Field scale biodiversity in relation to the intensity grade of agricultural practices on dairy farms

The present study has been aimed at finding relationships, at the field level, between indicators of the intensity of agricultural practices, computed as carbon footprint (CF), nutrient balance, and intensity use of agrochemicals and indicator-species groups, selected to be representative of biodiversity, such as herbaceous vascular plants, carabid beetles (Coleoptera: Carabidae), and butterflies (Lepidoptera: Papilionoidea). Fields were selected and monitored over 2 years, on six adjacent dairy farms, to represent three cropping systems and five different intensity levels of agricultural practices: mono-cropped maize with high (M-high) or low (M-low) chemical pressure; maize in rotation (M-rot) with alfalfa (Alf) and low chemical pressure; herbaceous strips with no chemical pressure (Strip). Plants (through whole field monitoring), carabids (through pitfall traps) and butterflies (through transects) were sampled in fields and field margins; CF was computed by considering emissions linked to direct and indirect CO2, CH4 and N2O; N, P2O5 and K2O balances, as the difference between the total imported and exported nutrients, were computed at the field level; the agrochemical use intensity was assessed by frequency of application and Load indexes for non-target organisms. The CF was lower in Alf (84 kg CO2-eq/t DM) and Strip (11 kg CO2-eq/t DM) than in all the maize fields (6357, 4499, and 4059 kg CO2-eq/t DM for M-high, M-low and M-rot respectively). The highest nitrogen surplus was observed in M-high (+413 kg/ha), while it was intermediate for M-low and M-rot and the lowest in Alf. The M-high fields showed the highest load indexes for all the selected non-target organisms (mammals, fishes, algae, bees, and earthworms), whereas the lowest load indexes were observed in Strip and Alf. Overall 179 herbaceous species (20 invasive), 45 carabid species, and 39 butterfly species were observed over the 2 years study. The numbers of herbaceous species, of species of interest for pollinators, and of plant families were higher in Strip, Alf and M-rot, whereas the non-native and invasive plants were higher in M-high and Strip. The highest number of carabid species was observed in the Alf and M-low fields, and in Alf, M-low and Strip when the field margins were considered. The highest number of butterfly species was observed in M-rot, Alf and Strip when a single year was considered, whereas it was observed in Alf when the data were pooled over two years. The CF (kg CO2-eq/ha) had a marked negative effect on the plant (r = - 0.90, P 0.01), carabid (r = - 0.90, P 0.01) and butterfly (r = - 0.79, P 0.01) species richness, as well as it positively affected the proportion of non-native invasive plants (r = 0.76, P 0.01). The N surplus negatively affected plant (r = - 0.71, P 0.01) and butterfly (r = - 0.81, P 0.01) species richness, while positively correlated with proportion of non-native invasive plants observed in the fields (r = 0.80, P 0.01). When all the species from the three indicator-species groups were considered together and plotted against CF (kg CO2-eq/t DM), and against N surplus (kg N/ha), great coefficients of determination were found (R2 = 0.81 and R2 = 0.88, respectively). The current findings provide evidence that well-implemented and diverse cropping systems constitute a benefit for biodiversity at the farm level, without compromising the yield potential, thus offering indications to European decision makers on how to design more specific agri-environmental measures.