Human disturbance and boreal vascular plant biodiversity in Alberta, Canada

Abstract

The worldwide biodiversity crisis has intensified the need to better understand how biodiversity and human disturbance are related. Yet this relationship lacks both consensus in theoretical expectations and consistency in observed empirical patterns. I present one of the largest extent studies of human impacts on boreal plant biodiversity to date, in the boreal ecoregion of Alberta, where disturbance in sites range ranged from 0 – 100 % area disturbed by varying land use types including agricultural fields, forestry cut-blocks, and petroleum extraction. The ‘intermediate disturbance hypothesis’ (IDH) suggests that disturbance regimes generate predictable non-linear patterns in species richness. But evidence often contradicts IDH at small scales and is generally lacking at large regional scales. I show that across a broad region species richness peaked in communities with intermediate anthropogenic disturbance, as predicted by IDH, even when accounting for many environmental covariates. IDH was consistently supported across trees, shrubs, forbs, and grasses and with temporary and perpetual disturbances. However, only native species fit this pattern; exotic species richness increased linearly with disturbance. A fundamental impediment to understanding the diversity-disturbance relationship is that both diversity and disturbance can depend on the scales at which they are sampled. To test the dependence of species richness on disturbance scale and the scale-dependence of the intermediate disturbance hypothesis, I hold the area over which species richness is measured constant and instead manipulate the area over which human disturbance is measured. I show the shape of richness-disturbance relationship is consistent across scales, but predictions of richness depend on the scale at which disturbance was measured. I also explore impacts of human disturbance on community composition and structure to both identify the risks facing communities and to assess the potential utility of these metrics for monitoring applications. In particular, I explore ranked species occupancy curves, species functional trait dispersion, and species specialization in relation to anthropogenic disturbance extent. Disturbed communities differed in functional traits, particularly in fruit and seed characteristics, but community structure was similar across communities regardless of disturbance class, and both species specialization and functional dispersion of traits were not strongly related to disturbance.