Mapping spatial zones of climate vulnerability and adaptive potential for major crops in the Texas High Plains

Mapping spatial zones of climate vulnerability and adaptive potential for major crops in the Texas High Plains | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Mapping spatial zones of climate vulnerability and adaptive potential for major crops in the Texas High Plains Ahmed Attia, Prem Woli, Charles R. Long, Francis M. Rouquette Jr., and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6864209/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 11 Sep, 2025 Read the published version in Modeling Earth Systems and Environment → Version 1 posted 7 You are reading this latest preprint version Abstract Climate change poses an increasing threat to agricultural productivity in the Texas High Plains (THP), a semi-arid region facing both warming trends and declining groundwater resources. This study integrates process-based crop modeling with geospatial analysis to identify spatial zones of climate vulnerability and adaptive potential for four major crops; winter wheat, cotton, maize, and grain sorghum under future climate scenarios. Using the DSSAT model, historical (1991–2020) and future yields (2031–2060 and 2070–2099) were simulated across 48 counties under Representative Concentration Pathway 4.5 and 8.5 (RCP 4.5 and RCP 8.5). Spatial clustering techniques, including Global Moran’s I and Getis-Ord Gi* statistics, were applied to classify counties into vulnerable, adaptive, stable, and more stable zones based on projected yield changes. Results revealed that wheat vulnerability was concentrated in southern counties, with projected yield decreases of 10–30% under RCP 8.5, while northern counties showed 30–50% yield increases under RCP 4.5 mid-century and RCP 8.5 end-century. In contrast, cotton yields are projected to increase by 20–40% across most counties under RCP 4.5 end-century and RCP 8.5 mid- and end-century, with localized vulnerability emerging in southwestern THP under RCP 8.5 by end-century. Grain sorghum yields are projected to increase by 10–20% in eastern and northern counties under RCP 4.5, but under RCP 8.5 widespread yield declines exceeding 40% are expected by end-century, attributed to reduced rainfall and increased temperature stress during the growing season. In contrast, maize showed greater resilience, with yield changes varying spatially but remaining positive in many southern counties. These spatially explicit findings underscore the need for targeted adaptation strategies, including the deployment of climate-resilient crop varieties, optimized irrigation management, crop diversification, and adaptive land use planning. The study offers actionable insights to support climate-resilient agricultural planning and inform precision adaptation policies for sustaining crop productivity in the THP under future climate scenario Climate change Spatial clustering Yield vulnerability Dryland agriculture climate impact zones Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction The Texas High Plains (THP) is a semi-arid region in northwest Texas, encompassing approximately 8.9 million ha. Major field crops grown in this area include cotton ( Gossypium hirsutum L.), grain sorghum ( Sorghum bicolor [L.] Moench), winter wheat ( Triticum aestivum L.), and maize ( Zea mays L.), cultivated under both irrigated and rainfed conditions. Irrigated agriculture accounts for about 44% of total cropland in THP in 2022, with maize fields accounts for 20%, and 4.5% each to cotton and winter wheat (Waller, 2023 ). Grain sorghum, predominantly grown under dryland conditions, planted in 587,000 ha across TX in 2024, highlighting its importance in low-input systems (USDA-NASS, 2024 ). Dryland agriculture constitutes the remaining percentage of cropland in the THP, rendering these areas particularly vulnerable to climate change due to limited water availability. The Ogallala Aquifer, a major groundwater source underlying the THP, has experienced significant depletion over the past decades, raising concerns about the sustainability of irrigation practices in the region (Southern SARE, 2024 ). Understanding the spatial heterogeneity of climate impacts on crop productivity is critical for designing region-specific adaptation strategies. In the THP, projected shifts in temperature, precipitation patterns, and growing season length under future climate scenarios are expected to alter crop performance unevenly across counties (Wen et al., 2024 ). Identifying climate impact zones such as vulnerable, adaptive, and stable areas based on yield response to climate change allows for more targeted management interventions and resource allocation. Vulnerable zones may require substantial adaptation investments, including drought-tolerant cultivars, optimized irrigation, and soil conservation practices. Conversely, adaptive and more stable zones offer opportunities to enhance resilience and maintain or even increase productivity under changing climatic conditions. Mapping these zones can thus inform and support both short- and long-term agricultural planning at the county scale. Recent advances in crop modeling and geospatial analysis offer powerful tools for evaluating the impacts of climate change on agricultural systems. Process-based models such as DSSAT (Decision Support System for Agrotechnology Transfer) (Hoogenboom et al., 2019 ; Jones et al., 2003 ) enable simulation of crop growth, yield, and soil processes under diverse climate and management scenarios. While several studies have explored future crop productivity in the Texas High Plains, most have been limited to specific sites or research stations (Adhikari et al., 2016 ; Attia et al., 2016a ; Kothari et al., 2019b ; Marek et al., 2017 ). These site-specific studies often yield aggregated findings that are generalized to surrounding areas under the assumption of minimal spatial variability. However, this can obscure localized drivers of production dynamics and lead to oversimplified recommendations. In contrast, spatially explicit assessments that capture crop yield responses at finer resolutions are invaluable for agronomic planning and climate adaptation. Previous research has demonstrated the effectiveness of geospatial techniques for supporting regionally targeted adaptation and resilience planning (Attia et al., 2024 ). Mapping zones of vulnerability and adaptive potential for major crops in the THP offers valuable insight into prioritizing localized adaptation strategies, optimizing resource allocation, and informing policy decisions. In this study, we integrated process-based crop modeling with county-scale spatial analysis to identify vulnerable and adaptive production zones for cotton, grain sorghum, wheat, and maize under historical and future climate scenarios. The specific objectives were to: (i) simulate historical and future crop yields using the DSSAT model across all counties in the THP; (ii) classify and map vulnerable and adaptive zones based on yield trajectories; and (iii) assess the relative impacts of climate drivers on productivity trends to inform regional adaptation planning. Materials and methods Study area The THP is a semi-arid agricultural region located in the northwestern part of Texas, covering approximately 8.9 million ha (Fig. S1 ). The region is characterized by a relatively flat terrain and a continental climate with hot summers, cold winters, and erratic precipitation averaging around 450–550 mm annually, most of which occurs during the summer growing season (April to September). The THP sits atop the Ogallala Aquifer, a vital but rapidly depleting groundwater source that primarily supports irrigated agriculture and livestock operations, while also indirectly benefiting dryland systems through supplemental irrigation and regional agricultural infrastructure (Waller, 2023 ). Major crops grown in the region include cotton ( Gossypium hirsutum L.), maize ( Zea mays L.), sorghum ( Sorghum bicolor L. Moench), and winter wheat ( Triticum aestivum L.), with varying degrees of reliance on irrigation. Irrigated agriculture accounts for roughly 44% of total cropland (Waller, 2023 ), though increasing aquifer depletion has led to a shift toward more water-efficient and rainfed systems. This makes the region especially vulnerable to climate variability and water constraints, necessitating improved understanding of spatial yield responses and adaptive agricultural practices. DSSAT model, calibration, and application DSSAT model is a process-based modeling platform that simulates crop growth, soil dynamics, and management interactions under variable climate conditions (Hoogenboom et al., 2019 ; Jones et al., 2003 ). In this study, we applied a 4-year improved rotation system specifically designed for the THP, integrating cover crops (CCs) based on the available windows between main crop cycles (Table S1 ). Cultivar parameters for grain sorghum ( Sorghum bicolor L. Moench), maize ( Zea mays L.), cotton ( Gossypium hirsutum L.), and wheat ( Triticum aestivum L.) were adopted from prior THP-specific calibrations (Adhikari et al., 2016 ; Kothari et al., 2019a ; Kothari et al., 2019b ; Marek et al., 2017 ). The following crop rotations and suggested improved N applications were used in the model: Sorghum : Planted June 25, harvested November 15, with 55 kg ha⁻¹ synthetic N applied at 10 days post-planting and another 55 kg ha⁻¹ at the tillering stage. Organic N application through manure: 40 kg ha⁻¹. Total N: 150 kg ha⁻¹. Winter legume CC (winter pea, Pisum sativum L. ) : Planted November 20, terminated April 20. No synthetic or organic N applied. Maize : Planted April 25, harvested September 15, with 50 kg ha⁻¹ synthetic N applied at 10 days post-planting and another 50 kg ha⁻¹ at the V6 stage (presence of six fully emerged leaves with visible leaf collars). Organic N application through manure: 50 kg ha⁻¹. Total N: 150 kg ha⁻¹. Winter legume CC (winter pea, Pisum sativum L. ) : Planted October 10, terminated April 20, fixing ~ 50 kg N ha⁻¹. No synthetic or organic N applied. Cotton : Planted May 1, harvested October 15, with 50 kg ha⁻¹ synthetic N applied at 10 days post-planting and another 50 kg ha⁻¹ at midseason. No organic N applied. Total N: 100 kg ha⁻¹. Winter wheat : Planted October 20, harvested June 20, with 30 kg ha⁻¹ synthetic N applied at 10 days post-planting, 30 kg ha⁻¹ at the tillering stage, and another 30 kg ha⁻¹ later in the season. Organic N application through manure: 50 kg ha⁻¹. Total N: 140 kg ha⁻¹. The winter pea was calibrated based on field experiments conducted at the Texas A&M AgriLife Chillicothe Research Station (DeLaune and Mubvumba, 2020 ), supplemented with literature data (Neugschwandtner et al., 2019 ; Wilson and Robson, 1996 ). Details on model calibration, initialization, and validation procedures are available in the supplementary materials (Table S2, Fig. S2 and S3). All simulations in this study were conducted under dryland (rainfed) conditions, with no supplemental irrigation applied to any crop. Climate models and soil data To assess climate impacts on crop productivity across the THP, simulations were conducted using five global circulation models (GCMs) and regionally specific soil datasets. Climate projections were based on two representative concentration pathways (RCPs): RCP 4.5, representing a stabilization scenario with emissions peaking around 2040, and RCP 8.5, a high-emissions scenario. CO 2 fertilization was considered according to each RCP scenario. Consequently, CO₂ fertilization effects were dynamically accounted for by incorporating representative CO 2 concentrations associated with each RCP scenario into the DSSAT weather files. We selected five GCMs from the Multivariate Adaptive Constructed Analogs (MACA) dataset, developed by the Climate Impacts Group and available via the Northwest Knowledge Network data portal (Abatzoglou and Brown, 2012 ; Taylor et al., 2012 ). These include IPSL-CM5A-MR, MIROC5, CCSM4, CNRM-CM5, and CSIRO-Mk3-6-0, representing a diverse range of climate sensitivities and modeling approaches (Table S3). All projected results presented in this study represent ensemble means averaged across these five GCMs. Gridded soil information was sourced from the WISE database (Batjes, 2009 ; Gijsman et al., 2007 ), developed by the ISRIC SoilGrids initiative, and further processed following the methods of Han et al. ( 2019 ) to ensure compatibility with DSSAT requirements. Soils were characterized and aggregated to match the simulation grid resolution used in climate modeling. A point shapefile layer representing a 12 km resolution grid was created to facilitate the spatio-temporal application of the DSSAT model across the THP. Simulations were conducted for three time periods: baseline (1991–2020), mid-century (2031–2060), and end-century (2070–2099), under two RCPs: 4.5 and 8.5. For each grid point, a Python-based pipeline was used to automate DSSAT simulations across spatial grid points, incorporating climate, soil, and crop management inputs. The pipeline also enabled batch post-processing of model outputs to evaluate temporal and spatial impacts of climate and crop rotation scenarios on productivity, soil carbon dynamics, and system-level sustainability. Climate change impact assessment analysis and spatial clustering To assess the spatial variability in wheat yield changes across the THP, yield changes were calculated for each county by comparing the baseline period to two future time periods: mid-century and end-century under RCP 4.5 and RCP 8.5 scenarios. These differences were expressed as percentage change in yield (future vs. baseline), and the spatial clustering and zoning analysis was applied to interpret patterns of vulnerability and stability. First, spatial yield changes were computed by interpolating gridded simulation results from DSSAT model outputs. A point shapefile with simulation output locations was joined with the county boundary shapefile of the THP. The shapefile contained county-level polygons (n = 48 counties) and was projected to WGS84 (EPSG:4326). The yield change data were prepared in a long format across four scenarios: RCP 4.5 mid-century, RCP 4.5 end-century, RCP 8.5 mid-century, and RCP 8.5 end-century. To calculate average yield changes per county, a spatial join was performed between point-level yield change estimates and county boundary polygons using the sf package in R software. After spatially linking each yield point to its respective county, the data were aggregated by county and scenario using the ddply() function from the plyr package to compute both the mean and standard deviation of yield change. These aggregated values enabled consistent comparisons across counties and scenarios. $$\:\text{Y}\text{i}\text{e}\text{l}\text{d}\:\text{c}\text{h}\text{a}\text{n}\text{g}\text{e}\:\left(\%\right)=\frac{\text{F}\text{u}\text{t}\text{u}\text{r}\text{e}\:\text{y}\text{i}\text{e}\text{l}\text{d}-\text{B}\text{a}\text{s}\text{e}\text{l}\text{i}\text{n}\text{e}\:\text{y}\text{i}\text{e}\text{l}\text{d}\:}{\text{B}\text{a}\text{s}\text{e}\text{l}\text{i}\text{n}\text{e}\:\text{y}\text{i}\text{e}\text{l}\text{d}}\:\:\times\:100\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(1\right)\:$$ Next, we applied Global Moran’s I (Moran, 1950 ) to determine the degree of spatial autocorrelation in the county-level yield change data. Moran’s I is defined as: $$\:I=\frac{N}{W}+\frac{{\sum\:}_{i=1}^{N}{\sum\:}_{j=1}^{N}{w}_{ij}({x}_{i}-\stackrel{-}{x})({x}_{j}-\:\stackrel{-}{x}\:)}{{\sum\:_{i=1}^{N}\left({x}_{i}-\:\stackrel{-}{x}\right)}^{2}}\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\:\left(2\right)$$ where \(\:\:N\) is the number of counties, \(\:{x}_{i},\:{x}_{j}\) are the yield change at locations i and j , respectively, \(\:\stackrel{-}{x}\) is the mean of all \(\:{x}_{i}\) values, \(\:{w}_{ij}\) is the spatial weight between location i and j , and \(\:W\) is the sum of all spatial weights. A significant positive Moran's I indicates that counties with similar yield changes are spatially clustered, while negative values indicate spatial dispersion. For finer resolution, the Getis-Ord Gi* statistic (Getis and Ord, 1992 ) was computed to classify local spatial clusters. Gi* analysis identifies hotspots and coldspots by measuring the statistical significance of high or low yield changes in the context of neighboring counties. The resulting Z-scores of the Gi* statistic were classified into four climate impact zones: vulnerable, adaptive, stable, and more stable. A vulnerable zone defines counties with negative GiZ greater than two standard deviations of the mean (95%) for all points within that county, i.e. GiZ ≤ -1.96 (statistically significant yield decline), adaptive zone defines counties with negative GiZ of spatial association equal to or greater than one standard deviation of the mean, i.e. -1.96 < GiZ ≤ -0.96. Crop yield is expected to be slightly reduced or remain neutral due to adaptation mechanisms in these adaptive zones. Stable zone defines counties with neutral to slight insignificant increase in future crop yield, where average historical level of production is expected to be maintained. The GiZ for those ranges from − 0.96 to 0.96. The more stable zone defines counties where future yield is expected to be significantly higher than the historical yield, GiZ is ≥ 0.96, suggesting beneficial climate change impacts. These classifications were applied for each crop and RCP scenario independently. The zone classification aids in identifying counties that are projected to benefit or suffer from climate change impacts on crop productivity, helping to prioritize adaptation strategies. In addition, growing season precipitation was extracted from climate input datasets used for model simulations. Precipitation data were aggregated specifically for each crop’s growing season, and changes between baseline and future periods were mapped under different RCP scenarios. These maps of spatial rainfall change were analyzed alongside yield change patterns, facilitating interpretation of crop responses to projected precipitation variability. This analysis provided insights into potential constraints or benefits of water availability during critical crop phenological stages, thus improving the understanding of climate-related impacts on crop performance across the THP. Results and discussion Spatial clustering and climate-impact zones for wheat and cotton The spatial distribution of simulated county-level yield changes for wheat and cotton exhibited significant positive spatial autocorrelation under both RCP 4.5 and RCP 8.5 scenarios at mid-century and end-century periods (Fig. 1 ). The global Moran's I values were high for wheat across all scenarios, ranging from 0.565 to 0.782 with p-values < 0.001, indicating strong spatial clustering of similar yield changes across counties. Similarly, cotton demonstrated significant but slightly lower Moran's I values ranging from 0.405 to 0.566 (p < 0.001), suggesting moderately strong spatial structure in yield change patterns. The county-level wheat yield projections revealed considerable variability across the THP (Fig. 2 ). Negative yield changes dominated the south and central counties, particularly under RCP 8.5 in mid-century, with yield losses of about 10–30% (Fig. 2 C-D). In contrast, northern counties exhibited positive yield changes, with increases ranging from 30 to 50% by end-century under RCP 8.5. Consequently, the vulnerable zones were predominantly located in the southern counties, while stable and more stable zones were clustered in the northern part of the region (Fig. 2 E-H). However, it should be noted that the baseline yields in these northern counties were generally lower than those in central and western counties (Fig. S5), which could be attributed to less amount of rainfall (Fig. S5). In this context, regulated deficit irrigation at grain-filling stage of wheat was reported to increase dryland yield by 68% in THP (Attia et al., 2016b ). This suggests that supplemental irrigation could help elevate yields to approximately 1.7 Mg ha − 1 in these water-limited areas, supporting both productivity and resilience under future climate stress. For cotton, the county-level yield projections showed a generally favorable outlook under future climate scenarios (Fig. 3 ). Under RCP 4.5 mid-century, most counties experienced seed yield increases of approximately 20–40% (Fig. 3 B). By end-century, regardless of RCP, yield gains became more substantial, with many counties experiencing increases ranging from 60% up to more than 100% (Fig. 3 D). Spatial climate impact zoning for cotton revealed a dominance of stable and more stable zones across the THP under both RCPs and time periods. Particularly under RCP 8.5 by end-century (Fig. 3 H), the expansion of more stable zones highlighted cotton’s higher adaptive capacity compared to wheat. Only isolated vulnerable zones were observed for cotton, mainly in a few southwestern counties under RCP 8.5 end-century (Fig. 3 H). It is noteworthy that despite an overall positive yield trend, spatial hotspot analysis (Getis-Ord Gi) revealed the emergence of localized vulnerable zones. This finding underscores that county-level yield increases do not always translate to uniform spatial resilience, emphasizing the importance of identifying spatial disparities even under favorable climate trajectories. This occurs because GiZ statistics detect spatial clustering patterns independent of overall mean shifts, areas with significantly lower-than-expected yield changes compared to their neighbors are flagged as vulnerable, even when general yield increases are observed. In statistical terms, while the mean yield change is positive, localized negative deviations relative to surrounding counties still trigger significant clustering of "low" values. Therefore, spatial disparity analysis complements broad-scale projections by identifying localized zones of vulnerability that may otherwise be masked by broad regional averages. Spatial clustering and climate-impact zones for maize and sorghum The spatial clustering of simulated maize and sorghum yields changes also demonstrated statistically significant spatial autocorrelation across the THP under all future climate scenarios (Fig. 4 ). For maize, the global Morans’s I values range from 0.77 (RCP 4.5 mid-century) to 0.911 (RCP 8.5 end-century), indicating very strong clustering of yield change responses. Sorghum showed similarly robust but slightly lower spatial autocorrelation, with Moran’s I values ranging from 0.434 to 0.764 (p < 0.001 in all cases), reflecting meaningful spatial structure across counties (Fig. 4 ). County-level maps revealed clear spatial contrasts in maize yield responses (Fig. 5 ). Under RCP 4.5, several southern counties were projected to experience up to 40–60% increase in maize yield by the end of the century. Under RCP 8.5, however, spatial variability was greater, with southern and southwestern counties showing gains exceeding 100% by mid-century (Fig. 5 C), while northern regions experienced more modest or stable trends. However, by end-century northern regions showed yield decrease by 50% (Fig. 5 D). These yield patterns translated into climate impact zones, with southern counties consistently classified as more stable under both scenarios, while several northern and Panhandle counties emerged as vulnerable or adaptive zones by mid- and end-century. This spatial differentiation of yield stability aligns with findings by Kipkulei et al. ( 2025 ), who used DSSAT-CERES-Maize modeling in Kenya to show yield declines of up to 41% and identified climate hotspots and stable zones to guide targeted adaptation strategies. For sorghum, the yield projections painted a more concerning picture under future climate scenarios, particularly under RCP 8.5 (Fig. 6 ). Under RCP 4.5, sorghum yields were projected to increase by 10–20% in several eastern and northern counties during the mid-century period, but these gains diminished by the end of the century, with most counties shifting to neutral or experiencing yield declines of around 10% (Fig. 6 A and B). In contrast, under RCP 8.5, widespread yield declines dominated across the THP, with several counties experiencing reductions exceeding 40% by end-century (Fig. 6 C and D). This could be attributed to relatively higher baseline precipitation amounts during sorghum growing season and thus higher baseline yield that could not be sustained under future climate conditions characterized by reduced rainfall and increased temperature stress (Figs. S4 and S5). In this context, adjusting planting dates may serve as an effective adaptation strategy. In this study, sorghum was planted in late June, but shifting the sowing window could help align critical growth stages with more favorable climatic conditions. Diawara et al. ( 2024 ) found that sorghum yields were significantly influenced by planting date, with late-May sowing resulting in higher yields than late-June planting, particularly for late-maturing hybrids. These findings suggest that optimizing planting time could help buffer sorghum yields against future climate stressors in semi-arid environments like the THP. The climate impact zone classification showed a somewhat similar but not identical pattern of vulnerability (Fig. 6 E–H). Under RCP 4.5, only a few southern counties were classified as vulnerable by end-century, while most counties remained categorized as stable or more stable. However, under RCP 8.5, the extent of vulnerable and adaptive zones expanded, and fewer counties retained the more stable classification (Fig. 6 H). Interestingly, in a few areas where projected yield reductions ranged between 20–40%, the GiZ classification identified these counties as adaptive or even stable zones (Fig. 6 C and G). This discrepancy highlights the advantage of using spatial clustering metrics like GiZ, which offer a more robust understanding of climate vulnerability than simple yield percentage changes alone. It underscores the importance of considering spatial patterns and statistical significance when interpreting the impact of climate change on crop productivity. Precipitation trends and implications for yield response The spatial patterns of precipitation change for wheat and cotton showed contrasting trends across RCPs and time periods (Fig. 7 ). For wheat, most counties exhibited a reduction in precipitation by mid- and end-century, especially under RCP 8.5, where declines exceeded 30% in southern areas (Fig. 7 C and D). However, this did not uniformly translate into yield losses. As shown earlier (Fig. 2 A–D), wheat yield responses varied, with some western and northern counties showing yield increases of up to 10–30% even under modest rainfall decline. This suggests that other compensatory mechanisms such as CO₂ fertilization effects and possible improvements in water use efficiency may have mitigated the impact of precipitation deficit under elevated CO₂ levels under RCP 8.5. In this context, Asseng et al. ( 2013 ) demonstrated that elevated CO₂ concentrations can increase wheat yields by enhancing photosynthesis and improving water-use efficiency, partially offsetting drought-related stress. Similarly, Rosenzweig et al. ( 2014 ) highlighted that CO₂ enrichment can mitigate yield losses under moderate water limitations in global wheat production systems. For cotton, the correlation between precipitation change and yield response was more pronounced under RCP 4.5 by end-century. In eastern counties, where projected precipitation increased by approximately 10–20% (Fig. 7 B), dryland cotton yields rose sharply, in some cases exceeding 60% (Fig. 3 B), indicating strong moisture sensitivity. In contrast, under RCP 8.5, several counties in the western and southern THP projected neutral or slightly increased cotton yields despite minimal changes or slight decreases in precipitation (Fig. 3 D and Fig. 7 D). This suggests that other climate-related factors, particularly elevated atmospheric CO₂ concentrations under RCP 8.5, may have offset the yield penalties typically associated with marginal water availability by enhancing photosynthetic efficiency and water use efficiency, a phenomenon commonly referred to as the CO₂ fertilization effect. Studies such as Reddy et al. ( 1995 ) and Kimball et al. ( 2002 ) confirm that elevated CO₂ can significantly boost cotton biomass and lint yield, particularly when water is not a limiting factor, supporting the patterns observed in the present study. Despite notable differences in spatial yield responses and precipitation patterns, both wheat and cotton being C 3 crops exhibited yield improvements in certain counties under RCP 8.5, even in the presence of reduced or stagnant precipitation trends during the growing season (Figs. 2 and 3 vs. Figure 7 ). This suggests a potential compensatory role of elevated atmospheric CO₂, which can enhance photosynthetic rates, improve water use efficiency, and partially buffer crops against mild to moderate water stress. Such CO₂ fertilization effects may explain the observed yield increases in drier western regions, where rainfall alone would not support substantial productivity gains. These findings underscore the complexity of climate-yield interactions and caution against interpreting future yield trends solely through changes in precipitation, especially for C 3 crops under high-emission scenarios. The projected rainfall patterns showed distinct spatial and temporal variability across the THP for maize and sorghum (Fig. 8 ). For maize, moderate rainfall increases of 10–20% were projected under RCP 4.5 and RCP 8.5 in southern and eastern counties by mid-century, aligning with yield gains of up to 100% in some areas (Fig. 5 C). These improvements were especially pronounced under RCP 8.5, where the combined effects of increased rainfall and CO₂ fertilization likely contributed to enhanced yield potential. However, by end-century, the northern counties experienced consistent rainfall declines of about 20% under both RCPs (Fig. 8 B, D), which corresponded with sharp yield decreases exceeding 50% (Fig. 5 B, D). This strong correlation reinforces the negative impacts of future drought scenarios on maize productivity in these areas. Importantly, the GiZ-based classification confirmed these observations, identifying the northern counties as vulnerable zones under both RCPs by end-century (Fig. 5 F, H), thus validating the spatially explicit yield–climate relationships and underscoring the need for targeted adaptation in this region. The strong correlation between projected rainfall declines and yield losses in northern counties by end-century aligns with findings from Adhikari et al. ( 2016 ), who highlighted the sensitivity of dryland maize yields to seasonal precipitation under climate change in the THP. Similar to our Gi* clustering results, spatially explicit modeling by Kellner and Niyogi ( 2015 ) in the U.S. Midwest showed that precipitation-driven yield trends can vary regionally, requiring tailored adaptation strategies. In contrast, sorghum showed a broader susceptibility to precipitation decline, particularly under RCP 8.5, where rainfall reductions of more than 30% were projected across much of the central and southern THP (Fig. 8 H). These deficits were accompanied by widespread yield losses, often exceeding 40% across the region (Fig. 6 D). Although the GiZ clustering for sorghum appeared more conservative compared to maize, it still captured the spatial trend of climate impact reasonably well. Vulnerable and adaptive zones were notably concentrated in the southeastern counties under RCP 8.5 (Fig. 6 H), corresponding with the areas of most pronounced rainfall decline and yield reduction. This further supports the spatial congruence between projected precipitation patterns, crop-specific sensitivity, and derived vulnerability zones. This agrees with the findings of Eggen et al. ( 2019 ), who noted that even minor shifts in seasonal rainfall patterns significantly influenced sorghum yield and stability across semiarid regions and how early-season deficits can significantly affect yields. The spatial clustering we observed also mirrors geospatial hotspot detection approaches used to identify vulnerable production zones. Mugiyo et al. ( 2022 ) employed geospatial modeling to map the potential distribution of sorghum under future climate scenarios in sub-Saharan Africa. Their use of the Maximum Entropy (MaxEnt) model to identify climatically suitable areas and production hotspots underlined the importance of spatially explicit approaches for assessing crop vulnerability and adaptive potential. Consistent with our findings, they observed that vulnerability was not uniformly distributed, but rather clustered in specific regions with unfavorable climatic conditions. This reinforces the value of spatial clustering techniques, such as Gi* analysis, for delineating localized zones of risk and opportunity, thereby informing regionally tailored adaptation strategies. Research needs and future directions Previous research has identified climate change adaption strategies that take into account the complex relationship between crop productivity and climatic factors (Adhikari et al., 2016 ; He et al., 2022 ; Kipkulei et al., 2025 ; Kothari et al., 2019a ). Our study takes an important step forward by integrating an improved crop rotation system specifically adapted to the THP. This includes the use of cover crops and DSSAT-calibrated cultivars suited for the region’s semi-arid climate. Nevertheless, further refinements are possible. For instance, in counties identified as vulnerable, especially in the southwestern THP for wheat and sorghum, there is a clear need to invest in breeding and deployment of drought and heat tolerant cultivars to buffer against projected climate extremes. Additionally, exploring the feasibility of supplemental irrigation strategies such as regulated or deficit irrigation during critical phenological stages could help stabilize yields without overburdening the Ogallala Aquifer (Adhikari et al., 2016 ; Attia et al., 2015 ; Kothari et al., 2019a ; Kothari et al., 2019b ; Woli et al., 2023 ). While our study assumed dryland conditions for broad applicability, future work could model adaptive water management scenarios that combine both technological and agronomic interventions. Furthermore, integrating socio-economic variables, e.g., land ownership, access to inputs, or water rights, and land use dynamics could enhance the realism of future projections and guide regionally grounded climate adaptation policies. Expanding the modeling framework to consider multi-risk overlays such as combining climate vulnerability with soil degradation or pest/disease risk could also sharpen the identification of priority intervention zones. Conclusions This study presents a spatially explicit assessment of climate change impacts on major crop yields in the THP, integrating DSSAT modeling and spatial clustering techniques. The findings highlight significant regional variation in crop responses to future climate scenarios. Wheat showed consistent vulnerability in southwestern counties under both RCPs by mid-century and under RCP 4.5 by end-century. Sorghum similarly exhibited localized vulnerability in the southeastern THP, particularly under the high-emission RCP 8.5. In contrast, cotton generally benefitted from climate change across most of the region, except for a few vulnerable zones in the southwestern counties under RCP 8.5 by end-century. Maize demonstrated higher resilience in southern counties, while vulnerability emerged in northern THP counties, especially by the end of the century. The application of spatial clustering metrics such as Moran’s I and Getis-Ord Gi* enabled the identification of statistically significant patterns in yield response across counties. These spatial analyses revealed hotspots of vulnerability and adaptation that may be masked in regional averages, providing a more nuanced understanding of climate risk. The combination of model-based simulations and spatial statistics offers a powerful approach to identifying priority areas for adaptation planning. While yield responses generally tracked with changes in growing season precipitation, some exceptions were observed. For instance, in a few southwestern counties, cotton yields declined under RCP 8.5 by end-century despite only modest reductions in precipitation. This divergence suggests that other factors such as elevated heat stress, soil limitations, or saturation of CO₂ fertilization benefits may play a role in localized yield reductions. These non-linear responses underscore the complexity of climate-crop interactions and the value of spatially disaggregated assessments. Altogether, this study provides a valuable framework for guiding climate-resilient agricultural planning in the THP. By identifying zones of vulnerability and adaptive capacity, the findings support more targeted adaptation strategies, optimized resource allocation, and improved resilience of agricultural systems in semi-arid environments under future climate conditions. Declarations Funding Funding for this research was provided by Texas A&M AgriLife Research. Data availability Climate data for the baseline and future projections were obtained from the Multivariate Adaptive Constructed Analogs (MACA) dataset, available at: https://climate.northwestknowledge.net/MACA/data_portal.php. Gridded soil input data were sourced from the Global High-Resolution Soil Profile Database for Crop Modeling Applications, accessible via the Harvard Dataverse repository: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/1PEEY0. Competing Interests The authors declare no competing interests. References Abatzoglou JT, & Brown TJ (2012). A comparison of statistical downscaling methods suited for wildfire applications. Int. J. Climatol. , 32(5), 772-780. Adhikari P, Ale S, Bordovsky JP, Thorp KR, Modala NR, et al. (2016). Simulating future climate change impacts on seed cotton yield in the Texas High Plains using the CSM-CROPGRO-Cotton model. Agric. Water Manag. , 164, 317-330. Asseng S, Ewert F, Rosenzweig C, Jones JW, Hatfield JL, et al. (2013). Uncertainty in simulating wheat yields under climate change. Nat. Clim. Change , 3(9), 827-832. Attia A, Marohn C, Shawon AR, de Kock A, Strassemeyer J, et al. (2024). Do rotations with cover crops increase yield and soil organic carbon?—A modeling study in southwest Germany. Agric. Ecosyst. Environ. , 375, 109167. Attia A, Rajan N, Nair SS, DeLaune PB, Xue Q, et al. (2016a). Modeling cotton lint yield and water use efficiency responses to irrigation scheduling using Cotton2K. Agron. J. , 108(4), 1614-1623. Attia A, Rajan N, Ritchie G, Cui S, Ibrahim A, et al. (2015). Yield, quality, and spectral reflectance responses of cotton under subsurface drip irrigation. Agron. J. , 107(4), 1355-1364. Attia A, Rajan N, Xue Q, Nair S, Ibrahim A, et al. (2016b). Application of DSSAT-CERES-Wheat model to simulate winter wheat response to irrigation management in the Texas High Plains. Agric. Water Manag. , 165, 50-60. Batjes N (2009). Harmonized soil profile data for applications at global and continental scales: updates to the WISE database. Soil Use Manag. , 25(2), 124-127. DeLaune P, & Mubvumba P (2020). Winter cover crop production and water use in Southern Great Plains cotton. Agron. J. , 112(3), 1943-1951. Diawara B, Diallo S, Traore B, Staggenbord S, & Prasad V (2024). Effect of Planting Date on Yield and Yield Components of Grain Sorghum Hybrids. American Journal of Plant Sciences , 15(5), 387-402. Eggen M, Ozdogan M, Zaitchik B, Ademe D, Foltz J, et al. (2019). Vulnerability of sorghum production to extreme, sub-seasonal weather under climate change. Environmental Research Letters , 14(4), 045005. Getis A, & Ord JK (1992). The analysis of spatial association by use of distance statistics. Geographical analysis , 24(3), 189-206. Gijsman AJ, Thornton PK, & Hoogenboom G (2007). Using the WISE database to parameterize soil inputs for crop simulation models. Comput. Electron. Agric. , 56(2), 85-100. Han E, Ines AV, & Koo J (2019). Development of a 10-km resolution global soil profile dataset for crop modeling applications. Environ. Model. Softw. , 119, 70-83. He Q, Li Liu D, Wang B, Li L, Cowie A, et al. (2022). Identifying effective agricultural management practices for climate change adaptation and mitigation: A win-win strategy in South-Eastern Australia. Agric. Syst. , 203, 103527. Hoogenboom G, Porter CH, Boote KJ, Shelia V, Wilkens PW, et al. (2019). The DSSAT crop modeling ecosystem. Advances in crop modelling for a sustainable agriculture (pp. 173-216). Burleigh Dodds Science Publishing. Jones JW, Hoogenboom G, Porter CH, Boote KJ, Batchelor WD, et al. (2003). The DSSAT cropping system model. Eur. J. Agron. , 18(3-4), 235-265. Kellner O, & Niyogi D (2015). Climate variability and the US Corn Belt: ENSO and AO episode-dependent hydroclimatic feedbacks to corn production at regional and local scales. Earth Interactions , 19(6), 1-32. Kimball B, Kobayashi K, & Bindi M (2002). Responses of agricultural crops to free-air CO2 enrichment. Adv. Agron. , 77, 293-368. Kipkulei HK, Bellingrath-Kimura SD, Lana M, Ghazaryan G, Baatz R, et al. (2025). Modeling the impact of climate change on maize (Zea mays L.) production at the county scale in Kenya. Reg. Environ. Change , 25(2), 62. Kothari K, Ale S, Attia A, Rajan N, Xue Q, et al. (2019a). Potential climate change adaptation strategies for winter wheat production in the Texas High Plains. Agric. Water Manag. , 225, 105764. Kothari K, Ale S, Bordovsky JP, Thorp KR, Porter DO, et al. (2019b). Simulation of efficient irrigation management strategies for grain sorghum production over different climate variability classes. Agric. Syst. , 170, 49-62. Marek GW, Marek TH, Xue Q, Gowda PH, Evett SR, et al. (2017). Simulating evapotranspiration and yield response of selected corn varieties under full and limited irrigation in the Texas High Plains using DSSAT-CERES-Maize. Trans. ASABE , 60(3), 837-846. Moran PA (1950). Notes on continuous stochastic phenomena. Biometrika , 37(1/2), 17-23. Mugiyo H, Chimonyo VGP, Kunz R, Sibanda M, Nhamo L, et al. (2022). Mapping the spatial distribution of underutilised crop species under climate change using the MaxEnt model: A case of KwaZulu-Natal, South Africa. Climate Services , 28, 100330. Neugschwandtner RW, Bernhuber A, Kammlander S, Wagentristl H, Klimek-Kopyra A, et al. (2019). Agronomic potential of winter grain legumes for Central Europe: Development, soil coverage and yields. Field Crops Res. , 241, 107576. Reddy KR, Hodges HF, & McKinion JM (1995). Cotton crop responses to a changing environment. Climate change and agriculture: Analysis of potential international impacts , 59, 3-30. Rosenzweig C, Elliott J, Deryng D, Ruane AC, Müller C, et al. (2014). Assessing agricultural risks of climate change in the 21st century in a global gridded crop model intercomparison. Proceedings of the national academy of sciences , 111(9), 3268-3273. Southern SARE (2024). Water conservation on the High Plains . Sustainable Agriculture Research and Education (SARE) Program. Avaliable online at: https://southern.sare.org/sare-in-your-state/texas/water-conservation-on-the-high-plains/. Accessed 23 April 2025. Taylor KE, Stouffer RJ, & Meehl GA (2012). An overview of CMIP5 and the experiment design. Bull. Am. Meteorol. Soc. , 93(4), 485-498. USDA-NASS (2024). STATE AGRICULTURE OVERVIEW. https://www.nass.usda.gov/Quick_Stats/Ag_Overview/stateOverview.php?state=TEXAS&utm_source=chatgpt.com. Accessed 10 April 2025. Waller S (2023). Irrigated Crops on the US High Plains . Drought, aquifer use, and the future of agriculture. Accessed 23-4-2025 2025. Wen N, Marek GW, Srinivasan R, Brauer DK, Qi J, et al. (2024). Assessing the impacts of long-term climate change on hydrology and yields of diversified crops in the Texas High Plains. Agric. Water Manag. , 302, 108985. Wilson D, & Robson M (1996) 'Pea phenology responses to temperature and photoperiod' Proceedings of the 8th Australian Agronomy Conference . Toowoomba, Queensland, Australia, pp. 590-593. Woli P, Smith GR, Long CR, Rudd JC, Xue Q, et al. (2023). Exploring the potential of cowpea-wheat double cropping in the semi-arid region of the Southern United States using the DSSAT crop model. Agric. Sci. , 14(1), 35-57. Additional Declarations No competing interests reported. Supplementary Files SupplementarymaterialsmappingMS.docx Cite Share Download PDF Status: Published Journal Publication published 11 Sep, 2025 Read the published version in Modeling Earth Systems and Environment → Version 1 posted Editorial decision: Revision requested 09 Jul, 2025 Reviews received at journal 04 Jul, 2025 Reviewers agreed at journal 29 Jun, 2025 Reviewers invited by journal 29 Jun, 2025 Editor assigned by journal 13 Jun, 2025 Submission checks completed at journal 13 Jun, 2025 First submitted to journal 10 Jun, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6864209","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":470763242,"identity":"cb3e0f0e-33c5-4d5e-845c-41c7854610d1","order_by":0,"name":"Ahmed Attia","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5ElEQVRIiWNgGAWjYDACCSDmATGYGRgfQMUMiNIiwcPMwGxwgDQtDAxsEkRp4Z/dnfjgDYNdnT0787PqDzXbEhvYm7dJ4LXkztnNhnMYkoEOYzO7ceDY7cQGnmNleLUw3MjdJs3DwAzyC1ALG1CLRI4ZXi3yN3K3/+ZhqAdqYf9WcOAfUIv8G/xaDIC2MPMwHAZq4TFjONgGsoUHvxbDG7mbJecYHJfsOcxTLHG277ZxG09asQU+LXI3cjd+eFNRzc/ef3zjh4pvt2X72Q9vvIFPC9R5SGw2wspHwSgYBaNgFBACAKgsRoDXDLA0AAAAAElFTkSuQmCC","orcid":"","institution":"Texas A\u0026M AgriLife Research","correspondingAuthor":true,"prefix":"","firstName":"Ahmed","middleName":"","lastName":"Attia","suffix":""},{"id":470763243,"identity":"dd34a15b-54e7-4dc9-9678-073d181172ec","order_by":1,"name":"Prem Woli","email":"","orcid":"","institution":"Texas A\u0026M AgriLife Research","correspondingAuthor":false,"prefix":"","firstName":"Prem","middleName":"","lastName":"Woli","suffix":""},{"id":470763244,"identity":"28db991e-767d-4e88-aa6f-f17511756e2d","order_by":2,"name":"Charles R. Long","email":"","orcid":"","institution":"Texas A\u0026M AgriLife Research","correspondingAuthor":false,"prefix":"","firstName":"Charles","middleName":"R.","lastName":"Long","suffix":""},{"id":470763247,"identity":"4c511ef7-8b38-4a2c-88b1-95ea95eed79a","order_by":3,"name":"Francis M. Rouquette Jr.","email":"","orcid":"","institution":"Texas A\u0026M AgriLife Research","correspondingAuthor":false,"prefix":"","firstName":"Francis","middleName":"M.","lastName":"Rouquette","suffix":"Jr."},{"id":470763248,"identity":"73b6e9c1-0418-455c-8f8c-2ebf33d83003","order_by":4,"name":"Gerald R. Smith","email":"","orcid":"","institution":"Texas A\u0026M AgriLife Research","correspondingAuthor":false,"prefix":"","firstName":"Gerald","middleName":"R.","lastName":"Smith","suffix":""},{"id":470763249,"identity":"cf008871-9bb1-43e0-82f6-a1bcde9aab61","order_by":5,"name":"Amir M.H. Ibrahim","email":"","orcid":"","institution":"Texas A\u0026M University","correspondingAuthor":false,"prefix":"","firstName":"Amir","middleName":"M.H.","lastName":"Ibrahim","suffix":""}],"badges":[],"createdAt":"2025-06-10 14:38:36","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6864209/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6864209/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s40808-025-02605-7","type":"published","date":"2025-09-11T15:56:57+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":84899957,"identity":"f15c8f64-5289-48c0-acf9-9a111d258046","added_by":"auto","created_at":"2025-06-18 14:52:33","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":82481,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of simulated Moran’s I values for wheat (left panels) and cotton (right panels) under RCP 4.5 and RCP 8.5 climate scenarios at mid-century and end-century periods in the Texas High Plains. The red dashed line represents the observed Moran’s I value for each scenario. The significant positive Moran’s I values (p \u0026lt; 0.001) indicate strong spatial autocorrelation of simulated yield changes across counties.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6864209/v1/cebb5ff9f0109d1ddf15c829.png"},{"id":84901497,"identity":"30480ed0-0071-4f88-9c86-9038e3a90e05","added_by":"auto","created_at":"2025-06-18 15:08:34","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":117153,"visible":true,"origin":"","legend":"\u003cp\u003eCounty-level yield change (%) and spatial climate impact zones for wheat in the Texas High Plains under RCP 4.5 and RCP 8.5 scenarios at mid-century (2031-2060) and end-century (2070-2099) compared with baseline (1991-2020). Left panels show the projected percentage change in wheat yields relative to baseline, while right panels classify counties into vulnerable, adaptive, stable, or more stable zones based on Gi* spatial clustering analysis.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6864209/v1/79c151abb291e7b7954c9dae.png"},{"id":84899959,"identity":"cbd3c656-4a7f-4c7c-9223-13919cfd5386","added_by":"auto","created_at":"2025-06-18 14:52:34","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":118748,"visible":true,"origin":"","legend":"\u003cp\u003eCounty-level yield change (%) and spatial climate impact zones for cotton in the Texas High Plains under RCP 4.5 and RCP 8.5 scenarios at mid-century (2031-2060) and end-century (2070-2099) compared with baseline (1991-2020). Left panels show the projected percentage change in wheat yields relative to baseline, while right panels classify counties into vulnerable, adaptive, stable, or more stable zones based on Gi* spatial clustering analysis.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6864209/v1/b0d7082a0eda556dd8aab3dc.png"},{"id":84900682,"identity":"cb5b9c84-ae46-4c10-9f78-4896cb33e259","added_by":"auto","created_at":"2025-06-18 15:00:34","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":78839,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of simulated Moran’s I values for maize (left panels) and sorghum (right panels) under RCP 4.5 and RCP 8.5 climate scenarios at mid-century and end-century periods in the Texas High Plains. The red dashed line represents the observed Moran’s I value for each scenario. The significant positive Moran’s I values (p \u0026lt; 0.001) indicate strong spatial autocorrelation of simulated yield changes across counties.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6864209/v1/e2ed92db777bbd8b0a724dfd.png"},{"id":84902788,"identity":"0f52aa93-27db-4efe-8e2e-bde1cbd9793f","added_by":"auto","created_at":"2025-06-18 15:16:34","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":116351,"visible":true,"origin":"","legend":"\u003cp\u003eCounty-level yield change (%) and spatial climate impact zones for maize in the Texas High Plains under RCP 4.5 and RCP 8.5 scenarios at mid-century (2031-2060) and end-century (2070-2099) compared with baseline (1991-2020). Left panels show the projected percentage change in wheat yields relative to baseline, while right panels classify counties into vulnerable, adaptive, stable, or more stable zones based on Gi* spatial clustering analysis.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6864209/v1/5f220ed1f6bac9bae0069a84.png"},{"id":84899963,"identity":"90c6e5c0-793a-4189-8763-f40ad54f38ad","added_by":"auto","created_at":"2025-06-18 14:52:34","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":115796,"visible":true,"origin":"","legend":"\u003cp\u003eCounty-level yield change (%) and spatial climate impact zones for sorghum in the Texas High Plains under RCP 4.5 and RCP 8.5 scenarios at mid-century (2031-2060) and end-century (2070-2099) compared with baseline (1991-2020). Left panels show the projected percentage change in wheat yields relative to baseline, while right panels classify counties into vulnerable, adaptive, stable, or more stable zones based on Gi* spatial clustering analysis.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-6864209/v1/c31423d4d7a998a14e0162ab.png"},{"id":84899964,"identity":"18f89fe6-b7b4-44fd-80c2-5024ed505932","added_by":"auto","created_at":"2025-06-18 14:52:34","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":117857,"visible":true,"origin":"","legend":"\u003cp\u003eCounty-level precipitation change (%) for wheat and cotton in the Texas High Plains under RCP 4.5 and RCP 8.5 scenarios at mid-century (2031-2060) and end-century (2070-2099) compared with baseline (1991-2020) as projected by five climate models. Left panels show the projected percentage change in precipitation during wheat growing season from 20 October to 20 June relative to baseline, while right panels illustrate percentage change in precipitation during cotton growing season from 1 May to 15 October relative to baseline.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-6864209/v1/ad2de029d7f497a8fbde8616.png"},{"id":84899970,"identity":"cf6b62a8-08ab-4599-a21d-dec88b3704ac","added_by":"auto","created_at":"2025-06-18 14:52:34","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":117324,"visible":true,"origin":"","legend":"\u003cp\u003eCounty-level precipitation change (%) for maize and sorghum in the Texas High Plains under RCP 4.5 and RCP 8.5 scenarios at mid-century (2031-2060) and end-century (2070-2099) compared with baseline (1991-2020) as projected by five climate models. Left panels show the projected percentage change in precipitation during maize growing season from 25 April to 15 September relative to baseline, while right panels show percentage change in precipitation during sorghum growing season from 25 June to 15 November relative to baseline.\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-6864209/v1/1c14d87920df896cf2806452.png"},{"id":91358947,"identity":"0b53d375-2568-4695-800b-2706dd23de3d","added_by":"auto","created_at":"2025-09-15 16:01:36","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1487344,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6864209/v1/6ef21ace-857b-416d-8656-26eaf35966b4.pdf"},{"id":84899967,"identity":"859eabcb-7afa-4cd7-8c5c-d6162161883c","added_by":"auto","created_at":"2025-06-18 14:52:34","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":915791,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementarymaterialsmappingMS.docx","url":"https://assets-eu.researchsquare.com/files/rs-6864209/v1/7602bc723ee1907d6417488c.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Mapping spatial zones of climate vulnerability and adaptive potential for major crops in the Texas High Plains ","fulltext":[{"header":"Introduction","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe Texas High Plains (THP) is a semi-arid region in northwest Texas, encompassing approximately 8.9\u0026nbsp;million ha. Major field crops grown in this area include cotton (\u003cem\u003eGossypium hirsutum\u003c/em\u003e L.), grain sorghum (\u003cem\u003eSorghum bicolor\u003c/em\u003e [L.] Moench), winter wheat (\u003cem\u003eTriticum aestivum\u003c/em\u003e L.), and maize (\u003cem\u003eZea mays\u003c/em\u003e L.), cultivated under both irrigated and rainfed conditions. Irrigated agriculture accounts for about 44% of total cropland in THP in 2022, with maize fields accounts for 20%, and 4.5% each to cotton and winter wheat (Waller, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Grain sorghum, predominantly grown under dryland conditions, planted in 587,000 ha across TX in 2024, highlighting its importance in low-input systems (USDA-NASS, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Dryland agriculture constitutes the remaining percentage of cropland in the THP, rendering these areas particularly vulnerable to climate change due to limited water availability. The Ogallala Aquifer, a major groundwater source underlying the THP, has experienced significant depletion over the past decades, raising concerns about the sustainability of irrigation practices in the region (Southern SARE, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eUnderstanding the spatial heterogeneity of climate impacts on crop productivity is critical for designing region-specific adaptation strategies. In the THP, projected shifts in temperature, precipitation patterns, and growing season length under future climate scenarios are expected to alter crop performance unevenly across counties (Wen et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Identifying climate impact zones such as vulnerable, adaptive, and stable areas based on yield response to climate change allows for more targeted management interventions and resource allocation. Vulnerable zones may require substantial adaptation investments, including drought-tolerant cultivars, optimized irrigation, and soil conservation practices. Conversely, adaptive and more stable zones offer opportunities to enhance resilience and maintain or even increase productivity under changing climatic conditions. Mapping these zones can thus inform and support both short- and long-term agricultural planning at the county scale.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003eRecent advances in crop modeling and geospatial analysis offer powerful tools for evaluating the impacts of climate change on agricultural systems. Process-based models such as DSSAT (Decision Support System for Agrotechnology Transfer) (Hoogenboom et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Jones et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2003\u003c/span\u003e) enable simulation of crop growth, yield, and soil processes under diverse climate and management scenarios. While several studies have explored future crop productivity in the Texas High Plains, most have been limited to specific sites or research stations (Adhikari et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Attia et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2016a\u003c/span\u003e; Kothari et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2019b\u003c/span\u003e; Marek et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). These site-specific studies often yield aggregated findings that are generalized to surrounding areas under the assumption of minimal spatial variability. However, this can obscure localized drivers of production dynamics and lead to oversimplified recommendations. In contrast, spatially explicit assessments that capture crop yield responses at finer resolutions are invaluable for agronomic planning and climate adaptation. Previous research has demonstrated the effectiveness of geospatial techniques for supporting regionally targeted adaptation and resilience planning (Attia et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eMapping zones of vulnerability and adaptive potential for major crops in the THP offers valuable insight into prioritizing localized adaptation strategies, optimizing resource allocation, and informing policy decisions. In this study, we integrated process-based crop modeling with county-scale spatial analysis to identify vulnerable and adaptive production zones for cotton, grain sorghum, wheat, and maize under historical and future climate scenarios. The specific objectives were to: (i) simulate historical and future crop yields using the DSSAT model across all counties in the THP; (ii) classify and map vulnerable and adaptive zones based on yield trajectories; and (iii) assess the relative impacts of climate drivers on productivity trends to inform regional adaptation planning.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy area\u003c/h2\u003e \u003cp\u003eThe THP is a semi-arid agricultural region located in the northwestern part of Texas, covering approximately 8.9\u0026nbsp;million ha (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The region is characterized by a relatively flat terrain and a continental climate with hot summers, cold winters, and erratic precipitation averaging around 450\u0026ndash;550 mm annually, most of which occurs during the summer growing season (April to September). The THP sits atop the Ogallala Aquifer, a vital but rapidly depleting groundwater source that primarily supports irrigated agriculture and livestock operations, while also indirectly benefiting dryland systems through supplemental irrigation and regional agricultural infrastructure (Waller, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Major crops grown in the region include cotton (\u003cem\u003eGossypium hirsutum\u003c/em\u003e L.), maize (\u003cem\u003eZea mays\u003c/em\u003e L.), sorghum (\u003cem\u003eSorghum bicolor\u003c/em\u003e L. Moench), and winter wheat (\u003cem\u003eTriticum aestivum\u003c/em\u003e L.), with varying degrees of reliance on irrigation. Irrigated agriculture accounts for roughly 44% of total cropland (Waller, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), though increasing aquifer depletion has led to a shift toward more water-efficient and rainfed systems. This makes the region especially vulnerable to climate variability and water constraints, necessitating improved understanding of spatial yield responses and adaptive agricultural practices.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eDSSAT model, calibration, and application\u003c/h3\u003e\n\u003cp\u003eDSSAT model is a process-based modeling platform that simulates crop growth, soil dynamics, and management interactions under variable climate conditions (Hoogenboom et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Jones et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). In this study, we applied a 4-year improved rotation system specifically designed for the THP, integrating cover crops (CCs) based on the available windows between main crop cycles (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Cultivar parameters for grain sorghum (\u003cem\u003eSorghum bicolor\u003c/em\u003e L. Moench), maize (\u003cem\u003eZea mays\u003c/em\u003e L.), cotton (\u003cem\u003eGossypium hirsutum\u003c/em\u003e L.), and wheat (\u003cem\u003eTriticum aestivum\u003c/em\u003e L.) were adopted from prior THP-specific calibrations (Adhikari et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Kothari et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019a\u003c/span\u003e; Kothari et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2019b\u003c/span\u003e; Marek et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The following crop rotations and suggested improved N applications were used in the model:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eSorghum\u003c/b\u003e: Planted June 25, harvested November 15, with 55 kg ha⁻\u0026sup1; synthetic N applied at 10 days post-planting and another 55 kg ha⁻\u0026sup1; at the tillering stage. Organic N application through manure: 40 kg ha⁻\u0026sup1;. Total N: 150 kg ha⁻\u0026sup1;.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eWinter legume CC (winter pea, Pisum sativum\u003c/b\u003e \u003cb\u003eL.\u003c/b\u003e\u003cb\u003e)\u003c/b\u003e: Planted November 20, terminated April 20. No synthetic or organic N applied.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eMaize\u003c/b\u003e: Planted April 25, harvested September 15, with 50 kg ha⁻\u0026sup1; synthetic N applied at 10 days post-planting and another 50 kg ha⁻\u0026sup1; at the V6 stage (presence of six fully emerged leaves with visible leaf collars). Organic N application through manure: 50 kg ha⁻\u0026sup1;. Total N: 150 kg ha⁻\u0026sup1;.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eWinter legume CC (winter pea, Pisum sativum\u003c/b\u003e \u003cb\u003eL.\u003c/b\u003e\u003cb\u003e)\u003c/b\u003e: Planted October 10, terminated April 20, fixing\u0026thinsp;~\u0026thinsp;50 kg N ha⁻\u0026sup1;. No synthetic or organic N applied.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eCotton\u003c/b\u003e: Planted May 1, harvested October 15, with 50 kg ha⁻\u0026sup1; synthetic N applied at 10 days post-planting and another 50 kg ha⁻\u0026sup1; at midseason. No organic N applied. Total N: 100 kg ha⁻\u0026sup1;.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eWinter wheat\u003c/b\u003e: Planted October 20, harvested June 20, with 30 kg ha⁻\u0026sup1; synthetic N applied at 10 days post-planting, 30 kg ha⁻\u0026sup1; at the tillering stage, and another 30 kg ha⁻\u0026sup1; later in the season. Organic N application through manure: 50 kg ha⁻\u0026sup1;. Total N: 140 kg ha⁻\u0026sup1;.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eThe winter pea was calibrated based on field experiments conducted at the Texas A\u0026amp;M AgriLife Chillicothe Research Station (DeLaune and Mubvumba, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), supplemented with literature data (Neugschwandtner et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Wilson and Robson, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e1996\u003c/span\u003e). Details on model calibration, initialization, and validation procedures are available in the supplementary materials (Table S2, Fig. S2 and S3). All simulations in this study were conducted under dryland (rainfed) conditions, with no supplemental irrigation applied to any crop.\u003c/p\u003e\n\u003ch3\u003eClimate models and soil data\u003c/h3\u003e\n\u003cp\u003eTo assess climate impacts on crop productivity across the THP, simulations were conducted using five global circulation models (GCMs) and regionally specific soil datasets. Climate projections were based on two representative concentration pathways (RCPs): RCP 4.5, representing a stabilization scenario with emissions peaking around 2040, and RCP 8.5, a high-emissions scenario. CO\u003csub\u003e2\u003c/sub\u003e fertilization was considered according to each RCP scenario. Consequently, CO₂ fertilization effects were dynamically accounted for by incorporating representative CO\u003csub\u003e2\u003c/sub\u003e concentrations associated with each RCP scenario into the DSSAT weather files.\u003c/p\u003e \u003cp\u003eWe selected five GCMs from the Multivariate Adaptive Constructed Analogs (MACA) dataset, developed by the Climate Impacts Group and available via the Northwest Knowledge Network data portal (Abatzoglou and Brown, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Taylor et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). These include IPSL-CM5A-MR, MIROC5, CCSM4, CNRM-CM5, and CSIRO-Mk3-6-0, representing a diverse range of climate sensitivities and modeling approaches (Table S3). All projected results presented in this study represent ensemble means averaged across these five GCMs. Gridded soil information was sourced from the WISE database (Batjes, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Gijsman et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), developed by the ISRIC SoilGrids initiative, and further processed following the methods of Han et al. (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) to ensure compatibility with DSSAT requirements. Soils were characterized and aggregated to match the simulation grid resolution used in climate modeling. A point shapefile layer representing a 12 km resolution grid was created to facilitate the spatio-temporal application of the DSSAT model across the THP. Simulations were conducted for three time periods: baseline (1991\u0026ndash;2020), mid-century (2031\u0026ndash;2060), and end-century (2070\u0026ndash;2099), under two RCPs: 4.5 and 8.5. For each grid point, a Python-based pipeline was used to automate DSSAT simulations across spatial grid points, incorporating climate, soil, and crop management inputs. The pipeline also enabled batch post-processing of model outputs to evaluate temporal and spatial impacts of climate and crop rotation scenarios on productivity, soil carbon dynamics, and system-level sustainability.\u003c/p\u003e\n\u003ch3\u003eClimate change impact assessment analysis and spatial clustering\u003c/h3\u003e\n\u003cp\u003eTo assess the spatial variability in wheat yield changes across the THP, yield changes were calculated for each county by comparing the baseline period to two future time periods: mid-century and end-century under RCP 4.5 and RCP 8.5 scenarios. These differences were expressed as percentage change in yield (future vs. baseline), and the spatial clustering and zoning analysis was applied to interpret patterns of vulnerability and stability.\u003c/p\u003e \u003cp\u003eFirst, spatial yield changes were computed by interpolating gridded simulation results from DSSAT model outputs. A point shapefile with simulation output locations was joined with the county boundary shapefile of the THP. The shapefile contained county-level polygons (n\u0026thinsp;=\u0026thinsp;48 counties) and was projected to WGS84 (EPSG:4326). The yield change data were prepared in a long format across four scenarios: RCP 4.5 mid-century, RCP 4.5 end-century, RCP 8.5 mid-century, and RCP 8.5 end-century.\u003c/p\u003e \u003cp\u003eTo calculate average yield changes per county, a spatial join was performed between point-level yield change estimates and county boundary polygons using the \u003cem\u003esf\u003c/em\u003e package in R software. After spatially linking each yield point to its respective county, the data were aggregated by county and scenario using the \u003cem\u003eddply()\u003c/em\u003e function from the \u003cem\u003eplyr\u003c/em\u003e package to compute both the mean and standard deviation of yield change. These aggregated values enabled consistent comparisons across counties and scenarios.\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\text{Y}\\text{i}\\text{e}\\text{l}\\text{d}\\:\\text{c}\\text{h}\\text{a}\\text{n}\\text{g}\\text{e}\\:\\left(\\%\\right)=\\frac{\\text{F}\\text{u}\\text{t}\\text{u}\\text{r}\\text{e}\\:\\text{y}\\text{i}\\text{e}\\text{l}\\text{d}-\\text{B}\\text{a}\\text{s}\\text{e}\\text{l}\\text{i}\\text{n}\\text{e}\\:\\text{y}\\text{i}\\text{e}\\text{l}\\text{d}\\:}{\\text{B}\\text{a}\\text{s}\\text{e}\\text{l}\\text{i}\\text{n}\\text{e}\\:\\text{y}\\text{i}\\text{e}\\text{l}\\text{d}}\\:\\:\\times\\:100\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\left(1\\right)\\:$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eNext, we applied Global Moran\u0026rsquo;s I (Moran, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1950\u003c/span\u003e) to determine the degree of spatial autocorrelation in the county-level yield change data. Moran\u0026rsquo;s I is defined as:\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:I=\\frac{N}{W}+\\frac{{\\sum\\:}_{i=1}^{N}{\\sum\\:}_{j=1}^{N}{w}_{ij}({x}_{i}-\\stackrel{-}{x})({x}_{j}-\\:\\stackrel{-}{x}\\:)}{{\\sum\\:_{i=1}^{N}\\left({x}_{i}-\\:\\stackrel{-}{x}\\right)}^{2}}\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\:\\left(2\\right)$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\:N\\)\u003c/span\u003e\u003c/span\u003e is the number of counties, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{x}_{i},\\:{x}_{j}\\)\u003c/span\u003e\u003c/span\u003e are the yield change at locations \u003cem\u003ei\u003c/em\u003e and \u003cem\u003ej\u003c/em\u003e, respectively, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{-}{x}\\)\u003c/span\u003e\u003c/span\u003e is the mean of all \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{x}_{i}\\)\u003c/span\u003e\u003c/span\u003e values, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{w}_{ij}\\)\u003c/span\u003e\u003c/span\u003e is the spatial weight between location \u003cem\u003ei\u003c/em\u003e and \u003cem\u003ej\u003c/em\u003e, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:W\\)\u003c/span\u003e\u003c/span\u003e is the sum of all spatial weights. A significant positive Moran's I indicates that counties with similar yield changes are spatially clustered, while negative values indicate spatial dispersion.\u003c/p\u003e \u003cp\u003eFor finer resolution, the Getis-Ord Gi* statistic (Getis and Ord, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e1992\u003c/span\u003e) was computed to classify local spatial clusters. Gi* analysis identifies hotspots and coldspots by measuring the statistical significance of high or low yield changes in the context of neighboring counties. The resulting Z-scores of the Gi* statistic were classified into four climate impact zones: vulnerable, adaptive, stable, and more stable. A vulnerable zone defines counties with negative GiZ greater than two standard deviations of the mean (95%) for all points within that county, i.e. GiZ \u0026le; -1.96 (statistically significant yield decline), adaptive zone defines counties with negative GiZ of spatial association equal to or greater than one standard deviation of the mean, i.e. -1.96\u0026thinsp;\u0026lt;\u0026thinsp;GiZ \u0026le; -0.96. Crop yield is expected to be slightly reduced or remain neutral due to adaptation mechanisms in these adaptive zones. Stable zone defines counties with neutral to slight insignificant increase in future crop yield, where average historical level of production is expected to be maintained. The GiZ for those ranges from \u0026minus;\u0026thinsp;0.96 to 0.96. The more stable zone defines counties where future yield is expected to be significantly higher than the historical yield, GiZ is \u0026ge;\u0026thinsp;0.96, suggesting beneficial climate change impacts.\u003c/p\u003e \u003cp\u003eThese classifications were applied for each crop and RCP scenario independently. The zone classification aids in identifying counties that are projected to benefit or suffer from climate change impacts on crop productivity, helping to prioritize adaptation strategies. In addition, growing season precipitation was extracted from climate input datasets used for model simulations. Precipitation data were aggregated specifically for each crop\u0026rsquo;s growing season, and changes between baseline and future periods were mapped under different RCP scenarios. These maps of spatial rainfall change were analyzed alongside yield change patterns, facilitating interpretation of crop responses to projected precipitation variability. This analysis provided insights into potential constraints or benefits of water availability during critical crop phenological stages, thus improving the understanding of climate-related impacts on crop performance across the THP.\u003c/p\u003e"},{"header":"Results and discussion","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eSpatial clustering and climate-impact zones for wheat and cotton\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe spatial distribution of simulated county-level yield changes for wheat and cotton exhibited significant positive spatial autocorrelation under both RCP 4.5 and RCP 8.5 scenarios at mid-century and end-century periods (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The global Moran's I values were high for wheat across all scenarios, ranging from 0.565 to 0.782 with p-values\u0026thinsp;\u0026lt;\u0026thinsp;0.001, indicating strong spatial clustering of similar yield changes across counties. Similarly, cotton demonstrated significant but slightly lower Moran's I values ranging from 0.405 to 0.566 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), suggesting moderately strong spatial structure in yield change patterns.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe county-level wheat yield projections revealed considerable variability across the THP (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Negative yield changes dominated the south and central counties, particularly under RCP 8.5 in mid-century, with yield losses of about 10\u0026ndash;30% (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC-D). In contrast, northern counties exhibited positive yield changes, with increases ranging from 30 to 50% by end-century under RCP 8.5. Consequently, the vulnerable zones were predominantly located in the southern counties, while stable and more stable zones were clustered in the northern part of the region (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE-H). However, it should be noted that the baseline yields in these northern counties were generally lower than those in central and western counties (Fig. S5), which could be attributed to less amount of rainfall (Fig. S5). In this context, regulated deficit irrigation at grain-filling stage of wheat was reported to increase dryland yield by 68% in THP (Attia et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016b\u003c/span\u003e). This suggests that supplemental irrigation could help elevate yields to approximately 1.7 Mg ha\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in these water-limited areas, supporting both productivity and resilience under future climate stress. For cotton, the county-level yield projections showed a generally favorable outlook under future climate scenarios (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Under RCP 4.5 mid-century, most counties experienced seed yield increases of approximately 20\u0026ndash;40% (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). By end-century, regardless of RCP, yield gains became more substantial, with many counties experiencing increases ranging from 60% up to more than 100% (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD).\u003c/p\u003e \u003cp\u003eSpatial climate impact zoning for cotton revealed a dominance of stable and more stable zones across the THP under both RCPs and time periods. Particularly under RCP 8.5 by end-century (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eH), the expansion of more stable zones highlighted cotton\u0026rsquo;s higher adaptive capacity compared to wheat. Only isolated vulnerable zones were observed for cotton, mainly in a few southwestern counties under RCP 8.5 end-century (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eH). It is noteworthy that despite an overall positive yield trend, spatial hotspot analysis (Getis-Ord Gi) revealed the emergence of localized vulnerable zones. This finding underscores that county-level yield increases do not always translate to uniform spatial resilience, emphasizing the importance of identifying spatial disparities even under favorable climate trajectories. This occurs because GiZ statistics detect spatial clustering patterns independent of overall mean shifts, areas with significantly lower-than-expected yield changes compared to their neighbors are flagged as vulnerable, even when general yield increases are observed. In statistical terms, while the mean yield change is positive, localized negative deviations relative to surrounding counties still trigger significant clustering of \"low\" values. Therefore, spatial disparity analysis complements broad-scale projections by identifying localized zones of vulnerability that may otherwise be masked by broad regional averages.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSpatial clustering and climate-impact zones for maize and sorghum\u003c/h3\u003e\n\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe spatial clustering of simulated maize and sorghum yields changes also demonstrated statistically significant spatial autocorrelation across the THP under all future climate scenarios (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). For maize, the global Morans\u0026rsquo;s I values range from 0.77 (RCP 4.5 mid-century) to 0.911 (RCP 8.5 end-century), indicating very strong clustering of yield change responses. Sorghum showed similarly robust but slightly lower spatial autocorrelation, with Moran\u0026rsquo;s I values ranging from 0.434 to 0.764 (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001 in all cases), reflecting meaningful spatial structure across counties (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eCounty-level maps revealed clear spatial contrasts in maize yield responses (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Under RCP 4.5, several southern counties were projected to experience up to 40\u0026ndash;60% increase in maize yield by the end of the century. Under RCP 8.5, however, spatial variability was greater, with southern and southwestern counties showing gains exceeding 100% by mid-century (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC), while northern regions experienced more modest or stable trends. However, by end-century northern regions showed yield decrease by 50% (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD). These yield patterns translated into climate impact zones, with southern counties consistently classified as more stable under both scenarios, while several northern and Panhandle counties emerged as vulnerable or adaptive zones by mid- and end-century. This spatial differentiation of yield stability aligns with findings by Kipkulei et al. (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), who used DSSAT-CERES-Maize modeling in Kenya to show yield declines of up to 41% and identified climate hotspots and stable zones to guide targeted adaptation strategies.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eFor sorghum, the yield projections painted a more concerning picture under future climate scenarios, particularly under RCP 8.5 (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). Under RCP 4.5, sorghum yields were projected to increase by 10\u0026ndash;20% in several eastern and northern counties during the mid-century period, but these gains diminished by the end of the century, with most counties shifting to neutral or experiencing yield declines of around 10% (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA and B). In contrast, under RCP 8.5, widespread yield declines dominated across the THP, with several counties experiencing reductions exceeding 40% by end-century (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC and D). This could be attributed to relatively higher baseline precipitation amounts during sorghum growing season and thus higher baseline yield that could not be sustained under future climate conditions characterized by reduced rainfall and increased temperature stress (Figs. S4 and S5). In this context, adjusting planting dates may serve as an effective adaptation strategy. In this study, sorghum was planted in late June, but shifting the sowing window could help align critical growth stages with more favorable climatic conditions. Diawara et al. (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) found that sorghum yields were significantly influenced by planting date, with late-May sowing resulting in higher yields than late-June planting, particularly for late-maturing hybrids. These findings suggest that optimizing planting time could help buffer sorghum yields against future climate stressors in semi-arid environments like the THP.\u003c/p\u003e \u003cp\u003eThe climate impact zone classification showed a somewhat similar but not identical pattern of vulnerability (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eE\u0026ndash;H). Under RCP 4.5, only a few southern counties were classified as vulnerable by end-century, while most counties remained categorized as stable or more stable. However, under RCP 8.5, the extent of vulnerable and adaptive zones expanded, and fewer counties retained the more stable classification (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eH). Interestingly, in a few areas where projected yield reductions ranged between 20\u0026ndash;40%, the GiZ classification identified these counties as adaptive or even stable zones (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC and G). This discrepancy highlights the advantage of using spatial clustering metrics like GiZ, which offer a more robust understanding of climate vulnerability than simple yield percentage changes alone. It underscores the importance of considering spatial patterns and statistical significance when interpreting the impact of climate change on crop productivity.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003ePrecipitation trends and implications for yield response\u003c/h3\u003e\n\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe spatial patterns of precipitation change for wheat and cotton showed contrasting trends across RCPs and time periods (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). For wheat, most counties exhibited a reduction in precipitation by mid- and end-century, especially under RCP 8.5, where declines exceeded 30% in southern areas (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eC and D). However, this did not uniformly translate into yield losses. As shown earlier (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA\u0026ndash;D), wheat yield responses varied, with some western and northern counties showing yield increases of up to 10\u0026ndash;30% even under modest rainfall decline. This suggests that other compensatory mechanisms such as CO₂ fertilization effects and possible improvements in water use efficiency may have mitigated the impact of precipitation deficit under elevated CO₂ levels under RCP 8.5. In this context, Asseng et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) demonstrated that elevated CO₂ concentrations can increase wheat yields by enhancing photosynthesis and improving water-use efficiency, partially offsetting drought-related stress. Similarly, Rosenzweig et al. (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) highlighted that CO₂ enrichment can mitigate yield losses under moderate water limitations in global wheat production systems.\u003c/p\u003e \u003cp\u003eFor cotton, the correlation between precipitation change and yield response was more pronounced under RCP 4.5 by end-century. In eastern counties, where projected precipitation increased by approximately 10\u0026ndash;20% (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB), dryland cotton yields rose sharply, in some cases exceeding 60% (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB), indicating strong moisture sensitivity. In contrast, under RCP 8.5, several counties in the western and southern THP projected neutral or slightly increased cotton yields despite minimal changes or slight decreases in precipitation (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD and Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eD). This suggests that other climate-related factors, particularly elevated atmospheric CO₂ concentrations under RCP 8.5, may have offset the yield penalties typically associated with marginal water availability by enhancing photosynthetic efficiency and water use efficiency, a phenomenon commonly referred to as the CO₂ fertilization effect. Studies such as Reddy et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e1995\u003c/span\u003e) and Kimball et al. (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2002\u003c/span\u003e) confirm that elevated CO₂ can significantly boost cotton biomass and lint yield, particularly when water is not a limiting factor, supporting the patterns observed in the present study.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDespite notable differences in spatial yield responses and precipitation patterns, both wheat and cotton being C\u003csub\u003e3\u003c/sub\u003e crops exhibited yield improvements in certain counties under RCP 8.5, even in the presence of reduced or stagnant precipitation trends during the growing season (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e vs. Figure\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). This suggests a potential compensatory role of elevated atmospheric CO₂, which can enhance photosynthetic rates, improve water use efficiency, and partially buffer crops against mild to moderate water stress. Such CO₂ fertilization effects may explain the observed yield increases in drier western regions, where rainfall alone would not support substantial productivity gains. These findings underscore the complexity of climate-yield interactions and caution against interpreting future yield trends solely through changes in precipitation, especially for C\u003csub\u003e3\u003c/sub\u003e crops under high-emission scenarios.\u003c/p\u003e \u003cp\u003eThe projected rainfall patterns showed distinct spatial and temporal variability across the THP for maize and sorghum (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). For maize, moderate rainfall increases of 10\u0026ndash;20% were projected under RCP 4.5 and RCP 8.5 in southern and eastern counties by mid-century, aligning with yield gains of up to 100% in some areas (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC). These improvements were especially pronounced under RCP 8.5, where the combined effects of increased rainfall and CO₂ fertilization likely contributed to enhanced yield potential. However, by end-century, the northern counties experienced consistent rainfall declines of about 20% under both RCPs (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB, D), which corresponded with sharp yield decreases exceeding 50% (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB, D). This strong correlation reinforces the negative impacts of future drought scenarios on maize productivity in these areas. Importantly, the GiZ-based classification confirmed these observations, identifying the northern counties as vulnerable zones under both RCPs by end-century (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eF, H), thus validating the spatially explicit yield\u0026ndash;climate relationships and underscoring the need for targeted adaptation in this region. The strong correlation between projected rainfall declines and yield losses in northern counties by end-century aligns with findings from Adhikari et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), who highlighted the sensitivity of dryland maize yields to seasonal precipitation under climate change in the THP. Similar to our Gi* clustering results, spatially explicit modeling by Kellner and Niyogi (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) in the U.S. Midwest showed that precipitation-driven yield trends can vary regionally, requiring tailored adaptation strategies.\u003c/p\u003e \u003cp\u003eIn contrast, sorghum showed a broader susceptibility to precipitation decline, particularly under RCP 8.5, where rainfall reductions of more than 30% were projected across much of the central and southern THP (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eH). These deficits were accompanied by widespread yield losses, often exceeding 40% across the region (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eD). Although the GiZ clustering for sorghum appeared more conservative compared to maize, it still captured the spatial trend of climate impact reasonably well. Vulnerable and adaptive zones were notably concentrated in the southeastern counties under RCP 8.5 (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eH), corresponding with the areas of most pronounced rainfall decline and yield reduction. This further supports the spatial congruence between projected precipitation patterns, crop-specific sensitivity, and derived vulnerability zones. This agrees with the findings of Eggen et al. (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), who noted that even minor shifts in seasonal rainfall patterns significantly influenced sorghum yield and stability across semiarid regions and how early-season deficits can significantly affect yields.\u003c/p\u003e \u003cp\u003eThe spatial clustering we observed also mirrors geospatial hotspot detection approaches used to identify vulnerable production zones. Mugiyo et al. (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) employed geospatial modeling to map the potential distribution of sorghum under future climate scenarios in sub-Saharan Africa. Their use of the Maximum Entropy (MaxEnt) model to identify climatically suitable areas and production hotspots underlined the importance of spatially explicit approaches for assessing crop vulnerability and adaptive potential. Consistent with our findings, they observed that vulnerability was not uniformly distributed, but rather clustered in specific regions with unfavorable climatic conditions. This reinforces the value of spatial clustering techniques, such as Gi* analysis, for delineating localized zones of risk and opportunity, thereby informing regionally tailored adaptation strategies.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eResearch needs and future directions\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003ePrevious research has identified climate change adaption strategies that take into account the complex relationship between crop productivity and climatic factors (Adhikari et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; He et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Kipkulei et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Kothari et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019a\u003c/span\u003e). Our study takes an important step forward by integrating an improved crop rotation system specifically adapted to the THP. This includes the use of cover crops and DSSAT-calibrated cultivars suited for the region\u0026rsquo;s semi-arid climate. Nevertheless, further refinements are possible. For instance, in counties identified as vulnerable, especially in the southwestern THP for wheat and sorghum, there is a clear need to invest in breeding and deployment of drought and heat tolerant cultivars to buffer against projected climate extremes. Additionally, exploring the feasibility of supplemental irrigation strategies such as regulated or deficit irrigation during critical phenological stages could help stabilize yields without overburdening the Ogallala Aquifer (Adhikari et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Attia et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Kothari et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019a\u003c/span\u003e; Kothari et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2019b\u003c/span\u003e; Woli et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). While our study assumed dryland conditions for broad applicability, future work could model adaptive water management scenarios that combine both technological and agronomic interventions. Furthermore, integrating socio-economic variables, e.g., land ownership, access to inputs, or water rights, and land use dynamics could enhance the realism of future projections and guide regionally grounded climate adaptation policies. Expanding the modeling framework to consider multi-risk overlays such as combining climate vulnerability with soil degradation or pest/disease risk could also sharpen the identification of priority intervention zones.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThis study presents a spatially explicit assessment of climate change impacts on major crop yields in the THP, integrating DSSAT modeling and spatial clustering techniques. The findings highlight significant regional variation in crop responses to future climate scenarios. Wheat showed consistent vulnerability in southwestern counties under both RCPs by mid-century and under RCP 4.5 by end-century. Sorghum similarly exhibited localized vulnerability in the southeastern THP, particularly under the high-emission RCP 8.5. In contrast, cotton generally benefitted from climate change across most of the region, except for a few vulnerable zones in the southwestern counties under RCP 8.5 by end-century. Maize demonstrated higher resilience in southern counties, while vulnerability emerged in northern THP counties, especially by the end of the century.\u003c/p\u003e \u003cp\u003eThe application of spatial clustering metrics such as Moran\u0026rsquo;s I and Getis-Ord Gi* enabled the identification of statistically significant patterns in yield response across counties. These spatial analyses revealed hotspots of vulnerability and adaptation that may be masked in regional averages, providing a more nuanced understanding of climate risk. The combination of model-based simulations and spatial statistics offers a powerful approach to identifying priority areas for adaptation planning.\u003c/p\u003e \u003cp\u003eWhile yield responses generally tracked with changes in growing season precipitation, some exceptions were observed. For instance, in a few southwestern counties, cotton yields declined under RCP 8.5 by end-century despite only modest reductions in precipitation. This divergence suggests that other factors such as elevated heat stress, soil limitations, or saturation of CO₂ fertilization benefits may play a role in localized yield reductions. These non-linear responses underscore the complexity of climate-crop interactions and the value of spatially disaggregated assessments. Altogether, this study provides a valuable framework for guiding climate-resilient agricultural planning in the THP. By identifying zones of vulnerability and adaptive capacity, the findings support more targeted adaptation strategies, optimized resource allocation, and improved resilience of agricultural systems in semi-arid environments under future climate conditions.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding \u003c/strong\u003eFunding for this research was provided by Texas A\u0026amp;M AgriLife Research. \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability \u003c/strong\u003eClimate data for the baseline and future projections were obtained from the Multivariate Adaptive Constructed Analogs (MACA) dataset, available at: https://climate.northwestknowledge.net/MACA/data_portal.php. Gridded soil input data were sourced from the Global High-Resolution Soil Profile Database for Crop Modeling Applications, accessible via the Harvard Dataverse repository: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/1PEEY0.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests \u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAbatzoglou JT, \u0026amp; Brown TJ (2012). A comparison of statistical downscaling methods suited for wildfire applications. \u003cem\u003eInt. J. Climatol.\u003c/em\u003e, 32(5), 772-780.\u003c/li\u003e\n \u003cli\u003eAdhikari P, Ale S, Bordovsky JP, Thorp KR, Modala NR, et al. (2016). Simulating future climate change impacts on seed cotton yield in the Texas High Plains using the CSM-CROPGRO-Cotton model. \u003cem\u003eAgric. Water Manag.\u003c/em\u003e, 164, 317-330.\u003c/li\u003e\n \u003cli\u003eAsseng S, Ewert F, Rosenzweig C, Jones JW, Hatfield JL, et al. (2013). Uncertainty in simulating wheat yields under climate change. \u003cem\u003eNat. Clim. Change\u003c/em\u003e, 3(9), 827-832.\u003c/li\u003e\n \u003cli\u003eAttia A, Marohn C, Shawon AR, de Kock A, Strassemeyer J, et al. (2024). Do rotations with cover crops increase yield and soil organic carbon?\u0026mdash;A modeling study in southwest Germany. \u003cem\u003eAgric. Ecosyst. Environ.\u003c/em\u003e, 375, 109167.\u003c/li\u003e\n \u003cli\u003eAttia A, Rajan N, Nair SS, DeLaune PB, Xue Q, et al. (2016a). Modeling cotton lint yield and water use efficiency responses to irrigation scheduling using Cotton2K. \u003cem\u003eAgron. J.\u003c/em\u003e, 108(4), 1614-1623.\u003c/li\u003e\n \u003cli\u003eAttia A, Rajan N, Ritchie G, Cui S, Ibrahim A, et al. (2015). Yield, quality, and spectral reflectance responses of cotton under subsurface drip irrigation. \u003cem\u003eAgron. J.\u003c/em\u003e, 107(4), 1355-1364.\u003c/li\u003e\n \u003cli\u003eAttia A, Rajan N, Xue Q, Nair S, Ibrahim A, et al. (2016b). Application of DSSAT-CERES-Wheat model to simulate winter wheat response to irrigation management in the Texas High Plains. \u003cem\u003eAgric. Water Manag.\u003c/em\u003e, 165, 50-60.\u003c/li\u003e\n \u003cli\u003eBatjes N (2009). Harmonized soil profile data for applications at global and continental scales: updates to the WISE database. \u003cem\u003eSoil Use Manag.\u003c/em\u003e, 25(2), 124-127.\u003c/li\u003e\n \u003cli\u003eDeLaune P, \u0026amp; Mubvumba P (2020). Winter cover crop production and water use in Southern Great Plains cotton. \u003cem\u003eAgron. J.\u003c/em\u003e, 112(3), 1943-1951.\u003c/li\u003e\n \u003cli\u003eDiawara B, Diallo S, Traore B, Staggenbord S, \u0026amp; Prasad V (2024). Effect of Planting Date on Yield and Yield Components of Grain Sorghum Hybrids. \u003cem\u003eAmerican Journal of Plant Sciences\u003c/em\u003e, 15(5), 387-402.\u003c/li\u003e\n \u003cli\u003eEggen M, Ozdogan M, Zaitchik B, Ademe D, Foltz J, et al. (2019). Vulnerability of sorghum production to extreme, sub-seasonal weather under climate change. \u003cem\u003eEnvironmental Research Letters\u003c/em\u003e, 14(4), 045005.\u003c/li\u003e\n \u003cli\u003eGetis A, \u0026amp; Ord JK (1992). The analysis of spatial association by use of distance statistics. \u003cem\u003eGeographical analysis\u003c/em\u003e, 24(3), 189-206.\u003c/li\u003e\n \u003cli\u003eGijsman AJ, Thornton PK, \u0026amp; Hoogenboom G (2007). Using the WISE database to parameterize soil inputs for crop simulation models. \u003cem\u003eComput. Electron. Agric.\u003c/em\u003e, 56(2), 85-100.\u003c/li\u003e\n \u003cli\u003eHan E, Ines AV, \u0026amp; Koo J (2019). Development of a 10-km resolution global soil profile dataset for crop modeling applications. \u003cem\u003eEnviron. Model. Softw.\u003c/em\u003e, 119, 70-83.\u003c/li\u003e\n \u003cli\u003eHe Q, Li Liu D, Wang B, Li L, Cowie A, et al. (2022). Identifying effective agricultural management practices for climate change adaptation and mitigation: A win-win strategy in South-Eastern Australia. \u003cem\u003eAgric. Syst.\u003c/em\u003e, 203, 103527.\u003c/li\u003e\n \u003cli\u003eHoogenboom G, Porter CH, Boote KJ, Shelia V, Wilkens PW, et al. (2019). The DSSAT crop modeling ecosystem. \u003cem\u003eAdvances in crop modelling for a sustainable agriculture\u003c/em\u003e (pp. 173-216). Burleigh Dodds Science Publishing.\u003c/li\u003e\n \u003cli\u003eJones JW, Hoogenboom G, Porter CH, Boote KJ, Batchelor WD, et al. (2003). The DSSAT cropping system model. \u003cem\u003eEur. J. Agron.\u003c/em\u003e, 18(3-4), 235-265.\u003c/li\u003e\n \u003cli\u003eKellner O, \u0026amp; Niyogi D (2015). Climate variability and the US Corn Belt: ENSO and AO episode-dependent hydroclimatic feedbacks to corn production at regional and local scales. \u003cem\u003eEarth Interactions\u003c/em\u003e, 19(6), 1-32.\u003c/li\u003e\n \u003cli\u003eKimball B, Kobayashi K, \u0026amp; Bindi M (2002). Responses of agricultural crops to free-air CO2 enrichment. \u003cem\u003eAdv. Agron.\u003c/em\u003e, 77, 293-368.\u003c/li\u003e\n \u003cli\u003eKipkulei HK, Bellingrath-Kimura SD, Lana M, Ghazaryan G, Baatz R, et al. (2025). Modeling the impact of climate change on maize (Zea mays L.) production at the county scale in Kenya. \u003cem\u003eReg. Environ. Change\u003c/em\u003e, 25(2), 62.\u003c/li\u003e\n \u003cli\u003eKothari K, Ale S, Attia A, Rajan N, Xue Q, et al. (2019a). Potential climate change adaptation strategies for winter wheat production in the Texas High Plains. \u003cem\u003eAgric. Water Manag.\u003c/em\u003e, 225, 105764.\u003c/li\u003e\n \u003cli\u003eKothari K, Ale S, Bordovsky JP, Thorp KR, Porter DO, et al. (2019b). Simulation of efficient irrigation management strategies for grain sorghum production over different climate variability classes. \u003cem\u003eAgric. Syst.\u003c/em\u003e, 170, 49-62.\u003c/li\u003e\n \u003cli\u003eMarek GW, Marek TH, Xue Q, Gowda PH, Evett SR, et al. (2017). Simulating evapotranspiration and yield response of selected corn varieties under full and limited irrigation in the Texas High Plains using DSSAT-CERES-Maize. \u003cem\u003eTrans. ASABE\u003c/em\u003e, 60(3), 837-846.\u003c/li\u003e\n \u003cli\u003eMoran PA (1950). Notes on continuous stochastic phenomena. \u003cem\u003eBiometrika\u003c/em\u003e, 37(1/2), 17-23.\u003c/li\u003e\n \u003cli\u003eMugiyo H, Chimonyo VGP, Kunz R, Sibanda M, Nhamo L, et al. (2022). Mapping the spatial distribution of underutilised crop species under climate change using the MaxEnt model: A case of KwaZulu-Natal, South Africa. \u003cem\u003eClimate Services\u003c/em\u003e, 28, 100330.\u003c/li\u003e\n \u003cli\u003eNeugschwandtner RW, Bernhuber A, Kammlander S, Wagentristl H, Klimek-Kopyra A, et al. (2019). Agronomic potential of winter grain legumes for Central Europe: Development, soil coverage and yields. \u003cem\u003eField Crops Res.\u003c/em\u003e, 241, 107576.\u003c/li\u003e\n \u003cli\u003eReddy KR, Hodges HF, \u0026amp; McKinion JM (1995). Cotton crop responses to a changing environment. \u003cem\u003eClimate change and agriculture: Analysis of potential international impacts\u003c/em\u003e, 59, 3-30.\u003c/li\u003e\n \u003cli\u003eRosenzweig C, Elliott J, Deryng D, Ruane AC, M\u0026uuml;ller C, et al. (2014). Assessing agricultural risks of climate change in the 21st century in a global gridded crop model intercomparison. \u003cem\u003eProceedings of the national academy of sciences\u003c/em\u003e, 111(9), 3268-3273.\u003c/li\u003e\n \u003cli\u003eSouthern SARE (2024). \u003cem\u003eWater conservation on the High Plains\u003c/em\u003e. Sustainable Agriculture Research and Education (SARE) Program. Avaliable online at: https://southern.sare.org/sare-in-your-state/texas/water-conservation-on-the-high-plains/. Accessed 23 April 2025.\u003c/li\u003e\n \u003cli\u003eTaylor KE, Stouffer RJ, \u0026amp; Meehl GA (2012). An overview of CMIP5 and the experiment design. \u003cem\u003eBull. Am. Meteorol. Soc.\u003c/em\u003e, 93(4), 485-498.\u003c/li\u003e\n \u003cli\u003eUSDA-NASS (2024). STATE AGRICULTURE OVERVIEW. https://www.nass.usda.gov/Quick_Stats/Ag_Overview/stateOverview.php?state=TEXAS\u0026amp;utm_source=chatgpt.com. Accessed 10 April 2025.\u003c/li\u003e\n \u003cli\u003eWaller S (2023). \u003cem\u003eIrrigated Crops on the US High Plains\u003c/em\u003e. Drought, aquifer use, and the future of agriculture. Accessed 23-4-2025 2025.\u003c/li\u003e\n \u003cli\u003eWen N, Marek GW, Srinivasan R, Brauer DK, Qi J, et al. (2024). Assessing the impacts of long-term climate change on hydrology and yields of diversified crops in the Texas High Plains. \u003cem\u003eAgric. Water Manag.\u003c/em\u003e, 302, 108985.\u003c/li\u003e\n \u003cli\u003eWilson D, \u0026amp; Robson M (1996) \u0026apos;Pea phenology responses to temperature and photoperiod\u0026apos; \u003cem\u003eProceedings of the 8th Australian Agronomy Conference\u003c/em\u003e. Toowoomba, Queensland, Australia, pp. 590-593.\u003c/li\u003e\n \u003cli\u003eWoli P, Smith GR, Long CR, Rudd JC, Xue Q, et al. (2023). Exploring the potential of cowpea-wheat double cropping in the semi-arid region of the Southern United States using the DSSAT crop model. \u003cem\u003eAgric. Sci.\u003c/em\u003e, 14(1), 35-57.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"modeling-earth-systems-and-environment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mese","sideBox":"Learn more about [Modeling Earth Systems and Environment](http://link.springer.com/journal/40808)","snPcode":"40808","submissionUrl":"https://submission.springernature.com/new-submission/40808/3","title":"Modeling Earth Systems and Environment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Climate change, Spatial clustering, Yield vulnerability, Dryland agriculture, climate impact zones","lastPublishedDoi":"10.21203/rs.3.rs-6864209/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6864209/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eClimate change poses an increasing threat to agricultural productivity in the Texas High Plains (THP), a semi-arid region facing both warming trends and declining groundwater resources. This study integrates process-based crop modeling with geospatial analysis to identify spatial zones of climate vulnerability and adaptive potential for four major crops; winter wheat, cotton, maize, and grain sorghum under future climate scenarios. Using the DSSAT model, historical (1991\u0026ndash;2020) and future yields (2031\u0026ndash;2060 and 2070\u0026ndash;2099) were simulated across 48 counties under Representative Concentration Pathway 4.5 and 8.5 (RCP 4.5 and RCP 8.5). Spatial clustering techniques, including Global Moran\u0026rsquo;s I and Getis-Ord Gi* statistics, were applied to classify counties into vulnerable, adaptive, stable, and more stable zones based on projected yield changes. Results revealed that wheat vulnerability was concentrated in southern counties, with projected yield decreases of 10\u0026ndash;30% under RCP 8.5, while northern counties showed 30\u0026ndash;50% yield increases under RCP 4.5 mid-century and RCP 8.5 end-century. In contrast, cotton yields are projected to increase by 20\u0026ndash;40% across most counties under RCP 4.5 end-century and RCP 8.5 mid- and end-century, with localized vulnerability emerging in southwestern THP under RCP 8.5 by end-century. Grain sorghum yields are projected to increase by 10\u0026ndash;20% in eastern and northern counties under RCP 4.5, but under RCP 8.5 widespread yield declines exceeding 40% are expected by end-century, attributed to reduced rainfall and increased temperature stress during the growing season. In contrast, maize showed greater resilience, with yield changes varying spatially but remaining positive in many southern counties. These spatially explicit findings underscore the need for targeted adaptation strategies, including the deployment of climate-resilient crop varieties, optimized irrigation management, crop diversification, and adaptive land use planning. The study offers actionable insights to support climate-resilient agricultural planning and inform precision adaptation policies for sustaining crop productivity in the THP under future climate scenario\u003c/p\u003e","manuscriptTitle":"Mapping spatial zones of climate vulnerability and adaptive potential for major crops in the Texas High Plains","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-18 14:52:29","doi":"10.21203/rs.3.rs-6864209/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-09T17:52:58+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-04T22:02:38+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"187738122263370958072378808219198403834","date":"2025-06-29T18:11:51+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-06-29T15:47:01+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-13T07:29:20+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-13T07:23:40+00:00","index":"","fulltext":""},{"type":"submitted","content":"Modeling Earth Systems and Environment","date":"2025-06-10T14:32:59+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"modeling-earth-systems-and-environment","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"mese","sideBox":"Learn more about [Modeling Earth Systems and Environment](http://link.springer.com/journal/40808)","snPcode":"40808","submissionUrl":"https://submission.springernature.com/new-submission/40808/3","title":"Modeling Earth Systems and Environment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"6a05b9af-f284-495a-956e-8ce1907af912","owner":[],"postedDate":"June 18th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-09-15T15:58:17+00:00","versionOfRecord":{"articleIdentity":"rs-6864209","link":"https://doi.org/10.1007/s40808-025-02605-7","journal":{"identity":"modeling-earth-systems-and-environment","isVorOnly":false,"title":"Modeling Earth Systems and Environment"},"publishedOn":"2025-09-11 15:56:57","publishedOnDateReadable":"September 11th, 2025"},"versionCreatedAt":"2025-06-18 14:52:29","video":"","vorDoi":"10.1007/s40808-025-02605-7","vorDoiUrl":"https://doi.org/10.1007/s40808-025-02605-7","workflowStages":[]},"version":"v1","identity":"rs-6864209","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6864209","identity":"rs-6864209","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}