Exploring the use of nitrogen fertilization and shifting of planting dates as adaptation strategies to climate change in the Coastal Savannah of Ghana Dilys S. MacCarthy a,* , Bright S. Freduah a, Folorunso M. Akinseye b , Samuel G.K. Adiku c , Daniel E. Dodor c , Alpha Y. Kamara d a Soil and Irrigation Research Centre, Kpong, School of Agriculture, University of Ghana, P. O. Box LG 68, Accra, Ghana b International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), c/o Centre d’étude régional pour l’amélioration de l’adaptation à la sécheresse (CERAAS), Thies Escale, Sénégal c Department of Soil Science, School of Agriculture, University of Ghana, Legon, Accra, Ghana d R4D Unit, International Institute of Tropical Agriculture, Ibadan, Nigeria A R T I C L E I N F O Keywords: West Africa Climate variability Maize DSSAT Process based model Climate scenarios A B S T R A C T Climate change poses a threat to the agriculture sector in Sub-Sahara Africa due to the reliance on weather for crop production and the low adaptive capacity of its farmers. While nitrogen fertilization and shifts in planting windows are widely promoted to increase maize production, their efficacy under future climates remains un certain. The CERES-maize model (hereafter reffered to as DSSAT) was used to assess the potential impacts of climate change on the yield of two maize varieties with different maturity duration and the potential of increased nitrogen fertilization and shifts in planting windows as climate change adaptation in the Coastal Savannah of Ghana. The impacts of these options were evaluated using climate projections by Coupled Model Intercompar ison Project—Phase 5 (CMIP5) climate models under two representative concentration pathways (RCPs) 4.5 and 8.5 in the mid-century (2040–2069) relative to the baseline climate (1980–2009). Yield reductions ranged from 14 % to 41 % for the Obatanpa variety (intermediate maturity) and between 18 % and 51 % for the Abontem (extra-early maturity) variety across the GCMs and RCPs. Reductions in yields were more severe under RCP 8.5 than RCP 4.5 scenario. Increasing nitrogen application from 30 kg N ha− 1 to 60 or 90 kg N ha− 1 improved yields and resulted in higher yield increase under the baseline climate than under future climate. This implies that the efficiency of fertilizers will decline under climate change and this would have a negative return on investments and environmental consequences. Conversely, delaying planting dates by 2–4 weeks in the main growing season significantly mitigated yield losses, resulting in yield gains of 4–23 % for Obatanpa and 8–29 % for Abontem across climate models. However, delayed planting in the minor season resulted in yield decline. Optimizing planting can enhance productivity in the major season and hence the livelihoods of farmers. Thus, optimizing planting schedules could be a viable adaptation strategy to sustain maize productivity under future climate whereas increased nitrogen fertilization may offer limited benefits. These findings are vital for policy planning and evidence-based decision making in the agriculture sector. 1. Introduction Agriculture, which is the mainstay of the economy of most countries in Sub-Saharan Africa is dominated by smallholder subsistent farmers. In West Africa in particular, low soil productivity and dependence on rainfall are the major constraints to crop production [1]. The low and declining soil fertility can be attributed to low levels of fertilizer use and poor crop and residue management practices among others. Tradition ally, long fallow systems were used to restore the fertility of soils, but this has become less common due to the increasing demand for land for many uses. In effect, lands, once cleared are cropped for many years without any conscious effort invested in soil improvement, which in turn leads to declining yields. It is conceivable that climate change would further exacerbate the current situation, with potentially negative con sequences for crop production in the region. Projected climate change scenarios for the West African sub-region indicate an increase in temperature of between 1.1 and 4.8 ◦C with greater rainfall variability by the end of this century [2]. This has * Corresponding author. E-mail address: dmaccarthy@ug.edu.gh (D.S. MacCarthy). Contents lists available at ScienceDirect Journal of Agriculture and Food Research journal homepage: www.sciencedirect.com/journal/journal-of-agriculture-and-food-research https://doi.org/10.1016/j.jafr.2025.102126 Received 18 March 2025; Received in revised form 15 June 2025; Accepted 22 June 2025 Journal of Agriculture and Food Research 22 (2025) 102126 Available online 23 June 2025 2666-1543/© 2025 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC license ( http://creativecommons.org/licenses/by- nc/4.0/ ). https://orcid.org/0000-0002-8062-3499 https://orcid.org/0000-0002-8062-3499 https://orcid.org/0000-0002-8677-6306 https://orcid.org/0000-0002-8677-6306 https://orcid.org/0000-0003-0696-2852 https://orcid.org/0000-0003-0696-2852 https://orcid.org/0000-0002-0640-2815 https://orcid.org/0000-0002-0640-2815 https://orcid.org/0000-0002-1844-2574 https://orcid.org/0000-0002-1844-2574 mailto:dmaccarthy@ug.edu.gh www.sciencedirect.com/science/journal/26661543 https://www.sciencedirect.com/journal/journal-of-agriculture-and-food-research https://doi.org/10.1016/j.jafr.2025.102126 https://doi.org/10.1016/j.jafr.2025.102126 http://crossmark.crossref.org/dialog/?doi=10.1016/j.jafr.2025.102126&domain=pdf http://creativecommons.org/licenses/by-nc/4.0/ http://creativecommons.org/licenses/by-nc/4.0/ consequences for regions such as Sub-Sahara Africa that rely predomi nantly on rainfed agriculture with low adaptive capacity. Climate change impact assessments on crop production in West Africa are becoming increasingly common. The results of the studies are however, varied in terms of the extent of yield reductions projected. An impact assessment study by Traore et al., [3] in Mali projected maize yield losses of between 51 and 57 % under current farmer fertilizer practices using RCP 4.5 and RCP 8.5, respectively by mid-century (2040–2069). Srivastava et al., [4] projected an increase of 57 % in crop yield for the semi-deciduous forest agro-ecological zone of Ghana, whereas Tachie-Obeng et al., [5] reported an average maize yield decline of 41.4 % based on nine GCMs (A2 SRES emission scenario) in the major season and an increased yield of 28.5 % in the minor season for the forest-savannah transition zone of Ghana. The emerging evidence of climate change with its attendant negative impacts on crop productivity highlights the need to explore strategies that can minimize or avert yield reductions. The Intergovernmental Panel on Climate Change (IPCC) defines climate change adaptation as “the adjustment in natural or human systems in response to actual or expected climatic stimuli or their effects which moderates harm or exploits beneficial opportunities” [6]. Several studies have reported various practices as adaptation strategies to reduce the negative impact of climate change on crop production [7–9]. For instance, Challinor et al., [10] reported on the use of shifting sowing dates and choice of cultivar maturity as adaptation strategies that reduced climate change impact. Muluneh et al., [11] also evaluated the shifting of sowing date, planting density, and supplementary irri gation as climate change adaptation strategies. Markos et al., [12] also explored the use of nitrogen, mulching, and shifts in planting dates as adaptation strategies to climate change. However, most of these studies only assessed the impacts of these strategies in increasing yields under future climates without comparing their relative impacts under the baseline climate. Thus, the impacts attributed to these strategies are often overestimated [6]. For instance, enhancing nitrogen fertilization under current or baseline climate results in higher yield increases than it does under future climate. Thus, though fertilization increased yields under future climate, the magnitude of increase is lower under climate change. To our knowledge, there are limited studies that evaluate these strategies under both baseline and projected future climate scenarios. This study, therefore, seeks to bridge this knowledge gap. In this study, we focus on maize because about 50 % of the popu lation of Sub-Saharan Africa depends on it as a source of carbohydrates, protein, iron, vitamin B, and minerals [13,14]. It is an important staple crop in the diet and contributes to improving the livelihoods of farmers as well [15]. Maize cultivation plays a vital role in ensuring food security in Ghana’s coastal savannahs, especially for smallholder farmers who mainly grow it under rainfed conditions [16]. As maize is the staple food for most Ghanaians, its production is crucial to household food security in the country [17]; accounting for more than 50 % of Ghana’s total food production and contributing 3.3 % of the value of the total agricultural production [18,19]. Despite the importance of maize cultivation, yield gaps are large, indicating the potential for improved production effi ciency and yields in the region [18]. In addition, the projected increase in temperature, and frequency of droughts due to intensified climate change pose greater challenges to enhancing crop productivity and ensuring food security in rural areas [20]. The extent to which man agement practices impact crop productivity in the future is essential to the design and implementation of adaptation strategies under the Na tionally Determined Contributions (NDCs) and the National Adaptation Plans. Few studies have evaluated the combined effects of nitrogen and planting dates in West Africa’s coastal savannahs. This study advances understanding by integrating process-based modeling with multi-GCM projections and contributes to addressing the paucity of knowledge on Fig. 1. Agro-ecological map of Ghana showing the location of the study. D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 2 climate change and adaptation impact assessment studies in the coastal savannah zones of West Africa, especially in relation to maize yields. To achieve this, we used the Decision Support Systems for Agro-Technological Transfer (DSSAT) [21] to simulate the impact of projected climate on the yield of maize and also assess the performance of management strategies with respect to reducing climate impact in the coastal savannah zone of Ghana. The DSSAT model is one of the most widely calibrated and used crop models in West Africa [22]. For instance, it has been used in Benin and Ghana to optimize fertilizer applications [23,24], to assess climate change impacts on crop produc tivity and farmers’ livelihoods in Burkina Faso, Ghana and Senegal [25, 26], and to optimize planting dates in Nigeria [27]. 2. Materials and methods 2.1. Description of study area The study was carried out at the University of Ghana, Soil and Irri gation Research Centre (SIREC), Kpong, located in the coastal savannah agro-ecological zone of Ghana (Fig. 1). The Research Centre is located at latitude 6◦ 09 N, longitude 00◦ 04’ E and at an altitude of 22 m above mean sea level. Rainfall is bimodal (Fig. 2) with a major season begin ning from March to mid-July with a mean rainfall of 682 mm (CV of 20 %) and a minor growing season from early September to mid-November with a mean annual rainfall of 342 mm and a CV of 30 % [28]. The dominant soil type found in the study area is Calcic Vertisol, according to the USDA soil taxonomy which is clayey in nature. It is a colluvial material derived from the weathering of garniteferous hornblende gneiss [29]. 2.2. Experiments for model calibration Before the use of the DSSAT model in this study, field trials were conducted at SIREC, Kpong, in 2014 to obtain datasets for model cali bration. Two maize cultivars were used for the trials, namely: Obatanpa and Abontem and the trials were established under rainfed conditions with supplementary irrigation under optimal nitrogen fertilization. Obatanpa is a white, open-pollinated, intermediate maize variety released in 1992 (International Institute for Tropical Agriculture (IITA)/ International Maize and Wheat Improvement Center (CIMMYT)/Crop Research Institute of Ghana (CRI)). It is widely cultivated in Ghana and accounts for more than 50 % of the land cultivated to cereals [30]. Abontem is a yellow, extra early maturing maize variety released (by the IITA/CRI of Ghana) in 2010 and is less widespread compared to Oba tanpa. Compost in the form of cow manure was applied two weeks prior to planting at a rate of 5000 kg ha− 1 and mineral fertilizers were applied at the rate of 120-45-45 kg ha− 1 using NPK (15-15-15) and sulphate of ammonia. Plant spacing was 75 × 40 cm on plot sizes of 36 m2. The cultivars served as the treatments and were randomized in three blocks. The treatments were replicated three times in each planting date and maize were planted on three different dates (26th June, 6th and September 26, 2014) to capture temporal variability. Data collected included crop phenology (date of emergence, date of anthesis, and date of maturity), grain and biomass yield data. Soil profile data (depths 0–15, 15–30, 30–45, 45–60, 60–75, 75–100 cm) were also collected in each of the experiments. For soil profile data, DSSAT allows the user to define thickness of soil layers, but not exceeding 20 cm [31]. Soil samples were analyzed for particle size distribution, organic carbon, and pH. Undisturbed samples were also taken to determine bulk density for each layer. A pedo-transfer function by Saxton and Rawls [32] embedded in the DSSAT shell was used to estimate the water charac teristics of the soil (field capacity, wilting point, and saturated water content). Weather data for the experimental period were obtained from a weather station close to the experimental site. 2.3. Experiments for the DSSAT model evaluation Apart from the calibration datasets, further trials were conducted in 2015 for model evaluation at Kpong (SIREC, University of Ghana). The two maize varieties were planted in a split-plot design with the fertilizer level being the main plot and the varieties being the subplots. Four different sowing dates (4th and 21st May, 13th and 28th September) were involved in the evaluation experiment. There were three replicates for each sowing date. The N levels used were 0, 45, 90 and 120 kg ha− 1. The N fertilizers were applied in splits in the ratio 2:3 at 10 and 30 days after emergence. Phosphorus and potassium were also applied at the rate of 45 kg ha− 1 P2O5 and K2O ten (10) days after emergence. Weeds were controlled manually. Data collected were the date of emergence, date of 50 % flowering, maximum leaf area index, date to 50 % physi ological maturity, and relative leaf appearance rate. Final grain yield and total biomass were taken from an area of 3 × 3 m, oven-dried to constant weight at a temperature of 70 ◦C and extrapolated to a kg ha− 1 basis. Soil samples were taken from soil profiles (0–15, 15–30, 30–45, 45–60, 60–75, 75–90, 90–105 cm), air-dried, and analyzed for pH, organic carbon, and particle size distribution (Table 1). Undisturbed samples were also taken using core samplers from each profile to determine bulk density. Weather data (daily rainfall, temperature (minimum and maximum), and solar radiation) were collected from the weather station at SIREC. Fig. 2. Baseline weather (30-year average 1980–2010) for Kpong, Ghana. Table 1 Description of soil data used for model evaluation. Soil Layer Depth (cm) Lower Limit (cm3/cm3) Drained Upper Limit (cm3/cm3) Saturated Water Content (cm3/cm3) Bulk Density (g/cm3) Organic Carbon (%) Clay (%) Silt (%) Total N (%) pH in H2O 15 0.142 0.331 0.418 1.21 0.7 47.6 5.35 0.129 5.6 30 0.164 0.331 0.373 1.22 0.66 54.7 5.6 0.120 5.7 45 0.166 0.32 0.377 1.37 0.59 52.1 4.6 0.073 6.1 60 0.174 0.301 0.396 1.39 0.53 55.4 3.6 0.095 6.1 75 0.174 0.303 0.396 1.42 0.39 55.4 3.6 0.095 6.2 90 0.181 0.305 0.435 1.44 0.23 56.9 4.6 0.095 6.5 105 0.194 0.309 0.452 1.46 0.2 56.9 4.6 0.095 6.5 D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 3 2.4. Model description The CERES-Maize module (hereafter referred to as DSSAT) [21] is a process-based mechanistic model that simulates crop growth and development, and soil processes. It utilizes information on soil profile, crop management information, crop genetic coefficients, and daily weather (maximum and minimum temperature, rainfall, and solar ra diation) to simulate soil processes and crop yield. Optimal plant growth is determined by photosynthetic capacity, radiation capture, thermal time, and photoperiod sensitivity whereas actual growth and develop ment are constrained by nutrient and water stress as well as sub-optimal temperatures [33]. Growth is constrained by nutrient and water sub-modules through stress factors. Plant nitrogen availability is via fertilizer input and mineralization of soil organic carbon. The Century model embedded in DSSAT was used to simulate soil organic matter mineralization. Further details can be found in Porter et al. [34]. The Ritchie [35] cascading water balance approach describes the movement of water between soil layers. It has a soil water sub-module that simu lates daily water balance as a function of precipitation, irrigation, transpiration, evaporation, saturated water flow, plant water uptake, drainage, and runoff. The nitrogen balance module simulates processes such as mineralization, immobilization, and N leaching. 2.5. Calibration of model Weather, soil, crop management, and experimental (phenology, biomass, and grain yield) data from the model calibration experiment (section 2.2.1) were used in version 4.6 of DSSAT for model calibration. The genetic coefficients (Table 2) that influence phenology; P1 and P5 were adjusted based on phenology data collected under optimal N and water conditions. Coefficients determining total biomass; PHINT was then used to adjust for total biomass. The coefficients G2 and G3 were used to adjust for grain yield. The calibration process was iterative until simulated parameters matched the observed data using Root Mean Squared Error (RMSE), Relative Root Mean Squared Error (RRMSE), and the modified coefficient of model efficiency (EF) as described in Eqns (1)–(3) below. The set of coefficients estimated for each of the two va rieties are presented in Table 2. The anthesis and physiological maturity were obtained with RMSEs (RRMSEs and EFs) of 0.5 days (1.0 % and 0.76) and 1.0 days (1.2 % and 0.72) for Abontem, while those of Oba tanpa were 0.5 days (1 % and 0.77) and 1.1 days (1.1 % and 0.82), respectively. The RMSE, RRMSE, and EF for maximum LAI were 0.07 m2m-2, 2.5 %, and 0.6 for Abontem, and 0.1 m2m-2, 2.0 %, and 0.60 for Obatanpa, respectively. The RMSE (RRMSE and EF) for grain and biomass yields were 49 kg ha− 1 (2 % and 0.58) and 233 kg ha− 1 (2.4 % and 0.68) for Abontem, while those of Obatanpa were 101 kg ha− 1 (2 % and 0.69), and 269 kg ha− 1 (2 % and 0.6), respectively. 2.6. Evaluation of model performance The genetic coefficients were evaluated with an independent dataset (soil, weather, crop management, phenology, grain, and biomass yield) collected from section 2.3 for model evaluation. Model performance in predicting plant parameters (maximum LAI, biomass, and grain yield) was assessed using a set of statistics; RMSE, RRMSE, and coefficient of model efficiency (EF). RMSE is defined as (Eqn (1)): RMSE= [ 1 n ∑( yieldsimulatedi − yieldobservedi )2 ]0.5 (1) Relative root mean squared error is defined as (Eqn (2)): RRMSE= ( RMSE Observed mean ) × 100 (2) The modified coefficient of model efficiency (EF) performance [36, 37] was also assessed using the equation below (Eqn (3)): EF=1 − ∑n i=1 |Observedi − Simulatedi| ∑n i=i |Observedi − Meanobserved| (3) A model efficiency of EF = 1 denotes a perfect prediction of the observed, while EF = 0 signifies poor agreement between observed and simulated data. In the modified version, squared difference terms are replaced by their respective absolute values, thus minimizing the sensitivity of the coefficient to outliers [38]. The model predicted days to anthesis (ADAT) satisfactorily with RMSE (RRMSE) of 5 days (9 %) and 4 days (9 %) for Obatanpa, and Abontem, respectively. The physiological maturity (MDAT) of the vari eties was adequately simulated with RMSE of 8 days (8 %) and 5 days (6 %) for Obatanpa, and Abontem, respectively. The maximum LAI, grain, and biomass yield were all satisfactorily simulated by the model as illustrated by the statistics of model performance (Table 3). Maximum LAI was adequately predicted with RMSE of 0.50 and 0.45 m2m-2 for Obatanpa, and Abontem, respectively. A comparison of simulated and measured maize yield for the two cultivars indicates the satisfactory performance of the model (Table 3). The model simulated grain yields with RMSE of 912 kg ha− 1 and 618 kg ha− 1 for Obatanpa and Abontem, respectively. Total biomass was equally well predicted. The RRMSE of the predictions was below 30 % across cultivars. 2.7. Assessment of the impact of climate change and adaptation/ management strategies on maize yield 2.7.1. Climate change scenarios To assess the impact of climate change on the yield of maize, 30 years of historically observed weather data (1980–2010) obtained from the Ghana Meteorological Agency were used to derive future weather pro jections using the mean change scenarios (delta) approach. It is a leading scenario-generation approach for crop modeling studies of climate im pacts [39]. The "delta approach" for generating future climate scenarios in CMIP5 was applied [39], which is a statistical downscaling method that applies changes (deltas) derived from global climate models (GCMs) to observed baseline climate data to produce localized future climate projections [40,41]. The delta method involves calculating the Table 2 Genetic coefficients developed for the Obatanpa and Abontem maize cultivars. Genetic coefficients Obatanpa Abontem Thermal time from emergence to end of juvenile (P1) (o C days) 266.9 217.2 Maximum possible grain number (G2) 591.0 585.0 Grain filling rate during the linear phase of the grain filling stage (G3) (mg grain− 1 day− 1) 7.5 8.5 Thermal time from silking to physiological maturity (P5) (oC days) 910.0 668.8 Phyllochron interval 45.00 45.25 Table 3 Performance of the model in simulating biomass, grain and maximum (Max) leaf area index (LAI). Parameter Biomass Grain Max LAI ​ Abontem RMSE (kg ha− 1) 1175 618 0.45a RRMSE (%) 14 26 19 EF 0.95 0.91 0.89 ​ Obatanpa RMSE (kg ha− 1) 1492 645 0.50a RRME (%) 15 20 18 EF 0.96 0.95 0.94 a Unit is m2m− 2, EF is coefficient of model efficiency and ranges from 0 to 1. D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 4 difference (delta or anomaly) between the future climate model output and the baseline climate for a given variable, such as temperature or precipitation. These deltas represent the projected changes in climate variables over time relative to a historical baseline (e.g., 1961–1990). The approach typically uses 30-year averages for both the baseline and future periods (e.g., near-term 2010–2039, mid-century 2040–2069, and end-century 2070–2099) to smooth out interannual variability and focus on long-term climate change signals. For temperature variables (minimum and maximum), absolute dif ferences between future and baseline means are calculated, while for precipitation, relative (percentage) changes are computed. These delta values are spatially interpolated from the coarse GCM grid points to a finer-resolution grids or station locations, allowing the anomalies to be applied to high-resolution baseline climate surfaces such as those from observational datasets like WorldClim. The interpolated deltas are then added to the observed baseline climate data, producing bias-corrected, high-resolution future climate scenarios suitable for impact assess ments in agriculture, hydrology, and ecology. The generation of future weather in this study involved five (5) contrasted General Circulation Models (GCMs); CanESM2, GFDL-ESM2M, HadGEM2-ES, MIROC5, and MPI-ESM-MR (R), considering a Representative Concentration Pathway (RCP) of 4.5 and 8.5 for the near future (2040–2069). The GCMs were selected to represent key regional climatic features. The RCP 4.5 and 8.5 represent low and high emission scenarios, respectively, with a radiation forcing of 8.5 Wm-2 by the year 2100 [42]. The RCP 4.5 and 8.5 also assume CO2 concentrations of 499 ppm and 571 ppm, compared with the current 390 ppm. 2.7.2. Assessing climate change impact under current farmers’ management practices The initial assessment of the impact of climate change focused on the current farmer practice which entails the application of 30 kg N ha− 1 on two widely used maize cultivars Obatanpa and Abontem. Maize perfor mance was assessed under the historical (baseline) and the future climate as generated by the 5 GCMs and 2 RCPs. For each year’s simu lation, planting was done automatically when the soil water content of the top 15 cm layer of the soil was above 50 % of field capacity. The planting window was set between 15th April and 15th May for the major season and 15th August – 15th September for the minor season. Fertil izer application was split, with two-thirds applied at 10 Days After Emergence (DAE) and the remaining one-third at 35 DAE. Soil data used for the simulations are presented on Table 1. The impact of climate change on yield was determined as (Eqn (4)): ΔYield= Yieldscenario − Yieldbaseline Yieldbaseline × 100 [%] (4) where ΔYield is the yield change due to climate change, YieldScenario and Yieldbaseline are yields obtained under future climate scenario and base line weather conditions, respectively. 2.7.3. Assessing increased fertilization and shifts in planting windows in mitigating climate change effect Two management strategies (i) increased N fertilizer application and (ii) shifting planting dates were assessed for their potential to offset yield losses under future climate. Under each management strategy, the same cultural practices were followed as previously described in section 2.7.2. The two maize varieties (Obatanpa and Abontem) were sown within the Fig. 3. Change in annual rainfall amount against temperature anomaly under (A) RCP 4.5 and (B) RCP 8.5, relative changes in rainfall events under (C) RCP 4.5 and (D) RCP 8.5 under different climate scenarios relative to the baseline period. D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 5 three planting windows for each season under three fertilizer applica tion rates: (30 kg N ha− 1 for current practice, 60 kg N ha− 1 for enhanced practice and, 90 kg N ha− 1 for intensive practice). The effectiveness of enhanced fertilization and shifting planting dates in reducing the negative (or enhancing positive) impact of projected climate change as defined by Lobell [6] was assessed as (Eqn (5)): CC AS= ( YieldAS CC − YieldBS CC YieldBS CC ×100 ) − ( YieldAS BC − YieldBS BC YieldBS BC ×100 ) (5) where CC_AS is the relative yield gain due to adapting the proposed strategy (increased fertilization and delayed planting) under climate change scenario compared to baseline climate. YieldAS_CC and YieldBS_CC are grain yields obtained under future climate scenarios using the alternate strategy and the base strategy, respectively. YieldAS_BC and YieldBS_BC are grain yields obtained under baseline (historical) climate using the alternate strategy and the base strategy, respectively. Analysis of variance (ANOVA) was performed using the 12th edition of GENSTAT to determine the effects of seasons, climate scenarios, RCPs and man agement practices on grain yield under projected climate conditions. 3. Results 3.1. Projected changes in climate variables The 5 GCMs under RCP 4.5 projected increases in temperatures of between 1.2 - 2.3 and 1.2–2.2 ◦C over the baseline average temperature of 28 ◦C for the major and minor seasons respectively. Projected changes under RCP 8.5 were much higher, ranging between 1.8 – 3.2 and 1.8–2.7 ◦C for the major and minor seasons respectively (Fig. 3). Unlike temperature, the direction of change in rainfall amounts varied among the GCMs, irrespective of the season and RCP. For those GCMs that projected increases in rainfall amounts, the magnitudes were generally higher in the minor season. Similarly, for the GCMs that projected a decline in rainfall amounts, the decline was generally higher in the major season (Fig. 3). The changes in the number of rainfall events were generally low across the two RCPs. The magnitude of decline was generally higher in the major season and under the higher emission scenario (Fig. 3). 3.2. Simulated baseline yields Simulated average baseline yields for the two maize varieties in the major season were 1950 kg ha− 1 and 1550 kg ha− 1 for Obatanpa and Abontem, respectively. Inter-annual variability in yield was slightly higher for Abontem (19 %) compared to Obatanpa (16 %). In the minor season, the average yield of Obatanpa was about 15 % lower than the major season yield. Similarly, yields produced by Abontem were marginally lower (2 %) in the minor season. The inter-annual variability in the yields was higher for Abontem (23 %) compared to those obtained for Obatanpa (16 %). 3.3. Projected future yields under climate change Simulated grain yields of the two maize varieties under the different climate scenarios were lower (p < 0.05) than their respective baseline yields (Table 4). The grain yields of Obatanpa reduced by between 33 Table 4 Simulated yield statistics of Obatanpa and Abontem maize varieties under different climate scenarios in the major and minor growing seasons. Max, Min, and CV represent maximum, minimum and coefficient of variation, respectively. Mean Max Min CV (%) Mean Max Min CV (%) (kg ha− 1) (kg ha− 1) (kg ha− 1) (kg ha− 1) (kg ha− 1) (kg ha− 1) Climate scenario Major season Minor season Obatanpa Baseline 1949 2796 1332 16 1625 2136 627 22 RCP 4.5 CanESM2 1501 2603 691 25 1072 1631 430 23 GFDL-ESM2M 1666 2726 904 21 1514 2144 617 19 HadGen2-ES 1320 2240 534 26 1060 1679 483 23 MIROC5 1611 2677 757 23 1308 1973 530 20 MPI-ESM-MR 1527 2555 721 23 1245 1868 510 21 RCP 8.5 CanESM2 1219 2168 489 29 759 1141 334 22 GFDL-ESM2M 1448 2396 671 23 1263 1947 558 20 HadGen2-ES 1168 2039 518 28 881 1341 420 22 MIROC5 1462 2368 702 23 1191 1805 534 22 MPI-ESM-MR 1344 2324 667 25 1093 1595 481 20 Abontem Baseline 1552 2171 889 19 1409 1942 527 21 RCP 4.5 CanESM2 1000 1537 606 23 909 1434 469 21 GFDL-ESM2M 1209 1831 730 22 1145 1707 515 19 HadGen2-ES 865 1252 449 24 789 1294 415 22 MIROC5 1158 1767 676 22 1050 1436 513 18 MPI-ESM-MR 1044 1569 595 23 959 1382 487 19 RCP 8.5 CanESM2 812 1239 432 25 634 1050 303 24 GFDL-ESM2M 1037 1666 588 24 925 1392 486 20 HadGen2-ES 759 1163 393 27 643 1025 306 23 MIROC5 1021 1496 617 22 920 1299 481 20 MPI-ESM-MR 884 1378 530 24 875 1242 450 19 D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 6 and 15 % across the five GCMs under the low-emission scenario (RCP 4.5) in the major season, whereas the yield reductions in the Abontem variety were more severe, ranging between 45 and 22 % (Fig. 4). The inter-annual variability in yield change ranged between 29 and 45 % for the Obatanpa variety while those simulated for the Abontem variety were much lower ranging between 14 and 33 %. Generally, the hottest and driest GCMs projected higher yield decline for both varieties with the shorter maturity variety experiencing greater yield loss than the inter mediate variety. Under the high-emission scenario, the projected yield decline was higher (p < 0.05) than that projected under the low- emission scenario (ranging between 26 and 41 % for the Obatanpa va riety and between 33 and 51 % for the Abontem variety across GCMs). The inter-annual variability in yield change was generally lower under the high emission climate scenario for both Obatanpa (24–30 %) and Abontem (15–21 %). The trend in yield losses between the two varieties and emission scenarios was similar in the minor season. However, the magnitude of yield decline was consistently lower in the minor season compared to the major season (Fig. 4). 3.4. Impact of fertilization under low and high emissions climate scenarios Applying 60 kg N ha− 1 fertilizer under climate change scenarios in the major season resulted in average yield increases (p < 0.05) of be tween 61 and 64 % across GCMs under the low-emission scenario for the Obatanpa variety, and between 57 and 58 % for the Abontem variety. Increasing fertilizer application to 90 kg N ha− 1 resulted in yield in creases ranging between 94 and 101 % (Obatanpa) and 87 and 92 % (Abontem). In the minor rainy season, the response of Obatanpa to N fertilization was generally higher (65–70 % and 104–115 % with 60 and 90 kg N ha− 1 application, respectively) compared to the response in the major rainy season. In contrast, the Abontem recorded a lower response to fertilization (53–55 % and 76–82 % across GCMs for the 60 and 90 kg N ha− 1 application rates, respectively) compared to those obtained in the major season (Fig. 5). Generally, the response to N fertilization was lower under the higher emission scenario compared to the lower emission scenario across GCMs, varieties, and seasons. In the major season, Obatanpa yielded between 58 and 63 %, which was higher than the response obtained by Abontem (55–59 %) with 60 kg N ha− 1 fertilization. With the application of 90 kg N ha− 1, yield increases were 86–99 % across GCMs for Obatanpa and 83–96 % for Abontem variety. As under RCP 4.5, the response of Obatanpa to N fertilization was higher (by 62–67 % and 98–110 % across GCMs for 60 and 90 kg N ha-1 fertilizer applications, respectively) in the minor season than in the major season. The yield response of Abontem was lower (52–55 % and 75–82 % for the 60 and 90 kg N ha− 1 appli cation rates, respectively) in the minor season compared to the major season. 3.5. Impact of shifting planting dates on yields under low and high emissions climate scenarios Changing planting dates generally influenced (p < 0.05) the per centage change in yield across GCMs, seasons, emission scenarios, and varieties (Fig. 6). The effect of shifting planting dates was more effective in the major seasons than in the minor seasons which are usually short. The benefits of shifting the planting window also diminished with delayed planting, especially in the minor seasons. The Abontem variety benefited more from the shift in planting date than the Obatanpa variety across planting dates. For instance, yield gains of between 17 and 30 % were obtained across GCMs for the Abontem compared to a 5–19 % yield gain for the Obatanpa when planting was delayed by two weeks. Shifting planting dates by two weeks in the minor season resulted in either smaller yield gains or yield declines. As expected, yield gains, particu larly in the minor seasons were greater for the Abontem compared to the Obatanpa variety. Shifting planting dates was generally, marginally more beneficial under the higher emission scenario than the lower emission scenario. Fig. 4. Impact of climate change on grain yield of Obatanpa under (A) RCP 4.5 and (B) 8.5 and Abontem under (C) RCP4.5 and (D) 8.5. D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 7 3.6. Using N fertilization as a climate change adaptation strategy Even though using N fertilizer results in yield gains across GCMs, seasons, and varieties (Fig. 7), the difference in the response of the two maize varieties in the major season to N fertilizer under projected future and historical periods is marginally positive in two out of five GCMs and negative in the other three GCMs when 60 kg N ha− 1 was applied under low emission scenarios. At 90 kg N ha− 1 fertilizer application rate, the difference in the response of the two varieties declined further, sug gesting that the application of N fertilizer resulted in higher grain yieldunder the historical climate than under the projected future climate scenarios. In the minor season, except under a few GCMs, the use of additional fertilizer resulted in a lower yield response in the Obatanpa under the future climate than under the baseline climate irrespective of the amount of fertilizer used (p < 0.05). The magnitude of the decline in yield in response to increased N fertilizer was consistently higher under the higher N fertilizer application rate. The inter-annual variability in the response to fertilizer was generally higher for the Abontem variety compared to that of the Obatanpa variety. 3.7. Assessing shifting planting date as a climate change adaptation Fig. 8 illustrates the impact of shifting the planting window in mitigating the negative impact of projected climate change on the yields of two maize varieties across climate emission scenarios, GCMs, and seasons. Generally, shifting the planting window by two weeks yielded a positive impact on grain yield for both maize varieties in the major seasons irrespective of the climate change emission scenario used (p < 0.05). The reverse was the case for the minor seasons as shifting planting dates by two weeks resulted in yield decline, suggesting that this would not be an effective adaptation to mitigate the climate change effect in the minor season. For instance, whereas shifting planting dates by two weeks in the major season yielded between 4 and 19 % gains in yield of Obatanpa, doing so in the minor season resulted in between one per centage gain in yield by one GCM to up to 3 % reduction in yield across the remaining GCMs. A further delay in planting dates by two weeks Fig. 5. Impact of fertilizer application on yields of two maize cultivars; Obatanpa and Abontem cultivated in two seasons under different future climate scenarios. Nitrogen fertilizer increased from 30 kg N ha− 1 to 60 kg N ha− 1under (A) RCP 4.5 and (B) RCP 8.5, and to 90 kg N ha− 1 under (C) RCP 4.5 and (D) 8.5 for the Obatanpa cultivar. E, F, represent increased fertilization to 60 kg N ha− 1 under RCP 4.5 and RCP 8.5, while G & H represent fertilizer rate of 90 kg N ha− 1 for the Abontem cultivar. D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 8 generally further increased yield gains in the major seasons across the GCMs, varieties, and emission scenarios in the major season. However, delaying planting by a further two weeks in the minor season generally resulted in further yield decline irrespective of variety and emission scenario. 4. Discussions 4.1. Impact of climate change on maize yield This study used a process-based model, DSSAT, in exploring adap tation strategies to climate change. The model has been satisfactorily used in earlier studies [43–45] in the West Africa sub-region and suc cessfully evaluated in this study, demonstrating its robustness. The ef fects of climate change are complex and are both location and crop-specific. Several studies have reported the negative effects of climate change on maize yield in Sub-Saharan Africa and these often vary in magnitude, depending on factors such as location, climate sce narios used, differences in the time horizon for which the studies were carried out, and even differences in crop models used [46]. Parry et al. [47] reported varied climate change impacts from − 98 to +16 % with the majority of the cases reporting yield losses for maize in SSA [48]. A recent review by MacCarthy et al. [49] on climate change impact studies on West African agriculture reported maize yield changes ranging from − 72 to +48 % with a median of − 17 %. A much smaller range of yield reduction of between 25 and 16 % was estimated for East Africa [50]. Thus, simulated reductions in maize yield in this study are within the range reported for locations within West Africa. The decline in yields in this study are as a result of increased temperature stress that translated into shortening of crop life cycle, with a consequent reduction in re sources (nutrients and water) uptake for optimum grain yield. The yield decline was more severe for the shorter maturity variety, a phenomenon that has earlier been reported by Kassie et al. [8], in their study on climate-induced yield variability and yield gaps of maize (Zea mays L.) in the Central Rift Valley of Ethiopia. Fig. 6. Impact of shifting planting dates on the yield of two maize cultivars; Obatanpa and Abontem cultivated in two seasons under different future climate scenarios. Delay planting by 2 weeks under (A) RCP 4.5 and (B) RCP 8.5, and by 4 weeks under (C) RCP 4.5 and (D) 8.5 for the Obatanpa cultivar. E, F, represent 2 weeks delay in planting under RCP 4.5 and RCP 8.5, while G & H represent 4 weeks delay in planting for the Abontem cultivar. D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 9 4.2. Mitigating climate change impact on maize with management practices Projected yield declines reemphasize the need for adequate plans to mitigate the projected negative climate change impact while maxi mizing any benefits. This supports the study’s aim to assess the appro priateness of management practices to achieve this. Increased use of fertilizer is a strategy being promoted to increase crop production in SSA as part of the push for increased productivity [51]. Given that most of the soils in SSA are poor in fertility, the promotion of soil fertility enhancement strategies such as the use of inorganic fertilizer cannot be overemphasized given the large yield gaps of maize that need to be bridged [1]. The large yield gap reinforces the need for intensification, and the use of inorganic fertilizer is a major component given their currently sub-optimal application, and the need to complement organic materials to provide the needed nutrients for optimal crop growth. Our study shows that even though increasing fertilizers will result in increased yield under future climate, the magnitude of yield gains was higher under the baseline climate. The decline in yield is due to the acceleration of crop growth in response to increased temperatures under climate change, relative to the baseline thereby reducing phenological phases and, hence, the uptake of resources such as nutrients and water needed to optimize growth and yield. Additionally, several studies have identified increased use of inorganic fertilizers as one of the manage ment practices that increase maize grain yields in SSA under current and historical weather [44,52]. Thus, increasing the use of fertilizer under climate change is not an effective climate change adaptation strategy as it does not offset the negative impact of climate change [6]. Hence, the proposal to use increased fertilizer application as a strategy for climate change adaptation may not be effective as the efficiency of fertilizer application is reduced under climate change. To the best of our knowl edge, this study is among the very few that compare the performance of a strategy under both current and projected future climates in assessing their potential to be used as a climate change adaptation strategy. Not comparing the potential of the strategy under historical weather often results in overestimating the magnitude of the impact of the strategy, Fig. 7. Effect of fertilizer application as a climate change adaptation strategy in cultivating Obatanpa and Abontem maize cultivars in two seasons under different future climate scenarios relative to baseline climate. Nitrogen fertilizer increased from 30 kg N ha− 1 to 60 kg N ha− 1under (A) RCP 4.5 & (B) RCP 8.5, and to 90 kg N ha− 1 under (C) RCP 4.5 & (D) 8.5 for the Obatanpa cultivar. E, F, represent increased fertilization to 60 kg N ha− 1 under RCP 4.5 & RCP 8.5, while G & H represent fertilizer rate of 90 kg N ha− 1 for the Abontem cultivar. D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 10 and hence, underestimating the cost of climate change [6]. Appropriate estimation of climate change adaptation is very important as this will feed into planning and decision-making of policymakers. Over estimating the gains of management practices under future climate may affect the nature and level of resources that will be allocated for climate change adaptation. Delaying planting by 2 or 4 weeks, relative to the base planting window, shifts the reproductive stage further into lower temperature regimes, hence, higher yields are obtained in the major season with delayed planting. Thus, planting time or date could be a strategy to minimize the effect of climate change in the major season. The impor tance of planting date, as a strategy to enhance crop productivity under current climatic conditions has earlier been reported [1,53]. Unlike ni trogen fertilization, delaying planting in the major seasons yielded greater benefits under the future climate than under the baseline climate and therefore, can be promoted as an adaptation to climate change. Shifting of planting windows in the minor season did not favour maize yields as the growing season is shorter in the minor season. Additionally, the reproductive period was shifted to higher temperature regimes relative to those for the base planting window. Tachie-Obeng et al. [5] also suggested an adjustment in the planting calendar as a potential adaptation strategy in the transition zone of Ghana. Our study, however, suggests that the general utility of planting date will be limited to the major season and may vary among GCMs due to the differences in their characteristics. This study contributes to the limited knowledge on climate change impact and adaptation strategies in West Africa needed to inform evidence-based policy and decision-making in national adaptation planning processes to improve farmers’ adaptive capacities. Addition ally, Ghana, as with many other countries, has committed to achieving the Sustainable Development Goals (SDGs) that are being challenged by climate change. This study provides insights that are useful in the un derstanding and development of appropriate climate change adaptation strategies that will contribute to ensuring the SDGs aimed at ensuring No poverty (SDG 1), Zero hunger (SDG 2), and Climate action (SDG 13) are attained. The findings of this study are also relevant to Ghana’s quest to Fig. 8. Effect of shifting planting dates as a climate change adaptation strategy in cultivating two maize cultivars; Obatanpa and Abontem in two seasons under different future climate scenarios relative to baseline climate. Delay planting by 2 weeks under (A) RCP 4.5 and (B) RCP 8.5, and by 4 weeks under (C) RCP 4.5 and (D) 8.5 for the Obatanpa cultivar. E, F, represent 2 weeks delay in planting under RCP 4.5 and RCP 8.5, while G & H represent 4 weeks delay in planting for the Abontem cultivar. D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 11 enhance resilience across food systems to climate change and reduce vulnerability in the agricultural sector as elaborated in its agricultural policies and its National Medium-Term Development Policy Framework (2022–2025). Soil degradation processes are known to negatively impact crop yields and the effects are usually visible in the long-term. Thus, short-term experiments cannot capture these processes, and data required to parameterize simulation models for long-term simulations are often lacking [54]. The study, therefore did not consider the effect of soil degradation processes such as erosion on yield of maize under projected climate scenarios. Thus, the climate change impacts and the efficacy of the management practices as climate change adaptation strategies should be interpreted with this in mind. 5. Conclusions The study examined the role of two management practices (nitrogen fertilization and shifting planting dates) in mitigating the negative impact of climate change on maize in the Coastal Savannah zone of Ghana. Our study suggests that even though increasing fertilizer use may lead to higher yields under future climate conditions, the gains are higher under current climate conditions. This suggests that relying on increased fertilizer application as an adaptation strategy to climate change may not be effective, as it does not fully offset the impact due to a decline in fertilizer efficacy under projected future climate. Unlike nitrogen fertilizer application, delayed planting can be an effective adaptation strategy for increasing maize productivity under changing climatic conditions. Our findings also show that the utility of delayed planting may depend on the length of the growing season and the climate scenario, due to the differences in their characteristics. The study provides insights into the effectiveness of fertilization and delayed planting as climate change adaptation strategies and hence would pro vide guidance to policymakers in terms of their investment decision- making aimed at providing interventions to improve the adaptive ca pacity of farming communities in low input systems under projected climate change. CRediT authorship contribution statement Dilys S. MacCarthy: Writing – review & editing, Writing – original draft, Visualization, Supervision, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptu alization. Bright S. Freduah: Writing – review & editing, Writing – original draft, Visualization, Formal analysis, Data curation. Folorunso M. Akinseye: Writing – review & editing, Methodology, Formal anal ysis, Data curation. Samuel G.K. Adiku: Writing – review & editing, Supervision, Methodology, Investigation, Data curation. Daniel E. Dodor: Writing – review & editing, Validation, Methodology, Data curation. Alpha Y. Kamara: Writing – review & editing, Resources, Project administration, Investigation, Data curation. Additional information No additional information is available for this paper. Funding statement The data used for this study was obtained from the CRP MAIZE, CIMMYT/CGIAR through the International Institute of Tropical Agri culture, Ibadan, Nigeria (A4032.09.34), and SARD-SC funded projects (2015–2017). The authors are grateful for the financial support. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments Authors are grateful to staff of the Soil and Irrigation Research Centre of the University of Ghana who help with experimental management and data collection. Data availability Data will be made available on request. References [1] D.S. MacCarthy, S.G. Adiku, B.S. Freduah, F. Gbefo, A.Y. Kamara, Using CERES- Maize and ENSO as decision support tools to evaluate climate-sensitive farm management practices for maize production in the northern regions of Ghana, Front. Plant Sci. 8 (2017) 31. [2] IPCC, Climate Change, The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the, Intergovernmental Panel on Climate Change, 2013, 2013. [3] B. Traore, K. Descheemaeker, M.T. van Wijk, M. Corbeels, I. Supit, K.E. Giller, Modelling cereal crops to assess future climate risk for family food self-sufficiency in southern Mali, Field Crops Res. 201 (2017) 133–145, https://doi.org/10.1016/j. fcr.2016.11.002. [4] A.K. Srivastava, C.M. Mboh, T. Gaiser, F. Ewert, Impact of climatic variables on the spatial and temporal variability of crop yield and biomass gap in Sub-Saharan Africa-a case study in Central Ghana, Field Crops Res. 203 (2017) 33–46. [5] E. Tachie-Obeng, P. Akponikpè, S. Adiku, Considering effective adaptation options to impacts of climate change for maize production in Ghana, Environ. Dev. 5 (2013) 131–145. [6] D.B. Lobell, Climate change adaptation in crop production: beware of illusions, Global Food Secur. 3 (2014) 72–76. [7] S.B. Bedeke, Climate change vulnerability and adaptation of crop producers in sub- Saharan Africa: a review on concepts, approaches and methods, Environ. Dev. Sustain. 25 (2023) 1017–1051. [8] G.W. Kassie, S. Kim, F.P. Fellizar Jr., Determinant factors of livelihood diversification: evidence from Ethiopia, Cogent Social Sciences 3 (2017) 1369490. [9] H. Webber, T. Gaiser, F. Ewert, What role can crop models play in supporting climate change adaptation decisions to enhance food security in Sub-Saharan Africa? Agric. Syst. 127 (2014) 161–177. [10] A.J. Challinor, J. Watson, D.B. Lobell, S.M. Howden, D. Smith, N. Chhetri, A meta- analysis of crop yield under climate change and adaptation, Nat. Clim. Change 4 (2014) 287–291. [11] A. Muluneh, L. Stroosnijder, S. Keesstra, B. Biazin, Adapting to climate change for food security in the Rift Valley dry lands of Ethiopia: supplemental irrigation, plant density and sowing date, J. Agric. Sci. 155 (2017) 703–724. [12] D. Markos, W. Worku, G. Mamo, Exploring adaptation responses of maize to climate change scenarios in southern central Rift Valley of Ethiopia, Sci. Rep. 13 (2023) 12949. [13] K. Tesfaye, S. Gbegbelegbe, J.E. Cairns, B. Shiferaw, B.M. Prasanna, K. Sonder, K. Boote, D. Makumbi, R. Robertson, Maize systems under climate change in sub- Saharan Africa: potential impacts on production and food security, International Journal of Climate Change Strategies and Management 7 (2015) 247–271. [14] M. Zeller, M. Sharma, C. Henry, C. Lapenu, An operational method for assessing the poverty outreach performance of development policies and projects: results of case studies in Africa, Asia, and Latin America, World Dev. 34 (2006) 446–464. [15] T. Stevens, K. Madani, Future climate impacts on maize farming and food security in Malawi, Sci. Rep. 6 (2016) 36241. [16] C.A. Wongnaa, D. Awunyo-Vitor, Achieving sustainable development goals on no poverty and zero hunger: does technical efficiency of Ghana’s maize farmers matter? Agric. Food Secur. 7 (2018) 71, https://doi.org/10.1186/s40066-018- 0223-z. [17] L. Scheiterle, R. Birner, Assessment of Ghana’s comparative advantage in maize production and the role of fertilizers, Sustainability 10 (2018) 4181, https://doi. org/10.3390/su10114181. [18] B.O. Asante, O. Temoso, K.N. Addai, R.A. Villano, Evaluating productivity gaps in maize production across different agroecological zones in Ghana, Agric. Syst. 176 (2019) 102650. [19] Ministry of Food and Agriculture, Agriculture in Ghana: Facts and Figures 2015, 2016. [20] J. Ellis-Jones, A. Larbi, I. Hoeschle-Zeledon, I. Dugje, I. Teli, S.S. Buah, R. Kanton, M. Kombiok, A. Kamara, I. Gyamfi, Sustainable Intensification of Cereal-Based Farming Systems in Ghana’s Guinea Savanna: Constraints and Opportunities Identified with Local Communities, 2012. [21] J.W. Jones, G. Hoogenboom, C.H. Porter, K.J. Boote, W.D. Batchelor, L. Hunt, P. W. Wilkens, U. Singh, A.J. Gijsman, J.T. Ritchie, The DSSAT cropping system model, Eur. J. Agron. 18 (2003) 235–265. [22] A. Kamara, M. Garba, A. Tofa, A. Mohamed, A. Souley, T. Abdoulaye, B. Kapran, Assessment of the impact of crop management strategies on the yield of early- maturing maize varieties in the drylands of Niger Republic: application of the DSSAT-CERES-Maize model, Heliyon 9 (2023) e17829. D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 12 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref1 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref1 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref1 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref1 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref2 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref2 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref2 https://doi.org/10.1016/j.fcr.2016.11.002 https://doi.org/10.1016/j.fcr.2016.11.002 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref4 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref4 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref4 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref5 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref5 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref5 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref6 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref6 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref7 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref7 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref7 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref8 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref8 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref9 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref9 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref9 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref10 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref10 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref10 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref11 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref11 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref11 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref12 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref12 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref12 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref13 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref13 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref13 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref13 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref14 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref14 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref14 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref15 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref15 https://doi.org/10.1186/s40066-018-0223-z https://doi.org/10.1186/s40066-018-0223-z https://doi.org/10.3390/su10114181 https://doi.org/10.3390/su10114181 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref18 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref18 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref18 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref19 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref19 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref20 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref20 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref20 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref20 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref21 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref21 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref21 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref22 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref22 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref22 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref22 [23] P.G. Tovihoudji, P.I. Akponikpè, E.K. Agbossou, C.L. Bielders, Using the DSSAT model to support decision making regarding fertilizer microdosing for maize production in the sub-humid region of Benin, Front. Environ. Sci. 7 (2019) 13. [24] D.S. MacCarthy, S.G.K. Adiku, A.Y. Kamara, B.S. Freduah, J.X. Kugbe, The role of crop simulation modeling in managing fertilizer use in maize production systems in northern Ghana, in: D.J. Timlin, S.S. Anapalli (Eds.), Enhancing Agricultural Research and Precision Management for Subsistence Farming by Integrating System Models with Experiments, first ed., Wiley, 2022, pp. 48–68, https://doi. org/10.1002/9780891183891.ch4. [25] P. Arumugam, A. Chemura, P. Aschenbrenner, B. Schauberger, C. Gornott, Climate change impacts and adaptation strategies: an assessment on sorghum for Burkina Faso, Eur. J. Agron. 142 (2023) 126655. [26] D.S. MacCarthy, M. Adam, B.S. Freduah, B.Y. Fosu-Mensah, P.A. Ampim, M. Ly, P. S. Traore, S.G. Adiku, Climate change impact and variability on cereal productivity among smallholder farmers under future production systems in West Africa, Sustainability 13 (2021) 5191. [27] A.A. Adnan, J.M. Jibrin, A.Y. Kamara, B.L. Abdulrahman, A.S. Shaibu, I.I. Garba, CERES–Maize model for determining the optimum planting dates of early maturing maize varieties in Northern Nigeria, Front. Plant Sci. 8 (2017) 1118. [28] D.S. MacCarthy, P. Akponikpe, S. Narh, R. Tegbe, Modeling the effect of seasonal climate variability on the efficiency of mineral fertilization on maize in the coastal savannah of Ghana, Nutrient Cycl. Agroecosyst. 102 (2015) 45–64. [29] H. Brammer, Detailed soil survey of the Kpong pilot irrigation area. https://www. cabidigitallibrary.org/doi/full/10.5555/19570700010, 1955. (Accessed 11 March 2024). [30] G.B. Adu, B. Badu-Apraku, R. Akromah, I.K. Amegbor, D.S. Adogoba, A. Haruna, K. A. Manigben, P.A. Aboyadana, A.N. Wiredu, Trait profile of maize varieties preferred by farmers and value chain actors in northern Ghana, Agron. Sustain. Dev. 41 (2021) 1–15. [31] J. Trenz, E. Memic, W.D. Batchelor, S. Graeff-Hönninger, Generic optimization approach of soil hydraulic parameters for site-specific model applications, Precis. Agric. 25 (2024) 654–680. [32] K.E. Saxton, W.J. Rawls, Soil water characteristic estimates by texture and organic matter for hydrologic solutions, Soil Sci. Soc. Am. J. 70 (2006) 1569–1578. [33] C.M.T. Soler, P.C. Sentelhas, G. Hoogenboom, Application of the CSM-CERES- Maize model for planting date evaluation and yield forecasting for maize grown off-season in a subtropical environment, Eur. J. Agron. 27 (2007) 165–177. [34] C.H. Porter, J.W. Jones, S. Adiku, A.J. Gijsman, O. Gargiulo, J. Naab, Modeling organic carbon and carbon-mediated soil processes in DSSAT v4. 5, Operational Research 10 (2010) 247–278. [35] J.T. Ritchie, Soil water balance and plant water stress. Understanding Options for Agricultural Production, 1998, pp. 41–54. [36] J.E. Nash, J.V. Sutcliffe, River flow forecasting through conceptual models part I—a discussion of principles, J. Hydrol. 10 (1970) 282–290. [37] K. Loague, R.E. Green, Statistical and graphical methods for evaluating solute transport models: overview and application, J. Contam. Hydrol. 7 (1991) 51–73. [38] D.R. Legates, G.J. McCabe Jr., Evaluating the use of “goodness-of-fit” measures in hydrologic and hydroclimatic model validation, Water Resour. Res. 35 (1999) 233–241. [39] A.C. Ruane, R. Goldberg, J. Chryssanthacopoulos, Climate forcing datasets for agricultural modeling: merged products for gap-filling and historical climate series estimation, Agric. For. Meteorol. 200 (2015) 233–248. [40] C. Navarro-Racines, J. Tarapues, P. Thornton, A. Jarvis, J. Ramirez-Villegas, High- resolution and bias-corrected CMIP5 projections for climate change impact assessments, Sci. Data 7 (2020) 7. [41] G.G. Wodaje, Z.E. Asfaw, M.A. Denboba, Statistical downscaling (Delta method) of precipitation and temperature for Bilate watershed, Ethiopia, Int. J. Water Resour. Environ. Eng. 13 (2021) 20–29. [42] D.P. Van Vuuren, J. Edmonds, M. Kainuma, K. Riahi, A. Thomson, K. Hibbard, G. C. Hurtt, T. Kram, V. Krey, J.-F. Lamarque, The representative concentration pathways: an overview, Clim. Change 109 (2011) 5–31. [43] F.M. Akinseye, M. Adam, S.O. Agele, M.P. Hoffmann, P.C.S. Traore, A. M. Whitbread, Assessing crop model improvements through comparison of sorghum (sorghum bicolor L. moench) simulation models: a case study of West African varieties, Field Crops Res. 201 (2017) 19–31, https://doi.org/10.1016/j. fcr.2016.10.015. [44] D.S. MacCarthy, P.L.G. Vlek, B.Y. Fosu-Mensah, The response of maize to N fertilization in a sub-humid region of Ghana: understanding the processes using a crop simulation model, in: J. Kihara, G. Hoogenboom, A. Bationo (Eds.), Improving Soil Fertility Recommendations in Africa Using the Decision Support System for Agrotechnology Transfer (DSSAT), Springer + Business Media B. V, 2012, pp. 61–75. [45] J. Ouedraogo, S. Youl, A. Mando, Combining the DSSAT model with experimentation to update recommendations of fertilizer rates for rice and maize in Burkina Faso, in: A. Bationo, D. Ngaradoum, S. Youl, F. Lompo, J.O. Fening (Eds.), Improving the Profitability, Sustainability and Efficiency of Nutrients through Site Specific Fertilizer Recommendations in West Africa Agro-Ecosystems, Springer International Publishing, Cham, 2018, pp. 1–22, https://doi.org/10.1007/978-3- 319-58792-9_1. [46] R.P. Rötter, T.R. Carter, J.E. Olesen, J.R. Porter, Crop–climate models need an overhaul, Nat. Clim. Change 1 (2011) 175–177. [47] M.L. Parry, C. Rosenzweig, A. Iglesias, M. Livermore, G. Fischer, Effects of climate change on global food production under SRES emissions and socio-economic scenarios, Glob. Environ. Change 14 (2004) 53–67. [48] A.J. Challinor, T.R. Wheeler, P.Q. Craufurd, C.A. Ferro, D.B. Stephenson, Adaptation of crops to climate change through genotypic responses to mean and extreme temperatures, Agric. Ecosyst. Environ. 119 (2007) 190–204. [49] D.S. MacCarthy, F.M. Akinseye, M. Ly, E.C. Timpong-Jones, I. Hathie, S.G.K. Adiku, Modelling the impact of climate change on agriculture in West Africa. https:// www.cabidigitallibrary.org/doi/pdf/10.5555/20230340603, 2023. (Accessed 11 March 2024). [50] P.K. Thornton, P.G. Jones, P.J. Ericksen, A.J. Challinor, Agriculture and food systems in sub-Saharan Africa in a 4 C+ world, Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 369 (2011) 117–136. [51] B. Vanlauwe, D. Coyne, J. Gockowski, S. Hauser, J. Huising, C. Masso, G. Nziguheba, M. Schut, P. Van Asten, Sustainable intensification and the African smallholder farmer, Curr. Opin. Environ. Sustain. 8 (2014) 15–22. [52] F. Aramburu-Merlos, F.A.M. Tenorio, N. Mashingaidze, A. Sananka, S. Aston, J. J. Ojeda, P. Grassini, Adopting yield-improving practices to meet maize demand in Sub-Saharan Africa without cropland expansion, Nat. Commun. 15 (2024) 4492, https://doi.org/10.1038/s41467-024-48859-0. [53] A.Y. Kamara, F. Ekeleme, D. Chikoye, L.O. Omoigui, Planting date and cultivar effects on grain yield in dryland corn production, Agron. J. 101 (2009) 91–98. [54] S.G.K. Adiku, D.S. MacCarthy, S.K. Kumahor, A conceptual modelling framework for simulating the impact of soil degradation on maize yield in data-sparse regions of the tropics, Ecol. Model. 448 (2021) 109525, https://doi.org/10.1016/j. ecolmodel.2021.109525. D.S. MacCarthy et al. Journal of Agriculture and Food Research 22 (2025) 102126 13 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref23 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref23 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref23 https://doi.org/10.1002/9780891183891.ch4 https://doi.org/10.1002/9780891183891.ch4 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref25 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref25 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref25 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref26 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref26 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref26 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref26 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref27 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref27 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref27 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref28 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref28 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref28 https://www.cabidigitallibrary.org/doi/full/10.5555/19570700010 https://www.cabidigitallibrary.org/doi/full/10.5555/19570700010 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref30 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref30 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref30 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref30 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref31 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref31 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref31 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref32 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref32 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref33 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref33 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref33 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref34 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref34 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref34 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref35 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref35 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref36 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref36 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref37 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref37 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref38 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref38 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref38 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref39 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref39 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref39 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref40 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref40 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref40 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref41 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref41 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref41 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref42 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref42 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref42 https://doi.org/10.1016/j.fcr.2016.10.015 https://doi.org/10.1016/j.fcr.2016.10.015 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref44 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref44 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref44 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref44 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref44 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref44 https://doi.org/10.1007/978-3-319-58792-9_1 https://doi.org/10.1007/978-3-319-58792-9_1 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref46 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref46 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref47 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref47 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref47 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref48 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref48 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref48 https://www.cabidigitallibrary.org/doi/pdf/10.5555/20230340603 https://www.cabidigitallibrary.org/doi/pdf/10.5555/20230340603 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref50 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref50 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref50 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref51 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref51 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref51 https://doi.org/10.1038/s41467-024-48859-0 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref53 http://refhub.elsevier.com/S2666-1543(25)00497-1/sref53 https://doi.org/10.1016/j.ecolmodel.2021.109525 https://doi.org/10.1016/j.ecolmodel.2021.109525 Exploring the use of nitrogen fertilization and shifting of planting dates as adaptation strategies to climate change in th ... 1 Introduction 2 Materials and methods 2.1 Description of study area 2.2 Experiments for model calibration 2.3 Experiments for the DSSAT model evaluation 2.4 Model description 2.5 Calibration of model 2.6 Evaluation of model performance 2.7 Assessment of the impact of climate change and adaptation/management strategies on maize yield 2.7.1 Climate change scenarios 2.7.2 Assessing climate change impact under current farmers’ management practices 2.7.3 Assessing increased fertilization and shifts in planting windows in mitigating climate change effect 3 Results 3.1 Projected changes in climate variables 3.2 Simulated baseline yields 3.3 Projected future yields under climate change 3.4 Impact of fertilization under low and high emissions climate scenarios 3.5 Impact of shifting planting dates on yields under low and high emissions climate scenarios 3.6 Using N fertilization as a climate change adaptation strategy 3.7 Assessing shifting planting date as a climate change adaptation 4 Discussions 4.1 Impact of climate change on maize yield 4.2 Mitigating climate change impact on maize with management practices 5 Conclusions CRediT authorship contribution statement Additional information Funding statement Declaration of competing interest Acknowledgments Data availability References