Using diatoms as indicators of ecological conditions in lotic systems: A regional assessment

Yangdong Pan, R. Jan Stevenson, Brian H Hill, Alan T. Herlihy, Gary B. Collins

Research output: Contribution to journalArticlepeer-review

247 Scopus citations


Benthic diatoms and water chemistry were sampled from 49 stream sites in the Mid-Atlantic Highlands region of the United States to evaluate the use of diatoms as indicators of environmental conditions in streams across varying geographic and ecoregional areas. Diatom samples were collected from depositional and erosional habitats in a randomly selected reach in each stream site. Patterns of diatom species distributions in relation to environmental variables were determined using canonical correspondence analysis. Diatom species in both habitats were highly correlated with a pH gradient. A second gradient was correlated with variables that were commonly associated with agricultural runoff such as turbidity and total phosphorus. The relationship between diatoms and major environmental variables was quantified with regression and calibration models. The correlation between diatom-inferred and observed pH was high (r 2 = 0.90). Cross-validation with jackknifing showed that pH models were reasonably robust (r 2 = 0.69 for depositional habitats, r 2 = 0.67 for erosional habitats). The regression and calibration models for the depositional habitats bad only slightly higher predictive powers than those of erosional habitats. The relationship between diatoms and important environmental variables was robust and quantifiable, and the sensitivity of diatom assemblages to environmental conditions did not differ between erosional and depositional habitats. Therefore we concluded that diatoms can be used as quantitative indicators of environmental conditions in lotic systems.

Original languageEnglish (US)
Pages (from-to)481-495
Number of pages15
JournalJournal of the North American Benthological Society
Issue number4
StatePublished - Dec 1996


  • Mid-Atlantic Highlands streams
  • benthic diatoms
  • biomonitoring
  • canonical correspondence analysis (CCA)
  • depositional and erosional habitats
  • pH
  • regression and calibration model


Dive into the research topics of 'Using diatoms as indicators of ecological conditions in lotic systems: A regional assessment'. Together they form a unique fingerprint.

Cite this