Next Article in Journal
Aquaculture Sludge as Co-Substrate for Sustainable Olive Mill Solid Waste Pre-Treatment by Anthracophyllum discolor
Next Article in Special Issue
The Border Effects of Dry Matter, Photosynthetic Characteristics, and Yield Components of Wheat under Hole Sowing Condition
Previous Article in Journal
Estimating Soil Hydraulic Parameters during Ponding Infiltration Using a Hybrid Algorithm
Previous Article in Special Issue
Winter Oilseed Rape: Agronomic Management in Different Tillage Systems and Seed Quality
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Establishing Optimal Planting Windows for Contrasting Sorghum Cultivars across Diverse Agro-Ecologies of North-Eastern Nigeria: A Modelling Approach

1
International Crop Research Institute for Semi-Arid Tropics (ICRISAT), Kano P.M.B. 349, Nigeria
2
Centre d’étude régional pour l’amélioration de l’adaptation à la sécheresse (CERAAS), Thies Escale BP 3320, Senegal
3
International Institute of Tropical Agriculture (IITA), Ibadan P.M.B. 5320, Nigeria
4
International Livestock Research Institute (ILRI), Dar es Salaam P.O. Box 34441, Tanzania
*
Author to whom correspondence should be addressed.
Agronomy 2023, 13(3), 727; https://doi.org/10.3390/agronomy13030727
Submission received: 22 December 2022 / Revised: 24 February 2023 / Accepted: 25 February 2023 / Published: 28 February 2023
(This article belongs to the Special Issue Cropping Systems and Agronomic Management Practices of Field Crops)

Abstract

:
In the context of climate change, the sowing date and cultivar choice can influence the productivity of sorghum, especially where production is constrained by low soil fertility and early terminal drought across the challenging agro-ecologies of north-eastern Nigeria. Planting within an optimal sowing window to fit the cultivar’s maturity length is critical for maximizing/increasing the crop yield following the appropriate climate-smart management practices. In this study, the APSIM crop model was calibrated and validated to simulate the growth and yield of sorghum cultivars with differing maturing periods sown within varying planting time windows under improved agricultural practices. The model was run to simulate long-term crop performance from 1985 to 2010 to determine the optimal planting windows (PWs) and most suitable cultivars across different agro-ecological zones (AEZs). The performance of the model, validated with the observed farm-level grain yield, was satisfactory across all planting dates and cropping systems. The model predicted a lower mean bias error (MBE), either positive or negative, under the sole cropping system in the July sowing month compared to in the June and August sowing months. The seasonal climate simulations across sites and AEZs suggested increased yields when using adapted sorghum cultivars based on the average grain yield threshold of ≥1500 kgha−1 against the national average of 1160 kgha−1. In the Sudan Savanna (SS), the predicted optimum PWs ranged from 25 May to 30 June for CSR01 and Samsorg-44, while the PWs could be extended to 10 July for ICSV400 and Improved Deko. In the Northern Guinea Savanna (NGS) and Southern Guinea Savanna (SGS), the optimal PWs ranged from 25 May to 10 July for all cultivars except for SK5912, for which predicted optimal PWs ranged from 25 May to 30 June. In the NGS zone, all cultivars were found to be suitable for cultivation with exception of SK5912. Meanwhile, in the SGS zone, the simulated yield below the threshold (1500 kgha−1) could be explained by the sandy soil and the very low soil fertility observed there. It was concluded that farm decisions to plant within the predicted optimal PWs alongside the use of adapted sorghum cultivars would serve as key adaptation strategies for increasing the sorghum productivity in the three AEZs.

1. Introduction

Nigeria is the largest producer of sorghum in West Africa, accounting for about 65–70% of the total sorghum production in the region [1]. Its sorghum production in 2018 was 6.9 million tonnes, accounting for 50% of the total cereal production and occupying about 45% of the total land area devoted to cereal crop production in Nigeria [2]. The production of sorghum in Nigeria, where it is predominantly cultivated in the northern region, has increased overall [3], reaching some seven million tons in 2021, with an average yield of 1160 kgha−1 meaning that it is one of the main crops for the country. The increase in production is associated with the dissemination of improved sorghum cultivars that are tolerant to drought and Striga [4]. These cultivars have been promoted through several initiatives by the Federal Government of Nigeria and other development partners. Landraces have long been recognized as a source of traits for local adaptation, stress tolerance, yield stability, and seed nutrition [5]. The long-term selection under variable and low-input environments has resulted in high crop diversity in landraces. The environmental factors contributing to production constraints and low yields include low fertility soils, the length of the growing periods, drought, and water-logging, as well as biotic stresses such as Striga parasitism and diseases attacking the foliage, stems, and/or grain [6]. Photoperiod sensitivity is an important trait of West African sorghum germplasm that allows farmers to cope with variations in the planting date (PD) and adapt to environmental constraints [7,8]. The triggering of flowering by day length effectively serves to synchronize the final developmental stages with the end of the rainy season [9]. A major problem in rainfed agriculture in semi-arid regions characterized by short rainy season, occasionally accompanied by in-season drought, is how to determine the optimum sowing date [10]. The delays in the onset of the rainfall, drought, unpredictable periodic dry spells, and shortened rainfall seasons have led to a slight shift in the traditionally recommended sorghum planting dates [11].
Crop management must not only adapt to changing climatic conditions to maintain sufficient production but must do so in a way that reduces greenhouse gas emissions as much as possible—i.e., cropping systems must be climate smart [12]. Transformative changes for climate-smart agriculture must include changes to crops, management, and systems that build resilience to climate change impacts and emit relatively low emissions [13]. Although limited data exist, the available studies have shown that the cultivation of sorghum is relatively low in agricultural emissions compared to other crops [14]. Despite the importance of understanding the potential of sorghum to contribute to a climate-smart future and to food security in Nigeria, as well as in the dryland West Africa region, the promotion of productivity-enhancing technologies (climate-smart strategies) among the farmers is becoming imperative for increasing productivity. Therefore, the choice of a sorghum cultivar with an appropriate planting date should be combined such that the productivity of the sorghum would be optimal when the flowering occurs at least 20 days before the terminal drought in the cropping season [7,15,16]. Thus, matching the phenology to the given biotic and abiotic conditions is a prerequisite for good varietal adaptation to a given environment [7]. Crops adapt to diverse environments through considerable plasticity of phenology, the main determinant of which is rainfall [17] in the semi-arid region; meanwhile, the temperature has a stronger effect in the temperate region. “Manipulating this climatic factor would require adequate knowledge of planting dates so as to accurately synchronize rainfall incidences with crop development” [18].
In north-eastern Nigeria, as applies to other semi-arid regions, the length of the growing period (LGP) is mainly a function of the date of the first rains [19,20], which is delayed as we moved northward and varies widely from year to year. The region is prone to climatic risk, and a good knowledge of the cultivar development cycles relative to the planting date is required for improved productivity. However, with the variable onset and distribution of rainfall as well as the frequent occurrence of drought within the growing season, the farmers’ choice of cultivars would depend mainly on their knowledge of the crop’s phenology and yield potential in relation to the local characteristics of the wet season [21,22].
In West Africa’s semi-arid agro-ecology, favourable conditions for sorghum cultivation usually extend from May to November [20]. Thus, floral initiation takes place under decreasing day length, and the growth duration of photoperiod-sensitive cultivars will be shortened when sowing is delayed [23]. Although photoperiod sensitivity benefits sorghum, in that flowering takes place at a relatively fixed calendar date and allows it to mature after the rains end, despite highly variable sowing dates [24], a high degree of poor grain filling is encountered among the late-planted and late-maturing varieties that run out of water if the sorghum is planted too late in the season [25,26]. In this situation, matching suitable cultivars with their optimal planting windows becomes an important management option. In addition, knowing the extent to which planting can be delayed and the likely yield penalty due to later than the optimal planting [27] is important for increasing the productivity of sorghum in a semi-arid environment.
In semi-arid environments, the planting date decision is important not only because of its effect on yield [28], but also because of the need to minimize the risk of establishment failures and ensure the availability of water for unrestricted plant growth and transpiration [17]. Recommendations concerning the planting dates of crops are usually based on agronomic field experiments that are specific to the fields and regions [29]. The majority of such trials cannot be temporally and spatially replicated across diverse agro-ecologies because of seasonal variations. The determination of the optimum sowing dates for sorghum by field experimentation entails repetition over long periods in order to capture the seasonal variability in the rainfall with the varying photoperiod sensitivity cultivars available. Thus, cropping system models (CSMs) have been a proven methodology for understanding the interactions between climate, soils, farming systems, and management [30,31]. These models, therefore, remain important diagnostic tools for decision-making, not only to capture the effects of variability of the rainfall and edaphic factors on crop productivity, but also to suggest sowing date rules and other crop management strategies for better and more sustainable agriculture [31,32]. Cropping system models such as Agricultural Production Systems sIMulator, APSIM [33,34], describe the dynamics of crop growth, soil water, soil nutrients, and plant residues as a function of climate, cropping history, and soil/crop management in a daily time step. Through the linking of crop growth with soil processes, APSIM is particularly suited for the evaluation of the likely impacts of alternative management practices such as varying planting dates on soil resources and crop productivity. The model has been used intensively in the search for strategies for more efficient production, improved risk management, crop adaptation, and sustainable production [33,35,36]. This work, therefore, seeks to establish the response of diverse sorghum cultivars to different planting windows in the three major agro-ecologies of north-eastern Nigeria. To achieve this, the following objectives were set: (i) evaluate the performance of the APSIM model for simulating the contrasting sorghum cultivars under different management systems, soils, and rainfall patterns; (ii) apply the model to determine the optimal PWs and adapted sorghum cultivars for higher grain yield and resilience in order to minimize crop failure across sites and AEZs.

2. Materials and Methods

2.1. Model Calibration (Experiments, Data Collection, Procedure for Model Calibration and Evaluation)

The experimental data used for the calibration were principally generated from on-station field experiments conducted between 2016 and 2018 under optimal conditions (i.e., no water and nitrogen stress) in two AEZs (Abuja, Southern Guinea savannah, and Kano, Sudan savannah) in northern Nigeria. The experiment was designed to evaluate the effects of sowing dates and nutrient responses on contrasted sorghum cultivars. In Abuja, the experiment was established at the International Institute of Tropical Agriculture (IITA) field station (Latitude 9.16° N, and Longitude 7.35° E), while, in Kano, the experiment was established in two locations: (i) the Bayero University Kano (BUK) Teaching and Research Farm (Latitude 12.98° N and Longitude 9.75° E) and (ii) the ICRISAT research field situated within the Institute for Agricultural Research (IAR) station, Wasai Village, Minjibir (Latitude 12.17° N and Longitude 8.65° E). The details of the experiment and the agronomic data collected have been reported [37,38]. Among the 20+ sorghum cultivars commercially available in Nigeria, five contrasting sorghum cultivars that were considered to be widely cultivated were tested based on their breeding selection history for phenology, photoperiod sensitivity, and grain yield productivity. According to a national cultivar report [39,40], ICSV-400 is an early maturing cultivar (85–90 days), is photoperiod-insensitive, and has a yield potential from 2.5 to 3.5 t/ha; Improved Deko is medium maturing (90–110 days) and has a low photoperiod sensitivity and a yield potential from 3.5 to 4.0 t/ha; Samsorg-44 and CSR01 are medium maturing and medium photoperiod-sensitive and have yield potential from 2.0 to 2.5 t/ha; and SK5912 is late maturing (165–175 days) and highly photoperiod-sensitive, with a potential yield of 2.5–3.5 t/ha when grown under optimum conditions.
The daily weather was obtained from an automatic weather station (AWS) installed within a 2 km radius of the experiment for the corresponding years of the experiment and was used for calibration. The parameters include the daily maximum and minimum temperature, the solar radiation, and the rainfall. Management practices such as planting dates, sowing depth, plant density, type and amount of fertilizer applied in form of NPK, and tillage (type, depth, and fraction of above-ground materials incorporated) were recorded and used for the model setup and simulation. The soil samples were taken before planting at each experimental site and were analysed for their physical and chemical properties. The agronomic data, such as dates of flowering and maturity, leaf number per plant, leaf area index (LAI), yield, and final biomass collected [4], were used to determine the cultivar-specific parameters.
The calibration of the APSIM-sorghum module was implemented within the APSIM 7.10 framework based on the phenology, morphology, yield, and aboveground biomass data described earlier. The model APSIM requires a number of inputs, which include the cultivar type, crop management practices/information, soil properties, and daily weather records (rainfall, minimum temperature, maximum temperature, and solar radiation). Crop development follows a thermal time approach with a reported base (Tb) and optimal (Topt) and maximum (Tm) temperatures of 11, 32, and 42 °C [41,42]. The thermal time target for the phase between emergence and panicle initiation is also a function of the day length, and its duration, when divided by the plastochron (°C degrees per leaf), determines the total leaf number. The total leaf number multiplied by the phyllochron (°C d per leaf) determines the thermal time to reach the flag leaf stage, which is thus an emergent property of the model. For parameterizing the genetic coefficients of previously undefined sorghum cultivars, the phenological and morphological stages were based on a combination of observed data and simulation to obtain a yield and above-ground biomass (AGB) that fell within the predefined error limits for each cultivar. Following this method, all coefficients were optimized for further simulation as defined in Table 1. Thereafter, the performance of the model in simulating the phenology (days to flowering and maturity), morphology (leaf number per plant and maximum leaf area index (Max_LAI)), grain yield, and AGB were compared with the observed values and assessed using mean bias error (MBE), root mean square error (RMSE), normalized root mean square error (RMSEn) and the traditional R2 regression statistic (least-squares coefficient of determination) [43]. RMSEn gives a measure (%) of the relative difference between the simulated versus observed data. The simulation was considered excellent with RMSEn < 10%, good if 10–20%, acceptable or fair if 20–30%, and poor >30% [44].
M B E = 1 - ( i = 1 n O i - i = 1 n P i ) i = 1 n O i
R M S E = i = 1   n ( P i - O i ) 2 n 0.5
R M S E n   % = i = 1 n ( P i - O i ) 2 m e a n   o f   o b s e r v e d   d a t a 0.5 × 100
where n is the number of observations, Pi is the predicted value for the ith measurement and Oi is the observed value for the ith measurement, and O and P represent the mean of the observed and predicted values for all of the parameters studied.

2.2. Model Validation (Experiments, Data Collection, Procedure for Model Validation, and Evaluation)

An independent dataset used for model validation was generated from multi-locational on-farm trials for improved sorghum production technology conducted through the farmers’ participatory program between 2013 and 2017. The dataset revealed three distinct cropping systems (intercropping, mixed cropping, and sole cropping) comprising a range of production technologies, including improved sorghum varietal demonstration, seed dressing techniques, conservation agriculture (minimum tillage and conventional tillage), and fertilization strategies aimed at increasing sorghum productivity at the farm level. The additional datasets were obtained from the ICRISAT breeding program from on-farm varietal experiments tested across northern Nigeria spanning four agroecological zones (Sahelian, Sudan Savanna, Northern Guinea, and Southern Guinea Savanna). All the data used are well-documented and include information about basic agronomic management practices such as the sowing date, fertilizer application rate, time of application, planting density, reference geographical coordinates of each farm plot/community, final grain yield, and stalk yield for the five (5) selected and calibrated sorghum cultivars. In addition, variations in the planting date across farms and cultivars were grouped under three months (referred to as “sowing month”), which revealed that 92% of farmers planted in the months of June and July, and only 8% of the farmers sowed in the month of August. Weather data were generated using the downscaled Climate Hazards Group InfraRed Precipitation with Station (CHIRPS) rainfall at a 5.5 km resolution and merged with NASA Power data (temperatures and solar radiation) from the database for Climatology Resource for Agroclimatology, National Aeronautics and Space Administration (NASA) (http://power.larc.nasa.gov, assessed on 25 April 2019) for the corresponding farm’s reference coordinates.
Two sources of soil information were obtained for soil parametrization. The first included field-measured soil characteristics and combined the reconnaissance soil survey of Nigeria reported in 1990 and the soil analysis by the Taking Maize Agronomy to Scale in Africa (TAMASA) project in Kano, Kaduna, and Katsina States, respectively. The second soil data source was downscaled ISRIC (International Soil Reference and Information Centre) soil data in 10 × 10 km grids, with the profile layers (in cm) being 5, 15, 30, 60, 100, and 200, used for the corresponding farm’s references coordinates. After bias correction of the gridded dataset using the available soil measurement, the soil information was extracted from the ISRIC database [45] for each farm’s reference coordinates (the nearest grid point) to run the simulation across the locations. Furthermore, R scripts were developed to (i) append the CHIRPS and NASA power data together and convert each location into a format readily ingestible by APSIM; and (ii) remap the ISRICS gridded soil from 5 cm to 15 cm for the top soil layer as required by APSIM, and then convert these soils into an APSIM SOIL readable format. Following the calibrated cultivar-specific coefficients, an excel executable file was developed that incorporated the management practices, cultivar name, soil, and weather records for the corresponding farm/plot alongside the reported observed grain yield. From the spreadsheet executable file, we created a 3266 APSIM simulation setup that defined different sowing dates, planting densities, and fertilizer applications as reported for the five sorghum cultivars. The model’s simulated and observed value was evaluated only for grain yield across the sowing and cropping system using the mean bias error (MBE) and root means square error (RMSE).

2.3. Bias Correction Methods: Daily Observed Rainfall Versus Gridded Rainfall Data (CHIRPS)

Data from nine (9) rainfall observation stations in northern Nigeria with long-term records (1983–2006) were obtained from the climatological unit of the Nigerian Meteorological Agency (NIMET). The Climate Hazards Group InfraRed Precipitation with Station (CHIRPS) data are satellite-based rainfall products with relatively high resolutions (0.05°) and quasi-global coverage (50° S-50° N) for their daily, pentadal, and monthly precipitation datasets [46]. The data were downscaled over the Nigeria grids and extracted for the reference coordinates of the 9 daily observed rainfall stations and 288 different farms coordinates used in the simulations. The bias correction of the gridded data using station-observed data has been shown to increase its applicability to daily time-step agricultural modelling [47]. Two techniques (linear scaling (LS) and empirical quantile mapping (EQM)) were applied to correct the biases in the dataset during validation process. The LS technique shows better accuracy than EQM and replicated the daily observed rainfall data following the study by [48,49].

2.4. Long-Term Simulations of the Contrasted Sorghum Cultivars under Varying Sowing Windows

The simulations were performed across 33 selected sites in Adamawa and Borno States in north-eastern Nigeria for the five calibrated sorghum cultivars. The sites represent the three agroecological zones of the SS, NGS, and SGS (Table 2). The SS has a long dry season followed by a mono-modal rainfall pattern with a distinct rainy season (May–October) and characterized by a high mean temperature (28–32 °C), short growing season (90–110 days), and low rainfall ranging from 600 to 800 mm [50]. Soils in the SS of Nigeria are highly weathered and fragile with low clay content [51]. The dominant soil class of the site is Alfisol, according to the USDA soil taxonomy [52]. In the NGS, the length of the growing period is between 151 and 180 days [53]. It has a mono-modal rainfall distribution ranging from 900 to 1000 mm annually, and its mean temperatures vary from 28 to 40 °C [54]. According to the world reference baseline, its soils are classified as leached ferruginous tropical soils with high clay content and overlying drift materials [55]. The dominant soil types found in the zone are Alfisols and Entisols, according to the FAO classification. In the SGS, the average maximum temperature in the growing season ranges from 26 to 28 °C, whereas the minimum temperature ranges between 18 and 22 °C [56,57]. The rainfall pattern is mono-modal, with an annual rainfall between 1000 mm and 1524 mm and spread over the 181–210 days that define the growing season [52,56]. The soils in this zone have been identified mainly as Lithosols, Ferralic combisols, Feric acrisols, Oxic haplustalfs and Luvisols [58].
The soil parameters used were obtained from on-site soil characterization using geospatial buffering points at a 20 km radius using an ArcGIS map of the reference indicating the sites/LGAs. For soil characterization and soil sampling, profile pits were dug in the 33 selected sites in Adamawa and Borno States. The profiles and soil types were classified using the FAO guidelines [59]. All laboratory analyses were carried out at the Analytical Services Laboratory of IITA. The total soil organic carbon (total C) was measured using a modified Walkley and Black chromic acid wet chemical oxidation and spectrophotometric method [60]. The total nitrogen (total N) was determined using a micro-Kjeldahl digestion method [61]. The soil pH in water (S/W ratio of 1:2.5) was measured using a glass electrode pH meter and the particle size distribution, following the hydrometer method [62]. The available phosphorus was extracted using the Bray-1 method [63]. The phosphorus in the extract was determined calorimetrically according to the molydo-phosphoric blue method, using ascorbic acid as a reducing agent. K was analysed based on the Mehlich 3 extraction procedure [64]. In Adamawa State, most of the topsoils were coarse-textured with higher sand content. In all, 72% had sandy loam, 17% had clay, and 11% had a sandy clay loam texture (Table 3). The soil pH for the selected communities in Adamawa ranged from 5.9 (Jonkolo-Lama in Shelleng) to 8.0 (Fufure). More than 55% of the soils had pH values for ideal plant growth, indicating neutral (6.1–6.5) to alkaline (8.1–8.3) soil reactions. The soil organic carbon (OC) content in ranged from 0.22% in Daneyel and Suktu to 0.90% in the Guyuk area. The distribution of soil in the study areas revealed that most of the soils had low (0.4–1.0%) OC levels. The total soil N content in the soils ranged from very low (<0.05%) to low (0.06–0.1%), with 67% of the study locations falling within the very low N class and 33% of the study sites indicating low N classes. The soil available P varied across the locations, with very low P (<3.0 mg kg−1) at Woroshi, Tawa, Chikila, Lakumna, Dulmava, Hushere-Zum, Jonkolo-Lama, Sabon-Gari, and Yelwa-Jambore. Low soil available P (3–7 mg kg−1) was found in Demsa-Nassarawa, Bare, Lakati-Libbo, and Suktu, while high P (11–32.1 mg kg−1) content was found in Mbula Kuli, Kikan_Kodomti and Fufure. The results showed that 50% of the study sites fell within the very low P fertility class, 28% of the sites fell within the low P fertility class, and 22% of the sites fell within the high P fertility class. The exchangeable K level across the sites ranged from low to high values, with 22% low (<0.15 cmol+ kg−1), 44% moderate (0.16–0.3 cmol+ kg−1), and 33% high (>0.3 cmol+ kg−1).
Similarly, in Borno state, the majority of the soils were coarse-textured with higher sand content. Out of the 15 sites, 47% had sandy loam, 27% had clay, and 26% had a silt loamy sand texture (Table 4). The soil pH of water for the communities in Borno State ranged from 6.1 to 8.4. More than 70% of the soils had neutral reactions (6.6–7.8), which is the ideal condition for plant growth. The soil OC content in the state ranged from 0.12% to 0.78%. Eight (8) communities equivalent to 53% of the study area had very low OC (<0.4%) levels. The total soil N content in the soils ranged from very low to low, with a very low (<0.05%) status found in the Balbaya, Bila Gusi, Briyel, Buratai, Gwaskara, Jara-Dali, Kubo, Kurba, Mathau, Puba Vidau, Sakwa-Shema, and Tum communities, while the Kwaya Bura, Kwajaffa, and Lakundum communities fell within the low (0.06–0.1%) N fertility class. With the exception of Gwaskara and Lakundum, the top soil available P at all the locations fell within very low (<3.0 mg kg−1) fertility class. The exchangeable K levels were 7% low (<0.15 cmol+ kg−1), 33% moderate (0.16–0.3 cmol+ kg−1), and 60% high (>0.3 cmol+ kg−1) across the sites.
The long-term (1985–2010) weather data used in the model application was a combination of downscaled CHIRPS (for daily rainfall) and the NASA database for Climatology Resource for Agroclimatology (for minimum and maximum air temperature and solar radiation respectively). The simulations were set up to run at different planting windows using the fertilizer N at the national fertilizer rate of recommendation (NPK 60:30:30 kghakgha−1) for sorghum. In the model, 30 kg N were applied at sowing (DAS), with Urea (46% N) top dressed at 30 kg of N ha−1 at 30 DAS. The simulation considered an optimum population to be at a 75 cm inter-row by 30 cm intra-row spacing given 44,444 hills/ha against the farmer’s lower rate of 22,222 hills/ha. Based on expert knowledge and a previous study [22] that found that the sowing period for sorghum across the three agro ecologies stretches over 60 days, we divided the entire sowing period into four equal planting windows to capture the photoperiod sensitivity of the cultivars. The model was set to consider four (4) planting windows as follows: 16–31 May (PW1), 1–15 Jun (PW2), 16–30 Jun (PW3), and 1–15 Jul (PW4), respectively. In addition, rule-based sowing within the sowing window was applied (cumulative rainfall of 20 mm in 3 rainy events) and implemented at the 33 sites. The sowing depth was set to 5 cm, with a sowing density of 4.5 plant m2. Considering the farmers’ practices in the region, a non-successive simulation (single season, non-rotation mode) was adopted, which implies that the water, organic matter, nitrogen, and phosphorus were reset a few weeks before the start of the growing season.
The optimal window for the sowing dates of the sorghum cultivar was based on the average simulated grain yield over the 26-year period and across the sites in each AEZ. Also, the coefficient of variation (CV%), as the ratio of the standard deviation to the mean simulated grain yield, was used to assess the suitable cultivar for each site and AEZ. The level of variability (high or low percentage) determined whether the cultivar had a high or low suitability for the site based on a mean grain yield of ≥1500 kgha−1 as the threshold. The threshold was determined as a break-even yield that farmers can produce for marginal economic benefit as described by [22]. The potential evapotranspiration based on the Penman-Monteith equation [37] in the APSIM model was computed as the addition of the simulated soil evaporation and crop transpiration, and, from that, the water use efficiency for the grain yield (WUEgrain) was calculated.

3. Results

3.1. Model Performance

As depicted in Table 1, there were differences in the cultivar-specific coefficients across the new sorghum cultivars, particularly in the thermal time that defined the crop vegetative and growth. ICSV400 and Improved Deko had a shorter thermal time requirement (in degree days) to attain the end of the juvenile stage compared to CSR01, Samsorg-44, and SK5912, respectively. Both cultivars (ICSV400 and Improved Deko) were originally bred for drought conditions, which could allow them to serve as a drought escaping mechanism compared to the other cultivars. Also, the calibrated photoperiod slope varied from 11.5 °C/h to 600 °C/H, indicating a shorter degree/hour for low photoperiod sensitivity cultivars such as ICSV400 and improved Deko, while a longer degree/hour was calibrated for the medium and high photoperiod sensitivity cultivars. The thermal time from flowering to physiological maturity above a base temperature of 10 °C was 560 °C days for ICSV400 and improved Deko, indicating a higher value than the degree days of CSR01 (460 °C days), Samsorg-44 (500 °C days), and SK5912 (450 °C days), respectively. The cultivar genetics coefficients for leaf appearance rate followed two steps, i.e., leaf appearance to the development of most leaf ligules (leaf_app_rate 1) and to the last leaf ligule (leaf_app_rate 2). The calibrated values (56 °C d/leaf and 28 °C d/leaf) were the same for all of the varieties except for ICSV400. These values justified the increase in the leaf number (>20) per plant for most West African sorghum cultivars that are photoperiod sensitive.
The performance of the model, presented in Table 5, shows that the simulated days to 50% flowering and to physiological maturity were good and reproduced the observed values with a mean bias error (MBE) ranging from −4 to 4 days (50% flowering) and from 1 to 2 days (physiological maturity). The RMSE of the mean observed estimate of ≤10% for all the cultivars confirmed the robustness of the predictions. The model’s adjustment of the leaf appearance rate for leaf ligules helps to get an accurate total leaf number (TLN) per plant close to the observed. The estimates of the MBE varied from one to five leaves, and RMSE (of the mean observed) ranged from a high model accuracy (6.4% for improved Deko) to a fairly low accuracy (26.2% for Samsorg-44) for TLN.
The simulated and observed maximum Leaf Area Index (Max_LAI) for all cultivars agrees well with RMSE (% of mean observed), indicating high accuracy for CSR01 and SK5912, low accuracy for improved Deko and Samsorg-44, and very low accuracy for ICSV400. The grain yield and total biomass were acceptably simulated for the contrasted sorghum cultivars within the bounds of statistical errors (Figure 1). For grain yield (Figure 1a), CSR01 had the lowest MBE of −48 kghakgha−1, which under-predicted the observed mean, followed by ICSV-400 (103 kghakgha−1) and improved Deko (114 kghakgha−1), while the highest yield (279 kgha−1) was shown by the cultivar Samsorg-44. The relative RMSE ranged from high accuracy for SK5912 (9.2%) to very low accuracy for ICSV-400 (28.7%). For total biomass (Figure 1b), the relative RMSE ranged from high accuracy for SK5912 (6.9%) to very low accuracy for improved Deko (36.8%).

3.2. Model Validation: Performance with Farm-Level Grain Yield

The performance of the model in simulating grain yield was compared to the observed values under varying planting dates and cropping systems for each sorghum cultivar (Table 6). The planting dates across farms and cultivars were grouped under three months (referred to as “sowing month”), and the number of observations/farms revealed that 92% of farmers planted in the months of June and July, and only 8% of the farmers sowed in the month of August. For ICSV-400, the model under-predicted the mean observed yield for intercropping and mixed cropping systems, but the model over-predicted the mean observed yield for the sole cropping system across the sowing months. The lowest MBE of −977 kgha−1 was estimated in theJuly sowing month under the intercropping system, followed by the mixed cropping system, while the highest MBE (781 kgha−1) was estimated under the sole cropping system in the month of June. The results showed that the model over-predicted the mean observed grain yield for Improved Deko across sowing months under the sole cropping system, with the lowest MBE (66 kgha−1) estimated for the July sowing month, while the highest MBE (548 kgha−1) was estimated for August sowing. The model over-predicted the grain yield across the sowing months and cropping systems except for the June sowing month under sole cropping system, for which lowest MBE of −234 kgha−1 was estimated. The highest MBE of 624 kgha−1 was estimated for July sowing under sole cropping. For CSR01, the model under-predicted the mean observed grain yield across sowing months and cropping systems except for the August sowing month under a mixed cropping system. Similarly, for SK5912, the model over-predicted the mean observed grain yield under the sole cropping system across sowing months, while the model under-predicted across sowing months for the mixed cropping system.
Figure 2 shows the model performance and the differences between the observed and simulated yield pooled together irrespective of the cropping systems and management practices for each cultivar. The mean observed grain yield for ICSV-400, CSR01, Improved Deko, Samsorg-44, and SK5912 are 1479, 1613, 1431, 1197, and 1446 kgha−1, respectively. Further statistical indices showed that the grain yield of the ICSV-400, Improved Deko, and Samsorg-44 cultivars, respectively, were over-predicted against the mean observed grain yield; meanwhile, the yields of the CSR01, and SK5912 cultivars were slightly under-predicted compared to the mean observed yield. The results revealed low MBEs for CSR01 (−228 kgha1), SK5912 (−241 kgha−1), and Samsorg-44 (102 kgha−1), respectively, with an RMSE of 642 kgha−1 estimated for improved Deko, and an RMSE of 655 kgha−1 estimated for Samsorg-44. The CV (%) described the level of variability for each cultivar simulated, which shows the lowest value of 8.9% for Samsorg-44, followed by Improved Deko (CV = 12.3%), while the highest variability was observed for CSR01 and SK5912 (CV = 25.5 and 18.4%).

3.3. Seasonal Rainfall and Temperature Trends across the Simulated Sites

The long-term (1985–2010) rainfall indicated that the rainy season starts in May and ends in October, with the highest peak observed in the month of August (Table 7 and Table 8). The tables further revealed that about 50–60% of the seasonal rainfall was observed in the months of July and August, with a high inter-seasonal variability indicated by the coefficients of variation (CV), ranging from 18 to 23%. All of the study sites showed a distinct mono-modal rainfall pattern and warming temperature throughout the year. Figure 3 and Figure 4 show the average monthly variations in the maximum and minimum temperatures across the selected sites in the Adamawa and Borno States. The maximum temperature uniformly decreases faster than the minimum temperature during the growing season (May–October). In addition, the estimated CV% values for the maximum temperature, ranging from 3.0 to 3.7%, are higher than those of minimum temperature, which range from 2.0 to 2.3% in both states, suggesting that no significant inter-annual variability was observed at the sites for either temperature.
In Adamawa State (Table 7), the seasonal rainfall (May-Oct.) for all of the sites over the 31-year period (1985–2010) ranged from 851 to 1104 mm. It was observed that the rainfall in Dulmava, Hushere Zum, and Guyaku and Tawa was slightly higher (>1000 mm) than in the other locations. The average monthly maximum temperature across the sites over the climatic period ranged from 27.5 to 39.1 °C (Figure 3a), while the average monthly minimum temperature ranged from 15.8 to 24.9 °C (Figure 3b). In Borno State (Table 8), the seasonal rainfall over the 31-year period (1985–2010) across the sites ranged from 883–998 mm with high inter-seasonal variability, varying from 18 to 22%. The average monthly maximum temperature across the sites over the climatic period ranged from 27.8 to 38.9 °C (Figure 4a), while the average monthly minimum temperature ranged from 15.5 to 24.7 °C (Figure 4b).

3.4. Seasonal Analysis of Planting Windows and Sorghum Cultivars on Simulated Grain Yield and Water Use Efficiency (WUEgrain)

Table 9 shows the mean simulated grain yield (GY) and the water use efficiency for grain yield (WUEgrain) of the sorghum cultivars across four different planting windows (PW1, PW2, PW3, and PW4) in the three agro-ecological zones (AEZs) between 1985 and 2010. The mean simulated grain yield and WUEgrain showed a decrease with delayed planting (PW1 to PW4) for all five sorghum cultivars. Following the sowing rule strategies implemented for the simulation, the model outputs indicate approximately 45 days of PW, from 25 May to 10 July across AEZs, for all sorghum cultivars except for SK5912, which has approximately 35 days of planting window varying from 25 May and 30 June in the NGS and SGS. A higher mean GY and WUEgrain were simulated in the NGS than in the SS and SGS zones. Additionally, the early and medium-maturing sorghum cultivars (ICSV400, Improved Deko, CSR01, and Samsorg44) had higher simulated GY and WUEgrain values than those of the late-maturing cultivar (SK5912).
For the SS zone, the optimal sowing window simulated ranged from 25 May to 30 June (PW 1 to PW3) for CSR01 and Samsorg-44 and from 25 May to 15 June (PW1 and PW2) for SK5912, while, for the ICSV400 and Improved Deko cultivars, sowing can extend to 10 July. In the NGS and SGS zones, the optimal planting window ranged from 25 May to 10 July for all sorghum cultivars except for SK5912, for which 25 May to 30 June was simulated to be the optimal planting window. The highest mean WUEgrain of 6.4–7.8 kgha−1 mm−1 was simulated for ICSV400. Next to it was improved Deko with a WUEgrain of 5.8–6.8 kgha−1 mm−1, and SK5912 was simulated to have the lowest WUEgrain (2.3–4.2 kgha−1 mm−1) across the three AEZs
Table 10 shows the mean simulated grain yield for evaluating the adapted sorghum cultivars across sites based on an increased yield threshold of ≥1500 kgha−1 and against the national average grain yield of 1160 kgha−1. In the SS zone, the simulated mean grain yield across the selected sites ranged from 2023 to 2673 kgha−1 for ICSV400, 1886–2509 kgha−1 for Improved Deko, 1022–3707 kgha−1 for CSR01, 939–3324 kgha−1 for Samsorg-44, and 730–2847 kgha−1 for SK5912, respectively. The CV shows the variability of the simulated GY across sites, with lower values estimated by ICSV400 (10%) and Improved Deko (9%) compared to higher values estimated by CSR01 (46%), Samsorg-44 (44%), and SK5912 (56%).
In NGS zone, the simulated mean grain yield across the sites ranged from 1926 to 2512 kgha−1 for ICSV400, 1841–2390 kgha−1 for Improved Deko, 1269–3761 kgha−1 for CSR01), 1174–3432 kgha−1 for Samsorg-44, and 940–3140 kgha−1 for SK5912. The variability of the GY across sites indicated low CV% for ICSV400 and Improved Deko (11%) compared to high CV% estimates for CSR01 (30%), Samsorg-44 (30%), and SK5912 (41%). In the SGS zone, the simulated mean grain yield across the sites ranged from 1306 to 2330 kgha−1 for ICSV400, 1165–2220 kgha−1 for Improved Deko, 1010–3049 kgha−1 for CSR01, 971–2707 kgha−1 for Samsorg-44, and 891–2325 kgha−1 for SK5912. The CV% was generally high for all cultivars, ranging from 24 to 43%. At the mean grain yield threshold of ≥1500 kgha−1, all cultivars simulated were found to be adapted for cultivation except at the Fufure site.

4. Discussion

This study contributes to efforts to develop climate risk strategies for the sorghum-based mixed farming systems in northern Nigeria. The evaluation of the model calibration and its validation with an independent dataset (farm-level yield) under different management, soils, and climatic conditions allow the APSIM-sorghum model to be applied to understanding the dynamics of this heterogeneous farming system. The application of crop modelling to develop adaptation strategies to changing climatic conditions was earlier demonstrated for sorghum by [22,31] and for maize by [65]. The predicted LAI-max and total leaf number (TLN) indicated a low accuracy (RMSE varied from 20 to 30%) due to the relatively higher values simulated for July sowing dates resulting in a higher mean grain yield simulated under calibration. However, the difficulty in predicting TLN could be linked to the fixed thermal time targets for each of the phases before flowering in the APSIM-sorghum module. These thermal time targets are not directly linked to leaf initiation and appearance [66]. The predictions of the grain yield (GY) and total biomass (TB) ranged from high accuracy RMSEn (SK5912: 9.2% for GY; 6.9% for TB) to very low accuracy RMSEn (ICSV400: 28.7% for GY; Improved Deko: 36.8% for TB) when evaluated against the observed mean. The low accuracy for GY and TB could be associated with the simulation of leaf initiation and leaf appearance, which are important for the accurate prediction of morphological traits [31,67].
The use of model evaluation using simple on-station trial datasets is the common procedure for developing new cultivar parameterizations. However, evaluating models with multi-locational, on-farm trial datasets has proven difficult, with many uncertainties, especially across the different soil, climate, and cropping systems considered [66]. The study presented here utilized comprehensive data from on-farm trials using different planting dates, cropping systems, fertilization strategies, soil types, and management regimes representing the heterogeneous farming system of northern Nigeria. The performance of the model was satisfactory under varying planting dates (referred to as “sowing month”), cropping systems, and sorghum cultivars as described in Table 6. With exception of the CSR01 and SK5912 cultivars, the model’s predictions had a lower MBE, either positive or negative, for the sole cropping system in the July sowing month compared to the June and August sowing months. These results could be explained by the pattern of rainfall that serves as a means of crop water utilization, which in turn determines the biomass accumulation for the grain yield. The high rainfall variability across the study sites suggests the importance of matching crop duration to the length of the growing period in the region because sorghum is a short-day crop and most West African cultivars are photoperiod sensitive that could only be produced under rainfed conditions [23,31]. These conditions place limits on the use of long-season sorghum cultivars in some locations even within the same AEZ, which permits the choice of early-medium maturing cultivars. Although the soil fertility composition across the sites suggested low values for organic carbon (OC) and nitrogen N, the pH values indicated ideal soils (neutral to alkaline conditions) suitable for plant growth of sorghum [51].
Our simulations revealed that the optimal PWs and suitable sorghum cultivars were influenced by the dates of sowing, soil types, rainfall amount, and pattern across sites and AEZs. In addition, this has to do with the cultivar’s sensitivity or insensitivity to photoperiod and inherently early/late flowering traits [68]. These results corroborate the findings by [23], who reported that inherent soil fertility and rainfall patterns can greatly influence the yield when sowing is delayed. Both early and medium-maturing sorghum cultivars (ICSV400, Improved Deko, CSR01, and Samsorg-44) produced higher GY and WUEgrain than those of the late-maturing cultivar (SK5912) at varying PWs and were found suitable to most sites across the AEZs. The optimal PWs slightly varied among the cultivars and AEZs. Our simulation results suggest an optimal sowing window for the ICSV400 and Improved Deko cultivars from 25 May to 10 July (45 days) and an optimal window for CSR01 and Samsorg-44 from 25 May to 30 June (35 days) in the SS zone. The results further revealed that the planting of CSR01, Samsorg-44, and SK5912 beyond these dates will significantly reduce the mean grain yield by 7%, 9%, and 11%, with no significant yield change estimated for the ICSV400 and Improved Deko cultivars. In the NGS and SGS zones, the optimal PWs ranged from 25 May to 10 July (45 days), except for SK5912, for which 25 May to 30 June (35 days) was simulated.
These results showed the use of early and medium maturing sorghum cultivars with higher yield and the most suitable cultivars to varying soil types simulated across the AEZs. In the SS zone, the level of variability suggests that ICSV400 and Improved Deko were highly suitable for cultivation across the sites; CSR01 and Samsorg-44 were suitable for cultivation in almost all the sites with exception of Guyaku, Balbaya, Mathau Puba Vidau, and Kwajaffa, while the late maturing cultivar (SK5912) adapted for cultivation only in 4 (Buratai, Kabura, Kurbo-Gayi, Lakundum) out of 14 sites. These results suggest only 4 out of the 5 sorghum cultivars may be suitable for cultivation under the current climatic conditions. In NGS, at a mean grain yield threshold of ≥1500 kgha−1 and the level of variability across the sites, all the cultivars were found to be adapted and suitable for cultivation in most sites, except for CSR01 and Samsorg-44 at Bare, and SK5912 at Bare, Daneyel, Hushere Zum, and Lakati-Libbo, respectively. The simulated yields of all the sorghum cultivars at the Fufure site in the SGS zone were found to be below the yield threshold of ≥1500 kgha−1, and these results could be associated with sandy soil in the area and the very low soil fertility resulting in low water retention for crop growth. Also, a late PW reduced the grain yield due to early terminal drought towards the cessation of the growing period, resulting in a high temperature that affects the grain filling period, i.e., slows the rate of grain filling and accelerates senescence, thereby decreasing the photosynthetic activities per unit leaf area [69]. In addition, the increased temperature and water deficit experienced in the late planting window, particularly in PW4, can reduce the crop canopy (leaves and tillers) and decrease the biomass production, which in turn reduces the grain yield.

5. Conclusions

The validation of the model with farm-level grain yield enhanced the predictive capacity of the model for simulating diverse climatically driven yields under different fertilization strategies, sowing dates, and planting densities for the contrasting sorghum cultivars. However, our model application used different PWs based on climate-smart management practices that include the recommended fertilizer application rate and optimal hill population against the farmer practices for sorghum production in Northern Nigeria, geared towards disseminating and increasing the adoption of climate-smart technology, which is the basis for higher productivity. The optimum PWs were simulated as being between 25 May and 30 June for CSR01 and Samsorg-44 but were extended to 10 July for ICSV400 and Improved Deko, while low yield was simulated for SK5912 for all planting windows in the SS zone. In the NGS and SGS zones, the optimal PWs ranged from 25 May to 10 July (45 days) for all cultivars except for SK5912, for which predicted optimal PWs ranged from 25 May to 30 June (35 days). The mean simulated GY for SK5912 fell below the threshold of ≥1500 kgha−1 in Bare, Daneyel, Hushere Zum, and Lakati-Libbo`. In addition, at the Fufure site in SGS, all of the sorghum cultivars were simulated to be below the yield threshold ≥1500 kg ha−1 due to sandy soil texture found in the area, with the very low soil fertility resulting in a low water retention capacity for growth. Under climate change, the adoption of appropriate climate-smart technology sorghum will improve food security and reduce greenhouse gas emissions. It may therefore be concluded that the predicted optimal PWs for sorghum would substantially assist the smallholder farmers and seed producers in the region in their choice of cultivars to promote for high yields relative to growing sites and agro-ecologies.

Author Contributions

F.M.A. and H.A.A.: experimental design. F.M.A. and H.A.A.: experimentation. F.M.A., H.A.A., A.M.W. and A.Y.K.: methods. F.M.A., A.O.O. and A.I.T.: statistical analysis. F.M.A., A.O.O. and A.I.T.: manuscript draft. F.M.A., H.A.A., A.M.W. and A.Y.K.: final manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

From 2017–2020, financial support came from the CGIAR Research Programs Grain Legumes and Dryland Cereals (GLDC), Climate Change, Agriculture and Food Security (CCAFS) carried out with support from the CGIAR Trust Fund and through bilateral funding agreements (for details, visit https://www.cgiar.org/funders/) is also acknowledged. The World Bank-funded AICCRA project (Accelerating Impacts of CGIAR Climate Research for Africa) Project ID 173398 is acknowledged for funding the involvement of F.M.A. and A.M.W. in this study. The Integrated Agriculture Activity project funded through the USAID is gratefully acknowledged.

Data Availability Statement

The raw data supporting the conclusions of this manuscript will be made available by the authors, without undue reservation.

Acknowledgments

The authors would like to thank the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) for their support of the study. We are grateful for the help of the field and laboratory staff of ICRISAT Kano for managing the field trials for this study.

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the collection, analysis, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References

  1. Ogbonna, A.C. Current Developments in Malting and Brewing Trials with Sorghum in Nigeria: A review. J. Inst. Brew. 2011, 117, 394–400. [Google Scholar] [CrossRef]
  2. FAOSTAT. FAO Statistical Database (online). Food and Agricultural Organization of the United Nations. Rome. 2019. Available online: http://www.fao.org/faostat/en/#data/ (accessed on 25 July 2020).
  3. FAOSTAT. Food and Agricultural Organization of the United Nations (FAO) FAO Statistical Database. 2021. Available online: http://faostat.fao.org (accessed on 19 March 2022).
  4. Ajeigbe, H.A.; Akinseye, F.M.; Ayuba, K.; Jonah, J. Productivity and Water Use Efficiency of Sorghum [Sorghum bicolor (L.) Moench] Grown under Different Nitrogen Applications in Sudan Savanna Zone, Nigeria. Int. J. Agron. 2018, 2018, 7676058. [Google Scholar] [CrossRef] [Green Version]
  5. Dwivedi, S.L.; Ceccarelli, S.; Blair, M.W.; Upadhyaya, H.D.; Are, A.K.; Ortiz, R. Landrace Germplasm for Improving Yield and Abiotic Stress Adaptation. Trends Plant Sci. 2016, 21, 31–42. [Google Scholar] [CrossRef]
  6. Ajeigbe, H.A.; Singh, B.; Emechebe, A. Field evaluation of improved cowpea lines for resistance to bacterial blight, virus and Striga under natural infestation in the West African Savannas. Afr. J. Biotechnol. 2008, 7, 3563–3568. [Google Scholar]
  7. Kouressy, M.; Dingkuhn, M.; Vaksmann, M.; Heinemann, A.B. Adaptation to diverse semi-arid environments of sorghum genotypes having different plant type and sensitivity to photoperiod. Agric. For. Meteorol. 2008, 148, 357–371. [Google Scholar] [CrossRef]
  8. Abdulai, A.L.; Kouressy, M.; Vaksmann, M.; Asch, F.; Giese, M.; Holger, B. Latitude and Date of Sowing Influences Phenology of Photoperiod-Sensitive Sorghums. J. Agron. Crop Sci. 2012, 198, 340–348. [Google Scholar] [CrossRef]
  9. Craufurd, P.Q.; Wheeler, T.R. Climate change and the flowering time of annual crops. J. Exp. Bot. 2009, 60, 2529–2539. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  10. Marteau, R.; Sultan, B.; Moron, V.; Alhassane, A.; Baron, C.; Traoré, S.B. The onset of the rainy season and farmers’ sowing strategy for pearl millet cultivation in Southwest Niger. Agric. For. Meteorol. 2011, 151, 1356–1369. [Google Scholar] [CrossRef] [Green Version]
  11. Yamusa, A.M.; Abu, S.T.; Yahaya, R.A.; Musa, I.J. Assessing the planting dates of sorghum in a changing climate at Samaru, Northern Nigeria. Int. J. Biol. Environ. Sci. Trop. 2013, 10, 21–26. [Google Scholar]
  12. FAO. Climate Smart Agriculture: Policies, Practices and Financing for Food Security, Adaptation and Mitigation; Food and Agriculture Organization: Rome, Italy, 2010. [Google Scholar]
  13. Lipper, L.; Thornton, P.; Campbell, B.M.; Baedeker, T.; Braimoh, A.; Bwalya, M.; Caron, P.; Cattaneo, A.; Garrity, D.; Henry, K.; et al. Climate-smart agriculture for food security. Nat. Clim. Chang. 2014, 4, 1068–1072. [Google Scholar] [CrossRef]
  14. Ayinde, T.B.; Ahmed, B.; Nicholson, C.F. Farm-Level Impacts of Greenhouse Gas Reductions for the Predominant Production Systems in Northern Nigeria. In African Handbook of Climate Change Adaptation; Oguge, N., Ayal, D., Adeleke, L., da Silva, I., Eds.; Springer: Cham, Switzerland, 2021. [Google Scholar] [CrossRef]
  15. Traore, S.B.; Reyniers, F.-N.; Vaksmann, M.; Kone, B.; Sidibe, A.; Yorote, A.; Yattara, K.; Kouressy, M. Adaptation a ‘la se’cheresse des e´cotypes locaux de sorghos du Mali. Se´cheresse 2000, 11, 227–237. [Google Scholar]
  16. Traoré, P.C.S.; Kouressy, M.; Vaksmann, M.; Tabo, R.; Maikano, I.; Traoré, S.B.; Cooper, P. Climate Prediction and Agriculture: What Is Different about Sudano-Sahelian West Africa? In Climate Prediction and Agriculture; Sivakumar, M.V.K., Hansen, J., Eds.; Springer: Berlin/Heidelberg, Germany, 2007. [Google Scholar] [CrossRef]
  17. Hadebe, S.T.; Mabhaudhi, T.; Modi, A.T. Water use of sorghum (Sorghum bicolor L. Moench) in response to varying planting dates evaluated under rainfed conditions. Water SA 2017, 43, 91. [Google Scholar] [CrossRef] [Green Version]
  18. Adetayo, A.O.; Dauda, T.O.; Adetayo, O.B.; Asiribo, O.E.; Issa, F. Rainfall instability difference in the effects of planting dates on growth and yield of maize (Zea mays) in forest savannah eco-climatic zone of Nigeria. Afr. J. Agric. Res. 2008, 3, 700–707. [Google Scholar]
  19. Sivakumar, M. Predicting rainy season potential from the onset of rains in Southern Sahelian and Sudanian climatic zones of West Africa. Agric. For. Meteorol. 1988, 42, 295–305. [Google Scholar] [CrossRef] [Green Version]
  20. Akinseye, F.M.; Agele, S.O.; Traore, P.C.S.; Adam, M.; Whitbread, A.M. Evaluation of the onset and length of growing season to define planting date—‘A case study for Mali (West Africa)’. Theor. Appl. Clim. 2015, 124, 973–983. [Google Scholar] [CrossRef] [Green Version]
  21. Lacy, S.M.; Cleveland, D.A.; Soleri, D. Farmer choice of sorghum cultivars in Southern Mali. Hum. Ecol. 2006, 34, 331–353. [Google Scholar] [CrossRef]
  22. Akinseye, F.M.; Ajeigbe, H.A.; Traore, P.C.; Agele, S.O.; Zemadim, B.; Whitbread, A. Improving sorghum productivity under changing climatic conditions: A modelling approach. Field Crop. Res. 2019, 246, 107685. [Google Scholar] [CrossRef]
  23. Folliard, A.; Traore, P.C.S.; Vaksmann, M.; Kouressy, M. Modelling of sorghum response to photoperiod: A threshold-hyperbolic approach. Field Crops Res. 2004, 89, 59–70. [Google Scholar] [CrossRef] [Green Version]
  24. Clerget, B.; Dingkuhn, M.; Goze, E.; Rattunde, H.F.W.; Ney, B. Variability of Phyllochron, Plastochron and Rate of Increase in Height in Photoperiod-sensitive Sorghum Varieties. Ann. Bot. 2008, 101, 579–594. [Google Scholar] [CrossRef] [Green Version]
  25. Haussmann, B.I.G.; Rattunde, H.F.; Weltzien-Rattunde, E.; Traoré, P.S.C.; Brocke, K.V.; Parzies, H.K. Breeding Strategies for Adaptation of Pearl Millet and Sorghum to Climate Variability and Change in West Africa. J. Agron. Crop. Sci. 2012, 198, 327–339. [Google Scholar] [CrossRef] [Green Version]
  26. Weltzien, E.; Rattunde, H.F.W.; van Mourik, T.A.; Ajeigbe, H.A. Sorghum cultivation and improvement in West and Central Africa. In Achieving Sustainable Cultivation of Sorghum Volume 2: Sorghum Utilization around the World; Rooney, W., Ed.; Burleigh Dodds Science Publishing: Cambridge, UK, 2018. [Google Scholar]
  27. Staggenborg, S.A.; Fjell, D.L.; Devlin, D.L.; Gordon, W.B.; Maddux, L.D.; Marsh, B.H. Selecting Optimum Planting Dates and Plant Populations for Dryland Corn in Kansas. J. Prod. Agric. 1999, 12, 85–90. [Google Scholar] [CrossRef]
  28. Santos, R.D.; Boote, K.J.; Sollenberger, L.E.; Neves AL, A.; Pereira LG, R.; Scherer, C.B.; Gonçalves, L.C. Simulated optimum sowing date for forage pearl millet cultivars in multi-location trials in Brazilian semi-arid region. Front. Plant Sci. 2017, 8, 02074. [Google Scholar] [CrossRef] [Green Version]
  29. Tofa, A.I.; Chiezey, U.F.; Babaji, B.A.; Kamara, A.Y.; Adnan, A.A.; Beah, A.; Adam, A.M. Modeling Planting-Date Effects on Intermediate-Maturing Maize in Contrasting Environments in the Nigerian Savanna: An Application of DSSAT Model. Agronomy 2020, 10, 871. [Google Scholar] [CrossRef]
  30. Bassu, S.; Brisson, N.; Durand, J.L.; Boote, K.J.; Lizaso, J.; Jones, J.W.; Rosenzweig, C.; Ruane, A.C.; Adam, M.; Baron, C.; et al. Do various maize crop models give the same responses to climate change factors? Glob. Change Biol. 2014, 20, 2301–2320. [Google Scholar] [CrossRef] [Green Version]
  31. Akinseye, F.M.; Adam, M.; Hoffmann, M.; Traore, P.; Agele, S.; Whitbread, A. Assessing crop model improvements through comparison of sorghum (Sorghum bicolor L. moench) simulation models: A case study for West African cultivars. Field Crop Res. 2017, 201, 19–31. [Google Scholar] [CrossRef] [Green Version]
  32. Amanullah, M.; Mohamed, C.K.; Safiullah, A.; Selvam, S.; Sivakumar, K. Crop simulation growth model in Cassava. Res. J. Agric. Biol. Sci. 2007, 3, 255–259. [Google Scholar]
  33. Keating, B.A.; Carberry, P.S.; Hammer, G.L.; Probert, M.E.; Robertson, M.J.; Holzworth, D.; Huth, N.I.; Hargreaves, J.N.; Meinke, H.; Hochman, Z.; et al. An overview of APSIM, a model designed for farming systems simulation. Eur. J. Agron. 2003, 18, 267–288. [Google Scholar] [CrossRef] [Green Version]
  34. Holzworth, D.P.; Huth, N.I.; Devoil, P.G.; Zurcher, E.J.; Herrmann, N.I.; McLean, G.; Chenu, K.; van Oosterom, E.J.; Snow, V.; Murphy, C.; et al. APSIM—Evolution towards a new generation of agricultural systems simulation. Environ. Model. Softw. 2014, 62, 327–350. [Google Scholar] [CrossRef]
  35. Van Ittersum, M.K.; Leffelaar, P.A.; Van Keulen, H.; Kropff, M.J.; Bastiaans, L.; Goudriaan, J. On approaches and applications of the Wageningen crop models. Eur J Agron. 2003, 18, 201–234. [Google Scholar] [CrossRef]
  36. Whitbread, A.; Robertson, M.; Carberry, P.; Dimes, J. How farming systems simulation can aid the development of more sustainable smallholder farming systems in southern Africa. Eur. J. Agron. 2010, 32, 51–58. [Google Scholar] [CrossRef] [Green Version]
  37. Allen, R.G.; Pereira, L.S.; Raes, D.; Smith, M. Crop evapotranspiration—Guidelines for computing crop water requirements. In FAO Irrigation and Drainage; Food and Agriculture Organization: Rome, Italy, 1998; Paper 56. [Google Scholar]
  38. Ajeigbe, H.A.; Akinseye, F.M.; Kunihya, A.; Jonah, J. Sorghum yield and water use under Phosphorus applications in Sudan Savanna zone of Nigeria. Glob. Adv. Res. J. Agric. Sci. 2018, 7, 245–257. [Google Scholar]
  39. NACGRAB, FMST. Crop Varieties Released and Registered in Nigeria; National Centre for Genetic Resources and Biotechnology, produced by Federal Ministry of Science and Technology: Ibadan, Nigeria, 2009; p. 42. [Google Scholar]
  40. NACGRAB, FMST. Guidelines for Registration and Release of New Crop Varieties in Nigeria; National Centre for Genetic Resources and Biotechnology, produced by Federal Ministry of Science and Technology: Ibadan, Nigeria, 2016; p. 23. [Google Scholar]
  41. Carberry, P.S.; Hammer, G.L.; Muchow, R.C. Modelling genotypic and environmental control of leaf area dynamics in grain sorghum. III. Senescence and prediction of green leaf area. Field Crops Res. 1993, 33, 329–351. [Google Scholar] [CrossRef]
  42. Carberry, P.S.; Muchow, R.C.; Hammer, G.L. Modelling genotypic and environmental control of leaf area dynamics in grain sorghum. II. Individual leaf level. Field Crops Res. 1993, 33, 311–328. [Google Scholar] [CrossRef]
  43. Loague, K.; Green, R.E. Statistical and graphical methods for evaluating solute transport models: Overview and application. J. Contam. Hydrol. 1991, 7, 51–73. [Google Scholar] [CrossRef]
  44. Jamieson, P.D.; Porter, J.R.; Wilson, D.R. A test of the computer simulation model ARCWHEAT1 on wheat crops grown in New Zealand. Field Crop. Res. 1991, 27, 337–350. [Google Scholar] [CrossRef]
  45. Funk, C.; Peterson, P.; Landsfeld, M.; Pedreros, D.; Verdin, J.; Shukla, S.; Husak, G.; Rowland, J.; Harrison, L.; Hoell, A.; et al. The climate hazards infrared precipitation with stations—A new environmental record for monitoring extremes. Sci. Data 2015, 2, 150066. [Google Scholar] [CrossRef] [Green Version]
  46. Batjes, N.H. Harmonized soil profile data for applications at global and continental scales: Updates to the WISE database. Soil Use Manag. 2009, 25, 124–127. [Google Scholar] [CrossRef]
  47. Gummadi, S.; Dinku, T.; Shirsath, P.B.; Kadiyala, M.D.M. Evaluation of multiple satellite precipitation products for rainfed maize production systems over Vietnam. Sci. Rep. 2022, 12, 485. [Google Scholar] [CrossRef]
  48. Teutschbein, C.; Seibert, J. Bias correction of regional climate model simulations for hydrological climate-change impact studies: Review and evaluation of different methods. J. Hydrol. 2012, 456–457, 12–29. [Google Scholar] [CrossRef]
  49. Crochemore, L.; Ramos, M.-H.; Pappenberger, F. Bias correcting precipitation forecasts to improve the skill of seasonal streamflow forecasts. Hydrol. Earth Syst. Sci. 2016, 11, 1145–1159. [Google Scholar] [CrossRef] [Green Version]
  50. Adnan, A.A.; Jibrin, J.M.; Kamara, A.Y.; Abdulrahman, B.L.; Shaibu, A.S.; Garba, I.I. CERES–Maize Model for Determining the Optimum Planting Dates of Early Maturing Maize Varieties in Northern Nigeria. Front. Plant Sci. 2017, 8, 1118. [Google Scholar] [CrossRef] [Green Version]
  51. Shehu, B.M.; Jibrin, J.M.; Samndi, A.M. Fertility Status of Selected Soils in the Sudan Savanna Biome of Northern Nigeria. Int. J. Soil Sci. 2015, 10, 74–83. [Google Scholar] [CrossRef] [Green Version]
  52. Dawaki, U.; Dikko, A.; Noma, S.; Aliyu, U. Heavy Metals and Physicochemical Properties of Soils in Kano Urban Agricultural Lands. Niger. J. Basic Appl. Sci. 2014, 21, 239. [Google Scholar] [CrossRef] [Green Version]
  53. Jagtap, S. Changes in annual, seasonal, and monthly rainfall in Nigeria during 1961-90 and consequences to agriculture. Discov. Innov. 1995, 7, 337–348. [Google Scholar]
  54. Atehnkeng, J.; Ojiambo, P.S.; Donner, M.; Ikotun, T.; Sikora, R.A.; Cotty, P.J.; Bandyopadhyay, R. Distribution and toxigenicity of Aspergillus species isolated from maize kernels from three agro-ecological zones in Nigeria. Int. J. Food Microbiol. 2008, 122, 74–84. [Google Scholar] [CrossRef]
  55. Aminu, Z.; Jaiyeoba, I.A. An assessment of soil degradation in Zaria area, Kaduna State, Nigeria. Ife Res. Publ. Geogr. 2015, 13, 27–37. [Google Scholar]
  56. Omotosho, J.B.; Agele, S.O.; Balogun, I.A.; Adefisan, E.A. Climate variability, crop-climate modelling and water ecophysiology research: Implications for plant’s capacities for stress acclimation, yield production and food security. Glob. J. Plant Ecophysiol. 2013, 3, 56–69. [Google Scholar]
  57. Ayanlade, A.; Adeoye, N.O.; Babatimehin, O. Intra-annual climate variability and malaria transmission in Nigeria. Bull. Geogr. Socio-Econ. Ser. 2013, 21, 7–19. [Google Scholar] [CrossRef] [Green Version]
  58. FAO/UNESCO. FAO–UNESCO Soil Map of the World; UNESCO: Paris, France, 1974; Volume 1. [Google Scholar]
  59. FAO. Base Reference Base for Soil Resources. A Framework for International Classification, Correlation and Communication; World Soil Resources Reports; FAO: Rome, Grace, 2006; Volume 103, p. 145. [Google Scholar]
  60. Heanes, D. Determination of total organic-C in soils by an improved chromic acid digestion and spectrophotometric procedure. Commun. Soil Sci. Plant Anal. 1984, 15, 1191–1213. [Google Scholar] [CrossRef]
  61. Bremmer, J.M. Nitrogen-Total. In Method of Soil Analysis Part 3—Chemical Methods; Sparks, D.L., Ed.; SSSA Book Series 5; SSSA: Madison, WI, USA, 1996; pp. 1085–1122. [Google Scholar]
  62. Gee, G.W.; Or, D. 2.4 Particle-Size Analysis. In SSSA Book Series; Dane, J.H., Topp, C.G., Eds.; Soil Science Society of America: Madison, WI, USA, 2002; pp. 255–293. [Google Scholar] [CrossRef]
  63. Bray, R.H.; Kurtz, L.T. Determination of total, organic, and available forms of phosphorus in soils. Soil Sci. 1945, 59, 39–45. [Google Scholar] [CrossRef]
  64. Mehlich, A. Mehlich 3 soil test extractant: A modification of Mehlich 2 extractant. Commun. Soil Sci. Plant Anal. 1984, 15, 1409–1416. [Google Scholar] [CrossRef]
  65. Beah, A.; Kamara, A.Y.; Jibrin, J.M.; Akinseye, F.M.; Tofa, A.I.; Adam, A.M. Simulating the Response of Drought–Tolerant Maize Varieties to Nitrogen Application in Contrasting Environments in the Nigeria Savannas Using the APSIM Model. Agronomy 2021, 11, 76. [Google Scholar] [CrossRef]
  66. Hammer, G.L.; Muchow, R.C. Quantifying climatic risk to sorghum in Australia’s semi-arid tropics and subtropics: Model development and simulation. In Climatic Risk in Crop Production: Models and Management for the Semi-arid Tropics and Subtropics; eds Muchow, R.C., Mellamy, J.A., Eds.; CAB International: Nosworthy Way, Wallingford OX10 8DE, UK, 1991; Chapter 16; pp. 205–232. [Google Scholar]
  67. Zhang, Y.; Feng, L.; Wang, E.; Wang, J.; Li, B. Evaluation of the APSIM-Wheat model in terms of different cultivars, management regimes and environmental conditions. Can. J. Plant Sci. 2012, 92, 937–949. [Google Scholar] [CrossRef] [Green Version]
  68. Craufurd, P.Q.; Mahalakshmi, V.; Bidinger, F.; Mukuru, S.; Chantereau, J.; Omanga, P.; Qi, A.; Roberts, E.; Ellis, R.; Summerfield, R.; et al. Adaptation of sorghum: Characterization of genotypic flowering responses to temperature and photoperiod. Theor. Appl. Genet. 1999, 99, 900–911. [Google Scholar] [CrossRef] [Green Version]
  69. Slafer, G.A.; Abeledo, L.G.; Miralles, D.J.; Gonzalez, F.G.; Whitechurch, E.M. Photoperiod sensitivity during stem elongation phase as an avenue to rise potential yield in wheat. Euphytica 2001, 119, 191–197. [Google Scholar] [CrossRef]
Figure 1. (a) Observed vs. simulated grain yield using experiment conducted in 2016–2018 growing seasons for cultivar ranges from early to late maturing. ICSV-400 (MBE = 103 kgha−1; RMSE = 617 kgha−1, RMSEn = 28.7%); Improved Deko (MBE = 114 kgha−1, RMSE = 370 kgha−1, RMSEn = 18.7%); Samsorg-44 (MBE = 279 kgha−1; RMSE = 377 kgha−1, RMSEn = 17.2%); CSR01 (MBE = −48 kgha−1, RMSE = 301 kgha−1, RMSEn = 13.8%); SK5912 (MBE = 234 kgha−1; RMSE = 254 kgha−1, RMSEn = 9.2%). (b) Observed vs. simulated total biomass using experiment conducted in 2016–2018 growing seasons for cultivar ranges from early to late maturing. ICSV-400 (MBE = 28 kgha−1, RMSE = 1249 kgha−1, RMSEn = 19.5%); Improved Deko (MBE = 2344 kgha−1, RMSE = 2621 kgha−1, RMSEn = 36.8%); Samsorg-44 (MBE = −1100 kgha−1; RMSE = 1432 kgha−1, RMSEn = 12.5%); CSR01 (MBE = −976 kgha−1, RMSE = 1687 kgha−1, RMSEn = 16.5%);SK5912 (MBE = −429 kgha−1; RMSE = 868 kgha−1, RMSEn = 6.9%).
Figure 1. (a) Observed vs. simulated grain yield using experiment conducted in 2016–2018 growing seasons for cultivar ranges from early to late maturing. ICSV-400 (MBE = 103 kgha−1; RMSE = 617 kgha−1, RMSEn = 28.7%); Improved Deko (MBE = 114 kgha−1, RMSE = 370 kgha−1, RMSEn = 18.7%); Samsorg-44 (MBE = 279 kgha−1; RMSE = 377 kgha−1, RMSEn = 17.2%); CSR01 (MBE = −48 kgha−1, RMSE = 301 kgha−1, RMSEn = 13.8%); SK5912 (MBE = 234 kgha−1; RMSE = 254 kgha−1, RMSEn = 9.2%). (b) Observed vs. simulated total biomass using experiment conducted in 2016–2018 growing seasons for cultivar ranges from early to late maturing. ICSV-400 (MBE = 28 kgha−1, RMSE = 1249 kgha−1, RMSEn = 19.5%); Improved Deko (MBE = 2344 kgha−1, RMSE = 2621 kgha−1, RMSEn = 36.8%); Samsorg-44 (MBE = −1100 kgha−1; RMSE = 1432 kgha−1, RMSEn = 12.5%); CSR01 (MBE = −976 kgha−1, RMSE = 1687 kgha−1, RMSEn = 16.5%);SK5912 (MBE = −429 kgha−1; RMSE = 868 kgha−1, RMSEn = 6.9%).
Agronomy 13 00727 g001
Figure 2. Yield (observed and simulated) using on-farm datasets from the 2013–2017 growing seasons from contrasting environments for five (5) sorghum cultivars ranged from early to late maturing. ICSV-400 (N = 1192; MBE = 535 kgha−1; RMSE = 971 kgha−1, CV = 13.8%); Improved Deko (N = 300; MBE = 960 kgha−1, RMSE = 1169 kgha−1, CV = 12.3%); Samsorg-44 (N = 100; MBE = 102 kgha−1; RMSE = 655 kgha−1, CV = 8.9%); CSR01 (N = 944; MBE = −228 kgha−1, RMSE = 755 kgha−1, CV = 25.5%); SK5912 (N = 731; MBE = −241 kgha−1; RMSE = 879 kgha−1, CV = 18.4%). Coefficient of variations (CV), N = number of observations.
Figure 2. Yield (observed and simulated) using on-farm datasets from the 2013–2017 growing seasons from contrasting environments for five (5) sorghum cultivars ranged from early to late maturing. ICSV-400 (N = 1192; MBE = 535 kgha−1; RMSE = 971 kgha−1, CV = 13.8%); Improved Deko (N = 300; MBE = 960 kgha−1, RMSE = 1169 kgha−1, CV = 12.3%); Samsorg-44 (N = 100; MBE = 102 kgha−1; RMSE = 655 kgha−1, CV = 8.9%); CSR01 (N = 944; MBE = −228 kgha−1, RMSE = 755 kgha−1, CV = 25.5%); SK5912 (N = 731; MBE = −241 kgha−1; RMSE = 879 kgha−1, CV = 18.4%). Coefficient of variations (CV), N = number of observations.
Agronomy 13 00727 g002
Figure 3. Average monthly variation of (a) maximum temperatures and (b) minimum temperatures between 1985 and 2010 across the simulation sites in Adamawa State. The coefficients of variation (CV) ranged from 3.0 to 3.7% for maximum temperature and 2.0 to 2.3% for minimum temperature.
Figure 3. Average monthly variation of (a) maximum temperatures and (b) minimum temperatures between 1985 and 2010 across the simulation sites in Adamawa State. The coefficients of variation (CV) ranged from 3.0 to 3.7% for maximum temperature and 2.0 to 2.3% for minimum temperature.
Agronomy 13 00727 g003
Figure 4. Average monthly variations of (a) maximum temperatures and (b) minimum temperatures between 1985 and 2010 across the simulation sites in Borno State. The coefficients of variation (CV) ranged from 3.0 to 3.7% for maximum temperature and 2.0 to 2.3% for minimum temperature.
Figure 4. Average monthly variations of (a) maximum temperatures and (b) minimum temperatures between 1985 and 2010 across the simulation sites in Borno State. The coefficients of variation (CV) ranged from 3.0 to 3.7% for maximum temperature and 2.0 to 2.3% for minimum temperature.
Agronomy 13 00727 g004
Table 1. Genetic coefficients of sorghum cultivars calibrated in the APSIM-sorghum model.
Table 1. Genetic coefficients of sorghum cultivars calibrated in the APSIM-sorghum model.
Description of ParameterUnitICSV400Impr. DekoCSR01Samsorg-44SK5912Calibration Method (A/B)
Thermal time from emergence to end of juvenile°C days180210100100100A
Thermal time from end of juvenile to floral initiation°C days160100100100120A
Photoperiod slope°C/hour150200500550600A
Thermal time from flag leaf to flowering°C days170170100100150A
Thermal time from flowering to start of grain filling°C days8080808080B
Thermal time from flowering to maturity°C days560560460500450A
Leaf appearance rate (leaf app rate 1)°C d/leaf4156565656[31]
Leaf appearance rate (leaf app rate 2)°C d/leaf2028282828[31]
Radiation use efficiency(RUE)g/MJ1.251.251.351.351.65A
Head grain number determinationg/grain0.000830.00880.000830.000830.0088A
Maximum grain filling (MaxGFrate)mg/grain/day0.090.030.050.050.09A
A: Manual tuning of parameter values; B: Model defaults values; [31] means the parameter calibrated based on the value reported.
Table 2. Summary of the selected sites for model application of sorghum cultivars under varying planting windows.
Table 2. Summary of the selected sites for model application of sorghum cultivars under varying planting windows.
S/NoStateLGASiteAEZLongitude (N)Latitude (E)
1AdamawaHongDulmavaSS12.982410.3014
2GombiGuyakuSS12.663410.3459
3DemsaMbula KuliNGS12.30169.45745
4GireiWuroshiNGS12.61649.46866
5GireiDaneyelNGS12.5149.54761
6GombiTawaNGS12.685610.1691
7GuyukChikilaNGS11.97199.77237
8GuyukLakumnaNGS11.98979.92083
9HongHushere ZumNGS13.080710.1038
10NumanBareNGS12.11089.5843
11NumanKikan_KodomtiNGS11.98789.46081
12ShellengJonkolo-Lama NGS12.1789.89965
13ShellengLakati-Libbo/NGS12.25029.69541
14SongSabon GariNGS12.59359.84049
15SongSuktuNGS12.42489.63746
16DemsaNassarawo DemsaSGS12.15019.29625
17Yola NorthYelwa-JamboreSGS12.50469.26165
18Yola SouthFufureSGS12.65049.1736
19BornoBayoBalbayaSS11.764810.5848
20BayoBriyelSS11.649710.371
21BayoJara-DaliSS11.731610.2759
22BiuBurataiSS12.415810.7675
23BiuKaburaSS12.265310.7392
24BiuMathauSS12.109710.7214
25BiuTumSS12.488110.8228
26HawulKwajaffaSS12.483110.5167
27HawulPuba VidauSS12.187910.5224
28HawulSakwa HemaSS12.389410.3867
29KwayakusarKurbo GayiSS11.957510.384
30ShaniLakundumSS12.050610.0556
31ShaniGwaskaraNGS12.15810.2271
32KwayakusarBila GusiNGS12.047610.5192
33ShaniKuboNGS12.085310.14
LGA—Local Government Area, AEZ—Agro-ecological zone; SS—Sudan Savannah, NGS—Northern Guinea savannah, SGS—Southern Guinea savannah.
Table 3. Physical and chemical properties used for model applications in Adamawa State.
Table 3. Physical and chemical properties used for model applications in Adamawa State.
Profile DepthBDOCSandSiltClaypHNMeh. PK
Site(cm)(g/cm3)(%)(%)(%)(%)(in H2O)(%)(ppm)cmol/kg
Mbula-Kuli 0–2001.760.845923187.80.0632.10.5
Demsa-Nassarawo24–1802.180.666515208.30.063.80.89
Daneyel 31–2001.760.22817127.00.0110.90.3
Woroshi 14–942.160.546519166.40.041.170.36
Guyaku19–1201.70.35799126.60.032.140.22
Tawa15–1271.790.627513126.70.053.380.21
Chikila 30–1802.180.901519668.50.082.550.13
Lakumna 20–2001.770.902523527.30.101.590.65
Dulmava 27–2011.820.516715187.50.061.030.17
Hushere-Zum 41–2051.930.46808126.30.032.410.40
Bare25–2001.620.35749176.60.024.070.20
Kikan_Kodomti 22–2001.760.66719207.30.0413.70.20
Lakati-Libbo 27–2001.830.30789137.40.015.040.20
Jonkolo-Lama 15–2002.060.337810125.90.020.890.14
Sabon-Gari 31–2001.730.662533426.20.041.450.4
Suktu35–2102.080.227111186.30.036.560.20
Yelwa-Jambore 24–1552.190.47711126.50.031.80.09
Fufure 20–1451.980.546517188.00.0232.10.10
BD = bulk density, OC = organic carbon content, N = percent Nitrogen, Meh P = Available Phosphorus, and K = potassium.
Table 4. Physical and chemical properties used for model applications in Borno State.
Table 4. Physical and chemical properties used for model applications in Borno State.
Profile DepthBDOCSandSiltClaypHNMeh. PK
Site(cm)(g/cm3)(%)(%)(%)(%)(H2O)(%)(ppm)cmol/kg
Balbaya 9–2001.590.29837106.10.011.030.0
Briyel15–2001.320.391929528.40.022.690.4
Jara-Dali 8–2001.550.335113366.60.021.720.3
Buratai29–1501.630.17748187.60.022.690.6
Kwaya Bura22–1011.360.783638267.10.060.899.0
Mathau 12.0–941.620.12900107.40.012.830.8
Tum 12–2001.400.192824487.40.011.170.6
Kwajaffa 30–1101.310.541627577.40.062.280.7
Puba Vidau 10–2001.320.41819638.30.020.890.6
Sakwa Hema 15–1701.570.52749177.00.040.760.1
Bila Gusi 80–2001.590.486715186.50.022.140.1
Kurba Gayi 10–2001.600.32759167.20.011.030.1
Gwaskara 19–2001.570.347213157.10.0111.50.1
Kubo 33–2001.54o.466413237.30.021.310.8
Lakundum 16–2001.520.737210187.30.0713.69.0
BD = bulk density, OC = organic carbon content, N = percent Nitrogen, Meh P = Available Phosphorus and K = potassium.
Table 5. Statistical evaluation of simulated phenology and morphological traits (LAI and total leaf number/plant) of contrasted sorghum cultivars calibrated from experiment conducted under optimum conditions in Southern Guinea and Sudan Savannah AEZs.
Table 5. Statistical evaluation of simulated phenology and morphological traits (LAI and total leaf number/plant) of contrasted sorghum cultivars calibrated from experiment conducted under optimum conditions in Southern Guinea and Sudan Savannah AEZs.
Parameters/
Cultivar
UnitNMBERMSEObserved RangeObserved Mean
Absolute Value% of Mean Observed
ICSV-400
50% FloweringDAP11−145.462–7568
Physiological MaturityDAP11254.690–10697
LAI-maxm2/m211−0.20.832.41.8–3.02.3
Leaf number 43.43.520.516–1817
Improved Deko
50% FloweringDAP7−467.975 9584
Physiological MaturityDAP7165.0101–122110
LAI-maxm2/m270.60.827.02.0–3.32.5
Leaf number 40.41.26.416–1918
Samsorg-44
50% FloweringDAP4133.085–11499
Physiological MaturityDAP4243.2112–140126
LAI-maxm2/m240.20.726.62.2–3.43.0
Leaf number 45.15.226.219–2320
CSR01
50% FloweringDAP8288.484–11295
Physiological MaturityDAP8176.1111–139123
LAI-maxm2/m280.30.414.72.3–3.73.0
Leaf number 844.119.519–2421
SK5912
50% FloweringDAP4454.495–122108
Physiological MaturityDAP4243.0122–149135
LAI-maxm2/m240.30.620.72.0–3.32.5
Leaf number 43.84.017.620.4–25.423
N—Number of observations; LAI-max: maximum leaf area index measured during growth; MBE = positive implies over-simulated mean observed; negative implies under-simulated the mean observed value.
Table 6. Statistical indices for model validation of contrasted sorghum cultivars across planting date and cropping system from on-farm production technology between 2013 and 2017.
Table 6. Statistical indices for model validation of contrasted sorghum cultivars across planting date and cropping system from on-farm production technology between 2013 and 2017.
Sowing Month/CultivarCropping SystemNSimulatedObservedMBERMSE
ICSV400 kgha−1
JuneSole535220114207811038
JulyIntercropping3720072084−77700
Mixed cropping2716982646−9481229
Sole46120521488564942
AugIntercropping1317782754−9771029
Mixed cropping1318502663−814959
Sole10818971537360936
Improved Deko
JuneSole17816561426231712
JulySole1111554148866617
AugSole111492943548598
SamSorg-44
JuneSole2214631697−234808
JulySole501586962624915
Intercropping11910750160161
AugSole1216231380244738
CSR01
JuneIntercropping1315732188−615624
Mixed cropping1815241729−206640
Sole45213351366−31726
JulyIntercropping2315171700−183700
Mixed cropping1312971973−6761203
Sole35615661886−320952
AugMixed cropping1515881388200258
Sole5514741932−458940
SK5912
JuneIntercropping2614331305128873
Mixed cropping1711571184−26834
Sole2631437142413848
JulyIntercropping1111471576−429873
Mixed cropping2211692225−10561285
Sole32015871408179764
AugIntercropping1013231858−535824
Mixed cropping811351603−744809
Sole5517861483303800
N—Number of observations/farms.
Table 7. Analysis of mean monthly, seasonal rainfall (mm) and level of variability across the simulation sites in Adamawa State (1985–2010).
Table 7. Analysis of mean monthly, seasonal rainfall (mm) and level of variability across the simulation sites in Adamawa State (1985–2010).
SiteMayJun.Jul.Aug.Sep.Oct.SeasonalStdevC.V (%)
Demsa-Nassarawo 102.1121.2189.3234.3172.773.589318821
Mbula Kuli 95.9115.7186.5225.8168.158.685118121
Daneyel99.8118.1202.9240.5156.954.487319122
Woroshi 103.3126.4216.5244.0156.455.790219121
Guyaku117.9155.9228.9308.8176.699.1108723021
Tawa 134.2149.6237.1293.3192.497.2110423922
Lakumna 91.8110.3167.5258.2174.968.987218521
Chikila 98.5106.5178.4249.7165.267.886618621
Hushere Zum120133.8211.7266.5196.7113104224123
Dulmava109.9150.6225.5302.8202.2113.1110424722
Bare91.9107.4176.9244.2162.980.686419422
Kodomti 91.1109.5176.8243.2170.275.086619422
Lakati-Libbo95.2109.6186.8250.2155.274.987219122
Jonkolo-Lama97.6115.0182.4268.6166.273.190319722
Sabon-Gari 99.8119.5211.3269.7181.882.196421222
Suktu99.6116.3211.4256.5157.861.590319922
Yelwa-Jambore 102.1125.4206.6218163.552.286818922
Fufure 103.8140.6220.6218.5160.551.489519021
Seasonal—average total seasonal rainfall from May to Oct.; Stdev—Standard deviation from mean; CV—coefficient of variations (in percentage).
Table 8. Analysis of mean monthly, seasonal rainfall (mm) and level of variability across the simulation sites in Borno State from 1985 to 2010.
Table 8. Analysis of mean monthly, seasonal rainfall (mm) and level of variability across the simulation sites in Borno State from 1985 to 2010.
SiteMayJun.Jul.Aug.Sep.Oct.SeasonalStdevC.V (%)
Balbaya 87.9141.3202.9287.9167.467.495520622
Briyel 93.2129.0174.2242.7182.761.188318221
Jara-Dali 78.4136.8202.8289.0204.480.399221721
Kabura 72.5142.4209.7316.1149.348.493918820
Mathau 78.3144.4204.4312.1165.651.995717418
Tum86.2149.8218.1317.4170.056.999820420
Buratai 77.4144.3210.9318.4148.545.694519120
Kwajaffa99.7142.3204.3306.7179.351.298318619
Puba Vidau96.6144.2199.6299.8188.360.398919119
Sakwa Hema 93.3144.2206.9307.4176.860.298918619
Bila-Gusi98.9124.5190.6268.6183.475.794218920
Kurba Gayi 85.5145.9213.1303.1166.261.197519920
Gwaskara 83.5142.1198.5295.4201.674.999619219
Kubo 97.3121.6181.9262.2192.272.392718620
Lakundum85.2146.0220.2307.1158.277.999521320
Seasonal—average total seasonal rainfall from May to Oct.; Stdev—Standard deviation from mean; CV—coefficient of variations (in percentage).
Table 9. Mean simulated grain yield and Water Use Efficiency for grain yield (WUEgrain) of sorghum cultivars across different planting windows (PWs) and agro ecological zones.
Table 9. Mean simulated grain yield and Water Use Efficiency for grain yield (WUEgrain) of sorghum cultivars across different planting windows (PWs) and agro ecological zones.
PW/CNOGrain Yield WUEgrain
ICSV400Impr. DekoCSR01Samsorg-44SK5912ICSV400Impr. DekoCSR01Samsorg-44SK5912
Sudan Savanna (SS)kgha−1kgha−1 mm−1
PW1420232122112340209717037.36.65.04.63.2
PW2420230921702205198115807.36.44.64.23.0
PW3420225521481895176012526.96.24.03.72.4
PW4420222821451778161311286.86.23.83.52.3
Mean 227821682054186314167.16.34.44.02.7
Northern Guinea Savanna(NGS)
PW1480232322342750253623587.77.06.66.15.0
PW2480231521882677244721287.86.86.05.54.2
PW3480223621712657238618567.26.35.85.33.7
PW4480222321382644237516547.16.56.05.43.5
Mean 227421822682243619997.56.76.15.64.1
Southern Guinea Savanna(SGS)
PW190195918652192196717336.76.05.55.13.9
PW290193918152091187816556.75.95.14.73.6
PW390192018412106189814886.45.85.14.73.3
PW490190318142059185015306.45.84.94.53.3
Mean90193018342112189816026.65.95.24.73.5
Impr.—improved; PW—planting windows [16–31 May (PW1), 1–15 Jun (PW2), 16–30 Jun (PW3), 1–15 Jul (PW4)]; C—Cultivar; NO—Number of observations.
Table 10. Mean simulated grain yield (kgha−1) for evaluating adapted sorghum cultivars across sites and AEZs based on increased yield threshold.
Table 10. Mean simulated grain yield (kgha−1) for evaluating adapted sorghum cultivars across sites and AEZs based on increased yield threshold.
AEZN-SiteICSV400Impr. DekoCSR01Samsorg-44SK5912
Sudan Savanna (SS)Balbaya22362132146713681106
Briyel22082118198317841450
Buratai22902130255524261926
Dulmava23632301224920011439
Guyaku21722069137112961110
Jara-Dali20922052170815451242
Kabura26362480370733242847
Kurbo Gayi26732509364532472445
Kwajaffa2328222116241448919
Lakundum22762151261222891621
Mathau2013188610671029730
Puba Vidau215720261022939783
Sakwa Hema22922180190217231112
Tum21572098184716531096
Mean22782168205418631420
CV(%)109464456
Northern Guinea Savanna (NGS)Bare1926180312691174940
Bila Gusi21262044232520541580
Chikila24022331315229142201
Daneyel20281945180716151247
Gwaskara24982372295725921789
Hushere Zum21802079177715991213
Jonkolo—Lama 22082141265725642149
Kikan_Kodomti20601992230120241557
Kubo24062336376134323140
Lakati-Libbo21232057195417431411
Lakumna24142344319932982790
Mbula Kuli22372176268523662120
Sabon Gari24952360338729992680
Suktu23142221287425382052
Tawa24582328312727532294
Wuroshi25122390367633102824
Mean22742182268224361999
CV(%)1111303041
Southern Guinea Savanna (SGS)Fufure130611651010971891
Nassarawo Demsa23302220304927072325
Yelwa-Jambore21542117227620171589
Mean19301834211218981602
CV(%)2427424043
Impr.—improved; CV(%)—Coefficients of variations in the percentage.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Akinseye, F.M.; Ajeigbe, H.A.; Kamara, A.Y.; Omotayo, A.O.; Tofa, A.I.; Whitbread, A.M. Establishing Optimal Planting Windows for Contrasting Sorghum Cultivars across Diverse Agro-Ecologies of North-Eastern Nigeria: A Modelling Approach. Agronomy 2023, 13, 727. https://doi.org/10.3390/agronomy13030727

AMA Style

Akinseye FM, Ajeigbe HA, Kamara AY, Omotayo AO, Tofa AI, Whitbread AM. Establishing Optimal Planting Windows for Contrasting Sorghum Cultivars across Diverse Agro-Ecologies of North-Eastern Nigeria: A Modelling Approach. Agronomy. 2023; 13(3):727. https://doi.org/10.3390/agronomy13030727

Chicago/Turabian Style

Akinseye, Folorunso M., Hakeem A. Ajeigbe, Alpha Y. Kamara, Akinrotimi O. Omotayo, Abdullahi I. Tofa, and Anthony M. Whitbread. 2023. "Establishing Optimal Planting Windows for Contrasting Sorghum Cultivars across Diverse Agro-Ecologies of North-Eastern Nigeria: A Modelling Approach" Agronomy 13, no. 3: 727. https://doi.org/10.3390/agronomy13030727

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop