Bryophyte response surfaces along ecological and climatic gradients

Bivariate response surfaces were fitted to abundance data for 6 mire bryophyte species in ecological and climatic space in western Canada. Response surfaces were derived by gridding data, eliminating outliers, and reducing effects of overly influential abundance values. Predicted values at each grid node were calculated as distance-weighted means of all observed values within a predetermined radius of each node. Response surfaces accurately delimited species ecological and climatic ranges. The gridding process consistently underestimated observed abundance values, and paired t-tests indicated significant differences between predicted and observed values for Sphagnum fuscum and Tomenthypnum nitens in ecological space, and for Sphagnum fuscum, Tomenthypnum nitens, and Scorpidium scorpioides in climatic space. Predicted values were significantly different only for those species that had large data sets, and for species that had clusters of points with an inordinately large number of data points within them. However, with the exception of Sphagnum tenellum, predicted and observed values were significantly correlated for all species in both ecological and climatic space. R² analyses for the linear relationship between predicted and observed values can be used as a diagnostic tool to determine the effectiveness of the gridding process to accurately predict observed abundance values.Abbreviations p statistical probability