleson 2000; Mateos et al.
2002; Hulsey et al. 2004; Mateos 2005; Devitt 2006; Huidobro et al. 2006; Morrone 2006).
Many lowland species are, thus, composed of highly differentiated groups occurring on either side of the MVB. Recent studies also suggest that orogenesis of the MVB drove rates of cladogenesis for montane species (Bryson & Riddle 2011; Bryson et al. 2011a,b,c,d). The
MVB is known to have formed in a west-east progression beginning at around 23 Ma and ending approximately 2.5 Ma (Moran-Zenteno 1994). The most extensive ridges of this belt are found in western regions throughout Jalisco, with the number and prominence of volcanoes dissipating towards Veracruz. However, phylogenetic evidence suggests that even the smaller ridges to the east can serve as major biogeographic barriers to lowland taxa
(Mulcahy & Mendleson 2000; Mulcahy et al. 2006). My molecular clock and biogeographic analysis cannot refute the hypothesis of vicariance due to the early orogenesis of the MVB in
Jalisco.
Event number two separates populations of P. tuberculosus saxatilis occurring in
Sonora and Sinaloa. Devitt (2006) provides a mtDNA study of the western lyresnake
(Trimorphodon biscutatus) and also reports a strong mtDNA discordance in the same general area. My study confirms these findings, suggesting a common biogeographic barrier between the states of Sonora and Sinaloa that may be a result of vicariance due to the Río Fuerte or the Sierra de Barabampo (Hafner & Riddle 2005).
Vicariant event three separates populations of P. tuberculosus magnus in Guerrero from populations in Oaxaca and Chiapas. These results are concordant with a phylogeographic study of black spiny-tailed iguanas that found a mtDNA break in the same general vicinity (Zarza et al. 2008). The biogeographic feature responsible for these vicariant CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 106 events is presently unknown. However, the Sierra Madre del Sur occurs in a belt throughout southern Mexico and is composed of numerous small mountain chains which may have served as a historical biogeographic barrier between lineages. Additional phylogenetic and phylogeographic studies are required to test the generality of a Guerrero/Oaxaca break.
An additional vicariant event previously hypothesized in the literature, but not corroborated by my biogeographic analysis separates populations of P. tuberculosus magnus east and west of the Isthmus of Tehuantepec. Unlike the MVB, there is still substantial controversy over whether there was a historical barrier across the isthmus in southern Oaxaca or not (Mulcahy et al. 2006). Some researchers suggest that a Pliocene seaway fragmented the isthmus into eastern and western components, in turn, isolating populations on either side
(Sullivan et al. 2000). Although my biogeographic analysis does not suggest a vicariant event, the phylogenetic analysis does suggest a possible historical barrier across the isthmus.
My biogeographic analysis for P. lanei suggests a single key vicariant event, separating populations in Guerrero from those in Jalisco and Michoacan. This division corresponds to the current taxonomy of P. lanei (Dixon 1964), with P. l. lanei being restricted to Guerrero and P. l. rupinus occurring in Jalisco and Michoacan. These subspecies are distributed on opposite sides of the Balsas Basin, a highly speciose region that contains many endemic taxa (Rzedowski 1993; Zaldívar-Riverón et al. 2004; Becerra & Venable
2008). Vicariance is most likely due to the Río Balsas, a large river that flows southwestwardly through the states of Mexico, Puebla, Morelos, and Guerrero. Alternatively, the Sierra de Taxco, which divides Balsas Basin into eastern and western units, may be responsible for the split within P. lanei (Becerra & Venable 1999). Conversely, my biogeographic analysis suggests an ancestral range in Jalisco with an associated vicariant CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 107 event dating to the formation of the MVB. Thus, it remains uncertain as to the relative importance of orogenesis of the MVB versus alternative hypotheses for the split between P. lanei lanei and P. lanei rupinus.
4.4 Population genetics and demographics
My population genetic results suggest low to moderate diversity within populations and
substantial differentiation between populations. Analyses suggest that almost every sampling
locality constitutes a highly-differentiated group, with sequence divergence values
approaching 32%. Summary statistics-based tests for demographic expansion fail to find
signs of expansion for any of the major mitochondrial lineages. Fu’s Fs test shows
significantly positive values for lineages A2 and B2 which may be a result of recent
fragmentation. However, both of these lineages include individuals from different areas that
are grouped together because of sample sizes. Thus, these significantly positive values are
likely the result of combining divergent lineages into a single unit for analysis.
My multilocus Bayesian estimates of demographic changes show weak signs of
increases in effective population sizes through time. The Extended Bayesian Skyline Plots
are a recent extension of the more-traditional single-locus skyline plots used to track
demographic changes through time (Heled & Drummond 2008). Although some of the
lineages show relatively wide 95% confidence intervals, results suggest that populations of
these geckos did not expand during or following Pleistocene climatic shifts. Conversely,
some of the lineages (e.g. A2, B2) showed trends of decreasing population size. However, I
feel this is due to combining genetically diverse individuals into a single unit for these analyses (Ho & Shapiro 2011) and not an actual demographic phenomenon. To maximize CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 108 genetic information for these tests I included all phased nuclear haplotypes. Evidence suggests that increasing the number of independent loci, versus increasing the number of individuals and lengths of sequences, has a drastic influence on parameter estimates (Heled
& Drummond 2008). In contrast to between-population comparisons, sequence diversity within populations is low to moderate for both nuclear loci, resulting in relatively little information to estimate historical demographic parameters. The collection of additional highly variable loci will likely decrease error estimates and provide a more comprehensive estimate of demographic trends through time for these geckos. Notwithstanding artefacts of sampling bias, these results indicate that the Pleistocene had little impact on demographic trends in leaf-toed geckos.
4.5 Migration
My MIGRATE-N analyses suggest both small effective population sizes and migrates rates
between populations and lineages. For P. lanei the number of migrants per generation (4Nm)
is less than one for most comparisons. Migration rates for P. tuberculosus/muralis appear
higher, with some populations exchanging approximately six migrants per generation.
However, these results should be taken with caution. Unlike the analysis for P. lanei,
confidence intervals for several parameters are relatively large for P. tuberculosus/muralis.
For example, the effective population size estimate for lineage A9 is 0.0797 with a 97.5%
confidence interval between 0.0518 and 0.1000. Because I multiply Ө and M to calculate the number of migrants per generation, these relatively large confidence intervals can affect
parameter estimates substantially. Further, confidence intervals for several M parameters are
high. Although this could be an artefact of analysis parameters, I specify multiple runs with CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 109 different heating parameters to better sample parameter space. ESS values are also large for all parameters suggesting that the chain was run long enough. I suggest that the large confidence intervals are due to the signal in the data and that the collection of additional independent loci will improve parameter estimates (Beerli & Felsenstein 2001; Beerli 2006).
Nevertheless, these results agree with my other analyses and suggest that these geckos exhibit limited dispersal abilities. These results are also concordant with previous genealogical studies of geckos that suggest high levels of differentiation over small spatial scales (Gübitz et al. 2000; Kasapidis et al. 2005; Thorpe et al. 2008)
4.6 IBD and drainage systems
Results suggest a mixed influence of Euclidean distance on genetic differentiation, providing
additional evidence for cryptic species. Contrary to expectations, partial Mantel tests fail to
find a significant correlation between drainage configuration and genetic divergence between
populations. Although visualizing my sampling localities on a map of Mexico’s drainage
systems may suggest a riverine effect on dispersal patterns, these results suggest that it is
more likely that these geckos simply encompass very small home range sizes. These
conclusions are also corroborated by the phylogenetic and migration rate analyses.
Unfortunately, I cannot perform partial Mantel tests for P. lanei due to the relatively small
number of populations. In addition, one of my populations (El Charco) is not adjacent to a
river-body, yet it is highly diverged from the nearby population of Chamela. Taken together,
these results suggest a limited influence of drainage configuration on genetic structure in
these geckos.
CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 110
4.7 Gecko diversity and taxonomic implications
Few studies examine geographic patterns of genetic differentiation in geckos, particularly throughout the New World. Most studies focusing of historical processes at and below the species level are restricted to North African, European, or Australian biotas (e.g. Gübitz et al.
2000; Kaspidis et al. 2005; Fujita et al. 2010; Rato et al. 2011). However, a general trend among these studies and mine is that geckos often harbour substantial cryptic diversity and that dispersal distances are presumably quite small. To my knowledge, this is the first study to statistically quantify migration rates between populations of geckos. My multilocus analysis of migration rates, coupled with the high levels of mtDNA sequence divergence
(32%) suggests that the taxonomy of the genus may be in need of substantial revision.
Because both the mtDNA and alpha-enolase data suggest that P. tuberculosus saxatilis is not
exchanging genetic information with other populations, I elevate this subspecies to species
level as Phyllodactylus saxatilis new combination. Further, my mtDNA and species-tree
analysis combined with previous karyotypic (Castiglia et al. 2009) and morphological data
(Dixon 1964) suggests that P. lanei rupinus should be elevated to species status and I consider it to be Phyllodactylus rupinus new combination. Additional data from variable nuclear markers, morphology, and ecology are needed to determine if additional maternal lineages warrant species status (e.g. mtDNA lineage A1). Regardless, these results suggest that cryptic molecular diversity in taxa distributed throughout the TDF of western Mexico is drastically underestimated. If we are to conserve this highly biodiverse ecoregion, additional studies are needed from diverse organismal groups to elucidate genetic patterns and the evolutionary processes generating them.
CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 111
4.8 Conclusions
The results of this study suggest that both the formation of TDF throughout western Mexico
in the Miocene and Pleistocene climatic shifts had a substantial influence on biotic evolution
throughout the region. Dry forests are one of Earth’s most diverse yet threatened ecosystems
(Myers et al. 2000) yet, compared to moist tropical forests, we still know relatively little
about the tempo and mechanisms of diversification (Werneck et al. 2011). By employing
sophisticated multilocus phylogenetic, molecular clock, demographic, and biogeographic
analyses, my results point to diversification processes operating prior-to, and including,
Pleistocene climate change. This study adds to a limited literature that illustrates to utility of
these approaches to understanding the biological history of species-rich areas (e.g. Che et al.
2010). If we are to adequately conserve Earth’s biodiversity at a time when natural and
anthropogenic change threaten future diversification, similar multi-tiered approaches are
required to elucidate the diverse processes contributing to the evolution of our planet’s rich
biota.
Acknowledgments
I thank Victor H. Jiménez Arcos, Roberto Lhemish Martínez Bernal, and Anibal Diaz de la
Vega for help in the field. I thank Oliver Haddrath, Kristen Choffe, and Amy Lathrop for assistance with lab work. Christina Davy, Jean-François Flot and Santiago Sánchez helped with data analysis. Alexei Drummond provided invaluable assistance with all BEAST
analyses. Funding was provided by the Natural Sciences and Engineering Research Council of Canada A3148 (RWM), the Theodore Roosevelt Memorial Fund (CB), and the Schad CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 112
Foundation at the Royal Ontario Museum. All research was conducted using approved
Animal Use Protocols. Collection (FLOR0030, SGPA/DGFS/047097/09,
SGPA/DGVS/02320/10, SGPA/DGVS/02372/11) and export permits (028, 059) were obtained from SEMARNAT through the Universidad Nacional Autónoma de México
(UNAM).
CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 113
References
Arenas M, Posada D (2010) The effect of recombination on the reconstruction of ancestral
sequences. Genetics, 184, 1133–1139.
Arevalo E, Davis SK, Sites JW (1994) Mitochondrial DNA sequence divergence and
phylogenetic relationships among eight chromosome races of the Sceloporus grammicus
complex (Phrynosomatidae) in central México. Systematic Biology, 43, 387–418.
Barrowclough GF, Zink RM (2009) Funds enough, and time: mtDNA, nuDNA, and the
discovery of divergence. Molecular Ecology, 18, 2934–2936.
Becerra JX (2005) Timing the origin and expansion of the Mexican dry forest. Proceedings
of the National Academy of Sciences USA, 102, 10919–10023.
Becerra JX, Venable E (1999) Macroevolution of insect-plant associations: the relevance of
host biogeography to host affiliation. Proceedings of the National Academy of Sciences
USA, 96, 12626–12631.
Becerra JX, Venable E (2008) Sources and sinks of diversification and conservation
priorities for the Mexican tropical dry forest. PLoS One, 3, e3436.
doi:10.1371/journal.pone.0003436.
Beerli P (2006) Comparison of Bayesian and maximum likelihood inference of population
genetic parameters. Bioinformatics, 22, 341–345.
Beerli P, Felsenstein J (2001) Maximum likelihood estimation of a migration matrix and
effective population sizes in n subpopulations by using a coalescent approach.
Proceedings of the National Academy of Sciences USA, 98, 4563–4568.
Blair C, Murphy RW (2011) Recent trends in molecular phylogenetic analysis: where to
next? Journal of Heredity, 102, 130–138. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 114
Blair C, Mendez de la Cruz FR, Ngo A, Lindell J, Lathrop A, Murphy RW (2009) Molecular
phylogenetics and taxonomy of leaf-toed geckos (Phyllodactylidae: Phyllodactylus)
inhabiting the peninsula of Baja California. Zootaxa, 2027, 28–42.
Brito PH, Edwards SV (2009) Multilocus phylogeography and phylogenetics using
sequence-based markers. Genetica, 135, 439–455.
Bryson RW, Riddle BR (2011) Tracing the origins of widespread highland species: a case of
Neogene diversification across the Mexican sierras in an endemic lizard. Biological
Journal of the Linnean Society, doi: 10.1111/j.1095-8312.2011.01798.x.
Bryson RW, Murphy RW, Lathrop A, Lazcano-Villareal D (2011a) Evolutionary drivers of
phylogeographical diversity in the highlands of México: a case study of the Crotalus
triseriatus species group of montane rattlesnakes. Journal of Biogeography, 38, 697–710.
Bryson RW, García-Vázquez UO, Riddle BR (2011b) Phylogeography of Middle American
gophersnakes: mixed responses to biogeographical barriers across the Mexican Transition
Zone. Journal of Biogeography, 38, 1570–1584.
Bryson RW, García-Vázquez UO, Riddle BR (2011c) Diversification in the Mexican horned
lizard Phrynosoma orbiculare across a dynamic landscape. Molecular Phylogenetics and
Evolution (in press).
Bryson RW, García-Vázquez UO, Riddle BR (2011d) Relative roles of Neogene vicariance
and Quaternary climate change on the historical diversification of bunchgrass lizards
(Sceloporus scalaris group) in México. Molecular Phylogenetics and Evolution (in
press).
Burgos A, Maass MJ (2004) Vegetation change associated with land-use in tropical dry
forest areas of western México. Agriculture, Ecosystems, & Environment, 104, 475–481. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 115
Castiglia R, García A, Bezerra AMR, Flores-Villa O, Gornung E (2009) Karyotypic
diversification due to Robertsonian rearrangements in Phyllodactylus lanei Smith, 1935
(Squamata, Gekkonidae) from México. Rendiconti Lincei, 20, 77–82.
Ceballos G, García A (1995) Conserving neotropical biodiversity: the role of dry forests in
western México. Conservation Biology, 9, 1349–1356.
Che J, Zhou WW, Hu JS, Yan F, Papenfuss TJ, Wake DB, Zhang YP (2010) Spiny frogs
(Paini) illuminate the history of the Himalayan region and Southeast Asia. Proceedings of
the Nationalo Academy of Sciences USA, 107, 13765–13770.
Degnan JH, Rosenberg NA (2006) Discordance of species trees with their most likely gene
trees. PLoS Genetics, 2, 762–768.
Degnan JH, Rosenberg NA (2009) Gene tree discordance, phylogenetic inference, and the
multispecies coalescent. Trends in Ecology and Evolution, 24, 332–340.
Devitt TJ (2006) Phylogeography of the western lyresnake (Trimorphodon biscutatus):
testing aridland biogeographical hypotheses across the Nearctic-Neotropical transition.
Molecular Ecology, 15, 4387–4407.
Dixon JR (1964) The systematics and distribution of lizards of the genus Phyllodactylus in
North and Central America. New México State University Research Center, Scientific
Bulletin, 64-1, 1–139.
Drummond AJ, Rambaut A (2007) BEAST: Bayesian evolutionary analysis by sampling
trees. BMC Evolutionary Biology, 7, 214.
Drummond AJ, Ho SYW, Phillips MJ, Rambaut A (2006) Relaxed phylogenetics and dating
with confidence. PLoS Biology, 4, e88. doi:10.1371/journal.pbio.0040088. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 116
Drummond AJ, Suchard MA, Xie D, Rambaut A (2012) Bayesian phylogenetics with
BEAUTi and the BEAST 1.7. Molecular Biology and Evolution. Advance Access.
Edwards SV (2009) Is a new and general theory of molecular systematics emerging?
Evolution, 63, 1–19.
Edwards SV, Beerli P (2000) Perspective: gene divergence, population divergence, and the
variance in coalescence times in phylogeographic studies. Evolution, 54, 1839–1854.
Edwards SV, Bensch S (2009) Looking forwards or backwards in avian phylogeography? A
comment on Zink and Barrowclough 2008. Molecular Ecology, 18, 2930–2933.
Flot JF (2010) SEQPHASE: a web tool for interconverting PHASE input/output files and FASTA
sequence alignments. Molecular Ecology Resources, 10, 162–166.
Friesen VL, Congdon BC, Walsh HE, Birt TP (1997) Intron variation in marbled murrelets
detected using analyses of single-stranded conformational polymorphisms. Molecular
Ecology, 6, 1047–1058.
Friesen VL, Congdon BC, Kidd MG, Birt TP (1999) PCR primers for the amplification of
five nuclear introns in vertebrates. Molecular Ecology, 8, 2147–2149.
Fu YX (1997) Statistical tests of neutrality of mutations against population growth,
hitchhiking and background selection. Genetics, 147, 915–925.
Fujita MK, Papenfuss TJ (2011) Molecular systematics of Stenodactylus (Gekkonidae), an
Afro-Arabian gecko species complex. Molecular Phylogenetics and Evolution, 58, 71–
75.
Fujita MK, McGuire JA, Donnellan SC, Moritz C (2010) Diversification and persistence at
the arid-monsoonal interface: Australia-wide biogeography of the Bynoe’s gecko
(Heteronotia binoei; Gekkonidae). Evolution, 64, 2293–2314. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 117
Gamble T, Bauer AM, Greenbaum E, Jackman TR (2008) Evidence for Gondwanan
vicariance in an ancient clade of gecko lizards. Journal of Biogeography, 35, 88–104.
García A (2006) Using ecological niche modelling to identify diversity hotspots for the
herpetofauna of Pacific lowlands and adjacent interior valleys of México. Biological
Conservation, 130, 25–46.
Gillespie TW, Grijalva A, Farris CN (2000) Diversity, composition, and structure of tropical
dry forests in Central America. Plant Ecology, 147, 37–47.
Grimaldi D (1995) On the age of Dominican amber. Amber, resinite, and fossil resins (ed
Anderson KB, Crelling JC), pp. 203–217. American Chemical Society, Washington D.C.
Gübitz T, Thorpe RS, Malhotra A (2000) Phylogeography and natural selection in the
Tanerife gecko Tarentola delalandii: testing historical and adaptive hypotheses.
Molecular Ecology, 9, 1213–1221.
Guindon S, Gascuel O (2003) A simple, fast, and accurate algorithm to estimate large
phylogenies by maximum likelihood. Systematic Biology, 52, 696–704.
Hafner DJ, Riddle BR (2005) Mammalian phylogeography and evolutionary history of
northern México’s deserts. pp. 225-245. In Biodiversity, ecosystems, and conservation in
northern México. (Cartron JL Ceballos G, Felger RS eds.). Oxford University Press, New
York.
Hall TA (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis
program for Windows 95/98/NT. Nucleic Acids Symposium Series, 41, 95–98.
Harris AJ, Xiang QY (2009) Estimating ancestral distributions of lineages with uncertain
sister groups: a statistical approach to dispersal-vicariance analysis and a case using CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 118
Aesculus L. (Sapindaceae) including fossils. Journal of Systematics and Evolution, 47,
349–368.
Hasbún CR, Gómez G, Köhler A, Lunt DH (2005) Mitochondrial DNA phylogeography of
the Mesoamerican spiny-tailed lizards (Ctenosaura quinquecarinata complex): historical
biogeography, species status and conservation. Molecular Ecology, 14, 3095–3107.
Heled J, Drummond AJ (2008) Bayesian inference of population size history from multiple
loci. BMC Evolutionary Biology, 8, 289.
Heled J, Drummond AJ (2010) Bayesian inference of species trees from multilocus data.
Molecular Biology and Evolution, 27, 570–580.
Hewitt G (2000) The genetic legacy of Quaternary ice ages. Nature, 405, 907–913.
Hillis DM, Bull JJ (1993) An empirical test of bootstrapping as a method for assessing
confidence in phylogenetic analysis. Systematic Biology, 42, 182–192.
Ho SYM (2005) Time dependency of molecular rate estimates and systematic overestimation
of recent divergence times. Molecular Biology and Evolution, 22, 1561–1568.
Ho SYM (2007) Calibrating molecular estimates of substitution rates and divergence times in
birds. Journal of Avian Biology, 38, 409–414.
Ho SYM, Phillips MJ (2009) Accounting for calibration uncertainty in phylogenetic
estimates of evolutionary divergence times. Systematic Biology, 58, 367–380.
Ho SYW, Shapiro B (2011) Skyline-plot methods for estimating demographic history from
nucleotide sequences. Molecular Ecology Resources, 11, 423–434.
Huidobro L, Morrone JJ, Villalobos JL, Álvarez F (2006) Distributional patterns of
freshwater taxa (fishes, crustaceans and plants) from the Mexican transition zone.
Journal of Biogeography, 33, 731–741. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 119
Hulsey DC, García de León FJ, Sánchez Johnson Y, Hendrickson DA, Near TJ (2004)
Temporal diversification of Mesoamerican cichlid fishes across a major biogeographic
boundary. Molecular Phylogenetics and Evolution, 31, 754–764.
Jensen JL, Bohonak AJ, Kelley ST (2005) Isolation by distance, web service. BMC Genetics,
6, 13.
Kasapidis P, Magoulas A, Mylonas M, Zouros E (2005) The phylogeography of the gecko
Cyrtopodion kotschyi (Reptilia: Gekkonidae) in the Aegean archipelago. Molecular
Phylogenetics and Evolution, 35, 612–623.
Klicka J, Zink RM (1997) The importance of recent ice ages in speciation: a failed paradigm.
Science, 12, 1666–1669.
Kluge AJ (1995) Cladistic relationships of sphaerodactyl lizards. American Museum
Novitates, 3139, 1–23.
Knowles LL (2009) Estimating species trees: methods of phylogenetic analysis when there is
incongruence across genes. Systematic Biology, 58, 463–467.
Kocher TD, Thomas WK, Meyer A, Edwards SV (1989) Dynamics of mitochondrial DNA
evolution in animals: amplification and sequencing with conserved primers. Proceedings
of the National Academy of Sciences USA, 86, 6196–6200.
Kubatko LS, Carstens BC, Knowles LL (2009) STEM: species tree estimation using
maximum likelihood for gene trees under coalescence. Bioinformatics, 25, 971–973.
Lee JY, Edwards SV (2008) Divergence across Australia’s Carpentarian barrier: statistical
phylogeography of the red-backed fairy wren (Malurus melanocephalus). Evolution, 62,
3117–3134. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 120
Librado P, Rozas J (2009) DnaSP v5: a software for comprehensive analysis of DNA
polymorphism data. Bioinformatics, 25, 1451–1452.
Liu L (2008) BEST: Bayesian estimation of species trees under the coalescent model.
Bioinformatics, 24, 2542–2543.
Liu L, Pearl DK, Brumfield RT, Edwards SV (2008) Estimating species trees using multiple-
allele DNA sequence data. Evolution, 62, 1–12.
Liu L, Yu L, Kubatko L, Pearl DK, Edwards SV (2009) Coalescent methods for estimating
phylogenetic trees. Molecular Phylogenetics and Evolution, 53, 320–328.
Macey JR, Wang Y, Anajeva NB, Larson A, Papenfuss TJ (1999) Vicariant patterns of
fragmentation among gekkonid lizards of the genus Teratoscincus produced by the Indian
collision: a molecular phylogenetic perspective and an area cladogram for Central Asia.
Molecular Phylogenetics and Evolution, 12, 320–332.
Mantel N (1967) The detection of disease clustering and a generalized regression approach.
Cancer Research, 27, 209–220.
Marshall CJ, Liebherr JK (2000) Cladistic biogeography of the Mexican transition zone.
Journal of Biogeography, 27, 203–216.
Martin DP, Rybicki E (2000) RDP: detection of recombination amongst aligned sequences.
Bioinformatics, 16, 562–563.
Martin DP, Lemey P, Lott M, Moulton V, Posada D, Lefeuvre P (2010) RDP3: a flexible and
fast computer program for analyzing recombination. Bioinformatics, 26, 2462–2463.
Mateos M (2005) Comparative phylogeography of livebearing fishes in the genera
Poeciliopsis and Poecilia (Poecilidae: Cyprinodontiformes) in central México. Journal of
Biogeography, 32, 775–780. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 121
Mateos M, Sanjur OI, Vrijenhoek RC (2002) Historical biogeography of the livebearing fish
genus Poeciliopsis (Poecilidae: Cyprinodontiformes). Evolution, 56, 972–984.
Maynard Smith J (1992) Analyzing the mosaic structure of genes. Journal of Molecular
Evolution, 34, 126–129
Moore WS (1995) Inferring phylogenies from mtDNA variation: mitochondrial gene trees
versus nuclear gene trees. Evolution, 49, 718–726.
Moran-Zenteno D (1994) The Geology of the Mexican Republic (American Association of
Petroleum Geologists, Cincinnati).
Morrone JJ (2006) Biogeographic areas and transition zones of Latin America and the
Caribbean Islands based on panbiogeographic and cladistic analyses of the entomofauna.
Annual Review of Entomology, 51, 467–494.
Mulcahy DG, Mendleson JR (2000) Phylogeography and speciation of the morphologically
variable, widespread species Bufo valliceps, based on molecular evidence from mtDNA.
Molecular Phylogenetics and Evolution, 17, 173–189.
Mulcahy DG, Morrill BH, Mendleson JR (2006) Historical biogeography of lowland species
of toads (Bufo) across the Trans-Mexican Neovolcanic Belt and the Isthmus of
Tehuantepec. Journal of Biogeography, 33, 1889–1904.
Murphy PG, Lugo AE (1986) Ecology of tropical dry forest. Annual Review of Ecology,
Evolution, and Systematics, 17, 67–88.
Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity
hotspots for conservation priorities. Nature, 403, 853–858. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 122
Nylander JAA, Olsson U, Alström P, Sanmartín I (2008) Accounting for phylogenetic
uncertainty in biogeography: a Bayesian approach to dispersal- vicariance analysis of the
thrushes (Aves: Turdus). Systematic Biology, 57, 257-268
Padidam M, Sawyer S, Fauquet CM (1999) Possible emergence of new geminiviruses by
frequent recombination. Virology, 265, 218–225.
Pennington RT, Prado DE, Pendry CA (2000) Neotropical seasonally dry forests and
Quaternary vegetation changes. Journal of Biogeography, 27, 261–273.
Pennington RT, Lavin M, Prado DE, Pendry CA, Pell SK, Butterworth CA (2004) Historical
climate change and speciation: neotropical seasonally dry forest plants show patterns of
both Tertiary and Quaternary diversification. Philosophical Transactions of the Royal
Society of London Series B, 359, 515–537.
Posada D (2001) Unveiling the molecular clock in the presence of recombination. Molecular
Biology and Evolution, 18, 1976–1978.
Posada D (2002) Evaluation of methods for detecting recombination from DNA sequences:
empirical data. Molecular Biology and Evolution, 19, 708–717.
Posada D (2008) jModelTest: phylogenetic model averaging. Molecular Biology and
Evolution, 25, 1253–1256
Pybus OG, Harvey PH (2000) Testing macro-evolutionary models using incomplete
molecular phylogenies. Proceedings of the Royal Society of London Series B, 267, 2267–
2272.
Rabosky DL (2006a) LASER: a maximum likelihood toolkit for detecting temporal shifts in
diversification rates from molecular phylogenies. Evolutionary Bioinformatics Online, 2,
247–250. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 123
Rabosky DL (2006b) Likelihood methods for inferring temporal shifts in diversification
rates. Evolution, 60, 1152–1164.
Rambaut A, Drummond AJ (2007) Tracer v1.6.1. Available from
http://beast.bio.ed.ac.uk/Tracer.
Ramos-Onsins SE, Rozas J (2002) Statistical properties of new neutrality tests against
population growth. Molecular Biology and Evolution, 19, 2092–2100.
Rato C, Carranza S, Harris DJ (2011) When selection deceives phylogeographic
interpretation: the case of the Mediterranean house gecko, Hemidactylus turcius
(Linnaeus, 1758). Molecular Phylogenetics and Evolution, 58, 365–373.
Rzedowski J (1993) Diversity and origins of the phanerogamic flora of México. In Biological
Diversity of México, origins and distribution (Ramamoorthy TP, Bye R, Lot A, Fa J
eds.). Oxford University Press, New York. P. 129–144.
Sauquet H, Ho SYW, Gandolfo MA, Jordan GJ, Wilf P, Cantrill DJ, Bayly MJ, Bromham L,
Brown GK, Carpenter RJ, Lee DM, Murphy DJ, Sniderman JMK, Udovicic F (2012)
Testing the impact of calibration on molecular divergence times using a fossil-rich group:
the case of Nothofagus (Fagales). Systematic Biology, 61, 289–313.
Smouse PE, Long JC, Sokal RR (1986) Multiple regression and correlation extensions of the
Mantel test of matrix correspondence. Systematic Biology, 35, 627–632.
Stamatakis A (2006) RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses
with thousands of taxa and mixed models. Bioinformatics, 22, 2688–2690.
Stamatakis A, Hoover P, Rougemont J (2008) A fast bootstrapping algorithm for the RAxML
web-servers. Systematic Biology, 57, 758–771. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 124
Stephens M, Smith N, Donnelly P (2001) A new statistical method for haplotype
reconstruction from population data. American Journal of Human Genetics, 68, 978–989.
Stephens M, Donnelly P (2003) A comparison of Bayesian methods for haplotype
reconstruction from population genotype data. American Journal of Human Genetics, 73,
1162–1169.
Sullivan J, Arellano E, Rogers DS (2000) Comparative phylogeography of Mesoamerican
highland rodents: concerted versus independent response to past climatic fluctuations.
American Naturalist, 155, 755–768.
Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (2011) MEGA5: Molecular
Evolutionary Genetics Analysis using Maximum Likelihood, Evolutionary Distance, and
Maximum Parsimony Methods. Molecular Biology and Evolution, 28, 2731–2739.
Tajima F (1989) Statistical method for testing the neutral mutation hypothesis by DNA
polymorphism. Genetics, 123, 585–595.
Thorpe RS, Jones AG, Malhotra A, Surget-Groba Y (2008) Adaptive radiation in Lesser
Antillean lizards: molecular phylogenetics and species recognition in the Lesser Antillean
dwarf gecko complex, Sphaerodactylus fantasticus. Molecular Ecology, 17, 1489–1504.
Trejo RI, Dirzo R (2000) Deforestation of seasonally dry tropical forest towards its northern
distribution: a national and local analysis in México. Biological Conservation, 94, 133–
142.
Weir JT (2006) Divergent timing and patterns of species accumulation in lowland and
highland neotropical birds. Evolution, 60, 842–855
Weir JT, Schluter D (2004) Ice sheets promote speciation in boreal birds. Proceedings of the
Royal Society of London Series B, 271, 1881–1887. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 125
Weisrock DW, Smith SD, Chan LM, Biebouw K, Kappeler PM, Yoder AD (2012)
Concatenation and concordance in the reconstruction of mouse lemur phylogeny: an
empirical demonstration of the effect of allele sampling in phylogenetics. Molecular
Biology and Evolution. Advance Access doi: 10.1093/molbev/mss008.
Werneck FP, Costa GC, Colli GR, Prado DE, Sites JW (2011) Revisiting the historical
distribution of seasonally dry tropical forests: new insights based on palaeodistribution
modelling and palynological evidence. Global Ecology and Biogeography, 20, 272–288.
Yu Y, Harris AJ, He XJ (2010) S-DIVA (Statistical Dispersal-Vicariance Analysis) a tool for
inferring biogeographic histories. Molecular Phylogenetics and Evolution, 56, 848–850.
Yu Y, Harris AJ, He XJ (2011) RASP (Reconstruct Ancestral State in Phylogenies) 2.0 beta.
Available at http://mnh.scu.edu.cn/soft/blog/RASP.
Zaldívar-Riverón A, León-Regagnon V, Nieto-Montes de Oca A (2004) Phylogeny of the
Mexican coastal leopard frogs of the Rana berlandieri group based on mtDNA
sequences. Molecular Phylogenetics and Evolution, 30, 38–49.
Zarza E, Reynoso VH, Emerson BC (2008) Diversification in the northern Neotropics:
mitochondrial and nuclear DNA phylogeography of the iguana Ctenosaura pectinata and
related species. Molecular Ecology, 17, 3259–3275.
Zink RM, Klicka J, Barber BR (2004) The tempo of avian diversification during the
Quaternary. Philosophical Transactions of the Royal Society of London Series B, 359,
215–219. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 126
Table 1. Primer information for loci used in this study.
Primer Annealing Temperature Sequence Reference
EnolL 731 60 5'-TGG ACT TCA AAT CCC CCG ATG ATC CCA GC-3' Friesen et al. (1997) EnolH 912 5'-CCA GGC ACC CCA GTC TAC CTG GTC AAA-3'
LamL724 60 5'-CCA AGA AGC AGC TGC AGG ATG AGA TGC-3' Friesen et al. (1997) LamH892 5'-CTG CCG CCC GTT GTC GAT CTC CAC CAG-3'
ND412931L 50 5'-CTA CCA AAA GCT CAT GTA GAA GC-3' Blair et al. (2009) ND413824H 5'-CAT TAC TTT TAC TTG GAT TTG CAC CA-3' Arevalo et al. (1994) ND4phyL 5'-CCT ATC CCC CAC AAC CCA AAC-3' Blair et al. (2009)
CytbB1L 50 5'-CCA TCC AAC ATC TCA GCA TGA TGA AA-3' Kocher et al. (1989) CytbB4H 5'-TTG CTG GGG TGA AGT TTT CTG GGT C-3'
CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 127
Table 2. Mitochondrial DNA sequence divergence between major lineages of leaf-toed geckos (Phyllodactylus) recovered in the phylogenetic analysis. Values below diagonal represent uncorrected p-distances and values above diagonal Tamura-Nei corrected distances.
Lineage A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 B1 B2 B3 B4 B5
A1 — 0.204 0.188 0.237 0.297 0.263 0.248 0.258 0.262 0.273 0.270 0.247 0.283 0.294 0.280 0.288 A2 0.173 — 0.107 0.189 0.248 0.252 0.245 0.242 0.244 0.274 0.252 0.266 0.253 0.252 0.282 0.276 A3 0.161 0.097 — 0.179 0.249 0.236 0.236 0.227 0.224 0.232 0.230 0.263 0.245 0.209 0.276 0.275 A4 0.198 0.163 0.156 — 0.252 0.229 0.224 0.225 0.230 0.229 0.239 0.237 0.257 0.258 0.274 0.269 A5 0.237 0.206 0.206 0.210 — 0.201 0.201 0.180 0.250 0.250 0.238 0.315 0.298 0.265 0.320 0.302 A6 0.217 0.208 0.195 0.193 0.171 — 0.087 0.167 0.175 0.071 0.218 0.311 0.301 0.323 0.310 0.301 A7 0.207 0.203 0.195 0.189 0.170 0.082 — 0.134 0.232 0.231 0.248 0.301 0.294 0.309 0.300 0.292 A8 0.213 0.201 0.191 0.190 0.156 0.146 0.119 — 0.221 0.229 0.223 0.294 0.286 0.287 0.281 0.262 A9 0.218 0.204 0.190 0.194 0.207 0.152 0.195 0.187 — 0.133 0.180 0.258 0.282 0.301 0.292 0.285 A10 0.225 0.224 0.195 0.193 0.207 0.064 0.195 0.193 0.117 — 0.134 0.294 0.303 0.303 0.311 0.293 A11 0.223 0.209 0.193 0.199 0.199 0.183 0.206 0.188 0.155 0.118 — 0.291 0.293 0.294 0.286 0.293 B1 0.205 0.218 0.215 0.197 0.250 0.246 0.240 0.235 0.214 0.237 0.235 — 0.061 0.216 0.151 0.152 B2 0.231 0.210 0.204 0.213 0.240 0.243 0.238 0.233 0.231 0.245 0.238 0.056 — 0.252 0.163 0.150 B3 0.238 0.209 0.177 0.213 0.218 0.257 0.248 0.235 0.244 0.246 0.239 0.179 0.206 — 0.245 0.240 B4 0.230 0.229 0.225 0.225 0.254 0.248 0.242 0.230 0.237 0.251 0.234 0.132 0.143 0.203 — 0.119 B5 0.235 0.225 0.224 0.221 0.243 0.243 0.237 0.218 0.233 0.239 0.239 0.132 0.132 0.199 0.107 — CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 128
Table 3. Diversity statistics and tests for demographic changes for each major mitochondrial lineage of leaf-toed geckos (Phyllodactylus) recovered in the phylogenetic analysis.
Species Lineage n H h S K π Tajima's D Fu's Fs R2 tuberculosus A1 6 5 0.933 10 5.067 0.0042 0.942 -0.327 0.232 tuberculosus A2 10 3 0.689 52 24.000 0.0460 1.496 13.235 0.231 tuberculosus A3 10 3 0.511 3 0.911 0.0016 -0.507 0.300 0.155 tuberculosus A4 4 2 0.500 5 2.500 0.0043 -0.797 2.598 0.433 tuberculosus A6 5 4 0.900 16 6.800 0.0076 -0.840 0.980 0.322 muralis A5 5 4 0.900 8 3.200 0.0057 -1.174 -0.226 0.187 muralis A7 5 4 0.900 5 2.600 0.0030 0.562 -0.567 0.205 muralis A8 9 3 0.667 2 0.778 0.0014 0.196 -0.108 0.210 tuberculosus A9 9 8 0.972 215 64.306 0.0595 -1.349 2.339 0.175 tuberculosus A10 6 2 0.333 1 0.333 0.0010 -0.933 -0.003 0.373 tuberculosus* A11 14 N/A N/A 176 59.516 0.0458 N/A N/A N/A lanei B1 8 2 0.429 1 0.429 0.0007 0.334 0.536 0.214 lanei B2 11 6 0.855 89 30.327 0.0341 -0.062 6.980 0.166 lanei B4 15 11 0.933 20 6.438 0.0054 0.191 -2.358 0.145 lanei B5 29 6 0.507 15 3.874 0.0034 -0.168 2.710 0.124
* signifies that statistics were calculated including sites with gaps or missing data. Bold values indicate significance at P < 0.05.
CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 129
Figure Legends
Fig. 1. Map of collecting localities for all species and subspecies used in this study. Shaded
landscape represents the present distribution of Mexican dry forest. Solid lines represent
previously documented biogeographic barriers.
Fig. 2. Genealogical relationships among representatives of the P. tuberculosus species group
based on maximum likelihood analysis of mtDNA. Values adjacent to nodes represent
bootstrap proportions estimated from 10000 replicates using the fast bootstrapping
algorithm in RAxML.
Fig. 3. Relaxed molecular clock analysis for the P. tuberculosus group based on mtDNA
sequences. Horizontal lines at nodes represent the 95% highest posterior density for
inferred nodal ages. Vertical line represents the hypothesized origin of the Mexican dry
forest (Becerra 2005). Horizontal scale is in million of years ago. Branch colors represent
rate of evolution (blue=slow; red=fast). Lineage Through Time plot illustrates the
accumulation of evolutionary lineages as a function of time.
Fig. 4. Species-tree analysis of the combined mitochondrial and nuclear data. Horizontal
lines at nodes represent the 95% highest posterior density for inferred nodal ages.
Horizontal scale is in million of years ago. Lineage Through Time plot illustrates the
accumulation of evolutionary lineages as a function of time.
Fig. 5. Historical biogeography of the P. tuberculosus species group based on the dated
mtDNA gene-tree. Colors represent the Mexican state where the animal was captured.
Letters at nodes represent ancestral areas with the highest likelihood from Bayesian
reconstructions. Green circles surrounding nodes represent postulated vicariant events. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 130
Dotted lines represent hypothesized biogeographic barriers. I=Sierra Madre Occidental;
II=Mexican Volcanic Belt; III=Isthmus of Tehuantepec.
Fig. 6. Extended Bayesian skyline plots showing demographic trends through time for each
major mtDNA lineage recovered in the phylogenetic analysis. Dashed line represents the
median population size, while solid lines denote 95% confidence intervals. Scale on the
x-axis represents time in millions of years before present, while the y-axis represents log
population size (Ne) x generation time (τ). CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 131
Fig. 1
Alamos (! N
Cosala (! (! VU
Chamela %2, %2, %2 GC Species El Charco Lombardia %2 %2 lanei %2 %, lanei/tuberculosus %2%2 %2 %2P!P!P! !P """"" " muralis P! P!" P!P! P""!"""""" P!P! P! t. magnus P!P! P! P!
(! t. saxatilis 0410820205 Kilometers
%2 TC %2%2%2 PEC%2 PLT%2 %2P! Acapulco P! P! SM JA SPT SJG """"" !P Tehuantepec km221,225 P! P! P!PRLA " Arriaga P! PT """"" P! PMH P!P!""" PN P! To na la P! PChiq PChip PMII P! P! km132 CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 132
Fig. 2 Phyllodactylus xanti CB587_JAL-ElCB587_JAL-El CharcoCharco 100 CB597_JAL-ElCB597_JAL-El CharcoCharco CB575_JAL-ElCB575_JAL-El CharcoCharco CB580_JAL-ElCB580_JAL-El CharcoCharco A1 Phyllodactylus sp. CB583_JAL-ElCB583_JAL-El CharcoCharco CB582_JAL-ElCB582_JAL-El CharcoCharco 100 CB302_SIN-CosalaCB302_SIN-Cosala CB301_SIN-CosalaCB301_SIN-Cosala CB300_SIN-CosalaCB300_SIN-Cosala 35415_SIN-VU35415_SIN-VU 100 35318_SIN-VU35318_SIN-VU 35319_SIN-VU35319_SIN-VU A2 35316_SIN-VU35316_SIN-VU 66 10000 35372_SIN-VU35372_SIN-VU P. tuberculosus 35317_SIN-VU35317_SIN-VU 35320_SIN-VU35320_SIN-VU CB94_SON-AlamosCB94_SON-Alamos 100 CB97_SON-AlamosCB97_SON-Alamos saxatilis CB98_SON-AlamosCB98_SON-Alamos CB101_SON-Alamos CB95_SON-Alamos CB96_SON-AlamosCB96_SON-Alamos CB93_SON-Alamos A3 CB92_SON-Alamos CB99_SON-AlamosCB99_SON-Alamos 80 CB100_SON-AlamosCB100_SON-Alamos 100 CB240_GR-SJGCB240_GR-SJG CB241_GR-SJG CB238_GR-SMCB238_GR-SM A4 P. tuberculosus magnus CB239_GR-SMCB239_GR-SM CB293_OAX-PChipCB293_OAX-PChip 1000 CB291_OAX-PChip CB292_OAX-PChipCB292_OAX-PChip AA55 CB294_OAX-PChipCB294_OAX-PChip P. muralis isthmus CB295_OAX-PChipCB295_OAX-PChip CB278_OAX-PMHCB278_OAX-PMH CB268_OAX-PChiqCB268_OAX-PChiq 87 CB270_OAX-PChiqCB270_OAX-PChiq CB269_OAX-PChiqCB269_OAX-PChiq A6 P. tuberculosus magnus CB271_OAX-PChiqCB271_OAX-PChiq 35269_OAX-SPT35269_OAX-SPT 944 35270_OAX-SPT35270_OAX-SPT 35268_OAX-SPT35268_OAX-SPT AA77 P. muralis muralis 35271_OAX-SPT35271_OAX-SPT 72 35272_OAX-SPT35272_OAX-SPT 35278_OAX-Tehuantepec35278_OAX-Tehuantepec CB286_OAX-PMIICB286_OAX-PMII CB288_OAX-PMIICB288_OAX-PMII 95 1000 CB287_OAX-PMIICB287_OAX-PMII CB280_OAX-PNCB280_OAX-PN CB283_OAX-PNCB283_OAX-PN A8 P. muralis isthmus CB281_OAX-PNCB281_OAX-PN CB284_OAX-PNCB284_OAX-PN CB282_OAX-PNCB282_OAX-PN CB279_OAX-PMHCB279_OAX-PMH CB265_OAX-PRLACB265_OAX-PRLA 919 CB266_OAX-PRLACB266_OAX-PRLA CB264_OAX-PRLACB264_OAX-PRLA CB261_OAX-MX200,CB261_OAX-MX200, km221km221 900 CB259_OAX-MX200,CB259_OAX-MX200, km225km225 A9 CB262_OAX-MX200, km221km221 CB260_OAX-MX200,CB260_OAX-MX200, km225km225 CB263_OAX-MX200,CB263_OAX-MX200, km221km221 CB276_OAX-MX200, km132km132 CB275_OAX-MX200,CB275_OAX-MX200, km132km132 CB267_OAX-PChiqCB267_OAX-PChiq P. tuberculosus CB273_OAX-MX200,CB273_OAX-MX200, km132km132 A10 CB272_OAX-MX200, km132km132 51 CB277_OAX-MX200,CB277_OAX-MX200, km132km132 magnus 100 35279_OAX-Tehuantepec35279_OAX-Tehuantepec 35280_OAX-Tehuantepec35280_OAX-Tehuantepec 100 CB285_OAX-PMIICB285_OAX-PMII 8484 CB289_OAX-PMIICB289_OAX-PMII 100 CB299_CHIAP-NETonalaCB299_CHIAP-NETonala CB298_CHIAP-NETonalaCB298_CHIAP-NETonala 35277_OAX-Arriaga35277_OAX-Arriaga (W)(W) 6464811 35276_OAX-Arriaga35276_OAX-Arriaga (W)(W) 35275_CHIAP-Tonala35275_CHIAP-Tonala AA1111 CB297_CHIAP-PTCB297_CHIAP-PT CB296_CHIAP-PTCB296_CHIAP-PT 35274_CHIAP-Tonala35274_CHIAP-Tonala 35273_CHIAP-Tonala35273_CHIAP-Tonala CB274_OAX-MX200,CB274_OAX-MX200, km132km132
CB581_JAL-ElCB581_JAL-El CharcoCharco 91 CB579_JAL-ElCB579_JAL-El CharcoCharco CB576_JAL-ElCB576_JAL-El CharcoCharco CB578_JAL-ElCB578_JAL-El CharcoCharco CB574_JAL-ElCB574_JAL-El CharcoCharco B1 CB584_JAL-ElCB584_JAL-El CharcoCharco CB573_JAL-ElCB573_JAL-El CharcoCharco CB577_JAL-ElCB577_JAL-El CharcoCharco 100 35440_MICH-Lombardia35440_MICH-Lombardia P. lanei rupinus 35439_MICH-Lombardia35439_MICH-Lombardia 67 CB422_JAL-ChamelaCB422_JAL-Chamela 100 CB427_JAL-ChamelaCB427_JAL-Chamela CB425_JAL-ChamelaCB425_JAL-Chamela CB421_JAL-ChamelaCB421_JAL-Chamela B2 CB424_JAL-ChamelaCB424_JAL-Chamela CB419_JAL-ChamelaCB419_JAL-Chamela CB420_JAL-ChamelaCB420_JAL-Chamela CB428_JAL-ChamelaCB428_JAL-Chamela 92 CB423_JAL-ChamelaCB423__JAL-Chamela 94 8852_GR-GC8852_GR-GC 8850_GR-GC8850_GR-GC B3 35298_GR-TC35298_GR-TC 35294_GR-TC35294_GR-TC 35290_GR-TC35290_GR-TC 35306_GR-TC35306_GR-TC 35307_GR-TC35307_GR-TC 8585 100 35310_GR-TC35310_GR-TC 35305_GR-TC35305_GR-TC 335291_GR-TC5291_GR-TC BB44 35308_GR-TC35308_GR-TC 35297_GR-TC35297_GR-TC 35296_GR-TC35296_GR-TC 35295_GR-TC35295_GR-TC 86 35287_GR-TC35287_GR-TC 35288_GR-TC35288_GR-TC 35289_GR-TC35289_GR-TC CB252_GR-AcapulcoCB252_GR-Acapulco 100 CB250_GR-AcapulcoCB250_GR-Acapulco CB254_GR-AcapulcoCB254_GR-Acapulco CB251_GR-AcapulcoCB251_GR-Acapulco 96 CB253_GR-Acapulco P. lanei lanei CB220_GR-PLTCB220_GR-PLT CB223_GR-PLTCB223_GR-PLT 622 CB219_GR-PLTCB219_GR-PLT CB214_GR-PLTCB214_GR-PLT CB222_GR-PLTCB222_GR-PLT CB216_GR-PLTCB216_GR-PLT CB218_GR-PLTCB218_GR-PLT CB236_GR-PECCB236_GR-PEC CB233_GR-PECCB233_GR-PEC CB217_GR-PLTCB217_GR-PLT CB235_GR-PECCB235_GR-PEC CB234_GR-PECCB234_GR-PEC BB55 CB237_GR-PECCB237_GR-PEC CB221_GR-PLTCB221_GR-PLT CB228_GR-PECCB228_GR-PEC CB229_GR-PECCB229_GR-PEC CB230_GR-PECCB230_GR-PEC CB224_GR-PECCB224_GR-PEC CB226_GR-PECCB226_GR-PEC CB232_GR-PECCB232_GR-PEC CB231_GR-PECCB231_GR-PEC CB227_GR-PECCB227_GR-PEC 0.07 CB225_GR-PECCB225_GR-PEC CB215_GR-PLTCB215_GR-PLT CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 133
Fig. 3
Coleonyx variegatus Phyllodactylus xanti CB581 0.4 1 CB579 B1 CB576 1 1 CB425 P. lanei rupinus 1 CB422 B2 CB427 1 1 8850 8852 B3 1 1 CB252 CB254 B5 P. lanei lanei 0.99 1 CB250 1 35298 35290 B4 35294 1 CB587 CB575 A1 P. sp. CB597 1 CB98 CB97 A3 0.58 1 CB94 P. tuberculosus 1 35319 35318 A2 saxatilis 35415 1 1 CB238 CB241 A4 CB240 1 CB265 CB266 A9 P. tuberculosus 0.81 1 CB264 1 CB289 magnus 1 CB299 A11 1 CB298 1 CB277 1 CB275 A10 CB276 1 CB293 CB291 A5 P. muralis isthmus CB292 1 1 CB270 P. tuberculosus CB271 A6 CB269 0.80.8 magnus 1 35268 35269 A7 0.8 35270 P. muralis muralis 1 CB286 35278 A8 P. muralis isthmus CB288
20.0
150.0 125.0 100.0 75.0 50.0 25.0 0.0
Cretaceous Eocene Oligocene Miocene Pliocene Paleocene Pleistocene
Log-Lineages Through Time
st = 2.67 Ma Log Lineages 1.0 1.5 2.0 2.5 3.0 3.5 4.0
140 120 100 80 60 40 20 0
Million Years Before Present CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 134
Fig. 4
Coleonyx variegatus Phyllodactylus xanti 1 0.98 B1 P. lanei rupinus 1 1 B2 B3
B4 P. lanei lanei 0.84 0.97 0.37 B5 A1 P. sp . A5 P. muralis isthmus 0.67 A3 P. tuberculosus saxatilis 0.53 A8 P. muralis isthmus 0.53 A2 P. tuberculosus saxatilis 0.99 0.46 0.89 A7 P. muralis muralis 0.49 A4
0.94 A9 A11 P. tuberculosus magnus A6 0.85 1 A10
20.0
150.0 125.0 100.0 75.0 50.0 25.0 0.0
Cretaceous Eocene Miocene Pliocene Paleocene Oligocene Pleistocene
Log-Lineages Through Time
st = 28 Ma Log Lineages 1.0 1.5 2.0 2.5 3.0
120 100 80 60 40 20 0
Million Years Before Present CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 135
LEGEND Fig. 5 A. Sonora
B. Sinaloa Coleonyx variegatus CB587 C C. Jalisco C CB575 CB597 P. s p. CB238 D D CB241 P. tuberculosus magnus D. Guerrero CB240 CB293 E E CB291 P. muralis isthmus E. Oaxaca CB292 CB270 I E E C D E CB271 P. tuberculosus magnus CB269 EF. Oaxaca/Chiapas E 35268 E E 35269 P. muralis muralis E 35270 CB286 E E E 35278 P. muralis isthmus CB288 EF CB299 EF CB298 C CB289 II E E CB275 E CB276 III E CB277 P. tuberculosus magnus C E CB266 E CB264 CB265 CB98 A A CB97 A CB94 35319 P. tuberculosus saxatilis B B 35318 35415 Phyllodactylus xanti D 8850 8852 D D CB254 D CB250 C CB252 P. lanei lanei I D D 35290 C D 35294 35298 C CB422 C C CB427 II CB425 III C CB579 P. lanei rupinus C CB576 CB581
140120 100 80 60 40 20 0 Million Years Before Present CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 136
A1 A2 A3 A5 A7 A8
A9 A10 A11 B2 B4 B5
Fig. 6 CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 137
Supplementary Material
Appendix I. Locality information and GenBank accession numbers for all species and populations used in this study.
GenBank Accession Numbers
Voucher α- Species Tissue # # State Location Code cytb ND4 enolase lamin A P. lanei lanei 8850 Guerrero Grutas de Cacahuamilpa GC P. lanei lanei 8851 Guerrero Grutas de Cacahuamilpa GC P. lanei lanei 8852 Guerrero Grutas de Cacahuamilpa GC P. lanei lanei CB0214 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0215 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0216 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0217 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0218 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0219 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0220 42797 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0221 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0222 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0223 42796 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0224 Guerrero MX 200; km 43; Puente Las Tranquitas I PLT P. lanei lanei CB0225 42798 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0226 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0227 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0228 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0229 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0230 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0231 42799 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0232 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0233 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0234 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0235 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0236 Guerrero MX 200; km 60; Puente El Cagaco PEC P. lanei lanei CB0237 Guerrero MX 200; km 60; Puente El Cagaco PEC CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 138
P. lanei lanei CB0250 42807 Guerrero Acapulco; Rio Camaron Acapulco P. lanei lanei CB0251 42808 Guerrero Acapulco; Rio Camaron Acapulco P. lanei lanei CB0252 42809 Guerrero Acapulco; Rio Camaron Acapulco P. lanei lanei CB0253 42810 Guerrero Acapulco; Rio Camaron Acapulco P. lanei lanei CB0254 42811 Guerrero Acapulco; Rio Camaron Acapulco P. lanei lanei ROM 35287 Guerrero Tierra Colorada, 1.1 km down road (hwy) to Ayutla TC P. lanei lanei ROM 35288 45105 Guerrero Tierra Colorada, 1.1 km down road (hwy) to Ayutla TC P. lanei lanei ROM 35289 Guerrero Tierra Colorada, 1.1 km down road (hwy) to Ayutla TC P. lanei lanei ROM 35290 45112 Guerrero Tierra Colorada, 9.1 km W hwy junction TC P. lanei lanei ROM 35291 Guerrero Tierra Colorada, 9.1 km W hwy junction TC P. lanei lanei ROM 35294 45113 Guerrero Tierra Colorada, 1.4 km W of hwy junction TC P. lanei lanei ROM 35295 Guerrero Tierra Colorada, 1.4 km W of hwy junction TC P. lanei lanei ROM 35296 45114 Guerrero Tierra Colorada, 1.4 km W of hwy junction TC P. lanei lanei ROM 35297 Guerrero Tierra Colorada, 1.4 km W of hwy junction TC P. lanei lanei ROM 35298 45109 Guerrero Tierra Colorada, 1.4 km W of hwy junction TC P. lanei lanei ROM 35305 45110 Guerrero Tierra Colorada, 5 km N on old hwy to Mexico City TC P. lanei lanei ROM 35306 Guerrero Tierra Colorada, ~ 1km SE of town centre TC P. lanei lanei ROM 35307 Guerrero Tierra Colorada, ~ 1km SE of town centre TC P. lanei lanei ROM 35308 Guerrero Tierra Colorada, ~ 1km SE of town centre TC P. lanei lanei ROM 35310 Guerrero Tierra Colorada, ~ 1km SE of town centre TC P. lanei rupinus CB0419 Jalisco Chamela Chamela P. lanei rupinus CB0420 Jalisco Chamela Chamela P. lanei rupinus CB0421 Jalisco Chamela Chamela P. lanei rupinus CB0422 Jalisco Chamela Chamela P. lanei rupinus CB0423 Jalisco Chamela Chamela P. lanei rupinus CB0424 Jalisco Chamela Chamela P. lanei rupinus CB0425 Jalisco Chamela Chamela P. lanei rupinus CB0427 Jalisco Chamela Chamela P. lanei rupinus CB0428 Jalisco Chamela Chamela P. lanei rupinus CB0573 Jalisco El Charco house El Charco P. lanei rupinus CB0574 Jalisco El Charco house El Charco P. lanei rupinus CB0576 Jalisco El Charco house El Charco P. lanei rupinus CB0577 Jalisco El Charco house El Charco P. lanei rupinus CB0578 Jalisco El Charco house El Charco P. lanei rupinus CB0579 Jalisco El Charco house El Charco P. lanei rupinus CB0581 Jalisco El Charco house El Charco CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 139
P. lanei rupinus CB0584 Jalisco El Charco house El Charco Michoaca Lombardi P. lanei rupinus ROM 35439 n Lombardia, 10 km N (by road, Hwy 37) a Michoaca Lombardi P. lanei rupinus ROM 35440 n Lombardia, 10 km N (by road, Hwy 37) a P. muralis isthmus CB0280 42837 Oaxaca MX 200; Puente Nizaburra PN P. muralis isthmus CB0281 42838 Oaxaca MX 200; Puente Nizaburra PN P. muralis isthmus CB0282 42839 Oaxaca MX 200; Puente Nizaburra PN P. muralis isthmus CB0283 42840 Oaxaca MX 200; Puente Nizaburra PN P. muralis isthmus CB0284 42841 Oaxaca MX 200; Puente Nizaburra PN P. muralis isthmus CB0286 42843 Oaxaca MX 200; Puente Merro II PMII P. muralis isthmus CB0287 42844 Oaxaca MX 200; Puente Merro II PMII P. muralis isthmus CB0288 42845 Oaxaca MX 200; Puente Merro II PMII P. muralis isthmus CB0291 42848 Oaxaca MX 200; Puente Chipehua Pchip P. muralis isthmus CB0292 42849 Oaxaca MX 200; Puente Chipehua Pchip P. muralis isthmus CB0293 42850 Oaxaca MX 200; Puente Chipehua Pchip P. muralis isthmus CB0294 42851 Oaxaca MX 200; Puente Chipehua Pchip P. muralis isthmus CB0295 42852 Oaxaca MX 200; Puente Chipehua Pchip P. muralis muralis ROM 35268 Oaxaca San Pedro Totolapan, 23.8 km E of town (by road) SPT P. muralis muralis ROM 35269 Oaxaca San Pedro Totolapan, 23.8 km E of town (by road) SPT P. muralis muralis ROM 35270 Oaxaca San Pedro Totolapan, 23.8 km E of town (by road) SPT P. muralis muralis ROM 35271 45088 Oaxaca San Pedro Totolapan, 23.8 km E of town (by road) SPT P. muralis muralis ROM 35272 Oaxaca San Pedro Totolapan, 23.8 km E of town (by road) SPT Tehuante P. muralis isthmus ROM 35278 Oaxaca Tehuantepec, ~16 km W pec P. sp. CB0575 Jalisco El Charco house El Charco P. sp. CB0580 Jalisco El Charco house El Charco P. sp. CB0582 Jalisco El Charco house El Charco P. sp. CB0583 Jalisco El Charco house El Charco P. sp. CB0587 Jalisco El Charco house El Charco P. sp. CB0597 Jalisco El Charco house El Charco P. tuberculosus magnus CB0238 42800 Guerrero MX 200; km 50; 9 km S San Marcos; Puente La Estancia I SM P. tuberculosus magnus CB0239 42801 Guerrero MX 200; km 50; 9 km S San Marcos; Puente La Estancia II SM P. tuberculosus magnus CB0240 42802 Guerrero MX 200; 3 km S San Juan Grande SJG P. tuberculosus magnus CB0241 42803 Guerrero MX 200; 3 km S San Juan Grande SJG P. tuberculosus magnus CB0259 42816 Oaxaca MX 200; km 225 at bridge km225 CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 140
P. tuberculosus magnus CB0260 42817 Oaxaca MX 200; km 225 at bridge km225 P. tuberculosus magnus CB0261 42818 Oaxaca MX 200; km 221 at bridge km225 P. tuberculosus magnus CB0262 42819 Oaxaca MX 200; km 221 at bridge km225 P. tuberculosus magnus CB0263 42820 Oaxaca MX 200; km 221 at bridge km225 P. tuberculosus magnus CB0264 42821 Oaxaca MX 200; Puente Rio Las Arenas PRLA P. tuberculosus magnus CB0265 42822 Oaxaca MX 200; Puente Rio Las Arenas PRLA P. tuberculosus magnus CB0266 42823 Oaxaca MX 200; Puente Rio Las Arenas PRLA P. tuberculosus magnus CB0267 42824 Oaxaca MX 200; Puente Chiquito Pchiq P. tuberculosus magnus CB0268 42825 Oaxaca MX 200; Puente Chiquito Pchiq P. tuberculosus magnus CB0269 42826 Oaxaca MX 200; Puente Chiquito Pchiq P. tuberculosus magnus CB0270 42827 Oaxaca MX 200; Puente Chiquito Pchiq P. tuberculosus magnus CB0271 42828 Oaxaca MX 200; Puente Chiquito Pchiq P. tuberculosus magnus CB0272 42829 Oaxaca MX 200; km 132 at bridge km132 P. tuberculosus magnus CB0273 42830 Oaxaca MX 200; km 132 at bridge km132 P. tuberculosus magnus CB0274 42831 Oaxaca MX 200; km 132 at bridge km132 P. tuberculosus magnus CB0275 42832 Oaxaca MX 200; km 132 at bridge km132 P. tuberculosus magnus CB0276 42833 Oaxaca MX 200; km 132 at bridge km132 P. tuberculosus magnus CB0277 42834 Oaxaca MX 200; km 132 at bridge km132 P. tuberculosus magnus CB0278 42835 Oaxaca MX 200; Puente Manialtepec Hidalgo PMH P. tuberculosus magnus CB0279 42836 Oaxaca MX 200; Puente Manialtepec Hidalgo PMH P. tuberculosus magnus CB0285 42842 Oaxaca MX 200; Puente Merro II PMII P. tuberculosus magnus CB0289 42846 Oaxaca MX 200; Puente Merro II PMII P. tuberculosus magnus CB0296 42853 Chiapas MX 200; Puente Tiltepec PT P. tuberculosus magnus CB0297 42854 Chiapas MX 200; Puente Tiltepec PT P. tuberculosus magnus CB0298 42855 Chiapas MX 200; 17.31 km NE of Tonala by bridge Tonala CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 141
P. tuberculosus magnus CB0299 42856 Chiapas MX 200; 17.31 km NE of Tonala by bridge Tonala P. tuberculosus magnus ROM 35273 Chiapas Tonala; 3-5 km W of Tonala on Hwy 200 to Arriaga Tonala P. tuberculosus magnus ROM 35274 45089 Chiapas Tonala; 3-5 km W of Tonala on Hwy 200 to Arriaga Tonala P. tuberculosus magnus ROM 35275 Chiapas Tonala; 3-5 km W of Tonala on Hwy 200 to Arriaga Tonala P. tuberculosus magnus ROM 35276 45106 Oaxaca 21.3 km (by road) W of Arriaga (in Chiapas) Arriaga P. tuberculosus magnus ROM 35277 Oaxaca 21.3 km (by road) W of Arriaga (in Chiapas) Arriaga P. tuberculosus Tehuante magnus ROM 35279 45096 Oaxaca Tehuantepec, ~30 km W pec P. tuberculosus Tehuante magnus ROM 35280 Oaxaca Tehuantepec, ~30 km W pec P. tuberculosus 10.7km W on Hwy 200 from where road from Playa Ventura (Juan magnus ROM 35283 45111 Guerrero Alvarez) JA P. tuberculosus 10.7km W on Hwy 200 from where road from Playa Ventura (Juan magnus ROM 35284 Guerrero Alvarez) JA P. tuberculosus saxatilis CB0300 Sinaloa Cosala Cosala P. tuberculosus saxatilis CB0301 Sinaloa Cosala Cosala P. tuberculosus saxatilis CB0302 Sinaloa Cosala Cosala P. tuberculosus saxatilis CB100 Sonora Arroyo Aduana Alamos P. tuberculosus saxatilis CB101 Sonora Arroyo Aduana Alamos P. tuberculosus saxatilis CB92 Sonora Arroyo Aduana Alamos P. tuberculosus saxatilis CB93 Sonora Arroyo Aduana Alamos P. tuberculosus saxatilis CB94 Sonora Arroyo Aduana Alamos P. tuberculosus saxatilis CB95 Sonora Arroyo Aduana Alamos P. tuberculosus saxatilis CB96 Sonora Arroyo Aduana Alamos P. tuberculosus saxatilis CB97 Sonora Arroyo Aduana Alamos P. tuberculosus saxatilis CB98 Sonora Arroyo Aduana Alamos P. tuberculosus saxatilis CB99 Sonora Arroyo Aduana Alamos P. tuberculosus saxatilis ROM 35316 Sinaloa Villa Union, 12 km E, (20 km E of Mazatlan) on hwy 40 toward Durango VU P. tuberculosus saxatilis ROM 35317 Sinaloa Villa Union, 12 km E, (20 km E of Mazatlan) on hwy 40 toward Durango VU CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 142
P. tuberculosus saxatilis ROM 35318 Sinaloa Villa Union, 12 km E, (20 km E of Mazatlan) on hwy 40 toward Durango VU P. tuberculosus saxatilis ROM 35319 Sinaloa Villa Union, 12 km E, (20 km E of Mazatlan) on hwy 40 toward Durango VU P. tuberculosus saxatilis ROM 35320 Sinaloa Villa Union, 12 km E, (20 km E of Mazatlan) on hwy 40 toward Durango VU P. tuberculosus saxatilis ROM 35372 Sinaloa Villa Union, 12 km E, (20 km E of Mazatlan) on hwy 40 toward Durango VU P. tuberculosus saxatilis ROM 35415 Sinaloa Villa Union, 12 km E, (20 km E of Mazatlan) on hwy 40 toward Durango VU CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 143
Supplementary Table 1. MIGRATE-N estimates of mutation-scaled effective population sizes and migration rates for Phyllodactylus tuberculosus and P. muralis along with confidence intervals. Locus 'All' signifies that all three loci were used to
calculate values. 4Nm = number of migrants per generation (ӨM).
Locus Parameter 2.50% 25.00% mode 75.00% 97.50% median mean 4Nm
All Theta_A2 0.0001 0.0015 0.0025 0.0033 0.0049 0.0026 0.0026 N/A All Theta_A3 0.0000 0.0003 0.0011 0.0018 0.0033 0.0015 0.0014 N/A All Theta_A5 0.0000 0.0014 0.0041 0.0087 0.0606 0.0098 0.0222 N/A All Theta_A7 0.0000 0.0000 0.0000 0.0005 0.0017 0.0006 0.0008 N/A All Theta_A8 0.0006 0.0028 0.0042 0.0059 0.0108 0.0049 0.0054 N/A All Theta_A9 0.0518 0.0824 0.0981 0.0995 0.1000 0.0830 0.0797 N/A All Theta_A10 0.0000 0.0000 0.0000 0.0005 0.0013 0.0005 0.0000 N/A All Theta_A11 0.0000 0.0017 0.0027 0.0036 0.0049 0.0027 0.0027 N/A All M_A3->A2 0.6700 58.0000 100.3300 246.6700 664.0000 219.6700 268.6500 0.6931 All M_A5->A2 47.3300 97.3300 189.6700 328.6700 796.6700 313.6700 360.4800 0.9300 All M_A7->A2 37.3300 81.3300 155.6700 302.6700 814.0000 338.3300 377.8200 0.9748 All M_A8->A2 0.0000 32.0000 65.6700 190.6700 590.6700 172.3300 217.4700 0.5611 All M_A9->A2 77.3300 253.3300 373.0000 521.3300 823.3300 425.6700 442.7500 1.1423 All M_A10->A2 0.0000 32.6700 76.3300 146.0000 540.0000 130.3300 189.7200 0.4895 All M_A11->A2 0.0000 8.0000 36.3300 78.0000 426.6700 71.0000 125.0700 0.3227 All M_A2->A3 0.0000 2.6700 44.3300 116.0000 590.0000 113.6700 193.4400 0.2611 All M_A5->A3 0.0000 14.6700 56.3300 132.0000 748.6700 259.6700 294.3900 0.3974 All M_A7->A3 0.0000 36.0000 73.0000 137.3300 632.0000 123.6700 193.3400 0.2610 All M_A8->A3 50.6700 80.0000 163.6700 333.3300 887.3300 307.6700 380.7600 0.5331 All M_A9->A3 19.3300 86.0000 167.6700 318.6700 853.3300 278.3300 354.8900 0.4968 All M_A10->A3 0.0000 24.0000 67.0000 116.0000 648.0000 101.0000 184.3600 0.2581 All M_A11->A3 0.0000 40.6700 94.3300 167.3300 737.3300 151.0000 239.8100 0.3357 All M_A2->A5 0.0000 12.6700 40.3300 76.6700 205.3300 66.3300 82.2800 1.8266 All M_A3->A5 28.0000 92.0000 134.3300 174.6700 304.6700 145.6700 155.8900 3.4608 All M_A7->A5 0.0000 25.3300 59.0000 102.6700 266.6700 87.0000 108.6200 2.4114 All M_A8->A5 0.0000 10.6700 55.6700 100.0000 154.0000 93.0000 170.8300 3.7924 All M_A9->A5 32.0000 128.0000 185.0000 243.3300 450.6700 206.3300 226.2800 5.0234 CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 144
All M_A10->A5 0.0000 52.0000 98.3300 148.0000 276.0000 108.3300 121.1100 2.6886 All M_A11->A5 0.0000 24.0000 55.0000 92.0000 254.0000 77.6700 97.9600 2.1747 All M_A2->A7 3.3300 181.3300 311.0000 385.3300 776.6700 295.0000 326.2000 0.2610 All M_A3->A7 0.0000 34.6700 83.0000 146.0000 419.3300 124.3300 159.4200 0.1275 All M_A5->A7 0.0000 23.3300 72.3300 212.6700 673.3300 199.0000 248.5500 0.1988 All M_A8->A7 51.3300 99.3300 142.3300 242.0000 664.0000 318.3300 332.4400 0.2660 All M_A9->A7 66.6700 114.0000 195.6700 428.0000 880.0000 389.0000 424.2300 0.3394 All M_A10->A7 12.6700 47.3300 87.0000 186.0000 548.6700 168.3300 215.0500 0.1720 All M_A11->A7 36.0000 118.0000 186.3300 271.3300 538.0000 232.3300 260.8300 0.2087 All M_A2->A8 0.0000 88.0000 169.0000 254.6700 558.0000 209.0000 241.6500 1.3050 All M_A3->A8 0.0000 23.3300 62.3300 134.6700 440.0000 121.0000 163.3600 0.8821 All M_A5->A8 34.6700 64.0000 121.6700 278.6700 752.0000 259.6700 314.4100 1.6978 All M_A7->A8 0.0000 96.6700 174.3300 260.6700 560.0000 215.6700 249.0500 1.3449 All M_A9->A8 24.0000 112.0000 205.6700 298.0000 556.6700 245.6700 268.2000 1.4483 All M_A10->A8 0.0000 15.3300 52.3300 177.3300 514.6700 166.3300 204.9900 1.1069 All M_A11->A8 7.3300 111.3300 187.6700 322.6700 627.3300 266.3300 297.0300 1.6040 All M_A2->A9 0.0000 23.3300 52.3300 73.3300 138.0000 57.0000 61.9600 4.9382 All M_A3->A9 0.0000 8.0000 25.0000 43.3300 118.6700 36.3300 44.0600 3.5116 All M_A5->A9 0.0000 18.6700 46.3300 79.3300 228.6700 67.0000 84.3100 6.7195 All M_A7->A9 6.6700 40.0000 64.3300 91.3300 179.3300 75.6700 83.9500 6.6908 All M_A8->A9 0.0000 8.6700 32.3300 58.0000 182.6700 50.3300 64.8000 5.1646 All M_A10->A9 0.0000 16.6700 41.6700 76.0000 230.0000 65.6700 85.5300 6.8167 All M_A11->A9 5.3300 44.6700 71.0000 101.3300 186.0000 82.3300 89.0200 7.0949 All M_A2->A10 0.0000 56.6700 119.6700 204.0000 468.6700 171.0000 203.5900 0.0061 All M_A3->A10 0.0000 27.3300 61.6700 166.0000 502.6700 149.6700 191.3400 0.0057 All M_A5->A10 20.0000 46.6700 108.3300 279.3300 799.3300 261.6700 325.3600 0.0098 All M_A7->A10 63.3300 116.0000 150.3300 335.3300 700.0000 304.3300 339.2200 0.0102 All M_A8->A10 34.6700 164.0000 244.3300 398.0000 685.3300 319.0000 341.7400 0.0103 All M_A9->A10 65.3300 156.0000 253.6700 450.0000 640.0000 381.0000 438.1400 0.0131 All M_A11->A10 0.0000 26.6700 72.3300 161.3300 521.3300 145.0000 192.3200 0.0058 All M_A2->A11 0.0000 45.3300 80.3300 164.0000 399.3300 140.3300 165.5000 0.4469 All M_A3->A11 0.0000 8.0000 35.6700 72.0000 218.0000 65.0000 81.8500 0.2210 All M_A5->A11 12.0000 50.0000 87.6700 189.3300 632.6700 170.3300 227.6400 0.6146 All M_A7->A11 0.0000 56.0000 96.3300 150.6700 415.3300 129.0000 164.9900 0.4455 All M_A8->A11 3.3300 48.0000 140.3300 226.0000 623.3300 202.3300 248.2900 0.6704 All M_A9->A11 0.0000 23.3300 109.0000 173.3300 571.3300 155.6700 213.7500 0.5771 CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 145
All M_A10->A11 18.0000 43.3300 110.3300 216.0000 579.3300 199.6700 252.5100 0.6818 CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 146
Supplementary Table 2. MIGRATE-N estimates of mutation-scaled effective population sizes and migration rates for Phyllodactylus lanei along with confidence intervals. Locus 'All' signifies that all three loci were used to calculate values. 4Nm = number of migrants per generation (ӨM).
Locus Parameter 2.50% 25.00% mode 75.00% 97.50% median mean 4Nm
All Theta_B1 0.0000 0.0000 0.0000 0.0008 0.0022 0.0008 0.0006 N/A All Theta_B2 0.0000 0.0009 0.0021 0.0028 0.0041 0.0020 0.0021 N/A All Theta_B4 0.0010 0.0032 0.0050 0.0072 0.0119 0.0060 0.0062 N/A All Theta_B5 0.0055 0.0086 0.0105 0.0127 0.0176 0.0111 0.0113 N/A All M_B2->B1 96.0000 639.3300 867.6700 982.6700 994.0000 549.0000 532.7700 0.3196 All M_B4->B1 0.0000 0.0000 23.6700 74.6700 427.3300 75.0000 126.4900 0.0759 All M_B5->B1 0.0000 0.0000 5.0000 46.6700 334.0000 47.0000 89.3900 0.0536 All M_B1->B2 0.0000 0.0000 13.0000 80.6700 676.0000 81.0000 181.7800 0.3817 All M_B4->B2 0.0000 4.6700 33.0000 62.6700 206.6700 58.3300 76.5300 0.1607 All M_B5->B2 0.0000 0.0000 21.0000 43.3300 194.6700 43.6700 62.2300 0.1307 All M_B1->B4 0.0000 0.0000 20.3300 53.3300 180.0000 53.6700 67.4400 0.4181 All M_B2->B4 0.0000 0.0000 24.3300 62.0000 200.0000 62.3300 76.3300 0.4732 All M_B5->B4 0.0000 3.3300 30.3300 57.3300 173.3300 54.3300 66.5100 0.4124 All M_B1->B5 0.0000 0.0000 25.6700 70.6700 236.0000 71.0000 88.8900 1.0045 All M_B2->B5 0.0000 6.6700 29.0000 50.6700 124.0000 44.3300 50.6000 0.5718 All M_B4->B5 0.0000 35.3300 71.0000 126.6700 267.3300 106.3300 119.8800 1.3546 CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 147
Supplementary Figure Legends
Fig. S1. Relationships among representatives of the P. tuberculosus species group based on
maximum likelihood analysis of phased alpha-enolase haplotypes. For ease of
visualization only unique haplotypes are presented. Values adjacent to nodes represent
bootstrap proportions estimated from 10000 replicates using the fast bootstrapping
algorithm in RAxML. I=Sierra Madre Occidental; II=Mexican Volcanic Belt; III=
Isthmus of Tehuantepec.
Fig. S2. Relationships among representatives of the P. tuberculosus species group based on
maximum likelihood analysis of phased lamin-A haplotypes. For ease of visualization
only unique haplotypes are presented. Values adjacent to nodes represent bootstrap
proportions estimated from 10000 replicates using the fast bootstrapping algorithm in
RAxML. I=Sierra Madre Occidental; II=Mexican Volcanic Belt; III= Isthmus of
Tehuantepec.
Fig. S3. Relationships among representatives of the P. tuberculosus species group based on
maximum likelihood analysis of the concatenated data. For concatenation with mtDNA a
single phased nuclear haplotype was selected from both intron alignments for each
individual. Values adjacent to nodes represent bootstrap proportions estimated from
10000 replicates using the fast bootstrapping algorithm in RAxML. CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 148
Fig. S1
Phyllodactylus xanti I
94 P. tuberculosus saxatilis 88 (Sonora, Sinaloa) II 59 III 96 58 67
69 70 P. lanei lanei (Guerrero) P. lanei rupinus(Jalisco, Michoacan)
88
92 I
II 53
III
51
100 55 100
P. tuberculosus magnus(Guerrero, Oaxaca, Chiapas) P. muralis muralis(Oaxaca) P. muralis isthmus(Oaxaca)
81 67
69
92
I 52 98
64 71 100 II 92 III
96
63 69
67
0.02 CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 149
Fig. S2
Phyllodactylus xanti 99 54 P. tuberculosus saxatilis (Sonora, Sinaloa, Jalisco) 97 73 P. tuberculosus magnus (Oaxaca) P. muralis muralis (Oaxaca) 100 P. muralis isthmus (Oaxaca) 99 71
68 72 54 I
57 100 II III 53 62
84 59 62
P. lanei lanei (Guerrero) 100 78 P. lanei rupinus (Jalisco, Michoacan) 88 95 P. muralis isthmus (Oaxaca) 66 61 I 87
II III
83 55 P. tuberculosus magnus (Guerrero, Oaxaxa, Chiapas) P. lanei lanei (Guerrero)
98 I
II 53 III
0.0060
CHAPTER 4—MOLECULAR PHYLOGENETICS OF P. TUBERCULOSUS GROUP 150
Fig. S3
Phyllodactylus xanti 335318_SIN-VU5318_SIN-VU 100 335415_SIN-VU5415_SIN-VU 35316_SIN-VU35316_SIN-VU 577 35317_SIN-VU35317_SIN-VU A2* 66 335320_SIN-VU5320_SIN-VU 1001 0 35372_SIN-VU35372_SIN-VU 335319_SIN-VU5319_SIN-VU CB302_SINCB302_SIN-Cosala-Cosala CCB300_SIN-CosalaB300_SIN-Cosala A2*A2* 85 CB301_SIN-CosalaCB301_SIN-Cosala P. tuberculosus saxatilis CB98CB98_SON-Alamos_SON-Alamos 71 CCB100_SON-AlamosB100_SON-Alamos 100 CCB99_SON-AlamosB99_SON-Alamos CB93_SON-Alamos CB101_SON-AlamosCB101_SON-Alamos CB92CB92_SON-Alamos_SON-Alamos A33 CB95_SON-AlamosCB95_SON-Alamos CB96_SON-AlamosCB96_SON-Alamos 66 CB97_SON-AlamosCB97_SON-Alamos CB94_SON-AlamosCB94_SON-Alamos 7788 CB239_GR-SMCB239_GR-SM CB238_GR-SMCB238_GR-SM 533 CB240_GR-SJGCB240_GR-SJG A4 P. tuberculosus magnus 1000 CB241CB241_GR-SJG_GR-SJG CB268_OAX-PChiqCB268_OAX-PChiq CB269CB269_OAX-PChiq_OAX-PChiq CB270_OAX-PChiqCB270_OAX-PChiq A6 P. tuberculosus magnus CB271_OAX-PChiqCB271_OAX-PChiq CB2CB27878 35270_OAX-SPT35270_OAX-SPT 1000 35269_OAX-SPT35269_OAX-SPT 92 335268_OAX-SPT5268_OAX-SPT 55 100 A7A7 335272_OAX-SPT5272_OAX-SPT P. muralis muralis 335271_OAX-SPT5271_OAX-SPT 800 1000 CB292_OAX-PChipCB292_OAX-PChip 97 CB294_OAX-PChipCB294_OAX-PChip CB293_OAXCB293_OAX-PChip-PChip A5A5 CB291CB291_OAX-PChip_OAX-PChip 82 CB295_OAX-PChipCB295_OAX-PChip 35278_OAX-Tehuantepec35278_OAX-Tehuantepec 99 1000CB28CB287_OAX-PMII7_OAX-PMII CB288_OAX-PMIICB288_OAX-PMII P. muralis isthmus 72 CB286_OAX-PMII CB281_OAX-PNCB281_OAX-PN A8 CB284_OAXCB284_OAX-PN-PN CB282_OAX-PNCB282_OAX-PN CB280CB280_OAX-PN_OAX-PN CB283CB283_OAX-PN_OAX-PN CB279_OAX-PMHCB279_OAX-PMH 588 CB265_OAX-PRLACB265_OAX-PRLA CCB264_OAX-PRLAB264_OAX-PRLA 89 977 CCB266_OAX-PRLAB266_OAX-PRLA CB260_OAX-MX200,CB260_OAX-MX200, km225km225 A9A9 777 CB263_OAX-MX200, km221km221 CB262_OAX-MX200, km22km2211 CB259_OAX-MX200,CB259_OAX-MX200, km225km225 CB261_OAX-MX200, km221km221 CB267_OAX-PChiqCB267_OAX-PChiq CB277_OAX-MX200,CB277_OAX-MX200, km132km132 CCB272_OAX-MX200,B272_OAX-MX200, km132km132 CB276_OAX-MX200,CB276_OAX-MX200, km132km132 A10 CB274_OAX-MX200,CB274_OAX-MX200, km132km132 CB273_OAX-MX200,CB273_OAX-MX200, km132km132 P. tuberculosus magnus CB275_OAX-MX200,CB275_OAX-MX200, km132km132 100 35279_OAX-Tehuantepec35279_OAX-Tehuantepec 35280_OAX-Tehuantepec35280_OAX-Tehuantepec 644 100 CCB289_OAX-PMIIB289_OAX-PMII 73 CB285_OAX-PMIICB285_OAX-PMII 100 CB299_CHIAP-NETonalaCB299_CHIAP-NETonala CB298_CHIAP-NETonalaCB298_CHIAP-NETonala 5151 35277_OAX-Arriaga35277_OAX-Arriaga (W) 35276_OAX-Arriaga35276_OAX-Arriaga (W) 35273_CHIAP-Tonala35273_CHIAP-Tonala A11A11 35275_CHIAP-Tonala35275_CHIAP-Tonala CB296_CHIAP-PTCB296_CHIAP-PT 35274_CHIAP-Tonala35274_CHIAP-Tonala CB297_CHIAP-PTCB297_CHIAP-PT 35283_GR-JA35283_GR-JA 35284_GR-JA35284 GR-JA CB587CB587_JAL-El_JAL-El CharcoCharco 1000 CBCB597_JAL-El597_JAL-El CharcoCharco CBCB575_JAL-El575_JAL-El CharcoCharco CCB583_JAL-ElB583_JAL-El CharcCharcoo A1*A1* P. s p . 59 CB580_JAL-ElCB580_JAL-El CharcoCharco CB582CB582_JAL-El_JAL-El CharcoCharco CB581_JAL-ElCB581_ JAL- El Charco Charco 959 CBCB579_JAL-El579_JAL-El ChCharcoarco CB578CB578_JAL-El_JAL-El CharcoCharco CB584_JAL-El CharcoCharco CBCB576_JAL-El576_JAL-El CharcoCharco B1 CB574_JAL-El CharcoCharco CCB573_JAL-ElB573_JAL-El CharcCharcoo CBCB577_JAL-El577_JAL-El CharcoCharco CCB427_JAL-ChamelaB427_JAL-Chamela 1000 CB422CB422_JAL-Chamela_JAL-Chamela CB425_JAL-ChamelaCB425_JAL-Chamela P. lanei rupinus 46 CB421_JAL-ChamelaCB421_JAL-Chamela CB424_JAL-ChamelaCB424_JAL-Chamela CCB419_JAL-ChamelaB419_JAL-Chamela B2 CB420_JAL-ChamelaCB420_JAL-Chamela 95 CB428_JAL-ChamelaCB428_JAL-Chamela CB423_JAL-ChamelaCB423_JAL-Chamela 1000 35440_MICH-Lombardia35440_MICH-Lombardia 35439_MICH-LombardiaCH Lombardia B2B2 * 96 888851_GR-GC51_GR-GC 8850_GR-GC8850_GR-GC B3 88852_GR-GC852_GR-GC 335295_GR-TC5295_GR-TC 10000 35296_GR-TC35296_GR-TC 35297_GR-TC35297_GR-TC 335287_GR-TC5287_GR-TC 93 335308_GR-TC5308_GR-TC 35288_GR-TC35288_GR-TC 35289_GR-TC35289_GR-TC 35291_GR-TC35291_GR-TC 335305_GR-TC5305_GR-TC B4 335310_GR-TC5310_GR-TC 91 35307_GR-TC35307_GR-TC 35306_GR-TC35306_GR-TC 35298_GR-TC35298_GR-TC 35294_GR-TC35294_GR-TC 35290_GR-TC35290_GR-TC CB250_GR-AcapulcoCB250_GR-Acapulco 100 CB254_GR-AcapulcoCB254_GR-Acapulco CB252_GR-AcapulcoCB252_GR-Acapulco CB251_GR-AcapulcoCB251_GR-Acapulco CB253CB253_GR-Acapulco_GR-Acapulco 98 CB229_GR-PEC CB226_GR-PEC P. lanei lanei CB214_GR-PLTCB214_GR-PLT CB235_GR-PEC 977 CB236_GR-PECCB236_GR-PEC CB231_GR-PEC CB234_GR-PEC CB232_GR-PECCB232_GR-PEC CCB233_GR-PECB233_GR-PEC CB217_GR-PLTCB217_GR-PLT CB221CB221_GR-PLT_GR-PLT B5 CB237_GR-PEC CB227_GR-PECCB227_GR-PEC CCB222_GR-PLTB222_GR-PLT CB224_GR-PECCB224_GR-PEC CB215_GR-PLTCB215_GR-PLT CB216_GR-PLT CB219_GR-PLTCB219_GR-PLT CB22CB225_GR-PEC5_GR-PEC CB230_GR-PECCB230_GR-PEC CB228_GR-PECCB228_GR-PEC CB218_GR-PLTCB218_GR-PLT CB220CB220_GR-PLT_GR-PLT 0.05 CB223_GR-PLTCB223_GR-PLT CHAPTER 5
Consideration of historical demographic processes in
landscape genetic studies of gene flow: a case study with
Mexican leaf-toed geckos
Abstract
It is critical to disentangle the relative influence of historical versus contemporary processes
that shape the spatial distribution of genetic variation. I employ different molecular marker
types and analytical methods in a landscape genetic context to understand the genetic
consequences of recent habitat fragmentation using leaf-toed geckos (Phyllodactylus) from
northern Mexico. Mitochondrial DNA sequences obtained from 77 individuals are
synthesized with microsatellite DNA data from 336 geckos to reveal substantial genetic
differentiation over a relatively small geographic area (<40 km). MtDNA sequence
divergences of up to 6% are resolved. I find a 10-fold increase in contemporary versus historical migration rates based on microsatellite data, and evidence that many populations suffer from reduced effective population sizes. Landscape genetic analyses find a stronger correlation between landscape structure and contemporary versus historical migration, suggesting that individuals are altering their dispersal routes in response to recent habitat changes. Finally, my analyses suggest that mtDNA is a poor metric for inferring contemporary landscape genetic relationships and that the utility of Bayesian methods needs to be evaluated further. This study suggests that both long-term female philopatry and recent
151 CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 152 fragmentation contribute to the high degree of differentiation observed over fine-spatial scales in leaf-toed geckos.
1 Introduction
Rates and patterns of dispersal, dispersion, and gene flow can be influenced by a variety of
intrinsic and extrinsic factors including species-specific philopatry, predation, competition,
physiological tolerances, and landscape characteristics (Ricketts 2001; Manel et al. 2003;
Storfer et al. 2010). Habitat fragmentation and loss due to both natural and anthropogenic
forces can be a substantial threat to the maintenance and connectivity of populations
(Sanderson et al. 2002; Ewers & Didham 2005). Maintaining genetic connectivity is essential
to allow the free exchange of neutral and adaptive alleles between demes. Processes that disrupt or prevent dispersal and gene flow could lead to inbreeding depression, the accumulation of deleterious alleles, reduced adaptive potential, and local extirpation (Fahrig
2003; Crispo et al. 2011).
The field of landscape genetics (Manel et al. 2003) is renewing interest in contemporary ecological processes that drive the genetic structure of natural populations
(Sork & Waits 2010; Storfer et al. 2010). Although many studies profoundly enhance our understanding of the spatial distribution of genetic variation, most studies do not examine the influence of historical demographic processes on contemporary genetic structure (e.g. Spear et al. 2005; Vignieri 2005; Wang 2009; Goldberg & Waits 2010). I note that by historical I
refer to processes operating over geologic time and not within the past few dozen
generations. Temporal components of landscape genetics studies that drive pattern-process CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 153
relationships need to be identified (Anderson et al. 2010; Wang 2010). For example, based on simulations Landguth et al. (2010) find that historical processes may confound contemporary landscape-genetic inference in species with small dispersal abilities. Thus, recent efforts have sought to disentangle the relative influence of historical and contemporary landscape processes on contemporary population genetic structure (Vandergast et al. 2007;
Zellmer & Knowles 2009; Dyer et al. 2010). One particularly valuable approach utilizes powerful statistical analyses to detecting migration rates operating over different timescales
(Beerli & Felsenstein 2001; Wilson & Rannala 2003; Beerli 2006), although this approach is not often applied in many contemporary studies. Combining estimates of historical and contemporary population genetic parameters provides a powerful means to determine if recent habitat fragmentation has had detrimental impacts on population size and connectivity
(e.g. Chiucchi & Gibbs 2010). Although the comparative demographics approach is promising, most studies directly comparing historical versus contemporary demographic parameters focus on temperate systems (Anderson et al 2009; Schmidt et al. 2009; Chiucchi
& Gibbs 2010). Few studies center explicitly on tropical habitats yet the majority of Earth’s biodiversity resides in these environments (Myers et al. 2000). Further, most studies have sought to compare historical and contemporary parameters in an aspatial context, disregarding the effects of landscape configuration on functional connectivity.
Although generally not as diverse as tropical rain forests, the tropical deciduous forests (TDF) of western Mexico are home to an enormous biodiversity including many endemic species (Robichaux &Yetman 2000; Myska 2007). In contrast to tropical moist or CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 154
rain forests, TDFs are characterized by distinct wet/dry seasons, with the latter lasting up to
10 months in some regions (Murphy & Lugo 1996; Robichaux & Yetman 2000). Functional and structural differences between these forest types include differences in plant phenology, canopy height, ground cover, soil chemistry, net primary productivity, and the proportion of thorny plant species (Murphy & Lugo 1996; Pennington et al. 2000). The geographic range of Mexican TDF spans most of the Pacific versant of the Sierra Madre Occidental and Sierra
Madre del Sur, with isolated stands scattered throughout eastern Mexico including the
Yucatán Peninsula (Robichaux & Yetman 2000; Becerra 2005; Cartron et al. 2005). These forests originated sometime during the Oligocene due to the uplift of the western sierras; forest maintenance relies heavily on these mountain ranges blocking cold fronts from the north (Becerra 2005). TDF is commonly encountered at elevations less than 2 000 m, where they gradually transition at higher elevations to pine/oak and pine forests.
The floral and faunal diversity within these forests is immense. Unfortunately, conservation efforts emphasize tropical rain forests and high elevation pine and pine-oak habitats, specifically throughout areas of the Mexican Volcanic Belt (Pennington et al. 2000;
Robichaux & Yetman 2000; García 2006). Whereas rain forest biodiversity is substantial, only a small portion of Mexico can be classified as tropical rain forest while TDF is a predominant habitat-type (Becerra 2005). In order to conserve Mexico’s TDF, both broad and fine-scale genetic studies are required to assess spatial patterns of diversity and the processes generating them. Molecular genetic data can be used to detect local and widespread patterns that can be used to inform conservation priorities, including distinguishing between CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 155
source and sink areas (Becerra & Venable 2008).
Leaf-toed geckos, genus Phyllodactylus (Phyllodactylidae), comprise approximately
50 species distributed throughout the New World Tropics, with most Mexican species concentrated in coastal Pacific states south of the Trans-Mexican Volcanic Belt (TMVB)
(Dixon 1964). These lizards are common on vertical surfaces throughout TDF habitat. Recent molecular, karyological and morphological work on the genus documents the presence of cryptic diversity, diagnosable only with a careful selection of molecular and morphological characters (Blair et al. 2009; Castiglia et al. 2009; Murphy et al. 2009; Blair et al. unpublished). Molecular evidence also suggests that the evolution of these lizards is tied to the formation of Mexican TDF during the Oligocene (Blair et al. unpublished).
Phyllodactylus tuberculosus Wiegmann 1835 (yellow-bellied gecko) currently exhibits the broadest geographic range of any leaf-toed gecko. The species is distributed throughout the tropical thornscrub and dry forests of the Neotropics from northwestern Mexico to Costa
Rica (Dixon 1964). Although Mosauer (1936) reports little morphological variation, Dixon
(1964) documents substantial variation in several meristic characters, leading to the possibility of potential cryptic species. In P. tuberculosus from near Alamos, Sonora rates of gene flow are correlated with forest structure, stream connectivity, undisturbed habitat, slope, and temperature (Blair et al. unpublished). A multilocus phylogenetic analysis of the P. tuberculosus species group (P. tuberculosus, P. lanei, P. muralis) throughout western
Mexico suggests that species richness is far greater than currently recognized (Blair et al. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 156
unpublished). Based on data from mtDNA, nDNA, and previous morphological characters, these authors elevate the subspecies P. tuberculosus saxatilis to species.
Herein I use microsatellite data and de novo mtDNA sequence data to test for confounding effects of genealogical history on landscape-genetic inference. Subsequently, I compare and contrast historical and contemporary demographic parameters for multiple populations of P. saxatilis throughout Alamos to test hypotheses related to the following: (i) cryptic mtDNA structure resulting from historical demographic processes; (ii) historical vs. contemporary migration rates and effective population sizes; (iii) bottlenecking; (iv) landscape structure correlates with contemporary vs. historical migration; and (v) response metric influences on landscape genetic relationships. This multi-tiered approach utilizing different molecular markers and analytical methods allows for robust conclusions to be drawn regarding the demographic consequences of recent habitat fragmentation.
2 Materials and methods
2.1 Study site and tissue sampling
All sampling localities were mapped in Figure 1. Microsatellite data (10 loci) for 336
individuals (mean=28 individuals per locality) were obtained from Blair et al. (unpublished).
I screened 77 individuals for mtDNA polymorphism to test for confounding historical
processes influencing population demography (Fig. 1; Supplementary Appendix I).
Individuals screened for mtDNA were chosen based on location and previous microsatellite CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 157
results. All tissues were preserved in the field in 95% ethanol. Vouchers were fixed in the field with 10% formalin and eventually transferred to 70% ethanol and deposited either in the
Laboratorio de Herpetologia, Instituto de Biologia, Universidad Nacional Autónoma de
México, or the Royal Ontario Museum. Fieldwork was conducted during the summers of
2007 and 2008. Sequences were also obtained from P. xanti (Blair et al. 2009) and de novo for P. homolepidurus to root all mtDNA networks.
2.2 DNA extraction, amplification and sequencing
Total genomic DNA was extracted from liver or muscle tissue using standard Proteinase K
and phenol-chloroform protocols (Sambrook et al. 1989; Hillis et al. 1996). I amplified and
sequenced 2400 base pairs (bp) of the mitochondrial genome encompassing partial sequences
of cytochrome c oxidase subunit I (COI), NADH dehydrogenase subunit 4 (ND4), and16S rRNA. Sequences were also obtained for the entire NADH dehydrogenase subunit I (NDI)
gene and the flanking tRNAs leucine, isoleucine, and glutamine. Amplified gene fragments,
number of corresponding bp, primer sequences, and references were presented in
Supplementary Table 1. PCR reactions (25 µl) were composed of the following reaction mix:
18.55 µl ddH20, 1 µl 10 mM primer (forward and reverse), 2.5 µl 10x buffer with 1.5 mM
MgCl2, 0.8 µl 10mM dNTPs, 0.15 µl 5 U Taq DNA Polymerase (Boehringer Mannheim),
and 1 µl template DNA. Amplification was performed on a GeneAmp PCR System 9700 or
Eppendorf AG 5345 thermal cycler using 38 cycles with the following conditions: initial
denaturation of 94°C (2 min) followed by 94°C (30 s), 50–55°C (45 s), 72°C (45 s), with a CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 158
final extension temperature of 72°C (7 min). Amplified DNA products were separated on a
1% agarose gel stained with ethidium bromide and visualized under ultraviolet (UV) light.
Visible bands were excised and filtered through spin columns for 10 min via centrifugation.
The resulting solution was used as the template for sequencing reactions.
Sequencing reactions (10 µl) were performed on an Eppendorf AG 5345 thermal cycler using the BigDye Terminator v 3.0 Cycle Sequencing Kit (Applied Biosystems) with the following reaction mix: 1 µl BigDye, 2 µl 5x BigDye Terminator Buffer, 2 µl ddH20, 1 µl of 10 mM primer solution (same as PCR), and 4 µl PCR product. Sequencing conditions were run with the following conditions: initial denaturation of 96°C (1 min), followed by 25 cycles of 96°C (10 s), 50°C (5 s), 60°C (4 min), and 4°C indefinitely. All fragments were sequenced in both directions. Following sequencing reactions, samples were cleaned, ethanol precipitated, and visualized on an ABI 3100 Automated Sequencer (Applied Biosystems).
2.3 Matrilineal genealogy
Mitochondrial DNA sequences were initially imported and edited in BIOEDIT 5.0.6 (Hall
1999). Forward and reverse gene fragments were manually combined and multiple sequence
alignments were performed manually in BIOEDIT. Individual datasets were exported into
MACCLADE 4.0 (Maddison & Maddison 2005) to identify and define codon positions for
each protein-coding gene. Because the amplified ND1 fragment encompassed several genes
(Leaché & McGuire 2006), I used the complete mitochondrial genome of Gekko gecko (NC
007627) to determine gene position. The MACCLADE nexus file was used for genealogical CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 159
inference.
I reconstructed genealogical relationships using both maximum parsimony (MP) and
Bayesian inference (BI). MP was based on the concatenated mtDNA dataset implemented using TNT 1.1 (Goloboff et al. 2008). I employed the parsimony ratchet (Nixon 1999) and tree drifting with subsequent rounds of tree fusing (Goloboff 1999; Goloboff et al. 2008) using default parameters. All characters were treated as nonadditive and equally weighted, and gaps were treated as missing data. Support for nodes was assessed from 10,000 nonparametric bootstrap replicates using a “fast” stepwise heuristic search (Felsenstein 1985) implemented in PAUP* v.4.0b10 (Swofford 2002), retaining groups with >50% nodal support. I assumed a bootstrap value >70 indicated reliability (Hillis & Bull 1993).
Bayesian inference (BI; Huelsenbeck & Ronquist 2001; Huelsenbeck et al. 2002;
Nylander et al. 2004) was used to reconstruct genealogical relationships. I implemented a partitioned analysis (Nylander et al. 2004; Brandley et al. 2005) with three strategies: the entire dataset, by gene, and by gene and codon position for protein-coding loci. Bayes
Factors (BF; Kass & Raftery 1995), obtained by using the sump command in MRBAYES 3.1.2
(Huelsenbeck & Ronquist 2001), were used to select the best partitioning scheme with a BF12
>10 indicating a preference one partition over another (Brandley et al. 2005). Because rate variation among partitions significantly influence parameter estimation in posterior distributions (Marshall et al. 2006) with regard to branch length estimates, I allowed rates to vary by using the command prset applyto=(all) ratepr=variable in MRBAYES.
MRMODELTEST 2.2 (Nylander 2004) was used to select the best evolutionary model CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 160
for each data partition using the Akaike Information Criterion (AIC; Akaike 1974, 1979). BI was performed using MRBAYES. Two simultaneous runs of six chains were implemented for
1 x 106–3 x 106 generations, depending on partitioning strategy, while sampling every 100
generations. Stationarity was assessed when log–likelihood scores versus generation reached
a stable equilibrium using the sump command in MRBAYES. I monitored convergence using
Tracer 1.5 (Rambaut & Drummond 2007) with target ESS values of at least 200. A burn-in
value of (25%) was used to discard topologies with sub-optimal likelihood scores prior to
generating a 50% majority rule consensus tree. Nodal support within the consensus tree was
assessed from the posterior distribution of topologies (Erixon et al. 2003). Although
Bayesian posterior probabilities (BPP) have often provided relatively high support for incorrect clades (Suzuki et al. 2002), a BPP value >95 with a bootstrap value >70 was assumed to indicate a real biological entity.
2.4 Population genetic analysis of mtDNA
DNASP 5.0 (Rozas et al. 2003) was used to estimate a suite of population genetic diversity
indices from mtDNA data for each population including the number of haplotypes, haplotype diversity and its variance (Nei 1987), nucleotide diversity (π; Nei 1987), the average number of nucleotide differences k (Tajima 1983) and total variance (stochastic + sampling)
assuming no recombination (Tajima 1993), and the number of segregating sites S (Tajima
1989). Fu’s Fs (Fu 1997) tests were used to test for the presence of demographic expansion.
Fu’s Fs was assumed to be the more powerful test for detecting population growth (Fu 1997) CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 161
than the raggedness index (Harpending 1994). Significantly negative values of Fu’s Fs
(compared to neutral expectations under the coalescent) indicated an excess number of rare alleles compared to θ (estimated from π) and suggested a recent population expansion (Fu
1997). DNASP was used to estimate the degree of genetic divergence between populations
using both the average number of nucleotide differences between populations and the
average number of nucleotide substitutions per site between populations (Nei 1987).
A Mantel test and Reduced Major Axis (RMA) regression analysis were implemented
to detect the presence of significant isolation-by-distance (IBD) using the IBD Web Service
3.16 (Jensen et al. 2005). The analysis was run for 30,000 randomizations using PhiST/(1-
PhiST) as a measure of genetic distance, with PhiST estimated from uncorrected p-distances
(Rousset 1997).
2.5 Historical and contemporary migration rates
A multi-tiered analytical approach was used to compare contemporary demographic
parameters and historical patterns. MIGRATE 3.2.16 (Beerli & Felsenstein 2001; Beerli 2006)
was used to estimate historical Ne and asymmetric migration rates between populations based
on the microsatellite data. MIGRATE employed coalescence in either a Bayesian or maximum likelihood framework to calculate Ө and M, where Ө represented the mutation-scaled effective population size (4Neµ for a nuclear locus, Neµ for mtDNA) and M represented the mutation-scaled immigration rate (m/µ). I estimated the effective number of immigrants per generation over thousands of years or approximately 4Ne generations in the past (Beerli
2008) by multiplying Ө and M. The FST calculation method was used to generate staring CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 162 values for both Ө and M while specifying uniform priors for both parameters with a minimum of zero, mean of 50, maximum of 100, and a delta of 10. The Bayesian method was used to infer Ө and M , specifying two independent runs, static heating with four chains using recommended temperatures, a sampling increment of 200, 5 000 recorded steps, and a burn-in of 100 000. Convergence was assessed by examination of ESS values with a target of at least 1 000. The analysis was run multiple times to confirm reaching the same posterior distribution of parameter estimates.
BAYESASS 3.0 (Wilson & Rannala 2003) was used to estimate contemporary
migration rates between populations. The method assumed linkage equilibrium between loci,
but allowed for deviations in Hardy-Weinberg proportions by introducing an additional
inbreeding (F) parameter. FORMATOMATIC (Manoukis 2007) was used to generate a
6 BAYESASS input file from a GENEPOP file. BAYESASS was run for 30 x 10 generations with a
burn-in of 3 x 106 and a sampling interval of 300. I modified the MCMC mixing parameters to obtain target values recommended by the program’s authors. Adequate convergence was
assessed by running the program multiple times starting with different random number seeds and by employing TRACER.
Estimated migration rate estimates from MIGRATE and BAYESASS were compared by
adapting an approach similar to that of Chiucchi & Gibbs (2010). Estimates of m from
-4 MIGRATE were obtained by multiplying all M-values by an estimated mutation rate of 5 x 10
(Garza & Williamson 2001). To statistically quantify the relationship between historical and contemporary migration rates I implemented a Mantel test (Mantel 1967) in the R package
ECODIST (Goslee & Urban 2008) using 10 000 permutations. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 163
2.6 Historical and contemporary effective population sizes
I tested the null hypothesis that historical and contemporary effective population sizes were
the same. The hypothesis would be rejected, Ne changed through time, assuming that such
indicated continual habitat fragmentation and anthropogenic affects. First, I estimated
historical effective population sizes in a Bayesian framework using the coalescent model
implemented in MIGRATE. To estimate contemporary Ne, I used the linkage disequilibrium method (Hill 1981) in LDNE 1.3.1 (Waples 2006; Waples & Do 2008). Whereas the temporal method required samples from multiple points in time (Waples 1989), the linkage disequilibrium method accommodated samples from a single time frame and corrected for potential bias in sample sizes (Waples 2006). I specified a random mating model with default critical values for allele frequencies. Jackknife 95% confidence intervals were used as a measure of variance and negative values were interpreted as infinity (Waples & Do 2008).
To test for reductions in Ne operating over different timescales, I used BOTTLENECK
1.2.02 (Piry et al. 1999) to compare estimates of observed heterozygosity to number of alleles at each locus. The null hypothesis was rejected, i.e. bottlenecking was indicated, if heterozygosity excess was observed (Cornuet & Luikart 1996) over the last 0.2–4.0 Ne generations. Heterozygosity excess was tested using all three mutation models for comparison, each with 1 000 iterations. For the two-phase mutation model (TPM), I specified a variance of 12 and set the proportion of stepwise mutation events to 95% (Piry et al. 1999).
Significance was assessed using the Wilcoxon signed rank test (Cornuet & Luikart 1996).
Second, I examined allele frequency distributions for each locus and population. Populations CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 164
in mutation-drift equilibrium were assumed to show L-shaped distributions representing a large proportion of alleles at low frequency (<0.1; Luikart et al. 1998). In contrast, recently bottlenecked populations were assumed to show a mode-shift in allele frequency distribution because low frequency alleles are lost more quickly versus alleles present in high frequency.
This test was most suitable for detecting bottlenecks occurring a few generations ago
(Luikart et al. 1998).
2.7 Landscape genetics
Landscape genetics was employed to understand the relationship between physiographic
structure, historical and contemporary migration rates, and mtDNA divergence. I tested the hypothesis that landscape-level processes did not acted over historical timescales. First, I created resistance surfaces in ARCMAP 10 (ESRI) based on landscape variables shown to be
important in shaping functional connectivity for this species (Blair et al., unpublished). Next,
I used CIRCUITSCAPE 3.5.7 (Shah & McRae 2008) to calculate pairwise resistance distances between populations based on each resistance surface and then used multiple regression on distance matrices (MRM; Legendre et al. 1994; Lichstein 2007) to correlate resistance distance with both historical and contemporary migration rates. Because MIGRATE and
BAYESASS produced asymmetrical pairwise rates, I averaged both values to obtain a single
rate between populations. If recent fragmentation was altering dispersal routes, then I
expected to observe a higher correlation between contemporary vs. historical migration and landscape. Conversely, a stronger correlation between landscape and historical migration was CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 165
interpreted to indicate the importance of landscape-level processes acting over older timescales. All MRM analyses were conducted in ECODIST. As a second measure of
historical migration I correlated resistance distances with pairwise mtDNA sequence
divergence (Dxy). The incorporation of clonally inherited mtDNA allowed me to examine
signs of long-term sex-biased dispersal. Finally, I tested the sensitivity of choice-of-response
in landscape genetic inference by utilizing a stepwise MRM procedure to correlate
contemporary migration rates with different combinations of landscape variables. I used a p-
to-enter value of < 0.05 and a p-to-remove of > 0.05 in all models. Significance was assessed
using 10 000 permutations of the response matrix. I also calculated AIC values for each
model to assess relative importance. Results were then compared to a previous study that
used pairwise Dest as the response variable (Blair et al., unpublished).
3 Results
3.1 Genetic diversity and phylogenetic model selection
I obtained microsatellite data for 10 loci from 336 individuals in 10 different populations
from a previous study (Blair et al. unpublished). All 10 loci were independent and in Hardy-
Weinberg equilibrium for each population. Within-site diversity was moderate as shown by
both expected heterozygosity and allelic richness. Expected heterozygosity over all loci
ranged from 0.7025 to 0.7812 and allelic richness from 54.2341 to 65.6058. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 166
I recovered 2400 bp of mtDNA sequences for 77 P. saxatilis from 11 populations distributed throughout the immediate area surrounding Alamos, Sonora. The total number of bp per partition, the proportion of potentially phylogenetically informative characters, and the selected model of sequence evolution were shown in Supplementary Table 2. No premature stop codons were found in the protein-coding genes. Variation was more pronounced at third positions in codons versus first and second, as was expected for protein- coding genes. Variation in the three tRNAs and 16S was moderate as compared to the protein-coding loci.
3.2 Matrilineal genealogy
Bayesian analyses using three alternate partitioning strategies were summarized in
Supplementary Table 3. The calculation of Bayes Factors suggested that the most heavily
partitioned dataset (by gene and codon position) was the most efficient approach based on
improved marginal likelihood values. Thus, analyses were based on this partitioning strategy.
Maximum parsimony analysis of the concatenated dataset resulted in eight most
parsimonious trees (MPTs) of 723 steps. The MP strict consensus tree with both bootstrap
proportions and posterior probabilities was shown in Fig. 2A. To illustrate branch lengths, I
also presented the Bayesian majority rule consensus tree (Fig. 2B). Both methods resulted in
a highly supported genealogical hypothesis comprised of four maternal lineages differing up
to 6% (uncorrected p-distances). Lineage A, comprised solely of individuals from
Choquincahui, was the most distinct and sister to the remaining populations with strong CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 167
statistical (BSP = 99; BPP = 1.0). The remaining three lineages received moderate to high statistical support (Lineage B = 53/0.80; Lineage C = 96/1.0; Lineage D = 100/1.0) and showed east-to-west geographic structuring. Lineage B was composed of individuals from
Arroyo Tabelo, Rio Cuchujaqui, Los Camotes, El Palmarito, and a population 55 km northeast of the Alamos cemetery. Lineage C contained haplotypes from La Sierrita Nature
Reserve, central Alamos, and Rio Cuchujaqui. Lineage D included haplotypes from western areas including Aduana, road to Navojoa (both populations), Arroyo Tabelo, Alamos, and La
Sierrita Reserve. Thus, populations central to the study sites served as a secondary contact zone for multiple divergent mitochondrial lineages (Fig. 3). Uncorrected p-distances between lineages B, C, and D approached 3%.
3.3 Population genetic analysis of mtDNA
Standard nucleotide diversity statistics for nine populations sampled for mtDNA polymorphism were presented in Table 1. Genetic diversity was greater for populations composed of haplotypes from multiple lineages. For example, nucleotide diversity estimates
(π) for populations at Alamos and La Sierrita are almost two-fold higher than values from other populations (0.0147 vs. 0.0032). Haplotype diversity was high in all populations, with a value of 1.0 recovered for Alamos, Road to Navojoa 1, Choquincahui, and Los Camotes.
Fu’s Fs test showed no significant deviations from neutral expectations in each population, indicating no demographic expansion.
As indicated by the genealogy, significant population differentiation was present throughout the region (Table 2). Choquincahui, on average, differed from all other CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 168 populations by 106.0525 substitutions or 0.0588 substitutions per site. Individuals from the
Alamos population were more similar genetically to the populations west of Alamos
(Aduana, Road to Navojoa) versus populations east of Alamos (Rio Cuchujaqui).
Conversely, individuals from La Sierrita Nature Reserve were generally just as related to western populations as they were to the eastern populations. Rio Cuchujaqui differed from western populations (Aduana, Road to Navojoa) by approximately 29.059 changes or 0.0239 substitutions per site.
The results of the Mantel and RMA tests showed a significantly positive relationship between geographic distance (km) and mtDNA genetic distance (Z = 5314.0176; r = 0.5007;
P = 0.0002). Model estimates for the RMA analysis included an intercept of –6.435, a slope of 0.4518, and an R2 value of 0.251 (Fig. 4).
3.4 Historical and contemporary migration rates
Multiple runs of BAYESASS and MIGRATE gave consistent results, indicating that the MCMC chain was mixing well and adequately sampling from the posterior distribution.
Contemporary migration rates estimated from BAYESASS were generally low to moderate
and for most population pairs 95% confidence intervals included zero (Table 3). However,
some pairs showed signs of exchanging a relatively high number of migrants per generation.
For example, migration rates from Aduana to the Road to Navojoa were 0.155, and rates
from El Quintero to Choquincahui were 0.256. Interestingly, rates from Choquincahui to El
Quintero and from the Road to Navojoa to Aduana were much lower (0.008 and 0.012,
respectively). The fraction of non-migrants per generation ranged from 0.675 in
Choquincahui to 0.940 in Rio Cuchujaqui. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 169
Estimates of M obtained from MIGRATE also suggested little migration between
populations with 95% confidence intervals including zero for most pairs (Supplementary
Table 4). I used M to calculate the number of migrants per generation (Nm) using the formula
Nm = (Mϴ)/4. Many of these values were less than one indicating little historical migration between populations. To compare historical rates to contemporary rates I multiplied M by a
mutation rate of 5 x 10-4 to obtain an estimate of historical m; historical estimates of m were
approximately one-tenth the size of contemporary m (Fig. 5). Because I found evidence for relatively high contemporary migration from El Quintero to Choquincahui compared to historical rates (mean 0.256 vs 0.000935) I performed Mantel tests both with and without this population. When including El Quintero there was a near-significant correlation between
historical and contemporary migration rates based on 10 000 permutations (Mantel r =
0.2592, p = 0.055). Excluding this population increased the p-value, indicating no significant correlation (Mantel r=0.2773, p=0.070).
3.5 Historical and contemporary effective population sizes
I used MIGRATE to estimate historical Ne and LDNe to estimate contemporary Ne.
Contemporary Ne was large for most populations with 95% confidence intervals approaching
infinity (Table 4). However, three populations (Road to Navojoa 1, Arroyo Tabelo, and
Mocuzari) showed relatively small sizes in comparison. Averaging over each allele
frequency models, estimated Ne at these three sites was 25.4, 35.6, and 32.1, respectively.
Some populations exhibited considerable variation in estimates of Ne depending on which CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 170
allele frequency model was used (0.05, 0.02, or 0.01). Estimates of historical and contemporary Ne were similar. However, mean estimates for historical Ne were generally
larger for most populations, but 95% confidence intervals were quite large (Table 4).
Tests for recent reductions in Ne using the excess heterozygosity method in
BOTTLENECK showed no evidence for a decrease in size for any population under the TPM or
SMM model (Table 5). However, under IAM, all populations showed recent reductions.
Tests for mode-shifts in allele frequency distribution showed an L-shaped curve for all
populations except for Road to Navojoa 1, which showed a shifted distribution.
3.6 Landscape genetics
Results from the MRM analyses indicated a higher correlation between landscape resistance and contemporary migration rates versus landscape resistance and both historical rates and
mtDNA divergence (Table 6). No significant correlation occurred between Euclidean
distance and historical migration rates. Slope was the only variable to show a significant
correlation with historical rates. Coefficient of determination (R2) values in the historical
models ranged from 0.0073 to 0.1145 (mean=0.0551). Correlations in the contemporary
models ranged from 0.0214 for slope to 0.1994 for Euclidean distance (mean=0.1268).
Further, four of the six landscape variables were significantly correlated with contemporary
migration rates. Euclidean distance was the only variable significantly correlated with
mtDNA divergence. The stepwise MRM analysis suggested that Euclidean distance and
anthropogenic disturbance were the best predictors of contemporary gene flow as estimated
with BAYESASS (AIC weight=0.1604; Supplementary Table 5). CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 171
4 Discussion
Population and landscape genetics seek to disentangle the relative roles historical and
contemporary processes play in shaping the spatial distribution of genetic variation
(Vandergast et al. 2007; Anderson et al. 2010; Burbrink 2010; Dyer et al. 2010; Wang 2010).
I use landscape genetics to investigate the temporal dynamics of demographic processes.
High levels of mtDNA divergence between lineages occur over small spatial scales, in the
absence of obvious geographic barriers, which is consistent with female philopatry and low
levels of historical gene flow. A 10-fold increase in contemporary versus historical migration
rates occurs yet contemporary versus historical migration appears to be better explained by
present-day landscape-level processes. I show that the choice of response-variable
(equilibrium vs. non-equilibrium metrics, microsatellites vs. mtDNA) can influence
subsequent landscape-genetic inference. Below I expand on these major findings and discuss
the advantages of adopting similar multi-tiered approaches when investigating the
demographic consequences of recent habitat fragmentation in a spatial context.
4.1 Sequence divergence and mitochondrial genealogy
Substantial lineage divergence occurs among populations of leaf-toed geckos distributed near
Alamos, Sonora, with samples from Choquincahui showing the highest divergence values
(approximately 6%). Pairwise comparisons between several other populations show divergence estimates (uncorrected p-distances) ranging from 1% to 4%. The MP and BI analyses yield highly compatible topologies and several nodes are clearly resolved (Fig. 2A, CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 172
B). Geographic structure and significant isolation-by-distance is detected. Whereas lineage D
(blue lineage) occurs predominantly west of Alamos the remaining lineages are distributed between Choquincahui and the western populations (green and yellow). Lineages B, C, and
D sometimes occur sympatrically (Arroyo Tabelo, Alamos, La Sierrita, and Rio Cuchujaqui).
The two sympatric mitochondrial lineages at each of these sites may result from contemporary female dispersal. If asymmetric female dispersal drives these patterns, then females from Rio Cuchujaqui must be dispersing to Alamos and females from Arroyo
Aduana and Road to Navojoa are dispersing to both Alamos and La Sierrita. Dispersal from western to eastern populations is non-existent. Thus, the home ranges must be quite small and dispersal capabilities limited to approximately 20 km. Indeed I find moderate, but significant isolation-by-distance with the mtDNA data (Fig. 4). This explanation also requires asymmetrical female dispersal via migration from both eastern and western populations into Alamos, but not vice versa. Asymmetrical migration rates appear to occur in these lizards (Table 3). Another explanation involves incomplete lineage sorting at sites containing sympatric lineages but, again, this requires asymmetrical female dispersal.
Finally, these sites might be composed of hybrids between eastern and western populations.
Blair et al. (unpublished) find evidence of substantial admixture within populations of these geckos. Taken together, this suggests a high degree of female philopatry and male-biased dispersal in these lizards.
The co-occurrence of both mitochondrial lineages throughout Alamos proper might suggest an ecological component is involved. Geckos often occur among human settlements, which provide additional microhabitat in the form of houses and other man-made structures.
These animals tend to prefer dark, moist areas on vertical substrate, a microhabitat that is CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 173 quite abundant in residential areas. However, Blair et al. (unpublished) find that leaf-toed geckos prefer to disperse via undisturbed habitat.
The discovery of significant cryptic lineage divergence is consistent with previous studies and provides further evidence that geckos possess extremely small home-range sizes
(Gübitz et al. 2000; Kaspidis et al. 2005; Rato et al. 2011). Whether these cryptic lineages warrant species status will require evidence from additional sources. Notwithstanding, the mtDNA and microsatellite evidence suggests that the population from Choquincahui should be recognized as a species.
4.2 Historical and contemporary migration rates and effective population sizes
Contrary to expectations, I find no reduction in contemporary versus historical migration rates but rather a 10-fold increase in contemporary migration rates. The Mantel test shows no significant correlation between these two metrics. These observations suggest that recent habitat fragmentation is not impacting the ability of individuals to disperse. Only a handful of studies have tested for similarities in historical and contemporary migration rates (Ross et al.
2007; Andersen et al. 2009; Schmidt et al. 2009; Muscarella et al. 2011) and few have
pursued a hypothesis-testing framework (Chiucchi & Gibbs 2010). Many of these studies inappropriately compare m estimated from BAYESASS with Nm as estimated with MIGRATE.
BAYESASS uses MCMC to estimate contemporary asymmetric migration rates (m) among
populations, whereas MIGRATE uses MCMC in a coalescent framework to estimate both
historical effective population sizes (4Neµ for an autosomal locus) and the mutation-scaled migration rate M (m/µ). Migration rate and the number of migrants per generation (Nm) are not equivalent. The accurate comparison of results from these two methods requires CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 174 calculating m from MIGRATE by multiplying M times the mutation rate µ. Thus, many
previous studies fail to elucidate temporal differences in migration rates and how these
correspond to recent landscape or environmental dynamics. The only previous study to use these methods in a strict hypothesis-testing framework has found no significant difference
between historical and contemporary rates, suggesting that populations have lived in small
isolated populations for a long period of time (Chiucchi & Gibbs 2010). Conversely, in leaf-
toed geckos contemporary gene flow is higher than historical gene flow. A non-
anthropogenic explanation for these patterns is wanting. Geckos commonly occur in and
around human settlements. Relatively recent colonization of humans appears to have
increased migration rates via hitchhiking. Similar studies conducted on geckos will determine
if similar patterns occur.
Lower contemporary versus historical population sizes are predicted when habitat
fragmentation is having negative consequences on populations. Estimates of contemporary
effective population sizes are high for most populations, with 95% confidence intervals
approaching infinity. However, Road to Navojoa, Arroyo Tabelo, and Mocuzari show signs
of small effective population sizes irrespective of the employed allele frequency model.
MIGRATE estimates large historical effective population sizes and historical population sizes
for these populations are substantially higher than contemporary estimates. In these cases,
recent fragmentation appears to have reduced population sizes. No evidence of recent
bottlenecks is detected using the models TPM and SMM but most populations show signs of a bottleneck when assuming IAM. Although microsatellites are traditionally assumed to follow an approximate SMM model (Di Rienzo et al. 1994, Piry et al. 1999), new evidence
suggests that recent bottlenecks may be better detected assuming an IAM when working with CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 175 microsatellites (Cristescu et al. 2010). Thus, all of the populations except La Sierrita have experienced a recent bottleneck. Tests for a mode-shift in allele frequency distribution yield a shift for Road to Navojoa.
4.3 Landscape genetics
Considerable recent debate has surrounded the utility of different molecular marker types and
analytical methods for landscape genetic inference (Manel et al. 2003; Vandergast et al.
2007; Storfer et al. 2010; Wang 2010, 2011; Bohonak & Vandergast 2011). Most landscape
genetic studies correlate effective and genetic distances, the later usually using FST-like
metrics based on microsatellites. I probe the power of the ability of different analytical
methods (MIGRATE vs. BAYESASS) and genetic marker-types (mtDNA vs. microsatellites) to
determine if and how the choice of method or marker influences landscape genetic inference.
MRM analysis more significantly correlate landscape with contemporary versus historical
migration rates, as expected when populations respond to recent habitat fragmentation. These
results would have been missed if I based my inference solely on differential historical and
contemporary migration rates in an aspatial context.
Correlations between landscape and mtDNA divergence contrast sharply with an
analysis based on microsatellites (Blair et al. unpublished). The stepwise MRM analysis
indicates that Euclidean distance is the only significant predictor of mtDNA divergence.
Although landscape genetic studies continue to rely predominantly on microsatellites (Storfer
et al. 2010), a handful of studies have relied solely on mtDNA to infer contemporary
ecological and evolutionary processes (e.g. Vandergast et al. 2007; Koscinski et al. 2009).
The prior employment of mtDNA may owe to cost and relative ease of data collection. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 176
However, this approach comes with costs that must be acknowledged by researchers (Wang
2010). Major issues include the stochasticity of the lineage sorting process, relatively slow mutation rates, and, most importantly, mtDNA is maternally inherited and does not document bi-parental gene flow. I obtain drastically different landscape-genetic relationships when using mtDNA versus microsatellites (Blair et al. unpublished). Similar results are found when comparing landscape-genetic relationships using different markers for wood frog populations (Lee-Yaw et al. 2009). Taken together, this begs the question as to the utility of mtDNA for inferring contemporary ecological processes. Although mtDNA may be adequate for testing simple barrier hypotheses using binary matrices (e.g. Vandergast et al. 2007), landscape-genetic relationships are often more complex (e.g. Cushman et al. 2006) and require the use of more sophisticated GIS-based modeling approaches. My analysis suggests that mtDNA is a poor choice for modeling landscape resistance based on a circuit theoretic approach. To avoid reaching erroneous conclusions regarding the influence of recent habitat fragmentation on genetic divergence, researchers should test explicit hypotheses and define the timeframe under consideration (Wang 2010).
My tests of landscape genetic inference based on different response metrics determine that contemporary migration rates differ substantially from those based on equilibrium approaches (Blair et al. unpublished). Most landscape genetic studies test for correlations between landscape variables and gene flow (Manel et al. 2003; Storfer et al. 2010), yet most have not used migration rates estimated with BAYESASS as the dependent variable. Perhaps
this owes to the difficulty of Bayesian MCMC algorithms used to estimate the posterior
distribution of demographic parameters (Wilson & Rannala 2003). The 95% confidence
intervals are quite large and this may confound subsequent inference. Further, the number of CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 177 populations, individuals, and loci can greatly impact the inference of contemporary migration rates (Wilson & Rannala 2003). As simulation studies continue to define and direct empirical landscape genetic research (e.g. Epperson et al. 2010; Landguth et al. 2010; Blair et al.
2012), future studies should test the power of migration rates estimated with BAYESASS
versus the more traditional equilibrium-based statistics such as FST. Additional Bayesian
methods that statistically correlate recent migration with environmental variables (e.g. BIMr;
Foll & Gaggiotti 2006) also show promise (Balkenhol et al. 2009).
I envision three different approaches to help disentangle the relative influences of
historical and contemporary processes that shape contemporary genetic structure. First,
methods that correlate genetic data with both contemporary and historical landscape and
environmental data can aid in determining if and when shifts in landscape structure
influenced genetic divergence (Vandergast et al. 2007; Zellmer & Knowles 2009). Second, it
is possible to partition out the effects of deep genealogical history prior to conducting
landscape genetic inference by incorporating residuals from regression analysis (Vandergast
et al. 2007; Dyer et al. 2010). However, the use of residuals for subsequent statistical
analysis is questioned due to biased parameter estimates as compared to multiple regression
approaches (Freckleton 2002). Third, multiple markers and analytical methods can be used in
a landscape genetic context to probe the population genetic consequences of recent habitat
fragmentation. This method is particularly attractive for areas generally lacking in fine-scale
geospatial data for multiple temporal periods. This approach is also useful for species that
show signs of strong genealogical structure at small spatial scales that cannot be attributed to
common ecological or biogeographic barriers.
CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 178
4.4 Implications for Mexican biodiversity
Mexico is hypothesized to house some of Earth’s richest biodiversity (particularly reptiles) due to its geographic location intersecting the Nearctic and Neotropical realms (Devitt 2006;
García 2006). Further, this diversity can be attributed to the dynamic geological and climatic histories of Mexico, which has resulted in numerous mountain ridges, river systems, and plateaus, contributing to a diverse array of habitats and ecosystems ranging from desert habitats in the north to tropical rain forest in the south and east (Cartron et al. 2005). This varied landscape, in turn, serves as an ideal region to address questions pertaining to population differentiation, landscape resistance, and barriers to gene flow. Species diversity in southern Sonora, for example, may be partly explained by drastic shifts in habitat characteristics, with the arid Sonoran Desert coming into contact with tropical thornscrub, deciduous, and pine-oak forest in the south (Robichaux & Yetman 2000; Cartron et al. 2005).
Further, orogenesis of the Sierra Madre Occidental during the Eocene/Oligocene transition has had a major impact on not only the geology and climate of western Mexico, but also on biogeographic patterns in several widespread groups (Cartron et al. 2005).
A multi-tiered analytical approach can help clarify both the historical and contemporary processes contributing to diversification throughout this ecosystem. Research is addressing phylogenetic, population genetic, and landscape genetic questions throughout
Mexico and Middle America utilizing a diverse array of taxa including lizards (Blair et al., unpublished), snakes (Mulcahy 2008), birds (Zink et al. 2001; Rojas-Soto et al. 2007), frogs
(Mulcahy & Mendelson 2000; Pfeiler & Markow 2008), fish (Hulsey et al. 2004; Huidobro CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 179
et al. 2006) and mammals (Harris et al. 2000; Ortega et al. 2009). Most of these studies attribute historical patterns to the orogenesis of Mexico’s major mountain systems including the TMVB or climatic fluctuations during the Pleistocene. However, most studies have not focused specifically on patterns of genetic diversity throughout the TDF of western Mexico
(but see Devitt 2006; McCormack et al. 2008; Zarza et al. 2008; Blair et al. unpublished) and, thus, we still know relatively little about the spatial and temporal processes contributing to diversification of species inhabiting this ecosystem.
4.5 Conclusions
I present a comprehensive method to understanding the population genetic consequences of
recent habitat fragmentation for leaf-toed geckos inhabiting Mexican TDF. My results
suggest that migration rates have increased in the recent past, but that geckos are altering
their dispersal routes in response to recent landscape change. I find evidence for recent
reductions in effective population sizes, likely due to continual anthropogenic infringement,
and a deep genealogical history over small spatial scales, which reflects low historical gene
flow and strong female philopatry. My analyses suggest that mtDNA is a poor marker for
inferring contemporary landscape-genetic relationships due to its low mutation rate, error
associated with the coalescent process, and matrilineal inheritance. Additional simulation
studies are needed to test the power of different measures of genetic connectivity (FST vs
BAYESASS) when inferring landscape genetic relationships.
Interesting, a high level of mtDNA variation occurs on a very small geographic scale.
Generally, DNA sequences lack the variation necessary to make demographic inferences on CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 180 such scales. Markers such as microsatellites or AFLPs are usually used in order to document such genetic patterning (Storfer et al. 2010). However, my analysis suggests that the examination of mtDNA sequence data provides insight into deeper historical divergences at the population genetic level, in contrast to markers such as microsatellites that are more suitable for detecting more recent genetic patterns. Thus, studies should use mtDNA data in addition to, rather in lieu of, other markers such as microsatellites and SNPs to understand the spatial and temporal dynamics of population connectivity. This may be of particular relevance to species likely to possess limited dispersal abilities. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 181
Acknowledgments
I would like to thank A. Ngo, D. Dewsbury, I. Conflitti, and C. Davy for providing insightful
comments on previous drafts of the manuscript. This research was supported by Discovery
Grant A3148 from the Natural Sciences and Engineering Research Council of Canada and by
the Royal Ontario Museum Board of Governors to RWM. Support was also provided by the
Schad Foundation of the Royal Ontario Museum and from a Theodore Roosevelt Memorial grant through the American Museum of Natural History (CB). Invaluable field assistance was provided by Martin G. Figueroa, Norberto Martinez Mendez, Rafael Lara Resendiz and
Arabel Escalona López. Fieldwork in reserves was conducted under the permit issued to Dr.
Francisco Molina Freaner. Stephanie A. Meyer provided invaluable advice and field logistics. All research was conducted using approved Animal Use Protocols. Collection
(FLOR0030, SGPA/DGFS/047097/09, SGPA/DGVS/02320/10, SGPA/DGVS/02372/11) and export permits (028, 059) were obtained from SEMARNAT through the Universidad
Nacional Autónoma de México (UNAM). CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 182
References
Akaike H (1974) A look at the statistical model identification. IEEE Transactions Automatic
Control AC, 19, 716–723.
Akaike H (1979) A Bayesian extension of the minimum AIC procedure of autoregressive
model fitting. Biometrika, 66, 237–242.
Andersen LW, Born EW, Doidge DW, Gjertz I, Wiig Ø, Waples RS (2009) Genetic signals
of historic and recent migration between sub-populations of Atlantic walrus
Odobenus rosmarus rosmarus west and east of Greenland. Endangered Species
Research, 9, 197–211.
Anderson CD, Epperson BK, Fortin MJ et al. (2010) Considering spatial and temporal scale
in landscape-genetic studies of gene flow. Molecular Ecology, 19, 3565–3575.
Arevalo E, Davis SK, Sites JW (1994) Mitochondrial DNA sequence divergence and
phylogenetic relationships among eight chromosomal races of the Sceloporus
grammicus complex (Phrynosomatidae) in Central Mexico. Systematic Biology, 43,
387–418.
Austin JJ, Arnold EN, Jones CG (2004) Reconstructing an island radiation using ancient and
recent DNA: the extinct and living day geckos (Phelsuma) of the Mascarene Islands.
Molecular Phylogenetics and Evolution, 31, 109–122.
Balkenhol N, Waits LP, Dezzani RJ (2009) Statistical approaches in landscape genetics: an
evaluation of methods for linking landscape and genetic data. Ecography, 32, 818–
830.
Becerra JX (2005) Timing the origin and expansion of the Mexican tropical dry forest.
Proceedings of the National Academy of Sciences USA, 102, 10919–10923. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 183
Becerra JX, Venable DL (2008) Sources and sinks of diversification and conservation
priorities for Mexican tropical dry forest. PLoS One, 3, 1–5.
Beerli P (2006) Comparison of Bayesian and maximum likelihood inference of population
genetic parameters. Bioinformatics, 22, 341–345.
Beerli P (2008) MIGRATE documentation (version 3.0). Technical Report.
http://popgen.sc.fsu.edu.
Beerli P, Felsenstein J (2001) Maximum likelihood estimation of a migration matrix and
effective population sizes in n subpopulations by using a coalescent approach.
Proceedings of the National Academy of Sciences of the USA, 98, 4563–4568.
Blair C, Méndez de la Cruz FR, Ngo A, Lindell J, Lathrop A, Murphy RW (2009) Molecular
phylogenetics and taxonomy of leaf-toed geckos (Phyllodactylidae: Phyllodactylus)
inhabiting the peninsula of Baja California. Zootaxa, 2027, 28–42.
Blair C, Weigel D, Balazik M, et al. (in press). A simulation-based evaluation of methods for
inferring linear barriers to gene flow. Molecular Ecology Resources.
Bohonak AJ, Vandergast AG (2011) The value of DNA sequence data for studying landscape
genetics. Molecular Ecology, 20, 2477–2479.
Brandley MC, Schmitz A, Reeder TW (2005) Partitioned Bayesian analysis, partition choice,
and the phylogenetic relationships of scincid lizards. Systematic Biology, 54, 373–
390.
Burbrink FT (2010) Historical versus contemporary migration in fragmented populations.
Molecular Ecology, 19, 5321–5323.
Cartron JE, Ceballos G, Felger RS (2005) Biodiversity, Ecosystems, and Conservation in
Northern Mexico. New York, New York: Oxford University Press. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 184
Castiglia R, García A, Bezerra AMR, Flores-Villela O, Gornung E (2009) Karyotipic
diversification due to Robertsonian rearrangements in Phyllodactylus lanei Smith,
1935 (Squamata, Gekkonidae) from Mexico. Rendiconti Lincei, 20, 77–82.
Chiucchi JE, Gibbs HL (2010) Similarity of contemporary and historical gene flow among
highly fragmented populations of an endangered rattlesnake. Molecular Ecology, 19,
5345–5358.
Cornuet JM, Luikart G (1996) Description and power analysis of two tests for detecting
recent population bottlenecks from allele frequency data. Genetics, 144, 2001–2014.
Crispo E, Moore JS, Lee-Yaw JA, Gray SM, Haller BC (2011) Broken barriers: human-
induced changes to gene flow and introgression in animals. BioEssays, 33, 508–518.
Cristescu R, Sherwin WB, Handasyde K, Cahill V, Cooper DW (2010) Detecting bottlenecks
using Bottleneck 1.2.02 in wild populations: the importance of the microsatellite
structure. Conservation Genetics, 11, 1043–1049.
Cushman SA, McKelvey KS, Hayden J, Schwartz MK (2006) Gene flow in complex
landscapes: testing multiple hypotheses with causal modeling. American Naturalist,
168, 486–499.
Devitt TJ (2006) Phylogeography of the western lyresnake (Trimorphodon biscutatus):
testing aridland biogeographical hypotheses across the Nearctic-Neotropical
transition. Molecular Ecology, 15, 4387–4407.
Di Rienzo A, Peterson AC, Garza JC et al. (1994) Mutational processes of simple-sequence
repeat loci in human populations. Proceedings of the National Academy of Sciences
USA, 91, 3166–3170. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 185
Dixon JR (1964) The systematics and distribution of lizards of the genus Phyllodactylus in
North and Central America. New Mexico State University Research Center, Scientific
Bulletin, 64–1, 1–139.
Dyer RJ, Nason JD, Garrick RC (2010) Landscape modeling of gene flow: improved power
using conditional genetic distance derived from the topology of population networks.
Molecular Ecology, 19, 3746–3759.
Epperson BK, McRae BH, Scribner K et al. (2010) Utility of computer simulations in
landscape genetics. Molecular Ecology, 19, 3549–3564.
Erixon P, Svennblad B, Britton T, Oxelman B (2003) Reliability of Bayesian posterior
probabilities and bootstrap frequencies in phylogenetics. Systematic Biology, 52, 665–
673.
Ewers RM, Didham RK (2005) Confounding factors in the detection of species responses to
habitat fragmentation. Biological Reviews, 81, 117–142.
Fahrig L (2003) Effects of habitat fragmentation on biodiversity. Annual Review of Ecology
and Systematics, 34, 487–515.
Felsenstein J (1985) Confidence limits on phylogenies: an approach using the bootstrap.
Evolution, 39, 783–791.
Foll M, Gaggiotti O (2006) Identifying the environmental factors that determine the genetic
structure of populations. Genetics, 174, 875–891.
Freckleton RP (2002) On the misuse of residuals in ecology: regression of residuals vs.
multiple regression. Journal of Animal Ecology, 71, 542–545.
Fu YX (1997) Statistical tests of neutrality of mutations against population growth,
hitchhiking and background selection. Genetics, 147, 915–925. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 186
Gamble T, Bauer AM, Greenbaum E, Jackman TR (2008) Out of the blue: a novel, trans-
Atlantic clade of geckos (Gekkota, Squamata). Zoologica Scripta, 37, 355–366.
García A (2006) Using ecological niche modeling to identify diversity hotspots for the
herpetofauna of Pacific lowlands and adjacent interior valleys of Mexico. Biological
Conservation, 130, 25–46.
Garza JC, Williamson EG (2001) Detection of reduction in population size using data from
microsatellite loci. Molecular Ecology, 10, 305–318.
Goldberg CS, Waits LP (2010) Comparative landscape genetics of two pond-breeding
amphibian species in a highly modified agricultural landscape. Molecular Ecology,
19, 3650–3663.
Goloboff PA (1999) Analyzing large data sets in reasonable times: solutions for composite
optima. Cladistics, 15, 415–428.
Goloboff PA, Farris JS, Nixon KC (2008) TNT, a free program for phylogenetic analysis.
Cladistics, 24, 774–786.
Goslee S, Urban D (2008) Dissimilarity-Based Functions for Ecological Analysis. CRAN
repository. Available from http://cran.r-project.org/web/packages/ecodist/.
Gübitz T, Thorpe RS, Malhotra A (2000) Phylogeography and natural selection in the
Tanerife gecko Tarentola delalandii: testing historical and adaptive hypotheses.
Molecular Ecology, 9, 1213–1221.
Hall TA (1999) BIOEDIT: a user-friendly biological sequence alignment editor and analysis
program for Windows 95/98/NT. Nucleic Acids Symposium Series, 41, 95–98.
Harris D, Rogers DS, Sullivan J (2000) Phylogeography of Peromyscus furvus (Rodentia;
Muridae) based on cytochrome b sequence data. Molecular Ecology, 9, 2129–2135. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 187
Harpending H (1994) Signature of ancient population growth in a low-resolution
mitochondrial DNA mismatch distribution. Human Biology, 66, 591–600.
Hill WG (1981) Estimation of effective population size from data on linkage disequilibrium.
Genetics Research, 38, 209–216.
Hillis DM, Mable BK, Larson A, Davis SK, Zimmer EA (1996) Nucleic acids IV:
sequencing and cloning. In: Molecular Systematics (eds Hillis DM, Moritz C, Mable
BK), pp. 321–383. Sinauer Associates, Sunderland, Massachusetts.
Hillis DM, Bull JJ (1993) An empirical test of bootstrapping as a method for assessing
confidence in phylogenetic analysis. Systematic Biology, 42, 182–192.
Huelsenbeck JP, Ronquist F (2001) MRBAYES v3.1.2: Bayesian inference of phylogeny.
Bioinformatics, 17, 754–755.
Huelsenbeck JP, Larget B, Miller RE, Ronquist F (2002) Potential applications and pitfalls of
Bayesian inference of phylogeny. Systematic Biology, 51, 673–688.
Huidobro L, Morrone JJ, Villalobos JL, Álvarez F (2006) Distributional patterns of
freshwater taxa (fishes, crustaceans and plants) from the Mexican Transition Zone.
Journal of Biogeography, 33, 731–741.
Hulsey CD, García de León FJ, Johnson YS, Hendrickson DA, Near TJ (2004) Temporal
diversification of Mesoamerican cichlid fishes across a major biogeographic
boundary. Molecular Phylogenetics and Evolution, 31, 754–764.
Jensen JL, Bohonak AJ, Kelley ST (2005) Isolation by distance, web service. BMC Genetics,
6, 13. v.3.16 http://ibdws.sdsu.edu/. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 188
Kasapidis P, Magoulas A, Mylonas M, Zouros E (2005) The phylogeography of the gecko
Cyrtopodion kotschyi (Reptilia: Gekkonidae) in the Aegean archipelago. Molecular
Phylogenetics and Evolution, 35, 612–623.
Kass RE, Raftery AE (1995) Bayes Factors. Journal of the American Statistical Association,
90, 773–795.
Koscinski D, Yates AG, Handford P, Lougheed SC (2009) Effects of landscape and history
on diversification of a montane, stream-breeding amphibian. Journal of
Biogeography, 36, 255–265.
Landguth EL, Cushman SA, Schwartz MK, McKelvey KS, Murphy M, Luikart G (2010)
Quantifying the lag time to detect barriers in landscape genetics. Molecular Ecology,
19, 4179–4191.
Larget B, Simon DL (1999) Markov Chain Monte Carlo algorithms for the Bayesian analysis
of phylogenetic trees. Molecular Biology and Evolution, 16 750–759.
Leaché AD, McGuire JA (2006) Phylogenetic relationships of horned lizards (Phrynosoma)
based on nuclear and mitochondrial data: Evidence for a misleading mitochondrial
gene tree. Molecular Phylogenetics and Evolution, 39, 628–644.
Lee-Yaw JA, Davidson A, McRae BH, Green DM (2009) Do landscape processes predict
phylogeographic patterns in the wood frog? Molecular Ecology, 18, 1863–1874.
Legendre P, Lapointe FJ, Casgrain P (1994) Modeling brain evolution from behaviour: a
permutational regression approach. Evolution, 48, 1487–1499.
Lichstein JW (2007) Multiple regression on distance matrices: a multivariate spatial analysis
tool. Plant Ecology, 188, 117–131. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 189
Luikart G, Allendorf FW, Cornuet JM et al. (1998) Distortion of allele frequency
distributions provides a test for recent population bottlenecks. Journal of Heredity,
89, 238–247.
Maddison DR, Maddison WP (2005) MACCLADE v.4.08 OSX: analysis of phylogeny and
character evolution. Sinauer Associates, Sunderland, Massachusetts.
Manel S, Schwartz MK, Luikart G, Taberlet P (2003) Landscape genetics: combining
landscape ecology and population genetics. Trends in Ecology and Evolution, 18,
189–197.
Manoukis NC (2007) FORMATOMATIC: a program for converting diploid allelic data between
common formats for population genetic analysis. Molecular Ecology Notes, 7, 592–
593.
Marshall DC, Simon C, Buckley TR (2006) Accurate branch length estimation in partitioned
Bayesian analyses requires accommodation of among-partition rate variation and
attention to branch length priors. Systematic Biology, 55, 993–1003.
McCormack JE, Peterson AT, Bonaccorso E, Smith TB (2008) Speciation in the highlands of
Mexico: genetic and phenotypic divergence in the Mexican jay. Molecular Ecology,
17, 2505–2521.
Mosauer W (1936) Description of a new Phyllodactylus from Mexico, with remarks on the
status of P. tuberculosus. Copeia, 1936, 141–146.
Mulcahy DG (2008) Phylogeography and species boundaries of the western North American
nightsnake (Hypsiglena torquata): revisiting the subspecies concept. Molecular
Phylogenetics and Evolution, 46, 1095–1115. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 190
Mulcahy DG, Mendelson JR III (2000) Phylogeography and speciation of the
morphologically variable widespread species Bufo valliceps, based on molecular
evidence from mtDNA. Molecular Phylogenetics and Evolution, 17, 173–189.
Murphy PG, Lugo AE (1996) Ecology of tropical dry forest. Annual Review of Ecology and
Systematics, 17, 67–88.
Murphy RW, Blair C, Méndez de la Cruz FR (2009) A new species of leaf-toed gecko, genus
Phyllodactylus (Squamata: Gekkota: Phyllodactylidae) from Guerrero, Mexico. South
American Journal of Herpetology, 4, 17–24.
Muscarella RA, Murray KL, Ortt D, Russell AL, Fleming TH (2011) Exploring
demographic, physical, and historical explanations for the genetic structure of two
lineages of Greater Antillean bats. PLoS One, 6, e17704.
doi:10.1371/journal.pone.0017704.
Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity
hotspots for conservation priorities. Nature, 409, 853–858.
Myska P (2007) Field Guide to the Amphibians, Reptiles, Birds, and Mammals of Western
Mexico. Peter Myska, Viva Natura.
Nei M (1987) Molecular Evolutionary Genetics. Columbia University Press, New York, New
York.
Nixon KC (1999) The parsimony ratchet, a new method for rapid parsimony analysis.
Cladistics, 15, 407–414.
Nylander JAA (2004) MRMODELTEST V2.2. Program distributed by author. Evolutionary
Biology Centre, Uppsala University. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 191
Nylander JAA, Ronquist F, Huelsenbeck JP, Nievees-Aldrey JL (2004) Bayesian
phylogenetic analysis of combined data. Systematic Biology, 53, 47–67.
Ortega J, Tschapka M, González-Terrazas TP et al. (2009) Phylogeography of Musonycteris
harrisoni along the Pacific coast of Mexico. Acta Chiropterologica, 11, 259–269.
Pennington TR, Prado DE, Pendry CA (2000) Neotropical seasonally dry forests and
Quaternary vegetation changes. Journal of Biogeography, 27, 261–273.
Pfeiler E, Markow TA (2008) Phylogenetic relationships of leopard frogs (Rana pipiens
complex) from an isolated coastal mountain range in southern Sonora, Mexico.
Molecular Phylogenetics and Evolution, 49, 343–348.
Piry S, Luikart G, Cornuet JM (1999) BOTTLENECK: a computer program for detecting recent
reductions in effective population size using allele frequency data. Journal of
Heredity, 90, 502–503.
Rambaut A, Drummond AJ (2007) TRACER v1.6.1. Available from
http://beast.bio.ed.ac.uk/Tracer.
Rato C, Carranza S, Harris DJ (2011) When selection deceives phylogeographic
interpretation: the case of the Mediterranean house gecko, Hemidactylus turcius
(Linnaeus, 1758). Molecular Phylogenetics and Evolution, 58, 365–373.
Ricketts TH (2001) The matrix matters: effective isolation in fragmented landscapes. The
American Naturalist, 158, 87–99.
Robichaux RH, Yetman DA (2000) The Tropical Deciduous Forest of Alamos. Biodiversity
of a Threatened Ecosystem in Mexico. Tucson: The University of Arizona Press.
Rojas-Soto OR, Espinosa de los Monteros A, Zink RM (2007) Phylogeography and patterns
of differentiation in the curve-billed thrasher. The Condor, 109, 456–463. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 192
Ross, KG, Krieger MJB, Keller L, Shoemaker DD (2007) Genetic variation and structure in
native populations of the fire ant Solenopsis invicta: evolutionary and demographic
implications. Biological Journal of the Linnean Society, 92, 541–560.
Rousset F (1997) Genetic differentiation and estimation of gene flow from F-statistics under
isolation by distance. Genetics, 145, 1219–1228.
Rozas J, Sanchez-Delbarrio JC, Messeguer X, Rozas R (2003) DNASP, DNA polymorphism
analyses by the coalescent and other methods. Bioinformatics, 19, 2496–2497.
Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory Manual,
second edition. New York: Cold Spring Harbor Laboratory Press, Cold Spring
Harbor.
Sanderson EW, Jaiteh M, Levy MA, et al. (2002) The human footprint and the last of the
wild. Bioscience, 52, 891–904.
Schmidt JI, Hundertmark KJ, Bowyer TR, McCracken KG (2009) Population structure and
genetic diversity of moose in Alaska. Journal of Heredity, 100, 170–180.
Shah VB, McRae BH (2008) CIRCUITSCAPE: a tool for landscape ecology. In: Varoquaux G,
Vaught T, Millman J (eds). Proceedings of the 7th Python in Science Conference
(SciPy 2008), pp. 62–66.
Sork VL, Waits LP (2010) Contributions of landscape genetics—approaches, insights, and
future potential. Molecular Ecology, 19, 3489–3495.
Spear SF, Storfer A (2008) Landscape genetic structure of coastal tailed frogs (Ascaphus
truei) in protected vs. managed forests. Molecular Ecology, 17, 4642–4656. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 193
Spear SF, Peterson CR, Matocq M, Storfer A (2005) Landscape genetics of the blotched tiger
salamander (Ambystoma tigrinum melanostictum). Molecular Ecology, 14, 25553–
2564.
Storfer A, Murphy MA, Spear SP, Holderegger R, Waits LP (2010) Landscape genetics:
where are we now? Molecular Ecology, 19, 3496–3514.
Suzuki Y, Glazko GV, Nei M (2002) Overcredibility of molecular phylogenies obtained by
Bayesian phylogenetics. Proceedings of the National Academy of Sciences USA, 99,
16138–16143.
Swofford DL (2002) PAUP*: Phylogenetic Analysis Using Parsimony (*and other methods).
Sinauer Associates, Sunderland, MA.
Tajima F (1983) Evolutionary relationships of DNA sequences in finite populations.
Genetics, 105, 437–460.
Tajima F (1989) Statistical method for testing the neutral mutation hypothesis by DNA
polymorphism. Genetics, 123, 585–595.
Tajima F (1993) Measurement of DNA polymorphism. In: Mechanisms of Molecular
Evolution (eds Takahata N, Clark AG), pp. 37–59. Sinauer Associates Inc.,
Sunderland, Massachusetts.
Trejo I, Dirzo R (2000) Deforestation of seasonally dry tropical forest: a national and local
analysis in Mexico. Biological Conservation, 94, 133–142.
Vandergast AG, Bohonak AJ, Weissman DB, Fisher RN (2007) Understanding the genetic
effects of recent habitat fragmentation in the context of evolutionary history:
phylogeography and landscape genetics of a southern California endemic Jerusalem CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 194
cricket (Orthoptera: Stenopelmatidae: Stenopelmatus). Molecular Ecology, 16, 977–
992.
Vignieri SN (2005) Streams over mountains: influence of riparian connectivity on gene flow
in the Pacific jumping mouse (Zapus trinotatus). Molecular Ecology, 14, 1925–1937.
Wang IJ (2010) Recognizing the temporal distinctions between landscape genetics and
phylogeography. Molecular Ecology, 19, 2605–2608.
Wang IJ (2011) Choosing appropriate genetic markers and analytical methods for testing
landscape genetic hypotheses. Molecular Ecology, 20, 2480–2482.
Wang IJ, Savage WK, Shaffer BH (2009) Landscape genetics and least-cost path analysis
reveal unexpected dispersal routes in the California tiger salamander (Ambystoma
californiense). Molecular Ecology, 18, 1365–1374.
Waples RS (1989) A generalized approach for estimating effective population size from
temporal changes in allele frequency. Genetics, 121, 379–391.
Waples RS (2006) A bias correction for estimates of effective population size based on
linkage disequilibrium at unlinked gene loci. Conservation Genetics, 7, 167–184.
Waples RS, Do C (2008) LDNE: a program for estimating effective population size from data
on linkage disequilibrium. Molecular Ecology Resources, 8, 753–756.
Wilson GA, Rannala B (2003) Bayesian inference of recent migration rates using multilocus
genotypes. Genetics, 163, 1177–1191.
Wright S (1943) Isolation by distance. Genetics, 28, 114–138.
Zarza E, Reynoso VH, Emerson BC (2008) Diversification in the northern Neotropics:
mitochondrial and nuclear DNA phylogeography of the iguana Ctenosaura pectinata
and related species. Molecular Ecology, 17, 3259–3275. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 195
Zellmer AJ, Knowles LL (2009) Disentangling the effects of historic vs. contemporary
landscape structure on population genetic divergence. Molecular Ecology, 18, 3593–
3602.
Zink RM, Kessen AE, Line TV, Blackwell-Rago RC (2001) Comparative phylogeography of
some aridland bird species. The Condor, 103, 1–10.
CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 196
Table 1. Nucleotide diversity statistics and their variances for all populations sampled for mtDNA.
Locality nmtDNA H h V(h) π k V(k) S Fu's Fs
Alamos 6 6 1 0.0093 0.0101 23.2 142.68 67 0.194 Rio Cuchujaqui 20 9 0.789 0.0074 0.0079 10.184 23.577 38 2.478 Road to Navojoa 1 4 4 1 0.0313 0.0017 4 6.338 7 -0.825 Choquincahui 5 5 1 0.016 0.0038 9 25.024 17 -0.48 La Sierrita 7 5 0.905 0.0107 0.0147 16.762 72.519 34 2.782 Los Camotes 4 4 1 0.0313 0.0081 17.667 100.111 35 0.978 Road to Navojoa 2 10 7 0.911 0.006 0.0032 5.533 8.442 15 -0.496 Arroyo Tabelo 14 7 0.879 0.0033 0.0075 11.209 29.352 41 3.106 Aduana 4 3 0.833 0.0495 0.0034 4.833 8.861 9 1.453 n = number of individuals; H = number of haplotypes; h = haplotype diversity; V(h) = variance in haplotype diversity; π = nucleotide diversity; k = average number of nucleotide differences; V(k) = variance in average number of nucleotide differences; S = number of segregating sites. Fu's Fs calculated from the total number of mutations. All Fu's Fs non-significant with an alpha=0.05. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 197
Table 2. DNA divergence between populations. Numbers above diagonal are the average number of nucleotide differences and numbers below correspond to the average number of substitutions per site (Dxy).
Population A B C D E F G H I
A. Alamos - 27.6330 13.9170 136.3000 13.8570 50.0830 12.1670 14.3450 12.4170 B. Rio Cuchujaqui 0.0214 - 31.5000 72.3700 7.6570 22.5750 27.5400 26.2710 30.9250 C. Road to Navojoa 1 0.0061 0.0244 - 135.1000 11.7140 51.2500 5.4500 9.0000 4.0000 D. Choquincahui 0.0594 0.0560 0.0587 - 70.5430 125.3000 106.5000 83.1570 119.1500 E. La Sierrita 0.0122 0.0145 0.0103 0.0619 - 20.7500 11.5140 8.6940 12.1430 F. Los Camotes 0.0230 0.0184 0.0235 0.0573 0.0182 - 41.7000 34.3210 44.7500 G. Road to Navojoa 2 0.0070 0.0245 0.0031 0.0614 0.0106 0.0240 - 7.5290 4.7500 H. Arroyo Tabelo 0.0096 0.0229 0.0060 0.0556 0.0122 0.0247 0.0058 - 8.3570 I. Aduana 0.0063 0.0239 0.0002 0.0602 0.0107 0.0242 0.0027 0.0056 - CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 198
Table 3. Estimates of contemporary migration rates between populations. Values in parentheses represent 95% confidence intervals. Bold values on diagonal represent the proportion of individuals derived from the source population each generation.
Locality 1 Road to Navojoa 2 Alamos 3 Tabelo 4 Aduana 5 Cuchujaqui 6 Sierrita 7 Mocuzari 8 El Quintero 9 Choquincahui 10 San Antonio
1 Road to Navojoa 0.685 0.018 0.020 0.155 0.031 0.021 0.018 0.018 0.017 0.017 (0.651–0.719) (-0.015–0.051) (-0.016–0.056) (0.080–0.229) (-0.021–0.083) (-0.017–0.059) (-0.015–0.052) (-0.015–0.050) (-0.015–0.049) (-0.014–0.049) 2 Alamos 0.007 0.834 0.030 0.011 0.059 0.018 0.017 0.010 0.007 0.007 (-0.007–0.021) (0.766–0.902) (-0.001–0.061) (-0.010–0.032) (0.001–0.117) (-0.013–0.048) (-0.011–0.044) (-0.008–0.028) (-0.007–0.021) (-0.007–0.022) 3 Tabelo 0.005 0.007 0.937 0.006 0.011 0.005 0.014 0.005 0.005 0.005 (-0.005–0.015) (-0.006–0.020) (0.900–0.974) (-0.005–0.017) (-0.005–0.027) (-0.005–0.015) (-0.011–0.039) (-0.005–0.015) (-0.005–0.015) (-0.005–0.015) 4 Aduana 0.012 0.028 0.020 0.824 0.022 0.031 0.026 0.013 0.012 0.012 (-0.011–0.035) (-0.020–0.077) (-0.016–0.056) (0.731–0.916) (-0.018–0.061) (-0.020–0.082) (-0.014–0.065) (-0.012–0.038) (-0.011–0.035) (-0.011–0.035) 5 Cuchujaqui 0.004 0.007 0.006 0.009 0.940 0.007 0.006 0.012 0.004 0.004 (-0.004–0.012) (-0.006–0.021) (-0.005–0.017) (-0.007–0.025) (0.907–0.973) (-0.006–0.020) (-0.005–0.016) (-0.003–0.027) (-0.004–0.012) (-0.004–0.013) 6 Sierrita 0.007 0.010 0.010 0.021 0.023 0.891 0.018 0.007 0.007 0.007 (-0.006–0.020) (-0.009–0.029) (-0.009–0.029) (-0.013–0.054) (-0.010–0.055) (0.841–0.941) (-0.009–0.046) (-0.006–0.020) (-0.006–0.020) (-0.007–0.021) 7 Mocuzari 0.008 0.013 0.021 0.010 0.011 0.011 0.898 0.012 0.008 0.008 (-0.007–0.023) (-0.010–0.035) (-0.011–0.054) (-0.009–0.030) (-0.009–0.031) (-0.009–0.031) (0.848–0.949) (-0.009–0.032) (-0.007–0.023) (-0.007–0.023) 8 El Quintero 0.008 0.009 0.008 0.009 0.014 0.009 0.008 0.917 0.008 0.008 (-0.008–0.024) (-0.008–0.027) (-0.008–0.024) (-0.008–0.027) (-0.010–0.037) (-0.008–0.026) (-0.008–0.025) (0.873–0.961) (-0.007–0.024) (-0.008–0.024) 9 Choquincahui 0.008 0.008 0.010 0.008 0.009 0.009 0.008 0.256 0.675 0.009 (-0.008–0.024) (-0.008–0.025) (-0.009–0.029) (-0.008–0.024) (-0.008–0.025) (-0.008–0.025) (-0.008–0.024) (0.208–0.303) (0.659–0.692) (-0.008–0.026) 10 San Antonio 0.014 0.095 0.014 0.061 0.043 0.049 0.014 0.014 0.014 0.684 (-0.012–0.039) (0.013–0.177) (-0.012–0.039) (-0.033–0.155) (-0.012–0.098) (-0.009–0.106) (-0.012–0.040) (-0.012–0.040) (-0.012–0.040) (0.649–0.718)
CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 199
Table 4. Estimates of contemporary and historical effective population sizes (Ne) and 95% confidence intervals. Results in first three columns represent Ne estimates based on the linkage disequilibrium method employed in the software LDNe. Results are reported based on lowest allele frequency values used (0.05, 0.02, 0.01). Estimates in last column represent historical Ne estimates using the coalescent model in -4 MIGRATE and calculated assuming a mutation rate of 5 x 10 (Garza & Williamson 2001).
Lowest allele frequency used
Locality n 0.05 0.02 0.01 MIGRATE
Road to Navojoa 1 10 25.4 (15.5–56.0) 25.4 (15.5–56.0) 25.4 (15.5–56.0) 483.355 (0–1333) Alamos 36 172.4 (65.2–infinity) 464.1 (104.8–infinity) 362.9 (106.5–infinity) 749.5 (0–1566) Arroyo Tabelo 57 26.8 (20.7–35.5) 36.4 (28.1–48.7) 43.7 (33.5–59.6) 703.34 (0–1500) Aduana 17 132.5 (33.6–infinity) -331 (74.1–infinity) -331 (74.1–infinity) 501.685 (0–1333) Rio Cuchujaqui 72 -3903.1 (308.0–infinity) 636.7 (225.3–infinity) 654.9 (226.4–infinity) 1456.435 (466–2400) Sierrita 38 148.2 (57.1–infinity) 142.9 (65.2–infinity) 239.7 (91.6–infinity) 668.235 (0–1500) Mocuzari 30 28.8 (19.6–47.3) 29.2 (21.0–43.9) 38.4 (26.7–62.1) 272.6 (0–1066) El Quintero 30 442.8 (62.5–infinity) -865.0 (110.9–infinity) 584.8 (90.6–infinity) 432.8 (0–1266) Choquincahui 31 1093.3 (57.4–infinity) 121.3 (43.9–infinity) 95.6 (41.7–infinity) 296.295 (0–1100) San Antonio 15 38.8 (15.3–infinity) 224.4 (45.2–infinity) 224.4 (45.2–infinity) 323.865 (0–1133) CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 200
Table 5. Results of tests for population bottlenecks performed using the software BOTTLENECK. Numbers represent P-values based on the corresponding mutation model. Allelic distribution represents the test for a shift in the relative abundance of alleles at different frequencies.
Allelic Locality n Wilcoxon TPM Wilcoxon SMM Wilcoxon IAM Distribution
Road to Navojoa 1 10 0.539 0.652 0.042 Shifted Alamos 36 0.216 0.348 0.001 Normal Arroyo Tabelo 57 0.385 0.722 0.001 Normal Aduana 17 0.385 0.577 0.001 Normal Rio Cuchujaqui 72 0.116 0.216 0.001 Normal La Sierrita 38 0.935 0.998 0.065 Normal Mocuzari 30 0.688 0.884 0.001 Normal El Quintero 30 0.813 0.935 0.005 Normal Choquincahui 31 0.991 0.998 0.001 Normal San Antonio 15 0.652 0.784 0.002 Normal
TPM = two-phase model; SMM = stepwise mutation model; IAM = infinite allele model CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 201
Table 6. Regression on distance matrix results showing the relationships between landscape resistance and historical migration, contemporary migration, and mtDNA sequence divergence (Dxy). Model M = MIGRATE; Model B = BAYESASS; Model MtDNA = mtDNA divergence. Models in italics significant at an alpha = 0.05.
Model Variables β Model R2 P
M1 Euclidean -8.2E-09 0.0337 0.231 M2 Temp -1.3E-05 0.0742 0.0795 M3 Forest -2.3E-05 0.0514 0.1447 M4 Stream -4E-05 0.0497 0.1578 M5 Anthro -3.1E-06 0.0073 0.6023 M6 Slope -1.3E-05 0.1145 0.0217
B1 Euclidean -8.1E-07 0.1994 0.0016 B2 Temp -0.0005 0.0587 0.0656 B3 Forest -0.0017 0.1768 0.0077 B4 Stream -0.0027 0.1342 0.008 B5 Anthro -0.0006 0.1703 0.002 B6 Slope -0.0002 0.0214 0.2446
Mt1 Euclidean 1.78E-06 0.6235 0.0193 Mt2 Temp 0.002 0.558 0.1135 Mt3 Forest 0.003 0.3978 0.0724 Mt4 Stream 0.003 0.1159 0.5357 Mt5 Anthro 0.0001 0.0029 0.8827 Mt6 Slope 0.001 0.5219 0.1012 CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 202
Figure Legends
Fig. 1. Map of the Alamos region of Sonora showing sampling sites for leaf-toed geckos,
Phyllodactylus, included in the study. Site names in italics represent sites where mtDNA sequences were obtained.
Fig. 2. A) Maximum parsimony (MP) strict consensus tree of eight most parsimonious trees
based on the concatenated mtDNA data for leaf-toed geckos, Phyllodactylus. Numbers above
nodes represent 10,000 nonparametric bootstrap proportions. Numbers below nodes represent
Bayesian posterior probabilities sampled from the posterior distribution of trees. B) Bayesian
majority rule phylogram resulting from the optimal partitioning strategy of gene + codon
position.
Fig. 3. Geographic distribution of major mitochondrial lineages for leaf-toed geckos,
Phyllodactylus, from throughout the Alamos region. Areas of sympatry are represented by
localities encircled by more than one color.
Fig. 4. Scatter plot with Reduced Major Axis (RMA) regression line illustrating isolation by
distance for leaf-toed geckos, Phyllodactylus, resulting from a Mantel test of 30 000
randomizations. PhiST/(1-PhiST) values used to calculate test statistic were based on the
uncorrected p-distances between haplotypes. Geographic distances shown in kilometers.
CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 203
Fig. 5. A) Frequency histogram of historical migration rates for leaf-toed geckos,
Phyllodactylus, estimated from MIGRATE. B) Frequency histogram of contemporary migration rates estimated from BAYESASS. CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 204
Fig. 1 CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 205
Fig. 2 CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 206
Fig. 3 Lineage A
Lineage B
1 Lineage C
Lineage D
0.6 CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 207
Fig. 4 CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 208
Fig. 5 A Histogram of Migrate Frequency 0 5 10 15 20 25
0.000 0.001 0.002 0.003 0.004 0.005 Migration Rate
Histogram of BayesAss B Frequency 0 20406080
0.00 0.05 0.10 0.15 0.20 0.25 0.30 Migration Rate CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 209
Supplementary Material
Appendix 1 Population and outgroup sampling with corresponding GenBank accession numbers for all Phyllodactylus sequenced for mtDNA. Population ID for each population follows that presented in all population genetic analyses. N/A = Not applicable.
GenBank Accession Numbers
Genus Species Country State Population Location Field # Latitude Longitude 16S COI Glutamine Isoleucine Leucine ND1 ND4
Phyllodactylus saxatilis Mexico Sonora A Alamos ROM 40772 27 01' 20.0" N 108 55' 18.0" W
Phyllodactylus saxatilis Mexico Sonora A Alamos ROM 40773 27 01' 20.0" N 108 55' 18.0" W
Phyllodactylus saxatilis Mexico Sonora A Alamos ROM 40770 27 01' 20.0" N 108 55' 18.0" W
Phyllodactylus saxatilis Mexico Sonora A Alamos ROM 40771 27 01' 20.0" N 108 55' 18.0" W
Phyllodactylus saxatilis Mexico Sonora A Alamos ROM 40779 27 01' 20.0" N 108 55' 18.0" W
Phyllodactylus saxatilis Mexico Sonora A Alamos ROM 40794 27 01' 20.0" N 108 55' 18.0" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui CB121 26 56' 31.2" N 108 53' 16.7" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui CB122 26 56' 31.2" N 108 53' 16.7" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 41121 26 56' 31.2" N 108 53' 16.7" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 41122 26 56' 31.2" N 108 53' 16.7" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 41125 26 56' 30.5" N 108 53' 16.3" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 41126 26 56' 30.5" N 108 53' 16.3" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 41127 26 56' 30.5" N 108 53' 16.3" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 41128 26 56' 30.5" N 108 53' 16.3" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40730 26 56' 30.5" N 108 52' 59.4" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40744 26 54' 45.5" N 108 54" 49.9" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40743 26 54' 45.5" N 108 54" 49.9" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40742 26 54' 45.5" N 108 54" 49.9" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40733 26 56' 30.5" N 108 52' 59.4" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40731 26 56' 30.5" N 108 52' 59.4" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40729 26 56' 30.5" N 108 52' 59.4" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40728 26 56' 30.5" N 108 52' 59.4" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40727 26 56' 30.5" N 108 52' 59.4" W CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 210
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40726 26 56' 30.5" N 108 52' 59.4" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40725 26 56' 30.5" N 108 52' 59.4" W
Phyllodactylus saxatilis Mexico Sonora B Rio Cuchujaqui ROM 40732 26 56' 30.5" N 108 52' 59.4" W
Phyllodactylus saxatilis Mexico Sonora C ca. 26 km W of Alamos on road to Navojoa; rock face in arroyo ROM 40749 27 02' 36.1" N 109 00' 08.3" W
Phyllodactylus saxatilis Mexico Sonora C ca. 26 km W of Alamos on road to Navojoa; rock face in arroyo ROM 40746 27 02' 36.1" N 109 00' 08.3" W
Phyllodactylus saxatilis Mexico Sonora C ca. 26 km W of Alamos on road to Navojoa; rock face in arroyo ROM 40747 27 02' 36.1" N 109 00' 08.3" W
Phyllodactylus saxatilis Mexico Sonora C ca. 26 km W of Alamos on road to Navojoa; rock face in arroyo ROM 40748 27 02' 36.1" N 109 00' 08.3" W
Phyllodactylus saxatilis Mexico Sonora D Choquincahui, at stream in arroyo ROM 40800 26 58' 57.6" N 108 40' 41.1" W
Phyllodactylus saxatilis Mexico Sonora D Choquincahui, at stream in arroyo ROM 40799 26 58' 57.6" N 108 40' 41.1" W
Phyllodactylus saxatilis Mexico Sonora D Choquincahui, at stream in arroyo ROM 40798 26 58' 57.6" N 108 40' 41.1" W
Phyllodactylus saxatilis Mexico Sonora D Choquincahui, at stream in arroyo ROM 40797 26 58' 57.6" N 108 40' 41.1" W
Phyllodactylus saxatilis Mexico Sonora D Choquincahui, at stream in arroyo ROM 40796 26 58' 57.6" N 108 40' 41.1" W
Phyllodactylus saxatilis Mexico Sonora E La Sierrita Nature Reserve ROM 40793 26 59' 44.1" N 108 56' 29.3" W
Phyllodactylus saxatilis Mexico Sonora E La Sierrita Nature Reserve ROM 40787 26 59' 44.1" N 108 56' 29.3" W
Phyllodactylus saxatilis Mexico Sonora E La Sierrita Nature Reserve ROM 40788 26 59' 44.1" N 108 56' 29.3" W
Phyllodactylus saxatilis Mexico Sonora E La Sierrita Nature Reserve ROM 40789 26 59' 44.1" N 108 56' 29.3" W
Phyllodactylus saxatilis Mexico Sonora E La Sierrita Nature Reserve ROM 40790 26 59' 44.1" N 108 56' 29.3" W
Phyllodactylus saxatilis Mexico Sonora E La Sierrita Nature Reserve ROM 40792 26 59' 44.1" N 108 56' 29.3" W
Phyllodactylus saxatilis Mexico Sonora E La Sierrita Nature Reserve ROM 40791 26 59' 44.1" N 108 56' 29.3" W
Phyllodactylus saxatilis Mexico Sonora F Los Camotes, ca. 40 km NE Alamos Cemetery by lower road to Milpillas; large arroyo ROM 40723 27 16' 07.4" N 108 50' 05.2" W
Phyllodactylus saxatilis Mexico Sonora F Los Camotes, ca. 40 km NE Alamos Cemetery by lower road to Milpillas; large arroyo ROM 40751 27 16' 07.4" N 108 50' 05.2" W
Phyllodactylus saxatilis Mexico Sonora F Los Camotes, ca. 40 km NE Alamos Cemetery by lower road to Milpillas; large arroyo ROM 40722 27 16' 07.4" N 108 50' 05.2" W
Phyllodactylus saxatilis Mexico Sonora F Los Camotes, ca. 40 km NE Alamos Cemetery by lower road to Milpillas; large arroyo ROM 40724 27 16' 07.4" N 108 50' 05.2" W
Phyllodactylus saxatilis Mexico Sonora G Near road to Navojoa from Alamos ROM 40763 27 02' 10.9" N 109 04' 59.6" W
Phyllodactylus saxatilis Mexico Sonora G Near road to Navojoa from Alamos ROM 40775 27 02' 36.7" N 109 00' 02.9" W
Phyllodactylus saxatilis Mexico Sonora G Near road to Navojoa from Alamos ROM 40762 27 02' 10.9" N 109 04' 59.6" W
Phyllodactylus saxatilis Mexico Sonora G Near road to Navojoa from Alamos ROM 40764 27 02' 10.9" N 109 04' 59.6" W
Phyllodactylus saxatilis Mexico Sonora G Near road to Navojoa from Alamos ROM 40767 27 02' 10.9" N 109 04' 59.6" W
Phyllodactylus saxatilis Mexico Sonora G Near road to Navojoa from Alamos ROM 40778 27 02' 36.7" N 109 00' 02.9" W
Phyllodactylus saxatilis Mexico Sonora G Near road to Navojoa from Alamos ROM 40777 27 02' 36.7" N 109 00' 02.9" W
Phyllodactylus saxatilis Mexico Sonora G Near road to Navojoa from Alamos ROM 40776 27 02' 36.7" N 109 00' 02.9" W
Phyllodactylus saxatilis Mexico Sonora G Near road to Navojoa from Alamos ROM 40768 27 02' 10.9" N 109 04' 59.6" W
Phyllodactylus saxatilis Mexico Sonora G Near road to Navojoa from Alamos ROM 40761 27 02' 10.9" N 109 04' 59.6" W CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 211
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 40760 27 09' 59.6" N 108 57' 43.0" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 40759 27 09' 59.6" N 108 57' 43.0" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 40758 27 09' 59.6" N 108 57' 43.0" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 40757 27 09' 59.6" N 108 57' 43.0" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 40756 27 09' 59.6" N 108 57' 43.0" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 40755 27 09' 59.6" N 108 57' 43.0" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 41109 27 09' 59.1" N 108 57' 42.7" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 41110 27 09' 59.1" N 108 57' 42.7" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 41111 27 09' 59.1" N 108 57' 42.7" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 41112 27 09' 59.1" N 108 57' 42.7" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 41113 27 09' 57.3" N 108 57' 43.1" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 41114 27 09' 57.3" N 108 57' 43.1" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 41115 27 09' 57.3" N 108 57' 43.1" W
Phyllodactylus saxatilis Mexico Sonora H Arroyo Tabelo ROM 41116 27 09' 57.3" N 108 57' 43.1" W
Phyllodactylus saxatilis Mexico Sonora I Arroyo Aduana ROM 41117 27 02' 10.5" N 109 00' 46.1" W
Phyllodactylus saxatilis Mexico Sonora I Arroyo Aduana ROM 41118 27 02' 10.5" N 109 00' 46.1" W
Phyllodactylus saxatilis Mexico Sonora I Arroyo Aduana ROM 41119 27 02' 10.5" N 109 00' 46.1" W
Phyllodactylus saxatilis Mexico Sonora I Arroyo Aduana ROM 41120 27 02' 10.5" N 109 00' 46.1" W
Phyllodactylus saxatilis Mexico Sonora N/A El Palmarito ROM 41135 27 01' 57.6" N 108 47' 04.9" W
Phyllodactylus saxatilis Mexico Sonora N/A ca. 55 km NE Alamos Cemetery by lower road to Milpillas, large boulders near stream ROM 40754 27 14' 44.1" N 108 45' 26.3" W
Phyllodactylus saxatilis Mexico Sonora N/A ca. 55 km NE Alamos Cemetery by lower road to Milpillas, large boulders near stream ROM 40753 27 14' 44.1" N 108 45' 26.3" W
Phyllodactylus homolepidurus Mexico Sonora N/A Isla San Pedro Nolasco ROM 38027 27 58' 05" N 111 25' 04" W
Phyllodactylus xanti Mexico Baja California Sur N/A San Bartolo, 1 km N of ROM 34047 23 44' 09" N 109 51' 12" W CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 212
Supplementary Table 1. Gene, number of sequenced bases, primer name and sequence, and reference for all mtDNA loci used.
Gene Base Pairs Primer Primer Sequence Source
COI 400 COI_m72L 5’–TGA TTC TTC GGT CAC CCA GAA GTG TA–3’ Austin et al. 2004 COI_m73H 5’– CCT ATT GAT AGG ACG TAG TGG AAG TG–3’ Austin et al. 2004
ND4 700 ND4_phyL 5’–CCT ATC CCC CAC AAC CCA AAC–3’ Blair et al. 2009 ND4_13824H 5’–CAT TAC TTT TAC TTG GAT TTG CAC CA–3’ Arevalo et al. 1994
ND1, 16S, 1300 ND1_16dR 5’–CTA CGT GAT CTG AGT TCA GAC CGG AG–3’ Leaché and McGuire 2006 tRNAleu, tRNAiso, tRNAglu ND1_tMet 5’–ACC AAC ATT TTC GGG GTA TGG GC–3’ Leaché and McGuire 2006
ND1 970 ND1_phyF 5’–GCC CTA GTA GTC TGA ACC CC–3’ This study ND1_phyR 5’–GTG CGA AGG AGC CTG CTG CG–3’ This study CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 213
Supplementary Table 2. Total and number of potentially phylogenetically informative characters for each data partition along with the chosen model of nucleotide substitution based on AIC.
Partition Characters Phylogenetically Model Informative
16S 114 10 GTR COI 393 31 SYM+Γ COI 1st 131 2 GTR COI 2nd 131 1 F81 COI 3rd 131 28 GTR+Γ Glutamine 67 6 HKY+I Isoleucine 73 4 K80 Leucine 74 2 JC ND1 966 107 GTR+I+Γ ND1 1st 322 18 HKY+I ND1 2nd 322 6 HKY ND1 3rd 322 83 GTR+Γ ND4 712 81 GTR+Γ ND4 1st 238 16 HKY+Γ ND4 2nd 237 6 GTR+I ND4 3rd 237 59 GTR+Γ All 2399 241 GTR+I+Γ CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 214
Supplementary Table 3. Alternate partitioning strategies tested using Bayes Factors.
Strategy Partition Arithmetic Mean Harmonic Mean 2ln Bayes factor
P1 All data -7504.25 -7597.52 858.7 P2 16S; COI, Glut; Iso; Leu; ND1; ND4 -7230.98 -7312.06 287.78
P3 16S; COI 1st; COI 2nd; COI 3rd; Glut; Iso; Leu; ND1 -7084.85 -7168.17 N/A 1st; ND1 2nd; ND1 3rd; ND4 1st; ND4 2nd; ND4 3rd
Bayes factors were calculated by comparing the harmonic means of different partitioning strategies. P3 served as the null hypothesis.
CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 215
Supplementary Table 4. MIGRATE estimates of historical asymmetric migration rates between populations. Locus 'All' signifies that all 10 microsatellite loci were used for calculations. ϴ = 4Neµ, M = m/µ, Nm = number of migrants per generation (ϴM/4). 1 = Road to Navojoa, 2 = Alamos, 3 = Arroyo Tabelo, 4 = Aduana, 5 = Rio Cuchujaqui, 6 = La Sierrita, 7 = Mocuzari, 8 = El Quintero, 9 = Choquincahui, 10 = San Antonio.
Locus Parameter 2.50% 25.00% mode 75% 97.50% median mean 4Nm Nm
All M_2->1 0 0.07 0.97 1.8 4.47 1.7 1.7 1.940.49 All M_3->1 2.4 9.13 10.97 12.6 15.73 10.63 10.22 9.88 2.47 All M_4->1 0.33 2.07 3.43 4.8 8.13 4.1 5.41 5.23 1.31 All M_5->1 1.73 3.47 4.5 6.13 12.2 6.77 6.84 6.61 1.65 All M_6->1 1.33 3.53 4.9 6.4 10.07 5.37 5.56 5.37 1.34 All M_7->1 0.13 1.73 2.83 3.8 5.67 2.97 2.94 2.84 0.71 All M_8->1 0 5.33 6.43 7.47 10.2 6.57 6.78 6.55 1.64 All M_9->1 0 1.67 2.77 3.93 6.2 3.1 3.11 3.01 0.75 All M_10->1 0 0.87 1.9 2.87 7.67 2.3 2.59 2.50.63 All M_1->2 0 0 0.5 1.2 2.93 1.23 1.05 1.570.39 All M_3->2 0.4 1.73 2.63 3.47 4.8 2.7 2.62 3.93 0.98 All M_4->2 0 0 0.03 1.07 2.6 1.1 0.85 1.270.32 All M_5->2 1.4 2.33 3.43 4.6 9.6 5.23 5.37 8.05 2.01 All M_6->2 0.4 1.73 2.63 3.47 4.87 2.77 2.69 4.03 1.01 All M_7->2 0 0 0.03 0.87 2.27 0.9 0.64 0.990.25 All M_8->2 0 0.53 1.3 2 3.13 1.5 1.32 1.98 0.50 All M_9->2 0.07 1.2 2.03 2.8 4 2.17 2.05 3.070.77 All M_10->2 0 0 0.03 1 2.67 1.03 0.88 1.320.33 All M_1->3 0 0 0.03 0.67 1.93 0.7 0.41 0.580.15 All M_2->3 0 0.13 0.83 1.47 2.8 1.3 1.06 1.49 0.37 All M_4->3 0 0 0.03 0.67 1.8 0.7 0.37 0.520.13 All M_5->3 0.6 1.93 2.83 3.6 4.93 2.9 2.83 3.98 1.00 All M_6->3 0 0 0.03 0.67 2.2 0.7 0.6 0.840.21 All M_7->3 0 0 0.03 0.87 2.27 0.9 0.63 0.890.22 All M_8->3 0 0 0.5 1.13 2.67 1.17 0.91 1.280.32 All M_9->3 0 0 0.03 0.8 2.13 0.83 0.54 0.760.19 All M_10->3 0 0 0.03 0.73 2.07 0.77 0.5 0.70.18 All M_1->4 0 0 0.03 0.8 2.33 0.83 0.64 0.640.16 All M_2->4 1.2 2.8 3.9 4.87 7.2 4.1 4.13 4.141.04 All M_3->4 0 0 0.17 1.2 4 1.23 1.26 1.260.32 All M_5->4 1.67 3.27 4.3 5.2 6.93 4.37 4.33 4.34 1.09 All M_6->4 4.13 6 7.17 8.2 10.13 7.23 7.19 7.211.80 All M_7->4 0 0 0.03 1.2 3.33 1.23 1.11 1.110.28 All M_8->4 0 2.4 3.77 5.07 6.27 3.57 3.28 3.290.82 All M_9->4 0 1.13 2.1 3 4.73 2.3 2.26 2.27 0.57 All M_10->4 0 0 0.23 1.13 2.8 1.17 0.97 0.970.24 All M_1->5 0 0 0.03 0.93 2.47 0.97 0.74 2.160.54 All M_2->5 0 0.73 1.57 2.27 3.33 1.7 1.56 4.54 1.14 All M_3->5 0 0.93 1.7 2.47 3.47 1.83 1.71 4.981.25 All M_4->5 0 0 0.03 0.73 2.07 0.77 0.49 1.430.36 All M_6->5 0 0.93 1.7 2.47 3.47 1.83 1.71 4.981.25 All M_7->5 0 0 0.5 1.13 2.6 1.17 0.88 2.560.64 CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 216
All M_8->5 0 0 0.1 1 2.47 1.03 0.78 2.270.57 All M_9->5 0 0.6 1.37 2.07 3.07 1.57 1.37 3.991.00 All M_10->5 0 0 0.03 1 2.47 1.03 0.76 2.210.55 All M_1->6 0 0 0.03 0.73 2.07 0.77 0.5 0.690.17 All M_2->6 0 0.87 1.7 2.47 3.6 1.83 1.71 2.290.57 All M_3->6 0 0.87 1.7 2.53 4 1.9 1.82 2.430.61 All M_4->6 0 0 0.03 0.8 2.27 0.83 0.62 0.830.21 All M_5->6 0.73 2.07 2.9 3.73 5.07 3.03 2.91 3.89 0.97 All M_7->6 0 0 0.03 1 2.6 1.03 0.83 1.110.28 All M_8->6 0 0.47 1.17 1.87 3 1.43 1.24 1.66 0.42 All M_9->6 0 0.53 1.3 2 3.2 1.57 1.35 1.80.45 All M_10->6 0 0 0.03 0.67 1.8 0.7 0.36 0.480.12 All M_1->7 0 0.87 1.9 2.8 4.93 2.23 2.19 1.19 0.30 All M_2->7 0 0.8 1.77 2.53 4.13 1.97 1.87 1.020.26 All M_3->7 4.4 6.13 7.23 8.27 10.8 7.5 7.53 4.11 1.03 All M_4->7 0 0 0.03 0.8 2.33 0.83 0.65 0.350.09 All M_5->7 1.47 3.13 4.17 5.2 7.07 4.3 4.28 2.33 0.58 All M_6->7 0.8 3.2 4.43 5.6 7.67 4.43 4.32 2.36 0.59 All M_8->7 0 0 0.03 1.13 3.33 1.17 1.12 0.610.15 All M_9->7 0 0.4 1.17 1.87 3.27 1.5 1.34 0.730.18 All M_10->7 0 1.27 2.17 3.07 4.47 2.3 2.26 1.23 0.31 All M_1->8 0 0 0.03 0.93 2.47 0.97 0.75 0.650.16 All M_2->8 0 0.13 0.77 1.4 2.8 1.23 1.02 0.88 0.22 All M_3->8 0.47 1.8 2.63 3.47 4.73 2.7 2.65 2.290.57 All M_4->8 0 0 0.03 0.87 2.47 0.9 0.73 0.630.16 All M_5->8 0 1.2 2.1 2.87 4.2 2.23 2.17 1.88 0.47 All M_6->8 0 0.87 1.63 2.4 3.53 1.83 1.66 1.44 0.36 All M_7->8 0 0.53 1.3 1.93 3 1.5 1.31 1.130.28 All M_9->8 1.53 3.07 3.97 4.87 6.4 4.03 4.01 3.47 0.87 All M_10->8 0 0 0.63 1.2 2.87 1.23 1.04 0.90.23 All M_1->9 0 0 0.3 1.13 2.67 1.17 0.91 0.540.14 All M_2->9 0 0.4 1.17 1.87 3.13 1.5 1.29 0.760.19 All M_3->9 0 1.13 2.03 2.87 4.27 2.23 2.11 1.25 0.31 All M_4->9 0 0 0.17 1.13 2.87 1.17 1 0.590.15 All M_5->9 0 1.13 1.97 2.8 4.07 2.17 2.03 1.20.30 All M_6->9 0 0.07 0.7 1.33 2.67 1.23 0.92 0.550.14 All M_7->9 0 0.8 1.63 2.4 3.8 1.83 1.72 1.020.26 All M_8->9 0 0.73 1.7 2.53 4.27 1.97 1.87 1.110.28 All M_10->9 0 0 0.7 1.33 2.87 1.3 1.06 0.630.16 All M_1->10 0 0 0.03 0.87 2.53 0.9 0.75 0.490.12 All M_2->10 2.53 5.67 7.7 9.47 13.73 7.9 8.07 5.23 1.31 All M_3->10 3.67 5.87 7.3 8.87 12.13 7.7 7.77 5.03 1.26 All M_4->10 0.47 2.33 3.5 4.67 6.87 3.7 3.69 2.390.60 All M_5->10 2.67 4.4 5.5 6.47 8.73 5.63 5.68 3.68 0.92 All M_6->10 0 0 0.03 1.13 3 1.17 1.03 0.670.17 All M_7->10 0 0 0.03 1.2 3.07 1.23 1.08 0.70.18 All M_8->10 0.6 2.33 3.43 4.47 6.6 3.63 3.61 2.34 0.59 All M_9->10 0 0.93 1.83 2.6 3.87 1.97 1.87 1.21 0.30 CHAPTER 5—HISTORICAL AND CONTEMPORARY DEMOGRAPHY 217
Supplementary Table 5. Multiple regression on distance matrices (MRM) results showing the relationship between contemporary migration rates (estimates using BAYESASS) and resistance distances incorporating landscape heterogeneity. Optimal cost values used to parameterize resistance surfaces prior to calculating resistance distances were selected based on Mantel r correlation coefficients. VIF = Variance Inflation Factor.
2 Model Variables β P Model R P VIF Model AICc ∆AIC Akaike Weight (wi) A1 Euclidean -8.09E-07 0.0016 0.1994 0.0016 -221.09 0 0.1604 A2 Temp -5.00E-04 0.0656 0.0587 0.0656 -213.81 7.28 0.0042 A3 Forest -1.70E-03 0.0077 0.1768 0.0077 -219.84 1.25 0.0859 A4 Stream -2.70E-03 0.008 0.1342 0.008 -217.57 3.52 0.0276 A5 Anthro -6.00E-04 0.002 0.1703 0.002 -219.49 1.6 0.0721 A6 Slope -2.00E-04 0.2446 0.0214 0.2446 -212.06 9.03 0.0018
B1 Euclidean -7.77E-07 0.0014 0.2004 0.0023 1.29 -218.74 2.35 0.0495 Temp -7.37E-05 0.7617
B2 Euclidean -5.53E-07 0.0883 0.211 0.0062 2.71 -219.34 1.75 0.0669 Forest -7.26E-04 0.3352
B3 Euclidean -6.76E-07 0.0169 0.205 0.0034 1.97 -218.99 2.1 0.0561 Stream -7.68E-04 0.495
B4 Euclidean -5.73E-07 0.0203 0.2412 0.0018 1.41 -221.09 0 0.1604 Anthro -3.61E-04 0.0361
B5 Euclidean -8.84E-07 0.0018 0.2052 0.0021 1.3 -219 2.09 0.0564 Slope 1.30E-04 0.468
C1 Euclidean -2.23E-07 0.4906 0.2727 0.0045 2.59 -220.46 0.63 0.1171 Anthro -5.73E-04 0.0121 2.05 Temp -4.76E-04 0.101 1.88
C2 Euclidean -5.72E-07 0.0675 0.2412 0.0028 2.71 -218.55 2.54 0.0451 Anthro -3.61E-04 0.0692 1.95 Forest -1.49E-06 0.9987 3.76
C3 Euclidean -5.24E-07 0.0748 0.2421 0.0031 2.15 -218.61 2.48 0.0464 Anthro -3.49E-04 0.0482 1.48 Stream -3.24E-04 0.7745 2.06
C4 Euclidean -4.47E-07 0.1739 0.2447 0.0021 2.77 -218.76 2.33 0.05 Anthro -4.41E-04 0.0544 2.22 Slope -1.27E-04 0.5907 2.04
CHAPTER 6
Molecular phylogenetics and taxonomy of leaf-toed geckos
(Phyllodactylidae: Phyllodactylus) inhabiting the peninsula of
Baja California
Abstract
Herein I assess the phylogenetic relationships and taxonomy of geckos of the genus
Phyllodactylus inhabiting the peninsula of Baja California, Mexico using five mitochondrial
and two nuclear genes. Phylogenetic analysis using maximum parsimony (MP) and Bayesian
inference (BI) recovered three distinct peninsular clades with high statistical support.
Sequence divergence estimates between peninsular taxa approached 13%. Two of the
species, P. unctus and P. xanti are Cape Region endemics, whereas P. nocticolus is
widespread throughout much of the peninsula and extreme southern California. Monophyly
of the peninsular taxa was strongly supported. In the MP analysis, P. unctus rooted at the
base of the peninsular clade, resolving P. xanti and P. nocticolus as sister taxa. Conversely,
BI placed P. nocticolus and P. unctus as sister taxa. These data provide further evidence for a trans-peninsular seaway near the Isthmus of La Paz, severing the Cape Region from the rest of the peninsula. The analysis also supports the validity of P. nocticolus as a distinct species and suggests a single invasion to the peninsula from mainland Mexico, presumably during tectonic activity during the Miocene.
218 CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 219
1 Introduction
The complex geological history of Baja California and the Gulf of California has contributed
significantly to the evolution of their regional biota. One hypothesis on the formation of the
peninsula and the origin of its flora and fauna suggests a trans-gulfian vicariance model. This
posits that tectonic activity during the Miocene led to rifting partitions of western Mexico off
the North American Plate (Murphy 1983a,b; Hausback 1984; Lonsdale 1989; Riddle et al.
2000). This biogeographic model, in part, explains the patterns of evolutionary relationships
observed between several taxa distributed on opposite sides of the Gulf of California
(Murphy 1983b).
The dynamic paleogeographic history of the peninsula has, in turn, led to numerous
taxonomic and phylogenetic studies conducted throughout Baja California, islands in the
Gulf of California, and mainland northwestern Mexico. Several of these studies evaluated
genetic breaks in mitochondrial DNA (mtDNA) lineages that coincide with trans-peninsular
seaways, one mid-way on the peninsula and one across the Isthmus of La Paz, just north of
the Cape Region (Upton & Murphy 1997; Riddle et al. 2000; Lindell et al. 2005, 2006,
2008).
The entire peninsula of Baja California is home to three endemic species of leaf-toed geckos: Phyllodactylus unctus, P. xanti and P. nocticolus. The latter species has been
considered by some authorities to be a subspecies of P. xanti based on morphological data
(Dixon 1964, 1966; Grismer 2002). Analyses, based on morphology and allozymes suggested
that P. nocticolus warrants full species status (Murphy 1983b). Disjunct populations of these
and other species of Phyllodactylus also occur on islands of the Gulf of California (Dixon
1966; Murphy & Aguirre León 2002). These populations and species are not reported from CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 220 mainland Mexico and, thus, it is suggested that island colonization resulted from over-water dispersal from the peninsula of Baja California (Savage 1960, Murphy 1983a,b), translocation by Seri Indians (Murphy & Aguirre León 2002) and/or paleotectonic activity
(Murphy 1983a; Lonsdale 1989; Carreño & Helenes 2002).
The taxonomy of Phyllodactylus remains controversial, mainly due to significant morphological variation within species (Dixon 1964). Several subspecies of Phyllodactylus were erected to assist in the classification of intraspecific morphological variability (Dixon
1964, 1966). However, many insular subspecies distributed throughout the Gulf of California
Ire synonymized with their mainland counterparts due to the lack of diagnosable characters differentiating them from the latter (Grismer 1999). Molecular data for Phyllodactylus are also generally lacking, making phylogenetic relationships and taxonomic designations difficult to ascertain.
There are two main objectives of the present study. First, I provide further molecular data indicating the validity of P. nocticolus and second, I determine the phylogenetic relationships of peninsular and representative mainland forms and compare the results with previously published data to infer and discuss concordant biogeographic patterns throughout the peninsula of Baja California.
2 Materials and Methods
2.1 Tissue collection
Specimens and tissues were collected from populations of each species throughout the
peninsula of Baja California (Table 1; Fig. 1). Taxon identification follows Murphy (1983b). CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 221
Fieldwork was conducted in 1978, 1985, and 2000. Geographic coordinates were obtained for the latter year only, although additional geographic coordinates were obtained using
Google Earth. Voucher specimens were deposited in collections housed at the National
Autonomous University of Mexico (UNAM), California Academy of Sciences (CAS), Los
Angeles County Museum (LACM) and the Royal Ontario Museum (ROM). Additional sequences were obtained from P. lanei, P. paucituberculatus, P. duellmani, and P. davisi from mainland Mexico as primary outgroup taxa. GenBank sequences were also obtained for
Gekko gecko (AY282753) to root all networks.
2.2 DNA extraction, amplification, and sequencing
Total genomic DNA was digested and extracted from liver or muscle tissue using standard
phenol–chloroform protocols (Sambrook et al. 1989; Hillis et al. 1996; Palumbi 1996).
Tissue samples were first mixed with Proteinase K and Laird’s buffer and digested overnight prior to extraction with phenol–chloroform. Following extraction, samples were re-suspended in ddH2O prior to amplification.
Approximately 4300 bp encompassing two nuclear genes and five mitochondrial
genes were amplified using the polymerase chain reaction (PCR; Saiki et al. 1988). Nuclear
genes amplified included brain-derived neurotrophic factor (BDNF, 670 bp) and proto-
oncogene C-mos (380 bp). Mitochondrial loci included cytochrome c oxidase subunit I (COI,
420 bp), and NADH dehydrogenase subunit 4 (ND4, 690 bp). In addition, two ribosomal
RNA (rRNA) fragments and the intervening transfer RNA tValine (tRNAVal) were amplified,
encompassing 1600 bp of 16S rRNA and 900 bp of 12S rRNA. Primers used for CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 222 amplification and sequencing, their corresponding sequences and references were presented in Appendix I.
PCR reactions (25 μl) were composed of the following reaction mix: 18.55 μl ddH20,
1 μl 10 mM of each primer, 2.5 μl 1.5 mM MgCl2 buffer, 0.8 μl 10mM dNTPs, 0.15 μl 5 U
Taq DNA Polymerase (Boehringer Mannheim), and 1 μl template DNA. Amplification was
performed on a Perkin Elmer GeneAmp 9700 (Applied Biosystems) or a PT200 DNA Engine
(MJ Research) thermal cycler. Amplification conditions were as follows: initial denaturation
of 94 °C (2 min) folloId by 39 cycles of 94 °C (30 sec), 49-50 °C (45 sec), 72 °C (45 sec),
with a final extension temperature of 72 °C for 6 min (Lindell et al. 2005). Following
amplification, DNA products were separated on a 1% agarose gel stained with ethidium
bromide under ultraviolet (UV) light. Visible bands were then excised from the gel and centrifuged through a filter pipette tip for 10 min. The resulting solution was used as the
template for sequencing reactions.
Sequencing reactions (10 μl) were performed on an Eppendorf AG 5345 thermal
cycler with the BigDye Terminator v 3.1 Cycle Sequencing Kit (Applied Biosystems) using
the following recipe: 1 μl BigDye, 2 μl 5x BigDye Terminator Buffer, 2 μl ddH20, 1 μl of 10
mM primer solution, and 4 μl PCR product. Reactions were run for 25 cycles with the following conditions: initial denaturation at 96 °C (1 min), 96 °C (10 sec), 50 °C (5 sec), 60
°C (4 min), and 4 °C indefinitely. All gene regions were sequenced in both forward and reverse directions. Following sequencing reactions, samples were cleaned and precipitated with sodium acetate and ethanol and visualized on an ABI 3100 Automated Sequencer
(Applied Biosystems).
CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 223
2.3 Sequence alignment and phylogenetic analysis
Forward and reverse sequences were initially imported and edited in BIOEDIT v.5.0.6 (Hall
1999) and subsequently combined for manual alignment. Aligned sequences were then imported into MACCLADE v.4.08 (Maddison & Maddison 2005) and checked by eye. The final MACCLADE alignment was then exported into PAUP* v.4.0b10 (Swofford 2002) for phylogenetic inference.
Maximum parsimony analysis (MP) was performed on the concatenated dataset using
PAUP* v.4.0b10 (Swofford 2002). Uninformative characters (3373) were excluded from the analysis. All characters were treated as unordered with equal weighting. Gaps were treated as missing data. Because of the relatively small dataset (24 specimens), I used a branch-and- bound search strategy with furthest addition sequences. Branches were collapsed if the maximum branch length equaled zero. Support for nodes was assessed via 10,000 nonparametric bootstrap replicates using “fast” stepwise additions and retaining groups with
>50% support (Felsenstein, 1985).
Bayesian inference (BI) was also used to infer the phylogenetic relationships among
Baja Californian Phyllodactylus. To better account for heterogeneous rate variation across sites both within and between regions, I employed a partitioned analysis for each gene
(Ronquist & Huelsenbeck 2003; Nylander et al. 2004; Brandley et al. 2005). Theoretical and empirical evidence suggests several benefits to employing a partitioned Bayesian analysis to explain different subsets of data (gene, codon position, etc.) including higher –lnL scores and clade credibility values (i.e. posterior probabilities). Combining different genes and/or codon positions into one evolutionary model ignores the fact that different partitions are presumably CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 224 evolving at very different rates, which will in turn increase the systematic error involved in model fitting (Brandley et al. 2005).
MrModeltest v.2.2 (Nylander 2004) was used to obtain evolutionary models for each gene partition using the Akaike Information Criterion (AIC; Akaike 1974, 1979). The AIC was chosen over hierarchical likelihood ratio tests (hLRTs) for model selection because the former tends to penalize more over-parameterized models more strongly than the latter and, thus, minimizes the degree of variance involved in model fitting (Akaike 1974, 1979).
Bayesian inference was performed using MrBayes v.3.1 (Huelsenbeck & Ronquist 2001).
Two simultaneous runs of six chains were run for 3 X 106 generations, sampling every 100
generations. Stationarity was assessed when likelihood scores reached a stable equilibrium. A
burn-in value of 7500 was then implemented to discard topologies with low likelihood scores
prior to generating a 50% majority rule consensus tree. Node support within the Bayesian
consensus tree was assessed from the posterior distribution of topologies (Erixon et al. 2003).
In order to assess potentially conflicting phylogenetic signals from nuclear and
mtDNA, I performed MP and BI on each dataset independently. For both MP analyses, a
heuristic search was run with 100 replicate additions and tree bisection-reconnection (TBR)
branch swapping. Nodal support for each MP analysis was also assessed by nonparametric
bootstrapping with 10,000 “fast stepwise addition replicates (Felsenstein 1985). BI was
performed on each dataset independently under the same conditions as the concatenated
dataset described above. For all nuclear data, Gekko gecko, L12, and L103 were removed
prior to phylogenetic analysis due to missing data. Trees were thus rooted using P. lanei.
CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 225
3 Results
3.1 Genetic diversity
For the concatenated dataset, 915 characters were potentially phylogenetically informative.
Variability within nDNA was low with only eight characters potentially phylogenetically
informative. There was a significant difference in average base frequencies (χ2 = 164.80; df =
69; P < 0.001) with an adenine/cytosine bias. Percent sequence divergence based on uncorrected p-distances revealed significant differentiation between major groups (Table 2).
Further, divergence estimates between P. nocticolus and P. xanti approached 10%, clearly indicating the mtDNA distinctiveness of both species. MrModeltest selected the GTR+I+G model of nucleotide evolution for the COI, ND4, 12S and 16S partitions, the HKY model for
C-mos and BDNF and GTR+G for tRNAval.
3.2 Phylogenetic analysis of concatenated data set
The branch-and-bound search resulted in 2 most parsimonious trees (MPTs) of 2596 steps
(CI = 0.5458; RI = 0.6837; RC = 0.3732). The only difference between the two trees
involved the relationships among samples L4, L15, and L33 of P. nocticolus. In order to
illustrate branch lengths I chose one of the MPTs (Fig. 2).
BI and MP analyses of the concatenated dataset resulted in highly compatible topologies with high support from both Bayesian posterior probabilities (BPP) and MP
bootstrap proportions (BSP) (Figs. 2,3). Both analyses resulted in three major well-
differentiated peninsular clades corresponding to species. Phyllodactylus lanei rooted as the CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 226 sister group of all ingroup taxa and P. paucituberculatus was sister to a clade containing P. davisi and P. duellmani. Both analyses supported the monophyly of peninsular species from mainland Mexican taxa (BSP = 92; BPP = 1.00). Phyllodactylus unctus rooted the base of the peninsular clade in the MP analysis, resolving P. xanti as sister to a clade containing P. nocticolus from across the peninsula. However, support for the sister relationship of P. xanti and P. nocticolus was relatively weak (BSP <50%). Conversely, BI placed P. nocticolus and
P. unctus as sister taxa with moderate support (0.94). Monophyly of P. nocticolus was strongly supported with both methods (BSP = 99; BPP = 1.00).
3.3 Phylogenetic analysis of nDNA
MP analysis of the two nuclear loci resulted in 42 MPTs of 28 steps (CI = 0.9643; RI =
0.9500; RC = 0.9161). Highly compatible topologies were recovered with both MP and BI
(Fig. 4). Phyllodactylus duellmani and P. davisi were again resolved as sister taxa. However,
unlike the total evidence analysis and mtDNA analysis (presented below), the relationship of
P. paucituberculatus to its congeners was unresolved. The phylogenetic relationships of all
peninsular species were also ambiguous. However, both MP and BI found strong support for
the monophyly of both P. xanti and P. nocticolus.
3.4 Phylogenetic analysis of mtDNA
MP analysis of the mitochondrial genes resulted in 2 MPTs of 3192 steps (CI = 0.6313; RI =
0.6826; RC = 0.4309). The strict consensus mtDNA tree was identical to the total evidence tree with strong support for three monophyletic peninsular clades (Figs. 2,5). As with the CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 227 total evidence topology, MP of mtDNA placed P. xanti and P. nocticolus as sister taxa, although support was relatively low (<50%). Monophyly of the three peninsular species was also strongly supported in the Bayesian topology (BPP = 1), although the phylogenetic relationships of these taxa were ambiguous (P. nocticolus + P. unctus = 0.84). Monophyly of
P. nocticolus and P. xanti was strongly supported.
4 Discussion
4.1 Taxonomy of Phyllodactylus
The phylogenetic analyses presented herein provide further genetic evidence for the validity
of P. nocticolus as a distinct species, separate from P. xanti. The MP analysis suggests that P.
nocticolus is sister of P. xanti, which together form a clade sister to P. unctus. In contrast, the more strongly supported BI analysis resolved P. nocticolus and P. unctus as sister taxa.
These results are of interest because, historically, P. nocticolus was considered a subspecies of P. xanti (Dixon 1964; Grismer 1999). Both P. nocticolus and P. xanti are easily distinguished from P. unctus because the latter lacks enlarged dorsal tubercles.
Estimates of pairwise mtDNA divergence approached 13% between peninsular
species of Phyllodactylus. This is in concordance with previous genetic work based on
allozyme data, which showed significant genetic differences between Cape Region geckos
and those north of La Paz (Murphy 1983b). Although previous researchers documented
morphological variation between P. xanti and P. nocticolus including color pattern, absence
of thigh tubercles in P. nocticolus, and scutellation differences, authors felt that it was not
significant enough to warrant full species status (Dixon 1964; Grismer 1994, 2002). Refer to CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 228
Dixon (1964) for a full list of morphological characters used to differentiate P. nocticolus from P. xanti and other congeners.
The taxonomy and delimitation of species ranges in Phyllodactylus has been ambiguous due to high intraspecific variability in several morphological characters (Dixon
1964). This variability, in conjunction with discordance between molecular and morphological characters, has hindered the elucidation of accurate phylogenetic hypotheses for the genus (Dixon 1964; Murphy 1983b). However, my molecular data, when taken in conjunction with previous allozyme and morphological investigations, provides comprehensive evidence for the validity of P. nocticolus as a distinct species
Phyllodactylus xanti occurs both in the Cape Region of the peninsula as well as on several islands in the Gulf of California (Dixon 1964, 1966; Murphy & Ottley 1984; Grismer
1999, 2002). Historically, the range of P. xanti encompassed most of the peninsula of Baja
California into extreme southern California (Dixon 1964). Populations residing in areas north of La Paz were designated as P. xanti nocticolus, whereas populations south of La Paz were classified as P. xanti xanti (Dixon 1964) with potential intergradation zones around Loreto and Comondu (Dixon 1964; Stebbins 2003). My genetic analysis combined with previous allozyme and morphological comparisons suggests that peninsular populations north of La
Paz be elevated to full species status. Thus, I recognize these populations as P. nocticolus.
Further, the lack of suitable habitat surrounding the La Paz region (Fig. 1) suggests that populations of P. xanti and P. nocticolus are likely not in contact. Additional molecular studies are needed to ascertain the taxonomic status of insular forms north of La Paz.
Phyllodactylus unctus has long been considered a monotypic species, even though insular populations exhibiting slight morphological variation occur on islands associated with CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 229 the Cape Region (Banks & Farmer 1962; Grismer 1999). On the peninsular mainland, the species is restricted to the Cape Region, where it occurs sympatrically and syntopically with
P. xanti (Dixon 1964). The high statistical support for the monophyly of the P. unctus clade is in concordance with previous investigations (Dixon 1964) and suggests that it constitutes a single species.
4.2 Biogeography
Several paleobiogeographic hypotheses have been put forth to explain patterns of
diversification in Phyllodactylus and other Baja Californian taxa (Murphy 1983a,b; Murphy
& Aguirre León 2002). Biogeographic and geological evidence suggests that the peninsula of
Baja California was continually fragmented into smaller landmasses and islands throughout
periods of the Miocene through the Pleistocene (Murphy 1983b; Smith 1991; Aguirre León et al. 1999; Riddle et al. 2000; Carreño & Helenes 2002; Murphy & Aguirre León 2002).
This fragmentation resulted in a trans-peninsular seaway during the late Miocene, which temporarily isolated the Cape Region from the remainder of the peninsula (Lindell et al.
2008). Furthermore, somewhat later during the late Miocene, a temporary mid-peninsular seaway separated northern and southern peninsular biotas (Upton & Murphy 1997; Murphy
& Aguirre León 2002; Lindell et al. 2005, 2006, 2008). My genetic data support the hypothesis of a trans-peninsular seaway that isolated the Cape Region from the rest of the peninsula (Schwennicke et al. 1996; Carreño & Helenes 2002]. This is shown in the high mtDNA sequence divergence estimates between the Cape Region’s endemic species (P. xanti and P. unctus) and more widespread species north of La Paz (P. nocticolus). CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 230
Several empirical studies have documented genetic breaks in maternal lineages that coincide with a seaway across the mid-peninsular region (Upton & Murphy 1997; Riddle et al. 2000; Murphy & Aguirre León 2002; Lindell et al. 2005, 2006, 2008; Lindell & Murphy
2008). In contrast, my genetic data for the widespread P. nocticolus shows no such break
(Figs. 2–5); there was no significant geographic structure throughout the peninsula for this species congruent with a mid-peninsular seaway. However, two reciprocally monophyletic lineages were detected, with a break between San Evaristo and Juncalito (Figs. 2–5). It is possible that this break in maternal lineages corresponds to the intergradation zone described in Dixon (1964).
Few studies have examined intraspecific trans-peninsular variation at nuclear loci.
However, Murphy (1983b) found substantial allozyme evidence for the validity of P. nocticolus north of the Isthmus of La Paz. Conversely, Adest (1987) found no significant genetic differentiation in the zebra tailed lizard, Callisaurus draconoides throughout seven peninsular populations. These data contrast significantly to those utilizing mtDNA as molecular markers. Lindell et al. (2005), for example, showed highly divergent maternal lineages for C. draconoides distributed across the peninsula of Baja California. A similar discordance between mitochondrial and allozyme data is evident in Urosaurus nigricaudus
(Aguirre León et al. 1999; Lindell et al. 2008). This mitochondrial-nuclear discordance clearly illustrates the need to incorporate bi-parentally inherited nuclear loci such as nuclear sequences and/or microsatellite DNA when examining biogeographic patterns of genetic differentiation throughout the peninsula of Baja California. Unfortunately, my nuclear data provided only eight potentially phylogenetically informative characters to the MP analysis. CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 231
The peninsular clade was resolved as monophyletic with high support from both MP and BI (Figs. 2,3). These results suggest a single invasion to the peninsula from mainland
Mexico, presumably resulting from the rifting of western Mexico off the North American
Plate during the increased tectonic activity associated with the Miocene (Murphy 1983b;
Lonsdale 1989; Carreño & Helenes 2002; Murphy & Aguirre León 2002; Lindell et al.
2005). This rifting isolated peninsular populations from their mainland counterparts through the formation and northward expansion of the Gulf of California.
This study is the first to incorporate molecular data to infer evolutionary relationships among species of Phyllodactylus. In my phylogenetic analysis P. lanei rooted the base of the ingroup taxa. Phyllodactylus duellmani was resolved as sister to P. davisi, which together formed a clade sister to P. paucituberculatus (Figs. 2,3). However, only seven of the approximately 20 Mexican species were included in the analysis and until a more comprehensive sampling is undertaken, phylogenetic relationships among Phyllodactylus inhabiting mainland Mexico should remain tentative. Regardless, the inclusion of these taxa supports a single origin for Baja Californian species with high support from both BPP and
BSP in the total evidence and mitochondrial analyses (P. paucituberculatus was unresolved with nDNA). Further, these results contrast to those of Dixon (1964), Grismer (1994), and
Murphy & Papenfuss (1980) who suggested multiple independent origins to the peninsula.
Researchers have postulated alternative vicariant hypotheses to explain the evolution of Baja Californian biota, suggesting that climatic oscillations during the Quaternary contributed significantly to speciation in the flora and fauna of the region (Hafner & Riddle
1997). However, Lindell et al. (2006) suggested that paleogeographic and geologic forces have played a more significant role since the age of the concordant patterns are older than the CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 232
Quaternary climatic fluctuations. Thus, it is highly probable that the biogeographic patterns observed in Phyllodactylus and other taxa distributed throughout the peninsula of Baja
California are primarily due to the geological history of the region and not Quaternary climatic phenomena. CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 233
Acknowledgments
I am grateful to a number of individuals and organizations. Norberto Martínez, Felipe
Rodriguez and Francisco Soto provided assistance with fieldwork. Oliver Haddrath, Ross
MacCulloch, Pat Ross and Tanya Trépanier helped with laboratory work. All permits were
granted to RWM by the Direccíon General de la Fauna Silvestre and Secretaría de Medio
Ambiente y Recursos Naturales, Mexico [65-78-866; 1-78-832; 122.4.3/128-33.26; 12140].
This work was funded by the Natural Sciences and Engineering Research Council (NSERC)
Discovery Grant A3148 to Robert W. Murphy, the ROM Foundation to Robert W. Murphy, the Theodore Roosevelt Foundation of the American Museum of Natural History to Johan
Lindell, the Royal Swedish Academy of Sciences to Johan Lindell, and Sigma Xi’s Grants- in-Aid programme to Johan Lindell. CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 234
References
Adest GA (1987) Genetic differentiation among populations of zebratail lizard, Callisaurus
draconoides (Sauria: Iguanidae). Copeia, 1987, 854–859.
Aguirre-León G, Morafka DJ, Murphy RW (1999) The peninsular archipelago of Baja
California: a thousand kilometers of tree lizard genetics. Herpetologica, 55, 369–381.
Akaike H (1974) A look at the statistical model identification. IEEE Transactions on
Automatic Control AC, 19, 716–723.
Akaike H (1979) A Bayesian extension of the minimum AIC procedure of autoregressive
model fitting. Biometrika, 66, 237–242.
Arevalo E, Davis SK, Sites JW (1994) Mitochondrial DNA sequence divergence and
phylogenetic relationships among eight chromosome races of the Sceloporus grammicus
complex (Phrynosomatidae) in Central Mexico. Systematic Biology, 43, 387–418.
Austin JJ, Arnold EN, Jones CG (2004) Reconstructing an island radiation using ancient and
recent DNA: the extinct and living day geckos (Phelsuma) of the Mascarene Islands.
Molecular Phylogenetics and Evolution, 31, 109–122.
Banks, R.C. & Farmer, W.M. (1962) Observations on reptiles of Cerralvo Island, Baja
California, Mexico. Herpetologica, 18, 246–250.
Brandley MC, Schmitz A, Reeder TW (2005) Partitioned Bayesian analyses, partition choice,
and the phylogenetic relationships of scincid lizards. Systematic Biology, 54, 373–390.
Buckley TR, Arensburger P, Simon C, Chambers GK (2002) Combined data, Bayesian
phylogenetics, and the origin of the New Zealand cicada genera. Systematic Biology, 51,
4–18. CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 235
Carreño AL, Helenes J (2002) Geology and Ages of the Islands. In: Case, T.J., Cody, M.L. &
Excurra, E. (Eds.), A New Island Biogeography of the Sea of Cortés. Oxford University
Press, New York, New York, pp. 14–40.
Dixon JR (1964) The systematics and distribution of lizards of the genus Phyllodactylus in
North and Central America. New Mexico State University Research Center, Scientific
Bulletin, 64–1, 1–139.
Dixon JR (1966) Speciation and systematics of the gekkonid lizard genus Phyllodactylus of
the islands in the Gulf of California. Proceedings of the California Academy of Sciences
series 4, 415–452.
Erixon P, Svennblad B, Britton T, Oxelman B (2003) Reliability of Bayesian posterior
probabilities and bootstrap frequencies in phylogenetics. Systematic Biology, 52, 665–
673.
Felsenstein J (1985) Confidence limits on phylogenies: an approach using the bootstrap.
Evolution, 39, 783–791.
Grismer LL (1994) Geographic origins for the reptiles on islands in the Gulf of California,
Mexico. Herpetological Natural History, 2, 17–40.
Grismer LL (1999) An evolutionary classification of reptiles on islands in the Gulf of
California, Mexico. Herpetologica, 55, 446–469.
Grismer LL (2002) Amphibians and Reptiles of Baja California, Including its Pacific Islands
and the Islands in the Sea of Cortés. University of California Press, Berkeley, California.
Gu X, Fu YX, Li WH (1995) Maximum likelihood estimation of the heterogeneity of
substitution rate among nucleotide sites. Molecular Biology and Evolution, 12, 546–57. CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 236
Hafner DJ, Riddle BR (1997) Biogeography of Baja California peninsular desert mammals.
In: Yates, T.L., Gannon, W.L. & Wilson, D.E. (Eds.), Life Among the Muses: Papers in
Honor of James S. Findley. University of New Mexico, Albuquerque, New Mexico, pp.
39–68.
Hall TA (1999) BIOEDIT: a user-friendly biological sequence alignment editor and analysis
program for Windows 95/98/NT. Nucleic Acids Symposium Series, 41, 95–98.
Hausback BP (1984) Cenozoic volcanic and tectonic evolution of Baja California Sur,
Mexico. In: Frizzell, V.A., Jr. (Ed.), Geology of the Baja California Peninsula. Pacific
Section, Society of Economic Paleontologists and Mineralogists, Los Angeles,
California, pp. 219–236
Hillis DM, Mable BK, Larson A, Davis SK, Zimmer EA (1996) Nucleic acids IV: In: Hillis,
D.M., Moritz, C. & Mable, B.K. (Eds.), Molecular Systematics. Sinauer, Sunderland,
Massachusetts, pp. 321–383.
Huelsenbeck JP, Bollback JP (2001) Empirical and hierarchical Bayesian estimation of
ancestral states. Systematic Biology, 50, 351–366.
Huelsenbeck JP, Ronquist F (2001) MRBAYES: Bayesian inference of phylogeny.
Bioinformatics, 17, 754–755.
Kocher TD, Thomas WK, Meyer A, et al. (1989) Dynamics of mitochondrial DNA evolution
in animals: Amplification and sequencing with conserved primers. Proceedings of the
National Academy of Sciences USA, 86, 6196–6200.
Leaché AD, McGuire JA (2006) Phylogenetic relationships of horned lizards (Phrynosoma)
based on nuclear and mitochondrial data: Evidence for a misleading mitochondrial gene
tree. Molecular Phylogenetics and Evolution, 39, 628–644. CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 237
Lindell J, Méndez-de la Cruz FR, Murphy RW (2005) Deep genealogical history without
population differentiation: discordance betIen mtDNA and allozyme divergence in the
zebra-tailed lizard (Callisaurus draconoides). Molecular Phylogenetics and Evolution,
36, 682–694.
Lindell J, Ngo A, Murphy RW (2006) Deep genealogies and the mid-peninsular seaway of
Baja California. Journal of Biogeography, 33, 1327–1331.
Lindell J, Méndez-de la Cruz FR, Murphy RW (2008) Discordance between deep
genealogical history and population differentiation in the black-tailed brush lizard
(Urosaurus nigricaudus) of Baja California. Biological Journal of the Linnean Society,
94, 89–104.
Lindell J, Murphy RW (2008) Simple identification of mitochondrial lineages in contact
zones based on lineage-selective primers. Molecular Ecology Resources, 8, 66–73.
Lonsdale P (1989) Geology and tectonic history of the Gulf of California. In: Winterer, E.L.,
Hussong, D.M. & Decker, R.W. (Eds.), The Geology of North America, Vol. N, The
Eastern Pacific Ocean and Hawaii. Geological Society of America, Boulder, Colorado,
pp. 499–521.
Maddison DR, Maddison WP (2005). MACCLADE, v4.08 OSX: Analysis of Phylogeny and
Character Evolution. Sinauer Associates, Sunderland, Massachusetts.
Murphy RW (1983a) The reptiles: origin and evolution. In: Case, T.J. & Cody, M.L. (Eds.),
Island Biogeography in the Sea of Cortez. University of California Press, Berkeley,
California, pp. 130–158.
Murphy RW (1983b) Paleobiogeography and genetic differentiation of the Baja California
herpetofauna. Occasional Papers of the California Academy of Sciences, 137, 1–48. CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 238
Murphy RW, Papenfuss TJ (1980) Biochemical variation of Phyllodactylus unctus and P.
paucituberculatus. Biochemical Systematics and Ecology, 8, 97–100.
Murphy RW, Aguirre-León G (2002) The nonavian reptiles: origins and evolution. In: Case,
T.J., Cody, M.L. & Ezcurra, E. (Eds.), A New Island Biogeography of the Sea of Cortés.
Oxford University Press, New York, New York, pp. 181–220.
Murphy RW, Ottley JR (1984) Distribution of amphibians and reptiles on islands in the Gulf
of California. Annals of Carnegie Museum, 53, 207–230.
Nylander JAA, Ronquist F, Huelsenbeck JP, Nievees-Aldrey JL (2004) Bayesian
phylogenetic analysis of combined data. Systematic Biology, 53, 47–67.
Palumbi SR (1996) Nucleic acids II: the polymerase chain reaction. In: Hillis, D.M., Moritz,
C. & Mable, B.K. (Eds.), Molecular Systematics. Sinauer, Sunderland, Massachusetts,
pp. 205–247.
Radtkey RR, Fallon SK, Case TJ (1997) Character displacement in some Cnemidophorus
lizards revisited: a phylogenetic analysis. Proceedings of the National Academy of
Sciences USA, 94, 9740–9745.
Riddle BR, Hafner DJ, Alexander LF, Jaeger JR (2000) Cryptic vicariance in the historical
assembly of a Baja California peninsular desert biota. Proceedings of the National
Academy of Sciences USA, 97, 14438–14443.
Ronquist F, Huelsenbeck JP (2003) MRBAYES 3: Bayesian phylogenetic inference under
mixed models. Bioinformatics, 19, 1572-1574.
Saiki RK, Gelfand DH, Stoeffel S (1988) Primer-directed enzymatic amplification of DNA
with a thermostable DNA polymerase. Science, 239, 487–491. CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 239
Saint KM, Austin CC, Donnellan SC, Hutchinson MN (1998) C-mos, A nuclear marker
useful for squamate phylogenetic analysis. Molecular Phylogenetics and Evolution, 10,
259–263.
Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory Manual,
second edition. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.
Savage JM (1960) Evolution of a peninsular herpetofauna. Systematic Zoology, 9, 184–212.
SchInnicke T, González-Barba G, DeAnda-Franco N (1996) LoIr Miocene marine and
fluvial beds at Rancho La Palma, Baja California Sur. Boletín del Departamento
Geológico de la Universidad de Sonora, 13, 1–14.
Smith JT (1991) Cenozoic marine mollusks and paleogeography of the Gulf of California. In:
Dauphin, J.P. & Simoneit, B.R.T. (Eds.), The Gulf and Peninsular Province of the
Californias. American Association of Petroleum Geologists Memoir, 47, Tulsa,
Oklahoma, pp. 637–666.
Stebbins RC (2003) A Field Guide to Istern Reptiles and Amphibians. Houghton Mifflin
Company, New York, New York.
Swofford DL (2002) PAUP*: Phylogenetic Analysis Using Parsimony (*and other methods).
Sinauer Associates, Sunderland, Massachusetts.
Upton DE, Murphy RW (1997) Phylogeny of the side-blotched lizards (Phrynosomatidae:
Uta) based on mtDNA sequences: support for a mid-peninsular seaway in Baja
California. Molecular Phylogenetics and Evolution, 8, 104–113.
Waddell P, Penny D (1996) Evolutionary trees of apes and humans from DNA sequences. In:
Lock, A.J. & Peters, C.R. (Eds.), C.R. Handbook of Symbolic Evolution. Oxford,
Clarendon, pp. 53–73. CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 240
Table 1. Taxon name, tissue number, sample ID, geographic location, GPS coordinates, and Genbank accession numbers for all species of Phyllodactylus sequenced. ROM = Royal Ontario Museum; RWM = Collections of Robert W.
Murphy housed at the ROM; N/A = No data available.
Genbank Accession
Numbers
Sam
Tissue ple BDN
Taxon no. ID Location Latitude Longitude 12S 16S COI ND4 F C-mos
RWM
P. nocticolus 1166 L8 Mexico; Baja California; San Felipe (near) 30 41' 33" N 114 57' 23" W
RWM
P. nocticolus 1879 L103 Mexico; Baja California; Santa Ines 29 43' 45" N 114 41' 47" W
RWM
P. nocticolus 2164 L5 Mexico; Baja California; Bahia de Los Angeles, 8 km W of 28 58' 44" N 113 38' 13" W
RWM
P. nocticolus 1275 L14 Mexico; Baja California Sur; San Ignacio 27 17' 55" N 112 53' 33" W
RWM
P. nocticolus 2297 L6 Mexico; Baja California Sur; Arroyo Santa Agueda 27 15' 29" N 112 20' 57" W
RWM
P. nocticolus 2347 L4 Mexico; Baja California Sur; Playa Coyote, 4 km S of 26 41' 40" N 111 54' 43" W
RWM
P. nocticolus 1281 L15 Mexico; Baja California Sur; Loreto, 42 km N of 26 23' 20" N 111 27' 48" W
RWM
P. nocticolus 1467 L33 Mexico; Baja California Sur; Juncalito, 1.5 km W of 25 50' 22" N 111 21' 01" W
ROM
P. nocticolus 35407 L54 Mexico; Baja California Sur; 6 km W of San Evaristo 24 52' 23" N 110 44' 27" W
RWM
P. nocticolus 920 L7 Mexico; Baja California Sur; San Pedro la Presa 24 50' 52" N 110 59' 32" W
RWM
P. nocticolus 754 L34 Mexico; Baja California Sur; San Pedro la Presa 24 50' 52" N 110 59' 32" W CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 241
RWM
P. nocticolus 34055 L28 Mexico; Baja California Sur; La Paz, 40 km N of 24 17' 45" N 110 39' 45" W
RWM
P. nocticolus 34062 L29 Mexico; Baja California Sur; La Paz, 80 km NW of 24 40' 07" N 110 47' 57" W
RWM
P. xanti 2307 L17 Mexico; Baja California Sur; San Bartolo 23 44' 16" N 109 51' 20" W
ROM
P. xanti 34047 L27 Mexico; Baja California Sur; San Bartolo 23 44' 16" N 109 51' 20" W
ROM
P. xanti 34065 L30 Mexico; Baja California Sur; San Bartolo 23 44' 16" N 109 51' 20" W
ROM
P. unctus 34045 L26 Mexico; Baja California Sur; La Paz, 30 km S of 23 51' 02" N 110 18' 32" W
RWM
P. unctus 633 L11 Mexico; Baja California Sur; Hwy 1, 11 km S of Todos Santos turnoff 23 21' 20" N 110 10' 12" W
RWM
P. unctus 697 L12 Mexico; Baja California Sur; Hwy 13, 2 miles SE of junction with Hwy 1 24 06’ 20” N 110 17’ 20” W
ROM
P. paucituberculatus 35455 L80 Mexico; Michoacan; Los Olivos vicinity; where Hwy 37 crosses Rio Tepalcatepec 18 50' 56" N 102 08' 05" W
ROM
P. davisi 46935 L97 Mexico; Colima; 2 km NE Santiago 19 08' 12" N 104 20' 18" W
ROM
P. duellmani 35451 L61 Mexico; Michoacan; Los Olivos vicinity; where Hwy 37 crosses Rio Tepalcatepec 18 50' 56" N 102 08' 05" W
ROM
P. lanei 35287 L42 Mexico; Guerrero; Tierra Colorada 17 09' 29" W 099 31" 53" W
CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 242
Table 2. Percent sequence divergence (uncorrected p-distances from mtDNA) for Baja Californian and mainland Mexican Phyllodactylus compared to Gekko gecko. N/A = undefined distance between sequences.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
1 Gekko gecko -
2 P. duellmani (L61) 28.80 -
3 P. davisi (L97) 28.60 11.30 -
4 P. paucituberculatus (L80) 29.00 12.70 13.00 -
5 P. lanei (L42) 28.30 14.50 14.40 14.20 -
6 P. unctus (L11) 27.90 11.90 12.10 11.80 13.90 -
7 P. unctus (L26) 28.00 11.80 12.10 11.70 13.80 0.90 -
8 P. unctus (L12) 29.70 14.50 14.90 15.60 18.00 1.00 1.00 -
9 P. xanti (L17) 28.10 12.00 12.20 12.10 13.70 9.70 9.60 11.70 -
10 P. xanti (L30) 28.00 11.90 12.50 12.20 13.90 9.90 9.80 12.10 1.50 -
11 P. xanti (L27) 28.30 12.00 12.40 12.10 13.80 9.70 9.70 11.90 0.56 1.10 -
12 P. nocticolus (L34) 28.00 11.90 12.80 12.60 14.00 9.80 10.00 12.20 10.10 9.90 10.10 -
13 P. nocticolus (L103) 27.80 15.80 15.50 14.00 16.20 11.30 11.60 N/A 12.10 12.10 12.00 3.90 -
14 P. nocticolus (L28) 28.80 12.80 12.90 12.20 14.10 10.10 10.10 12.80 10.20 10.30 10.10 2.10 2.70 -
15 P. nocticolus (L29) 28.80 12.80 13.00 12.20 14.20 10.10 10.10 12.60 10.20 10.20 10.10 2.00 2.80 0.25 -
16 P. nocticolus (L54) 28.50 12.60 13.00 12.40 14.10 9.80 10.00 12.50 9.80 9.80 9.90 0.20 3.10 2.00 2.00 -
17 P. nocticolus (L7) 28.40 12.60 13.00 12.30 14.10 9.80 10.00 12.40 9.70 9.70 9.70 0.34 3.10 2.00 2.00 0.35 -
18 P. nocticolus (L5) 28.20 13.10 13.00 12.40 14.10 10.20 10.40 12.70 10.00 10.20 9.90 2.60 2.80 2.80 2.80 2.60 2.70 -
19 P. nocticolus (L6) 28.90 13.00 13.30 12.50 13.90 9.90 10.10 12.80 9.80 10.00 9.70 3.10 2.70 3.10 3.10 3.10 3.20 2.50 -
20 P. nocticolus (L33) 28.70 13.00 13.20 12.60 14.40 10.30 10.30 12.60 10.30 10.20 10.10 3.20 2.90 2.90 2.90 3.40 3.30 2.70 2.60 -
21 P. nocticolus (L15) 28.60 13.00 13.00 12.60 14.10 10.30 10.40 13.00 10.00 10.00 9.90 3.20 2.80 3.10 3.10 3.30 3.30 2.60 2.40 1.00 -
22 P. nocticolus (L4) 28.80 13.00 13.20 12.50 14.20 10.10 10.20 12.50 9.90 9.90 9.90 3.20 2.90 3.00 3.00 3.20 3.20 2.70 2.40 1.10 1.10 -
23 P. nocticolus (L8) 28.60 13.10 12.90 12.40 14.10 9.90 10.10 12.60 10.00 10.20 10.10 2.90 0.87 2.80 2.80 2.90 2.90 2.30 2.50 2.80 2.70 2.80 -
24 P. nocticolus (L14) 28.00 13.00 13.10 12.30 14.00 9.80 9.90 12.60 9.90 10.10 9.90 2.80 2.40 3.00 3.00 2.90 2.90 2.50 2.70 3.00 2.60 2.80 2.30 CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 243
Figure Legends
Fig. 1. Map of Baja California, the Gulf of California, southwestern USA, and northwestern
Mexico illustrating the geographic location of samples of Phyllodactylus included in this
study. Sample numbers correspond to those presented in Table 1. The break in geographic
ranges in the Isthmus of La Paz region is due to unsuitable natural habitat (lack of rocky
outcrops). Gray area represents the geographic range of P. nocticolus, black area
represents the geographic range of P. xanti, striped area represents the range of P. unctus.
Fig. 2. Phylogenetic relationships of Mexican Phyllodactylus based on a branch-and-bound
maximum parsimony (MP) analysis of the concatenated dataset. Numbers above nodes
represent MP bootstrap proportions (BSP) resulting from 10,000 pseudoreplicates.
Fig. 3. Phylogenetic relationships of Mexican Phyllodactylus based on Bayesian inference
(BI) of the concatenated dataset. Numbers above and adjacent to nodes represent Bayesian
posterior probabilities (BPP; * = 1.0) sampled from the posterior distribution of trees.
Gekko gecko was removed from the tree due to its long branch length.
CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 244
Fig. 4. Phylogenetic relationships of Mexican Phyllodactylus based on maximum parsimony
(MP; A) and Bayesian Inference (BI; B) of nuclear DNA (nDNA). Numbers above nodes
on MP tree represent nonparametric bootstrap proportions (BSP). Numbers above nodes
on the Bayesian topology represent Bayesian posterior probabilities (BPP; * = 1.0). Refer
to text for a description of loci and total base pairs (bp) sequenced.
Fig. 5. Phylogenetic relationships of Mexican Phyllodactylus based on maximum parsimony
(MP; A) and Bayesian Inference (BI; B) of mitochondrial DNA (mtDNA). Numbers
above nodes on MP tree represent nonparametric bootstrap proportions (BSP). Numbers
above nodes on the Bayesian topology represent Bayesian posterior probabilities (BPP; *
= 1.0). Gekko gecko was removed from the BI tree due to its long branch length. Refer to
text for a description of loci and total base pairs (bp) sequenced. CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 245
Fig. 1 CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 246
Fig. 2 CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 247
Fig. 3 CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 248
Fig. 4 CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 249
Fig. 5 CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 250
Supplementary Material
Appendix I. Mitochondrial and nuclear loci sequenced for Phyllodactylus and their corresponding primer sequences. Different combinations of primers Ire used to obtain full reads of regions.
Genome Gene Primer Primer Sequence Source
Mitochondrial 12S rRNA 12sa 5'—AAA CTG GGA TTA GAT ACC CCA CTA T—3' Palumbi 1996
3395H 5'—AGT CTT CCC GTC CTT TTG CCA C—3' This study
504L 5'—GAT GAG GCA AGT CGT AAC ATG G—3' This study
12s2L 5'—ACA CAC CGC CCG TCA CCC TC—3' Kocher et al. 1989
12s2H 5'—AGG GTG ACG GGC GGT GT GT—3' Kocher et al. 1989
12S1L 5'—CAA ACT GGG ATT AGA TAC CCC ACT AT—3' This study
16S rRNA 1320L 5'—GGG GAA CCA CTG TTA AAA CTA G—3' This study
1715H 5'—TAA AGC TCC ACA GGG TCT TCT CGT C—3' This study
16s5H 5'—CTA CCT TTG CAC GGT TAG GAT ACC GCG GC—3' This study
3985L 5'—TTA CCA AAA ACA TGG CCT TTA GC—3' This study
16s1Lm 5'—CCG ACT GTT GAC CAA AAA CAT—3' This study*
16s3Lm 5'—CCC GAA CCC AGT CGA GCT AC—3' This study*
3456L 5'—GGT GAA ACG CCA AYC GCG CCT GG—3' This study
950L 5'—CCC TGT GGC AAA AGG GYG GG—3' This study
760L 5'—GTA CCA TAT AGG AAC ACT GAA AG—3' This study
990H 5'—CAT TTG ATG AGC AAC CAG CTA TCR C—3' This study
690H 5'—GTT CTT TCT CCA ATA CTG C—3' This study CHAPTER 6—PHYLOGENETICS OF BAJA CALIFORNIAN PHYLLODACTYLUS 251
16S2HM 5'—CCG GTC TGA ACT CAG ATC ACG—3' This study*
COI COIphyR 5'—CTA TTG ATA GGA CGT AGT GG—3' This study
COIphyL 5'—TCC TTC CCG GVT TYG GMA T—3' This study
COIphyL1 5'—CTC ATC CTC CCS GGR TTT GG—3' This study
COIm72L 5'—TGA TTC TTC GGT CAC CCA GAA GTG TA—3' Austin et al. 2004
COIm73H 5'—CCT ATT GAT AGG ACG TAG TGG AAG TG—3' Austin et al. 2004
ND4 Nd4phyL 5'—CCT ATC CCC CAC AAC CCA AAC—3' This study
Nd4phyH 5'—TTA GTG TTT TGG TTA AAY TAT—3' This study
ND412931L 5'—CTA CCA AAA GCT CAT GTA GAA GC—3' This study
ND413824H 5'—CAT TAC TTT TAC TTG GAT TTG CAC CA—3' Arevalo et al. 1994
Nuclear BDNF BDNFf 5'—GAC CAT CCT TTT CCT KAC TAT GGT TAT TTC ATA CTT—3' Leache & McGuire 2006
BDNFr 5'—CTA TCT TCC CCT TTT AAT GGT CAG TGT ACA AAC—3' Leache & McGuire 2006
CMOS CmosG73L 5'—GCG GTA AAG CAG GTG AAG AAA—3' Saint et al. 1998
CmosG74H 5'—TGA GCA TCC AAA GTC TCC AAT C—3' Saint et al. 1998
* Modified after Palumbi, 1996
CHAPTER 7 Conclusions and future directions
1 General Background
The continual reduction in DNA sequencing costs is revolutionizing molecular ecology and evolution and enabling us to ask more sophisticated questions regarding spatial and temporal patterns of biological diversification in non-model taxa. Accompanying these advances in
DNA sequencing technologies are new, powerful multilocus statistical methods that are able to provide a more comprehensive analysis of the processes influencing the evolutionary history of populations, species, and higher-level taxa (Blair & Murphy 2011). Conversely, the nascent field of landscape genetics seeks to determine the relative importance of contemporary ecological variables in shaping microevolutionary processes (Manel et al.
2003; Holderegger & Wagner 2006, 2008). Landscape genetic analysis can be seen as a complement to phylogenetic and phylogeographic inference by providing a temporal component regarding drivers of contemporary genetic structure (Manel et al. 2003; Storfer et al. 2010; Wang 2010). In this thesis, I utilize multiple types of molecular markers and statistical analyses to test hypothesis of interest to both molecular phylogenetics and landscape genetics. In the following sections I extrapolate on the main findings of the thesis, provide a historical background about why this research is both timely and necessary, and discuss future research directions.
2 Mexican Tropical Dry Forest
252 CHAPTER 7—CONCLUSIONS 253
Over the past few decades, tropical rain forests have garnered a substantial amount of attention from molecular ecologists and evolutionary biologists. This is due, in part, to the fact that these ecosystems contain the vast majority of Earth’s species and are disappearing at an alarming rate (Myers et al. 2000). Many studies seek to determine the relative importance of Neogene vicariance versus Quaternary climate change as engines for speciation and genetic diversification in taxa distributed thought rain forest habitat (Klicka & Zink 1997;
Moritz et al. 2000; Weir 2006). Conversely, tropical dry forests (TDFs) have received little attention from both an evolutionary and conservation perspective (Trejo & Dirzo 2000).
Although they are both tropical forests, a variety of structural and functional differences separate these habitat-types. For example, TDFs are seasonal with up to eight months of drought per year. The vegetation generally consists of short deciduous or semi-deciduous trees (~8–12 m in height) and thorny vegetation along with columnar cacti (Murphy & Lugo
1986). A primary distinction between TDF and tropical thornscrub lies with the height of the canopy in relation to cactus-height (i.e. canopy-height is higher than cactus-height in TDF;
Robichaux & Yetman 2000). Further, unlike rain forests, the soil in TDFs is generally nutrient rich with a moderate to high pH (Murphy & Lugo 1986).
Although the biological diversity in TDF is substantial (Ceballos 1995; Ceballos &
García 1995; Gillespie et al. 2000; García 2006), both natural and anthropogenic disturbance threaten the evolutionary potential for species inhabiting this ecosystem (Janzen 1988; Miles et al. 2006). At a global level, Neotropical dry forests suffered from a deforestation rate of
12% between 1980 and 2000 (Miles et al. 2006). Mexican TDF occurs in a semi-continuous belt on the west coast from southern Sonora to the Guatemalan border. It is hypothesized that the Mexican TDF originated between 30–20 million years ago in response to the orogenesis CHAPTER 7—CONCLUSIONS 254 of the Sierra Madre Occidental and Neovolcanic Axis (Becerra 2005). Trejo & Dirzo (2000) perform a time-series analysis and find that as of 1990 only 27% of Mexico’s TDF had remained intact. They estimate an annual deforestation rate of 1.4%. The majority of fragmentation is due to anthropogenic influences including slash-and-burn practices and the conversion of forest to pasturelands and agriculture fields (Burgos & Maass 2004).
Although we are now beginning to realize the biological importance of these tropical forests, few molecular genetic studies have focused explicitly on testing diversification hypotheses for species inhabiting this ecosystem. For example, most molecular phylogenetic and phylogeographic studies in Mexico focus on evolutionary processes throughout the peninsula of Baja California (e.g. Lindell et al. 2005, 2008; Blair et al. 2009), the Sonoran
Desert (e.g. Hafner & Riddle 2005), and the Mexican Highlands (e.g. Bryson et al.
2011a,b,c). The few studies focusing on genetic patterns throughout the TDF of western
Mexico find evidence for cryptic lineages, many of which correspond to previously hypothesized biogeographic barriers (e.g. Hasbún et al. 2005; Devitt 2006; Zarza et al.
2008). At present, there is a lack of landscape genetic studies focused on tropical systems
(Storfer et al. 2010) and to my knowledge there has been no published study that examines functional connectivity in species inhabiting TDF vegetation.
3 Leaf-toed Geckos
Leaf-toed geckos, genus Phyllodactylus, represent approximately 50 species that encompass a Neotropical distribution ranging from southern California to southwestern South America and the West Indies (Dixon 1964). Species are commonly found in arid and semi-arid environments throughout their ranges. In Mexico, the majority of species are confined to the CHAPTER 7—CONCLUSIONS 255
TDFs of the west coast and southern Baja California Sur. The genus is characterized be leaf- like extensions at the terminal ends of the digits. To date, little molecular work has been done with the genus with most phylogenetic studies utilizing a few representative species in order to understand higher-level gekkonid relationships (e.g. Gamble et al. 2008, 2011). Dixon
(1964) provides a comprehensive morphological investigation into the systematics of the genus throughout North and Middle America and suggests that diversity within the genus may be substantially underestimated. The fact that these lizards are primarily restricted to
TDF habitat makes them ideal to test both historical and contemporary ecological and evolutionary hypotheses operating throughout the ecosystem.
4 General Results of Thesis
By combining different molecular marker-types useful for testing ecological and
evolutionary hypotheses over multiple spatial and temporal scales (Wang 2010), I elucidate
the historical and contemporary processes influencing the spatial distribution of genetic
variation in leaf-toed geckos. In Chapter 2, I use a combination of molecular cloning and
next-generation 454 pyrosequencing to isolate and characterize 12 microsatellite loci for the
Mexican yellow-bellied gecko, P. tuberculosus. In Chapter 3, I use these loci to conduct a
landscape genetic study of P. tuberculosus throughout the landscape surrounding Alamos,
Sonora. Results suggest highly differentiated populations and that dispersal is influenced by variables including forest structure, stream networks, slope, temperature, and the degree of human perturbation.
Chapter 4 utilizes data from two mitochondrial genes and two intron markers in order
to understand the historical processes influencing rates and patterns of diversification of CHAPTER 7—CONCLUSIONS 256 geckos in the P. tuberculosus species group (P. t. saxatilis, P. t. magnus, P. muralis muralis,
P. m. isthmus, P. lanei lanei, P. l. rupinus). My results suggest that both the origin of
Mexican TDF and Pleistocene climate change were important drivers of cladogenesis for these geckos, although demographic expansion did not occur during or following the
Pleistocene. I also find low effective population sizes and migration rates between contemporary mtDNA lineages. Further, unlike my landscape genetic analysis, I find no significant correlation between riverine networks and genetic differentiation. These results are concordant with a recent analysis of the wood frog (Lithobates sylvaticus) and suggest that processes operating over a landscape genetic scale may not translate to processes over a phylogeographic scale (Lee-Yaw et al. 2009). The discrepancy in my results may be partly explained by inherent differences in molecular marker-types (mtDNA versus microsatellites), a result of specific ecological characteristics at field sites, or simply the fact that other ecological processes are more important over relatively broad spatial scales. For example, most of the streams and arroyos sampled for the landscape genetic study had extensive adjacent rocky outcroppings. Conversely, many of the larger rivers included in the phylogeographic analysis of geckos lacked extensive rocky habitat. As geckos are commonly encountered on rocky vertical substrate, the discordance may be partly explained by these microhabitat differences. Additional studies are required to test the ability of landscape genetic inference to predict similar phylogeographic patterns and processes.
In Chapter 5, I use the microsatellite data generated for Chapter 3 along with de novo mtDNA sequence data to compare and contrast historical and contemporary demographic parameters in geckos surrounding the region surrounding Alamos, Sonora. I find evidence for substantial cryptic lineage divergence over a small spatial scale (~40 km2), a 10-fold increase CHAPTER 7—CONCLUSIONS 257 in contemporary versus historical migration rates, and evidence for recent bottlenecks and smaller contemporary effective population sizes for many populations. Because the utility of mtDNA for testing landscape genetic hypotheses is still hotly debated (e.g. Vandergast et al.
2007; Wang et al. 2010,2011; Bohonak & Vandergast 2011), I created resistance surfaces based on variables previously shown to be important predictors of connectivity for the species (Chapter 3). I then use circuit theory to correlate different measures of landscape connectivity with mtDNA divergence, historical migration rate, and contemporary migration rate. My results show that landscape-genetic relationships based on mtDNA data contrast with those based on microsatellites. Higher correlations are also found between landscape structure and contemporary versus historical migration rate. Taken together, this suggests that although migration rates have increased since humans began modifying the landscape, geckos are altering their dispersal routes in response to recent habitat fragmentation. Thus, contemporary genetic structure is likely a product of low historical migration and strong female philopatry along with recent anthropogenic fragmentation to the landscape.
In Chapter 6, I examine the biogeographic and taxonomic consequences of the dynamic paleogeographic history of the peninsula of Baja California. Previous molecular genetic studies suggest that the peninsula was fragmented by trans-peninsular seaways at multiple locations throughout the Miocene–Pleistocene (e.g. Upton & Murphy 1997; Lindell et al. 2005, 2008). Evidence suggests that these seaways resulted in a vicariant event that separated populations on either side. Traditionally, the peninsula was home to two species of leaf-toed gecko, P. unctus and P. xanti, the latter being composed of multiple subspecies both on the peninsula and on islands within the Gulf of California (Dixon 1964). Phyllodactylus unctus and P. x. xanti are restricted to the TDFs of the southern Cape Region of the peninsula CHAPTER 7—CONCLUSIONS 258 whereas P. x. nocticolus has a more desert-like distribution ranging from just north of the La
Paz region all the way to extreme southern California. Using data from five mitochondrial and two nuclear genes, I find substantial genetic divergence both inter- and intraspecifically.
This provides further evidence that P. x. nocticolus deserves species-status and, thus, I elevate the taxon to be a species. Phylogenetic analyses also provide further evidence for a trans-peninsular seaway near the Isthmus of La Paz region that led to the separation and speciation of P. nocticolus from its Cape Region ancestor. Unfortunately, my analysis is unable to determine the sister group to P. nocticolus (due to low nodal support). Further, unlike other studies (e.g. Upton & Murphy 1997; Lindell et al. 2005, 2008) I find no evidence for a mid-peninsular seaway during the Miocene. However, my analysis recovers two reciprocally monophyletic lineages of P. nocticolus that may correspond to the previously hypothesized break near the Loreto region (Lindell et al. 2008). Unlike the biogeographic breaks commonly observed at the mid-peninsular and La Paz areas, few studies find deep genealogical breaks near Loreto (Lindell et al. 2005, 2008). However, the presence of marine sediments of Pliocene origin suggests a trans-peninsular seaway between
3.4–2.0 Ma (McLean 1989; Umhoefer et al. 1994). Additional data are required to test the generality of a trans-peninsular seaway near Loreto.
5 Future Directions
Although challenging, writing a multidisciplinary Ph.D. thesis does have its benefits. The
effort has taught me to think critically and always ask questions regarding methods of
analysis, philosophy, and assumptions. I am interested in applying the skills I have learned
during my Ph.D. to future projects in the diverse fields of molecular phylogenomics, CHAPTER 7—CONCLUSIONS 259 landscape genetics, and ecological genomics. Methods of phylogenetic inference have progressed rapidly over the past few years to handle the vast quantities of multilocus sequence data being generated (Edwards 2009; Blair & Murphy 2011) and I am interested in adopting a phylogenomic approach to better understand the evolutionary history of different vertebrate lineages. Applying these methods in a hypothesis-testing framework can illuminate the historical processes responsible for shaping Earth’s rich biota.
On a finer spatial and temporal scale, I am interested in continuing my work in landscape and conservation genetics. The field of landscape genetics is progressing rapidly since its inception almost 10 years ago (Manel et al. 2003; Storfer et al. 2010). I plan on pursuing additional empirical studies with small vertebrate species inhabiting fragmented tropical habitats to better understand how recent disturbance may be influencing functional connectivity. I am presently working on collaborations with Brazilian researchers to understand the genetic consequences of fragmentation throughout the Brazilian Atlantic forest. Simulation studies also continue to play a vital role in the development of the field of landscape genetics and I hope to continue my research in this area (Landguth et al. 2010;
Safner et al. 2011; Blair et al. 2012).
The continual reduction in costs for next-generation DNA sequencing is making ecological genomic studies in non-model taxa now within reach for many researchers. Many of these studies use genome-scans (microsatellites, AFLPs, or SNPs) to determine if and how populations may be adapting to specific environmental conditions (Holderegger et al. 2006;
Joost et al. 2007; Manel et al. 2010). Early methods utilized outlier approaches to test for signs of loci exhibiting differentiation higher than expected under neutrality, which may suggest that certain markers are linked to loci under selection (Luikart et al. 2003). Newer CHAPTER 7—CONCLUSIONS 260 methods continue to be developed that take the analysis one step further and statistically correlate specific loci or alleles with certain environmental features (e.g. Joost et al. 2007); a so-called ‘landscape genomics approach.’ I am interested in conducting a variety of landscape genomic studies in small vertebrates throughout the Mexican TDF to test for selection and adaptation at the molecular level. This will enable me to determine how species are responding to continual habitat change from both natural and anthropogenic sources. CHAPTER 7—CONCLUSIONS 261
References
Becerra JX (2005) Timing the origin and expansion of the Mexican dry forest. Proceedings
of the National Academy of Sciences USA, 102, 10919–10023.
Blair C, Murphy RW (2011) Recent trends in molecular phylogenetic analysis: where to
next? Journal of Heredity, 102, 130–138.
Blair C, Mendez de la Cruz FR, Ngo A, Lindell J, Lathrop A, Murphy RW (2009) Molecular
phylogenetics and taxonomy of leaf-toed geckos (Phyllodactylidae: Phyllodactylus)
inhabiting the peninsula of Baja California. Zootaxa, 2027, 28–42.
Blair C, Weigel DE, Balazik M, Keeley ATH, Walker FM, Landguth E, Cushman S, Murphy
M, Waits L, Balkenhol N (2012) A simulation-based evaluation of methods for inferring
linear barriers to gene flow. Molecular Ecology Resources, in press.
Bohonak AJ, Vandergast AG (2011) The value of DNA sequence data for studying landscape
genetics. Molecular Ecology, 20, 2477–2479.
Bryson RW, Murphy RW, Lathrop A, Lazcano-Villareal D (2011a) Evolutionary drivers of
phylogeographical diversity in the highlands of México: a case study of the Crotalus
triseriatus species group of montane rattlesnakes. Journal of Biogeography, 38, 697–710.
Bryson RW, García-Vázquez UO, Riddle BR (2011b) Phylogeography of Middle American
gophersnakes: mixed responses to biogeographical barriers across the Mexican Transition
Zone. Journal of Biogeography, 38, 1570–1584.
Bryson RW, García-Vázquez UO, Riddle BR (2011c) Diversification in the Mexican horned
lizard Phrynosoma orbiculare across a dynamic landscape. Molecular Phylogenetics and
Evolution (in press). CHAPTER 7—CONCLUSIONS 262
Burgos A, Maass MJ (2004) Vegetation change associated with land-use in tropical dry
forest areas of western México. Agriculture, Ecosystems, & Environment, 104, 475–481.
Ceballos G (1995) Vertebrate diversity, ecology, and conservation in neotropical dry forests.
In: Bullock SH, Mooney HA, Medina E, eds), Seasonally Dry Tropical Forests.
Cambridge University Press, Cambridge, pp 195–220.
Ceballos G, García A (1995) Conserving neotropical biodiversity: the role of dry forests in
western México. Conservation Biology, 9, 1349–1356.
Devitt TJ (2006) Phylogeography of the western lyresnake (Trimorphodon biscutatus):
testing aridland biogeographical hypotheses across the Nearctic-Neotropical transition.
Molecular Ecology, 15, 4387–4407.
Dixon JR (1964) The systematics and distribution of lizards of the genus Phyllodactylus in
North and Central America. New México State University Research Center, Scientific
Bulletin, 64-1, 1–139.
Edwards SV (2009) Is a new and general theory of molecular systematics emerging?
Evolution, 63, 1–19.
Gamble T, Bauer AM, Greenbaum E, Jackman TR (2008) Evidence for Gondwanan
vicariance in an ancient clade of gecko lizards. Journal of Biogeography, 35, 88–104.
Gamble T, Bauer AM, Colli GR, Greenbaum E, Jackman TR (2011) Coming to America:
multiple origins of New World geckos. Journal of Evolutionary Biology, 24, 231–244.
García A (2006) Using ecological niche modelling to identify diversity hotspots for the
herpetofauna of Pacific lowlands and adjacent interior valleys of México. Biological
Conservation, 130, 25–46. CHAPTER 7—CONCLUSIONS 263
Gillespie TW, Grijalva A, Farris CN (2000) Diversity, composition, and structure of tropical
dry forests in Central America. Plant Ecology, 147, 37–47.
Hafner DJ, Riddle BR (2005) Mammalian phylogeography and evolutionary history of
northern México’s deserts. pp. 225-245. In Biodiversity, ecosystems, and conservation in
northern México. (Cartron JL Ceballos G, Felger RS eds.). Oxford University Press, New
York.
Hasbún CR, Gómez G, Köhler A, Lunt DH (2005) Mitochondrial DNA phylogeography of
the Mesoamerican spiny-tailed lizards (Ctenosaura quinquecarinata complex): historical
biogeography, species status and conservation. Molecular Ecology, 14, 3095–3107.
Holderegger R, Wagner HH (2006) A brief guide to landscape genetics. Landscape Ecology,
21, 793–796.
Holderegger R, Wagner HH (2008) Landscape genetics. BioScience, 58, 199–207.
Holderegger R, Kamm U, Gugerli F (2006) Adaptive vs. neutral genetic diversity:
implications for landscape genetics. Landscape Ecology, 21, 797–807.
Janzen DH (1988) Tropical dry forests: the most endangered major tropical ecosystem.
Biodiversity (Wilson EO, ed.), pp 130–137. National Academy Press, Washington DC.
Joost S, Bonin A, Bruford MW, Després L, Conord C, Erhardt G, Taberlet P (2007) A spatial
analysis method (SAM) to detect candidate loci for selection: towards a landscape
genomics approach to adaptation. Molecular Ecology, 16, 3955–3969.
Klicka J, Zink RM (1997) The importance of recent ice ages in speciation: a failed paradigm.
Science, 12, 1666–1669. CHAPTER 7—CONCLUSIONS 264
Landguth EL, Cushman SA, Schwartz MK, McKelvey KS, Murphy M, Luikart G (2010)
Quantifying the lag time to detect barriers in landscape genetics. Molecular Ecology, 19,
4179–4191.
Lee-Yaw JA, Davidson A, McRae BH, Green DM (2009) Do landscape processes predict
phylogeographic patterns in the wood frog? Molecular Ecology, 18, 1863–1874.
Lindell J, Méndez de la Cruz FR, Murphy RW (2005) Deep genealogical history without
population differentiation: discordance between mtDNA and allozyme divergence in the
zebra-tailed lizard (Callisaurus draconoides). Molecular Phylogenetics and Evolution,
36, 682–694.
Lindell J, Méndez de la Cruz FR, Murphy RW (2008) Deep biogeographical history and
cytonuclear discordance in the black-tailed brush lizard (Urosaurus nigricaudus) of Baja
California. Biological Journal of the Linnean Society, 94, 89–104.
Luikart G, England PR, Tallmon D, Jordan S, Taberlet P (2003) The power and promise of
population genomics: from genotyping to genome typing. Nature Reviews Genetics, 4,
981–994.
Manel S, Schwartz MK, Luikart G, Taberlet P (2003) Landscape genetics: combining
landscape ecology and population genetics. Trends in Ecology and Evolution, 18, 189–
197.
Manel S, Poncet BN, Legendre P, Gugerli F, Holderegger R (2010) Common factors drive
adaptive genetic variation in Arabis alpina. Molecular Ecology, 19, 3824–3835.
McLean H (1989) Reconnaissance of a Pliocene marine embayment near Loreto, Baja
California Sur, Mexico. Pacific Section, Society of Economic Paleontologists and
Minerologists, Field Trip Guide Book, 63, 17–25. CHAPTER 7—CONCLUSIONS 265
Miles L, Newton AC, DeFries RS, Ravilious C, May I, Blyth S, Kapos V, Gordon JE (2006)
A global overview of the conservation status of tropical dry forests. Journal of
Biogeography, 33, 491–505.
Moritz C, Patton JL, Schneider CJ, Smith TB (2000) Diversification of rainforest faunas: an
integrated molecular approach. Annual Review of Ecology and Systematics, 31, 533–563.
Murphy PG, Lugo AE (1986) Ecology of tropical dry forest. Annual Review of Ecology,
Evolution, and Systematics, 17, 67–88.
Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity
hotspots for conservation priorities. Nature, 403, 853–858.
Robichaux RH, Yetman DA (2000) The Tropical Deciduous Forest of Alamos. Biodiversity
of a Threatened Ecosystem in Mexico. Tucson: The University of Arizona Press.
Safner T, Miller MP, McRae BH, Fortin MJ, Manel S (2011) Comparison of Bayesian
clustering and edge detection methods for inferring boundaries in landscape genetics.
International Journal of Molecular Sciences, 12, 865–889.
Storfer A, Murphy MA, Spear SP, Holderegger R, Waits LP (2010) Landscape genetics:
where are we now? Molecular Ecology, 19, 3496–3514.
Trejo RI, Dirzo R (2000) Deforestation of seasonally dry tropical forest towards its northern
distribution: a national and local analysis in México. Biological Conservation, 94, 133–
142.
Upton DE, Murphy RW (1997) Phylogeny of the side-blotched lizards (Phrynosomatidae:
Uta) based on mtDNA sequences: support for a mid-peninsular seaway in Baja
California. Molecular Phylogenetics and Evolution, 8, 104–113. CHAPTER 7—CONCLUSIONS 266
Umhoefer PJ, Dorsey RJ, Renne P (1994) Tectonics of the Pliocene Loreto basin, Baja
California Sur, Mexico, and evolution of the Gulf of California. Geology, 22, 649–652.
Vandergast AG, Bohonak AJ, Weissman DB, Fisher RN (2007) Understanding the genetic
effects of recent habitat fragmentation in the context of evolutionary history:
phylogeography and landscape genetics of a southern California endemic Jerusalem
cricket (Orthoptera: Stenopelmatidae: Stenopelmatus). Molecular Ecology, 16, 977–
992.
Wang IJ (2010) Recognizing the temporal distinctions between landscape genetics and
phylogeography. Molecular Ecology, 19, 2605–2608.
Wang IJ (2011) Choosing appropriate genetic markers and analytical methods for testing
landscape genetic hypotheses. Molecular Ecology, 20, 2480–2482.
Weir JT (2006) Divergent timing and patterns of species accumulation in lowland and
highland neotropical birds. Evolution, 60, 842–855
Zarza E, Reynoso VH, Emerson BC (2008) Diversification in the northern Neotropics:
mitochondrial and nuclear DNA phylogeography of the iguana Ctenosaura pectinata and
related species. Molecular Ecology, 17, 3259–3275.