COMMUNITY STRUCTURE (NICHE LIMITATION AND GUILD PROPORTIONALITY) IN RELATION TO THE EFFECT OF SPATIAL SCALE, IN A NOTHOFAGUS FOREST SAMPLEDWITH A CIRCULAR TRANSECT

Citation
Cm. Bycroft et al., COMMUNITY STRUCTURE (NICHE LIMITATION AND GUILD PROPORTIONALITY) IN RELATION TO THE EFFECT OF SPATIAL SCALE, IN A NOTHOFAGUS FOREST SAMPLEDWITH A CIRCULAR TRANSECT, New Zealand journal of ecology, 17(2), 1993, pp. 95-101
Citations number
28
Categorie Soggetti
Ecology
ISSN journal
01106465
Volume
17
Issue
2
Year of publication
1993
Pages
95 - 101
Database
ISI
SICI code
0110-6465(1993)17:2<95:CS(LAG>2.0.ZU;2-Z
Abstract
A Nothofagus-dominated rainforest in eastern Fiordland, New Zealand, w as sampled by shoot frequency in contiguous 1 x 1 m quadrats, along a topologically-circular transect. The data were analysed at five scales up to 5 x 1 m, to search for assembly rules, i.e., generalised restri ctions on species co-occurrences. There was no evidence of niche limit ation in terms of the whole community, at any scale examined. Rather, variance in species richness was greater than expected from a null mod el, suggesting environmental heterogeneity. This conclusion was confir med by using a patch-model. Guild structure was examined in terms of s ynusiae, based on vertical stratification. The proportion of species i n a quadrat that were from the Herb guild was significantly more const ant than expected under the null model. However, using a patch-model t his effect was reduced, and no longer significant. This suggests pools of species adapted to different micro-environments, but with parallel guild composition. Liane guild proportions were more variable than pr edicted by the null model; although this effect was not significant, i t parallels reports from other South Island forests. The work confirms conclusions from previous work, that plant assembly rules are to be f ound only at small spatial scales. Processes leading to high and low v ariance in species richness and in guild proportions are discussed. Se lection of a uniform site in the present work, and use of a patch-mode l, make it less likely that effects are due to environmental or histor ical effects. The use of guild proportionality as the criterion elimin ates the effect of limitation on the number of individuals that can co -occur locally.