Generalists versus specialists in fluctuating environments: a bet-hedging perspective

Bet-hedging evolves in fluctuating environments because long-term genotype success is determined by geometric (rather than arithmetic) mean fitness across generations. Diversifying bet-hedging produces different specialist offspring, whereas conservative bet-hedging produces similar generalist offspring. However, many fields, such as behavioral ecology and thermal physiology, typically consider specialist versus generalist strategies only in terms of maximizing arithmetic mean fitness benefits to individuals. Here we model how environmental variability affects optimal amounts of phenotypic variation within and among individuals to maximise genotype fitness, and we disentangle the effects of individual-level optimization and genotype-level bet-hedging by comparing long-term arithmetic versus geometric mean fitness. For traits with additive fitness effects within lifetimes (e.g. foraging-related traits), genotypes of similar generalists or diversified specialists perform equally well. However, if fitness effects are multiplicative within lifetimes (e.g. sequential survival probabilities), generalist individuals are always favored. In this case, geometric mean fitness optimization requires even more within-individual phenotypic variation than does arithmetic mean fitness, causing individuals to be more generalist than required to simply maximize their own expected fitness. In contrast to previous results in the bet-hedging literature, this generalist conservative bet-hedging effect is always favored over diversifying bet-hedging. These results link the evolution of behavioral and ecological specialization with earlier models of bet-hedging, and we apply our framework to a range of natural phenomena from habitat choice to host specificity in parasites.     

Smart,smarter, smartest: foraging information states and coexistence

Animals possess different abilities to gain and use information about the foraging patches they exploit. When ignorant of the qualities of encountered patches, a smart forager should leave all patches after the same amount of fixed search time. A smarter forager can be Bayesian by using information on cumulative harvest and time spent searching a patch to better inform its patch‐departure decision. The smartest forager has immediate and continuous knowledge about patch quality, and can make a perfect decision about when to leave each patch. Here we let each of these three strategies harvest resources from a slowly regenerating environment. Eventually a steady‐state distribution of prey among patches arises where the environment‐wide resource renewal just balances the environment‐wide harvest of the foragers. The fixed time forager creates a distribution with the highest mean and highest variance of patch qualities, followed by the Bayesian and the prescient in that order. The less informed strategies promote distributions with both more resources and more exploitable information than the more informed strategies. While it is true that a better‐informed strategy will always out‐perform a less well‐informed, its increase in performance may not compensate it for any costs associated with being better informed. We imagine that the fixed time strategy may be least expensive and the prescient strategy most expensive in terms of sensory organs and associated assess and respond capabilities. To consider competition between such strategies with varying costs, we introduced a single individual of each of the strategies into the environments created by populations of the other strategies. There are threshold costs associated with the better‐informed strategy such that it can or cannot outcompete a less‐informed strategy. However, over a relatively narrow range of foraging costs, less‐informed and better‐informed strategies will coexist. Furthermore, for the prescient and the Bayesian strategies, some combinations of foraging costs produce alternate stable states – whichever strategy establishes first remains safe from invasion by the other.     

Spatially heterogeneous stochasticity and the adaptive diversification of dormancy

Diversified bet‐hedging, a strategy that leads several individuals with the same genotype to express distinct phenotypes in a given generation, is now well established as a common evolutionary response to environmental stochasticity. Life‐history traits defined as diversified bet‐hedging (e.g. germination or diapause strategies) display marked differences between populations in spatial proximity. In order to find out whether such differences can be explained by local adaptations to spatially heterogeneous environmental stochasticity, we explored the evolution of bet‐hedging dormancy strategies in a metapopulation using a two‐patch model with patch differences in stochastic juvenile survival. We found that spatial differences in the level of environmental stochasticity, restricted dispersal, increased fragmentation and intermediate survival during dormancy all favour the adaptive diversification of bet‐hedging dormancy strategies. Density dependency also plays a major role in the diversification of dormancy strategies because: (i) it may interact locally with environmental stochasticity and amplify its effects; however, (ii) it can also generate chaotic population dynamics that may impede diversification. Our work proposes new hypotheses to explain the spatial patterns of bet‐hedging strategies that we hope will encourage new empirical studies of this topic.     

Short‐term insurance versus long‐term bet‐hedging strategies as adaptations to variable environments

Understanding how organisms adapt to environmental variation is a key challenge of biology. Central to this are bet‐hedging strategies that maximize geometric mean fitness across generations, either by being conservative or diversifying phenotypes. Theoretical models have identified environmental variation across generations with multiplicative fitness effects as driving the evolution of bet‐hedging. However, behavioral ecology has revealed adaptive responses to additive fitness effects of environmental variation within lifetimes, either through insurance or risk‐sensitive strategies. Here, we explore whether the effects of adaptive insurance interact with the evolution of bet‐hedging by varying the position and skew of both arithmetic and geometric mean fitness functions. We find that insurance causes the optimal phenotype to shift from the peak to down the less steeply decreasing side of the fitness function, and that conservative bet‐hedging produces an additional shift on top of this, which decreases as adaptive phenotypic variation from diversifying bet‐hedging increases. When diversifying bet‐hedging is not an option, environmental canalization to reduce phenotypic variation is almost always favored, except where the tails of the fitness function are steeply convex and produce a novel risk‐sensitive increase in phenotypic variance akin to diversifying bet‐hedging. Importantly, using skewed fitness functions, we provide the first model that explicitly addresses how conservative and diversifying bet‐hedging strategies might coexist.     

Spatial pattern of distribution of marine invertebrates within a subtidal community: do communities vary more among patches or plots?

Making links between ecological processes and the scales at which they operate is an enduring challenge of community ecology. Our understanding of ecological communities cannot advance if we do not distinguish larger scale processes from smaller ones. Variability at small spatial scales can be important because it carries information about biological interactions, which cannot be explained by environmental heterogeneity alone. Marine fouling communities are shaped by both the supply of larvae and competition for resources among colonizers—these two processes operate on distinctly different scales. Here, we demonstrate how fouling community structure varies with spatial scale in a temperate Australian environment, and we identify the spatial scale that captures the most variability. Community structure was quantified with both univariate (species richness and diversity) and multivariate (similarity in species composition) indices. Variation in community structure was unevenly distributed between the spatial scales that we examined. While variation in community structure within patch was usually greater than among patch, variation among patch was always significant. Opportunistic taxa that rely heavily on rapid colonization of free space spread more evenly among patches during early succession. In contrast, taxa that are strong adult competitors but slow colonizers spread more evenly among patches only during late succession. Our findings show significant patchiness can develop in a habitat showing no systematic environmental spatial variation, and this patchiness can be mediated through different biological factors at different spatial scales.     

Scale of dispersal in varying environments and its implications for life histories of marine invertebrates

Summary We present several models concerning the short term consequences of spreading offspring in varying environments. Our goal is to determine what patterns of spatial and temporal variation yield an advantage to increasing scale of dispersal. Of necessity, the models are somewhat artificial but we feel they are a reasonable approximation of and hence generalizable to natural systems. With these models we examine consequences of dispersal arising from environmental variation: increased environmental variance, different degrees of spatial and temporal correlation, some arbitrary spatial patterns of favorability and finally some patterns derived from long-term, large-scale weather data collected along a contiguous stretch of coastline from southern Oregon to northern Washington (USA). We examine the costs and benefits of increasing sclae of dispersal in both density dependent and density independent models.Several conclusions may be drawn from the results of these models. In the absence of any spatial or temporal order to favorability (where favorability is directly proportional to either fitness or carrying capacity) increasing scale of spread produces a higher tate of population increase. At larger scales, though, an asymptote of maximum relative advantage is approached, so each added increment of spread has a smaller contribution to fitness. This asymptote is higher and the approach to it relatively slower with increasing environmental variance. For a given environmental variance, increasing spatial correlation results in a slower approach to the same asymptote. In density independent models, increasing temporal correlation of fitness selects against increased dispersal if expected differences between sites are sufficiently great relative to variation within sites; but in this instance, density dependence yields a somewhat different result: dispersers have a refuge at sites of low carrying capacity or sites lacking non-dispersers. Finally, optimum intermediate scales of dispersal can occur where differences in expected fitness increase with increasing distance from the parental site, such as in a gradient, but where the environmental variation at a given site is fairly large relative to differences in expected fitness between adjacent sites.The foregoing results are extended for the following predictions. When greater longevity in a resistant phase of the life cycle reduces temporal variation in survival and fecundity, increased generation time should decrease the benefits of spreading offspring in an environment that would otherwise favor spread and could either increase or decrease the costs of spreading offspring in an environment selecting against spread. We speculate that if large scale patterns of varying survival and fecundity are similar to the variation in the physical environment which we examined with weather data, there should be little or no short term advantage to large scale spread of offspring (on the order of 50 kilometers or more) because expected differences increase and seldom if ever decrease with increasing distance between sites.This suggests that feeding larvae of benthic invertebrates with their concomitant long planktonic period, receive little if any advantage from increased scale of dispersal, and consequently that the advantages to planktotrophy over lecithotrophy must lie in other life history aspects, such as the ability to produce a greater number of smaller eggs.Order of authorship alphabetical and by increasing age and height     

Sex allocation and interactions between relatives in the bean beetle, Callosobruchus maculatus

When a small number of females contribute offspring to a discrete mating group, sex allocation (Local Mate Competition: LMC) theory predicts that females should bias their offspring sex ratio towards daughters, which avoids the fitness costs of their sons competing with each other. Conversely, when a large number of females contribute offspring to a patch, they are expected to invest equally in sons and daughters. Furthermore, sex ratios of species that regularly experience variable foundress numbers are closer to those predicted by LMC theory than species that encounter less variable foundress number scenarios. Due to their patterns of resource use, female Callosobruchus maculatus are likely to experience a broad range of foundress number scenarios. We carried out three experiments to test whether female C. maculatus adjust their sex ratios in response to foundress number and two other indicators of LMC: ovipositing on pre-parasitised patches and ovipositing with sisters. We did not find any evidence of the predicted sex ratio adjustment, but we did find evidence of kin biased behaviour.     

Parental effects and the evolution of phenotypic memory

Despite growing evidence for nongenetic inheritance, the ecological conditions that favour the evolution of heritable parental or grandparental effects remain poorly understood. Here, we systematically explore the evolution of parental effects in a patch‐structured population with locally changing environments. When selection favours the production of a mix of offspring types, this mix differs according to the parental phenotype, implying that parental effects are favoured over selection for bet‐hedging in which the mixture of offspring phenotypes produced does not depend on the parental phenotype. Positive parental effects (generating a positive correlation between parental and offspring phenotype) are favoured in relatively stable habitats and when different types of local environment are roughly equally abundant, and can give rise to long‐term parental inheritance of phenotypes. By contrast, unstable habitats can favour negative parental effects (generating a negative correlation between parental and offspring phenotype), and under these circumstances, even slight asymmetries in the abundance of local environmental states select for marked asymmetries in transmission fidelity.     

How should parents adjust the size of their young in response to local environmental cues?

Models of parental investment typically assume that populations are well mixed and homogeneous and have devoted little attention to the impact of spatial variation in the local environment. Here, in a patch‐structured model with limited dispersal, we assess to what extent resource‐rich and resource‐poor mothers should alter the size of their young in response to the local environment in their patch. We show that limited dispersal leads to a correlation between maternal and offspring environments, which favours plastic adjustment of offspring size in response to local survival risk. Strikingly, however, resource‐poor mothers are predicted to respond more strongly to local survival risk, whereas resource‐rich mothers are predicted to respond less strongly. This lack of sensitivity on the part of resource‐rich mothers is favoured because they accrue much of their fitness through dispersing young. By contrast, resource‐poor mothers accrue a larger fraction of their fitness through philopatric young and should therefore respond more strongly to local risk. Mothers with more resources gain a larger share of their fitness through dispersing young partly because their fitness in the local patch is constrained by the limited number of local breeding spots. In addition, when resource variation occurs at the patch level, the philopatric offspring of resource‐rich mothers face stronger competition from the offspring of other local mothers, who also enjoy abundant resources. The effect of limited local breeding opportunities becomes less pronounced as patch size increases, but the impact of patch‐level variation in resources holds up even with many breeders per patch.     

The transition to social inbred mating systems in spiders: role of inbreeding tolerance in a subsocial predecessor

The social spiders are unusual among cooperatively breeding animals in being highly inbred. In contrast, most other social organisms are outbred owing to inbreeding avoidance mechanisms. The social spiders appear to originate from solitary subsocial ancestors, implying a transition from outbreeding to inbreeding mating systems. Such a transition may be constrained by inbreeding avoidance tactics or fitness loss due to inbreeding depression. We examined whether the mating system of a subsocial spider, in a genus with three social congeners, is likely to facilitate or hinder the transition to inbreeding social systems. Populations of subsocial Stegodyphus lineatus are substructured and spiders occur in patches, which may consist of kin groups. We investigated whether male mating dispersal prevents matings within kin groups in natural populations. Approximately half of the marked males that were recovered made short moves (< 5m) and mated within their natal patch. This potential for inbreeding was counterbalanced by a relatively high proportion of immigrant males. In mating experiments, we tested whether inbreeding actually results in lower offspring fitness. Two levels of inbreeding were tested: full sibling versus non-sib matings and matings of individuals within and between naturally occurring patches of spiders. Neither full siblings nor patch mates were discriminated against as mates. Sibling matings had no effect on direct fitness traits such as fecundity, hatching success, time to hatching and survival of the offspring, but negatively affected offspring growth rates and adult body size of both males and females. Neither direct nor indirect fitness measures differed significantly between within patch and between-patch pairs. We tested the relatedness between patch mates and nonpatch mates using DNA fingerprinting (TE-AFLP). Kinship explained 30% of the genetic variation among patches, confirming that patches are often composed of kin. Overall, we found limited male dispersal, lack of kin discrimination, and tolerance to low levels of inbreeding. These results suggest a history of inbreeding which may reduce the frequency of deleterious recessive alleles in the population and promote the evolution of inbreeding tolerance. It is likely that the lack of inbreeding avoidance in subsocial predecessors has facilitated the transition to regular inbreeding social systems.     

Mating portfolios: bet-hedging,sexual selection and female multiple mating

Polyandry (female multiple mating) has profound evolutionary and ecological implications. Despite considerable work devoted to understanding why females mate multiply, we currently lack convincing empirical evidence to explain the adaptive value of polyandry. Here, we provide a direct test of the controversial idea that bet-hedging functions as a risk-spreading strategy that yields multi-generational fitness benefits to polyandrous females. Unfortunately, testing this hypothesis is far from trivial, and the empirical comparison of the across-generations fitness payoffs of a polyandrous (bet hedger) versus a monandrous (non-bet hedger) strategy has never been accomplished because of numerous experimental constraints presented by most ‘model’ species. In this study, we take advantage of the extraordinary tractability and versatility of a marine broadcast spawning invertebrate to overcome these challenges. We are able to simulate multi-generational (geometric mean) fitness among individual females assigned simultaneously to a polyandrous and monandrous mating strategy. Our approaches, which separate and account for the effects of sexual selection and pure bet-hedging scenarios, reveal that bet-hedging, in addition to sexual selection, can enhance evolutionary fitness in multiply mated females. In addition to offering a tractable experimental approach for addressing bet-hedging theory, our study provides key insights into the evolutionary ecology of sexual interactions.     

The pay-off from superparasitism in the egg parasitoid Trissolcus basalis, in relation to patch defence

1. In species such as quasi-gregarious egg parasitoids, which exploit defendable patches, the fitness pay-off from superparasitism is an important factor in considering the evolution of patch defence and optimal patch defence strategies.
2. When oviposition in a previously parasitized host has a non-zero fitness pay-off, competing parasitoids are not only able to diminish the future value of a patch to other females by depleting unparasitized hosts, but also devalue the previous investment of other females in the patch by superparasitizing.
3. In Trissolcus basalis , an egg parasitoid of pentatomid bugs, the fitness pay-off from superparasitism is higher than 0.5 when the time interval between ovipositions is less than 3 h, suggesting that defending a previous investment will be an important component of patch defence behaviour.
4. The data suggest a mechanism for the high early pay-off, which involves the superparasitizing female adjusting the sex ratio of its offspring in favour of males. Males develop more quickly than females and thus may have a competitive advantage.     

Bet hedging in a warming ocean: predictability of maternal environment shapes offspring size variation in marine sticklebacks

Bet hedging at reproduction is expected to evolve when mothers are exposed to unpredictable cues for future environmental conditions, whereas transgenerational plasticity (TGP) should be favoured when cues reliably predict the environment offspring will experience. Since climate predictions forecast an increase in both temperature and climate variability, both TGP and bet hedging are likely to become important strategies to mediate climate change effects. Here, the potential to produce variably sized offspring in both warming and unpredictable environments was tested by investigating whether stickleback (Gasterosteus aculeatus) mothers adjusted mean offspring size and within‐clutch variation in offspring size in response to experimental manipulation of maternal thermal environment and predictability (alternating between ambient and elevated water temperatures). Reproductive output traits of F1 females were influenced by both temperature and environmental predictability. Mothers that developed at ambient temperature (17 °C) produced larger, but fewer eggs than mothers that developed at elevated temperature (21 °C), implying selection for different‐sized offspring in different environments. Mothers in unpredictable environments had smaller mean egg sizes and tended to have greater within‐female egg size variability, especially at 21 °C, suggesting that mothers may have dynamically modified the variance in offspring size to spread the risk of incorrectly predicting future environmental conditions. Both TGP and diversification influenced F2 offspring body size. F2 offspring reared at 21 °C had larger mean body sizes if their mother developed at 21 °C, but this TGP benefit was not present for offspring of 17 °C mothers reared at 17 °C, indicating that maternal TGP will be highly relevant for ocean warming scenarios in this system. Offspring of variable environment mothers were smaller but more variable in size than offspring from constant environment mothers, particularly at 21 °C. In summary, stickleback mothers may have used both TGP and diversified bet‐hedging strategies to cope with the dual stress of ocean warming and environmental uncertainty.     

Variability in thermal and phototactic preferences in Drosophila may reflect an adaptive bet‐hedging strategy

Organisms use various strategies to cope with fluctuating environmental conditions. In diversified bet‐hedging, a single genotype exhibits phenotypic heterogeneity with the expectation that some individuals will survive transient selective pressures. To date, empirical evidence for bet‐hedging is scarce. Here, we observe that individual Drosophila melanogaster flies exhibit striking variation in light‐ and temperature‐preference behaviors. With a modeling approach that combines real world weather and climate data to simulate temperature preference‐dependent survival and reproduction, we find that a bet‐hedging strategy may underlie the observed interindividual behavioral diversity. Specifically, bet‐hedging outcompetes strategies in which individual thermal preferences are heritable. Animals employing bet‐hedging refrain from adapting to the coolness of spring with increased warm‐seeking that inevitably becomes counterproductive in the hot summer. This strategy is particularly valuable when mean seasonal temperatures are typical, or when there is considerable fluctuation in temperature within the season. The model predicts, and we experimentally verify, that the behaviors of individual flies are not heritable. Finally, we model the effects of historical weather data, climate change, and geographic seasonal variation on the optimal strategies underlying behavioral variation between individuals, characterizing the regimes in which bet‐hedging is advantageous.     

High Altitude Frogs (Rana kukonoris) Adopt a Diversified Bethedging Strategy in the Face of Environmental Unpredictability

Environmental unpredictability can influence strategies of maternal investment among eggs within a clutch. Models predict that breeding females should adopt a diversified bet-hedging strategy in unpredictable environments, but empirical field evidence from Asia is scarce. Here we tested this hypothesis by exploring spatial patterns in egg size along an altitudinal gradient in a frog species(Rana kukunoris) inhabiting the Tibetan Plateau. Within-clutch variability in egg size increased as the environment became variable(e.g., lower mean monthly temperature and mean monthly rainfall at higher altitudes), and populations in environments with more unpredictable rainfall produced eggs that were smaller and more variable in size. We provide support for a diversified bet-hedging strategy in high-altitude environments, which experience dynamic weather patterns and therefore are of unpredictable environmental quality. This strategy may be an adaptive response to lower environmental quality and higher unpredictable environmental variance. Such a strategy should increase the likelihood of breeding success and maximize maternal lifetime fitness by producing offspring that are adapted to current environmental conditions. We speculate that in high-altitude environments prone to physical disturbance, breeding females are unable to consistently produce the optimal egg size due to physiological constraints imposed by environmental conditions(e.g., duration of the active season, food availability). Species and populations whose breeding strategies are adapted to cope with uncertain environmental conditions by adjusting offspring size and therefore quality show a remarkable degree of ability to cope with future climatic changes.     

Should mothers provision their offspring equally? A manipulative field test

Within‐brood variation in offspring size is universal, but its causes are unclear. Theoretical explanations for within‐brood variation commonly invoke bet‐hedging, although alternatives consider the role of sibling competition. Despite abundant theory, empirical manipulations of within‐brood variation in offspring size are rare. Using a field experiment, we investigate the consequences of unequal maternal provisioning for both maternal and offspring fitness in a marine invertebrate. We create experimental broods of siblings with identical mean, but different variance, in offspring size, and different sibling densities. Overall, more‐variable broods had higher mean performance than less‐variable broods, suggesting benefits of unequal provisioning that arise independently of bet‐hedging. Complementarity effects drove these benefits, apparently because offspring‐size variation promotes resource partitioning. We suggest that when siblings compete for the same resources, and offspring size affects niche usage, the production of more‐variable broods can provide greater fitness returns given the same maternal investment; a process unanticipated by the current theory.     

The use of conspecific reproductive success for breeding patch selection in terrestrial migratory species

Classical models of breeding habitat selection rarely deal with the question of information gathering for patch quality assessment. In this paper, we present two models comparing the fitness outcomes of behavioural strategies based on conspecific reproductive success as a cue to assess local environmental quality before selecting a new breeding habitat. The models deal with two phases of the life-cycle of a territorial migratory species: recruitment to a breeding population (model 1) and breeding site fidelity of subsequent breeding attempts (model 2). The first model shows that prospecting breeding patches before recruiting is the best strategy if the environment is predictable and contains a low proportion of good patches, even if it implies losing a breeding opportunity. The second model shows that dispersing after a breeding attempt according to the patch's breeding success rather than the individual's own success is the bests own success is the best strategy if the environment is patchy. These results underline the importance of studying the spatio-temporal variations of factors affecting reproductive success when considering the importance of habitat selection strategies based on conspecifics. Moreover, they allow the understanding of individual behaviour patterns observed in natural populations and their potential consequences at the metapopulation level.     

Fast life history traits promote invasion success in amphibians and reptiles

Competing theoretical models make different predictions on which life history strategies facilitate growth of small populations. While ‘fast’ strategies allow for rapid increase in population size and limit vulnerability to stochastic events, ‘slow’ strategies and bet‐hedging may reduce variance in vital rates in response to stochasticity. We test these predictions using biological invasions since founder alien populations start small, compiling the largest dataset yet of global herpetological introductions and life history traits. Using state‐of‐the‐art phylogenetic comparative methods, we show that successful invaders have fast traits, such as large and frequent clutches, at both establishment and spread stages. These results, together with recent findings in mammals and plants, support ‘fast advantage’ models and the importance of high potential population growth rate. Conversely, successful alien birds are bet‐hedgers. We propose that transient population dynamics and differences in longevity and behavioural flexibility can help reconcile apparently contrasting results across terrestrial vertebrate classes.     

How to balance the offspring quality–quantity tradeoff when environmental cues are unreliable

Maternal effects on offspring size can have a strong effect on fitness, as larger offspring often survive better under harsh environmental conditions. Selection should hence favour mothers that find an optimal solution to the offspring size versus number tradeoff. If environmental conditions are variable, there will not be a single optimal offspring size, as predicted in a constant environment, but plastic responses can be favoured. To be able to adjust offspring size in an adaptive manner, mothers have to use environmental cues to predict offspring environmental conditions. Cues can be unreliable, however, particularly in species where individuals occupy different niches at different life stages. Here we model the evolution of plasticity of offspring size when the environmental cues mothers use to predict the conditions experienced by their offspring are not perfectly reliable. Our results show that plastic strategies are likely to be superior to fixed strategies in a stochastically varying environment when the environmental cues are at least moderately reliable, with the threshold depending on plasticity costs and the difference of resources available to mothers. Plasticity is more likely to occur if resource availability is not too different between environments. For any given scenario, plasticity in offspring size is favoured if offspring survival varies greatly between environmental states. Whenever plastic strategies are optimal, the occurring switches performed by mothers between small and large offspring are predicted to be substantial, as small adjustments are unlikely to reap fitness benefits great enough to overcome the costs of plasticity.     

The effects of serotiny and rainfall-cued dispersal on fitness: bet-hedging in the threatened cactus <Emphasis Type="Italic">Mammillaria pectinifera</Emphasis>

Serotiny—the retention of seeds in the mother plant for over a year—in unpredictable environments may increase the probability that at least some seeds are dispersed during favorable periods. Propagules may be expelled when environmental cues announcing favorable conditions occur, or be gradually released into the environment. This could be a bet-hedging strategy increasing the long-term fitness by reducing interannual variability in reproduction. However, the impact of seed retention on the population dynamics of serotinous species and its contribution to fitness has been barely explored under field conditions. We assessed these issues in the threatened Mammillaria pectinifera, a small globose cactus that gets established only in exceptionally rainy years. This species expels some seeds actively during unusually rainy periods, while dispersing others passively over several years. Dynamics of the seeds in the mother plant over two very contrasting years in terms of precipitation was incorporated into a stochastic matrix model. Seed retention was found to increase significantly the probability that some of the seeds retained in any given year are dispersed within a subsequent rainy period. Active seed-expulsion raises this probability even further. As expected in bet hedgers, seed retention increased fitness in the presence of temporal variability. Active fruit expulsion did not affect fitness, but reduced demographic stochasticity. The incomplete serotiny and fruit expulsion observed is the evolutionary outcome expected for the environment and life-history attributes of the species.