Identification of dragonflies (Odonata) as indicators of general species richness in boreal forest lakes

We argue the need to select indicator species on empirical data to avoid influence of personal opinions. To test an empirical selection process based on a nested subset matrix, we sampled partivoltine dragonfly larvae from 74 small lakes in central Sweden. A nestedness matrix was set up using the 'nestedness temperature calculator' program, selecting 11 species as potential indicators of species richness. These were tested against a known indicator species for water quality (the pool frog) and plant diversity through inventories and comparison to existing surveys of biological values ('rich' lakes vs. 'ordinary' lakes). We could only see a trend towards the pool frog occurring in dragonfly-rich lakes, but found a significant connection between the number of aquatic plants along the shore line and the number of dragonfly species present. A significantly higher number of indicators were encountered in lakes previously surveyed as 'rich' in plants than in lakes classified as 'ordinary'. Dragonfly species richness therefore appears to be positively associated with species richness of vascular plants. We propose nestedness matrices to be a good selecting tool for indicator species, particularly in groups where the biology of the species is not well known. However, it is important to define what such indicators really indicate.