Relationships between the presence of odonate species and environmental characteristics in lowland ponds of central Italy

A set of 21 ponds was sampled three times for odonate larvae during spring 2002. At the same time 17 environmental variables were recorded including area, wet phase duration, total nitrogen, total phosphorus, aquatic macrophytes and land use. A total of 16 odonate species belonging to Lestidae, Coenagrionidae, Aeshnidae and Libellulidae were recorded, and the total number of species per pond varied from zero to six. The relationships between species richness, assemblages and environmental variables were studied by simple and multiple correlation and by Canonical Correspondence Analysis (CCA). The results showed that permanent ponds were larger, deeper, had more macrophyte species, had more extensive macrophytes cover and lower concentrations of nitrogen and phosphorus than temporary ponds. Multiple regression analysis showed that the number of odonate species was positively affected firstly by the number of macrophyte species, and then by pond depth. However, pond depth appeared to be interchangeable with several others variables, such as pond area and water duration and negatively correlated with nitrogen concentration, variables which are all linked with the permanent or temporary status of the ponds. CCA analysis indicated that odonate species presence was linked with a few environmental variables, showing a tendency of Odonata to avoid ponds with higher nitrogen concentrations, with the exception of Lestes barbarus, a species typical of temporary water in central Italy. At the same time, the majority of species were linked with longer water phase duration and with greater macrophyte species richness. A comparison with previous studies, and in particular with those carried out in central Italy, confirmed the positive influence of macrophytes, water duration, and also the negative effect of nutrient load. However, several other variables, in particular land use, shade, presence of fish, which were influential in other studies, were not significant in this study.