The intensification of the global water cycle, driven by increasingly frequent and compound extreme weather events, threatens the sustainability of agri-food production. Coupled with population growth and rising food demand, this necessitates new knowledge, technologies, and practices for sustainable intensification. Robust projections of climate impacts on crop growth, underpinned by process-based models, are crucial for designing effective adaptation strategies. While most climate change assessments focus on drought, heat, or gradual climate change, our understanding of waterlogging's impacts on crop growth remains limited. Globally, significant cultivated land is affected by flooding annually, resulting in substantial economic losses. The intensification of the water cycle is expected to increase waterlogging prevalence, impacting the efficient use of economic, natural, and social capital. While genotype × environment × management (G × E × M) studies on climate change adaptation exist, their scalability and transferability are often limited. This study introduces a new approach for assimilating diverse crop model results into common groups characterized by daily stress trajectories plotted over the crop lifecycle. These groups, based on plant stress as an integrated measure of various interacting factors, allow for standardized contrasts across treatments. This approach facilitates intuitive identification of appropriate adaptations and their transferability across regions, improving the efficiency of climate change adaptation strategies.

Existing literature highlights the significant impact of extreme weather events, particularly drought and heat, on crop yields. Studies have shown substantial yield losses in major crops like maize and wheat due to extreme heat and drought conditions. However, the literature on the impact of waterlogging on crop production is relatively sparse, despite its widespread occurrence and potential for economic losses. Previous work has investigated the effects of waterlogging on specific crops and environments, but there's a lack of comprehensive, global assessments that account for the combined influence of genotype, environment, and climate change. Moreover, existing studies often focus on a single factor (waterlogging), without considering the complex interplay of environmental variables and genetic factors. This study bridges this gap by developing a novel approach to analyze the complex interactions affecting waterlogging stress and yield.

This study used a multi-faceted approach. First, a novel paradigm was developed to categorize waterlogging stress patterns based on phenological stages and daily stress trajectories, integrating various factors affecting plant stress responses. This enabled standardization across G × E × M factors and facilitated the identification of common stress patterns. Second, the APSIM farming systems model was enhanced with new algorithms to simulate the effects of waterlogging on photosynthesis and phenology, incorporating experimentally observed responses. The improved model was calibrated and validated using field data from five countries (Australia, Argentina, China, Canada, and Ireland). Third, the calibrated model, coupled with downscaled climate projections from 27 global circulation models (GCMs) under the SSP585 emissions scenario, was used to simulate crop growth and waterlogging stress under current and future (2040s and 2080s) climates. The simulations considered various factors including sowing time and different barley genotypes (spring and winter). Finally, unsupervised k-means clustering was applied to the model outputs to identify common waterlogging stress patterns and their frequencies under different climate scenarios. Statistical analyses were then conducted to assess the impact of waterlogging on yield, the effects of different management strategies (sowing time), and the potential benefits of waterlogging-tolerant genotypes. The yield loss percentage was calculated by comparing yields under waterlogged conditions with control yields. The study also examined the potential for adaptation strategies such as altering sowing times and using waterlogging-tolerant genotypes to mitigate yield losses.

The study's key findings include: 1. Yield penalties due to waterlogging are projected to significantly increase from the historical baseline (3-11%) to 10-20% by 2080. This increase is more pronounced in winter barley genotypes due to their longer growing seasons and higher susceptibility to early-season waterlogging. 2. Waterlogging-tolerant genotypes show greater potential for yield improvement in environments with longer growing seasons (e.g., UK, France, Russia, China), compared to environments with higher evapotranspiration-to-precipitation ratios (e.g., Australia). 3. Combining altered sowing times and waterlogging-tolerant genotypes reduces yield penalties by 18%. Earlier sowing of winter genotypes alleviates waterlogging by 8%. 4. A surprising finding is the similarity of historical and future waterlogging stress patterns, suggesting that adaptation strategies developed for current conditions may be effectively transferred to future climates. 5. K-means clustering revealed four distinct waterlogging stress patterns for spring barley and four for winter barley, capturing 71% and 80% of the variance respectively. These patterns reflect the timing and duration of waterlogging stress relative to crop phenology, influencing yield penalties. 6. Waterlogging tolerance demonstrably increased barley yield under wetter years and generally reduced the yield penalty under future climates. The yield benefit was greater for winter genotypes than spring genotypes. 7. Contextualized adaptation is crucial. The most effective approach varies depending on factors such as growing season length and climate characteristics. In regions with longer growing seasons (e.g., Europe, China), waterlogging-tolerant genotypes are expected to yield the largest benefit.

The findings of this study address the research question by quantifying the impact of climate change on crop waterlogging and identifying potential adaptation strategies. The significant increase in projected waterlogging-induced yield penalties highlights the urgency of developing and implementing effective adaptation measures. The identification of common waterlogging stress patterns across diverse environments and genotypes offers valuable insights for designing more targeted and effective interventions. The findings emphasize the importance of considering the interplay between genotype, environment, and management practices when developing climate change adaptation strategies for crop production. The study's success in identifying similar waterlogging stress patterns under present and future climates suggests that adaptations designed for current conditions can inform strategies for future climate scenarios, facilitating a more efficient and cost-effective approach to adaptation.

This study provides the first experimental quantification of waterlogging impacts under future climates across major barley-producing regions globally. A novel clustering approach categorized waterlogging stress patterns, revealing that despite increasing waterlogging frequency, the overall impact on yield will be moderated by regional variations and CO2 fertilization effects. This approach is widely applicable for analyzing large datasets in agroecology. Waterlogging-tolerant genotypes are a promising adaptation strategy, particularly in regions with longer growing seasons. Future research should focus on multi-model ensemble studies, investigating interactions between waterlogging and other stressors (e.g., nitrogen deficiency), and further characterizing genotypic responses to waterlogging stress for various crops and environments.

This study uses only one crop model (APSIM), although projections from an ensemble of 27 GCMs were used to account for climate variability. This limits the generalizability of the results and could potentially introduce uncertainties. The model's accuracy also depends on the accuracy of the input data, especially rainfall data, and further refinement of the model is still necessary. Although nitrogen stress was excluded to isolate the effects of waterlogging, future studies could explore the interplay between waterlogging and other stressors to get a more comprehensive picture.