Rapid Buffer Assessment Fails to Predict and Classify Wetland Floristic Quality in Oklahoma

Abstract

Rapid field assessment indicators of ecological stress and disturbance may or may not reflect actual biological conditions. We tested the ability of rapidly assessed buffer quality to predict and classify wetland floristic quality across 106 sites in Oklahoma, USA. We used buffer zone metrics from a national-level rapid assessment tool (USA-RAM) to evaluate applicability at the state level. The rapid assessment relying on buffer quality did not relate predictably (linearly) with floristic quality, but the relationship improved after ecoregional stratification. Focusing on the Central Great Plains ecoregion with 55 wetlands sampled, we found 16 floristically intact sites and 13 floristically altered sites based on multivariate testing of the Floristic Quality Assessment Index and several component metrics. Distributions of buffer metrics across these sites were similar between the floristically intact and altered classes, suggesting potential for misclassification of wetland biological condition using rapid buffer assessment. Some of the disconnect may be attributed to bias in the buffer measurements and to using a national assessment protocol at the state level. Buffer metrics in the USA-RAM may offer little insight about wetland biological condition in some regions, and the use of such metrics should only complement direct bioassessments and not substitute for them.