Methodological framework towards the synthesis of multiple biological indices of soil quality 

<p>A multi-factorial parametric characterization of soil health is crucial to monitor the correct provision of soil ecosystem services (ES) and to ensure its preservation over time in the current climate change scenario. In this framework, the implementation of biological indicators is still lagging the established use of physical and chemical indicators for soil quality assessment. Moreover, different biological groups are often analysed separately and not like a unique set of dynamic compartments that necessarily interact with each other, mutually determining soil biodiversity. Different initiatives at EU level have been undertaken to improve soil biodiversity monitoring practices, to finally promote their integration into Member States&#8217; national soil quality monitoring policies. Nevertheless, few initiatives clearly aimed at evaluating the multiple level integration of soil health monitoring indices. at multiple level, and established practices for determining and communicating multi-level soil biological health are still scarce.</p><p>The aim of this work is to ultimately develop a framework for the synthesis of multiple indices of soil quality, focused on biological indicators, to determine soil health and monitor its provision of ES. This work dwells at the intersection of the EXCALIBUR and EJP-Soil (MINOTAUR) European projects action and consortia. As a proof-of-concept, we selected "nutrient cycling" as ES to be evaluated. We identified the soil chemical and physical indicators that could be used for the monitoring of this ES under real field conditions within different pedoclimatic contexts, and we used those variables to build an "a priori" knowledge set and model of the nutrient cycling features in the considered samples. Next, we selected a set of possible biological indicators at different scales (Ii.e. synthetic indices or specific community components) and for different levels of soil biota (Ii.e. meso-fauna, micro-fauna, bacterial and fungal microbiota). Finally, we investigated the relationship between biological and chemical-physical indicators with the working hypothesis of identifying the most robust relationships. Different unsupervised data analysis methodologies were assessed for their ability in the evaluation ofuncovering the relationship between abiotic and biotic variables, also paying attention to the possibility of clear representation of the data and results for improved communication</p>