Decreased competitive interactions drive a reverse species richness latitudinal gradient in subarctic forests

Abstract

The tendency for species richness to decrease toward the poles is one of the best-characterized patterns in biogeography. The mechanisms behind this pattern have received much attention, yet very few studies have investigated very high-latitude communities. Here, using data from 134 permanent sample plots from 60 degrees to 68 degrees N, we show that boreal forest plant communities in northwestern Canada increase in richness toward the poles, despite a strong increase in climatic harshness. We hypothesized three possible explanations for this pattern: (1) historical biogeography, (2) reduced competition for light at high latitudes (biotic interactions), and (3) changes in soil characteristics with latitude. We used multidimensional scaling to investigate the community composition at each site and found no clustering of communities by latitude, suggesting that historical biogeography was not constraining site diversity. We then investigated the mechanisms behind this gradient using both abiotic (climate and soil) and biotic (tree stand characteristics) variables in a multiple factor analysis. We found that the best predictor of species richness is an environmental gradient that describes an inverse relationship between temperature and tree-stand density, suggesting that reduced competition for light due to reduced tree growth at low temperatures at higher latitudes allows greater species richness. This study shows that low energy availability and climatic harshness may not be limiting species richness toward the poles, rather, abiotic effects act instead on the strength of biotic interactions.

Mean conditions predict salt marsh plant community diversity and stability better than environmental variability

“…Species turnover decreased with increases in minimum temperature on both coasts but was unaffected by variability in temperature. We also observed greater richness in marshes with colder long‐term means of minimum temperatures on the east coast of North America, unlike most ecosystems in which richness increases at low latitudes (Wallace , Schall and Pianka , Hillebrand , but see Canepuccia et al , Marshall and Baltzer ). Turnover was lower at higher mean temperatures, potentially because greater richness results in more potential species that can arrive or leave a site in any year.…”

Aspect and slope influence plant community composition more than elevation across forest–tundra ecotones in subarctic Canada

“…Although they did not explicitly survey forest–tundra ecotones, Lippok et al. () found that local topography exerted a larger influence on plant community composition than elevation across a 600‐m gradient in Bolivia, and Marshall & Baltzer () found a complete lack of clustering of communities by latitude in ordination space when they surveyed plant communities across a latitudinal gradient extending from 60° to 68° N in northwestern Canada. These studies indicate that our findings are not restricted to elevational gradients in topographically complex regions.…”

Discovering biogeographic and ecological clusters with a graph theoretic spin on factor analysis

“…Despite the fact that step‐across improves multidimensional scaling, it is of no use for CA, PCA, FA or more modern methods (Walker and Jackson , Warton et al ) because none of them operate directly on a distance matrix. Now, CA and PCA can be mimicked by PCoA because they respectively assume that samples are separated by Euclidean and chi‐squared distances.…”

“…Warton et al () explained how Bayesian methods could be used to generate latent variable models from ecological data, which is to say, they showed that FA could be reinvented as a special case of Bayesian modelling. Warton et al () were primarily interested in using latent variables to predict abundance matrices from environmental variable matrices.…”