The observation that genetic variation within a species may be correlated with the ecological properties of associated communities has done much to integrate our understanding of ecological pattern and evolutionary process. The theoretical underpinning to this field of community or ecosystem genetics considers the individual’s environment as one that comprises other interacting individuals, which is therefore akin to a genotype–environment interaction in which the environment is made up of genes of others. In communities, similar interactions among species are known as interspecific indirect genetic effects (IIGEs). This thesis examines some unresolved issues in community genetics using a series of field experiments in a benthic marine system, using the encrusting bryozoan, Hippopodina iririkiensis, as interacting species.
That prior residence by a species influences community assembly is widely accepted, but the role of priority effects in the context of natural environmental heterogeneity is rarely considered. In my first experiment, I found that variation in substrate orientation explained the great majority of variation in community assembly. Where present, priority effects of a resident species usually interacted with environmental effects, such that the impact of a resident species on community assembly varied with orientation and also spatially.
In spite of increasing empirical evidence for IIGEs on community assembly, their multigenerational impact remains comparatively unresolved. Persistent, directional change in communities deriving from IIGEs of a resident species can occur only when community properties covary genetically with fitness in genotypes of those species. Estimates of such covariance are crucial for understanding the potential for eco-evolutionary feedback between resident species and their communities, but remain rare. In my second experiment, genotypes of Hippopodina explained up to 35% of variation in community assembly in the field. Uniquely, I detected significant covariances between the performance of genotypes and the abundances of other community members, representing an indirect genetic effect of Hippopodina that can potentially drive multigenerational dynamics of co-existing species. Under these circumstances, different genotypes will associate with different communities across generations.
IIGEs have rarely been considered in the context of natural environmental variation. In a third factorial experiment, effects of genotypic variation in Hippopodina on community assembly were greatly outweighed by the effects of variation in substrate orientation: less than 5% of variation in any community response was explained by IIGEs, compared to as much as 73% of that explained by orientation. Genetic influence on community assembly was isolated to an indirect effect on one species of the community that significantly covaried with the abundance of another in a single orientation. Therefore, indirect genetic effects on community assembly may be slight relative to the effects of environmental heterogeneity, and driven by comparatively isolated effects on only a minority of species present in an assemblage. At the same time, IIGEs may show genotype–environment effects, confounding straightforward prediction of response to selection in heterogeneous environments.
Last, I examined IIGEs within the context of variation in the successional stage of communities. In nature, primary space is rare, and species recruiting in a landscape are subject to powerful forces of environmental filtering and species interactions that may limit the potential for indirect effects to act on communities with an established successional trajectory. I found variation in community age (0, 3, or 6 weeks old) played a pivotal role in explaining variation in diversity, abundance and species composition of communities, whereas IIGEs explained little (0–7% of the variation modelled). Idiosyncratic responses among a minority of species were, again, the likely cause of a unique genotype–environment interaction, whereby significant covariances of sabellid worms with Bugula stolonifera changed sign between 0 and 6 week-old communities, but showed no covariance in 3-week old communities.
Overall, my results suggest that the contribution of priority effects to the structuring of communities may be secondary to the effects of environmental variation. Within that fraction of variation explained by the presence of a focal species, a small proportion again is explained by genetic variation within that species. Mismatches in the scale of genetic to environmental effects had different implications in relation to each of the two sources of environmental variation examined. In response to the first environmental factor, substrate orientation, genotypes may influence communities in different ways in different orientations, but selection in one orientation will indirectly select in another. In relation to successional stage, in contrast, genetic responses to variation in the age of communities were limited, but showed altered correlations among communities at different successional stages. Accordingly, in response to fine-scale variation in this second environmental factor, IIGEs are likely to have idiosyncratic effects on assembly in this system, showing reduced effects in established communities. Overall, even where the relative contribution of genetic effects on communities may be less than those generated by environmental variation, non-additive effects may occur, confounding simple predictions about the effects of different sources of variation.