A standardised method for estimating environmental and agronomic covariates to discriminate the explanatory variables effects on bioindicators: a case study on soil fauna

A method was developed for the standardisation and analysis of environmental and agronomic covariates to discriminate the effects of specific explanatory variables on a given bioindicator. To test it, the effects of plant protection products (PPP) was assessed on soil fauna sampled in organic and conventional hazelnut orchards. More than 100 standardised covariates were numerically reduced, by Principal coordinates analysis (PCoA), to two derived covariates. Then, redundancy analysis (RDA) was applied using, as explanatory variables, two indexes referred to PPP input and the derived covariates. The results showed a marked differentiation of the soil fauna communities between the two groups of sampled sites and their clear response to the use of PPP. The procedure proved to be effective in reducing the 'background noise' determined by a great number of covariates. This method can be successfully applied in monitoring activities concerning the effects on biodiversity of several initiatives aimed at reducing PPP use.