Interobserver error in grassland vegetation surveys: sources and implications

Aims Observer error is an unavoidable aspect of vegetation surveys involving human observers. We quantified four components of interobserver error associated with long-term monitoring of prairie vegetation: overlooking error, misidentification error, cautious error and estimation error. We also evaluated the association of plot size with pseudoturnover due to observer error, and how documented pseudochanges in species composition and abundance compared with recorded changes in the vegetation over a 4-year interval.