Odonata communities in retrodunal ponds: a comparison of sampling methods

Dragonflies are commonly used as indicators of environmental quality and different methods have been employed to monitor odonate assemblages, such as surveys of all adults, evaluations based on breeding adults, sampling of larvae and collection of exuviae. Results obtained with different sampling methods may not be interchangeable, as the different life stages (e. g. larvae, adults) differ in mobility (aquatic, aerial) and as they are subjected to different ecological constraints. Therefore generalization about habitat quality based on only one survey method might be questionable. Additionally, detectability of species might vary when different methods are used. In this study, nine retrodunal ponds in the Migliarino, San Rossore, Massaciuccoli Regional Park (Tuscany, Italy) were repeatedly and contemporaneously sampled during May-September 2008 with the following methods: all adults, breeding adults, larvae and exuviae. In total, 22 species were detected and the results showed that the four methods were not interchangeable. First, some species were only found using certain methods. Second, univariate measures of diversity obtained with the four sampling methods were considerably different. Alpha diversity was maximal when computed on all adults and minimal with exuviae; breeding adults and larval collection had intermediate values. Beta diversity showed an inverse trend, with the lowest value for "all adults" surveys and higher values for all the others. Finally, congruence among the assemblages revealed by the four methods was generally low. The results show that the four survey techniques are not interchangeable and that monitoring of Odonata has to be based on a carefully chosen method, which should reflect the aim of the study.