Ecology, Evolution, and Organismal Biology
A fluctuating environment drives coexistence in five non-pollinating fig wasps
Author(s): Duthie, AB, Abbott, KC, Nason, JD
The principle of competitive exclusion states that species competing for identical resources cannot coexist, but this appears paradoxical given myriad ecologically similar competitors. Theory shows coexistence is possible if a fluctuating environment changes the competitive dominance of species, but only when environmental variation leads to non-linear or non-additive population growth. Non-additivity facilitates coexistence when competition and environment covary in their effects on growth such that competition is weaker in poor environments. The variation required to facilitate coexistence is typically interpreted as a physical aspect of environment. We investigate a system that might be considered constant under such interpretations, but show a hidden source of temporal variability relying on fluctuating dispersal distances between resource patches. We model populations that use resources in discrete, ephemeral, patches. Such ephemeral patch systems often support many competing species. We frame our model with respect to the highly diverse non-pollinating fig wasp communities that oviposit and develop within fig fruit. Using numerical and individual-based models, we show that temporal storage of larval wasps leads to long-term community coexistence under a wide range of biologically realistic parameter values when wasps face a trade-off between dispersal ability and fecundity. We empirically test whether or not such a trade-off exists in a community of five competitor species of non-pollinators associated with Ficus petiolaris. We find strong evidence of a negative correlation between species dispersal abilities and fecundities, which we present as an extreme case in which a single fluctuating environmental variable appears to mediate coexistence in a community of competitors. We suggest that fluctuating environmental conditions may drive coexistence more generally, especially among competitors regularly dispersing between ephemerally available habitat patches.
Department of Biodiversity and Evolutionary Biology
Cichlid evolution in crater lakes and ecological opportunity
Author(s): Barluenga, M, Magalhaes, IS
Ecological opportunity fuels the generation of biodiversity. When empty habitats are colonized ecological release favors species niche expansion and eventually the divergence of taxa via intraspecific resource competition and character release. The very young cichlid fish radiations from the Nicaraguan lakes in Central America are a powerful model for the study of very rapid diversification. Multiple recently formed crater lakes exist in this area, and all of them have been independently colonized by a subset of the fish fauna from the larger and older close by Nicaraguan Great Lakes. This setting is excellent to test the idea of ecological and character release following ecological opportunity. To this end we have studied the entire fish fauna of several crater lakes and characterized the ecological role of each of the species. We investigate the consequence of the absence of some of these species in the community for the remaining species, and the ability of the latter to expand their niche, release their morphological design and ultimately speciate.
Departement of Ecology and Evolution
Co-evolutionary branching of dispersal and sociality in structured populations
Author(s): Mullon, CDL, Keller, L, Lehmann, L
Dispersal has antagonistic effects for the evolution of altruism, cooperation and social behavior. On one hand, dispersal of individuals from their native patches decreases relatedness between locally interacting individuals, and thus disfavors the evolution of social traits. On the other hand, dispersal reduces local competition among kin, thereby favoring cooperation. It is traditionally thought that these two antagonistic effects balance, and as a consequence, an intermediate level of dispersion and sociality evolves. However, recent numerical experiments have suggested that when social traits and dispersal evolve together, evolutionary branching may take place, leading to the stable coexistence of social and asocial individuals with different dispersal strategies. In order to clarify the effects of dispersal on the evolution of cooperation, we developed analytical tools to study the evolution of multiple phenotypic traits in structured populations. By tracking changes in the phenotypic distribution in a population over time, we are able to explicitly predict the conditions that lead to evolutionary branching, and thus the coexistence of different dispersal and cooperating strategies. More generally, our method can be applied to study the emergence of highly differentiated life histories in structured populations.
Department of Zoology
Disentangling the social, parental and genetic influences on natal dispersal in great tits
Author(s): Garroway, CJ, Hinde, C, Verhelst, B, Sheldon, BC
Natal dispersal is a key process underpinning the structure and dynamics of populations. Individual variation in dispersal behaviour is substantial, but we know very little about the causes of individual variation in dispersal, and the extent to which this variation is influenced by social processes. Here, we integrate longitudinal data collected over five decades from a wild great tit (Parus major) population with two cross-fostering experiments to disentangle the influence of social processes, parental effects and genetics on dispersal behaviour. We show first that parental dispersal phenotype, whether they are themselves locally hatched or immigrants, has scale-independent effects on dispersal by offspring. Birds with immigrant parents dispersed further within patches and were more likely to disperse outside patches. Using an index of the composition of early social environments, with reference to the immigrant and locally hatched status of neighbours, we then show that dispersal is independent of the local social environment in which birds are raised. We used two large-scale cross-fostering experiments to demonstrate that parental effects on dispersal are primarily intrinsic to offspring, and therefore independent of parental behaviour. Finally, we show that parental dispersal phenotypes show similar fledging success but differing rates of local recruitment of offspring, implying that dispersal phenotypes will be genetically structured across landscapes. Our findings suggest that understanding the underpinning genetics of dispersal will be important for understanding the behaviour of populations in fragmented landscapes. Non-random dispersal of particular types of individuals suggests scope for the emergence of fine-scale population structure and has important consequences for interpretations of selection studies and dispersal theory.
Evolution, Ecology and Genetics
Dispersal in a world with edges: hypermobility, inertia, and the population cost of evolution
Author(s): Shaw, AK, Jalasvuori, M, Kokko, H
Dispersal is a fundamental characteristic of organisms -- it determines the spatial, genetic, and demographic structure of population and is necessary for the persistence of any species. However, from the perspective of the individual, dispersal often carries substantial costs that can come in the currencies of energetics, time, risk and opportunities. Often the dispersal strategy favored by individuals will not be the strategy that would be ‘best’ for the population (best in terms of maximizing size or distribution). This leads to one of two results 1) individuals disperse less than is best for the population (“inertia”) or 2) individuals disperse more than is best for the population (“hypermobility”). Models of dispersal are often constructed to avoid edge effects, however, the real world is full of edges. A major risk of dispersal, the possibility of landing in unfavorable habitat, becomes exacerbated in habitats that are not only potentially fragmented but also have sharp boundaries (e.g., an archipelago). One might therefore predict inertia to be a problem in worlds of small size where most habitat is near an edge. Here we construct a model to address these questions. We find that small worlds, small patch carrying capacities, and high dispersal mortality all lead to inertia whereas large worlds, large patch carrying capacities, and low dispersal mortality lead to hypermobility. We also show that although heterogeneity in individual dispersal behavior can increase long-distance dispersal events, this is not guaranteed to be so. In a subset of parameter space heterogeneity actually decreases the probability of long-distance dispersal, in contrast to previous thinking.
Dispersal: causes, mechanisms and population consequences
Author(s): Clobert, J
Dispersal is a three steps process. The social and non social environments can act either similarly or dissimilarly at all these stages. We will review the actual knowledge on dispersal ecology and evolution and try to sketch priorities for future research. We will also try to examine how dispersal might interact with global changes and in turn influence the dynamic of fragmented populations.
Department of Ecology and Evolution
Fitness returns of inflorescence architecture in a wind pollinated annual plant
Author(s): Santos del Blanco, L, Tudor, E, Pannell, JR
Male fitness in outcrossing wind-pollinated plants is indirectly influenced by the efficiency of pollen delivery. Many wind-pollinated plant species present adaptations for efficient pollen dispersal. A typical example is provided by Mercurialis annua, where male individuals disperse pollen from erect inflorescences held above the plant. In contrast, hermaphrodite individuals usually release their pollen from sessile axillary inflorescences (axillary hermaphrodites), but pollen production and the length of inflorescences on hermaphrodites in some populations resembles that found in males (pedunculate hermaphrodites). The ability of males to invade and establish in populations with hermaphrodites is expected to depend directly on their relative ability to disperse pollen successfully. Thus, enhanced pollen production and dispersal ability by pedunculate hermaphrodites might prevent male invasion. Here, we estimated the pollen production and siring success of males and hermaphrodites of M. annua that varied in their inflorescence architecture in experimental mating arrays. Males sired four times more offspring when growing in populations with axillary hermaphrodites than with pedunculate hermaphrodites. We use these results, and those from previous work, to model the fate of males and hermaphrodites with different inflorescence architectures. Our results have important implications for the maintenance of males with hermaphrodites in this species (androdioecy). They also raise questions about the genetic architecture and the geographic distribution of inflorescence morphology in M. annua.
Hybrid songbirds employ intermediate routes in migratory divide
Author(s): Delmore, KE, Irwin, DE
Seasonal migration may play a significant role in speciation; many divergent populations breed adjacent to one another while using different routes to reach their wintering grounds (i.e., form migratory divides). Migratory orientation is largely genetically determined in these populations and often involves navigation around geographic barriers. Accordingly, hybrids have been predicted to employ intermediate routes that are inferior to those of parental forms. We provide the first direct test of this hypothesis here, by attaching light-level geolocators to birds in a narrow hybrid zone between two groups of Swainson’s thrushes that form a migratory divide in western North America. Most of these birds employed intermediate routes, navigating over large geographic barriers. The remainder employed a mixed strategy, using the same route as one parental form on fall migration and the other on spring migration. Data from age ratios and cline analyses further suggests that hybrids survive at lower rates than parental forms and are selected against. Together, these results provide strong support for the migratory divide hypothesis and represent one of few established examples in which a behavioral trait reduces hybrid fitness, thereby promoting speciation.
Department of Biosciences
Life-history of the Glanville fritillary butterfly in fragmented versus continuous landscapes
Author(s): Duplouy, A, Ikonen, S, Hanski, I
Habitat loss and fragmentation threaten the long-term viability of innumerable species of plants and animals. At the same time, habitat fragmentation may impose strong natural selection on populations and lead to evolution of life-histories with possible consequences for demographic dynamics. The Baltic populations of the Glanville fritillary butterfly (Melitaea cinxia) inhabit landscapes with highly fragmented habitat (networks of small dry meadows) as well as landscapes with extensive continuous habitat (calcareous alvar grasslands). Here, we report the results of common garden studies on butterflies originating from two highly fragmented landscapes in Finland and Sweden and from two continuous landscapes in Sweden and Estonia, conducted in a large outdoor population cage and in the laboratory. We investigated a comprehensive set of 63 life-history traits, including components of larval development, adult behavior and reproductive performance. Habitat fragmentation shows association with several life-history traits. Most notably, the growth rate of the post-diapause larvae and several measures of flight capacity were higher in butterflies from fragmented than continuous landscapes. These results support theoretical predictions about high rate of population turnover in fragmented habitats selecting for increased rate of dispersal. Females from continuous landscapes had shorter intervals between consecutive egg clutches and somewhat higher life-time egg production, but shorter longevity, than females from fragmented landscapes. These results are likely to reflect the constant opportunities for oviposition in females living in continuous habitats, and possibly also the more dispersive females from fragmented landscapes allocating more resources to dispersal capacity at the cost of egg maturation rate
Quantifying the demographic cost of selection in a changing environment
Author(s): Reed, TE, Gienapp, P, Visser, ME
The lag load, or reduction in mean fitness caused by incomplete adaptive tracking of a moving phenotypic optimum, is a key quantity in models of evolution in changing environments. Despite its central theoretical importance, empirical studies quantifying the lag load and the factors affecting it are lacking. Here we explore these issues in a Dutch population of great tits Parus major, where warming springs have generated a mismatch between annual breeding time and the timing of an important food resource, providing an ideal opportunity to examine how the resulting directional selection for earlier breeding time impacts population mean fitness and related demographic parameters. First, we show that interannual variation in mismatch over almost four decades has surprisingly not affected population growth, despite it having led to intensified directional selection. We demonstrate an important mechanism contributing to this uncoupling, whereby fitness losses associated with mismatch are counteracted by fitness gains due to relaxed competition. Next, we parameterised a quantitative genetic model to predict the theoretical ‘critical rate of environmental change’, beyond which increased maladaptation leads to population extinction. Results imply that even ‘mild’ rates of climate change would be close to the critical rate. However, individual-based simulations that account for both evolutionary processes and density dependence revealed that the expected time to extinction, although highly uncertain, is on the order or centuries, rather than decades. These findings imply that microevolution would only rescue the population from mild rates of sustained climate change. On the other hand, they also illustrate that considerable maladaptation can be demographically tolerated in the short term through density dependence, without immediate population declines, effectively ‘buying time’ for microevolution to restore adaptation until the environment (climate) stabilises.