A probability-based indicator of ecological condition

Robert W. Howe, Ronald R. Regal, Gerald J Niemi, Nicholas P. Danz, Jo Ann M. Hanowski

Research output: Contribution to journalArticlepeer-review

28 Scopus citations

Abstract

We introduce a new method for quantifying the ecological condition (C) of sites based on documented species' responses to environmental stress. Preliminary research is needed to establish species-specific logistic functions, representing probabilities of finding individual species across an explicit reference gradient, ranging from maximally stressed (C = 0) to minimally stressed (C = 10) localities. Each function takes into account the species' tolerance to stress, the species' overall ubiquity, and the probability of detecting the species when it is present. Given a set of standardized species-specific functions, the ecological condition of any site can be derived by iteration, converging on the value of C that best "predicts" the species that are actually present. Species from multiple taxonomic groups can be included in the calculations, and results are not directly affected by species richness or sampling area. We demonstrate a successful application of this method for bird species assemblages in the U.S. portion of the Great Lakes coastal zone. Approximately, 28% of the bird species observed in the Eastern Deciduous Forest Ecological Province and 35% of the species in the Laurentian Mixed Forest Ecological Province showed strong relationships with a reference gradient of land cover variables. Functional stress-response relationships of these species can be used effectively to estimate ecological condition at new sites. The estimated condition based on bird species generally mirrors the reference condition, but deviations from the expected 1:1 relationship provide meaningful insights about ecological condition of the target areas. Sensitivity analysis using different numbers of species shows that our method is robust and can be applied consistently with 25-30 species exhibiting strong stress-response functions.

Original languageEnglish (US)
Pages (from-to)793-806
Number of pages14
JournalEcological Indicators
Volume7
Issue number4
DOIs
StatePublished - Nov 1 2007

Keywords

  • Bird
  • Environmental stress
  • Great Lakes
  • Indicator method
  • Landscape
  • Probability model

Fingerprint Dive into the research topics of 'A probability-based indicator of ecological condition'. Together they form a unique fingerprint.

Cite this