Environmental effects on variation in leaf morphology, stem length, and leaf density of Andreaea blyttii are found to be statistically significant. This study is based on a numerical analysis of the morphological variation using principal components analysis (PCA). The results from this analysis were correlated with environmental parameters using generalized linear models (GLM) and generalized additive models (GAM) with a cubic smooth spline estimator. The magnitude of the morphological variation was large, and four specimens differed considerably from the others in their extremely broad leaves and relatively short limbs. This deviation in leaf shape is not correlated with any environmental variation, and may therefore be due to a rare genetic combination. Although there is lack of significant environmental diference, the morphological variation related to the main bulk of specimens shows a significant distribution along a complex moisture gradient composed of elevation, snow-persistence, shelter, slope, and aspect. With increased availability of moisture the leaves become longer, more curved, wider, with longer cells in the basal marginal region, wider costae, and longer stems, whereas the number of leaves per stem decreases. This increased relative growth is, however, not correlated with the frequency of occurrence along the moisture gradient. The lack of correlation is interpreted as a decrease in the probability of survival and/or establishment in areas of autumnal high water availability, although a higher degree of growth seems to be characteristic of such habitats. There was also a negative correlation of the above-mentioned morphological parameters with the calcium concentration in the supporting rock. Together with a reduced number of occurrences observed under such conditions, this indicates a strong avoidance of substrates with medium or high calciumcontent. Although there is a significant relationship between the morphology and environment, there is considerable residual variation in the statistical models. Adjustments with geographical distribution as an estimate of large scale genetic differences reduced the residual variation somewhat, but it was still a considerable part of the model. This residual variation is interpreted as being caused either by genetic differences on a spatially local scale or by a plastic morphological response to annual and inter-annual environmental history. Variation in morphology of mosses in relation to environmental characteristics of their habitats has, for a long time, been of interest to bryologists. There has been a considerable effort in classification of growth-forms (Gimmingham & Birse 1957; Gimmingham & Roberts 1950) and in branching patterns (La Farge-England 1996; Proctor & Smith 1995). These variations have been correlated with the environmental conditions under which particular species occur (Birse 1957, 1958a, b; Gimmingham & Birse 1957), and strong relationships have been identified between different forms and water availability within their habitats (Maigdefrau 1982). However, more recently emphasis has been put on the phenology, life expectancy, and dispersal ability, for the classification of life forms in mosses (During 1992). Studies related to morphological structures and their significance in relation to the environment have also been undertaken, e.g., the relationship between papillae and water flux (Procr 1984), and the significance of leaf surface wax to gas exchange (Proctor 1979). Other general investigations into the relationship of bryophyte structure to environmental factors stress morphological adaptations and significance (Longton 1980; Proctor 1982, 1984). However, most of these papers consider differences at an interspecific level, and the environmental characterizations are often presented in general terms at relatively crude scales. There are to my knowledge, few studies that consider intraspecific morphological variation of bryophytes (Ststad & Flatberg 1993) in relation to the environment to which these specimens have been exposed, based on quantitative field observations. The primary aim of this paper is to analyze the morphological variation of Andreaea blyttii Schimp, and to correlate this intraspecific variation with the environmental features of the sites supporting these plants. This analysis is based on finescale environmental measurements which have 0007-2745/97/308-323$1.75/0 This content downloaded from 207.46.13.101 on Sat, 08 Oct 2016 06:10:13 UTC All use subject to http://about.jstor.org/terms 1997] HEEGAARD: ANDREAEA BLYTTII 309