PARENT SESSION
Poster Session #15: Vegetation Analysis.
Tuesday, August 7, 2001. Presentation from 10:30 AM to 12:00 PM. Exhibition Hall

Testing the scale-dependence of species associations among ferns.

KARST, JUSTINE¹, ARII, KEN¹, LECHOWICZ, MARTIN¹, ¹

ABSTRACT- We studied the constancy of associations among fern species at varying scales within a single hectare of unusually high species diversity. The hectare contains 19 of the 38 fern species found in a 10 km² tract of old-growth deciduous forest at Mont St. Hilaire, Quebec. We surveyed the percent cover of fern species in each m² of this hectare as the basis for a series of DCA ordinations at different spatial scales. At each scale (defined by quadrats of 4, 9, 16, 25, 36, 49 and 64m²) the data were resampled at stratified random positions within the hectare to generate a series of ordinations; we then assessed the mean and variance of species relationships from the ordinations at a given scale and across scales. The total area sampled was held constant for each scale, therefore the number of ordinations increased as scale decreased. Observed patterns were compared to those from a null model of random distribution of species among m² within the hectare. Associations among the fern species are non-random; the nature and stability of these associations change with scale. We identify an optimal scale for sampling associations among these ferns and affirm the need to consider scale in evaluating associations among species.

KEY WORDS: ferns, DCA ordination, scale