In the context of the global decline in biodiversity there is a pressing need for simple methods to assess biodiversity and community composition. Identification of phytoplankton to species level is difficult, expensive and time-consuming and requires high levels of expertise. Thus, the search for parsimonious predictors of organism diversity based on simplified taxonomy or approaches representing community structure, has received much attention. Few studies have focused on identifying surrogates for predicting both phytoplankton richness and community composition. Here we examined the suitability of several taxonomic and ecological classifications in summarising phytoplankton diversity and community structure from 87 stratified-random selected Andalusian artificial ponds. Taxa based approaches at genus and family level, as well as functional groups predicted relatively well both phytoplankton richness and assemblage composition. Size classes could be used as a reasonable predictor of richness and environmental conditions, but it was a weak predictor of community composition. The morphology-based approach was the poorest proxy for richness patterns and environmental conditions, but more suitable than the size class approach as a proxy for assemblage composition.