Detecting the effects of water regime on wetland plant communities: Which plant indicator groups perform best?

• We examined vegetation composition and frequency data from 51 plots.
• We tested for trends at species and functional group level based on hydrology.
• Three functional group classification methods were compared.
• Water plant functional group and growth form abundances showed the clearest trends.

Water regime is a primary driver of patterns in wetland vegetation composition. Differences in composition can be used as indicators of differences in water regime. We used vegetation point intercept data collected from 51 wetland monitoring plots in the Blue Mountains, south-eastern Australia, to determine which of three indicator group classifications, growth forms, water plant functional groups (WPFGs) or wetland indicator categories (WICs), demonstrated the most consistent differences between vegetation communities from plot sample groups differing in location (wetland edge or core) and surface water availability (typically inundated or damp). PERMANOVA tests showed significant differences between core and edge plot communities analysed by growth form or WIC relative frequencies, but only when tree canopy data (higher in edge plots, which were abutting woodland) was included. Significant differences in communities (PERMANOVA, p ≤ 0.02) were detected between inundation categories for all classification methods when tree data were included, but not for WIC data when tree data were excluded. Overall, ordination plots and ANOSIM R values showed the most consistent community-level differences (least overlap in sample groups) between inundation categories when data were classified by WPFGs, followed by growth forms. ANOVA tests on individual indicator group relative frequencies showed that WPFG classification provided the most indicator groups differing significantly in relative frequency between inundation categories, with these groups also collectively comprising a much higher proportion of the total vegetation recorded per plot than the growth forms or WICs that differed between categories.