The effects of environmental variation on bryophytes at a regional scale

The distribution of bryophytes in central Belgium was investigated using species grid-mapping superimposed on a series of maps which included information on soil conditions and land use. Our objectives were to assess the influence of environmental variation on the bryoflora at a regional scale, to examine how bryophytes respond to environmental variation, and to assess the extent to which species ecological and life-history traits determine the accuracy of the predictability of species occurrence in order to provide comprehensive lists of species based on environmental conditions. The first two axes of a correspondence analysis (CA) of the floristic data explained 14.6% of the total chi(2). CA1 was significantly correlated with loamy-sandy soils on a sand layer (r = -0.74, p < 0.001), forest cover (r = -0.80, p < 0.001), loamy soils (r = 0.79, p < 0.001), and agricultural fields cover (r = 0.61, p < 0.001). CA2 had a low but significant correlation coefficient with pebbly soils cover (r = 0.38, p < 0.001). The probability of occurrence of 59% of the investigated species could be significantly predicted by logistic regression from the sets of environmental variables. About 55% of the species exhibited an increasing probability of occurrence with increasing forest cover and loamy-sandy soils cover, 1% with agricultural fields and loamy soils cover, and 3% with pebbly soils cover. The predictability of species occurrence varied as a function of four life-history traits (minimum spore size, life expectancy, type of gametophyte and papillose leaf cell walls) and three ecological traits (indicator values of light, temperature and soil acidity). The most predictable species, including a number of leafy liverworts, were characteristic for acidic, fresh and shaded conditions and displayed a strong preference for forest habitats. Taxa with limited predictability included epiphytes and mosses characteristic of pebbly soils due to the ability of these species to efficiently disperse and adapt to various ecological conditions. Species for which the distribution range could not successfully be predicted were either ubiquitons, characteristic for ephemeral habitats, or highly successful in a very common habitat.