Assessing Invertebrate Responses to Human Activities: Evaluating Alternative Approaches

Abstract

The goal of biological monitoring is to evaluate the effect of human activities on biological resources. In this study, we linked human activities across landscapes to specific changes in assemblages of benthic macroinvertebrates in streams that drain those landscapes. We used data from 2nd- to 4th-order streams in southwestern Oregon to test approximately 30 hypotheses about how macroinvertebrates respond to several common human actions, especially logging and associated road construction. We found 10 attributes of macroinvertebrate assemblages to be reliable indicators of disturbance. Data from a subsequent year confirmed those results. We used simple graphical methods to evaluate land-use data and to relate the data to observed responses of invertebrates. We constructed a multimetric index (benthic index of biological integrity, or B-IBI) from component metrics that distinguished disturbed stream sites from minimally disturbed sites. Using an independent data set, we found that B-IBI scores were significantly lower for streams whose watersheds were more degraded by human activities. We also tested rapid bioassessment protocol (RBP) III as modified by Oregon Department of Environmental Quality. RBP III failed to detect differences among sites that B-IBI did detect Because biologists continue to debate the relative merits of a multimetric vs. a multivariate approach for interpreting biomonitoring data, we also used principal components analysis (PCA) to explore patterns in our multidimensional data. We chose PCA based on species abundance because it has been the most common technique used by state and federal agency biologists to interpret biomonitoring data. PCA failed to detect clear differences in our data set (e.g., between most and least disturbed sites). We suggest that multivariate statistical analyses are most appropriate for exploratory analysis when the investigator has limited knowledge of an ecological system and wants to generate testable hypotheses. We demonstrate that sufficient information about the natural history of stream invertebrates is available to support more direct tests of how invertebrates respond to human disturbance. The components of a good multimetric index are founded on tested hypotheses and thus provide a sound scientific assessment of a stream site. We argue that a method that incorporates ecological information is more suitable for biomonitoring than one relying on statistical algorithms.