Search

Nutrient availability and plant diversity are two important factors determining crop productivity in agricultural ecosystems, but little is known about the underlying mechanisms shaping microbial communities and their regulatory roles in soil biological activity and function. Here, we explored the impacts of fertilization regimes and crop rotations on soil physicochemical properties, crop yield and bacterial and fungal community structures in a 26-year field experiment. The critical determinants for regulating soil enzyme activity profiles involved in carbon (C), nitrogen (N) and phosphorus (P) cycling were identified by the partial least squares path model (PLS-PM). Long-term inorganic or organic fertilization significantly increased soil total N by 27%–77% and crop yield by 237%–419% and decreased soil pH by an average of 0.4 units when compared with non-fertilized control. Soil bacteria were more sensitive than fungi to the fertilization practices. Nutrient additions enriched copiotrophic taxa affiliated to the Pseudomonadaceae and Cytophagaceae bacterial families, but reduced some Acidobacteria such as subgroup 4 RB41, which was the most sensitive biomarker responding to no fertilization. Conversely, fungi were more active in response to crop conversion from wheat-maize to wheat-soybean rotation, leading to a 3-fold enhancement of an unclassified Sordariomycetes family in soybean-based rotation. PLS-PM revealed that fertilization-induced increases in soil enzyme activities were regulated by the bacterial community, while plant-driven alterations in yield, organic C input and soil aggregate-size distribution played an important role for fungal development, which, however, had no significant link to soil enzyme activity profiles. Our results suggest that different response patterns of soil bacteria and fungi to agricultural practices might have consequences for ecosystem function.

Crop nitrogen nutrition is an important indicator for evaluating crop growth. Rapid and non-destructive estimation of nitrogen accumulation in wheat leaves is of great significance for crop nitrogen fertilizer management. Based on field test data from multiple wheat varieties for different locations, years, nitrogen levels, and growth periods, the relationship between 11 canopy hyperspectral parameters and nitrogen accumulation in wheat plants was studied. According to the results of correlation and regression analysis, the flowering period of wheat was selected as the most suitable growth period for crop growth evaluation (the average R2 was 0.732, and the root mean square error (RMSE) was 0.354). A new vegetation index, NDchI*DDN (referred to as the nitrogen accumulation vegetation index, abbreviated as NAVI), was constructed based on the pairwise combination of traditional vegetation index products. This parameter had a high correlation with plant nitrogen accumulation (R2 = 0.856), and the root mean square error (RMSE) was 0.296. Tested by independent experimental data, the fitting degree of the plant nitrogen accumulation inversion model established with NAVI as the variable was R2 = 0.861, the relative error RE = 9.3%, RMSE = 0.398, and the prediction accuracy was significantly higher than other models. Therefore, construction of a NAVI-based plant nitrogen accumulation monitoring model gave ideal test results, which could reduce the limitations of experimental conditions and is expected to provide new important technical support for precise fertilization.