The Different Sources of Variation in Inbreeding Depression, Heterosis and Outbreeding Depression in a Metapopulation of Physa Acuta

The Different Sources of Variation in Inbreeding Depression, Heterosis and Outbreeding Depression in a Metapopulation of Physa acuta

Juan Sebastia´n Escobar,1 Antoine Nicot2 and Patrice David3 Centre d’Ecologie Fonctionnelle et Evolutive UMR 5175, 34293 Montpellier, France Manuscript received June 17, 2008 Accepted for publication September 6, 2008

ABSTRACT Understanding how parental distance affects offspring ﬁtness, i.e., the effects of inbreeding and outbreeding in natural populations, is a major goal in evolutionary biology. While inbreeding is often associated with ﬁtness reduction (inbreeding depression), interpopulation outcrossing may have either positive (heterosis) or negative (outbreeding depression) effects. Within a metapopulation, all phenomena may occur with various intensities depending on the focal population (especially its effective size) and the trait studied. However, little is known about interpopulation variation at this scale. We here examine variation in inbreeding depression, heterosis, and outbreeding depression on life-history traits across a full-life cycle, within a metapopulation of the hermaphroditic snail Physa acuta. We show that all three phenomena can co-occur at this scale, although they are not always expressed on the same traits. A large variation in inbreeding depression, heterosis, and outbreeding depression is observed among local populations. We provide evidence that, as expected from theory, small and isolated populations enjoy higher heterosis upon outcrossing than do large, open populations. These results emphasize the need for an integrated theory accounting for the effects of both deleterious mutations and genetic incompatibilities within metapopulations and to take into account the variability of the focal population to understand the genetic consequences of inbreeding and outbreeding at this scale.

HE importance of inbreeding and outbreeding on known as inbreeding depression (Charlesworth T fitness has been recognized for a long time and Charlesworth 1987, 1999). Inbreeding depres- (Darwin 1876; Dobzhansky 1936; Wright 1937; sion has been observed in most plant and animal Crow 1948) and has more recently reclaimed impor- species (Husband and Schemske 1996; Crnokrak tance on theoretical (Lynch 1991; Schierup and and Roff 1999; Keller and Waller 2002) and has Christiansen 1996; Merila¨ and Sheldon 1999; been usually attributed to the expression of recessive Bierne et al. 2002; Edmands and Timmerman 2003; lethal and sublethal mutations maintained in appre- Shpak 2005) and conservational (Frankham 1999; ciable frequency in large outbred populations (Mukai Quilichini et al. 2001; Marr et al. 2002; Tallmon et al. et al. 1974; Deng et al. 1998; Keller and Waller 2002). 2004; Aspi et al. 2006; Willi et al. 2007) grounds. It is To a lesser extent, overdominance (Charlesworth and widely assumed that there is some optimal genetic Charlesworth 1999; Li et al. 2001) and epistasis (Lynch distance (or degree of relatedness) between parents, 1991; Lynch and Walsh 1998) may also contribute to the above and below which offspring fitness decreases low performance in inbred individuals. (Price and Waser 1979; Waser 1993). However, little The same category of mutations is thought to be can be predicted about this optimum, essentially responsible for heterosis, i.e., the higher fitness in in- because the two sides are explained by different terpopulation hybrids compared to offspring of within- theories and assumed to result from different catego- population crosses (Lynch and Walsh 1998). Heterosis ries of mutations. seems to be mainly produced by the presence of com- Within populations, mating between relatives (in- plementary sets of deleterious recessive alleles within breeding) is often associated with reduction in perfor- both parental populations and the masking of their ef- mance when compared to outbreeding, a phenomenon fects in the F1 heterozygotes (Lynch and Walsh 1998; Remington and O’Malley 2000). Such alleles generally 1Present address: INRA–Centre de Montpellier, UMR Diversiteét have small effects on fitness; hence they can approach Adaptation des Plantes Cultiveés, 34130 Mauguio, France. fixation by random genetic drift in some isolated pop- 2Present address: Institut de Recherche pour le De´veloppement, GEMI ulations (Kimura et al. 1963; Lynch 1991; Whitlock UMR 2724 CNRS–IRD, 34394 Montpellier, France. et al. 2000). As for inbreeding depression, overdomi- 3Corresponding author: Centre d’Ecologie Fonctionnelle et Evolutive UMR 5175, 1919 Route de Mende, Campus CNRS, 34293 Montpellier nance and epistatic interactions may also be involved in Cedex 5, France. E-mail: [email protected] heterosis (Lynch 1991; Lynch and Walsh 1998).

Genetics 180: 1593–1608 (November 2008) 1594 J. S. Escobar, A. Nicot and P. David

However, hybridization between divergent popula- of either A or a alleles within demes (while migration tions may eventually result in a loss in fitness, a phe- among demes tends to restore intermediate frequen- nomenon known as outbreeding depression (Lynch cies). In small demes, the effect of random genetic drift 1991; Edmands 1999), considered as a first step in the adds up to that of selection and favors even more rapid way to speciation. Such a decline in fitness may be divergence and fixation of alternative alleles. Thus, apparent in F1 hybrids, which have a complete haploid interpopulation hybridization should lead to stronger set of genes from each parent, but it is usually consid- outbreeding depression in small demes than in larger ered more likely to be observed in F2 and later ones. This principle remains valid with other types of generations, when deleterious interactions among re- genetic incompatibilities than underdominant loci (e.g., cessive alleles at different loci get exposed to selection epistasis). However, because heterosis and outbreeding due to recombination (Lynch 1991; Edmands 1999, depression both increase when population size is small, 2007). On the other hand, outbreeding depression can they may partly cancel each other, and thus the overall appear in the F1 as a result of the disruption of local effect of population size is not easy to predict. adaptation (gene 3 environment interactions), under- The expected effects of decreasing selection intensity dominance, or epistatic interactions such as the are similar, to a large extent, to those of decreasing breakup of favorable additive 3 additive epistatic effects population size. This is because in both cases selection (Burton 1990; Lynch 1991; Edmands 1999), all of becomes weaker relative to genetic drift. Therefore, the which can be subsumed under the term ‘‘genetic consequences of inbreeding and outbreeding may not incompatibilities.’’ Unlike the alleles responsible for be the same depending on the trait under consideration inbreeding depression and heterosis (which are un- and are likely to vary among traits expressed at different conditionally deleterious), alleles creating genetic in- life stages. For example, selection is expected to be compatibilities become deleterious only when they weaker on late-expressed traits (Hamilton 1966; Rose interact with a foreign genetic background. In among- 1991). Another possible origin of variation among traits population crosses, either heterosis or outbreeding is that some traits are more mutable than others: for depression, depending on the pairs of populations con- example, many semilethal mutations on basic functions sidered, has been observed in natural populations of will not be expressed further than early life stages plants (Waser and Price 1994; Trame et al. 1995; Byers because mutant individuals die early. Early-expressed 1998; Fenster and Galloway 2000; Waser et al. 2000; traits therefore appear as larger ‘‘mutational targets’’ Pe´labon et al. 2005; Willi and Van Buskirk 2005) and than late-expressed traits. animals (Burton 1990; Edmands 1999; Gharrett et al. Most studies of the effects of inbreeding and out- 1999; Aspi 2000; Andersen et al. 2002; Gilk et al. 2004). breeding (or, more generally, parental divergence) on Many parameters determine the expected levels of offspring fitness in natural populations are performed inbreeding depression, heterosis, and outbreeding de- by crossing a set of individuals from a single population pression in natural populations, including effective to increasingly distant mates. However, given the in- population size, selection coefficients, dominance co- trinsically stochastic nature of genetic drift and the efficients, migration rates, and mutation rates (Whitlock process of fixation of deleterious mutations, the varia- et al. 2000). Among these, the effective population size is tion in the effective size, in the degree of isolation, and of great importance since genetic drift will be greatest in the intensity of selection among populations, it is not when the effective size of local populations is smallest clear to which extent the observed patterns (e.g., (Hartl and Clark 1997), which increases the likeli- heterosis or outbreeding depression at a certain dis- hood of fixation of alleles (Kimura et al. 1963; Bataillon tance) are general or specific to the focal population and Kirkpatrick 2000). For this reason, one expects and traits. For example, even in a metapopulation where smaller populations to exhibit reduced inbreeding on average there is neither heterosis nor outbreeding depression compared to larger ones (Bataillon and depression, different populations may have different Kirkpatrick 2000; Gle´min 2003). On the other hand, average breeding values for a given trait. Such traits will since genetic drift and mutation cause allele frequen- tend to increase upon outcrossing (heterosis) if the focal cies to vary among local populations, smaller popula- population is below average and will tend to decrease tion size leads to a higher level of genetic differentiation (outbreeding depression) if it is above average. We there- among populations and hence increases heterosis fore need more studies to understand the amount of caused by deleterious recessives (Whitlock et al. 2000). variation of the effects of inbreeding and outbreeding For the same reason, small population sizes may in- among demes in a metapopulation. crease outbreeding depression due to genetic incom- In this article we examine how the effects of in- patibilities. Consider, for instance, an underdominant breeding and outbreeding vary (and covary) among locus with alleles A and a, with fitness AA . aa . Aa.In traits expressed at different life stages, and among this case, outbreeding depression is due to the increased populations, in the hermaphroditic self-fertile freshwa- heterozygosity in interpopulation F1 compared to within- ter snail Physa acuta. These populations are taken within population crosses. Natural selection favors the fixation a few kilometers in a single metapopulation, harboring Inbreeding and Outbreeding in a Metapopulation 1595

TABLE 1 Origin, coordinates, and neutral-genetic variation parameters for the studied populations

Habitat Population Coordinates N alleles Allelic richness Hexp Hobs Fis s River Buz 43.46 N 3.59 E 4.0 6 1.6 3.53 6 1.32 0.56 6 0.18 0.51 6 0.24 0.086 6 0.097* 0.004, NS River Lam 43.47 N 3.43 E 4.1 6 2.0 3.59 6 1.46 0.59 6 0.14 0.53 6 0.12 0.075 6 0.054, NS 0.005, NS River Lez 43.43 N 3.49 E 6.2 6 2.7 3.99 6 1.40 0.60 6 0.22 0.50 6 0.23 0.139 6 0.062*** 0.009, NS River Mos 43.40 N 3.46 E 6.2 6 2.9 4.42 6 1.84 0.64 6 0.17 0.55 6 0.23 0.111 6 0.084** 0.03, NS River Sal 43.41 N 3.55 E 5.6 6 2.8 4.43 6 2.02 0.58 6 0.26 0.53 6 0.21 0.087 6 0.061* 0.08, NS Pond Vio 01 43.44 N 3.42 E 5.2 6 2.3 4.07 6 1.82 0.60 6 0.21 0.44 6 0.28 0.203 6 0.090*** 0.30** Pond Vio 02 43.45 N 3.43 E 2.9 6 0.8 2.28 6 1.13 0.54 6 0.17 0.39 6 0.22 0.246 6 0.100** 0.26, NS Pond Vio 07 43.45 N 3.46 E 5.3 6 2.3 4.00 6 1.49 0.55 6 0.21 0.49 6 0.20 0.087 6 0.076* 0.008, NS Pond Vio 11 43.46 N 3.47 E 3.4 6 1.2 2.94 6 0.89 0.48 6 0.21 0.54 6 0.25 0.119 6 0.058, NS 0.03, NS Pond Vio 12 43.46 N 3.48 E 3.6 6 2.1 3.01 6 1.84 0.42 6 0.23 0.37 6 0.21 0.130 6 0.072* 0.20, NS All rivers 5.2 6 1.1 3.99 6 0.43 0.59 6 0.03 0.52 6 0.02 0.11 6 0.03 0.002 All ponds 4.1 6 1.1 3.26 6 0.76 0.52 6 0.07 0.45 6 0.07 0.14 6 0.17 0.17

N alleles, number of alleles; Hexp, gene diversity; Hobs, observed heterozygosity; s, selﬁng rate. *P , 0.05; **P , 0.01; ***P , 0.001; NS, P . 0.05.

different local habitats characterized by different ex- brought alive into the laboratory. A subsample from each pected population sizes and degrees of isolation. We use population (30 individuals) was preserved in 70% ethanol controlled crosses and molecular analyses to ask several for subsequent molecular analyses. In the laboratory, G0 individuals spent 1 week in high-density conditions (20–30 questions: (i) Do inbreeding depression, heterosis, individuals per 3-liter aquariums) to ensure cross-fertilization. and/or outbreeding depression occur simultaneously Behavioral and molecular evidence suggests that self-fertiliza- at the same geographic scale?; (ii) If so, do they affect in tion is very unlikely whenever mates are abundant ( Jarne et al. the same way different traits expressed early and late in 2000; Bousset et al. 2004; Henry et al. 2005; David et al. 2007). the life-cycle?; and (iii) Do local populations differ in After this, 24–90 individuals per population were isolated in 75-ml plastic boxes and reproduction was checked every the amount of inbreeding depression and heterosis or 2 days. Additional snails, not related to experimental families, outbreeding depression, and are these differences re- were kept as large stock colonies (one colony per population) lated to local population characteristics represented by in optimal conditions to serve as mating partners later. the habitat type? After some mortality occurred, 13–40 families per population were created by isolating egg capsules from G0 parents in 75-ml boxes. The G1 offspring of a given clutch were raised MATERIALS AND METHODS together for 2 weeks. At 15 days, density was standardized (five individuals per box) and at 22 days, prior to sexual maturity, G1 Studied species and sampled populations: The studied offspring of each family were isolated (one individual per species, P. acuta, is a simultaneously hermaphroditic freshwa- box). This experimental protocol avoids mating among ter snail exhibiting both high outcrossing rates (80–100%, relatives while reducing the number of boxes to handle. At Jarne et al. 2000; Henry et al. 2005) and high inbreeding 30 days, offspring from each family were randomly split into depression (78–90%, Jarne et al. 2000; Escobar et al. 2007). It three treatments: selfing and intra- and interpopulation is an oviparous species laying egg capsules, typically containing outcrossing. a few tens of eggs. Hatching occurs within 6–7 days after egg Selfers (N ¼ 503) were kept isolated throughout the laying, and sexual maturity is attained within 6–8 weeks at experiment until they reproduced. Intrapopulation outcross- 22°24° in laboratory conditions. ers (N ¼ 401) were given adult mates, randomly chosen from Ten populations of P. acuta were sampled within 25 km the stock colony corresponding to their own population. around Montpellier, southern France, in October–November Interpopulation outcrossers (N ¼ 872) were given adult mates, 2005 (Table 1). Five of these were from isolated ponds and 5 randomly chosen from a stock colony corresponding to a from rivers. Pond populations are assumed to have lower different population. For this treatment, each population was effective size and to be more isolated from the rest of the crossed with at least two different populations, one pond and metapopulation than are river populations. A previous study one river, following the crossing scheme shown in Figure 1. performed on 24 French populations of P. acuta has provided Four types of crosses (female 3 male) were therefore created: support for this assumption by finding more variation at pond 3 pond, pond 3 river, river 3 pond, and river 3 river. In microsatellite loci (number of alleles and gene diversity) in both intra- and interpopulation outcrossing treatments, each rivers than in ponds and higher Fst among ponds than G1 individual was mated with at least three different randomly among rivers (Bousset et al. 2004). This is in line with the chosen adult partners by periods of 48 hr each. To identify fact that rivers are open habitats harboring large-sized pop- partners, we marked them with a dot of varnish on the shell ulations that experience few bottlenecks, whereas ponds are (Henry and Jarne 2007). Eggs obtained during mating closed habitats with small-sized populations subject to fre- periods were discarded because of uncertainty in maternity. quent bottlenecks (Bousset et al. 2004; Henry et al. 2005). Only eggs obtained after the removal of partners were con- Experimental setting: Wild mature individuals (G0) were served for subsequent measures. Reproduction in all treat- sampled in each population (100–200 individuals) and ments was checked three times a week. 1596 J. S. Escobar, A. Nicot and P. David

three times a week. Late survival of marked snails was then quantified as the age at death minus 49 days. This measure is thought to represent the reproductive life span (RLS) because the first reproduction occurred at 46–49 days for most G2 individuals (72%), constrained by the fact that we provided the mates only at this time. Every 2 weeks, marked G2 snails were isolated in 75-ml boxes during 72 hr to quantify their fecundity. At that moment, snails were measured by means of digital pictures, as described above. Selfed and intra- and interpopulation outcrossed 0- to 15- day survival was recorded from 60,396 G2 eggs; 15- to 22- and 22- to 49-day survivals were estimated from 5837 and 2322 individuals, respectively, and RLS was estimated in 1253 individuals. Fecundity was estimated from 1801 individuals. Size was measured in 5150 individuals at 22 days and in 1117 individuals at 150 days. Molecular analyses: DNA was extracted using the DNeasy blood and tissue kit (QIAGEN, Valencia, CA) from 100 mg of 30 G0 individuals per population (N ¼ 276), maintained in 70% ethylic alcohol, and eluted in 200 ml of sterilized water. Ten loci were used for the PCR amplification: Pac1, Pac2, and Pac5 (Sourrouille et al. 2003); locus 27, locus 32-B, and locus 59-B (Monsutti and Perrin 1999); and Pasu1-02, Pasu1- 09, Pasu1-11, and Pasu1-12 ( J. Goudet, personal communi- cation). The loci used were multiplexed in three subsets for igure F 1.—Crossing scheme for interpopulation hybridiza- genotype analysis: (i) Pac1, Pac2, Pac5, Pasu1-12, and locus 27; tions. Each population was crossed with at least one river and (ii) locus 32-B and locus 59-B; and (iii) Pasu1-02, Pasu1-09, and one pond population. Arrows are from females to males. Pasu1-11. PCRs were conducted in a Mastercycler thermocy- Thirty-three types of interpopulation crosses were performed cler (Eppendorf) using 10.0-ml final volumes, including 0.2 mm (note that the female Mos 3 male Lez cross could not be per- of each primer and 10 ng of genomic DNA, using the QIAGEN formed). multiplex PCR kit. PCR conditions were as follows: a 15-min activation of the HotStartTaq DNA polymerase at 95°; 30 cycles including Snails were maintained at 25° under a 12 hr:12 hr light:dark 30 sec initial denaturation at 94°, 90 sec annealing, and 60 photoperiod and fed with boiled lettuce. Food and water were sec extension at 72°; followed by a 30-min final extension at changed twice a week. 60°. For genotyping, 3.0 ml of diluted multiplexed amplicon Traits measured: G2 eggs (mean 6 SD: 78 6 37 eggs per were pooled with 13.0 ml of deionized formamide and 0.2 ml mother) were counted and isolated in 75-ml boxes. An GeneScan-500 LIZ size standard and analyzed on an ABI estimate of early survival was obtained as the number of living Prism 3100 genetic analyzer. G2 individuals 15 days after the egg laying divided by the Statistical analyses: G2 individuals are characterized by their number of eggs in the clutch (hereafter, 0- to 15-day survival). treatment (selfing and intra- and interpopulation outcross- As in the previous generation, density was adjusted at 15 days ing), maternal habitat (pond, river), maternal population (five individuals per box) and individuals were isolated at (nested within maternal habitat), and family (nested within 22 days. Survival between 15 and 22 days was computed. Size maternal population). In addition to the above factors, (height from the base to the apex of the shell) was measured in interpopulation-hybrid characterization includes the paternal G2 individuals at isolation (22 days) by means of digital pictures habitat (pond, river) and paternal population (nested within (Nikon coolpix 5900) analyzed with the ImageJ 1.34s software paternal habitat). asband (R 2007). At 45 days, virgin G2 individuals were pair Estimates of survival (0–15, 15–22, and 22–49 days, and mated with a single adult partner from the stock colony RLS) were combined into a single cumulative measure (re- corresponding to the original maternal population during productive life expectancy, RLE) representing the expected 24 hr. As above, partners were marked with a dot of varnish on number of days during which an individual is able to lay eggs. the shell and clutches laid throughout the mating period were RLE was calculated as the product of survival rates from 0 to discarded. At 49 days, fecundity was measured as the number of 49 days and RLS. Fecundity (number of eggs per day of eggs laid after the removal of the partner (i.e., through 72 hr). reproductive life) was estimated by dividing the total number Additionally, we computed 22- to 49-day survival. of eggs a G2 individual laid during all 72-hr isolation periods At 50 days, 1–26 G2 individuals (mean 6 SD: 6 6 4 before dying and the number of days in isolation. Finally, the individuals) per family and population were individually expected lifetime reproductive success was calculated as the marked using a numbered, colored plastic tag for honeybees product of RLE and fecundity. (Ickowicz Apiculture, Bolle`ne, France) and a dot of varnish on Deviance analyses were performed on survival and fecun- the shell (Henry and Jarne 2007). This double marking dity measures with R 2.4.0 (R Development Core Team provided unique combinations for .1200 snails. Tagged 2006). Survival up to 49 days was analyzed by fitting individuals (N ¼ 1253) were randomized and grouped in generalized linear models (GLM), using a binomial error 3-liter plastic aquariums at constant density (100 individuals distribution (logit-link function). Data on RLS (ages at per aquarium). Density was adjusted within aquariums every death minus 49 days, because all individuals were 49 days week, using when necessary unmarked nonexperimental old when measures started) were analyzed using a GLM with individuals randomly taken from stock colonies. Individuals Gamma error distribution (inverse-link function), which is were randomized across aquaria every week to avoid common- convenient to analyze survival data without censoring environment effects. Mortality within aquaria was checked (Crawley 2005). Fecundity per day of reproductive life Inbreeding and Outbreeding in a Metapopulation 1597 was analyzed by a GLM using Poisson error distribution (log- (David et al. 2007). The RMES software estimates selfing rates link function). on the basis of the distribution of multilocus heterozygosity; Deviance analyses were performed in two subsets: (i) selfed these estimates, unlike those derived from Fis,havebeenshown and intrapopulation outcrossed performance and (ii) intra- to be insensitive to technical artifacts such as null alleles (David and interpopulation outcrossed performance, to test for in- et al. 2007). Differences in allelic richness and expected breeding depression and heterosis (or outbreeding depres- heterozygosity between ponds and rivers were calculated, and sion), respectively. In both cases analyses started with a significance of the test was assessed using a scheme of 1000 complete model including all main effects: treatment, mater- permutations (keeping the number of samples in each habitat nal habitat, maternal population (nested within maternal type constant) with FSTAT 2.9.3 (Goudet 1995). habitat), family (nested within maternal population), and all Components of phenotypic variation for life-history traits interactions. Then, terms were sequentially deleted, starting were estimated among and within populations, as well as from higher-order interactions. To account for overdispersion, between habitats, using data from the intrapopulation out- deviance ratios were computed to determine whether dropping crossing treatment. ANOVA were computed to estimate a term from the model significantly reduced the explained variance components. Analyses included habitat, population, deviance. These ratios are approximately F-distributed and are and family as random effects. Survival and fecundity measures equivalent to F-values in ordinary ANOVA (Crawley 2005). were log-transformed, whereas size measures were not. Vari- Size measures were analyzed separately. We fitted individual ance components served to calculate quantitative indexes of Bertalanffy’s growth curves, using phenotypic divergence (Spitze 1993) among populations within habitats (Qst) and between habitats (Qct), using kðt22daysÞ Lt ¼ L‘ ðL‘ L22Þe ; ð1Þ 2 spopulation where L represents the mean maximum size, L represents Qst ¼ ð4Þ ‘ 22 ðs2 1 s2 Þ the size at 22 days (i.e., our first measure), and k is a relative population 2 families growth rate. Only individuals with four or more size measures and were taken into account for this analysis. Growth parameters were estimated using maximum-likelihood nonlinear fits s2 performed with Mathematica 5.1 (Wolfram Research). Q ¼ habitats : ð5Þ ct s2 1 s2 1 s2 We computed size at 150 days (L150) using Equation 1. Both ð habitats population 2 familiesÞ L22 and L150 were analyzed with mixed-model ANOVA using the same factors as above. Treatment and habitat were We additionally performed deviance analyses to distinguish analyzed as fixed effects, whereas population, family, and all between additive and nonadditive effects on hybrid fitness, as interactions containing these two factors were considered as well as between symmetric and asymmetric nonadditive effects, random effects. using data from both intra- and interpopulation outcrossing Survival, fecundity, and size traits were log-transformed to treatments. We constructed an additive design matrix contain- express values in both a relative and an additive scale. The ing the number of genes (0, 1, or 2) contributed by each of the difference in the logarithm of fitness traits is a natural and 10 populations to each cross (for example, in the cross female statistically well-behaved measure of inbreeding depression A 3 male B, A and B contribute one gene each; while in the and heterosis, in contrast to the traditional ratio measures 3 ohnston choen intrapopulation cross A A, population A contributes two genes (e.g.,J and S 1994). Log-inbreeding depres- and population B contributes none). We then fitted GLMs on (i) sion was estimated for each character, family, population, and the above-described matrix of additive effects (in this way fitted maternal habitat using the estimator values of interpopulation hybrids are forced to equal the midparent value); (ii) the identity of both parental populations, ðdÞ¼ ð Þ ð Þ; ð Þ ln ln Wintra ln Wself 2 irrespective of which population acted as mother or father, i.e., A 3 B ¼ B 3 A (one value is fitted for each cross except that where Wintra corresponds to the mean fitness of intrapopula- reciprocal crosses are forced equal); and (iii) maternal 3 tion-outcrossed offspring and Wself holds for the mean fitness of selfed offspring. Note that Equation 2 is not a simple paternal interaction in which we did distinguish reciprocal logarithmic transformation of the classic estimator of in- crosses (i.e., A 3 B 6¼ B 3 A). The first fit indicates additive effects breeding depression, but a measure of the inbreeding load on hybrid fitness, the second model includes symmetric non- (Charlesworth and Charlesworth 1987; Willis 1999). additive effects, and the third term indicates asymmetric non- Likewise, log heterosis was estimated using additive effects (e.g., maternal effects). Significance of the terms was obtained by means of deviance ratios, as described above. Finally, three types of correlation analyses were performed: lnðHÞ¼lnðWinterÞlnðWintraÞ; ð3Þ (i) between inbreeding depression, heterosis, and neutral where Winter holds for the mean fitness of interpopulation diversity parameters (allelic richness and gene diversity); (ii) hybrids. Note that log-transformed estimates can be back between population genetic divergence (calculated as lnðX Þ transformed to recover classic estimates using Y ¼ 1 e ; Fst=½1 Fst;Rousset 1997) and interpopulation hybrid per- Y represents either inbreeding depression (d) or heterosis formance and heterosis; and (iii) between inbreeding de- (H). For survival traits, in some cases W ¼ 0 (no egg or pression and heterosis for all measured traits. Tests for individual survived) and logarithmic transformation was associations were performed using Pearson’s r and Spearman’s impossible. To avoid biasing the data set by dropping these r with JMP 7.0 (SAS Institute, Cary, NC). values, we used the transformation ln(0.01 1 W) instead of ln(W)(Escobar et al. 2007). Microsatellite data served to estimate neutral variation within populations and between habitat types, fixation indexes RESULTS (within and among populations and between habitat types, Timing of inbreeding depression, heterosis, and namely Fis, Fst, and Fct, respectively), and intrapopulation selfing rates. Molecular data were analyzed with Arlequin 3.11 outbreeding depression: The magnitude of inbreeding (Excoffier et al. 2005), FSTAT 2.9.3 (Goudet 1995), and RMES depression, heterosis, and outbreeding depression 1598 J. S. Escobar, A. Nicot and P. David markedly changed across the life cycle. We detected

strong inbreeding depression early in the life cycle F (mean value of 1 ½Wself =Wintra 6 SEM over populations [range]: 0- to 15-day survival, 0.56 6 0.06 [0.15– 0.77]), no inbreeding depression at middle-age stages

(15–22 days, 0.01 6 0.03 [0.28–0.12]; 22–49 days, 0.08 MSQ 6 0.04 [0.13–0.30]), and signiﬁcant inbreeding de-

pression for adult survival (RLS: 0.27 6 0.05 [0.01– with the corresponding 0.64]) (Table 2; Figure 2A). When combining survival at F all stages (RLE), we detected a very strong inbreeding depression (0.71 6 0.05 [0.38–0.87]), mostly due to the high value at 0–15 days (Figure 2A). Inbreeding depression on fecundity was also signiﬁcant (0.29 6 0.06 MSQ [0.01–0.52]) (Table 2; Figure 2A). Our cumulative

fitness measure (the lifetime reproductive success), F representing the number of eggs an individual lays across a full-life cycle, exhibited a very strong inbreeding depression (0.79 6 0.05 [0.45–0.91]) (Figure 2A). Size measures exhibited slight, though significant inbreed- Dev ing depression (22 days, 0.08 6 0.03 [0.07–0.24]; 150 days, 0.09 6 0.02 [0.02–0.16]) (Table 2; Figure 2A). On the other hand, significantly negative heterosis F (i.e., outbreeding depression) for survival was detected in the two earliest age classes (mean value of 1

½Wintra=Winter 6 SEM over populations [range]: 0–15 days, Dev 0.29 6 0.12 [1.03–0.18]; 15–22 days, 0.06 6 0.03 [0.32–0.02]), no heterosis at 22–49 days (0.004 6 0.06 [ 0.45–0.19]) and significant and positive heterosis in F adult survival (RLS: 0.13 6 0.07 [0.17–0.34]) (Table 3; Figure 2B). Combining all survival estimates, we de- TABLE 2 tected negative heterosis on RLE (0.20 6 0.22 [1.96– 0.39]) (Figure 2B). In contrast, significant positive heterosis was detected in fecundity (0.15 6 0.08 Dev [0.29–0.47]). Remarkably, when combining RLE and fecundity, we found neither heterosis nor outbreeding depression on the lifetime reproductive success (0.02 6 0.18 [1.41–0.59]) (Figure 2B). Size at 22 days exhibited F no significant heterosis (0.05 6 0.04 [0.12–0.19]), and the heterosis observed on size at 150 days was very low 0.05. Note that for size at 150 days, there was no selfed progeny from population Vio 12. Therefore, the degrees of ., change in deviance between models with and without that factor); MSQ, mean square; RLS, reproductive life span. The Inbreeding depression for survival, fecundity and size measures . i.e though significant (0.03 6 0.02 [–0.04–0.11]) (Table 3; P

Figure 2B). Dev Neutral and phenotypic divergence among populations: We ﬁrst analyzed the neutral genetic variation 0.05; NS,

within populations and we found that rivers exhibited , F more neutral variation than ponds (Table 1): differ- P (1; 892)(8; 884) (1; 618) (8; 610) (1; 584) (8; 576) (1; 580) (8; 572) (1; 580) (8; 572) (1; 70) (8; 62) (1; 10) (7; 55) (1; 902)(9; 893) (1; 628) (9; 619) (1; 594) (9; 585) (1; 590) (9; 581) (1; 590) (9; 581) (1; 70) (8; 62) (1; 10) (7; 10) (80; 638) (58; 400) (55; 371) (45; 389) (45; 389) (55; 1298) (44; 182) ences in gene diversity were signiﬁcantly greater in (166; 718) (152; 458) (150; 426) (138; 434) (138; 434) (55; 55) (44; 44) rivers than in ponds (one-sided P-value ¼ 0.006) and 0.01; * nearly signiﬁcant for the allelic richness (one-sided P- , P

value ¼ 0.061). The multilocus selfing rate was signifi- 0- to 15-day survival 15- to 22-day survival 22- to 49-day survival RLS Fecundity Size 22 days Size 150 days cant in one pond population (Vio 01, s ¼ 0.30), whereas other pond populations and all river populations exhibited selfing rates not significantly different from 0.001; ** , zero (Table 1). Analysis of genetic and geographical P treatment 54.0 3.23, NS 2.9 1.51, NS 2.2 1.44, NS 0.2 0.63, NS 15.1 3.35, NS 0.024 1.52, NS 0.001 0.07, NS divergence among populations did not provide evi- treatment 1678.9 2.23*** 114.5 1.23, NS 94.9 1.28, NS 15.2 1.20, NS 254.0 1.30, NS 0.022 4.49*** 0.015 1.16, NS 3 treatment 679.0 5.27*** 20.4 1.33, NS 15.7 1.31, NS 4.2 1.83, NS 45.9 1.28, NS 0.009 0.48, NS 0.015 1.06, NS dence for isolation by distance (Mantel test, Z ¼ 81.74, 3 3 r ¼0.18, P ¼ 0.73). Genetic divergence (F ) among all st Dev, deviance explained by the factor considered ( -values. *** freedom associated with population are 7 instead of 8. F treatment factor has two modalities: selfed and intrapopulation outcrossed. Numerator and denominator degrees of freedom are given in parentheses Pop. FamilyFamily 6550.8 3.69*** 413.9 1.65*** 266.0 1.27* 37.9 0.96, NS 630.9 1.02, NS 0.022 0.97, NS 0.022 1.51, NS Source Dev populations was 0.12 6 0.01. It was stronger among TreatmentPopulationHabitat 2584.0 144.72*** 1147.0 7.61*** 1.9 148.4 0.89, NS 8.55*** 3.7 101.6 2.27, NS 7.53*** 10.9 18.4 34.58*** 168.8 7.12*** 36.26*** 118.4 0.101 2.91*** 6.37* 0.062 3.47** 0.033 22.01*** 0.101 4.63* Inbreeding and Outbreeding in a Metapopulation 1599 F MSQ lation refers to the F MSQ

Figure 2.—Inbreeding depression and heterosis for life-his- F tory traits. Left y-axes, the log-inbreeding depression (ln ID in A), and the log heterosis (ln H in B); right y-axes, classic inbreeding depression (A) and heterosis (B). Shaded bars, mean

values over the two habitat types; open bars, means for ponds; Dev and solid bars, means for rivers. Error bars were calculated as SEM over populations. RLS, reproductive life span; RLE, reproductive life expectancy; Fec., fecundity; LTRS, lifetime re- F productive success; L22, size at 22 days; L150, size at 150 days. Note that ln ID and ln H (left y-axes) do not represent a simple logarithmic transformation of values presented in the right y-axes (see materials and methods for details). Dev pond populations (0.20 6 0.07) than among river 6

populations (0.08 0.02). The genetic differentiation F between the two habitat types (pond vs. river) was quite low, though significant (AMOVA F : 0.02 6 0.03, P , ct TABLE 3 0.05) (Table 4). We then analyzed population variance for life-history traits using the intrapopulation outcrossing treatment Dev alone. We found that the habitat type significantly affected both early survival and RLS: rivers exhibited on average slightly greater 0- to 15-day survival than ponds (0.58 6 0.07 and 0.49 6 0.06, respectively) and F Heterosis for survival, fecundity, and size measures much longer RLS (51.03 6 3.19 and 37.57 6 1.24 days, respectively) (Table 5). Among-population variation (within habitats) was significant for all measured traits,

whereas within-population (among-family) variation Dev was signiﬁcant only for 0- to 15- and 15- to 22-day survival and the two size measures (Table 5). Among- and within-population variances for life-history traits served to calculate the quantitative index of population F (1; 1261)(8; 1253)(73; 984) (1; 982) (8; 974) (62; 722) (1; 932) (8; 924) (1; 983) (62; 674) (8; 975) (1; (47; 981) 748) (8; 973) (47; 746) (1; 73) (8; 68) (62; 1807) (1; 17) (8; 48) (44; 288) divergence (Qst) between habitat types and to compare (1; 1271)(8; 1263)(1; 1262) (1; 992) (8; 984) (1; 983) (1; 942) (8; 934) (1; 933) (1; 993) (8; 985) (1; 984) (1; 991) (8; 983) (1; 982) (1; 74) (8; 68) (1; 73) (1; 17) (8; 7) (1; 17) (196; 1057) (190; 784) (188; 736) (180; 795) (180; 793) (62; 62) (44; 44) this measure with the among-population ﬁxation index on neutral markers (Fst). On the one hand, rivers had

mean Qst almost always greater than mean Fst (the only 0- to 15-day survival 15- to 22-day survival 22- to 49-day survival RLS Fecundity Size 22 days Size 150 days exception corresponds to size at 22 days). On the other hand, ponds exhibited mean Qst lower than mean Fst for 0- to 15- and 15- to 22-day survival, RLS, and size at 22 days and Qst . Fst for 22- to 49-day survival, fecundity, treatment 106.0 5.67* 7.1 3.57, NS 0.6 0.43, NS 3.8 16.01*** 0.2 0.06, NS 0.020 0.81, NS 0.032 1.43, NS treatment 2891.0 3.06*** 96.9 0.90, NS 113.6 1.40* 10.2 0.97, NS 161.6 1.21, NS 0.035 5.64*** 0.017 1.27, NS and size at 150 days (Table 4). Finally, Fct was almost 3 treatment 612.0 4.18*** 14.0 0.88, NS 15.0 1.34, NS 5.3 2.85** 153.6 6.22*** 0.012 0.43, NS 0.028 1.71, NS always .Q , with the exception of RLS, where Q ? F 3

ct ct ct 3 (Table 4), consistent with the large difference observed Deﬁnitions are the same as in Table 2. Treatment now includes two modalities: within-population outcrossing and interpopulation outcrossing. Popu maternal populations (common to both treatments). Source Dev Pop. FamilyFamily 7317.0 2.52*** 579.6 1.76*** 291.1 1.14, NS 47.5 1.18*** 723.6 1.40** 0.032 0.92, NS 0.016 0.95, NS between the two habitats for this trait. HabitatPopulationTreatmentHabitat 1891 46 12.42*** 2.26, 356.0 NS 237.1 18.98*** 2.6 14.73*** 21.2 1.16, NS 158.7 10.64** 14.21*** 8.1 11.5 0.5 5.86*** 5.22* 0.36, NS 497.9 19.24*** 7.4 5.1 0.043 31.17*** 20.09*** 1.62, NS 24.7 0.167 2.7 7.69** 0.72, 6.02* NS 0.053 0.033 2.12, 1.42, NS NS 0.150 0.054 0.56, NS 6.70* 1600 J. S. Escobar, A. Nicot and P. David

TABLE 4 Genetic variances associated with the maternal habitat type on measured life-history traits

0- to 15- to 22- to Variance 15-day 22-day 49-day Size 22 Size 150 Mean Mean AMOVA Habitat component survival survival survival RLS Fecundity days days Fst Qst Fct Pond Interpopulation 0.050 0.024 0.634 1.905 5.141 0.0008 0.005 Intrapopulation 1.334 0.075 0.334 42.748 3.838 0.002 0.001 Qst 0.018 0.138 0.487 0.022 0.401 0.159 0.689 0.20 6 0.07 0.27 6 0.26 River Interpopulation 0.194 0.142 0.323 34.348 0.453 0.0003 0.003 Intrapopulation 0.296 0.129 0.283 89.322 0.735 0.002 0.003 Qst 0.247 0.355 0.363 0.161 0.236 0.064 0.335 0.08 6 0.02 0.25 6 0.11 Qct 0.020 0.060 0.030 0.410 0.100 0.020 0.110 0.02 6 0.03 Variance components were calculated using the intrapopulation outcrossing treatment alone. Survival and fecundity variances were calculated after logarithmic transformation, whereas sizes were not. RLS, reproductive life span.

Population variability for inbreeding depression, similar effects of outcrossing between ponds and rivers heterosis, and outbreeding depression: Among-popu- (Figure 2B). Analysis of the lifetime reproductive lation variation in inbreeding depression was significant success (LTRS) showed that rivers tended to exhibit only for 0- to 15-day survival (population-by-treatment negative heterosis (i.e., outbreeding depression) on the interaction in Table 2). However, this variation was not cumulative fitness measure, whereas ponds tended to related with the maternal habitat (mean d 6 SEM have positive heterosis (Figure 2B; Table 6). However, ponds, 0.56 6 0.11; rivers, 0.55 6 0.05; Table 2; Figure this trend was not significant due to large variation 2A). In all other traits inbreeding depression neither among populations within each habitat. varied between habitats nor varied among populations For each population, we plotted the fitness of inbred within habitats (Table 2; Figure 2A). Importantly, almost individuals and interpopulation hybrids against the all populations exhibited inbreeding depressions on fitness of within-population cross-fertilized individuals the lifetime reproductive success .0.5, the threshold to know whether differences in inbreeding depression above which cross-fertilization is assumed to be favored and heterosis among populations could be due to in most theoretical models. The only exception corre- additive effects. Under additive effects, populations sponds to the Buz population (0.45) (Table 6). This having above-average values for fitness components in result is in line with the fact that P. acuta is a the intrapopulation outcrossing treatment would tend preferentially cross-fertilizing species. to exhibit less heterosis (or more outbreeding depres- For each population i, the effect of outbreeding can sion) than populations below average. This is because be measured, keeping maternal effects constant, by (additive) breeding values of interpopulation hybrids using the relative difference (H ¼ 1 ½Wi;intra=Wi;inter) are the average of their two parental populations and between the performance of offspring from crosses therefore lower than the best, and higher than the within population i (Wi,intra) and that of interpopula- worst, of these two parental populations. With regard to tion hybrids whose mothers belong to population i and inbreeding depression, the same principle holds for fathers belong to another population (Wi,inter). This different families within a population: families with relative difference significantly varied among popula- above-average breeding values will tend to show less tions for early survival (0–15 days), RLS, and fecundity inbreeding depression than below-average families. (maternal population-by-treatment interaction; Table Because we sampled a finite number of families within 3). Importantly, this variation was significantly related to each population, the average of our sample can be, by the maternal habitat type for two of these three chance, either higher or lower than the true population characters (maternal habitat-by-treatment interaction; mean, resulting in a downward or an upward bias Table 3): 0- to 15-day survival and RLS. Early survival (respectively) on our estimate of inbreeding depression decreased more upon interpopulation crossing when and a lower apparent inbreeding depression in above- the mother population was in a river than when it was in average populations. No such tendencies were, however, a pond (H ¼0.45 6 0.18 in rivers and 0.15 6 0.12 in observed in our analyses (Figure 3 and supplemental ponds). RLS increased more upon outbreeding when Figure 1). the mother population was in a pond than when it was in Additive, nonadditive, and asymmetric effects of a river (H ¼ 0.22 6 0.09 and 0.02 6 0.09, respectively). interpopulation hybrids: We further analyzed data from These interactions were still significant when the two both intra- and interpopulation outcrossing treatments types of interpopulation crosses (with a pond and with a to determine the relative importance of additive, non- river paternal population) were distinguished in the additive, and asymmetric effects of interpopulation analysis (data not shown). All other fitness traits showed hybrids. By nonadditivity, we mean that interpopulation Inbreeding and Outbreeding in a Metapopulation 1601

crosses differ from the average of parental populations. By asymmetry, we mean a difference in the ﬁtness of a cross between two populations, depending on which population assumes the maternal role. Signiﬁcant nonadditive effects were present for all traits, and 0.05.

. signiﬁcant asymmetries between reciprocal crosses MSQ F

P weredetectedinalltraitsbutRLSandsizeat150days (Table 7). This is visible in Figure 4 (and supplemental Figure 2), which shows that most interpopulation F

0.05; NS, crosses differ from their midparent value and often

, lie outside the parental range. Nonadditive effects are P either positive (heterosis) or negative (outbreeding depression), depending on the trait and population.

0.01; * Figure 4 also illustrates that the two reciprocal crosses MSQ , often yield contrasted results. P Correlation analyses: Three types of association tests were performed: (i) across maternal populations, be- F tween neutral-diversity parameters and inbreeding de- 0.001; **

, pression or heterosis; (ii) across pairs of populations,

P between population-genetic divergence and interpopulation-hybrid performance; and (iii) across maternal Dev populations, between inbreeding depression and heterosis for all measured traits. None of the three types of -values. ***

F correlations were signiﬁcant, when considering our

F global measure of fitness (lifetime reproductive success). The same was true for all other traits, after correction for multiple comparisons (supplemental Tables 1–3). Dev TABLE 5 DISCUSSION F In this study we manipulated the relatedness among parental individuals, from selfing to interpopulation survival RLS Fecundityoutcrossing, Size 22 days Size 150 days as well as the direction of hybridization 22- to 49-day (e.g., we distinguished between female A 3 male B and Dev female B 3 male A crosses), and evaluated fitness effects in F1 in the form of inbreeding depression,

., change in deviance between models with and without that factor); MSQ, mean square; RLS, reproductive life span. heterosis, and outbreeding depression expressed dur- i.e F ing a full-life cycle. We further analyzed theoretical Deviance analyses for the intrapopulation outcrossing treatment alone predictions concerning the effective population size, neutral genetic diversity, and genetic variance on all survival

15- to 22-day three parameters.

Dev Inbreeding depression, heterosis, and outbreeding depression simultaneously occur in the F1 of a structured population: Our results show that within a metapopulation, inbreeding depression, heterosis, and F outbreeding depression are simultaneously expressed (1; 399)(8; 391) (1; 343) (8; 335) (1; 330) (8; 322) (1; 332) (8; 324) (1; 332) (8; 324) (1; 130) (8; 123) (1; 8) (8; 130)

(120; 271) (114; 221)on different (112; 210) traits. (102; 222) Inbreeding (102; 222) depression (112; 1337) affects basi- (98; 195)

survival cally all fitness traits and is generally high (0.79 on the 0- to 15-day cumulative fitness measure on average), which is not surprising in a predominantly outcrossing species (70– 100% outcrossing from our molecular data, in agreement with previous data on the same species; Jarne et al. 2000; Escobar et al. 2007). To a certain extent heterosis, the increase in fitness traits upon interpopulation Dev, deviance explained by the factor considered ( Numerator and denominator degrees of freedom are given in parentheses with the corresponding Population 962.4Family 6.69*** 4242.8 132.0 3.43*** 10.12*** 272.1 76.5 1.93*** 6.76*** 184.5 4.1 1.27, NS 2.32* 23.9 1.09, 132.1 NS 423.5 4.03*** 1.02, 0.079 NS 0.028 3.43** 5.70*** 0.112 0.019 6.35*** 1.48* Habitat 152.4 7.60** 0.23 0.12, NS 3.74 2.32, NS 7.9 34.73*** 1.0 0.23, NS 0.033 1.59, NS 0.013 0.13, NS Source Dev crossing, reflects the same type of mutations as in- 1602 J. S. Escobar, A. Nicot and P. David

TABLE 6 Log-inbreeding depression and log heterosis on measured life-history traits per population

0- to 15-day 15- to 22-day 22- to 49-day Size 22 Size 150 Population survival survival survival RLS RLE Fecundity LTRS days days Buz 0.5378 0.2453 0.1184 0.3086 0.4826 0.1220 0.6047 0.0417 0.1061 (0.4282) (0.0068) (0.1806) (0.1341) (0.3750) (0.3550) (0.0199) (0.0334) (0.0624) Lam 0.9963 0.0097 0.2470 0.2319 1.4654 0.5617 2.0271 0.0108 0.1137 (0.4405) (0.0323) (0.0804) (0.0297) (0.5828) (0.4686) (0.1142) (0.1791) (0.0958) Lez 1.1252 0.0633 0.0720 0.4380 1.6985 0.7451 2.4436 0.2248 0.0441 (0.4119) (0.0530) (0.0604) (0.4018) (0.0027) (0.2522) (0.2548) (0.0221) (0.0171) Mos 0.8016 0.0139 0.1431 0.2828 1.2136 0.4534 1.6671 0.0083 0.1456 (0.7071) (0.0174) (0.2075) (0.1541) (0.6363) (0.2508) (0.3855) (0.1511) (0.0161) Sal 0.5352 0.0511 0.0254 0.2949 0.8559 0.2379 1.0937 0.0298 0.0765 (0.1327) (0.0094) (0.0650) (0.0261) (0.0844) (0.0183) (0.0661) (0.1600) (0.0006) Vio 01 1.4636 0.0817 0.0302 0.2534 1.8289 0.2338 2.0627 0.2461 0.1047 (0.4186) (0.0712) (0.0555) (0.3395) (0.0949) (0.2320) (0.3269) (0.0819) (0.0405) Vio 02 1.0051 0.0402 0.1730 0.0110 1.2293 0.5188 1.7481 0.4001 0.0738 (0.1125) (0.0634) (0.0979) (0.3225) (0.2445) (0.6434) (0.8880) (0.0497) (0.1189) Vio 07 0.1637 0.0309 0.3504 0.0984 0.6434 0.0583 0.7017 0.1627 0.0241 (0.2041) (0.0444) (0.0915) (0.4229) (0.4912) (0.0753) (0.5665) (0.0372) (0.0211) Vio 11 0.5757 0.0174 0.1097 0.2661 0.9341 0.4977 1.4317 0.0200 0.1106 (0.0725) (0.0938) (0.0279) (0.3049) (0.1106) (0.1459) (0.2565) (0.1173) (0.0244) Vio 12 0.9936 0.1271 0.1263 1.0220 2.0164 0.0121 2.0044 0.2351 NA (0.2943 (0.2801) (0.3726) (0.1367) (1.0837) (0.2031) (0.8806) (0.2150) (0.0774) Log-inbreeding depression is provided outside and log heterosis within brackets. Note that for size at 150 days there was no surviving selfed progeny from population Vio 12; therefore the inbreeding depression could not be estimated (NA). breeding depression; i.e., it also stems from directional demes by genetic drift, even without long-term com- dominance, most likely due to a constant mutational plete isolation, in such a way as to generate substantial input of deleterious, recessive alleles (Lynch 1991). An outbreeding depression in the F1. The existence of such additional force is needed to generate heterosis: genetic incompatibilities adds complexity to the interpretation drift, allowing allele frequencies to diverge among of the observed pattern, because interpopulation F1 will populations (Whitlock et al. 2000). The importance express the balance between two classes of alleles with of drift in our snail metapopulation is indicated by opposite effects: deleterious recessives (creating heter- significant and relatively high neutral microsatellite Fst osis) and genetic incompatibilities (creating outbreed- (at such a small scale, all populations within 25 km), ing depression). Similar data have been obtained in mostly in the range 10–20% (overall 12%). This degree plants from demes separated by short distances (Waser of population structure appears sufficient for popula- and Price 1989; Fenster 1991; Waser and Price 1994; tions to fix (or nearly fix) different sets of mildly Fischer and Matthies 1997). deleterious mutations, thus producing heterosis upon Inbreeding depression, heterosis, and outbreeding crossing. depression reveal differential impacts of mutations The occurrence of significant outbreeding depres- along the life cycle: The magnitude of inbreeding sion in the F1 of demes a few kilometers apart is more depression, heterosis, and outbreeding depression was surprising. Outbreeding depression is usually consid- very variable along the life cycle in P. acuta. Inbreeding ered an important issue when divergent populations, depression was strongest early in life, inexistent at often recognized as distinct subspecies or well on the intermediate stages, and increased again after sexual way to speciation, are brought into contact (e.g., during maturity (RLS and fecundity). Such stage-specific differ- reintroduction programs) rather than for two demes in ences in inbreeding depression have already been a local metapopulation. Moreover, outbreeding depres- observed in other predominantly outcrossing organ- usband chemske sion is considered more likely to be expressed in the F2, isms, especially plants (H and S 1996; because different populations can more easily accumu- Koelewijn et al. 1999). They are thought to arise late genetic incompatibilities (i.e., alleles with negative because both mutational pressure and selection inten- epistatic interaction) when the latter are completely sity vary among stages. Strong early-acting inbreeding recessive (the Dobzhansky–Muller model; Orr and depression reflects the expression of recessive lethal and Turelli 2001). Our data show that genetic incompat- sublethal mutations affecting basic functions ( Jarne ibilities that are not purely recessive can accumulate in and Charlesworth 1993; Husband and Schemske metapopulations and diverge in frequency among 1996; Remington and O’Malley 2000). In organisms Inbreeding and Outbreeding in a Metapopulation 1603

Figure 3.—Log-fitness traits for self-fertilized offspring (ln selfing, left) and interpopulation hybrids (ln inter, right) as a function of log-fitness traits for intrapopulation outcrossed offspring (ln intra). Only the most important fitness components are represented. (A) 0- to 15-day survival; (B) reproductive life span (RLS); (C) fecundity; (D) lifetime reproductivesuccess(LTRS). Lines represent constant values of log-inbreeding depression (ln d, left and log heterosis (ln H, right). Open squares, ponds; solid squares, rivers. Open areas represent positive values and shaded areas negative values of inbreeding depression and heterosis, respectively (note that negative heterosis represents outbreeding depression). Note change in scales among graphs.

that survive the juvenile stage, the inbreeding depres- breeding depression among life stages, suggesting that sion observed in later life stages is probably due to a genes causing inbreeding depression are to some extent more restricted set of partially recessive mildly deleteri- independent across the life cycle. ous genes with age-specific expression (e.g., genes in- The fitness in F1 interpopulation hybrids reveals the volved in late growth or reproduction). In agreement balance between heterosis and outbreeding depression. with this idea, we found no correlation between in- Theory states that the fitness loss caused by genetic 1604 J. S. Escobar, A. Nicot and P. David

incompatibilities has to exceed twice the beneﬁt from

F dominance for the F1 population to exhibit outbreeding depression (Lynch 1991). Our results show that outbreeding depression predominates at early stages while heterosis predominates at the reproductive stage (RLS Dev and fecundity). This suggests that viability is relatively more affected by genetic incompatibilities while muta-

F tions affecting speciﬁcally the reproductive stage are e performances equal to

nally the asymmetric non- more of the deleterious recessive type. The expression of strong incompatibilities affecting viability is surprising because, in the context of speciation, it is often assumed Dev that hybrid fertility is compromised earlier than hybrid viability during the process of population divergence resgraves

F (P 2003). For example, interspeciﬁc F1’s often show heterosis for survival and condition, while at least one sex is sterile (Orr and Turelli 2001). Interestingly, in our data set, late-acting heterosis and

Dev early-acting outbreeding depression cancel out when analyzing our cumulative ﬁtness measure (the lifetime reproductive success); i.e., averaging over the entire life F cycle, interpopulation F1 hybrids are as ﬁt as offspring from local residents. As for inbreeding depression, heterosis was not correlated among life stages, suggest-

Dev ing independence among genes at different stages of the life cycle. Furthermore, inbreeding depression and heterosis were not correlated to each other at any stage

F of the life cycle, which is expected because inbreeding depression is due to mutations segregating within demes, while heterosis is more affected by mutations that reach ﬁxation in some demes. TABLE 7 Genetic divergence among populations and habitat Dev types: As expected, and in agreement with previous studies (Bousset et al. 2004; Henry et al. 2005), rivers exhibited more neutral genetic variation and less

F genetic divergence (Fst) than ponds. River populations also outperformed pond populations with respect to early (0–15 days) and late (RLS) survival (intrapopulation outcrossing treatment). This suggests that small population size and isolation enhance genetic drift

Dev within populations, which in turn has parallel effects on neutral variation (less variation within and more variation among demes) and on nonneutral variation

Deviance analyses testing for asymmetries between reciprocal crosses in ﬁtness traits (larger standing genetic load). F Comparisons between phenotypic (Qst) and molecu- (9; 1263) (9; 984) (9; 934) (9; 985) (9; 983) (9; 1941) (9; 386) (17; 1246)(16; 1230) (17; 967) (16; 951)lar (17; 917) (Fst) (16; 901)divergence (17; 968) (16; 952) are (17; 966) usually (16; 950) (17;interpreted 1924) (15; 1909) (16; 370) as (13; 357) character- izing the type of selection (spatially heterogeneous selection if Qst . Fst and homogeneous selection when

0- to 15-day survival 15- to 22-day survival 22- to 49-day survival RLS Fecundity Size 22 daysFst . Size 150Q days st;McKay and Latta 2002; Goudet and Bu¨chi 2006). However, this interpretation should be viewed with caution here because the traits involved (survival and fecundity) are likely to be positively selected in all populations rather than selected for intermediate optima, and we did not detect correlations with other traits that could be indicative of trade-offs. Although heterogeneous selection cannot be excluded, the fact that the habitat type (river vs. pond) explains more Deﬁnitions are the same as in Table 2. Data include intra- and interpopulation crosses. The additive model constrains interpopulation crosses to hav AdditiveSymmetric nonadditive 1523Asymmetric nonadditive 1133 4.90*** 1662 4.02*** 71.3 9.60*** 58.2 2.10** 214 1.84* 43.5 11.67*** 42.0 148.7 1.84* 11.67*** 1.91* 19.0 18.0 4.88*** 8.16*** 213.3 3.7 396.3 3.95*** 1.02, NS 13.18*** 1.18 182.1 0.59 3.74*** 8.73*** 7.69*** 0.45 0.71 1.34 3.88*** 3.20*** 9.76*** 0.21 1.15, NS Effect Dev their midparent value. Theadditive symmetric model nonadditive ﬁts model one relaxes value this for constraint, each but constrains cross reciprocal including interpopulation reciprocals. crosses to be equal. Fi variance in RLS than in molecular traits (Qct . Fct), as Inbreeding and Outbreeding in a Metapopulation 1605

Figure 4.—Intra- and interpopulation outcrossing performances for the most important fitness components. (A) 0- to 15-day survival; (B) reproductive life span; (C) fecundity. Solid squares and lines represent intrapopulation outcrossing; shaded squares and lines represent interpopulation outcrossing. Results are presented for each pair of populations (two shaded squares and two solid squares). For instance, the first two populations on the left (Lam and Mos) should be read as follows: the solid square on the left represents the value for the Lam 3 Lam cross, and the shaded square represents the value for the cross female Lam 3 male Mos. The solid square on the right represents the value for the Mos 3 Mos cross, and the shaded square the value for the female Mos 3 male Lam cross. In this representa- tion, differences between parental populations are indicated by the slope of the solid line, and asymmetries between reciprocal crosses are indicated by the slope of the shaded line.Different shadings de- note heterosis (interpopulation hybrids fitterthan intrapopulation crosses) and outbreeding depression (vice versa) as indicated. By convention, the river is on the left and the pond on the right in all mixed pairs of populations. Note that one cross (female Mos 3 male Lez) could not be performed.

mentioned above, likely reflects a differential impact of drift in pond populations (so that even selected traits drift (fixation of mildly deleterious alleles by drift being behave more like neutral ones in ponds). At any rate, more frequent in small isolated populations) rather one should remain cautious in comparing Fst and Qst than different selection regimes. Similarly, the larger Qst here, both because of the imprecision in Qst estimates for most traits in the pond group of populations, relative (based on relatively few populations) and because of to the river group, probably reflects a larger influence of dominance, the presence of deleterious recessives, and 1606 J. S. Escobar, A. Nicot and P. David epistasis, all of which can affect classical predictions Additive and nonadditive effects on variation in (Whitlock 1999; Lo´pez-Fanjul et al. 2003; Goudet heterosis and outbreeding depression: The variation in and Bu¨chi 2006). heterosis or outbreeding depression among popula- Because the mean and variance in mutation fre- tions could be due in part to additive effects, i.e., quency across demes depend on genetic drift, inbreed- differences in the average breeding value among ing depression, heterosis, and outbreeding depression populations for a given trait. For example, although are all expected to vary stochastically among popula- alleles with purely additive effects generate neither tions. This is basically what we found. However, variation heterosis nor outbreeding depression on average, they in inbreeding depression seems limited to the early contribute to the variance among populations because survival stage, and basically all populations show strong populations with below-average breeding values (e.g., inbreeding depression. On the other hand, the balance pond populations) will tend to gain fitness, while between outbreeding depression and heterosis seems populations with above-average breeding value (e.g., more variable, with positive values for some population river populations) will tend to lose fitness, when mixing pairs and negative values for others, for both early and their genes with genes from other populations. If late traits. More specific predictions can be made on additive effects were mainly responsible for differences interpopulation variation assuming that local popula- in heterosis/outbreeding depression among populations vary with respect to effective size and isolation. In tions, we expect (i) that these differences would be small and isolated populations, deleterious allele fre- determined mainly by the performances in the within- quencies should more easily drift away from the de- population crosses, (ii) that the highest heterosis terministic equilibrium and from the metapopulation should be observed in ponds, when crossing them with mean (up to local fixation) (Kimura et al. 1963; a river population, though not (or less) when crossing Bataillon and Kirkpatrick 2000; Gle´min et al. them with another pond population. We have no 2001). They should therefore exhibit less inbreeding evidence for either i or ii. The interaction between depression and more heterosis than larger ones. How- maternal habitat type and cross type (i.e., the variation ever, considering that heterosis is expected to arise from in heterosis among maternal habitat types) remains milder mutations than inbreeding depression on aver- significant once the paternal habitat is controlled for in age, the sensitivity to population sizes should differ: pop- the model. Moreover, very significant nonadditive ulations must be very small to lose a substantial part of interactions among the paternal and maternal popula- their inbreeding depression, while even moderately small tions are observed, and heterosis occurs in pond 3 effective sizes allow heterosis to increase (Kirkpatrick pond crosses just as it does in pond 3 river crosses. and Jarne 2000; Gle´min et al. 2001; Gle´min 2003). Interactions in a given population pair seem highly Habitat (pond vs. river), our proxy for population size, idiosyncratic, with some heterotic crosses and some did not explain interpopulation variation in inbreeding crosses showing outbreeding depression, sometimes depression in our data set; however, it did have a sig- involving the same maternal populations (Figure 4 nificant effect on heterosis for RLS in the expected and supplemental Figure 2). In addition, many of the direction; i.e., ponds benefited more from interpopula- crosses show asymmetry between reciprocal crosses, tion outcrossing than did rivers. This suggests that the which is globally significant for all fitness traits but range of variation in effective size among the studied RLS. Asymmetries can result from several mechanisms, populations was sufficient to generate differential rates including maternal effects (Roach and Wulff 1987), of fixation of mildly deleterious recessive alleles, but nucleocytoplasmic gene interactions (Galloway and populations were still large enough to avoid the fixation Fenster 1999; Pe´labon et al. 2005), and locally adapted of stronger mutations and thus retain their inbreeding alleles interacting differently with native and foreign depression. Interpreting why the effect of interpopula- genetic backgrounds (Wright 1931; Wade 1992). tion outcrossing on early survival was less detrimental to With respect to early survival, the observed asymmetry ponds than to rivers is more complicated. Assuming that is certainly strongly influenced by maternal effects. Even this effect represents the sum of a negative term (out- if eggs were separated from their mother immediately breeding depression) and a positive one (heterosis), after oviposition, maternal effects can be mediated this observation could be explained (as for RLS) by the through egg cytoplasm (e.g., amount of nutritive re- increase in heterosis expected in small populations. How- serves) and such mechanisms may dominate all other ever, small population size, by enhancing the influence sources of asymmetry in early-expressed characters of genetic drift and genetic divergence among popula- (Roach and Wulff 1987). Conversely, maternal effects tions, should in principle increase both heterosis and are probably limited when evaluating adult traits, such outbreeding depression, and the effect could therefore as fecundity. In this case, hypotheses invoking genetic go both ways. Unfortunately, we lack precise theoretical background, such as nucleocytoplasmic incompatibili- expectations to predict whether both outbreeding de- ties, commonly observed for fecundity traits (reviewed pression and heterosis could respond to the same range in Budar et al. 2003), are probably involved. Further of variation in population size. research is needed to determine whether the genetic Inbreeding and Outbreeding in a Metapopulation 1607 divergence between P. acuta populations involves other Charlesworth, D., and B. Charlesworth, 1987 Inbreeding de- mechanisms than nuclear genes. pression and its evolutionary consequences. Annu. Rev. Ecol. Evol. Syst. 18: 237–268. In conclusion, we provide evidence for the simulta- Crawley, M. J., 2005 Statistics: An Introduction Using R. Wiley Pub- neous expression of inbreeding depression, heterosis, lishers, London. rnokrak off and outbreeding depression in first-generation off- C , P., and D. A. R , 1999 Inbreeding depression in the wild. Heredity 83: 260–270. spring, at the scale of local demes within a metapopu- Crow, J. F., 1948 Alternative hypotheses of hybrid vigor. Genetics lation. We show that the relative importance of the three 33: 477–487. arwin phenomena varies along the life cycle, i.e., among D , C., 1876 The Effects of Cross and Self Fertilisation in the Vegeta- ble Kingdom. John Murray, London. different fitness traits, which are exposed to different David, P., B. Pujol,F.Viard,V.Castella and J. Goudet, impacts of both mutation pressure and selection. 2007 Reliable selfing rate estimates from imperfect population Specifically, inbreeding depression and outbreeding genetic data. Mol. Ecol. 16: 2474–2487. Deng, H. W., Y. X. Fu and M. Lynch, 1998 Inferring the major ge- depression are more intense in early-expressed traits nomic mode of dominance and overdominance. Genetica 102/ while heterosis is moderate and more visible during the 103: 559–567. reproductive stage. In addition, we provide evidence Dobzhansky, T., 1936 Studies on hybrid sterility. II. Localization of sterility factors in Drosophila pseudoobscura hybrids. Genetics 21: that the effects of interpopulation crossing largely vary 113–135. depending on the focal maternal populations. A large Edmands, S., 1999 Heterosis and outbreeding depression in inter- part of this variation is not additive and is influenced by population crosses spanning a wide range of divergence. Evolu- tion 53: 1757–1768. maternal or cytoplasmic effects, reflecting unpredict- Edmands, S., 2007 Between a rock and a hard place: evaluating the able interactions between genetic backgrounds that relative risks of inbreeding and outbreeding for conservation and have diverged through stochastic processes. However, management. Mol. Ecol. 16: 463–475. Edmands, S., and C. C. Timmerman, 2003 Modeling factors affect- as predicted by theory, habitat type, through its effect on ing the severity of outbreeding depression. Conserv. Biol. 17: effective population size and/or immigration rates, 883–892. explains a part of the variation in heterosis among Escobar, J. S., G. Epinat,V.Sarda and P. David, 2007 No correla- populations. tion between inbreeding depression and delayed selfing in the freshwater snail Physa acuta. Evolution 61: 2655–2670. We acknowledge P. Jarne and D. Promislow for comments on the Excoffier, L., G. Laval and S. Schneider, 2005 Arlequin ver. 3.0: manuscript; J. Goudet for unpublished microsatellite primers; and an integrated software package for population genetics data analysis. Evol. Bioinform. Online 1: 47–50. V. Sarda, G. Hinet, J. Terraube, and E. Dincuff for help in the labo- enster ratory. J.S.E. was supported by the Programme Alban (grant F , C. B., 1991 Effect of male pollen donor and female seed parent on allocation resources to developing seeds and fruit in E04D045840CO). Funds for this research were given by the Centre Chamaecrista fasciculata (Leguminosae). Am. J. Bot. 78: 12–23. National pour la Recherche Scientifique. Fenster, C. B., and L. F. Galloway, 2000 Population differentiation in an annual legume: genetic architecture. Evolution 54: 1157–1172. Fischer, M., and D. Matthies, 1997 Mating structure and inbreed- LITERATURE CITED ing and outbreeding depression in the rare plant Gentianella ger- manica (Gentianaceae). Am. J. Bot. 84: 1685–1692. Andersen, D. H., C. Pertoldi,V.Scali and V. Loeschcke, Frankham, R., 1999 Quantitative genetics in conservation biology. 2002 Intraspecific hybridization, developmental stability and Genet. Res. 74: 237–244. fitness in Drosophila mercatorum. Evol. Ecol. Res. 4: 603–621. Galloway, L. F., and C. B. Fenster, 1999 The effect of nuclear and Aspi, J., 2000 Inbreeding and outbreeding depression in male cytoplasmic genes on fitness and local adaptation in an annual courtship song characters in Drosophila montana. Heredity 84: legume, Chamaecrista fasciculata. Evolution 53: 1734–1743. 273–282. Gharrett, A. J., W. W. Smoker,R.R.Reisenbichler and S. G. Aspi, J., E. Roininen,M.Ruokonen,I.Kojola and C. Vila, Taylor, 1999 Outbreeding depression in hybrids between 2006 Genetic diversity, population structure, effective popula- odd- and even-broodyear pink salmon. Aquaculture 173: 117– tion size and demographic history of the Finnish wolf popula- 129. tion. Mol. Ecol. 15: 1561–1576. Gilk, S. E., I. A. Wang,C.L.Hoover,W.W.Smoker,S.G.Taylor Bataillon, T., and M. Kirkpatrick, 2000 Inbreeding depression et al., 2004 Outbreeding depression in hybrids between spatially due to mildly deleterious mutations in finite populations: size separated pink salmon, Oncorhynchus gorbuscha, populations: ma- does matter. Genet. Res. 75: 75–81. rine survival, homing ability, and variability in family size. Envi- Bierne,N.,T.Lenormand,F.Bonhomme and P. David, ron. Biol. Fish. 69: 287–297. 2002 Deleterious mutations in a hybrid zone: Can mutational Gle´min, S., 2003 How are deleterious mutations purged? Drift ver- load decrease the barrier to gene flow? Genet. Res. 80: 197– sus nonrandom mating. Evolution 57: 2678–2687. 204. Gle´min, S., T. Bataillon,J.Ronfort,A.Mignot and I. Olivieri, Bousset, L., P.-Y. Henry,P.Sourrouille and P. Jarne, 2001 Inbreeding depression in small populations of self-incom- 2004 Population biology of the invasive freshwater snail Physa patible plants. Genetics 159: 1217–1229. acuta approached through genetic markers, ecological character- Goudet, J., 1995 FSTAT (Version 1.2): a computer program to cal- ization and demography. Mol. Ecol. 13: 2023–2036. culate F-statistics. J. Hered. 86: 485–486. Budar, F., P. Touzet and R. De Paepe, 2003 The nucleo-mitochon- Goudet,J.,andL.Bu¨chi, 2006 The effects of dominance, regular in- drial conflict in cytoplasmic male sterilities revisited. Genetica breeding and sampling design on QST, an estimator of population 117: 3–16. differentiation for quantitative traits. Genetics 172: 1337–1347. Burton, R. S., 1990 Hybrid breakdown in developmental time in Hamilton, W. D., 1966 The moulding of senescence by natural se- the copepod Tigriopus californicus. Evolution 44: 1814–1822. lection. J. Theor. Biol. 12: 12–45. Byers, D. L., 1998 Effect of cross proximity on progeny fitness in a Hartl, D. L., and A. G. Clark, 1997 Principles of Population Genetics. rare and a common species of Eupatorium (Asteraceae). Am. J. Sinauer Associates, Sunderland, MA. Bot. 85: 644–653. Henry, P. Y., and P. Jarne, 2007 Marking hard-shelled gastropods: Charlesworth, B., and D. Charlesworth, 1999 The genetic basis tag loss, impact on life-history traits, and perspectives in biology. of inbreeding depression. Genet. Res. 74: 329–340. Invertebr. Biol. 126: 138–153. 1608 J. S. Escobar, A. Nicot and P. David

Henry, P. Y., L. Bousset,P.Sourrouille and P. Jarne, 2005 Partial Rasband, W., 2007 ImageJ 1.34s. Image processing and analysis in selfing, ecological disturbance and reproductive assurance in an Java. National Institutes of Health, Bethesda, MD. invasive freshwater snail. Heredity 95: 428–436. Remington, D. L., and D. M. O’Malley, 2000 Whole-genome char- Husband, B., and D. Schemske, 1996 Evolution of the magnitude and acterization of embryonic stage inbreeding depression in a selfed timing of inbreeding depression in plants. Evolution 50: 54–70. loblolly pine family. Genetics 155: 337–348. Jarne, P., and D. Charlesworth, 1993 The evolution of the selfing Roach, D. A., and R. D. Wulff, 1987 Maternal effects in plants. An- rate in functionally hermaphrodite plants and animals. Annu. nu. Rev. Ecol. Evol. Syst. 18: 209–235. Rev. Ecol. Evol. Syst. 24: 441–466. Rose, M. R., 1991 Evolutionary Biology of Aging. Oxford University Jarne, P., M. A. Perdieu,A.F.Pernot and P. David, 2000 The in- Press, New York. fluence of self-fertilization and grouping on fitness attributes in Rousset, F., 1997 Genetic differentiation and estimation of gene the freshwater snail Physa acuta: population and individual in- flow from F-statistics under isolation by distance. Genetics 145: breeding depression. J. Evol. Biol. 13: 645–655. 1219–1228. Johnston, M. O., and D. J. Schoen, 1994 On the measurement of Schierup, M. H., and F. B. Christiansen, 1996 Inbreeding depres- inbreeding depression. Evolution 48: 1735–1741. sion and outbreeding depression in plants. Heredity 77: 461– Keller, L. F., and D. M. Waller, 2002 Inbreeding effects in wild 468. populations. Trends Ecol. Evol. 17: 230–241. Shpak, M., 2005 The role of deleterious mutations in allopatric spe- Kimura, M., T. Maruyama and J. F. Crow, 1963 The mutation load ciation. Evolution 59: 1389–1399. in small populations. Genetics 48: 1303–1312. Sourrouille, P., C. Debain and P. Jarne, 2003 Microsatellite vari- Kirkpatrick,M.,andP.Jarne, 2000 The effect of a bottleneck on in- ation in the freshwater snail Physa acuta. Mol. Ecol. Notes 3: 21– breeding depression and the genetic load. Am. Nat. 155: 154–167. 23. Koelewijn, H. P., V. Koski and O. Savolainen, 1999 Magnitude Spitze, K., 1993 Population structure in Daphnia obtusa: quantitative and timing of inbreeding depression in Scots pine (Pinus sylvestris genetic and allozymic variation. Genetics 135: 367–374. L.). Evolution 53: 758–768. Tallmon, D. A., G. Luikart and R. S. Waples, 2004 The alluring Li, Z. K., L. J. Luo,H.W.Mei,D.L.Wang,Q.Y.Shu et al., simplicity and complex reality of genetic rescue. Trends Ecol. 2001 Overdominant epistatic loci are the primary genetic basis Evol. 19: 489–496. of inbreeding depression and heterosis in rice. I. Biomass and Trame, A. M., A. J. Coddington and K. N. Paige, 1995 Field and grain yield. Genetics 158: 1737–1753. genetic-studies testing optimal outcrossing in Agave schottii,a Lo´pez-Fanjul, C., A. Fernańdez and M. A. Toro, 2003 The effect of long-lived clonal plant. Oecologia 104: 93–100. neutral nonadditive gene action on the quantitative index of Wade, M. J., 1992 Sewall Wright: gene interaction and the shifting population divergence. Genetics 164: 1627–1633. balance theory. Oxf. Surv. Evol. Biol. 8: 35–62. Lynch, M., 1991 The genetic interpretation of inbreeding depres- Waser, N. M., 1993 Population structure, optimal outbreeding, and sion and outbreeding depression. Evolution 45: 622–629. assortative mating in angiosperms, pp. 173–199 in The Natural Lynch, M., and B. Walsh, 1998 Genetics and Analysis of Quantitative History of Inbreeding and Outbreeding, edited by N. W. Thornhill. Traits. Sinauer Associates, Sunderland, MA. University of Chicago Press, Chicago. Marr, A. B., L. F. Keller and P. Arcese, 2002 Heterosis and out- Waser, N. M., and M. V. Price, 1989 Optimal outcrossing in Ipomop- breeding depression in descendants of natural immigrants to sis aggregata: seed set and offspring fitness. Evolution 43: 1097– an inbred population of song sparrows (Melospiza melodia). Evo- 1109. lution 56: 131–142. Waser, N. M., and M. V. Price, 1994 Crossing-distance effects in Del- McKay, J. K., and R. G. Latta, 2002 Adaptive population diver- phinium nelsonii, outbreeding and inbreeding depression in prog- gence: markers, QTL and traits. Trends Ecol. Evol. 17: 285–291. eny fitness. Evolution 48: 842–852. Merila¨, J., and B. C. Sheldon, 1999 Genetic architecture of fitness Waser, N. M., M. V. Price and R. G. Shaw, 2000 Outbreeding de- and nonfitness traits: empirical patterns and development of pression varies among cohorts of Ipomopsis aggregata planted in ideas. Heredity 83: 103–109. nature. Evolution 54: 485–491. Monsutti, A., and N. Perrin, 1999 Dinucleotide microsatellite loci Whitlock, M. C., 1999 Neutral additive variance in a metapopula- reveal a high selfing rate in the freshwater snail Physa acuta. Mol. tion. Genet. Res. 74: 215–221. Ecol. 8: 1075–1092. Whitlock, M. C., P. K. Ingvarsson and T. Hatfield, 2000 Local Mukai, T., T. K. Watanabe and O. Yamaguchi, 1974 Genetic struc- drift load and the heterosis of interconnected populations. He- ture of natural populations of Drosophila melanogaster. 12. Linkage redity 84: 452–457. disequilibrium in a large local population. Genetics 77: 771–793. Willi, Y., and J. Van Buskirk, 2005 Genomic compatibility occurs Orr, H. A., and M. Turelli, 2001 The evolution of postzygotic iso- over a wide range of parental genetic similarity in an outcrossing lation: accumulating Dobzhansky-Muller incompatibilities. Evo- plant. Proc. R. Soc. B Biol. Sci. 272: 1333–1338. lution 55: 1085–1094. Willi, Y., M. Van Kleunen,S.Dietrich and M. Fischer, Pe´labon, C., M. L. Carlson,T.F.Hansen and W. S. Armbruster, 2007 Genetic rescue persists beyond first-generation outbreed- 2005 Effects of crossing distance on offspring fitness and devel- ing in small populations of a rare plant. Proc. R. Soc. B Biol. Sci. opmental stability in Dalechampia scandens (Euphorbiaceae). Am. 274: 2357–2364. J. Bot. 92: 842–851. Willis, J. H., 1999 The contribution of male-sterility mutations to Presgraves, D. C., 2003 A fine-scale genetic analysis of hybrid in- inbreeding depression in Mimulus guttatus. Heredity 83: 337– compatibilities in Drosophila. Genetics 163: 955–972. 346. Price, M. V., and N. M. Waser, 1979 Pollen dispersal and optimal Wright, S., 1931 Evolution in Mendelian populations. Genetics 16: outcrossing in Delphinium nelsoni. Nature 277: 294–297. 97–159. Quilichini, A., M. Debussche and J. D. Thompson, 2001 Evidence Wright, S., 1937 The distribution of gene frequencies in popula- for local outbreeding depression in the Mediterranean island en- tions. Proc. Natl. Acad. Sci. USA 23: 307–320. demic Anchusa crispa Viv. (Boraginaceae). Heredity 87: 190–197. RDevelopment Core Team, 2006 R: a language and environment for statistical computing. R Foundation for Statistical Comput- ing, Vienna. Communicating editor: M. K. Uyenoyama