Alexandra Jacquelyn Burgess

Summary: Wind‐driven plant movement generates rapid light fluctuations (windflecks), which can impact canopy photosynthesis. Targeting crop photosynthesis in dynamic light provides a potential path towards boosting yield. Here, we quantified how plant architecture and biomechanics modulate such windflecks across 10 high‐yielding cultivars of winter wheat ( Triticum aestivum ). Using synchronized high‐frequency measurements of irradiance, wind speed, and canopy motion (quantified by frame differencing from video), we assessed the propensity of wheat cultivars to move (motion sensitivity), and the ability for movement to produce windflecks (light modulation efficiency) in the field. There was up to 10‐fold variation in the quantity of motion between cultivars under identical wind speeds. Cultivars also exhibited structural trade‐offs and specific in canopy windfleck properties. Some had low motion under wind but produced frequent windflecks when moving, whereas others exhibited high motion under similar wind but varied in windfleck frequency. Overall, windfleck properties were best explained by aerodynamic traits: cultivars with narrower leaves and lower leaf‐to‐stem mass ratios were associated with more intense windflecks. These findings establish that wheat cultivars actively modulate their light environment through biomechanical traits. By integrating plant motion into crop models, favouring motion–light relationships, which could provide a critical route to yield improvements in turbulent environments.

Stomata regulate gas exchange between plants and the atmosphere, but analysing their morphology is challenging due to anatomical variability and artefacts during image acquisition. Deep learning (DL) can address these challenges but often requires large and diverse datasets, which are costly and error prone to produce. Generative adversarial networks (GANs) offer a solution by generating artificial data via unsupervised learning. However, GANs often suffer from problems including mode collapse, vanishing gradients, and network failure, particularly with small datasets. Here, we present StomaGAN, a deep convolutional GAN (DCGAN) with tailored modifications to address common GAN issues. We collected 559 stomatal impressions of field, or faba bean (Vicia faba) consisting of ~3000 stoma, 80% of which were used to train StomaGAN. Evaluation metrics, including generator and discriminator loss progression and a mean Fréchet Inception Distance (FID) score of 61.4 across eight experimental runs confirm successful training. To validate StomaGAN, we generated artificial images to train a deep convolutional neural network (DCNN) based on the DeepLabV3 framework for stomata detection from real, unseen images. The DCNN achieved a mean Interception over Union (IoU) of 0.95 on artificial training images and 0.91 on real, unseen, images across varying magnifications. Our results demonstrate that StomaGAN effectively generates high-quality synthetic datasets, enabling reliable stomatal detection and enhancing phenotypic analysis. This approach reduces the need for extensive manual data collection and simplifies complex morphological assessments.

Plant physiology and metabolism rely on the function of stomata, structures on the surface of above-ground organs that facilitate the exchange of gases with the atmosphere. The morphology of the guard cells and corresponding pore that make up the stomata, as well as the density (number per unit area), are critical in determining overall gas exchange capacity. These characteristics can be quantified visually from images captured using microscopy, traditionally relying on time-consuming manual analysis. However, deep learning (DL) models provide a promising route to increase the throughput and accuracy of plant phenotyping tasks, including stomatal analysis. Here we review the published literature on the application of DL for stomatal analysis. We discuss the variation in pipelines used, from data acquisition, pre-processing, DL architecture, and output evaluation to post-processing. We introduce the most common network structures, the plant species that have been studied, and the measurements that have been performed. Through this review, we hope to promote the use of DL methods for plant phenotyping tasks and highlight future requirements to optimize uptake, predominantly focusing on the sharing of datasets and generalization of models as well as the caveats associated with utilizing image data to infer physiological function.

This article comments on: Yixiang Shan and Colin P. Osborne, Diversification of quantitative morphological traits in wheat, Annals of Botany, Volume 133, Issue 3, 1 March 2024, Pages 413–426 https://doi.org/10.1093/aob/mcad202

Alexandra J Burgess, José Manuel Ugalde; SOS: Speed of Stomata opening and closing is influenced by vapour pressure deficit, Plant Physiology, , kiae066, https:

The crop ideotype was first defined by Donald in 1968 as an idealised structural model of a plant which can theoretically attain maximum yield quantity and quality within a given environment. Many of the major crop species have proposed ideotypes. For example, in wheat (Triticum aestivum), suggested characteristics include a short strong stem, few erect, small leaves and large ears. Breeding for crop ideotypes is a continuous process as beneficial traits are updated to match current needs. Government organisations such as the Agriculture and Horticulture Development Board (AHDB) in the United Kingdom, publish a recommended list of varieties each year, and so we hypothesise that these recommended varieties will express traits of the ideotype. However, through greenhouse experiments, we show that three recent AHDB recommended wheat varieties differ to that of the ideotype and to each other. This either indicates that the concept may not be universal, or the main target for breeding programmes; that traits may not have successfully been bred yet; or, that other conflicting reasons may be preventing the expression of the ideotype. To this end, we review the factors that may influence plant structure before discussing new tools and methods that can aid ideotype design. Through this paper, we argue that whilst there are benefits to the concept of the ideotype, the ‘power’ of the concept is lost when integrating the multitude of factors that, in combination, determine plant structure and function in a given environment. Nevertheless, regardless of these limitations, we argue that there is still a role for the ideotype in crop breeding, but that this concept needs to be expanded to emphasise the genetic and environmental interactions that influence plant physiology.

Future climate change scenarios predict changes in temperature, water availability, and the frequency and severity of extreme weather events. These climate variables often occur in combination; for example, drought and heat stress generally occur simultaneously. The synergistic impacts of these climate variables often mean that their combined impact on plant physiology is much greater than that of the individual stressors. Therefore, understanding both the individual and compound effect of stressors on carbon assimilation will be vital to both predict and improve crop yield.

Crop yields must increase in order to meet food security goals. Estimating crop yield-related traits accurately and rapidly is therefore of utmost importance, with applications in breeding, precision farming, and yield monitoring. Hight-throughput phenotyping (HTPP) refers to the use of non-invasive techniques to measure physiological and agronomical traits of crop plants. Optical remote-sensing techniques are widely used in HTPP. They encompass data captured from areas of the visible (400–700 nm, captured using red green blue sensors), near infrared (700–1350 nm), red edge (680–730 nm), and shortwave infrared (1350–2500 nm) regions of the spectrum ( Gamon et al. 2019). This spectral information has numerous applications within crop monitoring direct prediction of a trait of interest such as pigment content or biomass area, or through using spectral indices, otherwise known as vegetative indices, which correlate spectral reflectance with physiological traits such as green area, portrayed by the normalized difference vegetation index ( Fu et al. 2020; Tattaris et al. 2016).

Plants grow in diverse environments: on freezing islands close to the Antarctic, in sweltering deserts, or in humid tropical rainforests. In contrast to animals such as migratory birds, they are sessile and must adapt to environmental changes, including differences in temperature, rainfall, or changes in daylength (photoperiod) throughout the seasons of the year. This ability to adapt can be controlled by diverse mechanisms and could provide potential targets to improve crop performance.

Photosynthesis uses sunlight energy to assimilate CO₂ into organic compounds. This remarkable process is the source of most of our food and fuel. Optimizing photosynthesis to adapt crops to climate change and to meet future food demands has resulted in substantial research in this area in recent years. Most research has been focused on photosynthesis in foliar tissue (i.e. leaves), where most of the light harvesting and CO₂ fixation is catalyzed by the enzyme Rubisco. However, nonfoliar tissues like stem tissue, seeds, or ears of wheat also perform photosynthesis ( Schwender et al. 2004; Maydup et al. 2010). Rubisco activity has been detected in nonfoliar tissue of soybean, oilseed rape, and broad bean and in the pod walls of seeds. The Rubisco concentration in these tissues is approximately 10 to 100 times lower than that of leaves ( Hedley et al. 1975). Furthermore, nonfoliar tissues have all the required components for light harvesting, such as PSI and PSII, the cytochrome b₆f complex and ATP synthase ( Weber et al. 2005; Allorent et al. 2015; Kohzuma et al. 2017). Although both types of tissue fix CO₂ that has entered the plant through stomata, nonfoliar tissues tend to use a greater amount of CO₂ released during mitochondrial respiration ( Simkin et al. 2020).