Grantee Research Project Results

Enhancing a landscape assessment with intensive data: a case study in the Upper Juniata watershed.

Citation:

Hychka KC, Wardrop DH, Brooks RP. Enhancing a landscape assessment with intensive data: a case study in the Upper Juniata watershed. Wetlands 2007;27(3):446-461.

Abstract:

A multi-level approach to wetland assessment and monitoring has been developed to incorporate information from multiple spatial scales and varying levels of effort. In this approach, wetland condition is evaluated in an intensive assessment through detailed, on-site measurement of physical and biological condition, and is inferred in a landscape assessment from a wetland's landscape setting characterized with available spatial data. This study assessed a comprehensive set of landscape metrics to improve an existing landscape assessment using wetland condition measures from the Upper Juniata intensive assessment data. On-site measures of wetland state (n = 10) were compared with landscape metrics (n = 47) measured at multiple spatial scales using Pearson's correlation coefficients. Landscape metrics enhanced the existing landscape assessment if they were correlated with condition metrics not correlated with the existing landscape assessment. Finally, landscape metrics identified through the correlation analysis were used to place sites in categories of condition based on the Floristic Quality Assessment Index (FQAI) using classification and regression tree analysis (CART). Results showed the existing landscape assessment metric is correlated with multiple measures of wetland state. The study identified landscape metrics that could enhance the existing landscape assessment, including measures of near-stream land use measured at an upstream scale, the percent of agriculture on steep slopes in a 250-m-radius circle or upstream area, and a measure of interior forest measured at a 250-m landscape circle or an upstream scale. Finally, the CART analysis showed the prediction of the FQAI was significantly (p < 0.001) improved by the addition of the landscape metrics identified in this study.