Indicators of what? Linking plant indicator species to wetland ecological condition

Background/Question/Methods

Humans have a nearly ubiquitous impact on global ecosystems, and we rely on monitoring to understand, quantify, and sometimes rectify these impacts. To assess the overall condition of an ecosystem, monitoring programs often include field observations of biota that respond to environmental stressors. However, these taxa-environment relationships are not always demonstrated. Previous work in Alberta, Canada identified vascular plants associated with human development, but has yet to link their occurrence to wetland ecology. Here, we sought to determine which wetland characteristics could be explained by the occurrence of these presumptive indicator species (IS). We used published lists of IS and publicly available data on fourteen wetland ecological characteristics relating to vegetation diversity (e.g. total vascular plant richness, richness of culturally important species), soil biogeochemistry (e.g. organic matter depth, total soil carbon), and water physiochemistry (e.g. total nitrogen, dissolved oxygen). We then tested for linear relationships between these wetland characteristics and IS richness. Finally, we repeated the analyses using only a subset of IS (i.e. those with the highest indicator values) to determine if a rapid assessment could provide a similar characterization of wetland ecology as the intensive assessment.

Results/Conclusions

We identified 1255 bogs, fens, marshes, shallow lakes, and wet meadows with measures of at least one of the 14 ecological characteristics of interest. The strength of the relationships between IS richness and wetland ecological characteristics varied among wetland classes, and depended on the type of IS and which characteristic was being predicted. In general, IS richness was a strong predictor of vegetation diversity; a moderate, but infrequent, predictor of soil biogeochemistry; and a poor predictor of water physiochemistry. For example, the richness of high-development IS was strongly positively correlated with the richness of culturally important species in bogs (p < 0.001, R2=0.68). Organic matter depth increased with low-development IS richness, but only in bogs (p < 0.001, R2=0.29) and fens (p < 0.001, R2=0.34). Across all wetland classes, water pH declined as low-development IS richness increased, but predictive ability was low (p < 0.050 for all wetland classes; R2 ranged from 0.04 in fens to 0.17 in marshes). Importantly, models that used the richness of only the ten strongest IS showed similar relationships with wetland characteristics as models that used the richness of all IS. This suggests that these strong IS can provide valuable, cost-effective information about a range of wetland ecological characteristics.