Feeding diversity in macroinvertebrate communities: A contribution to estimate the ecological status in shallow waters

The feeding diversity of subtidal samples of macroinvertebrates from Ria Formosa was estimated with Shannon–Wiener information index and complementary evenness. The results were compared with other commonly used methodologies under the European Water Framework Directive, such as diversity indices, AMBI and ITI.Assuming that in a healthy environment all feeding groups are present, and that no group clearly dominates, the feeding diversity is expected to be maximal and the evenness feeding diversity will be close to 1. In degraded environments some feeding groups might be absent or having low relative abundance, and generally with one or two groups dominating the community. In this way the evenness feeding diversity index would measure deviations from expected values due to a degradation of the environment. Although confirmation of this approach needs to be tested in other shallow waters, the results obtained show interesting features.To each of the 297 species belonging to the Ria Formosa data matrix a feeding group was assigned, among six groups: surface deposit feeders, subsurface deposit feeders, herbivores, suspension feeders and suspension/deposit feeders (species which have the two feeding modes depending on food availability). The carnivorous, parasites, omnivorous and scavengers were all grouped together, forming the sixth group. Most of the stations of Ria Formosa showed high feeding diversity, which could correspond to a good or high ecological status (ES) except at one location, that occasionally showed low feeding diversity. This poor condition was essentially due to low water renewal and extreme environmental variation of some parameters, such as salinity. At some locations an intermediate feeding diversity was observed mainly due to natural accumulation of organic matter. Other commonly used indices also point out to the same tendencies.We propose the evenness feeding diversity estimate approach as a practical and apparently robust method to estimate the ES of shallow waters, which can be used together with other common indicators. This approach has also the advantage of showing low sensibility to small samples and to low taxonomic identification effort.