Combined use of redundancy and indicator species analyses for the selection of optimal pressure-specific bioindicators: a case study on Carabidae

This article presents a method for identifying highly effective bioindicator species within a taxonomic group, even in the absence of full knowledge of their ecology or tolerance to specific stresses. This method was tested using trapping data of ground beetle species (Coleoptera: Carabidae) in organic and conventional hazelnut Corylus avellana) orchards in order to select the species most sensitive to plant protection products. The dataset includes 10,565 individuals belonging to 57 different ground beetle species, collected in the Piedmont and Latium regions (Italy), where hazelnut cultivation is widespread. A comparison was made between the results obtained through Redundancy Analysis conducted on the entire sample and the same analysis conducted on a portion of the sample obtained by selecting, through the Indicator Value Index, the species with the highest bioindication capacity for the considered pressure. The species were ordered according to the difference of the Indicator Value Index of each, respectively for conventional and organic cultivation, and a battery of Redundancy Analyses was conducted by progressively excluding the species with lower bioindication capacity. The results with the best combination of explained inertia and statistical significance (permutation test) were then selected, defining this approach as Backward Redundancy Analysis Sequence (BRS). Partial Redundancy Analyses were subsequently conducted to isolate the impact of plant protection products from that produced by other factors. Regarding the bioindication capacity of ground beetles on the effects of plant protection products on hazelnut orchards, the results obtained confirm a significantly clearer and more consistent response by applying the BRS compared to the classical approach.