Wheat yield as a measure of the residual fertility after 20 years of forage cropping systems with different manure management in Northern Italy

After 20 years of application of different manure types, cropping systems and additional nitrogen (N) levels, their residual fertility effects were compared by measuring the yield of a following unfertilised wheat crop (Experiment 1), which was sown on exactly the same plots of the previous long-term trial. All previously applied factors caused significant differences in wheat yield. Wheat yielded more on plots that had received farmyard manure (FMY) compared to those where semi-liquid manure (SLM) was previously applied. Long-term application of a semi-intensive rotation, with three years of annual double cropping of autumnsown Italian ryegrass and spring-sown silage maize followed by three years of mown lucerne (R6), resulted in higher wheat yield than application of just the annual double cropping of Italian ryegrass and silage maize (R1). Application of further mineral N fertilisation to previous cropping systems caused higher yield of the subsequent wheat crop. The difference in wheat yield between the R6 and R1 systems was greater with SLM (+28%) than FYM application (+11%) resulting in a significant manure × system interaction. A companion experiment (Experiment 2) was carried out to compute the nitrogen agronomic efficiency (NAE) from the yield of wheat plots that were sown after ploughing a nearby 20- year unfertilised grassland and received four levels of mineral N fertilisation. NAE was further used to empirically estimate the productive advantage (PA) conferred by previous manure-systemmineral nitrogen combinations in the long-term trial. PA was measured as equivalent kg of mineral N to be applied to wheat to achieve the yield level recorded after any previous combination. The estimated PA values were much higher when wheat followed FYM compared to SLM application, and when it followed R6 compared to R1 system. The SLM-R1 combination had negative PA values, indicating a productive disadvantage on wheat of this preceding combination. The enhancement of residual soil fertility by long-term application of FYM compared to SLM could be attributed to greater nutrient provision during the years by FYM than by SLM. However, further fertility advantages of FYM are discussed. Despite lower nutrient supply by organic fertilisers in R6 than in R1 system, the former had higher residual fertility. The presence of lucerne in the R6 rotation likely enriched the soil in nitrogen and increased its availability for following cropping. Possible benefits of the legume on the soil suppressiveness might have been a further asset of the R6 system.