Climate scenarios and their impacts on soybean grain yield in the Central-West region of Brazil1ABSTRACT
The objective of this study was to assess the variability of annual and monthly precipitation under current and future climate conditions, and to characterize its effects on soybean grain yield in the municipality of Tangará da Serra, Mato Grosso state, Brazil, using future climate projections from General Circulation Models (GCMs) and the Decision Support System for Agrotechnology Transfer (DSSAT), employing the CROPGRO-Soybean module. Model calibration was based on climate, soil, and management data, as well as genetic coefficients and field observations from the 2015/2016 growing season, considering cultivars with different maturity groups. Simulations were performed for six different sowing dates. Results indicated both positive and negative monthly fluctuations in precipitation under the current scenario. Among these, significant reductions were observed in September, October, and December, along with a general decreasing trend in annual precipitation for the region. Future scenarios (RCPs 4.5 and 8.5) showed both positive and negative variations, though not statistically significant. Earlyand intermediate-maturing cultivars exhibited lower yields and greater variability, whereas the late-maturing cultivar showed superior yield across all scenarios and sowing dates. Early sowings (September 20 to October 20) resulted in the lowest yields, while late sowings (November 1 and 10) were more favorable. Annual average precipitation was 1711 mm (current), 1494 mm (RCP 4.5), and 1515 mm (RCP 8.5). It is concluded that climate variability significantly affects soybean grain yield, with late-maturing cultivars and delayed sowings being recommended to enhance adaptation under future climate conditions.
Key words:
climate; general circulation models; DSSAT-CROPGRO
HIGHLIGHTS:
The CROPGRO-Soybean model effectively simulates soybean grain yield for the projected scenarios from 2030 to 2060.
Future soybean grain yield is dependent on management, late sowings and late-maturing cultivars showing increased yield.
Late sowing periods are characterized by higher yields under both Representative Concentration Pathway (RCP) scenarios.
RESUMO
O objetivo deste estudo foi avaliar a variabilidade da precipitação anual e mensal sob condições climáticas atuais e futuras, e caracterizar seus efeitos na produtividade da soja no município de Tangará da Serra, Mato Grosso. Foram utilizadas projeções climáticas futuras de Modelos de Circulação Geral (GCMs) e o Decision Support System for Agrotechnology Transfer (DSSAT), empregando o módulo CROPGRO-Soybean. A calibração do modelo baseou-se em dados climáticos, de solo e de manejo, bem como em coeficientes genéticos e observações de campo da safra 2015/2016, considerando cultivares com diferentes grupos de maturação. As simulações foram realizadas para seis diferentes datas de semeadura. Os resultados indicaram flutuações mensais positivas e negativas na precipitação no cenário atual. Dentre estas, observaram-se reduções significativas em setembro, outubro e dezembro, além de uma tendência geral de decréscimo na precipitação anual para a região. Os cenários futuros (RCPs 4.5 e 8.5) apresentaram tanto variações positivas quanto negativas, embora não estatisticamente significativas. As cultivares precoce e intermediária exibiram menores produtividades e maior variabilidade, enquanto a cultivar tardia apresentou produtividade superior em todos os cenários e datas de semeadura. As semeaduras precoces (20 de setembro a 20 de outubro) resultaram nas menores produtividades, enquanto as semeaduras tardias (1 e 10 de novembro) foram mais favoráveis. A precipitação média anual foi de 1711 mm (atual), 1494 mm (RCP 4.5) e 1515 mm (RCP 8.5). Conclui-se que a variabilidade climática afeta significativamente a produtividade da soja, sendo recomendadas cultivares tardias e semeaduras tardias para melhorar a adaptação sob as condições climáticas futuras.
Palavras-chave:
clima; modelos de circulação geral; DSSAT-CROPGRO
INTRODUCTION
Brazil is the world’s largest soybean producer, having reached a record output of 156 million metric tons in the 2022/2023 season, surpassing the United States and Argentina. In the 2023/2024 season, the country produced 147.35 million metric tons, accounting for approximately 37% of global soybean production (USDA, 2023).
The state of Mato Grosso accounts for 29% of Brazil’s total soybean production. During the 2022/2023 period, it produced 45.6 million metric tons, while forecasts for 2023/2024 indicated a 13.72% reduction in output for the region. The main factor behind these losses was climatic variability, especially during sowing carried out from September to October, when this period was characterized by low rainfall and high temperatures (CONAB, 2024).
Climate is the primary determinant of agricultural activities, as meteorological variables influence all phases of the phenological cycle of crops, directly affecting vegetative development rates, yield, and the incidence of pests and diseases (Hoogenboom et al., 2019). Climate change represents the foremost global challenge for the future, with alterations in air temperature, rainfall patterns, extreme weather events, and increased frequency of droughts potentially leading to various impacts on agriculture and human life (Song et al., 2022).
Forecasting future climate conditions is therefore considered essential for obtaining more accurate information on the behavior of meteorological variables under different scenarios. Moreover, assessing the potential impacts on agricultural crops has become a strategic approach to measure the necessary adaptations to possible climate change scenarios (Song et al., 2022).
General Circulation Models (GCMs) are regarded as one of the best tools for simulating past climate and projecting future conditions. Based on emission scenarios, this methodology enables the analysis of specific changes in each meteorological variable, allowing the identification of major projected changes in the region and supporting a more detailed assessment of future conditions (Salehie et al., 2023).
Considering the effects of climate change on agricultural systems, studies focused on future simulation models are essential for understanding climate behavior patterns, soil conditions, and how these factors influence crop performance. These models enable the simulation of known processes and the prediction of possible system responses, also assisting in decision-making regarding optimal sowing times and locations across different regions (Hoogenboom et al., 2019).
The Decision Support System for Agrotechnology Transfer (DSSAT) performs simulations focusing on crop growth, development, and yield. The core modules within the software address climate, soil, plant dynamics, and the interactions between climate, plant, and atmosphere (Jones et al., 2003). The main factors affecting plant growth and development include crop genotypes, climate conditions, soil properties, and field management practices (Wang et al., 2023). CROPGRO-Soybean stands out as a model that simulates vegetative and reproductive development, growth, and crop yield, accounting for the interaction of factors such as crop characteristics, climatic conditions, soil properties, and diverse management strategies (Bhatia et al., 2008).
From this perspective, the objective of this study was to assess the variability of annual and monthly precipitation under current and future climate conditions, and to characterize its effects on soybean grain yield in the municipality of Tangará da Serra, Mato Grosso state, Brazil, using future scenarios derived from General Circulation Models (GCMs) and the DSSAT crop simulation model.
MATERIAL AND METHODS
The research was conducted at the State University of Mato Grosso, at the Professor Eugênio Carlos Stieler Campus, located in the municipality of Tangará da Serra in the year 2025, within the facilities of the Technological Center for Geoprocessing and Remote Sensing (CETEGEO-SR). The variability of annual and monthly precipitation under current and future climate conditions and its effect on soybean grain yield were analyzed for Tangará da Serra, Mato Grosso state (Figure 1).
Spatial location and geographical boundaries of Tangará da Serra municipality within the state of Mato Grosso, Brazil
According to the Köppen classification, the region is characterized as megathermal or tropical with dry winters (Aw) (Santos et al., 2024). The predominant soil type is classified as Typic Haplorthox (Oxisol) according to the USDA Soil Taxonomy, featuring a very fclayey texture with a mean clay content of 641 g kg⁻1. The soil exhibits water availability and penetration resistance values of 1.649 mm cm⁻1 and 1.94 MPa, respectively (Daniel et al., 2022). The municipality is situated within a transition zone between the Cerrado and Amazon biomes, with terrain ranging from flat to gently undulating (Moreira & Vasconcelos, 2007)
Yield simulations for the soybean cultivars were carried out using the CROPGRO-Soybean module, developed for simulating growth, development, and yield, within the Decision Support System for Agrotechnology Transfer (DSSAT).
To calibrate the CROPGRO-Soybean model, a minimum dataset is required for operation. Climate, soil, and management parameters, as well as genetic coefficients of the varieties, were used along with field experiment data regarding growth rates and yield collected in Tangará da Serra during the 2015/2016 growing season. These data were sourced from an experiment conducted by Barbieri et al. (2020). In addition, information on sowing date, emergence, flowering, physiological maturity, 1000-grain weight, and yield were obtained from the same study.
The indices used for calibration and validation of the models were derived from three soybean varieties: SoyTech 815 RR (early maturity), SoyTech 820 RR (intermediate maturity), and Tropical Melhoramento & Genética Ltda (late maturity). The cultivars were sown on four different planting dates: date 1 (09/22/2015), date 2 (10/06/2015), date 3 (10/21/2015), and date 4 (11/05/2015), respectively, for each cultivar.
Harvest dates followed the growth cycle of each cultivar: First sowing date: 01/11/2016, 01/20/2016, 02/08/2016; Second sowing date: 02/01/2016, 02/05/2016, 03/03/2016; Third sowing date: 02/18/2016, 02/24/2016, 03/20/2016; Fourth sowing date: 03/10/2016, 03/15/2016, 03/30/2016.
These harvest dates refer respectively to the three cultivars. For further details on the experiment and the procedures used to collect growth and yield data, refer to Barbieri et al. (2020).
Calibration of the crop’s genetic characteristics followed the recommendations of Jones et al. (2003) and Hoogenboom et al. (2019), using field-collected data, beginning with the best-performing sowing date (date 3). Validation was carried out across all sowing dates. Table 1 presents the genetic coefficients used to simulate soybean growth and development. These coefficients are applied to simulate and predict daily crop responses to climate, soil conditions, and management practices. The genetic coefficients were implemented in the CROPGRO-Soybean module, following the methodology outlined by Barbieri et al. (2019).
The climatic variables considered as input for the CROPGRO-Soybean model, used in the simulations, were: daily maximum and minimum air temperature (°C), daily rainfall (mm), and solar radiation (MJ m⁻2). These data were obtained through the execution of a script in the Code Editor on the Google Earth Engine platform, referring to the latest version of the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP), conducted by the Coupled Model Intercomparison Project Phase 6 (CMIP6) for RCPs 4.5 and 8.5. These datasets have a spatial resolution of 0.25º × 0.25º for all variables within the period from 2030 to 2060, for the MIROC6 model (Thrasher et al., 2022).
The simulated variables, input into the DSSAT models, served to analyze the potential effects of future climate changes on the yield of early-, intermediate-, and late-maturing soybean cultivars with different sowing dates. Trends were evaluated using the Mann-Kendall statistical test, which is characteristic for time series analysis, indicating a smooth and monotonic increase or decrease in the series indices. A positive Z (Tau) result indicates an increasing trend in the time series, while a negative Z (Tau) implies a decreasing trend; both scenarios are considered significant according to the Mann-Kendall test.
Certain soil input data are required for calibration, as they regulate the water balance within the module (Ritchie, 1998). In this study, soil data regarding field capacity, permanent wilting point, soil depth, and total soil water availability were entered into DSSAT. These data, along with proportions of sand, silt, and clay, were obtained from the work of Daniel et al. (2022) for the municipality of Tangará da Serra, considering soil depths of 0-10, 10-20, 20-30, and 30-40 cm.
Simulations were performed for six different sowing dates to determine yield variation over these intervals, considering the calibrated soil parameters and climatic variables. The simulated sowing dates were: Date 1 (20 September), Date 2 (1 October), Date 3 (10 October), Date 4 (20 October), Date 5 (1 November), and Date 6 (10 November). Statistical analyses were performed considering a significance level of 5% (p < 0.05).
RESULTS AND DISCUSSION
The annual distribution of precipitation in Tangará da Serra is characterized by fluctuations between periods with rainfall volumes above and below the region’s climatological average across all studied scenarios. The highest annual average precipitation is observed in the current scenario (Baseline), with 1,711.91 mm per year (Figure 2A). Future projections indicate a reduction in total rainfall, with averages of 1,494.45 mm for RCPs 4.5 (Figure 2B) and 1,515.37 mm for RCP 8.5 (Figure 2C).
Annual precipitation distribution in Tangará da Serra (MT, Brazil). Historical series from 1970 to 2023 (A), Future projection (2030-2060) under the intermediate emissions scenario (RCP 4.5) (B), Future projection (2030-2060) under the high emissions scenario (RCP 8.5) (C)
Viana et al. (2025) updated the studies for the municipality of Tangará da Serra and pointed out the occurrence of changes in rainfall patterns in the last 50 years, in which: the average annual precipitation for the municipality of Tangará da Serra was 1,774.69 mm, behavior possibly associated with the frequency of occurrence of ENSO events. This phenomenon causes impacts around the world with different intensities, as also observed by Yen & Nguyet (2025), evaluating rice yield under these phenomena and working with a historical data series from 1985 to 2022.
Given that water is essential for agricultural development, changes in rainfall patterns may negatively affect yield levels, reducing water availability and increasing the incidence of droughts, thus impacting all life within an ecosystem (Hoogenboom et al., 2019).
In studying the feasibility of double cropping of soybean and maize in certain regions of Mato Grosso state, Andrea et al. (2020), using future scenarios and the same RCPs as in the present study, already identified a reduction in rainfall levels for several municipalities. The authors reported that 17% of these areas may become unsuitable for agriculture due to changes in rainfall patterns.
Rainfall is a key climatic variable for the functioning and maintenance of global ecosystems. Estimating rainfall patterns is essential for planning a wide range of human activities, particularly in the agricultural sector, where it plays a vital role in assessing water availability (Dallacort et al., 2011). Rain is considered the main source of water input in the hydrological cycle, and its distribution and variability are directly related to atmospheric dynamics and regional topography (Barbieri et al., 2019).
Considering a reduction in annual rainfall volume for Tangará da Serra (MT, Brazil), trend lines show a marked decline in the Baseline scenario, with a negative Mann-Kendall Tau statistic, indicating a lower volume of precipitation compared to previous years (Figure 3A).
Annual precipitation trend analysis (Mann-Kendall test) in Tangará da Serra (MT, Brazil) for the historical period and future projections: historical period (Baseline; 1970-2023) (A), future projection (RCP 4.5; 2030-2060) (B), and future projection (RCP 8.5; 2030-2060) (C)
Moreover, this reduction is statistically significant (p ≤ 0.05) for the Baseline (Figure 3A). For future scenarios, the downward trend and negative Tau value indicate a decline in annual precipitation for RCP 4.5 (Figure 3B), although not statistically significant (p > 0.05). For RCP 8.5 (Figure 3C), the trend lines suggest an increase in annual precipitation; however, as with RCP 4.5, this trend is also not statistically significant (p > 0.05).
Tangará da Serra is located in a transition zone between the Cerrado and Amazon biomes, which may result in greater climatic variability, particularly in rainfall patterns. It is well established that ecotone regions such as these exhibit higher rainfall levels (Campos & Chaves, 2020). The climatic sensitivity of the Amazon to variability can be attributed to both natural system changes and anthropogenic activities, such as the increase in greenhouse gas (GHG) concentrations. Land use and land cover changes, deforestation, and agricultural activities impact the hydrological cycle and, consequently, rainfall patterns (Marengo et al., 2024).
The precipitation analyses for January and March (Figures 4A and C), corresponding to the peak of the rainy season and the end of summer, did not exhibit statistically significant trends. During these months, persistent high interannual variability was observed in both the historical series and future projections (RCP 4.5 and RCP 8.5).
Monthly precipitation distribution in Tangará da Serra, MT, Brazil, under future RCP 4.5 and RCP 8.5 scenarios. The panels represent distributions for different months: January (A), February (B), March (C), April (D), May (E), June (F), July (G), August (H), September (I), October(J), November (K), and December (L)
The monthly distribution of precipitation in Tangará da Serra is characterized by two well-defined seasons: a wet season from October to April (Figures 4J and D), and a dry season from June to August (Figures 4F and H). The months of May and September (Figures 4E and I) are classified as transitional, marking the shift between seasons, with decreasing rainfall in May and increasing rainfall in September.
In contrast, July (Figure 4G), which marks the peak of the dry season, showed a marked decline in Baseline precipitation, suggesting a historical intensification of drought conditions. For November (Figure 4K), a period characterized by the establishment of rainfall, a decreasing trend was identified in the historical series, indicating a potential delay in the onset of the rainy season. However, this trend diverges in the RCP 8.5 scenario, which projects an increasing trend in precipitation.
When evaluating the RCP scenarios separately, positive and negative fluctuations in monthly rainfall levels are evident, particularly during the wet season, with values below 200 mm in several years under both scenarios. This variability is more pronounced under RCP 8.5, with monthly rainfall peaks reaching 400 mm in December, and lows of 74 mm in February, highlighting the uncertainty in monthly rainfall distribution for this scenario (Figures 4L and B).
These findings are consistent with those reported by Barbieri et al. (2019), who also characterized the precipitation regime in the region as consisting of two distinct seasons: a dry season and a rainy season. Similar results were reported by Dallacort et al. (2011), although their study identified December as having the lowest standard deviation during the rainy season. This characterizes December as the month with the most uniform rainfall distribution.
Positive and negative trends are observed in several months throughout the year for the Baseline, RCP 4.5, and RCP 8.5 scenarios. When evaluating the entire time series, although trend lines suggest increases or decreases in some months, these tendencies are not statistically significant (p > 0.05), except for September, October, and December (Figures 4I, J and L)
Due to the extension of the dry season in the region, monthly levels show a sharp decrease in precipitation from September onwards (the transition to the rainy season). Furthermore, this negative trend is statistically significant (p ≤ 0.05). For the future scenarios, the time series trend shows a similar pattern of reduction, though the changes under RCPs 4.5 and 8.5 are not statistically significant.
In addition to monthly rainfall decline, December also exhibits a significant negative trend for the Baseline scenario (p ≤ 0.05), indicating a decrease in rainfall in recent years compared to previous decades. In the RCP scenarios, both increases and decreases in December precipitation are observed; however, these trends are not statistically significant (p > 0.05) in either scenario.
For the dry season under future scenarios, a greater amplitude in monthly precipitation values is expected, compared to previously cited studies. During the study period, for both RCPs, monthly values exceeded 100 mm in August, and reached 50 mm in June and July. This distribution may suggest a possible shift in the onset or delay of the rainy season, as the months that follow frequently showed either higher values (indicative of earlier onset) or lower values (indicative of delay).
In their assessment of the impacts of climate change on rainfall across the Amazon region, with projections covering the whole of Brazil, Rocha et al. (2019) identified a reduction in precipitation levels, particularly in the Amazon, Pantanal, Central-West, South, and Southeast regions. Furthermore, the authors found that the impacts are most pronounced during the rainy season. These results further contribute to the understanding of how such changes affect meteorological variables, with implications for the water balance, ecosystem dynamics, and increased frequency of droughts and floods.
The CROPGRO-Soybean model effectively simulated the yield of cultivars in Tangará da Serra under projected conditions from 2030 to 2060. The early and intermediate cultivars showed greater variability in yield across both RCP scenarios, while the late-maturing cultivar consistently achieved higher yield across all sowing dates. Early sowings (20 September, 01 October, 10 October, and 20 October) were the most negatively affected in terms of yield, whereas later sowings (01 November and 10 November) exhibited the most stable yield for all three cultivars under both RCPs (Figures 5A and D).
Simulated soybean grain yield under future climate scenarios for Tangará da Serra (MT, Brazil). Panels show projections for: A-C) the intermediate emissions scenario (RCP 4.5) and D-F) the high emissions scenario (RCP 8.5). Columns represent maturity groups: Early (A, D), Mid (B, E), and Late (C, F)
The earliest sowing dates were associated with reduced yield for all cultivars, particularly for the early and intermediate cultivars, in sowing windows 1, 2, and 3 (Figures 5A, B, D and E). These sowing windows displayed the highest volatility in yield, with minimum yield dropping below 2,000 kg ha⁻1 across both RCPs. Conversely, for the same sowing periods and cultivars, maximum yields exceeded 3,000 kg ha⁻1, with the intermediate cultivar exceeding 4,000 kg ha⁻1 in sowing window 3 under both RCP 4.5 and RCP 8.5.
The late-maturing soybean cultivar showed superior grain yields in the early sowing dates (1, 2, and 3) under both RCP 4.5 and RCP 8.5 scenarios, with yield values exceeding 5,000 kg ha⁻1 (Figures 5C and F). These results may be explained by a delay in certain phenological stages, allowing the cultivar to avoid the adverse effects of elevated temperatures and reduced water availability over short periods, as simulated in this study’s climate scenarios.
In their analysis of soybean grain yield in different sowing windows under ENSO events, Barbieri et al. (2019) similarly observed that early sowings were more detrimental to yield. Furthermore, the late-maturing cultivar in that study also showed the highest yield levels in response to climate variability. Another notable aspect is the high variability in yield, as seen in the current study, especially for the early and intermediate cultivars, which showed larger gaps between maximum and minimum yields.
Future water availability conditions may become more variable due to potential shifts in the onset or end of the rainy season. Silva et al. (2021), investigating soybean yields under climate change, identified increasing temperature trends and decreasing precipitation in some Brazilian regions, including the Central-West. These authors reported an increase of 1.55 to 3.02 ºC in air temperature and a variation of -4.5 to +7.9% in future precipitation, under both RCP 4.5 and 8.5 scenarios.
Agricultural crops are highly sensitive to any climatic changes, particularly fluctuations in air temperature and rainfall, which can significantly affect plant phenology. Responses to climate change are speciesand stage-dependent, and impacts may include alterations in photosynthetic rates, yield variability, and shortening of phenological cycles (Battisti et al., 2017).
The late sowing dates (windows 4, 5, and 6) were associated with higher yields across both RCP scenarios and all maturity groups. In sowing window 4, variability in yield is still observed for the early and intermediate cultivars, with minimum values falling below 1,000 kg ha⁻1 under both RCPs.
The final two sowing dates stood out with the highest yields across all cultivars and both RCPs. These periods also showed lower yield variability, establishing them as the most suitable sowing windows for the region under future conditions. Maximum and minimum yields ranged between 3,000 and 5,000 kg ha⁻1, highlighting their greater consistency. Battisti et al. (2017) had already noted the higher yield risks associated with early sowing, recommending later sowing dates to enhance cultivar performance, owing to greater water availability linked to higher rainfall during these periods.
The average soybean grain yield, as a function of the evaluated scenarios and sowing dates (Table 2), was lower in the early sowing windows. For the early-maturing cultivar, the first two sowing dates produced yields below 3,000 kg ha⁻1. The highest yield for this group was 3,304.3 kg ha⁻1 under RCP 8.5 in sowing window 4. The intermediate-maturing cultivar reached yields above 3,000 kg ha⁻1 starting from the second sowing date, with a maximum of 3,645.8 kg ha⁻1 in window 5 under RCP 4.5. The late-maturing cultivar showed the highest overall average yields, even in early sowing windows, achieving top yields across both RCPs, with a maximum yield of 4,665.5 kg ha⁻1.
Average soybean grain yield (kg ha⁻1) of three cultivars under RCP 4.5 and RCP 8.5 climate scenarios in Tangará da Serra, MT, Brazil
When evaluating the main soybean-producing regions in Brazil across different sowing dates, using CROPGRO-Soybean, Battisti & Sentelhas (2019) reported average yields ranging from 2,801 to 3,400 kg ha⁻1 for the Central-West region during the period from 1980 to 2013. Their study considered different maturity groups and sowing dates from 15 September to 15 February.
Considering the intervals between sowing window 1 (20 October) and window 6 (10 November), several yield values fall within this range. In some RCP scenarios, values exceed this scale, especially for the late-maturing cultivar. Moreover, the authors highlighted that early sowing dates may negatively impact yield, particularly due to water deficits and extreme temperature events.
Under the RCP 4.5 scenario (Figure 6), grain yield probability is characterized by greater instability during the early sowing windows (Figures 6A, B, and C), particularly for shorter cycles. The analysis reveals differences among maturity groups: at the 75% probability level, early-maturing cultivars exhibited yields ranging from 2,300 to 2,700 kg ha⁻1, while intermediate cultivars showed yields ranging from 2,500 to 2,960 kg ha⁻1. In contrast, late-maturing cultivars exhibited greater stability and yield potential, with yields ranging from 3,610 to 4,080 kg ha⁻1 during these same periods.
Probability distribution of soybean grain yield (kg ha-1) under the RCP 4.5 climate scenario for Tangará da Serra (MT, Brazil). The panels represent projections for different sowing dates: Sowing 1 (A); Sowing 2 (B); Sowing 3 (C); Sowing 4 (D); Sowing 5 (E); and Sowing 6 (F)
Although performance improved in the later sowing windows (Figures 6D, E, and F), limitations were still observed for the early-maturing cycle, which remained near the 3,000 kg ha⁻1 threshold (ranging from 3,070 to 3,140 kg ha⁻1). Conversely, the intermediate cultivar surpassed these levels (3,430 kg ha⁻1), while the late-maturing cultivar maintained high yields, ranging from 4,090 to 4,140 kg ha⁻1.
Under the RCP 8.5 scenario (Figure 7), earlyand intermediate-maturing cultivars exhibit yield instability during the early sowing windows. At the 75% probability level, yields for the early-maturing cultivar range from 2,220 to 2,610 kg ha⁻1 (sowing dates 1 to 3), while the yield for the intermediate cultivar ranges between 2,410 and 2,850 kg ha⁻1, approaching 3,000 kg ha⁻1 only for the third sowing date (Figures 7A, B, and C).
Probability distribution of soybean grain yield (kg ha-1) under the RCP 8.5 climate scenario for Tangará da Serra (MT, Brazil). The panels represent projections for different sowing dates: Sowing 1 (A); Sowing 2 (B); Sowing 3 (C); Sowing 4 (D); Sowing 5 (E); and Sowing 6 (F)
From the fourth sowing window onwards (Figures 7D, E, and F), risk decreases, yet differences persist: the early-maturing cultivar stabilizes just above the 3,000 kg ha⁻1 threshold (ranging from 3,010 to 3,110 kg ha⁻1), whereas the intermediate cultivar displays superior performance, reaching up to 3,390 kg ha⁻1.
In contrast, the late-maturing cultivar exhibits superior performance and consistency across all scenarios. It maintains a 75% probability of achieving yields above 3,800 kg ha⁻1 even in the earliest sowing windows, with its performance becoming even more stable and elevated in the late sowing windows.
Overall, considering the future scenarios, a potential for increased yield levels is observed for the municipality of Tangará da Serra. However, achieving yields that exceed the local average of 3,000 kg ha⁻1 reported by Barbieri et al. (2019), and the state average of 3,320.8 kg ha⁻1 (CONAB, 2024), is highly dependent on the choice of cultivar. The results indicate that this increase is primarily associated with the late-maturing cultivar, especially when sown in later windows, as the earlyand intermediate-maturing cultivars present significant risks of not reaching these averages.
When studying the El Niño Southern Oscillation phenomenon in relation to sowing dates and soybean grain yield in southern Brazil, Nóia Júnior et al. (2020) identified that, for most studied regions, the highest yields occurred for sowing dates from late October to early November. This finding corroborates the present study, where sowing windows 4 and 5 (20 October and 1 November) recorded the highest yields. The authors also emphasize that sowings outside this period, either earlier or later than November, tend to negatively affect yield. Their simulations were based on historical data from 1961 to 2016 using a cultivar with a maturity group of 6.5, classified as intermediate.
The increase in soybean grain yield under future climate scenarios has also been investigated by Silva et al. (2021) for the state of Mato Grosso. Both studies conducted simulations using various RCP scenarios and models; however, due to climate change, these yields, after an initial increase, tend to decline subsequently, primarily as a result of elevated air temperatures.
CONCLUSIONS
-
1. The annual average precipitation for the baseline scenario is 1,711.9 mm. For future conditions, a reduction of 217.5 and 196.6 mm is observed for RCPs 4.5 and 8.5, respectively.
-
2. The months of September, October, and December present a significant reduction (p ≤ 0.05) in precipitation for the baseline scenario. The same pattern is observed under future conditions, although the reductions are not significant (p > 0.05).
-
3. Yield simulations for future scenarios indicate that late-maturing cultivars show the highest average yields for both RCPs. Intermediateand early-maturing cultivars are more susceptible to water deficit and show high yield variability. Although sowing windows 4, 5, and 6 (20 October to 10 November) are characterized as more suitable, the yield gains and stability in these windows are primarily associated with the late-maturing cultivar; sowing after 20 October is recommended.
Data Availability Statement:
The data that support the findings of this study are available on request from the corresponding author.
Acknowledgments:
To the Coordination for the Improvement of Higher Education Personnel (CAPES), for granting the first author a master’s scholarship.
LITERATURE CITED
-
Andrea, M. C. S.; Dallacort, R.; Tieppo, R. C.; Barbieri, J. D. Assessment of climate change impact on double-cropping systems. SN Applied Sciences, v.2, p.1-13, 2020. https://doi.org/10.1007/s42452-020-2325-z
» https://doi.org/10.1007/s42452-020-2325-z -
Barbieri, J. D.; Dallacort, R.; Freitas, P. S. L. de; Araújo, D. V. de; Tieppo, R. C.; Fenner, W. Effects of the ENSO on the variability of precipitation and air temperature in agricultural regions of Mato Grosso State. Journal of Agricultural Science, v.11, p.91, 2019. https://doi.org/10.5539/jas.v11n9p91
» https://doi.org/10.5539/jas.v11n9p91 -
Barbieri, J. D.; Dallacort, R.; Freitas, P. S. L.; Tieppo, R. C.; Andrea, M. C. S.; Seabra Junior, S. Produtividade da soja e milho safrinha simulada para os eventos ENOS em diferentes épocas de semeadura no estado de Mato Grosso. Acta Iguazu, v.9, p.45-66, 2020. https://doi.org/10.48075/actaiguaz.v9i1.21600
» https://doi.org/10.48075/actaiguaz.v9i1.21600 -
Battisti, R.; Sentelhas, P. C.; Boote, K. J.; Câmara, G. M. de S.; Farias, J. R. B.; Basso, C. J. Assessment of soybean grain yield with altered water-related genetic improvement traits under climate change in Southern Brazil. European Journal of Agronomy, v.83, p.1-14, 2017. https://doi.org/10.1016/j.eja.2016.11.004
» https://doi.org/10.1016/j.eja.2016.11.004 -
Battisti, R.; Sentelhas, P. C. Characterizing Brazilian soybean-growing regions by water deficit patterns. Field Crops Research, v.240, p.95-105 2019. https://doi.org/10.1016/j.fcr.2019.06.007
» https://doi.org/10.1016/j.fcr.2019.06.007 -
Bhatia, V. S.; Singh, P.; Wani, S. P.; Chauhan, G. S.; Rao, A. K.; Mishra, A. K.; Srinivas, K. Analysis of potential yields and yield gaps of rainfed soybean in India using CROPGRO-Soybean model. Agricultural and Forest Meteorology, v.148, p.1252-1265, 2008. https://doi.org/10.1016/j.agrformet.2008.03.004
» https://doi.org/10.1016/j.agrformet.2008.03.004 -
Campos, J. O.; Chaves, H. M. Tendências e variabilidades nas séries históricas de colheitas mensais e anuais no Bioma Cerrado no período 1977-2010. Revista Brasileira de Meteorologia, v.35, p.157-169, 2020. http://dx.doi.org/10.1590/0102-7786351019
» http://dx.doi.org/10.1590/0102-7786351019 -
CONAB - Companhia Nacional de Abastecimento. Boletim da safra de grãos: 9º Levantamento de grãos - Safra 2024. Available on: https://www.conab.gov.br/ Accessed on: Jun. 2024.
» https://www.conab.gov.br/ -
Dallacort, R.; Martins, J. A.; Inoue, M. H.; Freitas, P. S. L.; Coletti, A. J. Distribuição das chuvas no município de Tangará da Serra, médio norte do Estado de Mato Grosso, Brasil. Acta Scientiarum. Agronomy, v.33, p.193-200, 2011. https://doi.org/10.4025/actasciagron.v33i2.5838
» https://doi.org/10.4025/actasciagron.v33i2.5838 -
Daniel, D. F.; Dallacort, R.; Barbieri, J. D.; Freitas, P. S. L de; Carvalho, M. A. C.; Modolo, A. J.; Yamashita, O. M. Caracterização de um Latossolo Vermelho distroférrico no município de Tangará da Serra, Mato Grosso. Caderno de Ciências Agrárias, v.14, p.1-10, 2022. https://doi.org/10.35699/2447-6218.2022.39613
» https://doi.org/10.35699/2447-6218.2022.39613 -
Hoogenboom, G.; Porter, C. H.; Shelia, V.; Boote, K. J.; Singh, U.; White, J. W.; Hunt, L. A.; Ogoshi, R.; Lizaso, J. I.; Koo, J.; Asseng, S.; Singels, A.; Moreno, L. P.; Jones, J. W. The DSSAT crop modeling ecosystem. In: Boote, K. J. Advances in crop modelling for a sustainable agriculture. Cambridge: Burleigh Dodds Science Publishing. 2019. Cap.8, p.173-216. https://doi.org/10.19103/AS.2019.0061.10
» https://doi.org/10.19103/AS.2019.0061.10 -
Jones, J. W.; Hoogenboom, G.; Porter, C. H.; Boote, K. J.; Batchelor, W. D.; Hunt, L. A.; Wilkens, P. W.; Singh, U.; Gijsman, A. J.; Ritchie, J. T. The DSSAT cropping system model. European Journal of Agronomy, v.18, p.235-265, 2003. https://doi.org/10.1016/S1161-0301(02)00107-7
» https://doi.org/10.1016/S1161-0301(02)00107-7 -
Marengo, J. A.; Cunha, A. P. M. do A.; Espinoza, J. C.; Fu, R.; Schöngart, J.; Jimenez, J. C.; Costa, M.; Ribeiro, J.; Wongchuig, S.; Zhao, S. The drought of Amazonia in 2023-2024. American Journal of Climate Change, v.13, p.567-597, 2024. https://doi.org/10.4236/ajcc.2024.133026
» https://doi.org/10.4236/ajcc.2024.133026 - Moreira, M. L. C.; Vasconcelos, T. N. N. Mato Grosso: Solos e Paisagens. 1.ed. Cuiabá: Entrelinhas/Secretaria de Estado de Planejamento e Coordenação Geral - SEPLAN, 272p. 2007.
-
Nóia Júnior, R. S.; Fraisse, C. W.; Karrei, M. A. Z.; Cerbaro, V. A.; Perondi, D. A. Effects of the El Niño Southern Oscillation phenomenon and sowing dates on soybean grain yield and on the occurrence of extreme weather events in southern Brazil. Agricultural and Forest Meteorology, v.290, e108038, 2020. https://doi.org/10.1016/j.agrformet.2020.108038
» https://doi.org/10.1016/j.agrformet.2020.108038 -
Ritchie, J. T. Soil water balance and plant water stress. In: Tsuji, G. Y.; Hoogenboom, G.; Thornton, P. K. Understanding options for agricultural production. Dordrecht: Springer, 1998. p.41-54. https://doi.org/10.1007/978-94-017-3624-4_3
» https://doi.org/10.1007/978-94-017-3624-4_3 -
Rocha, V. M.; Correia, F. W. S.; Gomes, W. de B. Avaliação dos impactos da mudança do clima na precipitação da Amazônia utilizando o modelo RCP 8.5 Eta-HadGEM2-ES. Revista Brasileira de Geografia Física, v.12, p.2051-2065, 2019. https://doi.org/10.26848/rbgf.v12.6.p2051-2065
» https://doi.org/10.26848/rbgf.v12.6.p2051-2065 -
Salehie, O.; Hamed, M. M.; Ismail, T. B.; Tam, T. H.; Shahid, S. Selection of CMIP6 GCM with projection of climate over the Amu Darya River Basin. Theoretical and Applied Climatology, v.151, p.1185-1203, 2023. https://doi.org/10.1007/s00704-022-04332-w
» https://doi.org/10.1007/s00704-022-04332-w -
Santos, J. A. de S.; Wanderley, H. S.; Amorim, R. F. C. de; Delgado, R. C.; Fernades, R. C. The longest multiannual drought in Northeastern Brazil. Journal of South American Earth Sciences, v.143, 104976, 2024. https://doi.org/10.1016/j.jsames.2024.104976
» https://doi.org/10.1016/j.jsames.2024.104976 -
Silva, E. H. F. M. da; Antolin, L. A. S.; Zanon, A. J.; Andrade Junior, A. S.; Souza, H. A. de; Carvalho, K. dos S.; Vieira Junior, N. A.; Marin, F. R. Impact assessment of soybean grain yield and water yield in Brazil due to climate change. European Journal of Agronomy, v.129, p.126-329, 2021. https://doi.org/10.1016/j.eja.2021.126329
» https://doi.org/10.1016/j.eja.2021.126329 -
Song, Y.; Zhang, B.; Wang, J.; Kwek, K. The impact of climate change on China’s agricultural green total factor yield. Technological Forecasting and Social Change, v.185, e122054, 2022. https://doi.org/10.1016/j.techfore.2022.122054
» https://doi.org/10.1016/j.techfore.2022.122054 -
Thrasher, B.; Wang, W.; Michaelis, A.; Melton, F.; Lee, T.; Nemani, R. NASA global daily downscaled projections, CMIP6. Scientific Data, v.9, 262, 2022. https://doi.org/10.1038/s41597-022-01393-4
» https://doi.org/10.1038/s41597-022-01393-4 -
USDA - United States Department of Agriculture Foreign Agriculture Service. Available on: https://usdabrazil.org.br//. Accessed on: Jun. 2024.
» https://usdabrazil.org.br/ -
Viana, E. N.; Dallacort, R.; Dias, V. R. de M.; Souza, M. D.; Barbieri, J. D.; Tieppo, R. C.; Fenner, W. Rainfall variability in the region of Tangará da Serra, Mato Grosso, using the standardized precipitation index. Revista Brasileira de Engenharia Agrícola e Ambiental, v.29, e291712, 2025. http://dx.doi.org/10.1590/1807-1929/agriambi.v29n9e
» http://dx.doi.org/10.1590/1807-1929/agriambi.v29n9e -
Wang, H.; Cheng, M.; Liao, Z.; Guo, J.; Zhang, F.; Fan, J.; Feng, H.; Yang, Q.; Wu, L.; Wang, X. Performance evaluation of AquaCrop and DSSAT-SUBSTOR-Potato models in simulating potato growth, yield and water productivity under various drip fertigation regimes Agricultural Water Management, v.276, 108, 2023. https://doi.org/10.1016/j.agwat.2022.108076
» https://doi.org/10.1016/j.agwat.2022.108076 -
Yen, N. T. H.; Nguyet, L. T. Climate variability induced the impact of El-Niño Southern Oscillation events on rice-growing areas in the Mekong Delta region. Revista Brasileira de Engenharia Agrícola e Ambiental, v.29, e288921, 2025. http://dx.doi.org/10.1590/1807-1929/agriambi.v29n5e288921
» http://dx.doi.org/10.1590/1807-1929/agriambi.v29n5e288921
Edited by
-
Editors:
Ítalo Herbet Lucena Cavalcante & Carlos Alberto Vieira de Azevedo
