Observer Variability in Classifying Habitat Types in Stream Surveys

We report on the ability of trained observers to independently classify habitat units within stream reaches into primary (pools, riffles, and glides) and secondary (types of pools and types of riffles) habitat types. Differences among observers in classifying habitat types increased with the number of habitat types and decreased with level of observer training. Observer variability also seemed to he affected by reach-specific physical attributes, such as gradient and the amount of wood in the stream channel. Attempts to classify stream habitats will be more consistent and useful if observers receive sufficient uniform training and are required to distinguish between fewer habitat types.