| Summary: | Neighbouring plants can interact strongly, competing for resources including light, water, animal mutualists, and local germination sites. From an evolutionary perspective, this implies that a plant’s best resource acquisition strategy will usually depend on the traits of its neighbours, and for plants in particular, neighbours are often genealogical relatives. Here, I use a combination of theory and experiments to expose some important consequences of social interactions among plants. The first model analyzes selection on traits used to attract pollinators, showing that competitive interactions (in the absence of local relatedness) can select for exaggerated secondary sexual characters. To complement this model, I performed experiments that confirm the mechanisms by which adaptive pollinator foraging naturally leads to interactions among plants. The observed foraging behaviour (of bumble bees) also provides unique evidence for ‘Bayesian foraging’, a sophisticated type of resource assessment that depends on prior experience in a particular environment. A second model considers how selection on the sex allocation of cosexual, animal-dispersed plants leads to competition and cooperation over local germination sites, sometimes leading to the origin of gender dimorphism. The model reveals novel ecological contexts in which disruptive selection on sex allocation can arise, and in general, illustrates how selection for cooperation can facilitate or inhibit evolutionary diversification. In the models considered here, cooperation is indiscriminant, but plants might also assess the relatedness of neighbours and cooperate with kin over non-kin. In the final chapter, I present experimental evidence that is consistent with preferential cooperation over soil resources among sibling plants. This study is the first to link a potentially cooperative resource allocation strategy with an increase in the mean fitness of related plants.
|
|---|
