| The population of stream reference sites in south-western Ontario was defined by considering 6979 candidate sampling locations' exposures to agricultural stress. Least-disturbed sites (those occupying several catchment-scale-stressor variables' percentile-based least-impacted tails) were considered reference sites. On the basis of benthic community composition, water chemistry, and exposure to agricultural stress, the condition of these Percentile-based Reference Sites was compared against that of a set of reference sites selected using best professional judgement, and against test sites known to be impacted by agriculture. Percentile-based Reference Sites more closely aligned with natural land-cover than Professional-judgement Reference Sites and test sites did, their pollutant concentrations were lower, and their benthic-invertebrate communities showed fewer signs of impairment. Given this evidence of their least-disturbed condition, a random selection of Percentile-based Reference Sites was used to set biocriteria (i.e., to compile a table of critical values for selected benthic-invertebrate-community indices), which may be used in bioassessments of south-western Ontario streams.Keywords: reference sites, criteria for least disturbed, biocriteria, agricultural streams, Ontario, Canada, biomonitoring, reference condition, benthic macroinvertebrates, benthic-macroinvertebrate indices. |
