Fertilization using manure minimizes the trade-offs between biodiversity and forage production in agri-environment scheme grasslands

Estimation of the number of livestock that can be supporting by the different fertiliser treatments We estimated the number of livestock that could be fed with forage produced under each type of fertilizer application regime. We consider both the amount of forage produced (yield) and the nutritional content of the forage. Because plots in the experiment are too small to support livestock, we used information from the literature to make estimations. We did this for Limou- sin suckler cows, assuming an average weight of 600kg per cow [47]; these cattle are mainly bred for meat production and breeding stock in our experiment as well as in upland grassland systems in England more generally [48]. Given recommendation for farmers on hay nutrient content and food requirement of livestock are reported per hectare in in the literature [49,50], we extrapolated the yield for each of our plots, to that from one hectare [see 12,35,51]. We, then, used the software FarmIQ (2016) to calculate the forage required for feeding cows over winter ahead of spring calving [49], while adjusting the equation parameters to reflect the mea- sured metabolizable energy (ME) content of the forage. Based on the calculated allowance of forage, a spring-calving suckler cow with a body weight of 600 kg would require an allowance of 9.2 kg DM/d (dry matter per day); see S1 File Section 2 for the details of this calculation. We also calculated the nutritional value of the forage available (i.e. metabolizable energy content) to sheep, another livestock animal commonly kept in the uplands [48]. We followed

the same method used for the estimation of cattle production but considering that the equiva- lent forage allocation for mixed aged ewes with an average scanning percentage of 150% and due to lamb in mid-April is 1.4 kg DM/d of forage. Scanning percentage is a performance mea- sure used to tailor management (i.e., nutritional needs) and production (lambing), e.g., [52] and in this case assumes half the ewes will be carrying single lambs, and half twins [47]. Results Comparing the performance of the four different fertilizer regimes with respect to indicators relating to plant community, insect community, forage production and forage nutrient content When comparing different indicators we observed that there was no single fertilizer treatment that had a constant performance across indicators, with treatments showing high and low val- ues for indicators within and between forage production, plant communities and insect com- munities. Moreover, all treatments had a weighted mean of approximately 0.5, showing strong trade-offs between biodiversity and forage production under all treatments (Fig 1). Neverthe- less, FYM showed the highest weighted mean (0.52) followed by no-fertilizer (0.47) and NPK (0.46) and NPK + FYM (0.46). The values of individual indicators are reported in Table 2 in S1 File. The effect of fertilization regimes on the trade-offs between the plant community, insect community and livestock forage For forage related indicators, we observed lower crude protein in the no-fertilizer control com-

pared to FYM, NPK and NPK +FYM treatments, and lower values in NPK-only compared to NPK+FYM. Similarly, we observed lower ruminant metabolizable energy in no-fertilizer than in FYM and NPK treatments compared to FYM. Finally, mean dry content of hay 2011–2014 was lower in the no-fertilizer control than in FYM, NPK and NPK+FYM. Although, not signif- icant at the 5% level, the probability that FYM had a higher crude protein content in forage than NPK was 93% while the probability that FYM has a higher mean dry content was of 93%. We observed no significant differences in ash content among fertilizers. When comparing the PLOS ONE Manure minimizes trade-offs between biodiversity and forage production in agri-environment scheme PLOS ONE | https://doi.org/10.1371/journal.pone.0290843 October 4, 2023 7 / 18