In the quest for sustainable intensification of crop production, we discuss the option of extending the root depth of crops to increase the volume of soil exploited by their root systems. We discuss the evidence that deeper rooting can be obtained by appropriate choice of crop species, by plant breeding, or crop management and its potential contributions to production and sustainable development goals. Many studies highlight the potentials of deeper rooting, but we evaluate its contributions to sustainable intensification of crop production, the causes of the limited research into deep rooting of crops, and the research priorities to fill the knowledge gaps.

Cropping systems comprising winter catch crops followed by spring wheat could reduce N leaching risks compared to traditional winter wheat systems in humid climates. We studied the soil mineral N (N_inorg) and root growth of winter– and spring wheat to 2.5 m depth during 3 years. The roots of the winter and spring wheat penetrated the soil at a similar rate (1.3 mm ^oC day⁻¹) and by virtue of its longer growing period, winter wheat reached depths of 2.2 m, twice that of spring wheat (1.1 m). The deeper rooting of winter wheat was related to much lower amounts of N_inorg left in the 1 to 2.5 m layer after winter wheat (81 kg N_inorg ha⁻¹ less). When growing winter catch crops before spring wheat, N content in the 1 to 2.5 m layer after spring wheat was not different from that after winter wheat. The results suggest that due to its deep rooting, winter wheat may not lead to as high levels of leaching as it is often assumed in humid climates. Deep soil and root measurements (below 1 m) in this experiment were essential to answer the questions we posed.

Summary: Convolutional neural networks (CNNs) are a powerful tool for plant image analysis, but challenges remain in making them more accessible to researchers without a machine‐learning background. We present RootPainter, an open‐source graphical user interface based software tool for the rapid training of deep neural networks for use in biological image analysis. We evaluate RootPainterby training models for root length extraction from chicory (Cichorium intybusL.) roots in soil, biopore counting, and root nodule counting. We also compare dense annotations with corrective ones that are added during the training process based on the weaknesses of the current model. Five out of six times the models trained using RootPainterwith corrective annotations created within 2 h produced measurements strongly correlating with manual measurements. Model accuracy had a significant correlation with annotation duration, indicating further improvements could be obtained with extended annotation. Our results show that a deep‐learning model can be trained to a high accuracy for the three respective datasets of varying target objects, background, and image quality with < 2 h of annotation time. They indicate that, when using RootPainter, for many datasets it is possible to annotate, train, and complete data processing within 1 d.