Rare species drive local trait diversity in two geographically disjunct examples of a naturally rare alpine ecosystem in New Zealand

• Published in: Journal of vegetation science Vol. 23; no. 4; pp. 626 - 639
• Main Authors: Richardson, Sarah J., Williams, Peter A., Mason, Norman W. H., Buxton, Rowan P., Courtney, Shannel P., Rance, Brian D., Clarkson, Beverley R., Hoare, Robert J. B., St. John, Mark G., Wiser, Susan K.
• Format: Journal Article
• Language: English
• Published: Blackwell Publishing Ltd 01-08-2012
• Subjects: Assembly rules; Biological taxonomies; Diversity partitioning; Ecosystems; Functional diversity; Granite gravel field; Gravel; Leaf nutrient concentrations; Phylogenetic diversity; Phylogenetics; Plant functional trait; Plants; Rare species; Species; Species diversity; Synecology; Vegetation
• ISSN: 1100-9233; 1654-1103
• DOI: 10.1111/j.1654-1103.2012.01396.x

Question: Is there evidence for similar community assembly processes in two geographically disjunct examples of a rare alpine ecosystem? Location: Two alpine granite gravel fields 610 km apart along a fault line in western South Island, New Zealand — the Lookout Range and Mt Titiroa. Methods: Plot-based vascular plant composition and traits for 86 species (height, seed length, leaf size and nutrient concentrations) were used to examine community structure reflecting assembly processes. We examined species richness, trait values averaged across species and plots, relationships between trait pairs using standardized major axis (SMA) regression and phylogenetically independent contrasts (PIC), relationships between abundance and trait values, and functional diversity indices against null models. Lastly, we partitioned taxonomic, functional and phylogenetic diversity within and across alpine ranges. Results: Functional and phylogenetic turnover of diversity between the two locations was low (2.9% and 6.3%, respectively) relative to turnover of taxonomic diversity (75%). Species-level traits were similar between the two locations, except leaf P, which was higher at Mt Titiroa. Plot-level average traits were all significantly higher on Mt Titiroa. Relationships between species-level traits were typically non-significant at both locations, or significant only at a single location. In contrast, relationships between plot-level trait values were frequently significant and consistent across the two locations. At both locations, dominant species had a narrow range of similar trait values, while rare species had a wide range of values. Within plots, we found both trait convergence and divergence. Associations of trait values among dominant and rare species were largely non-random but inconsistent between the two locations. Conclusions: The two locations had a similar pool of trait values but often differed in how those traits were assembled. Results demonstrate that dominant species are subject to rigorous filters during community assembly leading to trait convergence, but rare species contributed to trait diversity and trait divergence. Models of community assembly can often predict traits of dominant species but must consider the importance of rare species as a source of local trait diversity, even in species-poor communities under severe environmental conditions.