Seasonal soil health dynamics in soy-wheat relay intercropping.

There is growing interest in intercropping as a practice to increase productivity per unit area and ecosystem functioning in agricultural systems. Relay intercropping with soy and winter wheat may benefit soil health due to increased diversity and longer undisturbed soil cover, yet this remains largely unstudied. Using a field experiment in Eastern Germany, we studied the temporal dynamics of chemical, biological, and physical indicators of soil health in the topsoil over a year of cultivation to detect early effects of soy-wheat relay intercropping compared to sole cropping. Indicators included microbial abundance, permanganate-oxidizable carbon, carbon fractions, pH, and water infiltration. Relay intercropping showed no unique soil health benefits compared to sole cropping, likely affected by drought that stressed intercropped soy. Relay intercropping did, however, maintain several properties of both sole crops including an increased MAOM C:N ratio and higher soil water infiltration. The MAOM C:N ratio increased by 4.2 and 6.2% in intercropping and sole soy and decreased by 5% in sole wheat. Average near-saturated soil water infiltration rates were 12.6, 14.9, and 6.0 cm hr-1 for intercropping, sole wheat, and sole soy, respectively. Cropping system did not consistently affect other indicators but we found temporal patterns of these indicators, showing their sensitivity to external changes.

Does weed diversity mitigate yield losses?

While intensive control of weed populations plays a central role in current agriculture, numerous studies highlight the multifaceted contribution of weeds to the functionality and resilience of agroecosystems. Recent research indicates that increased evenness within weed communities may mitigate yield losses in contrast to communities characterized by lower diversity, since weed species that strongly affect crop yields, also dominate weed communities, with a concurrent reduction of evenness. If confirmed, this observation would suggest a paradigm shift in weed management towards promoting higher community diversity. To validate whether the evenness of weed communities is indeed linked to higher crop productivity, we conducted two field experiments: one analyzing the effects of a natural weed community in an intercrop of faba bean and oat, and the other analyzing the effects of artificially created weed communities, together with the individual sown weed species, in faba bean, oats and an intercrop of both crops. The evenness of the weed communities ranged from 0.2 to 0.9 in the natural weed community, from 0.2 to 0.7 in faba bean, from 0 to 0.8 in the intercrop and from 0.3 to 0.9 in oats. Neither the natural nor the artificial weed community showed significant effects of evenness on crop grain yield or crop biomass. The results of this study do not validate a positive relationship of crop productivity and weed evenness, possibly due to low weed pressure and the absence of competitive effects but suggest that also less diverse weed communities may be maintained without suffering yield losses. This is expected to have far reaching implications, since not only diverse weed communities, but also higher abundances of few weed species may contribute to ecosystem functions and may support faunal diversity associated with weeds.

Data-driven crop growth simulation on time-varying generated images using multi-conditional generative adversarial networks.

BACKGROUND: Image-based crop growth modeling can substantially contribute to precision agriculture by revealing spatial crop development over time, which allows an early and location-specific estimation of relevant future plant traits, such as leaf area or biomass. A prerequisite for realistic and sharp crop image generation is the integration of multiple growth-influencing conditions in a model, such as an image of an initial growth stage, the associated growth time, and further information about the field treatment. While image-based models provide more flexibility for crop growth modeling than process-based models, there is still a significant research gap in the comprehensive integration of various growth-influencing conditions. Further exploration and investigation are needed to address this gap. METHODS: We present a two-stage framework consisting first of an image generation model and second of a growth estimation model, independently trained. The image generation model is a conditional Wasserstein generative adversarial network (CWGAN). In the generator of this model, conditional batch normalization (CBN) is used to integrate conditions of different types along with the input image. This allows the model to generate time-varying artificial images dependent on multiple influencing factors. These images are used by the second part of the framework for plant phenotyping by deriving plant-specific traits and comparing them with those of non-artificial (real) reference images. In addition, image quality is evaluated using multi-scale structural similarity (MS-SSIM), learned perceptual image patch similarity (LPIPS), and Fréchet inception distance (FID). During inference, the framework allows image generation for any combination of conditions used in training; we call this generation data-driven crop growth simulation. RESULTS: Experiments are performed on three datasets of different complexity. These datasets include the laboratory plant Arabidopsis thaliana (Arabidopsis) and crops grown under real field conditions, namely cauliflower (GrowliFlower) and crop mixtures consisting of faba bean and spring wheat (MixedCrop). In all cases, the framework allows realistic, sharp image generations with a slight loss of quality from short-term to long-term predictions. For MixedCrop grown under varying treatments (different cultivars, sowing densities), the results show that adding these treatment information increases the generation quality and phenotyping accuracy measured by the estimated biomass. Simulations of varying growth-influencing conditions performed with the trained framework provide valuable insights into how such factors relate to crop appearances, which is particularly useful in complex, less explored crop mixture systems. Further results show that adding process-based simulated biomass as a condition increases the accuracy of the derived phenotypic traits from the predicted images. This demonstrates the potential of our framework to serve as an interface between a data-driven and a process-based crop growth model. CONCLUSION: The realistic generation and simulation  of future plant appearances is adequately feasible by multi-conditional CWGAN. The presented framework complements process-based models and overcomes their limitations, such as the reliance on assumptions and the low exact field-localization specificity, by realistic visualizations of the spatial crop development that directly lead to a high explainability of the model predictions.