Beware your neighbours: interactions shape population dynamics in natural grassland communities
Thesis DisciplineBiological Sciences
Degree GrantorUniversity of Canterbury
Degree NameMaster of Science
Interactions between species have long been recognised as being of major importance to patterns of diversity and abundance, yet our knowledge of how interactions operate and vary in natural plant communities remains incomplete. In such diverse system, interactions are both numerous and highly variable, and they depend on both abiotic and biotic context. This makes quantifying interaction strengths between species in natural communities a difficult and complex task, but it is also a necessary one if we are to uncover the driving forces underpinning this variation.
Species interactions are an important component of grassland population dynamics, where plants compete for sunlight, space, nutrients and water. Furthermore, the short generation times of most grassland plant species makes them a useful system for examining the effects of interactions. The National Vegetation Survey databank, operated by Landcare Research, hosts time-series of grassland plant abundances from several hundred plots and collected over the span of several decades. I proposed to exploit the potential of this data set in order to examine the effects of interactions on plant population dynamics.
My first hypothesis was that including species interactions in models of population dynamics would improve our ability to predict changes in plant abundance in natural grassland communities. To test this hypothesis, I compared the fits of a series of models which included interactions in increasingly complex and biologically meaningful ways. I found that including interactions improved predictions for the overwhelming majority of species. The most successful model included every pairwise interaction, which allowed us to estimate measures of interaction strengths between species. In turn, the distribution of these interaction strengths provided insights into potential community-level differences in stability. The results presented here suggest these differences were driven by elevation, which weakened interactions, and the presence of exotic species, who tended to have stronger interactions than native species.
My second study investigated how interactions between species varied along multiple abiotic gradients. I estimated interaction strengths between focal species and four guilds of competitors — forbs, graminoids, woody species and non-woody others — occurring over a range of elevations and latitudes. Both gradients had varying effects on the mean interaction strengths of each competitor guild. On average, increasing elevation and latitude made interactions with forbs facilitative, whereas interactions with graminoids and woody species became less facilitative and more competitive. This variation in interaction effects might be a reflection of the different optima conditions for each group of species.
Together, the work that makes up this thesis suggests that interactions between species can be regarded as an important driver of changes in plant abundance in these grassland communities. Plantplant interactions should be included in models of population dynamics in order to improve predictions of changes in abundance. Furthermore, including interactions also uncovers how variable their effects are, to both environmental conditions and identity of the interaction partner. In particular, the relationships between elevation, species functional guild and biological status affected species interactions in complex, and at times unexpected ways. This has important implications for our understanding of how plant interactions shape grassland community dynamics, and thereby how these communities might respond to biotic and abiotic threats.