QUADRAT COVARIANCE ANALYSIS AND THE SCALES OF INTERSPECIFIC ASSOCIATION DURING PRIMARY SUCCESSION

Three term covariance analysis is used to investigate interspecific association at a range of spatial scales, using both density and presence/absence data to evaluate the strength of species interactions. This procedure also investigates the relationship between the spatial patterns of species. These methods are applied to the vegetation on a series of glacial moraines near Mt. Robson, British Columbia, Canada, in order to test whether the plants of different species and their patterns become more independent and more random during succession. Another hypothesis tested is that if a species exerts a positive influence on others by increasing soil nutrients, this effect decreases with surface age. The vegetation analysed supports both these hypotheses to the extent that most species pairs have maximum covariance intensity on medium aged surfaces. The covariance-blocksize graphs derived from density data and from presence/absence data were qualitatively similar in their trends; however, the results are sufficiently different to suggest that the relationships among species can be subtle, affecting density rather than presence. The results also show that the affect of scale on species associations is more complex than simple models predict.