https://cpw.cvlcollections.org/files/original/29d27c2483c075ba9eb38767b4639a05.pdf 0b714fc2cde571cf10afd3c82bd7dd84 PDF Text Text The research in this publication was partially or fully funded by Colorado Parks and Wildlife. Dan Prenzlow, Director, Colorado Parks and Wildlife • Parks and Wildlife Commission: Marvin McDaniel, Chair • Carrie Besnette Hauser, Vice-Chair Marie Haskett, Secretary • Taishya Adams • Betsy Blecha • Charles Garcia • Dallas May • Duke Phillips, IV • Luke B. Schafer • James Jay Tutchton • Eden Vardy �Evolutionary Applications Evolutionary Applications ISSN 1752-4571 ORIGINAL ARTICLE Fine-scale genetic correlates to condition and migration in a wild cervid Joseph M. Northrup,1 Aaron B. A. Shafer,2 Charles R. Anderson Jr,3 David W. Coltman4 and George Wittemyer1 1 2 3 4 Department of Fish, Wildlife, and Conservation Biology, Colorado State University, Fort Collins, CO, USA Department of Evolutionary Biology, Evolutionary Biology Centre, Uppsala University, Uppsala, Sweden Mammals Research Section, Colorado Parks and Wildlife, Grand Junction, CO, USA Department of Biological Sciences, University of Alberta, Edmonton, AB, Canada. Keywords genetic differentiation, heterozygosity fitness correlation, migration, mule deer, multilocus heterozygosity, Odocoileus hemionus, singlelocus heterozygosity, wildlife. Correspondence Joseph M. Northrup, Department of Fish, Wildlife, and Conservation Biology, Colorado State University, 1474 Campus Delivery, Fort Collins, CO 80523, USA. Tel.: +970 491 2302; fax: +970 491 5091; e-mail: joe.northrup@colostate.edu Received: 7 January 2014 Accepted: 30 June 2014 doi:10.1111/eva.12189 Abstract The relationship between genetic variation and phenotypic traits is fundamental to the study and management of natural populations. Such relationships often are investigated by assessing correlations between phenotypic traits and heterozygosity or genetic differentiation. Using an extensive data set compiled from freeranging mule deer (Odocoileus hemionus), we combined genetic and ecological data to (i) examine correlations between genetic differentiation and migration timing, (ii) screen for mitochondrial haplotypes associated with migration timing, and (iii) test whether nuclear heterozygosity was associated with condition. Migration was related to genetic differentiation (more closely related individuals migrated closer in time) and mitochondrial haplogroup. Body fat was related to heterozygosity at two nuclear loci (with antagonistic patterns), one of which is situated near a known fat metabolism gene in mammals. Despite being focused on a widespread panmictic species, these findings revealed a link between genetic variation and important phenotypes at a fine scale. We hypothesize that these correlations are either the result of mixing refugial lineages or differential mitochondrial haplotypes influencing energetics. The maintenance of phenotypic diversity will be critical to enable the potential tracking of changing climatic conditions, and these correlates highlight the need to consider evolutionary mechanisms in management, even in widely distributed panmictic species. Introduction Understanding variation in phenotypic traits related to fitness in wild populations is fundamental to the study of evolution and ecology. Such traits can be related to genetic variation at relatively fine spatial scales, and knowledge of these relationships can provide insight into important ecoevolutionary processes such as inbreeding depression, local adaptation, population structure, and speciation (Kupper et al. 2010; Olano-Marin et al. 2011; Shafer and Wolf 2013; Shafer et al. 2014). Moreover, these relationships can have implications for developing and implementing conservation and management plans that strive to account for evolutionary processes (e.g., maintenance of gene flow through protection of corridors or minimizing possible effects of inbreeding). Relationships between fine-scale genetic variation and phenotypic traits have been identified using a variety of methods. Chief among these in wild populations are heterozygosity–fitness correlations (HFCs; see Chapman et al. 2009) and correlations among genetic differentiation and phenotypic or ecological divergence (Shafer and Wolf 2013). Heterozygosity–fitness correlations are typically calculated between heterozygosity at neutral loci and phenotypic traits presumed to be proxies for fitness (Szulkin et al. 2010). Correlations can occur with a multilocus heterozygosity (MLH) metric, indicating a general genomewide effect of inbreeding, or heterozygosity at a single locus (single-locus heterozygosity; SLH), indicating local (either direct or indirect) effects due to linkage to a gene that affects fitness (Hansson et al. 2004). For the latter, individual neutral markers are hypothesized to show associative © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 937 �Genetic correlates in a wild cervid Northrup et al. overdominance as a result of the consequences of deleterious alleles or a fitness advantage at those linked loci (Frydenberg 1963; Houle 1989; David et al. 1995; David 1997; Pamilo and Palsson 1998). Screening for HFCs can be described as a tantalizing pursuit; significant relationships are rarely found and care must be used with interpretation as overall effect sizes often are variable and small (Chapman et al. 2009; Kardos et al. 2014), and numerous concerns (but also caveats) related to the HFC exist (Szulkin et al. 2010). Given the potential for false positives with SLH correlations, confidence in these relationships can be bolstered by appropriate statistical analyses and by examining the location of loci on the annotated genome of a related species that might provide post hoc links to causative agents (Von Hardenberg et al. 2007; Kupper et al. 2010; Kardos et al. 2014). In slight contrast, correlations between genetic differentiation and phenotypic (or ecological) divergence have been identified across taxa and appear to be relatively robust (Shafer and Wolf 2013; Sexton et al. 2014). While this pattern is generally regarded as evidence for local adaptation (Nosil 2012), ancestral (allopatric) divergence and secondary contact can confound interpretations of this correlation (Bierne et al. 2013) and, similar to HFCs, must be factored into interpretations and models. But beyond these caveats, correlations between phenotypic traits and both genetic diversity and differentiation can provide important indications of inbreeding and local adaptation that should be considered by managers (Shafer et al. 2014). Mule deer ecology and evolution Cervids (family Cervidae) are an ecologically important group of ungulate that have been the focus of numerous investigations into the relationship between genetic variation and phenotypic traits. Da Silva et al. (2009) showed that juvenile roe deer (Capreolus capreolus L.) survival was correlated with MLH; likewise, red deer (Cervus elaphus L.) birth weight, neonatal survival, and lifetime breeding success increased significantly with heterozygosity (Coulson et al. 1998; Slate et al. 2000), and individuals with the smallest antlers tended to have lower heterozygosity (Perez-Gonzalez et al. 2010). Furthermore, studies have shown correlations between genetic differentiation and social groups in white-tailed deer (Odocoileus virginianus Zimm.; Miller et al. 2010), and niche overlap in mule deer (Odocoileus hemionus Raf.; Pease et al. 2009). Among cervids, mule deer present an interesting species for which to examine correlations between phenotypic traits and genetic variation. Latch et al. (2009, 2014) showed that across their range, there are multiple phylogeographic lineages that presumably represent different refugia, although the species shows minimal population-level 938 genetic structure at large geographic scales (Cullingham et al. 2011b; Powell et al. 2013). Female mule deer also display fine-scale genetic structuring, likely due to the existence of related social groups (Cullingham et al. 2011b; Colson et al. 2013). In addition, hybridization with whitetailed deer can occur, resulting in fairly widespread genetic introgression (Carr et al. 1986; Cathey et al. 1998). Mule deer also exhibit substantial variation in important phenotypic traits such as body size and migratory behavior, both across their range (Anderson 1981; Wallmo 1981), and within populations (Monteith et al. 2011a; Lendrum et al. 2013). Lastly, mule deer are the subject of extensive management programs throughout North America, due to their importance as a game species [e.g., it was estimated that more than 30 000 mule deer were harvested in the state of Colorado in 2013 (Colorado Parks and Wildlife 2014)]. Both the aforementioned phenotypic traits are of paramount importance for survival and reproduction in this species. Condition is a fitness proxy as individuals rely heavily on fat and protein stores for survival on winter range when forage quality is low (Wallmo et al. 1977; Torbit et al. 1985). Body fat also influences annual survival of adult females (Bender et al. 2007), pregnancy and twinning rates (Johnstone-Yellin et al. 2009; Tollefson et al. 2010), and the probability of a female rearing a fawn through the summer (Johnstone-Yellin et al. 2009). Deer across much of their range migrate from high altitude, productive summer range to low altitude winter range and back again in the spring. Migrations typically match changes in resource availability (Fryxell and Sinclair 1988), with mule deer attempting to optimize migratory timing relative to both plant productivity and weather (snow depth and temperature) on their summer range (Monteith et al. 2011a; Lendrum et al. 2013). The timing of migratory onset is clustered around a few weeks each year, but individuals show different strategies in terms of early or late onset dates (Monteith et al. 2011a; Lendrum et al. 2013). Thus, migration timing is of clear interest in understanding the ecology of this species and, importantly, recent work has identified a clear genetic component to differences in this trait in other taxa (Ruegg et al. 2014; Toews et al. 2014). Both individual condition and migration are of interest to wildlife managers as recent anthropogenic development may threaten migratory routes for mule deer (Sawyer et al. 2005, 2009), and climate change could cause trophic mismatches (Post and Forchhammer 2008), with phenotypic plasticity in migration being suggested as a potential buffer for mule deer against this process (Monteith et al. 2011b). The importance of winter condition to deer survival has led to active research into means of improving winter condition through habitat manipulation and supplemental feeding (Bishop et al. 2009; Bergman et al. 2014). The existence of genetic correlations to these traits could pro- © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 �Northrup et al. vide insight into the effectiveness of management programs and aid managers in making decisions in light of evolutionary processes. Here, we examined the relationship between genetic variability and phenotypic traits in a wild mule deer population of the Piceance Basin, Colorado. Using an extensive data set consisting of more than 100 individual animals, we combined phenotypic, behavioral (global positioning system [GPS]), and genetic data to (i) examine whether genetic differentiation was correlated with migration timing, (ii) screen for specific mitochondrial haplotypes associated with migration timing, and (iii) test whether heterozygosity (multilocus and single locus) was associated with body mass and fat. We discussed the results in light of the phylogeographic history of mule deer and the metabolic role of the mitochondrion and highlight the importance of considering evolutionary processes in the management of this species. Genetic correlates in a wild cervid (A) NR NM RG SM (B) Materials and methods Sample collection and DNA extraction We captured adult (>1 year old) female mule deer using helicopter net gunning in four winter range study areas in the Piceance Basin of Northwestern Colorado (Fig. 1). Deer were captured in either December 2010 or March 2011. These dates were chosen because during December, deer have recently migrated from summer range and typically are in their best physical condition, while March represents the end of winter when deer typically are in their worst condition. Deer were transferred to processing sites where we weighed them using a portable scale, estimated body condition by palpating the rump (Cook et al. 2001, 2007, 2010) and measured the thickness of their subcutaneous rump fat and longissimus dorsi muscle using a portable ultrasound (Stephenson et al. 1998, 2002; Cook et al. 2001). The above measurements were used to calculate the percent ingesta-free body fat (hereafter fat) of each deer following Cook et al. (2010). Deer were fit with store-onboard GPS radio collars (Advanced Telemetry Systems, Isanti MN, USA) set to attempt a relocation on one of three schedules (once every 5 h, once every 60 min, or once every 30 min – meaning the relocation schedules varied by individual). Blood samples were taken for genetic analysis, and DNA was extracted using the DNeasyTM Blood and Tissue Kit (Qiagen, Inc., Valencia, CA, USA) following the manufacturer’s protocol. Microsatellite genotyping and DNA sequencing We amplified 17 microsatellite loci using a previously optimized multiplex reaction from Cullingham et al. (2011a) and single PCRs. The mitochondrial control region was Figure 1 (A) Winter range areas Ryan Gulch [RG], South Magnolia [SM], North Magnolia [NM] and North Ridge [NR]) and simplifications of migratory routes, with arrows indicating general location of summer ranges for mule deer in the Piceance Basin and (B) location of study within the United States. Adapted with permission from Lendrum et al. (2013). sequenced using both the primers from Latch et al. (2009) and LGL215 and ISM015 from Purdue et al. (2006). PCR conditions and basic population genetic analyses are available in Appendices S1 and S2. For the microsatellite data, we first used STRUCTURE 2.3.3 (Pritchard et al. 2000) to assess genetic structure (1 000 000 iterations with 25% removed as a burn-in repeated five times for each number of possible populations [k] ranging from 1 to 5). We assumed an admixed model with correlated allele frequencies (Falush et al. 2003) and used the LOCPRIOR parameter to allow location information to assist in the clustering. Next, we calculated overall MLH as the average of heterozygosity at each locus and SLH as binary variables indicating heterozygosity (1) or homozygosity (0) at each locus. Pairwise relatedness between all individuals was estimated with the Queller and Goodnight (QG) relationship coefficient using the software SPAGEDI v.1.3 (Hardy and Vekemans 2002). We also constructed a coancestry matrix using the software MOL_COAN v.3 (Fernandez and Toro 2006). Here, a simulated annealing approach was used to create virtual common ancestors of the genotyped individuals, producing pedigree-like rela- © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 939 �Genetic correlates in a wild cervid Northrup et al. tionship coefficients. Model parameters consisted of 200 steps with 5000 solutions tested per step, an initial temperature of 0.01 and increase of 0.75. We simulated two previous generations, each consisting of 1000 males and 1000 females. For mitochondrial DNA (mtDNA; conducted on a subset of individuals), we constructed a minimum-spanning tree among haplotypes using ARLEQUIN v. 3.5.1.3 (Excoffier and Lischer 2010) and edited it with HAPSTAR v0.7 (Teacher and Griffiths 2011). Neighbor-joining analysis using pairwise deletion and both P and K2 distances was conducted using the software package MEGA v.5 (Tamura et al. 2011). Bayesian analysis was conducted in MRBAYES v.3.1.2 (Huelsenbeck and Ronquist 2001) with a model of nucleotide substitution determined from MODELTEST v.3.07 (Posada and Crandall 1998). For the Bayesian phylogenetic analysis, we used default priors with two independent runs of four chains (three heated) run for 10 000 000 generations, with the first 25% discarded as a burn-in. Confidence in topologies was evaluated based on 1000 bootstrap replicates (for the neighbor joining) or posterior distributions. All three methods were compared to identify common mitochondrial haplogroups. Genetic correlates to phenotypic traits Both migration and body condition are phenotypic traits that are important to the fitness of mule deer. However, only condition can be thought of as a proxy for fitness. Thus, we used two separate analytical frameworks to examine genetic correlations with these traits. For migration, we examined the relationship between mitochondrial haplotypes and genetic differentiation to determine whether there was a genetic component to the timing of migration (an isolationby-ecology analysis, sensu Shafer and Wolf 2013). For body condition, a fitness proxy, we followed the general HFC framework discussed in Chapman et al. (2009). Genetic–migration correlates After GPS radio collars were recovered and data were downloaded, we calculated the initiation and termination dates of spring and fall migration (i.e., the dates at which deer started or finished their migration) in ARCMAP 10.1 (Environmental Systems Research Institute, Redlands, CA, USA). Migration was demarcated as the time period during which deer travelled between their winter and summer home ranges. Home ranges were determined by outlining a minimum convex polygon around all locations that occurred prior to directed movement, without return, away from the summer or winter range areas. We first examined the relationship between mtDNA haplogroup (derived from haplotype and phylogenetic 940 analyses) and the dates of spring and fall migrations. For this analysis, we corrected the Julian date of migration to the earliest date among all individuals. The resulting data represented a count of the number of days since the earliest arriving or leaving migrant had terminated or initiated their migration. These data were analyzed using negative binomial regression (see Appendix S3 for model formulation). We included covariates for the mtDNA haplogroup to which each deer was assigned (categorical) as well as a covariate for the age of the animal and binary covariates indicating winter range study area (i.e., three separate covariates indicating whether the deer was from a winter range study area [1] or not [0]). Before models were run, correlations among covariates were examined to assess collinearity (no predictors were correlated at |r| > 0.7) and age was standardized ½ðx � �xÞ=rx �, a common procedure in regression to aid in interpretability of coefficient estimates (Gelman and Hill 2007). We fitted all models under a Bayesian framework in JAGS (Plummer 2012) and R3.0.1 (R Core Team 2013), using the ‘rjags’ package (Plummer 2013). See Appendix S3 for specifics of model runs and assessment of convergence. To assess the fit of the models, we calculated residuals (observed – predicted values) and plotted them against the fitted values to examine any potential patterns in residuals. Secondly, we examined correlations between genetic relatedness metrics and similarity in migration using Mantel tests. For this analysis, we calculated absolute pairwise distances (calculated in days) between each individual’s migration termination or initiation dates leaving us with four matrices representing differences in migration timing for spring and fall. The relationships between relatedness indices (QG and coancestry) and migratory behavior (dates) were evaluated in R3.0.1 (R Core Team 2013) using Mantel tests (Mantel 1967) under 10 000 permutations as implemented by the Ecodist library (Goslee and Urban 2007). Here, a comparison is made between relatedness and the difference in migration timing, and thus, a negative relationship is expected if there is a genetic signature – that is, more closely related individuals have more similar migration timing. To account for similarities among individuals inhabiting similar areas or grouping together, we ran two partial Mantel tests controlling for the distance between the centroids of individuals’ winter range and summer range (Fig. 1). Significance was assessed by examining 95% confidence intervals. Genetic–condition correlates We next examined whether there was a relationship between either MLH or SLH and condition metrics (mass and fat) using the HFC framework. We fit hierarchical (i.e., random effects) models in a Bayesian framework. The presence or the absence of a relationship was determined © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 �Northrup et al. by examining the posterior probability distributions of each coefficient to determine the probability that either MLH or heterozygosity at any single locus was related to condition. In all models, we included covariates for either MLH or SLH, the age of the animal, a binary variable for if the data came from a March capture (both mass and fat are expected to be lower in March), and binary variables indicating which of the four winter range areas the deer was captured in (as in the migration analysis, above). We tested between models with solely a linear effect or a quadratic effect of age using the deviance information criteria (DIC; Spiegelhalter et al. 2002; but with the effective number of parameters calculated as in Plummer 2012). Identity disequilibrium among loci (i.e., covariance of heterozygosity among loci) was used to infer the validity of MLH correlation: Accordingly, we calculated g2, where a value of zero means no variance in inbreeding (Szulkin et al. 2010). We examined the relationship between heterozygosity and mass or fat using linear regression and beta regression, respectively. Mass was natural log-transformed to ensure proper support (i.e., untransformed mass is strictly positive, while linear regression allows for negative values; log transformation addresses this issue), while beta regression was used because it is proper for dependent variables ranging between 0 and 1 as percent body fat does. Because there were multiple condition measures for certain deer (i.e., those captured in both March and December), for both analyses, we allowed the intercept to vary by individual, estimating a population-level intercept (i.e., we fit a random intercept by individual), with all other coefficient values fixed. See Appendix S3 for details of model parameters and convergence assessment. To assess the fit of the models, we calculated residuals (observed – predicted values) and plotted them against the fitted values to examine any potential patterns in residuals. Genetic correlates in a wild cervid most likely (i.e., had the lowest likelihood score). Based on winter range, FIS values were as follows: NR = �0.05 (P = 0.02), NM = �0.02 (P = 0.16), RG = �0.03 (P = 0.07), and SM = 0.01 (P = 0.31). The MOL_COAN analysis produced a matrix of pedigree-like coefficients for all individuals; we note the one suspected mother–daughter pairing had a coefficient of 0.50, suggesting the results were indeed reflective of pedigree data. We sequenced the mitochondrial control region in a subset of animals (n = 81). For comparison with data from Latch et al. (2009), we parsed the data set down to 545 base pairs (GenBank submission KM061069–KM061151). Examining the mtDNA, 37 unique haplotypes were observed (Fig. 2). The neighbor-joining and Bayesian phylogeny (based on a GTR + I + G substitution model) produced essentially the same topology (Appendix S4): a major split between two clades was highly supported, while a third, more tenuous clade was evident in the neighbor analysis with some support in the Bayesian analysis (posterior probability = 0.60). The three groupings are identified in the haplotype network (Fig. 2). Genetic correlates to phenotypic traits Condition and migration data We obtained mass and fat measures on 134 adult female mule deer. Migration data were not obtained for all deer due to mortalities, collar failure, or because some deer were 2 3 Results Genotype and mitochondrial sequence data A total of 134 adult female deer were captured with 30 captured in the NM area, 30 in the NR area, 44 in the RG area, and 30 in the SM area (102 in December, and 79 in March, with 47 caught during both capture periods; see Appendix S5 for details). Deer ranged in age from yearlings to more than 11 years old, with a median age of 5.5 years old (See Appendix S5). All 134 deer were genotyped at 17 loci producing a data set that was 99% complete (data available from the Dryad Digital Repository: http://datadryad.org/ resource/doi:10.5061/dryad.3vc1b). All markers were in Hardy–Weinberg Equilibrium, and there was no evidence of linkage (diversity statistics by loci are presented in Appendix S2). The STRUCTURE-based analysis of the microsatellites suggested a single, homogenous population was RG 11 13 1 NM SM NR Figure 2 Mitochondrial control region haplotype network and winter range area assignments. Circle size is proportional to the haplotype frequency with small black circles representing undetected, intermediate haplotypes. Haplotypes are colored according to winter range area. The dashed circle outlines and corresponding numbers are in reference to the phylogenetic clades (Appendix S4). © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 941 �Genetic correlates in a wild cervid Northrup et al. not collared during capture. Thus, our total sample for microsatellites analyses examining relationships with migration consisted of 104 and 95 deer for spring and fall migration, respectively. Our total sample for mtDNA analyses consisted of 65 and 59 deer for spring and fall migration, respectively. In addition, two deer did not leave summer range, while collars were still attached and thus were excluded from the fall migration analyses. During spring, deer initiated migration between April 11 and June 1 and terminated migration between April 19 and June 21. During the fall, deer initiated migration between October 4 and November 8 and terminated migration between October 6 and November 14. Genetic–migration correlates For all regression models, hereafter, we made inference based on the proportion of the posterior distributions that fell to one side of 0. Winter range area was related to fall migration termination and initiation dates, while age was not related to migration timing in any of the analyses (Table 1; Appendix S5). The mtDNA haplogroups were related to both fall termination and initiation, although Table 1. Covariates, median coefficient (coeff.) values, and the probability (prob.) of either a negative or positive effect of the covariate from negative binomial regression model on mule deer fall migration termination dates from deer in the Piceance basin, Colorado. Covariate Median coeff. value Neighbor-joining clades Intercept 3.08 Age �0.09 Winter range NR* 0.16 RG† �0.38 SM‡ �0.56 mtDNA Haplogroup 2§ �0.46 Haplogroup 3§ �0.33 Bayesian clades Intercept 2.932 Age �0.095 Winter range NR* 0.1768 RG† �0.270 SM‡ �0.440 mtDNA Haplogroup 2§ �0.350 Prob. coeff. is negative Prob. coeff. positive 0.00 0.88 1.00 0.12 0.22 0.96 0.99 0.78 0.04 0.01 0.99 0.94 0.01 0.06 0.000 0.90 1.000 0.10 0.22 0.90 0.97 0.78 0.10 0.03 0.97 0.03 *Deer captured in the NR winter range, with NM as the reference category. †Deer captured in the RG winter range, with NM as the reference category. ‡Deer captured in the SM winter range, with NM as the reference category. §mtDNA haplogroup 1 is the reference category. 942 both the effect itself and the probability of an effect were lower for fall initiation (Table 1; Appendix S5). For haplogroups identified by the Bayesian phylogenetic analysis, our models predicted that deer in haplogroup 2 terminated migration 6 days earlier on average than those in haplogroup 1 (see Fig. 2 for haplogroups), while for the neighbor-joining analysis, models predicted that deer in haplogroups 2 and 3 terminated migration on average 7 and 9 days earlier than those in haplogroup 1. Plots of residuals against fitted values showed no trend, although the six largest negative residuals were all from the NR winter range area, indicating the potential for a missing covariate (Appendix S5). The microsatellites analyses showed that related individuals generally migrated at similar times regardless of the distance between them on summer or winter range (Table 2). Genetic–condition correlates There was weak evidence for identity disequilibrium (g2 = 0.01, P = 0.07); however, MLH was a poor predictor Table 2. Mantel test models, Mantel’s r and lower and upper confidence limits (CL), calculated through randomization, for models examining correlation between relatedness metrics (Queller-Goodnight [QG] and coancestry) and migration dates, for mule deer in the Piceance basin, Colorado. End spring and end fall indicate the termination of spring and fall migration, respectively. Start spring and start fall indicate the initiation of spring and fall migration, respectively. Winter distance and summer distance indicate the distance between winter and summer range centroids. All values are presented as Mantel r (lower CL, upper CL). Vertical lines (|) indicate partial Mantel tests with the covariate that is controlled for following the vertical line. Migratory metric QG Coancestry End spring End spring | winter distance End spring | summer distance End fall End fall | winter distance End fall | summer distance Start spring Start spring | winter distance Start spring | summer distance Start fall Start fall | winter distance Start fall | summer distance �0.04 (�0.06, �0.01) �0.06 (�0.09, �0.01) �0.02 (�0.04, �0.001) �0.07 (�0.10, �0.03) �0.03 (�0.05, �0.01) �0.05 (�0.08, �0.02) �0.04 (�0.06, �0.01) �0.04 (�0.06, �0.01) �0.02 (�0.05, 0.02) �0.02 (�0.05, 0.02) �0.04 (�0.06, �0.01) �0.02 (�0.05, 0.01) 0.002 (�0.02, 0.02) 0.01 (�0.02, 0.02) �0.03 (�0.07, 0.01) �0.03 (�0.06, 0.01) 0.01 (�0.01, 0.04) �0.02 (�0.06, 0.01) �0.05 (�0.07, �0.03) �0.05 (�0.06, �0.01) �0.05 (�0.08, �0.03) �0.05 (�0.08, �0.03) �0.05 (�0.07, �0.03) �0.05 (�0.08, �0.03) © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 �Northrup et al. Genetic correlates in a wild cervid of both body mass and fat in all models (Appendix S5 Table 2), while heterozygosity at individual loci was strongly related to condition measures (Table 3; Appendix S5 Table 2). Because heterozygosity at individual loci was the only significant correlates to the phenotypic traits, we continued with this model only. When examining the relationship between SLH and body mass, models with a quadratic term for age fit the data slightly better than those with a linear term, with evidence for greater body mass for middle aged deer compared with young or old deer (Appendix S5 Table 2). When examining fat, models with a linear effect of age fit the data slightly better, and age was a poor predictor of fat (Table 3; Appendix S5 Table 2). Winter range area was weakly related to both body mass and fat (<95% of posterior on one side of 0; Table 3; Appendix S5 Table 2). Heterozygosity at two loci (RT30 and P) was strongly related to fat (>95% of posterior on one side of 0; Table 3. Covariates, median coefficient (coeff.) values, and the probability (prob.) of either a negative or positive effect of the covariate from multilevel beta regression on the percent body fat of mule deer in the Piceance basin, Colorado. Covariate Intercept Age March capture Winter range NR* RG† SM‡ Microsatellite loci INRA011 RT30 BBJ K BL25 BM6438 BM848 RT7 N ETH152 BM6506 P D BM4107 RT5 OCAM R Median coeff. value Prob. coeff. is negative Prob. coeff. positive �2.15 �0.05 �0.52 1 0.89 1 0 0.11 0 �0.10 �0.11 �0.06 0.80 0.82 0.69 0.20 0.18 0.31 �0.13 �0.24 0.08 �0.03 0.07 �0.001 �0.11 �0.08 0.09 �0.004 0.02 0.18 0.092 0.05 0.15 0.02 �0.08 0.93 0.99 0.22 0.65 0.27 0.50 0.87 0.72 0.22 0.52 0.40 0.04 0.13 0.32 0.13 0.41 0.81 0.07 0.01 0.78 0.35 0.73 0.50 0.13 0.28 0.78 0.48 0.60 0.96 0.87 0.68 0.87 0.59 0.19 *Deer captured in the NR winter range, with NM as the reference category. †Deer captured in the RG winter range, with NM as the reference category. ‡Deer captured in the SM winter range, with NM as the reference category. Table 3; Fig. 3). Plots of residuals against fitted values showed a positive trend, with all of the largest fitted values showing positive residuals (Appendix S5). To guard against false positives, we refit the models with a strong mean 0 multivariate normal prior on the coefficients, which shrinks coefficient estimates toward 0 (the standard deviation on the prior was taken as the standard deviation of the median coefficient values; approximately 0.14; Gelman et al. 2012). Discussion We documented relationships between phenotypic traits recognized as being critical to fitness and genetic variation at a very fine spatial scale in female mule deer. These results provide insight into the genetic structuring of the population and the possible genetic drivers shaping the diversity of phenotypes and migration strategies seen in this important game species. These findings have potential implications for conservation and management, particularly in light of contemporary climatic changes and white-tailed deer expansion (Latham et al. 2011), as both migration timing and body condition are influential traits for mule deer survival and reproduction that vary among individuals in a population (Monteith et al. 2011b, 2013). Examining these traits conjointly provided a more complete picture of the genetic contributions to important phenotypic traits in this population and cervids in general. Genetic-migration correlations Fall and spring migration dates were more similar among related females. An individual’s mtDNA haplogroup also was a stronger predictor of fall than spring migration – even when controlling for winter or summer range. The mtDNA haplotype effect is particularly striking, given there appears to be virtually no spatial clustering of haplotypes (Fig. 2). Female philopatry and relatedness among social groups would explain the pattern in the form of learning (e.g., the majority of white-tailed deer fawns follow their mother’s migration route; Nelson 1998); however, our model accounted for such effects through the range covariates (i.e., if daughters were following their mother’s migration path they would also share a winter and summer range), and the diversity of haplotypes suggests many different matrilines. In addition, upon examination of individual migratory routes, we found only two deer that shared an identical route. An analysis including males could test this hypothesis (sensu Nielsen et al. 2013) or at least be viewed as an independent replicate as males are more prone to disperse (Nelson 1993). Interestingly, Colorado represents a confluence of several different refugial lineages (Latch et al. 2009), with recolonization routes and so-called hybrid hot spot clusters falling © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 943 �Northrup et al. 0.0 −0.2 −0.4 Coefficient value 0.2 0.4 Genetic correlates in a wild cervid INRA011 RT30 BBJ K BL25 BM6438 BM848 RT7 N ETH152 BM6506 P D BM4107 RT5 OCAM R Locus Figure 3 Box plots of coefficients for effect of microsatellite loci on mule deer body fat percent. Coefficients were obtained through beta regression model in a Bayesian hierarchical framework. Box plots represent median (black line) interquartile range (box bounds) and upper and lower 95% bounds (whiskers) of coefficient values. directly in northwestern Colorado (Swenson and Howard 2005). We hypothesize that the mtDNA effect we documented is either: (i) reflective of different refugial histories and biogeography of the mtDNA lineages (Latch et al. 2009), where for example, mule deer originating in northern regions would have locally adapted phenotypes and distinct haplotypes linked to earlier migration times than those from the south (a carryover effect); or (ii) due to differences in energetics related to mtDNA, where, for example, Toews et al. (2014) showed that mitochondrial introgression (where different haplotypes had different energetic outputs) was responsible for differing migratory behavior in a warbler transition zone. Monteith et al. (2011a) and Lendrum et al. (2013) showed that spring migration timing is closely linked to plant phenology, as deer aim to arrive on their summer range close in time to the peak of plant productivity. Spring arrival dates are more likely to follow plant phenology on individual deer summer ranges, whereas fall migration is linked to weather (temperature and snow on summer range) and individual characteristics such as age and condition. Monteith et al. (2011a) suggested that prime age individuals in the best condition can adopt a strategy by which they stay on summer range for longer time periods to consume higher quality vegetation in spite of the potential for being caught in adverse weather, while poorer quality individuals cannot take on such risks. The individual characteristic hypothesis of Monteith et al. (2011b) provides support for the energetics scenario (ii above), whereby individuals with certain haplotypes might be better suited for taking on the risks associated with remaining on summer range later in the season due to associated differences in energetics. Fine-scale natal dispersal has been shown to have a heritable basis in albatross (Charmantier et al. 2011), 944 and genotype–phenotype associations are thought to be important next steps in migration studies (Liedvogel et al. 2011). For the carryover effect to be true, the mtDNA lineages must reflect nuclear differences that (at least partially) encode for differences in migratory behavior or have a physiological effect. While our results cannot tease apart a specific nuclear or mitochondrial effect, given the mtDNA migration effect shown in warblers (Toews et al. 2014), we think this is worth following up on using both biochemical modeling and genome-wide scans (i.e., with mtDNA haplotype as the response measure or interaction term). Importantly, recent development in the western United States has raised concerns over the sustainability of mule deer migratory routes (Sawyer et al. 2005, 2009), and under climate change, there is the potential for trophic mismatch for migratory species, whereby migrations occur asynchronously with plant phenology (Post et al. 2008). Monteith et al. (2011b) suggested that plasticity in mule deer migration might allow the species to avoid such mismatches; however, if there is a genetic basis for the variability in migration among individuals, there may be less plasticity and more natural selection at work (Nelson 1998). Mitochondrial introgression with white-tailed deer is likely to be unidirectional (Carr et al. 1986), which could jeopardize the adaptive potential if hybridizations increase. However, we note that there is no evidence of white-tailed deer presence in our study area, and thus, hybridization is not a concern at this point. The potential for loss of migratory routes to development combined with climate change and hybridization highlight the importance of maintaining the existing genetic variability in diverse migratory phenotypes. © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 �Northrup et al. Genetic–condition correlations Fat is an important determinant of fitness for mule deer (Bender et al. 2007; Johnstone-Yellin et al. 2009; Tollefson et al. 2010). We identified two genetic markers as having relationships with fat, although the relationships were antagonistic (i.e., one had a positive relationship with fat and the other negative). Similar results have been seen in studies of both the Kentish plover (Charadrius alexandrinus L.; Kupper et al. 2010) and the blue tit (Parus caeruleus L.; Olano-Marin et al. 2011). With the contrasting signals of the two markers, interpretations of what these relationships represent become muddled. Olano-Marin et al. (2011) viewed the negative correlation as evidence for direct effects of the neutral loci, with the positive correlation due to inbreeding. Inbreeding in our study area is not supported by the FIS values and difficult to imagine given the population size and deer ecology. Based on the evidence for a mixing of different mitochondrial lineages and effect sizes, the negative relationship to body fat of RT30 (0.99 probability and nearly double the effect size as all other loci) is the most likely to be genuine. However, given the concern over spurious HFCs, we must still consider the possibility of Type I errors (i.e., false positives). The potential for type I errors is of particular concern when detecting local effects and examining multiple models (Szulkin et al. 2010). In light of this concern, we highlight three points of support for the recorded relationship. First, the effect sizes of the significant coefficients were substantially greater than those of the other loci (Fig. 3). Second, we refit all models that had significant coefficients, but with a strong mean 0 multivariate normal prior on the coefficients. This approach shrinks all estimates toward 0, acting as a penalty and reducing the number of significant covariates (Gelman et al. 2012). In the case of the SLH – fat correlation, all significant results (probability of an effect > 0.95) remained. Lastly, the proximity of a locus in question relative to genes of known effect can be taken as supportive evidence for understanding single-locus HFCs (Von Hardenberg et al. 2007; Kupper et al. 2010). Slate et al. (2002) observed considerable synteny in ruminants, and more than half of the microsatellites used in their deer linkage map had been used for the same purposes in cow and sheep. When we screened RT30 against the annotated cow genome (using BLAST), both primers colocalized with 100% identity to a region with the closest known gene being that of TBC1D1. Interestingly, this gene regulates cell growth and differentiation and has been shown to influence fat metabolism in mice and humans (Stone et al. 2006; Chadt et al. 2008). Given the combination of divergent mtDNA lineages in our study area and panmixia (k = 1), a slight disruption of co-adapted alleles that are linked to fat metabolism could explain the negative correlation between Genetic correlates in a wild cervid this locus and fat (we emphasize these results represent a small effect as body fat was predicted to decrease body fat by <0.2% in the model). This is predicted to outcome when locally adapted lineages mix, and it has been recently suggested for grizzly bears in an area where they are subject to large-scale human-assisted migration (Shafer et al. 2014). While the above lines of evidence offer support to the effect of RT30 on fat being genuine, given the small number of loci examined, we must remain skeptical about this relationship. Rather, we present these findings as noteworthy and in need of confirmation by studies with larger samples and with genomic methods. Conclusions and evolutionary applications We have shown fine-scale relationships between genetic variation and phenotypic traits in mule deer that have not been found in previous work on this species. Our study identified fine-scale genetic correlates to both migration timing and body fat that are likely overlooked (and probably unexpected) in this species. These results have potential management implications for mule deer, which are under substantial human pressure from a multitude of stressors (Sawyer et al. 2006). The genetic polymorphisms in this population that are linked to phenotypic traits related to phenology and metabolic variation could prove important in the face of climate change and other anthropogenic stressors that are likely to affect both optimal timing of migration and the role of fat stores in survival and reproduction. Monitoring hybridization with white-tailed deer should also be considered with respect to the mtDNA effect, as introgression is likely to go from white-tailed to mule deer (Carr et al. 1986) and could alter the adaptive potential. Efforts should be made to better characterize additional drivers behind this phenotypic and genetic variation in an effort to maintain a diversity of phenotypes that might best be able to adapt to novel conditions. Screening for similar associations in more imperiled deer populations (and cervid species) may help shed light on local population dynamics and better inform management decisions. Acknowledgements Mule deer capture and monitoring was funded by Colorado Parks and Wildlife (CPW), White River Bureau of Land Management, ExxonMobil Production/XTO Energy, WPX Energy, Shell Exploration and Production, EnCana Corp., Marathon Oil Corp., Federal Aid in Wildlife Restoration (W-185-R), the Colorado Mule Deer Foundation, the Colorado Mule Deer Association, Safari Club International, Colorado Oil and Gas Conservation Commission, and the Colorado State Severance Tax. Genetic analysis was funded by startup funds to GW. We thank L. Wolfe, C. © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 945 �Genetic correlates in a wild cervid Northrup et al. Bishop, D. Finley (CPW) and numerous field technicians for capture expertise and field assistance. We thank Quicksilver Air, Inc. for deer captures, and L. Gepfert (CPW) and Coulter Aviation, Inc. for fixed-wing aircraft support. M. Alldredge and M. Rice (CPW), along with 2 anonymous reviewers provided helpful comments that greatly improved the manuscript. Data archiving statement Data for this study are available from GenBank Accession Numbers KM061069–KM061151 and from the Dryad Digital Repository: http://doi.org/10.5061/dryad.3vc1b. Literature cited Anderson, A. E. 1981. Morphological and physiological characteristics. In: O. C. Wallmo, ed. Mule and Black-Tailed Deer of North America, pp. 27–98. University of Nebraska Press, Lincoln, NE. Bender, L. C., L. A. Lomas, and J. Browning 2007. Condition, survival, and cause-specific mortality of adult female mule deer in north-central New Mexico. Journal of Wildlife Management 71:1118–1124. Bergman, E. J., C. J. Bishop, D. J. Freddy, G. C. White, and P. F. Doherty 2014. Habitat management influences overwinter survival of mule deer fawns in Colorado. The Journal of Wildlife Management 78:448– 455. Bierne, N., P.-A. Gagnaire, and P. David 2013. The geography of introgression in a patchy environment and the thorn in the side of ecological speciation. Current Zoology 59:72–86. Bishop, C. J., G. C. White, D. J. Freddy, B. E. Watkins, and T. R. Stephenson 2009. Effect of enhanced nutrition on mule deer population rate of change. Wildlife Monographs 172:1–28. Carr, S. M., S. W. Ballinger, J. N. Derr, L. H. Blankenship, and J. W. Bickham 1986. Mitochondrial DNA analysis of hybridization between sympatric white-tailed deer and mule deer in west Texas. Proceedings of the National Academy of Sciences 83:9576–9580. Cathey, J. C., J. W. Bickham, and J. C. Patton 1998. Introgressive hybridization and nonconcordant evolutionary history of maternal and paternal lineages in North American deer. Evolution 52:1224– 1229. Chadt, A., K. Leicht, A. Deshmukh, L. Q. Jiang, S. Scherneck, U. Bernhardt, T. Dreja et al. 2008. Tbc1d1 mutation in lean mouse strain confers leanness and protects from diet-induced obesity. Nature Genetics 40:1354–1359. Chapman, J. R., S. Nakagawa, D. W. Coltman, J. Slate, and B. C. Sheldon 2009. A quantitative review of heterozygosity–fitness correlations in animal populations. Molecular Ecology 18:2746–2765. Charmantier, A., M. Buoro, O. Gimenez, and H. Weimerskirch 2011. Heritability of short-scale natal dispersal in a large-scale foraging bird, the wandering albatross. Journal of Evolutionary Biology 24:1487– 1496. Colorado Parks and Wildlife. 2014. Statistics. http://cpw.state.co.us/ thingstodo/Pages/Statistics.aspx (accessed on 25 June 2014). Colson, K. E., T. J. Brinkman, D. K. Person, and K. J. Hundertmark 2013. Fine-scale social and spatial genetic structure in Sitka blacktailed deer. Conservation Genetics 14:439–449. Cook, R. C., J. G. Cook, D. L. Murray, P. Zager, B. K. Johnson, and M. W. Gratson 2001. Development of predictive models of nutritional 946 condition for Rocky Mountain elk. Journal of Wildlife Management 65:973–987. Cook, R. C., T. R. Stephenson, W. L. Myers, J. G. Cook, and L. A. Shipley 2007. Validating predictive models of nutritional condition for mule deer. Journal of Wildlife Management 71:1934–1943. Cook, R. C., J. G. Cook, T. R. Stephenson, W. L. Myers, S. M. Mccorquodale, D. J. Vales, L. L. Irwin et al. 2010. Revisions of rump fat and body scoring indices for deer, elk, and moose. Journal of Wildlife Management 74:880–896. Coulson, T. N., J. M. Pemberton, S. D. Albon, M. Beaumont, T. C. Marshall, J. Slate, F. E. Guinness et al. 1998. Microsatellites reveal heterosis in red deer. Proceedings of the Royal Society B-Biological Sciences 265:489–495. Cullingham, C. I., E. H. Merrill, M. J. Pybus, T. K. Bollinger, G. A. Wilson, and D. W. Coltman 2011a. Broad and fine-scale genetic analysis of white-tailed deer populations: estimating the relative risk of chronic wasting disease spread. Evolutionary Applications 4:116–131. Cullingham, C. I., S. M. Nakada, E. H. Merrill, T. K. Bollinger, M. J. Pybus, and D. W. Coltman 2011b. Multiscale population genetic analysis of mule deer (Odocoileus hemionus hemionus) in western Canada sheds new light on the spread of chronic wasting disease. Canadian Journal of Zoology-Revue Canadienne De Zoologie 89:134–147. Da Silva, A., J. M. Gaillard, N. G. Yoccoz, A. J. M. Hewison, M. Galan, T. Coulson, D. Allaine et al. 2009. Heterozygosity-fitness correlations revealed by neutral and candidate gene markers in roe deer from a long-term study. Evolution 63:403–417. David, P. 1997. Modeling the genetic basis of heterosis: tests of alternative hypotheses. Evolution 51:1049–1057. David, P., B. Delay, P. Berthou, and P. Jarne 1995. Alternative models for allozyme-associated heterosis in the marine bivalve Spisula-ovalis. Genetics 139:1719–1726. Excoffier, L., and H. E. L. Lischer 2010. Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources 10:564–567. Falush, D., M. Stephens, and J. K. Pritchard 2003. Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164:1567–1587. Fernandez, J., and M. A. Toro 2006. A new method to estimate relatedness from molecular markers. Molecular Ecology 15:1657–1667. Frydenberg, O. 1963. Population studies of a lethal mutant in Drosophila melanogaster. 1. Behaviour in populations with discrete generations. Hereditas 50:89–116. Fryxell, J. M., and A. R. E. Sinclair 1988. Causes and consequences of migration by large herbivores. Trends in Ecology & Evolution 3:237– 241. Gelman, A., and J. Hill 2007. Data Analysis Using Regression and Multilevel/Hierarchical Models. Cambridge University Press, Cambridge, New York. Gelman, A., J. Hill, and M. Yajima 2012. Why we (usually) don’t have to worry about multiple comparisons. Journal of Research on Educational Effectiveness 5:189–211. Goslee, S. C., and D. L. Urban 2007. The ecodist package for dissimilarity-based analysis of ecological data. Journal of Statistical Software 22:1–19. Hansson, B., H. Westerdahl, D. Hasselquist, M. Akesson, and S. Bensch 2004. Does linkage disequilibrium generate heterozygosity-fitness correlations in great reed warblers? Evolution 58:870–879. Hardy, O. J., and X. Vekemans 2002. SPAGEDi: a versatile computer program to analyse spatial genetic structure at the individual or population levels. Molecular Ecology Notes 2:618–620. © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 �Northrup et al. Houle, D. 1989. Allozyme-associated heterosis in Drosophila melanogaster. Genetics 123:789–801. Huelsenbeck, J. P., and F. Ronquist 2001. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17:754–755. Johnstone-Yellin, T. L., L. A. Shipley, W. L. Myers, and H. S. Robinson 2009. To twin or not to twin? Trade-offs in litter size and fawn survival in mule deer. Journal of Mammalogy 90:453–460. Kardos, M., F. W. Allendorf, and G. Luikart 2014. Evaluating the role of inbreeding depression in heterozygosity-fitness correlations: how useful are tests for identity disequilibrium? Molecular Ecology Resources 14:519–530. Kupper, C., A. Kosztolanyi, J. Augustin, D. A. Dawson, T. Burke, and T. Szekely 2010. Heterozygosity-fitness correlations of conserved microsatellite markers in Kentish plovers Charadrius alexandrinus. Molecular Ecology 19:5172–5185. Latch, E. K., J. R. Heffelfinger, J. A. Fike, and O. E. Rhodes 2009. Species-wide phylogeography of North American mule deer (Odocoileus hemionus): cryptic glacial refugia and postglacial recolonization. Molecular Ecology 18:1730–1745. Latch, E. K., D. M. Reding, J. R. Heffelfinger, C. H. Alcal�a-Galv�an, and O. E. Rhodes 2014. Range-wide analysis of genetic structure in a widespread, highly mobile species (Odocoileus hemionus) reveals the importance of historical biogeography. Molecular Ecology 23:3171– 3190. Latham, A. D. M., M. C. Latham, N. A. McCutchen, and S. Boutin 2011. Invading white-tailed deer change wolf–caribou dynamics in northeastern Alberta. The Journal of Wildlife Management 75:204–212. Lendrum, P. E., C. R. Jr Anderson, K. L. Monteith, J. A. Jenks, and R. T. Bowyer 2013. Migrating mule deer: effects of anthropogenically altered landscapes. PLoS One 8:e64548. Liedvogel, M., S. � Akesson, and S. Bensch 2011. The genetics of migration on the move. Trends in Ecology & Evolution 26:561–569. Mantel, N. 1967. Detection of disease clustering and a generalized regression approach. Cancer Research 27:209–220. Miller, B. F., R. W. DeYoung, T. A. Campbell, B. R. Laseter, W. M. Ford, and K. V. Miller 2010. Fine-scale genetic and social structuring in a central Appalachian white-tailed deer herd. Journal of Mammalogy 91:681–689. Monteith, K. L., V. C. Bleich, T. R. Stephenson, B. M. Pierce, M. M. Conner, R. W. Klaver, and R. T. Bowyer 2011a. Timing of seasonal migration in mule deer: effects of climate, plant phenology, and lifehistory characteristics. Ecosphere 2:art47. Monteith, K. L., V. C. Bleich, T. R. Stephenson, B. M. Pierce, M. M. Conner, R. W. Klaver, and R. T. Bowyer. 2011b. Timing of seasonal migration in mule deer: effects of climate, plant phenology, and lifehistory characteristics. Ecosphere 2: art47. Monteith, K. L., T. R. Stephenson, V. C. Bleich, M. M. Conner, B. M. Pierce, and R. T. Bowyer 2013. Risk-sensitive allocation in seasonal dynamics of fat and protein reserves in a long-lived mammal. Journal of Animal Ecology 82:377–388. Nelson, M. E. 1993. Natal dispersal and gene flow in white-tailed deer in northeastern Minnesota. Journal of Mammalogy 74:316–322. Nelson, M. E. 1998. Development of migratory behavior in northern white- tailed deer. Canadian Journal of Zoology 76:426–432. Nielsen, S. E., A. B. A. Shafer, M. S. Boyce, and G. B. Stenhouse 2013. Does learning or instinct shape habitat selection? PLoS One 8: e53721. Nosil, P. 2012. Ecological Speciation. OUP, Oxford. Olano-Marin, J., J. C. Mueller, and B. Kempenaers 2011. Heterozygosity and survival in blue tits (Cyanistes caeruleus): contrasting effects of Genetic correlates in a wild cervid presumably functional and neutral loci. Molecular Ecology 20:4028– 4041. Pamilo, P., and S. Palsson 1998. Associative overdominance, heterozygosity and fitness. Heredity 81:381–389. Pease, K. M., A. H. Freedman, J. P. Pollinger, J. E. McCormack, W. Buermann, J. Rodzen, J. Banks et al. 2009. Landscape genetics of California mule deer (Odocoileus hemionus): the roles of ecological and historical factors in generating differentiation. Molecular Ecology 18:1848–1862. Perez-Gonzalez, J., J. Carranza, J. Torres-Porras, and J. L. FernandezGarcia 2010. Low heterozygosity at microsatellite markers in iberian red deer with small antlers. Journal of Heredity 101:553–561. Plummer, M. 2012. JAGS Version 3.3.0 User Manual. International Agency for Research on Cancer, Lyon, France. Plummer, M. 2013. rjags: Bayesian Graphical Models Using MCMC. R package version 3-10. http://CRAN.R-project.org/package=rjags (accessed on 10 June 2013). Posada, D., and K. A. Crandall 1998. MODELTEST: testing the model of DNA substitution. Bioinformatics 14:817–818. Post, E., and M. C. Forchhammer 2008. Climate change reduces reproductive success of an Arctic herbivore through trophic mismatch. Philosophical Transactions of the Royal Society B-Biological Sciences 363:2369–2375. Post, E., C. Pedersen, C. C. Wilmers, and M. C. Forchhammer 2008. Warming, plant phenology and the spatial dimension of trophic mismatch for large herbivores. Proceedings of the Royal Society B-Biological Sciences 275:2005–2013. Powell, J. H., S. T. Kalinowski, M. D. Higgs, M. R. Ebinger, N. V. Vu, and P. C. Cross 2013. Microsatellites indicate minimal barriers to mule deer Odocoileus hemionus dispersal across Montana, USA. Wildlife Biology 19:102–110. Pritchard, J. K., M. Stephens, and P. Donnelly 2000. Inference of population structure using multilocus genotype data. Genetics 155:945–959. Purdue, J. R., T. K. Oleksyk, and M. H. Smith 2006. Independent occurrences of multiple repeats in the control region of mitochondrial DNA of white-tailed deer. Journal of Heredity 97:235–243. R Core Team. 2013. R: A Language and Environment for Statistical Computing, Reference Index Version 3.0.1. R Foundation for Statistical Computing, Vienna, Austria. Ruegg, K., E. C. Anderson, J. Boone, J. Pouls, and T. B. Smith. 2014. A role for migration-linked genes and genomic islands in divergence of a songbird. Molecular Ecology, DOI: 10.1111/mec.12842. Sawyer, H., F. Lindzey, and D. McWhirter 2005. Mule deer and pronghorn migration in western Wyoming. Wildlife Society Bulletin 33:1266–1273. Sawyer, H., R. M. Nielson, F. Lindzey, and L. L. McDonald 2006. Winter habitat selection of mule deer before and during development of a natural gas field. Journal of Wildlife Management 70:396–403. Sawyer, H., M. J. Kauffman, R. M. Nielson, and J. S. Horne 2009. Identifying and prioritizing ungulate migration routes for landscape-level conservation. Ecological Applications 19:2016–2025. Sexton, J. P., S. B. Hangartner, and A. A. Hoffmann 2014. Genetic isolation by environment or distance: which pattern of gene flow is most common? Evolution 68:1–15. Shafer, A. B. A., and J. B. W. Wolf 2013. Widespread evidence for incipient ecological speciation: a meta-analysis of isolation-by-ecology. Ecology Letters 16:940–950. Shafer, A. B. A., S. E. Nielsen, J. M. Northrup, and G. B. Stenhouse 2014. Linking genotype, ecotype, and phenotype in an intensively managed larg carnivore. Evolutionary Applications 7:301–312. © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 947 �Genetic correlates in a wild cervid Northrup et al. Slate, J., L. E. B. Kruuk, T. C. Marshall, J. M. Pemberton, and T. H. Clutton-Brock 2000. Inbreeding depression influences lifetime breeding success in a wild population of red deer (Cervus elaphus). Proceedings of the Royal Society B-Biological Sciences 267:1657–1662. Slate, J., T. C. Van Stijn, R. M. Anderson, K. M. McEwan, N. J. Maqbool, H. C. Mathias, M. J. Bixley et al. 2002. A deer (subfamily Cervinae) genetic linkage map and the evolution of ruminant genomes. Genetics 160:1587–1597. Spiegelhalter, D. J., N. G. Best, B. R. Carlin, and A. van der Linde 2002. Bayesian measures of model complexity and fit. Journal of the Royal Statistical Society Series B-Statistical Methodology 64:583–616. Stephenson, T. R., K. J. Hundertmark, C. C. Schwartz, and V. Van Ballenberghe 1998. Predicting body fat and body mass in moose with ultrasonography. Canadian Journal of Zoology-Revue Canadienne De Zoologie 76:717–722. Stephenson, T. R., V. C. Bleich, B. M. Pierce, and G. P. Mulcahy 2002. Validation of mule deer body composition using in vivo and postmortem indices of nutritional condition. Wildlife Society Bulletin 30:557–564. Stone, S., V. Abkevich, D. L. Russell, R. Riley, K. Timms, T. Tran, D. Trem et al. 2006. TBC1D1 is a candidate for a severe obesity gene and evidence for a gene/gene interaction in obesity predisposition. Human Molecular Genetics 15:2709–2720. Swenson, N. G., and D. J. Howard 2005. Clustering of contact zones, hybrid zones, and phylogeographic breaks in North America. American Naturalist 166:581–591. Szulkin, M., N. Bierne, and P. David 2010. Heterozygosity-fitness correlations: a time for reappraisal. Evolution 64:1202–1217. Tamura, K., D. Peterson, N. Peterson, G. Stecher, M. Nei, and S. Kumar 2011. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Molecular Biology and Evolution 28:2731– 2739. Teacher, A. G. F., and D. J. Griffiths 2011. HapStar: automated haplotype network layout and visualization. Molecular Ecology Resources 11:151–153. Toews, D. P. L., M. Mandic, J. G. Richards, and D. E. Irwin 2014. Migration, mitochondria, and the yellow-rumped warbler. Evolution 68:241–255. Tollefson, T. N., L. A. Shipley, W. L. Myers, D. H. Keisler, and N. Dasgupta 2010. Influence of summer and autumn nutrition on body condition and reproduction in lactating mule deer. Journal of Wildlife Management 74:974–986. Torbit, S. C., L. H. Carpenter, D. M. Swift, and A. W. Alldredge 1985. Differential loss of fat and protein by mule deer during winter. Journal of Wildlife Management 49:80–85. Von Hardenberg, A., B. Bassano, M. Festa-Bianchet, G. Luikart, P. Lanfranchi, and D. Coltman 2007. Age-dependent genetic effects on a secondary sexual trait in male Alpine ibex, Capra ibex. Molecular Ecology 16:1969–1980. Wallmo, O. C. 1981. Mule and Black-Tailed Deer of North America. University of Nebraska Press, Lincoln, NE. 948 Wallmo, O. C., L. H. Carpenter, W. L. Regelin, R. B. Gill, and D. L. Baker 1977. Evaluation of deer habitat on a nutritional basis. Journal of Range Management 30:122–127. Supporting Information Additional Supporting Information may be found in the online version of this article: Appendix S1. PCR conditions. Appendix S2. Microsatellite diversity statistics Table 1. Microsatellite loci, number of individuals genotyped (N), number of alleles present at each loci (Na), observed heterozygosity (Ho), expected heterozygosity (He), and fixation index (F). Appendix S3. Bayesian model formulations. Appendix S4. Phylogenetic trees. Appendix S5. Supplemental results. Table 1. Identification (ID) numbers, ages, study area, and whether individual was captured in December, March or both for mule deer captured in the Piceance basin of Colorado. Table 2. DIC values for multi-level linear regression models on mule deer body mass, and multi-level beta regression models on mule deer body fat relative to MLH (multi-locus heterozygosity) or SLH (singlelocus heterozygosity). Table 3. Covariates, median coefficient (coeff.) values, and the probability (prob.) of either a negative or positive effect of the covariate on mule deer body mass and body fat estimated from multi-level linear or beta regression respectively. Table 4. Covariates, median coefficient (coeff.) values, and the probability (prob.) of either a negative or positive effect of the covariate on mule deer Spring migration termination date estimated from negative binomial regression model from mule deer captured in the Piceance basin, Colorado. Table 5. Covariates, median coefficient (coeff.) values, and the probability (prob.) of either a negative or positive effect of the covariate on mule deer Spring migration initiation date estimated from negative binomial regression model from mule deer captured in the Piceance basin, Colorado. Table 6. Covariates, median coefficient (coeff.) values, and the probability (prob.) of either a negative or positive effect of the covariate on mule deer Fall migration initiation date estimated from negative binomial regression model from mule deer captured in the Piceance basin, Colorado. Figure 1. Fitted values versus residuals from negative binomial model fit to migration timing of mule deer in the Piceance basin of Colorado. Figure 2. Fitted values versus residuals from negative binomial model fit to migration timing of mule deer in the Piceance basin of Colorado. The residuals were calculated from the model including mtDNA clades determined from the neighbor joining analysis. Figure 3. Fitted values versus residuals from hierarchical beta regression fit to percent body fat of mule deer in the Piceance basin of Colorado. © 2014 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd 7 (2014) 937–948 � https://cpw.cvlcollections.org/files/original/c8b40986b4f775d081d91501a5c1ee16.pdf bb0e3d57670c051d730b62a2da6ca7e5 PDF Text Text APPENDIX S1: PCR conditions The multiplex microsatellite reaction consisted of 25 ng DNA, 5 μL 2× Qiagen Multiplex mix, 2 μL primer mix, and 0.5 μL distilled water. The 10 μl single reactions contained 0.8 μL of MgCl2 (20 mM), 1 μL 10× PCR buffer, 2 μL of dNTPs (0.2 mM each), a 20× primer mix diluted to between 0.24 and 0.34 μL each, 0.08 μL of Taq (0.5 units), 1 μL of DNA template (~10 ng) and Milli-Q water. One primer per pair was fluorescently labelled. The multiplex PCR parameters followed Cullingham et al. (2011) and the single-PCRs began with an initial 3-minute denaturation at 95°C, followed by 38 cycles of 30 seconds denaturation at 94°C, 90 seconds annealing at 49°C, and 30 seconds extension at 72°C. The microsatellite amplicons were loaded on an ABI 3730 DNA sequencer (Applied Biosystems, Foster City, CA, USA) with a GS500LIZ size standard (Applied Biosystems). Microsatellite alleles were scored using GENEMAPPER version 4.0 (Applied Biosystems) and deviations from Hardy-Weinberg equilibrium (HWE) were tested using the exact test (Guo and Thompson 1992) implemented in Genepop v.4.0 (Rousset 2008)and FSTAT v.2.9.3 (Goudet 1995)was used to test for linkage disequilibrium. The mitochondrial control region was amplified in a 25 μl PCR reaction containing ~50 ng of template DNA, 0.2 mM each dNTP, 1× PCR buffer, 0.2 μM each primer, 1.6 mM MgCl2, 0.1 U Taq DNA polymerase, and Milli-Q water. The PCR profile was as follows: hot-start followed by an initial 2-minute denaturation at 94°C, followed by 38 cycles of 30 seconds denaturation at 94°C, 58°C, 72°C. The run concluded after 5 minutes at 72°C. PCR success was determined from gel electrophoresis. PCR product (10 μl) was treated with 5 μl of ExoSAP (USB Corporation, OH, USA) and incubated at 37°C for 15 minutes followed by 80°C for 15 minutes. The ExoSAP treated PCR product was used in a sequencing reaction. Amplicons were directly �sequenced in both directions using a Big Dye Terminator Kit (Applied Biosystems, Foster City, CA) and generated on an ABI 3730. Sequences were aligned using the ClustalW algorithm (Thompson et al. 1994). � https://cpw.cvlcollections.org/files/original/756a0fecdec8773a041d00e580978b54.pdf 7c57289a08f68dfe55dc174934ddea2c https://cpw.cvlcollections.org/files/original/e6c788ed6c9ab9d2e5caea6f90526293.pdf e885e81ea37d1183496bde9a8fccbd25 PDF Text Text APPENDIX S3: Microsatellite diversity statistics Table 1. Microsatellite loci, number of individuals genotyped (N), number of alleles present at each loci (Na), observed heterozygosity (Ho), expected heterozygosity (He), and fixation index (F). Locus N Na Ho He F INRA011a 134 6.000 0.500 0.489 -0.023 RT30a 132 14.000 0.788 0.794 0.008 BBJ2a 134 8.000 0.739 0.781 0.054 Ka 135 5.000 0.748 0.725 -0.032 BL25a 133 6.000 0.767 0.706 -0.086 BM6438a 134 10.000 0.784 0.732 -0.071 BM848a 135 9.000 0.748 0.755 0.010 RT7a 133 8.000 0.827 0.786 -0.052 Na 135 12.000 0.852 0.881 0.033 ETH152a 134 10.000 0.791 0.803 0.015 BM6506a 135 5.000 0.741 0.701 -0.056 Pa 132 7.000 0.538 0.550 0.021 Da 132 6.000 0.462 0.463 0.001 BM4107a 134 11.000 0.828 0.838 0.011 RT5a 134 10.000 0.836 0.777 -0.075 OCAMa 129 8.000 0.628 0.558 -0.125 Ra 131 6.000 0.634 0.619 -0.024 Mean 133.294 8.294 0.718 0.703 -0.023 SE 0.400 0.629 0.029 0.030 0.012 � https://cpw.cvlcollections.org/files/original/5864da9f56ed27240ab3e2b9031d2aa6.pdf b0004f3abcf67ded8549a40f1f7db121 PDF Text Text Appendix D: Phylogenetic trees �                                                                                                                            �                                                                                                                                                                                                                                      � https://cpw.cvlcollections.org/files/original/e7a0d1fce8344e3a89be5e0daf647cd9.pdf bbcbf4fe4f58349fa9a809fcdac7fcd2 PDF Text Text APPENDIX S5: supplemental results Table 1. Identification (ID) numbers, ages, study area, and whether individual was captured in December, March or both for mule deer captured in the Piceance basin of Colorado. ID number Age Study area Capture period 11890 4.5 SM December 102352 5.5 NM December 102353 9.5 NM December 102354 5.5 NM December 102357 10.5 NM December 102358 2.5 NM December 102363 5.5 NM December 102365 5.5 NM December 102368 3.5 NM December 102370 3.5 NM December 102371 10.5 NM December 102374 8.5 SM December 102375 2.5 SM December 102376 10.5 SM December 102377 5.5 SM December 102380 2.5 SM December �102384 7.5 SM December 102385 1.5 SM December 102386 4.5 SM December 102390 7.5 SM December 102391 4.5 SM December 102392 4.5 RG December 102396 7.5 RG December 102398 3.5 RG December 102407 9.5 RG December 102408 6.5 RG December 102409 4.5 RG December 102410 1.5 RG December 102411 4.5 RG December 102412 7.5 RG December 102413 4.5 RG December 102416 3.5 RG December 102417 6.5 RG December 102418 3.5 RG December �102419 5.5 RG December 102420 4.5 RG December 102421 10.5 RG December 102423 2.5 RG December 102426 7.5 RG December 102427 9.5 RG December 102429 7.5 RG December 102430 9.5 RG December 102431 10.5 RG December 102434 5.5 RG December 102435 4.5 RG December 102436 7.5 NR December 102442 7.5 NR December 102443 10.5 NR December 102444 4.5 NR December 102448 6.5 NR December 102449 3.5 NR December 102452 2.5 NR December �102455 3.5 NR December 102456 4.5 NR December 102457 8.5 NR December 11758 9.5 NR Both 11760 5.5 NR Both 11761 2.5 NR Both 11767 4.5 NR Both 11768 6.5 NR Both 11769 8.5 NR Both 11771 3.5 NR Both 11772 8.5 NR Both 11778 5.5 RG Both 11779 5.5 RG Both 11780 4.5 RG Both 11781 4.5 RG Both 11782 6.5 RG Both 11783 4.5 RG Both 11784 5.5 RG Both �11787 8.5 RG Both 11788 3.5 RG Both 11789 6.5 RG Both 11791 4.5 RG Both 11792 3.5 RG Both 11885 2.5 SM Both 11886 5.5 SM Both 11887 6.5 SM Both 11888 3.5 SM Both 11889 5.5 SM Both 11891 10.5 SM Both 11894 10.5 SM Both 11899 4.5 SM Both 11903 3.5 NM Both 11905 3.5 NM Both 11906 2.5 NM Both 11908 3.5 NM Both 11910 8.5 NM Both �11911 4.5 NM Both 11913 6.5 NM Both 11916 10.5 NM Both 11919 2.5 NM Both 11920 2.5 NM Both 102393 6.5 RG Both 102395 8.5 RG Both 102397 6.5 RG Both 102399 4.5 RG Both 102401 3.5 RG Both 102432 3.5 RG Both 102433 9.5 RG Both 102437 5.5 NR Both 11756 7.5 NR March 11759 7.5 NR March 11762 1.5 NR March 11764 7.5 NR March 11765 2.5 NR Both �11766 5.5 NR Both 11770 6.5 NR March 11773 6.5 NR March 11774 5.5 NR March 11775 3.5 NR March 11776 2.5 NR March 11777 11.5 RG March 11882 5.5 SM March 11883 7.5 SM March 11884 8.5 SM March 11892 4.5 SM March 11893 8.5 SM March 11895 10.5 SM March 11896 4.5 SM March 11897 3.5 SM March 11898 6.5 SM March 11900 7.5 SM March 11901 3.5 SM March �11902 7.5 NM March 11904 3.5 NM March 11907 7.5 NM March 11909 4.5 NM March 11912 6.5 NM March 11914 4.5 NM March 11915 8.5 NM March 11917 4.5 NM March 11918 8.5 NM March 11921 6.5 NM March �Table 2. DIC values for multi-level linear regression models on mule deer body mass, and multilevel beta regression models on mule deer body fat relative to MLH (multi-locus heterozygosity) or SLH (single-locus heterozygosity). Body mass and fat were calculated from deer captured via helicopter net-gunning on their winter range. Dependent Model Model structure DIC Mass M1 MLH + Age + March Capture + Study area -338.9 Mass M2 MLH + Age + Age2 + March Capture + Study area -339.2 Mass M3 SLH* + Age + March Capture + Study area -334.5 Mass M4 SLH* + Age + Age2 + March Capture + Study area -334.6 Fat F1 MLH + Age + March Capture + Study area -695.9 Fat F2 MLH + Age + Age2 + March Capture + Study area -695.8 Fat F3 SLH* + Age + March Capture + Study area -691.6 Fat F4 SLH* + Age + Age2 + March Capture + Study area -690.9 variable SLH indicates a set of 17 dummy variables indicating if the individual was heterozygous (1) or * not (0) at a specific locus. �Table 3. Covariates, median coefficient (coeff.) values, and the probability (prob.) of either a negative or positive effect of the covariate on mule deer body mass and body fat estimated from multi-level linear or beta regression respectively. Models presented are lowest DIC models for both MLH and SLH models of mass and body fat. Mass model M2 Covariate Median coeff. value Prob. coeff. is negative Prob. coeff. positive MLH 0.117 0.24 0.76 Age 0.036 0.03 0.97 Age2 -0.021 0.91 0.09 March Capture -0.1 1 0 NR* -0.014 0.61 0.39 RG† -0.02 0.67 0.33 SM‡ -0.007 0.56 0.44 Median coeff. value Prob. coeff. is negative Prob. coeff. positive Mass model M4 Covariate �Age 0.033 0.06 0.94 Age2 -0.021 0.89 0.11 March Capture -0.100 1 0 NR* -0.014 0.60 0.40 RG† -0.027 0.69 0.31 SM‡ -0.027 0.69 0.31 INRA011 -0.024 0.73 0.27 RT30 -0.060 0.90 0.10 BBJ 0.024 0.30 0.70 K 0.036 0.20 0.80 BL25 -0.004 0.53 0.47 BM6438 0.010 0.42 0.58 BM848 0.028 0.27 0.73 RT7 0.030 0.29 0.71 N 0.085 0.06 0.94 ETH152 -0.034 0.77 0.23 BM6506 0.036 0.21 0.79 �P 0.002 0.48 0.52 D 0.010 0.40 0.60 BM4107 0.044 0.21 0.79 RT5 0.032 0.27 0.73 OCAM 0.027 0.25 0.75 R -0.038 0.82 0.18 Covariate Median coeff. value Prob. coeff. is negative Prob. coeff. positive MLH 0.094 0.39 0.61 Age -0.043 0.87 0.13 March Capture -0.515 1 0 NR* -0.078 0.74 0.26 RG† -0.026 0.60 0.4 SM‡ 0.041 0.36 0.64 Fat model F1 Fat model F3 �Covariate Median coeff. value Prob. coeff. is negative Prob. coeff. positive Age -0.051 0.89 0.11 March Capture -0.518 1 0 NR* -0.104 0.80 0.20 RG† -0.105 0.82 0.18 SM‡ -0.059 0.69 0.31 INRA011 -0.127 0.93 0.07 RT30 -0.240 0.99 0.01 BBJ 0.077 0.22 0.78 K -0.034 0.65 0.35 BL25 0.074 0.27 0.73 BM6438 -0.001 0.50 0.50 BM848 -0.108 0.87 0.13 RT7 -0.078 0.72 0.28 N 0.087 0.22 0.78 ETH152 -0.004 0.52 0.48 BM6506 0.024 0.40 0.60 �P 0.175 0.04 0.96 D 0.092 0.13 0.87 BM4107 0.053 0.32 0.68 RT5 0.145 0.13 0.87 OCAM 0.020 0.41 0.59 R -0.076 0.81 0.19 Indicates deer captured in the NR study area, with NM as the reference category * † Indicates deer captured in the RG study area, with NM as the reference category ‡ Indicates deer captured in the SM study area, with NM as the reference category �Table 4. Covariates, median coefficient (coeff.) values, and the probability (prob.) of either a negative or positive effect of the covariate on mule deer Spring migration termination date estimated from negative binomial regression model from mule deer captured in the Piceance basin, Colorado. Neighbor joining clades Covariate Median coeff. value Prob. coeff. is negative Prob. coeff. positive Intercept 3.48 0 1 Age 0.01 0.45 0.55 NR* -0.23 0.93 0.07 RG† -0.01 0.52 0.48 SM‡ -0.02 0.56 0.44 mtDNA cluster 2§ 0.04 0.38 0.62 mtDNA cluster 3§ 0.04 0.39 0.61 Covariate Median coeff. value Prob. coeff. is negative Prob. coeff. positive Intercept 3.49 0 1 Bayesian clades �Age 0.01 0.44 0.56 NR* -0.23 0.93 0.07 RG† -0.02 0.55 0.45 SM‡ -0.03 0.58 0.42 mtDNA cluster 2§ 0.03 0.41 0.59 Indicates deer captured in the NR study area, with NM as the reference category * † Indicates deer captured in the RG study area, with NM as the reference category ‡ Indicates deer captured in the SM study area, with NM as the reference category mtDNA cluster 1 is the reference category § �Table 5. Covariates, median coefficient (coeff.) values, and the probability (prob.) of either a negative or positive effect of the covariate on mule deer Spring migration initiation date estimated from negative binomial regression model from mule deer captured in the Piceance basin, Colorado. Neighbor joining clades Covariate Median coeff. value Prob. coeff. is negative Prob. coeff. positive Intercept 3.41 0 1 Age -0.03 0.7 0.3 NR* -0.67 1 0 RG† -0.14 0.79 0.21 SM‡ -0.08 0.67 0.33 mtDNA cluster 2§ 0.01 0.48 0.52 mtDNA 3 -0.12 0.76 0.24 Covariate Median coeff. value Prob. coeff. is negative Prob. coeff. positive Intercept 3.36 0 1 Bayesian clades �Age -0.04 0.76 0.24 NR* -0.66 1 0 RG† -0.11 0.74 0.26 SM‡ -0.05 0.32 0.38 mtDNA cluster 2§ 0.04 0.38 0.62 Indicates deer captured in the NR study area, with NM as the reference category * † Indicates deer captured in the RG study area, with NM as the reference category ‡ Indicates deer captured in the SM study area, with NM as the reference category mtDNA cluster 1 is the reference category § �Table 6. Covariates, median coefficient (coeff.) values, and the probability (prob.) of either a negative or positive effect of the covariate on mule deer Fall migration initiation date estimated from negative binomial regression model from mule deer captured in the Piceance basin, Colorado. Neighbor joining clades Covariate Median coeff. value Prob. coeff. is negative Prob. coeff. positive Intercept 3.07 0 1 Age -0.17 0.92 0.08 NR* -0.1 0.62 0.38 RG† -0.44 0.91 0.09 SM‡ -0.78 0.98 0.02 mtDNA cluster 2§ -0.52 0.96 0.04 mtDNA cluster 3§ -0.55 0.95 0.05 Covariate Median coeff. value Prob. coeff. is negative Prob. coeff. positive Intercept 2.84 0 1 Bayesian clades �Age -0.18 0.93 0.07 NR* -0.08 0.59 0.41 RG† -0.28 0.81 0.19 SM‡ -0.59 0.95 0.05 mtDNA cluster 2§ -0.35 0.9 0.1 Indicates deer captured in the NR study area, with NM as the reference category * † Indicates deer captured in the RG study area, with NM as the reference category ‡ Indicates deer captured in the SM study area, with NM as the reference category mtDNA cluster 1 is the reference category § �Fig. 1. Fitted values versus residuals from negative binomial model fit to migration timing of mule deer in the Piceance basin of Colorado. The residuals were calculated from the model including mtDNA clades determined from the Bayesian analysis. �Fig. 2. Fitted values versus residuals from negative binomial model fit to migration timing of mule deer in the Piceance basin of Colorado. The residuals were calculated from the model including mtDNA clades determined from the neighbor joining analysis. �Fig. 3. Fitted values versus residuals from hierarchical beta regression fit to percent body fat of mule deer in the Piceance basin of Colorado. Residuals are from best model as determined by DIC (deviance information criteria). � https://cpw.cvlcollections.org/files/original/1c24b0f7fd262f5676ca63bb0bc580da.xls 5567bf75d11a54a447b390de1bd544b3 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Journal Articles Description An account of the resource CPW peer-reviewed journal publications Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Fine-scale genetic correlates to condition and migration in a wild cervid Rights Information about rights held in and over the resource <div class="element"> <div class="element-text"><a href="http://rightsstatements.org/vocab/InC-NC/1.0/" target="_blank" rel="noreferrer noopener">In Copyright - Non-Commercial Use Permitted</a></div> </div> <a href="https://creativecommons.org/licenses/by-nc/4.0/" target="_blank" rel="noreferrer noopener">Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)</a> Date Created Date of creation of the resource. 2014-08-28 Description An account of the resource The relationship between genetic variation and phenotypic traits is fundamental to the study and management of natural populations. Such relationships often are investigated by assessing correlations between phenotypic traits and heterozygosity or genetic differentiation. Using an extensive data set compiled from free-ranging mule deer (<em>Odocoileus hemionus</em>), we combined genetic and ecological data to (i) examine correlations between genetic differentiation and migration timing, (ii) screen for mitochondrial haplotypes associated with migration timing, and (iii) test whether nuclear heterozygosity was associated with condition. Migration was related to genetic differentiation (more closely related individuals migrated closer in time) and mitochondrial haplogroup. Body fat was related to heterozygosity at two nuclear loci (with antagonistic patterns), one of which is situated near a known fat metabolism gene in mammals. Despite being focused on a widespread panmictic species, these findings revealed a link between genetic variation and important phenotypes at a fine scale. We hypothesize that these correlations are either the result of mixing refugial lineages or differential mitochondrial haplotypes influencing energetics. The maintenance of phenotypic diversity will be critical to enable the potential tracking of changing climatic conditions, and these correlates highlight the need to consider evolutionary mechanisms in management, even in widely distributed panmictic species. Creator An entity primarily responsible for making the resource Northrup, Joseph M. Shafer, Aaron B. A. Anderson Jr, Charles R. Coltman, David W. Wittemyer, George Language A language of the resource English Is Part Of A related resource in which the described resource is physically or logically included. Evolutionary Applications Extent The size or duration of the resource. 12 pages Bibliographic Citation A bibliographic reference for the resource. Recommended practice is to include sufficient bibliographic detail to identify the resource as unambiguously as possible. Northrup, J. M., A. B. A. Shafer, C. R. Anderson Jr., D. W. Coltman, and G. Wittemyer. 2014. Fine–scale genetic correlates to condition and migration in a wild cervid. Evolutionary Applications 7:937–948. <a href="https://doi.org/10.1111/eva.12189" target="_blank" rel="noreferrer noopener">https://doi.org/10.1111/eva.12189</a> Source A related resource from which the described resource is derived <a href="https://datadryad.org/stash/dataset/doi:10.5061/dryad.3vc1b" target="_blank" rel="noreferrer noopener">Data from: Fine scale genetic correlates to condition and migration in a wild Cervid</a> Subject The topic of the resource Genetic differentiation Heterozygosity fitness correlation Migration Mule deer Multilocus heterozygosity <em>Odocoileus hemionus</em> Single-locus heterozygosity Wildlife Format The file format, physical medium, or dimensions of the resource application/pdf Type The nature or genre of the resource Article