A Selective Advantage to Immigrant Genes in a Daphnia Metapopulation

See allHide authors and affiliations

Science  18 Jan 2002:
Vol. 295, Issue 5554, pp. 485-488
DOI: 10.1126/science.1067485


Immigrants to habitats occupied by conspecific organisms are usually expected to be competitively inferior, because residents may be locally adapted. If residents are inbred, however, mating between immigrants and residents results in offspring that may enjoy a fitness advantage from hybrid vigor. We demonstrate this effect experimentally in a natural Daphnia metapopulation in which genetic bottlenecks and local inbreeding are common. We estimate that in this metapopulation, hybrid vigor amplifies the rate of gene flow several times more than would be predicted from the nominal migration rate. This can affect the persistence of local populations and the entire metapopulation.

Gene flow between populations can be both a creative and a constraining force in evolution (1–3). The introduction of new genetic material into a population increases local genetic diversity and helps the spread of favorable alleles across metapopulations. On the other hand, it reduces genetic variation between populations and hinders local adaptation. In inbred populations, the consequences of migration may be particularly important: If the hybrid offspring of immigrants and residents are competitively superior, their hybrid vigor will amplify the gene flow caused by migration (4–6). Furthermore, the demographic consequences of increased vigor could prevent the decline and even the extinction of populations (7–12). The magnitude of hybrid vigor is, however, controversial. Highly inbred populations may have low genetic loads because inbreeding exposes recessive deleterious alleles to purging by natural selection (13–17).

In subdivided populations with local extinctions and colonizations, genetic bottlenecks can be frequent, leading to increased homozygosity (15, 16, 18). If homozygosity results in a fitness reduction (inbreeding depression), then small amounts of immigration can have disproportional effects on the vigor and persistence of local populations. This will in turn influence the extinction and colonization dynamics of the whole metapopulation (4–6).

To test whether hybrid vigor has an effect on the genetic structure of local populations, we studied the consequences of immigration in small populations of the water flea Daphnia magna. This planktonic crustacean occurs in various bodies of water, ranging from large ponds, with genetically diverse populations, to small intermittent pools, characterized by low genetic diversity. In our metapopulation, D. magna inhabits small pools (0.5 to 20 m2 and 0.1 to 0.5 m deep) on the rocky banks of islands along the Swedish and Finnish Baltic Sea coast (19–22). The number of these rock pool populations is thought to be larger than 105, with one to a few hundred pools per island. In our study area around Tvärminne in southern Finland, local extinction rates are about 20% per year, estimated over an 18-year study including 507 pools (21, 23). About 5% of the empty pools are colonized per year, and the proportion of occupied pools has remained around 20% since 1983. The life cycle of D. magna begins with the hatching of females from resting eggs in spring, followed by asexual reproduction for up to 12 generations. At the end of the season, sexual reproduction produces resting eggs that survive the winter and also serve as dispersal stages.

In 1998, a population genetic survey of 96 D. magnapopulations with known age indicated that local populations are often founded by a single clone, followed by a population increase due to clonal expansion (24). Such genetic bottlenecks lead to substantial inbreeding during sexual reproduction, which is most extreme when a single founder clone produces resting eggs by mating with clonal brothers and sisters (known as selfing) (25). The frequent occurrence of genetic bottlenecks makes this metapopulation an ideal site to study the impact of local inbreeding on the success of immigrants.

To test whether hybrid vigor occurs in this metapopulation, we introduced immigrant genotypes into 22 D. magna populations and monitored the fate of their outbred offspring over one summer season. We chose 22 rock pool populations on nine islands in July 1999. These rock pools were known to be continuously inhabited by D. magna for at least 2 years before the beginning of the study. Of the five genetic loci known to be polymorphic in these metapopulations, the 22 studied populations were polymorphic at none (12 populations), one (six populations), or two loci (four populations), which is typical of the metapopulation. In July 1999, we removed all D. magnafrom these populations and kept 200 to 500 females from each population with minimal selection in the laboratory. To remove resting eggs that may have been produced earlier in that season, we removed the water and most of the soft sediments from the pools. Comparable disturbances occur naturally in this metapopulation, where storms on the Baltic Sea frequently wash out parts of or even entire pools (19,21). We also collected 22 clones from different pools within the study area and propagated them in the laboratory. These we designated as our experimental immigrant clones.

Two weeks later, after rain had refilled the pools, we brought 200 individuals of the populations back into their original pools and added 200 individuals of one immigrant clone into each pool (26). In each experimental pool, the immigrant clone differed from the local clones at at least one allozyme marker locus. To avoid bias due to potential fitness effects of the genetic markers, we used multiple alleles at four loci. The genetic markers allowed us to distinguish between hybrids, offspring of local residents, and selfed offspring of the immigrant clone after the populations had undergone one round of sexual recombination, i.e., in spring of the following year.

Rock pools were left undisturbed until we took the first samples in May 2000. From these samples, we genotyped 66 to 122 animals and also founded laboratory populations using 200 to 300 females each. This was done to duplicate the natural rock pool experiment under controlled laboratory conditions so that we could decouple the effect of inbreeding from uncontrolled environmental effects, such as the local pool environment, predators, and further immigrants from the same or other Daphnia species (27). Second and third samples were taken from all populations about 60 and 100 days after the first samples had been collected.

Our hypothesis was that the outbred offspring would increase in frequency as the result of hybrid vigor. Figure 1 shows that this was the case in all rock pool populations where we recovered hybrids in May 2000 (all but pool 15). It was also the case in 17 of 18 laboratory populations (Fig. 1) (28). The changes in genotype frequencies in the field and the laboratory were highly correlated with each other (r = 0.71 to 0.93 for the three offspring types,P < 0.002, n = 16), and the increase of outbred genotypes did not differ significantly among them (paired t test: t = 0.52,P = 0.61, n = 16). Hybrids also increased significantly in the time period between the second and the third sample (P < 0.01 in the field and the laboratory) at a time when genotypes with very low fitness (e.g., due to castrating homozygous recessive mutations) had already been purged. Therefore, the observed hybrid vigor is at least partially due to deleterious effects of alleles with weak to intermediate effect (6, 29). As selection against such alleles is weak, they can accumulate to high frequencies. High genetic loads have been reported for Daphnia (30–32).

Figure 1

Frequency changes of hybrids, residents, and inbred immigrants during asexual competition. The left plot of each pair of area plots shows frequency changes in the natural rock pools, whereas the right plots show changes in the laboratory. The population numbers are given on the right. All plots have the same y axis (genotype frequencies ranging from 0 to 1) and x axis (sampling events at 0, 60, and 100 days). The first sampling date was between 18 May and 5 June 2000 (depending on the emergence ofDaphnia from resting eggs), the second between 18 and 21 July 2000, and the third on 28 August 2000. Empty places for two of the rock pool area plots and area plots including only the first and the second sample indicate extinction of these populations after the first or the second sample, respectively. Only 19 of the 22 populations are shown, because populations 4 and 11 went extinct during the winter and in population 18 only residents were found (sample size >300). No hybrids were detected in population 15 and no inbred immigrants in populations 12, 13, and 22. In all other populations, we found all three offspring types.

We believe that hybrid vigor best explains our results, as a number of alternative hypotheses can be excluded. First, because we used a different immigrant clone for each population, it is unlikely that all of them carried superior alleles (33). Second, although rare genotypes could have had an advantage by, for example, exploiting unoccupied niches, we rejected this hypothesis because hybrids increased in frequency irrespective of their initial frequencies in May 2000 (Fig. 1). Third, although residents may suffer from locally adapted predators or parasites (34,35), predators were excluded from the laboratory, and the increase of hybrids did not differ significantly between populations infected with a parasite or not (36). Fourth, mating success cannot explain our results (37–39) as hybrid success was measured during the asexual phase of the Daphnialife cycle.

One key effect of hybrid vigor is an increase in effective gene flow (4). We estimate that 2 years after a natural immigration event (where immigrants represent a much smaller fraction of the population than in our experiments), the effective rate of gene flow is about 35 times larger than would be predicted by the number of immigrants alone, and it will even increase further in the following years (40). The amount by which gene flow is amplified depends on the causes and magnitude of hybrid vigor. It may be much smaller in genetically more diverse Daphnia populations but may be high in other systems characterized by frequent extinctions and recolonizations.

Amplified gene flow caused by hybrid vigor may account for lower observed levels of population differentiation than predicted by models based on neutral effects (4, 6,41–43) and may influence the evolution of dispersal (44). An important effect of hybrid vigor is the “genetic rescue” of populations from extinction, because it may influence extinction and colonization dynamics of the whole metapopulation (5). Although we did not show a link between population persistence and inbreeding depression, such a relation has been shown before (9, 11, 12) and is likely to apply to Daphnia metapopulations as well. In this light, gene flow is an essential component for the persistence of metapopulations. Thus, our study gives clear empirical support for the need to maintain gene flow in the management and conservation of subdivided populations.

  • * These authors contributed equally to this work.

  • To whom correspondence should be addressed. E-mail: dieter.ebert{at}


View Abstract

Navigate This Article