^{1}

^{2}

^{1}

^{2}

^{3}

^{4}

^{1}

^{5}

^{6}

^{‡}

^{2}

^{6}

^{‡}

The authors have declared that no competing interests exist.

‡ These authors are joint senior authors on this work.

Geographically limited dispersal can shape genetic population structure and result in a correlation between genetic and geographic distance, commonly called isolation-by-distance. Despite the prevalence of isolation-by-distance in nature, to date few studies have empirically demonstrated the processes that generate this pattern, largely because few populations have direct measures of individual dispersal and pedigree information. Intensive, long-term demographic studies and exhaustive genomic surveys in the Florida Scrub-Jay (

Dispersal is a fundamental component of the life history of most organisms and therefore influences many biological processes. Dispersal is particularly important in creating genetic structure on the landscape. We often observe a pattern of decreased genetic relatedness between individuals as geographic distances increases, or isolation-by-distance. This pattern is particularly pronounced in organisms with extremely short dispersal distances. Despite the ubiquity of isolation-by-distance patterns in nature, there are few examples that explicitly demonstrate how limited dispersal influences spatial genetic structure. Here we investigate the processes that result in spatial genetic structure using the Florida Scrub-Jay, a bird with extremely limited dispersal behavior and extensive genome-wide data. We take advantage of the long-term monitoring of a contiguous population of Florida Scrub-Jays, which has resulted in a detailed pedigree and measurements of dispersal for hundreds of individuals. We show how limited dispersal results in close genealogical relatives living closer together geographically, which generates a strong pattern of isolation-by-distance at an extremely small spatial scale (<10 km) in just a few generations. Given the detailed dispersal, pedigree, and genomic data, we can achieve a fairly complete understanding of how dispersal shapes patterns of genetic diversity over short spatial scales.

The movement of individuals over the landscape (dispersal) influences biological processes and diversity at many levels [

Despite the fact that correlations between genetic differentiation and geographic distance are common across many types of organisms, to date, there are few existing empirical demonstrations of how contemporary patterns of dispersal generate spatial patterns of genetic variation and contribute to observed patterns of isolation-by-distance. This is, in part, because dispersal is hard to estimate empirically, as it requires monitoring many individuals over long periods of time across the full range of potential dispersal distances [

Patterns of isolation-by-distance can reflect underlying biological processes. Since the early development of isolation-by-distance theory, differences in mating systems and dispersal propensity have both been known to generate differences in isolation-by-distance patterns [

Here we examine the causes of fine-scale isolation-by-distance in a non-migratory bird, the Florida Scrub-Jay (

Previous work on Florida Scrub-Jays using microsatellite markers has shown isolation-by-distance across multiple populations [

We documented natal dispersal distances for 382 male and 290 female Florida Scrub-Jays that were born and established as breeders within the population at Archbold Biological Station between 1990–2013. Dispersal curves for both males and females were strongly leptokurtic, consistent with previous studies (^{−16}). Florida Scrub-Jays disperse extremely short distances compared with other bird species [

(A) Natal dispersal distances for Florida Scrub-Jays born and breeding within Archbold Biological Station (1990–2013, ^{−16}). Median values are shown with arrows at top of plot. Florida Scrub-Jay drawing by JWF. (B) Map of breeding territories (gray polygons) for a representative year (2008) within Archbold with individual breeders colored by PC1 values shows isolation-by-distance from north to south. (C) Isolation-by-distance patterns in autosomal SNPs shown with standard error bars. The decline in identity-by-descent with geographic distance is stronger in male-male (blue) and male-female (purple) pairwise comparisons than in female-female comparisons (salmon). δ values, the distance where identity-by-descent drops halfway to the mean (see text for details), are shown as dashes on the x-axis.

To explore the genetic implications of this limited, sex-biased dispersal, we genotyped all breeding adults in the Archbold population in 2003, 2008, and 2013 (

To test for isolation-by-distance, we quantified autosomal genetic relatedness between all possible pairs of individuals in the dataset as the estimated proportion of the genome shared identical-by-descent. Under a model of isolation-by-distance, the proportion of the genome shared identical-by-descent should decrease as the distance between individuals in a pair increases. Plotting genetic relatedness against geographic distance for all unique pairs across all years, we found a clear pattern of isolation-by-distance (

To measure the strength of isolation-by-distance in different subsets of the data, we fitted loess curves and used them to estimate the distance (δ) where the proportion of the genome shared identical-by-descent drops halfway to the mean from its maximum value. A lower δ indicates a more rapid decay of genetic relatedness by geographic distance,

Because of the detailed pedigree information available for the Florida Scrub-Jay population within Archbold, we have a rare opportunity to decompose the isolation-by-distance patterns found in this population by familial relationship. The Florida Scrub-Jay pedigree from our study population consists of 12,738 unique individuals over 14 generations and is largely complete (see

Identity-by-descent in autosomal SNPs versus distance for all possible (A) male-male, (B) male-female, and (C) female-female comparisons is, in part, generated by highly related individuals remaining physically close together. Loess curves are shown in each panel. Isolation-by-distance patterns are significantly stronger in (A) male-male and (B) male-female comparisons than in (C) female-female comparisons. Points are colored by specific pedigree relationship or, for more distant relationships, grouped into a single coefficient of relationship (

Another way of visualizing how dispersal generates the observed pattern of isolation-by-distance is to plot the distribution of geographic distances separating pairs of individuals with different pedigree relationships (

(A) Distances between all possible male-male (blue), male-female (purple), and female-female (salmon) comparisons separated by pedigree relationship. Significant differences using the Kolmogorov-Smirnov Test are indicated with two asterisks when all three comparisons were significantly different (MM-FF, MM-MF, MF-FF) and a single asterisk when only MM-FF and MM-MF comparisons were significantly different. The distance between parent-offspring pairs is significantly shorter than the distance between full-siblings (Wilcoxon rank sum test, ^{−9}) and the distance between full-siblings is significantly shorter than the distance between pairs with

We can further assess the contribution of various relationship types by sequentially removing pedigree relationship classes and observing the resulting isolation-by-distance curves (

Patterns of genetic diversity on the Z chromosome are expected to differ from those on the autosomes because of the difference in inheritance patterns and sex-specific dispersal behavior [

We separately assessed patterns of isolation-by-distance in 277 Z-linked SNPs. PCA results for Z-linked markers are similar to those observed in autosomes. We found significant correlations for PC1 and PC2 with UTM northing, though correlations between PC2 and UTM northing were significant only for all breeders and male only comparisons (

Similar to autosomal SNPs, isolation-by-distance patterns in Z-linked SNPs are stronger in male-male comparisons (δ = 615 m, [592, 639]) than in either female-female (δ = 979 m, [673, 2048]) or male-female comparisons (δ = 637 m, [601, 674];

Unbiased identity-by-descent for autosomal (black) and Z-linked (gray) SNPs versus geographic distance for all possible unique pairwise comparisons showing higher mean identity-by-descent in Z-linked SNPs (0.014) than in autosomal SNPs (0.0027). Here we use untransformed estimates of identity-by-descent to avoid biases introduced by the different numbers of autosomal and Z-linked SNPs (see text for details). Identity-by-descent values are binned across 10 distance quantiles and shown as mean ± SE. δ values, the distance where identity-by-descent drops halfway to the mean (see text for details), are shown as dashes on the x-axis.

To test our understanding of the population mechanisms leading to fine-scale isolation-by-distance, we used simulations to explore whether observed patterns could be predicted strictly by dispersal curves and other population parameters. We first conducted simulations of local dispersal in a contiguous population to determine how well the observed distribution of geographic distances between individuals of known pedigree relationships was predicted by the observed natal dispersal curves. Assuming that the dispersal curves are constant and that dispersal distance has negligible heritability, we simulated the distance between individuals of a known, close pedigree relationship using random draws from the sex-specific dispersal curves. For example, for two female first cousins, we first simulated the dispersal distances of the parental siblings from the grandparental nest (randomly picking their sexes). We then simulated dispersal distances of the two female cousins from their respective parental nests and calculated the distance (

(A) An example schematic of a dispersal simulation for two female first cousins. Our simulations were over a two-dimensional space, but here we show dispersal on a one-dimensional line for visualization purposes. For the dispersal simulations, we start with the most recent common ancestor for a pair of individuals of known pedigree relationship and simulate dispersal events forward in time until the present. In this case, we start at the grandparental nest, simulate dispersal distances (and angles) of the parents, and then dispersal of the two cousins. Light blue birds are females and dark blue are males. Arrows indicate male (blue) and female (salmon) dispersal events drawn from the dispersal curves. In most simulations, sexes of all ancestors are randomly determined. (B) The gametic kinship chain from Malécot’s theory of isolation-by-distance. A locus from individual _{n} are identical-by-descent if both are descended from the same locus in their common ancestor in generation _{n-p.} Under Malécot’s model, genetic relatedness of individuals should decrease as the distance between them increases. Redrawn from [_{1} and trace their ancestral lineages backwards in time until we either reach a common ancestor or one of the ancestors was an immigrant. In each generation, the probability a given pair coalesces is sampled directly from the pedigree.

We found that the dispersal simulations generally nicely reconstruct the observed distribution of geographic distances between related individuals up to second cousins (

Simulated (gray line) and observed (colored histograms) dispersal values for full-sibling, aunt/uncle-nibling, first cousin, and second cousin comparisons. Male-male comparisons are shown in blue, male-female comparisons in purple, and female-female comparisons in salmon. Median values for the simulated (gray) and observed (colored) distributions are indicated by arrows above each plot. Simulated distributions that were significantly different from the observed distribution using the Kolmogorov-Smirnov Test are marked with asterisks above the median arrows. See

The distance simulations described above randomized the sexes for all ancestral individuals and therefore averaged across all possible lineages for a given pedigree relationship. However, given the strongly sex-biased dispersal in the Florida Scrub-Jay, we expect the geographic distance between a given pair of individuals to also depend on the sexes of the ancestors. For example, two females can be cousins because their mothers are siblings (four female dispersal events), their mother and father are siblings (three female and one male dispersal events), or because their fathers are siblings (two female and two male dispersal events).

To assess the relationship between the sex of the ancestors and geographic distance between a pair of individuals of a given pedigree relationship, we conducted additional simulations of first cousins in which we fixed the sexes for the two common ancestors (aunts or uncles) in addition to the focal individuals (the cousins). As predicted, we found that the median geographic distance between two cousins strongly correlates with the number of female dispersal events in the lineage (Spearman rank correlation: ρ = 0.8208,

In nature, we know that dispersal movements are largely restricted to the bounded area that is the study population. Because our natal dispersal curves include only within-population dispersal events, we do not think a violation of this assumption is problematic for simulations of closely related pairs, which involve just a few dispersal events. To accurately simulate distances between more distantly related pairs, we would need to consider the spatial extent of the population and not allow dispersal movements outside of population boundaries.

Malécot envisioned identity-by-descent as being due to the chain of ancestry running from present day individuals back to their shared ancestors (“les chaînes de parenté gamétique”;

To empirically demonstrate the underlying mechanisms behind Malécot’s model, we calculated the expected identity-by-descent values as a function of geographic distance for male-male, male-female, and female-female pairs using a spatially-explicit coalescent model. We parameterized these simulations using the observed pedigree, dispersal curves, immigration rate, and basic demographic information about the study system. Our simulations extended Malécot’s framework to include immigration from other populations because previous work has demonstrated a non-negligible rate of immigration into our study population [_{k} apart in generation _{k} apart who are full-siblings, half-siblings, or parent-offspring pairs (see

We ran five different simulations to investigate how increasing the complexity of the model improved our fit to the observed isolation-by-distance patterns in male-male, male-female, and female-female pairs (^{2} = 0.90), but not for male-male and female-female comparisons (^{2} = 0.61 and -0.10, respectively; ^{2} = 0.88–0.90 for model M3;

Simulated (black crosses) and observed (colored circles and line) autosomal isolation-by-distance patterns for male-male (blue), male-female (purple), and female-female comparisons (salmon). We ran five different simulations using the observed pedigree, dispersal curves, and immigration rate. Results are shown for two models: the simplest model with no sex-specific parameters (M0) on top and our final model with sex-specific parameters and isolation-by-distance in immigrants (M4) on bottom. By increasing the biological realism of our models, we can recover the observed pattern of isolation-by-distance. The coefficient of determination for the final model is 0.98 for male-male comparisons, 0.96 for male-female comparisons, and 0.78 for female-female comparisons. See

Model number | Model description | Autosomal | Z-linked | ||||
---|---|---|---|---|---|---|---|

MM | MF | FF | MM | MF | FF | ||

M0 | No sex-specific parameters | 0.61 (0.60) | 0.90 (0.88) | -0.10 (0.06) | 0.46 (0.44) | 0.65 (0.61) | 0.16 (0.12) |

M1 | M0 + sex-specific dispersal | 0.64 (0.64) | 0.92 (0.91) | 0 (0.10) | 0.48 (0.47) | 0.66 (0.64) | 0.16 (0.10) |

M2 | M1 + sex-specific relatedness | 0.86 (0.86) | 0.90 (0.88) | 0.84 (0.66) | 0.68 (0.68) | 0.66 (0.65) | 0.16 (0.13) |

M3 | M2 + sex-specific immigration | 0.88 (0.88) | 0.90 (0.91) | 0.88 (0.73) | 0.61 (0.58) | 0.63 (0.58) | 0.22 (0.17) |

M4 | M3 + isolation-by-distance in immigrants | 0.98 (0.94) | 0.96 (0.97) | 0.78 (0.63) | 0.89 (0.85) | 0.93 (0.91) | 0.41 (0.37) |

Values listed are from models with parameter estimates from the full dataset. Mean values from five cross-validation runs are included in parentheses. MM = male-male pairs, MF = male-female pairs, FF = female-female pairs. See

A number of studies have compared direct estimates of dispersal obtained from field observations to indirect estimates of dispersal obtained by regressing pairwise genetic differentiation on geographic distance [

Here we have used single-marker estimates of genome-wide identity-by-descent to study relatedness. Additional power to infer recent demography and dispersal history can be gained by studying shared identity-by-descent blocks—linked segments of the genome that are shared identical-by-descent between pairs of individuals [

Isolation-by-distance is a commonly observed pattern in nature. Despite its ubiquity and the frequent use of isolation-by-distance patterns to indirectly estimate dispersal in diverse organisms, few studies to date have deconstructed the causes of isolation-by-distance. Here, we have shown how limited dispersal can result in isolation-by-distance in the Florida Scrub-Jay. The extremely short dispersal distances of this species allow us to detect a signal of isolation-by-distance within a single, small contiguous population over just a few generations. In systems with longer dispersal distances, patterns of isolation-by-distance will likely only be observed over larger spatial scales, and reflect relatedness over potentially much longer timescales. The extensive dispersal, pedigree, and genomic data in this well-studied system provided a rare opportunity to empirically unpack Malécot’s isolation-by-distance model [

The Florida Scrub-Jay is a cooperatively breeding bird endemic to Florida oak scrub habitat [

Here, we measured dispersal distances of individuals banded as nestlings within Archbold and that subsequently bred within Archbold between 1990 and 2013 (382 males and 290 females). We began our sampling in 1990 because the study site was expanded to its current size by 1990; hence, dispersal measures before this year are systematically shorter (

As part of a previous study, 3,984 individuals have been genotyped at 15,416 genome-wide SNPs using Illumina iSelect Beadchips [

To determine genetic relatedness, we estimated the proportion of the genome shared identical-by-descent relative to the population frequency for all individual pairwise comparisons within and across years using the ‘genome’ option in PLINK v1.07 [

Additionally, we estimated relatedness of all individual pairwise comparisons using the pedigree. We calculated the coefficient of relationship by using the ‘kinship’ function within the package kinship2 [

We used three approaches to test for isolation-by-distance patterns in our data. First, we conducted principal component analysis on the autosomal and Z-linked genomic data using custom Perl and R scripts. We conducted separate analyses on males only, females only, and all individuals. We then compared the first two PC axes from each analysis with the UTM northing values of the territory centroids for each individual using Spearman rank correlations. To ensure these patterns were not driven by differences in genetic diversity within the study site, we estimated observed heterozygosity and inbreeding coefficients (^{III} from [

Second, we conducted Mantel correlogram tests using the ‘mantel.correlog’ function in the vegan package [

Finally, we fitted a loess curve to the scatterplot of identity-by-descent and geographic distance between pairs of individuals. We tested for isolation-by-distance by determining whether identity-by-descent at the smallest distance interval was larger than the overall mean. To measure the strength of isolation-by-distance, we estimated the distance where identity-by-descent drops halfway to the mean from its maximum value, which we define as δ. To assess uncertainty in these estimates, we used a bootstrapping method in which we randomly resampled pairs with replacement, fitted a loess curve, and estimated identity-by-descent at distance bin 0, mean identity-by-descent, and δ. We repeated this procedure 1,000 times to obtain 95% bias-corrected and accelerated bootstrap confidence intervals.

We used simulations to determine whether we could generate the observed distribution of geographic distances between related pairs using only the natal dispersal curve. For each of several focal pairwise relationships (full-siblings, aunt/uncle-nibling, first cousins, and second cousins), we simulated dispersal events starting at their common ancestral nest and then recorded the resulting distance between the two focal individuals using a custom script in R (

We generated the expected isolation-by-distance pattern for the autosomes and the Z chromosome given the observed dispersal curves and immigration rate using spatially-explicit pedigree-based simulations that are extensions of Malécot’s model of isolation-by-distance [_{g} between the two individuals. Here, we assumed that there is no genetic variation for dispersal distance, and sampled a dispersal distance for all individuals of a given sex from the same distribution. After the first dispersal event, we randomly assigned sexes for all ancestors. We then calculated the probability that the two lineages located at distances (_{1}, …, _{g}) did not coalesce (share a common ancestor) or have an immigrant ancestor in the previous _{p}(_{f}(_{h}(_{p}), E(_{f}), and E(_{h}) are the expected identity-by-descent values for parent-offspring, full-sibling, and half-sibling pairs, respectively (_{m}), and E(_{r}) are the sex-specific empirical mean identity-by-descent values for immigrant-immigrant and immigrant-resident pairs, respectively. Because we found a pattern of isolation-by-distance in immigrant-immigrant pairs, we used expected identity-by-descent values for immigrant-immigrant and immigrant-resident pairs conditional on distance. We binned distances into 15 quantiles and ran 1,000 simulations for each distance bin. To evaluate the fit of our model, we calculated the coefficient of determination ^{2} for each type of comparison as follows:
_{i} is the mean observed identity-by-descent value in distance bin ^{2} when the model performs so poorly that the mean of the data provides a better fit than our model. We ran simulations using parameters estimated from the full dataset, and then performed two-fold cross-validation to check for over-fitting. As results from both sets of models were similar, we discuss results from the full dataset in the text. See

(TXT)

(TXT)

(R)

(R)

(R)

(PDF)

(PDF)

(PDF)

Map of breeding territories (gray polygons) for a representative year (2008) within Archbold Biological Station with individual breeders colored by PC2 values.

(TIF)

PC score for PCAs conducted on the autosomal SNPs versus UTM northing of the territory centroid for all breeders, males only, and females only. Significant comparisons are shown with a blue linear regression line. See

(TIF)

PC score for PCAs conducted on the Z-linked SNPs versus UTM northing of the territory centroid for all breeders, males only, and females only. Significant comparisons are shown with a blue linear regression line. See

(TIF)

Results are shown for all possible (A) male-male, (B) male-female, and (C) female-female comparisons. Loess curves are shown in each panel. Isolation-by-distance patterns are significantly stronger in male-male (A) comparisons than in either male-female (B) or female-female (C) comparisons. Points are colored by specific pedigree relationship or grouped into a single coefficient of relationship class. Gray points indicate no known pedigree relationship. Note that the scale of the y-axis differs between panels. Pedigree relationship abbreviations: PO = parent-offspring, FS = full-siblings, G1 = grandparent-grandchild, HS = half-siblings, N1 = aunt/uncle-nibling, C1 = first cousins, DC1 = double first cousins. (“Nibling” is a gender-neutral term for niece and nephew).

(TIF)

(A) The probability two randomly sampled alleles coalesce in the previous generation for autosomal (black) and Z-linked (gray) markers at different distance bins estimated directly from the population pedigree. The probability two Z-linked alleles coalesce in the previous generation is higher than the probability two autosomal alleles coalesce only at short distances. (B) Unbiased identity-by-descent for autosomal (black) and Z-linked (gray) SNPs versus geographic distance for all possible unique pairwise comparisons, separated into immigrant-immigrant and immigrant-resident pairs (solid line) versus resident-resident pairs (dotted line). Pairs containing an immigrant have higher observed levels of identity-by-descent for Z-linked markers compared to autosomal markers. Here we use untransformed estimates of identity-by-descent to avoid biases introduced by the different numbers of autosomal and Z-linked SNPs. Identity-by-descent values are binned across 10 distance quantiles and shown as mean ± SE. See

(TIF)

Simulated (blue crosses) and observed (circles and lines) isolation-by-distance patterns for autosomal (black) and Z-linked (gray) markers. Results are shown for three models: the full autosomal model (left), a model with all Z parameters except for identity-by-descent values for immigrants (middle), and the full Z model (right). Substituting in Z values for immigrant identity-by-descent greatly improves the fit of the model, suggesting that the observed pattern of isolation-by-distance on the Z is largely driven by immigration. Untransformed identity-by-descent values are binned across 10 distance quantiles and shown as mean ± SE. See

(TIF)

(A) Simulated isolation-by-distance patterns for autosomal (black and blue) and Z-linked (gray and salmon) markers under two different dispersal scenarios. Simulations with sex-biased dispersal are shown as solid lines and simulations with equal dispersal between sexes are shown as dashed lines. Female-biased dispersal slightly elevates predicted identity-by-descent at short distance bins for both types of markers, with a larger effect for Z-linked markers. (B) Simulated isolation-by-distance patterns for four different models: the full autosomal model (black), a model with Z parameters but autosomal coalescent probabilities (blue), a model with Z parameters but autosomal coalescent probabilities divided by 0.75 to simulate the smaller effective population size of the Z (orange), and the full Z-linked model (gray). Both the smaller effective population size of the Z and sex-biased dispersal contribute to higher identity-by-descent for Z-linked markers compared to autosomal markers at short distances. We use autosomal identity-by-descent values for immigrants in all simulations here to generate predictions for autosomal and Z differences without immigration. See

(TIF)

Simulated (gray line) and observed (colored histograms) dispersal values for sex-specific first cousin dispersal simulations. Median values for the simulated (gray) and observed (colored) distributions are indicated by arrows above each plot. Simulated distributions that were significantly different from the observed distribution using the Kolmogorov-Smirnov Test are marked with asterisks above the median arrows.

(TIF)

Proportion of various pedigree relationship classes within 500 m distance bins for all unique individual comparisons with (A) including individuals unrelated in the pedigree (NA, gray bars) and (B) excluding individuals unrelated in the pedigree. Colors indicate pedigree relationship and match those in

(TIF)

Simulated (black crosses) and observed (colored circles and line) identity-by-descent values are indicated for male-male (blue), male-female (purple), and female-female comparisons (salmon). The coefficient of determination for each model can be found in

(TIF)

Simulated (black crosses) and observed (colored circles and line) identity-by-descent values are indicated for male-male (blue), male-female (purple), and female-female comparisons (salmon). The coefficient of determination for each model can be found in

(TIF)

Black line indicates a 1:1 relationship, solid blue line shows the linear regression of all of the data, and the dashed blue line shows the linear regression with

(TIF)

Results are shown for all possible (A) male-male, (B) male-female, and (C) female-female comparisons. Loess curves are shown in each panel. Points are grouped by color into autosomal identity-by-descent classes. Gray points indicate no known pedigree relationship.

(TIF)

Results are shown for all possible (A) male-male, (B) male-female, and (C) female-female comparisons. Loess curves are shown in each panel. Points are grouped by color into autosomal identity-by-descent classes. Gray points indicate no known pedigree relationship.

(TIF)

Spearman rank correlations comparing the first two PC axes and the UTM northing value of the territory centroid. Each cell contains the Spearman’s ρ value followed by the

(DOCX)

Mantel correlogram results for all unique individual comparisons of all breeders, male breeders only, and female breeders only across all years. Each cell contains the Mantel correlation coefficient followed by the corrected

(DOCX)

Counts of different known relationship pairs in the full pedigree of the Archbold population of Florida Scrub-Jays. The expected coefficient of relationship for each pair of relatives is also included. Note that the actual pedigree-based coefficient of relationship for a given pair may be higher than the expected value due to inbreeding in the population. Pedigree relationship abbreviations: PO = parent-offspring, FS = full-siblings, HS = half-siblings, G1 = grandparent-grandchild, N1 = aunt/uncle-nibling, DC1 = double first cousins, C1 = first cousins, G2 = great-grandparent-great-grandchild, GN = great-aunt/uncle-grand-nibling, N2 = half-aunt/uncle-half-nibling, C0.5 = half-cousins, C1r1 = first cousins once removed, G3 = great-great-grandparent-great-great-grandchild, C2 = second cousins, C2r1 = second cousins once removed, C3 = third cousins, C4 = fourth cousins.

(DOCX)

Wilcoxon rank sum test results comparing the distribution of distances between pairs in different pedigree relationship classes. Sequential comparisons were performed between pairs with pedigree coefficient of relationship (

(DOCX)

Wilcoxon rank sum test results comparing the distribution of distances between all individuals within specific pedigree relationship classes (as shown in boxplots in

(DOCX)

Tests of autosomal isolation-by-distance (measured as the difference between identity-by-descent (IBD) at distance bin 0 and the overall mean IBD) and the strength of isolation-by-distance (measured by δ) for different subsets of the data. Mean values are listed with 95% CI in parentheses.

(DOCX)

Kolmogorov-Smirnov (KS) and Wilcoxon rank sum (WRS) test results comparing simulated and observed distributions and medians, respectively, between individuals within specific pedigree relationship classes (as shown in

(DOCX)

Kolmogorov-Smirnov (KS) and Wilcoxon rank sum (WRS) test results comparing simulated and observed distributions and medians, respectively, between sex-specific first cousin comparisons (as shown in

(DOCX)

Expected identity-by-descent values are the coefficient of relationship for different pairs. For the expected Z-linked identity-by-descent of male-female parent-offspring pairs, we averaged that of father-daughter and mother-son pairs. For the expected Z-linked identity-by-descent of half-siblings, we averaged that of paternal half-siblings and maternal half-siblings. See [

(DOCX)

We thank the many students, interns, and staff at Archbold Biological Station and the University of Memphis who collected the dispersal data. We thank the members of the Harrison, Clark, Lovette, and Coop labs and anonymous reviewers for thoughtful comments on this manuscript. In memory of Richard G. Harrison without whom this collaboration would not have been formed.