Revealing the biodiversity of wild bees (Hymenoptera: Apoidea: Anthophila) in flower strips in Mediterranean floodplains. Which monitoring method fits best?

Context The ongoing pollinator decline may threaten and compromise the resilience of terrestrial ecosystems. Implementing conservation action requires monitoring pollinator populations' actual status, but this is particularly difficult for pronubes insects such as wild bees. Their monitoring is difficult and time-consuming but crucial for assessing their health status. Objectives Here we compared and evaluated the efficiency of three different monitoring methods to evaluate wild bee biodiversity in lowland areas sown with entomophilous flowers to support pollinating insects in a Mediterranean environment. Methods We sampled wild bees for two years by using hand netting, pan traps and artificial nests. We compared species richness and abundance among these methods with a particular focus on how flowering coverage affects the efficiency of walking transects and pan traps and discussed the attractiveness of the different colours of pan traps. Results Hand netting captured a higher abundance of wild bees than the other two methods but a similar number of species to pan traps. Artificial nests captured fewer specimens and species. Bee assemblages were significantly different between pan traps and hand netting, and pan traps had greater potential in capturing the whole bee biodiversity, but their attractiveness is negatively influenced by the flowering coverage contrary to hand netting sampling. Conclusions Like other studies, the three sampling methods are complementary regarding species assemblages. The juxtaposition of several monitoring methods is essential to assess the biodiversity status of species with such particularly different ecological traits.