Influences of Meteorological Factors on Maize and Sorghum Yield in Togo, West Africa

Abstract

This paper explores the effect of meteorological factors such as rainfall, temperature, sunshine, wind speed, and relative humidity on the yield of maize (Zea mays L.) and sorghum (Sorghum bicolor L.) at different growth stages in Togo’s Plateau, Central, and Savannah regions. For this purpose, data from 1990 to 2019 on weather variables and maize and sorghum yields were used. The study applied Fisher’s meteorological regression and Chebyshev polynomial function. Our findings revealed that rainfall had a more beneficial than detrimental effect on maize and sorghum yield across stages and regions. Contrariwise, temperature influence was as beneficial as detrimental and more significant across all growth stages of maize and sorghum in the Savannah and Plateau regions. Furthermore, the sunshine effect on maize yield was more significant in the Central and Savannah regions, while negative on sorghum yield in all the growth stages in the Central region. Similarly, the wind speed was also beneficial and detrimental to maize and sorghum yields, although it was more significant for sorghum in Plateau and Savannah regions. Lastly, relative air humidity positively and negatively influenced maize and sorghum yields in all the growth stages and regions for maize and the Plateau and Savannah regions for sorghum. Therefore, there is a need for real-time agricultural meteorological information to help farmers plan crop production more efficiently and increase crop yield.

1. Introduction

In this era of climate change dispensation, achieving increased crop yield in agriculture is quite challenging. Assessment of crop yield data of any country in the world is crucial for investigating its food security, grain import and export, and land management, which are key indicators of the agricultural economy [1]. Crop yield is influenced by numerous factors, including genetics, agronomic management, and environmental conditions [2,3]. Crop performance from the early stages of growth to the predicted outputs is endangered mainly by changes in environmental factors such as rainfall, temperature, sunshine [4,5], wind speed, relative humidity, and evapotranspiration [6], thus leading to food insecurity around the world [7]. Boosting food security in the world often hinges on the production of cereals, including rice, wheat, maize, and sorghum which are considered a staple food in most countries of the world. Providing evidence that cereals are essential for humans, global cereals utilization is estimated at an annual value of 2.7 billion tons in 2021 and could exceed 3 billion tons by 2030 [8].

In Togo, like other African countries, cereal production and consumption play a massive role in the population’s daily calorie intake; thus, it is paramount to the country’s food security. According to the Togolese agricultural data center (DSID), maize and sorghum are Togo’s first and second most produced and consumed cereals. In 2021, maize and sorghum production was estimated at 885,030 tons and 279,105 tons, respectively [9]. Nevertheless, Togo’s cereal productivity is still very low (1.2 tons per hectare) compared to 5.5 tons per hectare and 7.2 tons per hectare in European Union and Northern America, respectively [8]. Weather fluctuations could partially explain this gap in the yield.

Previous studies have demonstrated the association between the cumulative effects of climatic conditions (rainfall, temperature, and CO₂ emission) on agricultural yield and food security [10]. A study reported a strong link between maximum and minimum temperatures and rice yield in China [11]. Another study also found that climate-related variables (temperature, solar radiation, humidity, and rainfall) significantly impact corn yield [12]. Furthermore, relative humidity, wind speed, evaporation, and sunshine duration have been reported to influence crop (wheat, maize, and rice) yield in China [6]. Research conducted in the Ashanti Region of Ghana showed that rainfall, temperature, and relative humidity affect cassava, yam, and maize yield [13]. A similar study in Togo by Boansi [14] reported that cassava yield is affected by both normal climate variables and within-season rainfall variability. Additionally, Ali [15] concluded that maize is the most affected by the inter-seasonal and intra-seasonal variability of temperature and precipitation compared to sorghum and rice in Northern Togo. Likewise, some researchers reported that rainfall and temperature variability affect crop production and yield in Southern and Northern Togo, even though the effects vary across crops [16,17]. Metrological variables influence overall yield, and the total output reflects changes during various crop growth stages. For example, the weather positively and negatively affected rice and wheat yield [4,18], according to similar research in China. Furthermore, Kulyakwave et al. [5] reported that rainfall, the intensity of sunshine, and minimum and maximum temperature significantly impact rice yield during growth stages in the Mbeya region of Tanzania.

According to the United Nations Development (UNDP), Togo can be divided into three vulnerable zones to climate change impact. The first zone comprises the Maritime and Plateau regions; zone 2 includes the Central and Kara regions, and zone 3 is the Savannah region. It was predicted that zones 1 and 2 will be sensitive to a decline in rainfall at the 2025 horizon, while zone 3 will potentially witness a temperature rise [19]. Currently, the country is experiencing the above-predicted climate change effects, with temperatures reaching 40 °C locally. In addition, the country witnesses strong winds, which can reach a speed of 100/115 Km per hour, causing crop lodging. These adverse effects cause a shift in crop growing seasons, thus jeopardizing crop yields [20]. Therefore, this study aims to assess the effect of meteorological factors such as rainfall, average temperature, sunshine, and additional meteorological factors, including wind and humidity, on the growth stages and yield of maize and sorghum in the Plateau, Central, and Savannah regions of Togo. This study contributes to the literature in two ways. First, our findings emphasize the need to include a broader range of meteorological factors when evaluating the effects of climate change on crop yields by growth stage. Our results will be conducive to future studies in improving estimates of the effects of climate change on economic well-being as the literature on this topic continues to increase fast. Second, this research and data from meteorological sensors may allow real-time agricultural yield forecasting.

2. Materials and Methods

2.1. Study Area

Located between 6° N and 11° N latitudes and 0° E and 2° E longitudes and with a surface area of 56,600 km², Togo is a small West African country. The country is divided into five administrative regions: the Savannah and the Kara regions in the north, the Central region, and the Plateau and Maritime regions in the south. Togo has a tropical climate, which is characterized as a part of the hot and humid inter-tropics marked by two main wind flows: the monsoon, which blows from the southwest and is associated with the rainy season, and the airstream harmattan from the northeast, which brings with it cool, dry weather between November and March; periodic droughts occur in the north while floods are more recurrent in the south part of the country [21].

The north–south configuration of the country comes with a diversity of climates, apportioning the country into three main climatic zones. The Sub-Equatorial Region extends from the coast to the 8° north latitude and represents the southern regions (Maritime and Plateau). The rainfall ranges from 800–1400 mm, divided into two seasons: from mid-March to late July and from early September to early mid-November. The average annual temperature is 27 °C. The relative humidity is high and fluctuates around 90%. The Guinea–Sudan region lies between 8° and 10° north in latitude and is equivalent to the Central and Kara regions. It is a climatic transition zone where annual rainfall ranges from 1400 mm to 1500 mm, falling in a single rainy season between June and October. Annual temperatures average is 26.5 °C (15–37 °C), and relative humidity ranges between 60% and 80%. The Sudan region lies between the 10° and 11° north in latitude and represents the Savannah region. It is semi-arid and has the lowest rainfall, ranging from 900–1100 mm. The region experiences a single rainfall regime which expands from May to October. Temperatures vary from 17–41 °C in the dry season. The humidity fluctuates around 60% [21].

The study is conducted on one major maize and sorghum-producing region from each of the three climatic zones mentioned above. The Plateau region represents zone 1, whereas the Central and Savannah regions represent zones 2 and 3, respectively. In 2020, these three regions contributed up to 77.22% and 76.68% to maize and sorghum production output, respectively. The annual national maize production was 885.03 tons, with a yield of 1.23 tons per hectare, while sorghum production amounted to 279.11 tons for a productivity of 0.9 tons per hectare [9]. In the Plateau region, maize production reached 438.83 tons with an average yield of 1.41 tons per hectare, while sorghum output was estimated at 75.04 tons (1.0141 tons per hectare). The Central region recorded 140.79 tons of maize production with a total productivity of 1.54 tons per hectare. The total sorghum output was 54.11 tons (0.8 tons per hectare yield). In the Savannah region, maize production was 103.77 tons, with 1.27 tons per hectare yield. As for sorghum, the total production was 84.88 tons, with a yield of 1.09 tons per hectare [9].

Maize and sorghum are similar in their genetic organization, plant morphology, developmental physiology, and uses [22]. Both crops are C4 plants (fixes carbon dioxide into four-carbon acids) and require less water than C3 plants (rice, soybean, wheat). However, sorghum is more tolerant to water stress than maize and has a competitive edge over other rainfed crops such as maize and rice [23,24]. They thrive better in tropical climates such as Togo’s.

2.2. Data Sources

The study used secondary time series data from 1990 to 2019. Maize and sorghum yield data on the three studied regions (Plateau, Central, and Savannah) were acquired from the Togolese agricultural data center (DSID). The meteorological data, including average monthly rainfall (mm), temperature (°C), sunshine (hours), average monthly wind speed (m/s), and relative humidity (%) for 30 years were obtained from the Togolese meteorological center. The spatial distribution of the selected meteorological stations is represented in Figure 1. Figure 2 represents maize and sorghum yield across the Plateau, Central, and Savannah regions.

The study considered five growth stages for maize using phenological data from previous studies [25,26,27,28]: emergence stage (Stage 1), jointing stage (Stage 2), tasseling stage (Stage 3), milk stage (Stage 4), and physiological maturity stage (Stage 5). Sorghum growth stages are also five: the emergence stage (Stage 1), booting stage (Stage 2), flowering stage (Stage 3), grain filling stage (Stage 4), and physiological maturity stage (Stage 5) [29].

Table 1. Growth stages of maize/sorghum in the Plateau, Central, and Savannah regions.

Region	Growth Stage
	Emergence	Jointing (Booting)	Tasselling (Flowering)	Milk (Grain-Filling)	Physiological Maturity
Maize
Plateau	March	April	May	June	July
Central	June	July	August	September	October
Savannah	May	June	July	August	September
Sorghum
Plateau	May	June	July	August	September
Central	June	July	August	September	October
Savannah	May	June	July	August	September

summarizes the different growing stages of maize and sorghum in the selected regions of Togo. We used monthly average data to represent each growth stage because daily data were unavailable for most meteorological variables, and planting dates were not uniform across each study region.

2.3. Research Method

This study used econometric methods such as simple regression to calculate the trend yield, while multiple regression is employed to assess the impact of the meteorological factors during the whole growth period on the maize and sorghum yield. Furthermore, the meteorological function is modeled using Fisher Integral Regression Model, and lastly, Chebyshev polynomial function and stepwise regression served in the computation of yield coefficients for weather-related factors on maize and sorghum yields by growth stage.

2.3.1. Actual, Trend, and Meteorological Yield Model

The study used the traditional classical production function with maize yield (Y_M) and sorghum yield (Y_S) as dependent variables. (Y_M/Y_S) is a function of various independent variables [30].

Crop yield is influenced by a variety of factors, including environmental factors (rainfall, temperature, sunshine, wind speed, and relative humidity) and non-environmental factors (genetics, agronomic management, technology level, and pesticide and chemical fertilizer use [2].

The actual yield is defined as the sum of trend and meteorological-related yield (Equation (1)).

$$Y=Y_T+Y_W+\epsilon$$

(1)

where $Y$ is the actual maize/sorghum yield, $Y_T$ is the trend yield, $Y_W$ represents the weather yield, and $\epsilon$ is an error term.

The yield trend is a long-term trend due to changes in genetics, agronomic management, technology level, and pesticide and chemical fertilizer use (Equation (2)) [4,5].

To obtain the trend yield data, we first run a simple linear regression with maize/sorghum yield as the dependent variable and the time variable (year) as the explanatory variable for each study region. The regression results were then used to predict the trend yield ($Y_T$) (Equation (3)).

$$Y_T=f\left(t\right)$$

(2)

$${\widehat{Y}}_T=\alpha +\beta Year$$

(3)

where f (t) shows the function of a particular year t, ${\widehat{Y}}_T$ is the predicted value of the trend yield, $\alpha$ represents the constant term, and $\mathsf{\beta }$ is the year coefficient.

The meteorological yield is the short-term fluctuation induced by natural factors [4,5]. It is represented by Equation (4) below:

$$Y_M={{\displaystyle \sum }}_{q=0}^i{{\displaystyle \sum }}_{p=0}^{j}f \left(W_{p,q}\right)$$

(4)

q is the growth stages of the maize or sorghum plant, p is the weather factors. $W_{p,q}$ is the weather variable by growth stage, and f($W_{p,q}$) is the function of weather variables and individual yield variations.

After obtaining the trend yield data, the meteorological yield was deducted from Equation (1):

$${\widehat{Y}}_M=\widehat{Y}-{\widehat{Y}}_T$$

(5)

The standard yield is introduced to simplify the model calculation, and the following relationship in Equations (6)–(8) can be deduced:

$$Y_0=Y_{T_0}+Y_{W_0}+\epsilon$$

(6)

$$Y_{T_0}=f(t_0)$$

(7)

$$Y_{M_0}={{\displaystyle \sum }}_{q=0}^n{{\displaystyle \sum }}_{p=0}^{m}f \left({\overline{W}}_{p,q}\right)$$

(8)

$Y_0$ is the standard maize/sorghum yield, $Y_{T_0}$ denotes the base trend yield or the yield given by non-natural sources. $Y_{W_0}$ represents the weather yield, which results from the average weather variables.

2.3.2. Fisher Integral Regression Model and Chebyshev Orthogonal Polynomial Function

The Fisher Integral Regression Model [31] is a reliable statistical regression model for calculating the association between weather yields and weather parameters at various crop growth stages. It also gives options for determining the quantitative association between variables and their coefficients in the model. The rainfall, temperature, and sunshine model will further be addressed as model 1, and the rainfall, temperature, sunshine, wind speed, and relative humidity as model 2. For example, Equations (9a) (model 1) and (9b) (model 2) can be derived from Equation (4) as follows:

$${\widehat{Y}}_M={\mathsf{\alpha }}_0+{{\displaystyle \int }}_0^{\mathsf{\tau }}{\mathrm{a}}_{1\mathrm{j}}\left(\mathrm{t}\right){\mathrm{X}}_1\left(\mathrm{t}\right)\mathrm{dt} +{{\displaystyle \int }}_0^{\mathsf{\tau }}{\mathrm{a}}_{2\mathrm{j}}\left(\mathrm{t}\right){\mathrm{X}}_2\left(\mathrm{t}\right)\mathrm{dt}+{{\displaystyle \int }}_0^{\mathsf{\tau }}{\mathrm{a}}_{3\mathrm{j}}\left(\mathrm{t}\right){\mathrm{X}}_3\left(\mathrm{t}\right)\mathrm{dt}$$

(9a)

$$\begin{gathered} {\widehat{Y}}_M={\mathsf{\alpha }}_0+{{\displaystyle \int }}_0^{\mathsf{\tau }}{\mathrm{a}}_{1\mathrm{j}}\left(\mathrm{t}\right){\mathrm{X}}_1\left(\mathrm{t}\right)\mathrm{dt} +{{\displaystyle \int }}_0^{\mathsf{\tau }}{\mathrm{a}}_{2\mathrm{j}}\left(\mathrm{t}\right){\mathrm{X}}_2\left(\mathrm{t}\right)\mathrm{dt}+{{\displaystyle \int }}_0^{\mathsf{\tau }}{\mathrm{a}}_{3\mathrm{j}}\left(\mathrm{t}\right){\mathrm{X}}_3\left(\mathrm{t}\right)\mathrm{dt} \\ +{{\displaystyle \int }}_0^{\mathsf{\tau }}{\mathrm{a}}_{4\mathrm{j}}\left(\mathrm{t}\right){\mathrm{X}}_4\left(\mathrm{t}\right)\mathrm{dt} +{{\displaystyle \int }}_0^{\mathsf{\tau }}{\mathrm{a}}_{5\mathrm{j}}\left(\mathrm{t}\right){\mathrm{X}}_5\left(\mathrm{t}\right)\mathrm{dt} \end{gathered}$$

(9b)

where ${\widehat{Y}}_M$ denotes the meteorological yield; ${\mathrm{a}}_0$ is a constant. τ = growth stage of maize/sorghum from 0 as sowing stage with j^th independent period; ${\mathrm{a}}_1, {\mathrm{a}}_2,{\mathrm{a}}_3,{\mathrm{a}}_{4}and {\mathrm{a}}_5$ are the meteorological yield of independent variables [X1 = rainfall (RF), X2 = average temperature (T), X3 = sunshine (SS), X4 = wind speed (WS), and X5 = relative humidity (RH)]; ${\mathrm{X}}_n$ denotes the function of the independent variable in the model. t (year) is the maize/sorghum j^th growth durations) the independent variable in the model. From the emergence to the physiological maturity stage, the effect of each meteorological element was calculated using Equations (9a) and (9b).

The Chebyshev orthogonal polynomial function was then used to calculate regression coefficients in linear function form, as shown in Equation (10):

$$a_i\left(t\right)={{\displaystyle \sum }}_j{\alpha }_{ij}{\phi }_{ji}$$

(10)

whereby ${\phi }_{ji}$ is j = 1, 2, 3, …5. and $a_i$(t) is a regression coefficient of ${\mathrm{X}}_n$, as a functional form of maize/sorghum growth stage and time, which is approximated by Chebyshev polynomial as in Equation (11):

$${\widehat{Y}}_M={\alpha }_0+{{\displaystyle \sum }}_i{{\displaystyle \sum }}_j{\alpha }_{ij}{P}_{ji}$$

(11)

This resulted in Equation (12) for independent variable effect on yield:

$$P_{ji}={{\displaystyle \int }}_0^{\tau }{x}_i\left(t\right){\phi }_{j}dt$$

(12)

φ_j could be derived from the Chebyshev orthogonal polynomial Equation (13)

$${\mathsf{\phi }}_{\mathrm{k}+1}\left(\mathrm{x}\right)={\mathsf{\phi }}_1\left(\mathrm{x}\right){\mathsf{\phi }}_{\mathrm{k}}\left(\mathrm{x}\right)-\frac{{\mathrm{k}}^2\left({\mathrm{n}}^2-{\mathrm{k}}^2\right)}{4\left(4{\mathrm{k}}^2-1\right)}{\mathsf{\phi }}_{\mathrm{k}-1}\left(\mathrm{x}\right)$$

(13)

3. Results

3.1. Descriptive Analyses

Table 2. Descriptive statistics.

Crop	Obs	Mean	Min	Max	Variance	Std. Dev	Kurtosis	Skew
Plateau
Maize	30	1392.85	963.15	1962.42	59,143.35	243.19	3.48	0.57
Sorghum	30	1038.02	591.1	1433.92	55,500.29	235.59	2.28	−0.29
Central
Maize	30	1423.57	943.76	1812.07	54,246.48	237.29	2.20	−0.54
Sorghum	30	839.365	488.25	1538.94	38,443.11	196.07	7.49	1.90
Savannah
Maize	30	1218.99	606.16	1745.35	54,228.26	232.87	4.10	0.18
Sorghum	30	915.91	303.39	1435.34	99,545.95	315.51	1.94	−0.33

portrays the maize and sorghum yield statistics in the Plateau, Central, and Savannah regions from 1990 to 2019. Results indicate skewness values ranging between −2 and 2. However, some kurtosis values are higher than 3 [32]. These results imply that variables sorghum yield in the Plateau and Savannah and maize yield in the Central region follow a normal distribution, whereas variables maize yield in the Plateau and Savannah and sorghum yield in the Central are not normally distributed. Furthermore, autocorrelations of the meteorological variables were performed for each study area and can be found in Appendix A, Appendix B and Appendix C.

3.2. Trend Yield Determination

The trend yield is the yield variation over time (1990 to 2019). It is essential because it shows the level of technological advancement and acceptance and other non-natural attributes, including (farm management, fertilizer application, seed varieties, pesticides, policy, and labor) which increase yields with time [5]. The regression analysis was performed to check whether maize and sorghum yield in the Plateau, Central, and Savannah regions in Togo are a function of time (Equation (2)). The results in

Table 3. Regression results of maize and sorghum yield over time (years).

Crop	Variable	Coef	Std. Err	R2
Plateau
Maize	year	1.40	5.21	0.00
	cons	−1411.40	10,451.43
Sorghum	year	−4.21	4.99	0.02
	cons	9468.74	10,011.44
Central
Maize	year	17.37 ***	3.89	0.42
	cons	−33,399.22 ***	7807.13
Sorghum	year	8.98 **	3.85	0.16
	cons	−17,151.12 **	7721.66
Savannah
Maize	year	−2.57	4.98
	cons	6372.51	9973.14
Sorghum	year	27.14 ***	4.42	0.57
	cons	−53,482.85 ***	8867.71

***, ** significant at 1% and 5%, respectively.

show that maize yield in the Plateau region is positively and significantly influenced by time. However, time had a significantly negative effect on the maize yield in the Savannah region. Furthermore, there was no relationship between time and the maize yield from 1990 to 2019 in the Central region. On the contrary, a linear and negative relationship was observed between the sorghum yield and time in the Plateau region. Lastly, no relationship was found between sorghum yield and time in the Central and Savannah regions. The trend and actual yields of maize and sorghum in the three study areas are presented in Figure 3.

3.3. Determination of Coefficients of the Variables: Stepwise Regressions on Orthogonal Polynomial

Table 4. Stepwise regression results on the impact of weather variables on maize and sorghum yield over the entire growth.

Maize					Sorghum
Variable	Model 1		Model 2		Variable	Model 1		Model 2
	Coefficient	Std.Err	Coefficient	Std.Err		Coefficient	Std.Err	Coefficient	Std.Err
Plateau
RH			−68.29	31.23	RF	2.32 **	1.10	2.32 **	1.10
cons	1392.85 ***	44.4	6796.13 **	2471.06	cons	589.13 **	216.16	589.13 **	216.16
F test	F(0, 29) = 0		F(1, 28) = 4.78		F test	F(1, 28) = 4.47		F(1, 28) = 4.47
R2	0		0.15		R2	0.14		0.14
Adj R2	0		0.12		Adj R2	0.11		0.11
Central
WS			−590.69 ***	195.31	WS			−530.41 ***	155.43
cons	1423.57 ***	43.32	1841.38 ***	143.35	cons	839.36***	35.80	1214.37 ***	114.07
F test	F(0, 29) = 0		F(1, 28) = 9.15		F test	F(0, 29) = 0		F(1, 28) = 11.65
R2	0		0.25		R2	0		0.2938
Adj R2	0		0.22		Adj R2	0		0.2685
Savannah
T	−192.71 *	104.48	−184.26 *	91.16	RH			6.8 *	3.44
cons	6518.09 **	2873.25	6306.42 **	2506.77	T	396.91 ***	129.80	427.63 ***	124.53
F test	F(1, 28) = 3.4		F(1, 27) = 4.09		cons	−9998.13 ***	3569.44	−11,347.7 ***	3465.97
R2	0.11		0.13		F test	F(1, 28) = 9.35		F(2, 27) = 7.11
Adj R2	0.08		0.10		R2	0.25		0.34
					Adj R2	0.22		0.30

Note: RF (rainfall), T (temperature), SS (sunshine), WS (wind speed), RH (relative humidity). ***, **, * significant at 1%, 5% and 10%, respectively.

illustrates the stepwise regression results for the maize and sorghum growing season with respect to the main and additional weather factors. The output revealed that the rainfall, temperature, and sunshine do not influence the maize growing season in the Plateau and Central regions. However, in the Savannah region, the temperature had a significantly negative effect on maize yield. Furthermore, when wind speed and relative humidity were added to the equation, wind speed, relative humidity, and temperature were harmful to maize yield in the Plateau, Central, and Savannah regions.

Rainfall was beneficial to sorghum in the Plateau region. In the Savannah region, temperature exhibited a positive effect on sorghum yield. In the additional weather variables model, the effect of rainfall was still positive and significant in the Plateau region. However, in the Central region, the wind speed had a negative effect on sorghum yield. Lastly, temperature and relative humidity positively and significantly impacted the sorghum yield in the Savannah region.

The results in

Table 5. Stepwise regression results for the weather- yield with weather factors on the maize and sorghum growth stages.

Maize					Sorghum
Variable	Model 1		Model 2		Variable	Model 1		Model 2
	Coef	Std.Err	Coef	Std.Err		Coef	Std.Err	Coef	Std.Err
Plateau
RF0	1.92 **	0.31	1.11 ***	0.23	RF0	0.50**	0.21	0.4 **	0.17
RF1	−1.23 *	0.12			RF1			−0.30 *	0.11
T3	−31.25 *	15.59	−46.25 ***	12.02	WS1			−128.69 *	68.63
SS1	32.91 **	14.22	38.31 ***	10.12	RH3			−31.14 **	12.9
WS2			−83.46 **	37.97	cons	−486.42 **	209.97	−136.26 **	61.87
RH0			−17.79 ***	5.16	F test	F(1, 28) = 5.56		F(3, 26) = 3.05
cons	241.39 *	115.75	7348.21 ***	2065.44	R2	0.17		0.26
F test	F(4, 25) = 3.25		F(5, 24) = 9.77		Adj R2	0.14		0.18
R2	0.34		0.67
Adj R2	0.24		0.6
Central
RF0	0.35 *	0.18	0.54 **	0.21	RF0	0.35 **	0.16	0.48 ***	0.15
RF3			−0.35 *	0.17	RF3	−0.13 **	0.04
T1			52.98 *	28.58	T2	24.17 **	10.79	29.57 **	10.1
T2			−39.17 **	17.28	T3	−134.47 ***	35.2	−150.04 ***	31.91
SS0	−39.59 *	19.28	−77.67 ***	21.61	T4	15.91 *	8.56	20.19 **	7.67
SS1			−26.69 *	14.66	SS0			−26.97 *	14.74
SS3	−26.28 *	12.45	−50.39 ***	13.07	WS0			−63.47 **	27.13
WS2			−84.97 *	34.33	cons	−181.06 **	65.52	653.84 *	366.25
WS3			50.33 **	26.66	F test	F(5, 24) = 5.06		F(7, 22) = 5.97
WS4			−76.79 **	36.39	R2	0.51		0.66
RH2			−10.65 **	3.89	Adj R2	0.41		0.55
cons	941.20 *	456.35	2137.80 ***	552.48
F test	F(3, 26) = 4		F(11, 18) = 4.12
R2	0.32		0.72
Adj R2	0.24		0.54
Savannah
			−0.91 *	0.12	T3 SS2	−29.54 **	12.86	−51.08 *
RF4 T2			−0.76 *** −153.72 ***	0.15 28.02	WS2			101.98 **
T3			−81.38 **	27.04	WS3			91.90 **
SS0			−67.38 ***	16.38	RH2			8.01 *
SS2			36.46 **	11.27	cons	110.42 *	59.54	246.44 **
SS4			−20.56 ***	4.49	F test	F(1,28) = 5.27		F(5, 23) = 2.66
WS1			−76.05 ***	36.36	R2	0.16		0.37
WS2			109.12 ***	47.82	Adj R2	0.13		0.23
WS3			95.42 **	50.41
RH0			5.02 **	1.82
RH3			4.12 **	1.35
cons	0	42.31	649.37	843.24
F test	F(0, 29) = 0	F(11, 17) = 6.88
R2	0		0.82
Adj R2	0		0.7

***, **, * significant at 1%, 5% and 10%, respectively.

present the study regions’ weather yield (Y_M) models. Chebyshev orthogonal coefficients (φij), derived from Equation (13), whereas Equation (14) to (24) are adapted from Equation (11) and presented in

Table 6. Chebyshev orthogonal n = 5.

Month	φ0	φ1	φ2	φ3	φ4
1	1	−2	2	−1	1
2	1	−1	−1	2	−4
3	1	0	−2	0	6
4	1	1	−1	−2	−4
5	1	2	2	1	1

. Lastly, the resulting yield coefficients are presented in

Table 7. The yield coefficients for weather-related factors on maize yield by growth stage from the Chebyshev polynomial function.

Variable	Region	Unit	Growth Stage
			Emergence	Jointing	Tasselling	Milk	Physiological Maturity
Model 1
Rainfall	Plateau	Kg/ha/mm	4.38	3.15	1.92	0.69	−0.54
	Central		0.35	0.35	0.35	0.35	0.35
Temperature	Plateau	Kg/ha/°C	31.25	−62.51	0	62.51	−31.25
Sunshine	Plateau	Kg/ha/h	−65.83	−32.91	0	32.91	65.83
	Central		−13.31	−92.15	−39.59	12.96	−65.87
Model 2
	Plateau	Kg/ha/mm	1.11	1.11	1.11	1.11	1.11
Rainfall	Central		0.89	0.89	0.54	0.19	−0.16
	Savannah		0.15	1.22	0	1.06	−1.43
Temperature	Plateau	Kg/ha/°C	46.25	−92.51	0	92.51	−46.25
	Central		−184.31	−13.8	78.35	92.15	27.61
	Savannah		−226.05	−9.05	307.43	316.48	−388.82
Sunshine	Plateau		−76.63	−38.31	0	38.31	76.63
	Central	Kg/ha/h	26.09	−151.77	−77.67	−3.57	−181.44
	Savannah		−15.03	−21.59	−263.67	−21.59	−15.03
Wind	Plateau	Kg/ha/m/s	−166.93	83.46	166.93	83.46	−166.93
	Central		−297.06	492.79	−290.80	291.47	−196.39
	Savannah		274.92	157.77	−218.24	−376.01	161.56
Humidity	Plateau	Kg/ha/%	−17.79	−17.79	−17.79	−17.79	−17.79
	Central		−21.31	10.65	21.31	10.65	−21.31
	Savannah		0.9	13.25	5.02	−3.21	9.13

and Table 8. Y_MP denotes the weather yield in the Plateau region, Y_MC (weather yield in the Central region), and Y_MS (weather yield in the Savannah region).

• Maize

Model 1

$${\mathrm{Y}}_{\mathrm{M}}\mathrm{P}==241.39+1.92\times \mathrm{RF}0-1.23\times \mathrm{RF}1-31.25\times \mathrm{T}3+32.91\times \mathrm{SS}1$$

(14)

$${\mathrm{Y}}_{\mathrm{M}}\mathrm{C}=941.20+0.35\times \mathrm{RF}0-39.59\times \mathrm{SS}0-26.28\times \mathrm{SS}3$$

(15)

Model 2

$$\begin{array}{c}{\mathrm{Y}}_{\mathrm{M}}\mathrm{P}=7348.21-1.11\times \mathrm{RF}0-46.25\times \mathrm{T}3+38.31\times \mathrm{SS}1-83.46\times \mathrm{WS}2\hfill \\ -17.79\times \mathrm{RH}0 \hfill \end{array}$$

(16)

$$\begin{array}{c}{\mathrm{Y}}_{\mathrm{M}}\mathrm{C}=0.54\times \mathrm{RF}0-0.35\times \mathrm{RF}3+52.98\times \mathrm{T}1-39.17\times \mathrm{T}2-77.67\times \mathrm{SS}0\hfill \\ -26.69\times \mathrm{SS}1-50.39\times \mathrm{SS}3-84.97\times \mathrm{WS}2+50.33\times \mathrm{WS}3\hfill \\ --76.79\times \mathrm{WS}4-10.65\times \mathrm{RH}2\hfill \end{array}$$

(17)

$$\begin{array}{c}{\mathrm{Y}}_{\mathrm{M}}\mathrm{S}=-0.91\times \mathrm{RF}3-0.76\times \mathrm{RF}4-153.72\times \mathrm{T}2-81.38\times \mathrm{T}3-67.38\times \mathrm{SS}0\hfill \\ +36.46\times \mathrm{SS}2-20.56\times \mathrm{SS}4-76.05\times \mathrm{WS}1+109.12\times \mathrm{WS}2\hfill \\ +95.42\times \mathrm{WS}3+5.02\times \mathrm{RH}0+4.12\times \mathrm{RH}3\hfill \end{array}$$

(18)

• Sorghum

Model 1

$${\mathrm{Y}}_{\mathrm{M}}\mathrm{P}=-486.42+0.50\times \mathrm{RF}0$$

(19)

$${\mathrm{Y}}_{\mathrm{M}}\mathrm{C}=-181.06+0.35\times \mathrm{RF}0-0.13\times \mathrm{RF}3+24.17\times \mathrm{T}2-134.47\times \mathrm{T}3+15.91\times \mathrm{T}4$$

(20)

$${\mathrm{Y}}_{\mathrm{M}}\mathrm{S}=110.42-29.54\times \mathrm{SS}2$$

(21)

Model 2

$${\mathrm{Y}}_{\mathrm{M}}\mathrm{P}=-136.26+0.4\times \mathrm{RF}0+0.20\times \mathrm{RF}1-128.69\times \mathrm{WS}1-31.14\times \mathrm{RH}3$$

(22)

$$\begin{array}{c}{\mathrm{Y}}_{\mathrm{M}}\mathrm{C}=653.84+0.48\times \mathrm{RF}3-0.09\times \mathrm{RF}4+29.57\times \mathrm{T}2-150.04\times \mathrm{T}3+20.19\times \mathrm{T}4\hfill \\ -26.97\times \mathrm{SS}0-63.47\times \mathrm{WS}0\hfill \end{array}$$

(23)

$${\mathrm{Y}}_{\mathrm{M}}\mathrm{S}=246.44-51.08\times \mathrm{T}3+101.98\times \mathrm{WS}2+91.90\times \mathrm{WS}3+8.01\times \mathrm{RH}2$$

(24)

3.4. Impact of Meteorological Factors on Maize and Sorghum Yields by Growth Stages

3.4.1. Maize

Table 7 shows the meteorological-related influence coefficients of maize across the Plateau, Central, and Savannah regions for models 1 and 2.

• Model 1

Rainfall was beneficial to yield in the Plateau region during the emergence, jointing, tasseling, and milk but detrimental during physiological maturity. In the Central region, rainfall positively affected yield in all five stages. However, no relationship was found between rainfall and yield in the Savannah region. The effect of temperature on yield was only found in the Plateau region, where a slight increase led to yield improvement during the emergence and milk stages. The opposite effect was observed during the jointing and physiological maturity. Furthermore, sunshine exhibited a negative correlation with yield during the emergence and jointing stages and a positive effect in the milk and physiological maturity stages of the Plateau region. There was a negative relationship between yield and sunshine in the Central region except during the milk stage, where a positive trend was observed.

• Model 2

As clearly presented in the second part of Table 7, model 2 included wind speed and relative humidity representing our additional variables. The outcome revealed that rainfall was beneficial to maize yield in all five growth stages in the Plateau region. A similar trend was observed in the Central and Savannah regions except during the physiological maturity, where a negative effect was reported.

The effect of temperature on yield fluctuated across different growth periods and regions. For instance, a beneficial effect was observed during the emergence and milk stages in the Plateau region, tasseling, milk, and physiological maturity stages in the Central, and tasseling and milk stages in the Savannah. However, there was a negative relationship in other stages, except for the tasseling stage in the Plateau, with no observed effect.

Sunshine duration affected maize yield negatively at the emergence and jointing stages in the Plateau and at all five stages in the Central and Savannah regions except for the emergence stage in the Central region, where the effect was positive, similar to the milk and physiological maturity stages in the Plateau.

Furthermore, wind speed depicted a beneficial effect on yield in the: Plateau (jointing, tasseling, and milk stages), Central (jointing and milk stages), and Savannah (emergence, jointing, and physiological maturity stages). The opposite trend was observed in other growth stages.

Lastly, the effect of relative humidity on yield followed the same trend as in the cases above. The Plateau region observed a detrimental effect on yield in all five growth stages. However, relative humidity benefited yield at the jointing, tasseling, and milk stages, while during the emergence and physiological maturity stages in the Central region, it portrayed the opposite effect. In the Savannah region, during the first three stages (emergence, jointing, and tasseling), relative humidity was beneficial to yield, whereas the opposite trend was observed at the milk and physiological maturity stages.

3.4.2. Sorghum

Table 8 presents the meteorological-related influence coefficients of sorghum across the Plateau, Central, and Savannah regions for models 1 and 2.

• Model 1

Rainfall positively influenced sorghum yield in all five stages in the Plateau and Central region. The temperature had a positive effect on yield in the emergence, flowering, and grain filling stages and was harmful in the booting and physiological maturity stages in the Central region. The effect of sunshine on yield was detrimental in the Savannah region at the emergence and physiological maturity stages and beneficial in the booting, flowering, and grain filling stages.

Table 8. Yield coefficients for weather-related factors on sorghum yield by growth stage from the Chebyshev polynomial function.

Variable	Region	Unit	Growth Stage
			Emergence	Booting	Flowering	Grain Filling	Physiological Maturity
Model 1
Rainfall	Plateau	Kg/ha/mm	0.5	0.5	0.5	0.5	0.5
	Central		0.48	0.09	0.35	0.61	0.22
Temperature	Central	Kg/ha/°C	198.72	−356.75	47.12	181.13	−70.22
Sunshine	Savannah	Kg/ha/h	−59.07	29.54	59.07	29.54	−59.07
Model 2
Rainfall	Plateau	Kg/ha/mm	1	0.7	0.4	0.1	−0.2
	Central		0.48	0.48	0.48	0.48	0.48
Temperature	Central	Kg/ha/°C	229.36	−410.39	61.97	189.77	−70.72
	Savannah		51.08	−102.16	0	102.16	−51.08
Sunshine	Central	Kg/ha/h	−26.97	−26.97	−26.97	−26.97	−26.97
Wind	Plateau	Kg/ha/m/s	257.38	128.69	0	−128.69	−257.38
	Central		−63.47	−63.47	−63.47	−63.47	−63.47
	Savannah		112.06	81.82	−203.96	−285.78	295.86
Humidity	Plateau	Kg/ha/%	31.14	−62.27	0	62.27	−31.14
	Savannah		21.83	−19.62	−16.02	3.6	10.22

Table 8. Yield coefficients for weather-related factors on sorghum yield by growth stage from the Chebyshev polynomial function.

Variable	Region	Unit	Growth Stage
			Emergence	Booting	Flowering	Grain Filling	Physiological Maturity
Model 1
Rainfall	Plateau	Kg/ha/mm	0.5	0.5	0.5	0.5	0.5
	Central		0.48	0.09	0.35	0.61	0.22
Temperature	Central	Kg/ha/°C	198.72	−356.75	47.12	181.13	−70.22
Sunshine	Savannah	Kg/ha/h	−59.07	29.54	59.07	29.54	−59.07
Model 2
Rainfall	Plateau	Kg/ha/mm	1	0.7	0.4	0.1	−0.2
	Central		0.48	0.48	0.48	0.48	0.48
Temperature	Central	Kg/ha/°C	229.36	−410.39	61.97	189.77	−70.72
	Savannah		51.08	−102.16	0	102.16	−51.08
Sunshine	Central	Kg/ha/h	−26.97	−26.97	−26.97	−26.97	−26.97
Wind	Plateau	Kg/ha/m/s	257.38	128.69	0	−128.69	−257.38
	Central		−63.47	−63.47	−63.47	−63.47	−63.47
	Savannah		112.06	81.82	−203.96	−285.78	295.86
Humidity	Plateau	Kg/ha/%	31.14	−62.27	0	62.27	−31.14
	Savannah		21.83	−19.62	−16.02	3.6	10.22

• Model 2

Rainfall amount was beneficial to yield. For instance, a slight increase in rainfall in all five stages would increase yield in the Plateau and Central regions, except during the physiological maturity in the Plateau region. Likewise, temperature depicted a positive effect during the emergence, flowering, and grain filling stages, while the effect was negative during the Central region’s booting and physiological maturity stages. The effect was accentuated during the stages of emergence and booting. The same pattern was observed in the Savannah region except for the flowering stage, where no effect was registered.

The influence of sunshine on yield was consistently negative in all five growth stages in the Central region. The effect of wind speed was observed across all three regions. In the Plateau region, the wind speed positively influenced yield during the emergence and booting stages. However, the effect was negative during the grain filling and physiological maturity stages.

Wind speed was detrimental to yield in the Central region in all five growth stages. Yield in the Savannah region was positively and negatively affected by wind speed. The positive correlation was noticed during the emergence, booting, and physiological maturity stages, whereas the detrimental effect was observed during the flowering and grain filling stages.

Relative humidity was favorable to yield in the Plateau region during the emergence and grain filling stages. However, the opposite effect was observed during the booting and physiological maturity stages. Lastly, the correlation between yield and relative humidity was positive during the emergence, grain filling, and physiological maturity stages and negative during the booting and flowering stages in the Savannah region.

4. Discussion

Weather plays a vital role in yield improvement. This study comprehensively evaluated the impact of meteorological factors on maize and sorghum yield by growth stages in Togo’s Plateau, Central, and Savannah regions. The results from this study are substantial to previous studies, which found that weather factors’ effects are not uniform across different regions and growth stages [5,33,34,35].

Maize is a warm season crop that requires 200 to 450 mm of rainfall throughout the growing season [36]. This study found that rainfall influenced maize yield during the various growth stages across the three regions. Rainfall was beneficial in all three study areas, mainly during the emergence, jointing, tasseling, and milk stages, but the impact was negative during the physiological maturity stage. These results imply that holding other factors constant, an increase in the rainfall amount in the emergence stage sets an excellent and healthy foundation for the other four growth stages of maize. This finding substantiates previous research, which concluded that drought stress adversely influences plant physiological processes (dry matter partitioning, germination stage, tiller stage, vegetative growth, and reproductive organ development) [37], and grain filling [38,39]. During the rapid vegetative growth period, a deficit in water supply could result in a 28–32% loss of final dry matter weight and a consequent decrease in output yield [40]. Similarly, water shortage during the pre-anthesis, which occurs within the tasseling stage, leads to delayed leaf tip emergence and reduced leaf area. This induces long-term adverse effects, including reduced final diameters of leaves and internodes and yield losses of 15–25% [41].

Similarly, the study revealed a positive relationship between rainfall and sorghum yield during the emergence, booting, flowering, and grain filling stages in the Plateau and Central regions. This suggests that, like maize, sorghum requires appropriate and sufficient water for germination in the emergence stage. However, previous studies found that, like most crops, sorghum has an inverted U-shape relationship with cumulative precipitation, indicating that both low and high precipitation years are associated with yield reductions [42]. This explains why the effect of rainfall is negligible. Additionally, although sorghum tolerates both high and low precipitations, the reproductive stages (panicle development, flowering, and grain filling stages) [43,44] are susceptible to drought stress. Our results shed some light on the main growth stages during which water management strategies need to be implemented to increase maize and sorghum yield in Togo. It is also worth highlighting that the quantity of water available for plant use is also influenced by other weather parameters such as temperature, sunshine, evaporation, and relative humidity. Therefore, the above-cited factors must be considered when designing an efficient water use strategy. Sorghum is more tolerant to water stress than maize and has a competitive edge over other rainfed crops such as maize and rice [23,24]. This could explain the rationale behind the non-significant effect of rainfall on sorghum yield in all growth stages in the Savannah region, known for its hot climate. Water infrastructure has been proven to have a high and significant impact on economic growth in Sub-Saharan Africa [45], where agriculture’s share in the GDP is still high.

The temperature has a double effect on crop production. First, an increase or decrease in temperature might affect crop growth positively or negatively, depending on many factors (geographical situation, the crop tolerance to heat, the growth stage, etc.). Our findings indicated a positive impact of temperature on maize yield at various growth stages: (emergence and milk), (tasseling, milk, and physiological maturity), and (tasseling and milk) for Plateau, Central, and Savannah regions, respectively. However, a negative relationship was observed between temperature and maize yield during growth stages; (jointing and physiological maturity), (emergence and jointing), and (emergence, jointing, and physiological maturity) in Plateau, Central, and Savannah regions, respectively. This implies that higher temperatures harm crops at specific growth stages. Located in the southern part of Togo, the Plateau region is the country’s coldest region, with temperatures ranging from 19 °C to 36 °C. The Central and Savannah regions are in Northern Togo. The Savannah region is the hottest, with temperatures ranging from 20 °C to 40 °C, according to recent data from the Togolese meteorological center [46].

Our results acquiescently align with previous findings that high temperature could induce seedling death during the seedling stage owing to extreme dehydration of leaves beyond the permanent wilting point, chlorophyll content, photosynthesis, and respiration rate [47]. Maize plants are intolerant to heat stress and exposure of plants to heat beyond the tolerance threshold (>30 °C) for an extended period leads to a drastic reduction in maize yield [48]. Equally, high-temperature stress at the grain filling stage reduced the yield and quality of maize [49,50].

In the case of sorghum, an increase in temperature was detrimental to sorghum yield during the booting and physiological maturity stages and beneficial in the other three stages in the Central and Savannah regions. These results agree with previous research findings [51,52]. Our findings suggest that more sorghum production should be promoted in hotter areas such as the Savannah region. Sorghum is more tolerant to heat stress, especially during the vegetative stages, and because of sorghum’s tolerance to heat, West African production is mainly done in semi-arid and sub-humid areas [53]. However, higher temperatures in an already warm climate decrease sorghum yield during the reproductive stages, which are critical in sorghum plant growth and can adversely affect the final output. For example, high temperatures during anthesis can cause sorghum flower and embryo abortion [54]. Likewise, sorghum is susceptible to heat stress before and during the critical flowering and grain filling stages. The physiological processes (panicle initiation, reproductive organ development (gametogenesis), and pre-flowering photosynthetic efficiency) are disrupted by heat stress [51,52].

Sunshine is the key driver for photosynthesis, which plays a significant role in plant growth and development; thus, it is a significant predictor of crop output and quality and consequently provides a baseline for farm economic sustainability. In this study, we found that sunshine was positively correlated with maize yield during the milk and physiological maturity stages and negatively correlated during the other stages in the Plateau region. In the Central and Savannah regions, the correlation was negative except in the emergence stage in the Central region. On the other hand, the effect of sunshine on sorghum yield was negative in all the stages of growth evaluated, which was only observable in the Central region. The detrimental effect of sunlight on crops was also the main finding of authors [5,55], who confirmed that although plants needed sunlight for growth during the vegetative stage, if daylight duration was longer than what was necessary, the plant could lose its color and hibernate on the ground, and can delay the plant growth or worse cause its death in case the situation persists. Furthermore, high sunlight intensities can hinder flower formation during the tasseling or flowering stage, leading to fewer grains. This result is consistent with previous findings [5,56], who reported that minimum day sunshine is beneficial during early photosynthesis and floral formation. Moreover, drier and sunnier days decrease as we move from the northern (Savannah and Central regions) to the southern (Plateau region) part of Togo. This explains why the adverse effect of sunshine is more accentuated in the Savannah and Central regions. However, compared to maize, sorghum absorbs solar radiation heat more efficiently than maize [57]; thus, the effect of sunshine on sorghum yield is negligible compared to maize yield. This corroborates the findings of our study. Conversely, some studies reported that low solar radiation during the emergence stage could lead to abnormal seedling development (cotyledons remain small and yellow, diverting energy to stem elongation), disrupting the proper plant growth and, consequently, crop yield [55]. The beneficial effect of sunshine on yield during the physiological maturity stage in the Plateau region is in line with Kulyakwave et al. findings on rice [5]. Sunshine is very conducive to seed maturity, thus preparing them for harvest. However, in the Central and Savannah regions, the effect was disadvantageous. Long hours of sunlight are likely to cause maize and sorghum’s early ripening, which will decrease the total yield. Similar studies on wheat [58] came to the same conclusion.

Wind speed is an important yet understudied meteorological factor that affects plant growth and yield. Our results revealed that wind speed benefits maize yield during the Plateau region’s jointing, tasseling, and milk stages. However, the opposite effect was observed in the Central region during the emergence, tasseling, and physiological maturity stages, with a higher impact during the emergence stage. Further analysis showed a positive correlation between wind speed and sorghum yield during various growth stages: (emergence and jointing) and (emergence, jointing, and physiological maturity) for Plateau and Savannah regions, respectively, whereas the opposite effect was found in all five growth stages in the Central region. This suggests the importance of including wind speed in the climate–yield models for an accurate and timely prediction. Wind speed negatively and positively impacts crop growth and development across different regions. High wind speed can cause: stems and branches break, changes in plant motion, leaves damage, crop uprooting and dislocation, and growth suppression; thus, reducing crops yield [6,59]. However, it can also be advantageous to crop yield (dissemination of pollen, seeds, and other propagules, an increased supply of CO₂, and photosynthesis) [60].

Relative humidity is also an important climate factor that denotes the ratio of air vapor pressure to saturated vapor pressure. Our study found a negative correlation between relative humidity and maize yield in the Plateau region in all the growth stages. However, the negative correlation was only observed during the emergence and physiological maturity in the Central region and the milk stage in the Savannah region. Furthermore, relative humidity positively affected sorghum yield in the emergence and milk stages in the Plateau region and during the emergence, milk, and physiological maturity in the Savannah region. On the other hand, the negative effect of relative humidity on sorghum yield was during the booting and physiological maturity stages in the Plateau region and during the booting and flowering stages in the Savannah region. These findings suggest that plants are susceptible to relative humidity. The relative humidity can influence the leaves’ rate of transpiration and affect the water balance in crops. Changes in the relative air humidity affect the variations in the rate of photosynthetic activity [61] through its direct effect on leaf expansion (the enabler of light absorption, photosynthesis, and biomass production) [62]. In addition, it can affect dry matter, leaf area [63], and spikelet sterility [64]. Even though only the plant’s transpiration (which occurs only during the day because the stomatal opening is triggered by light) is directly impacted by relative humidity [65], the effect is widespread across the plant’s productivity. Low relative humidity increases evapotranspiration, enhancing the water requirements of both rainfed and irrigated crops, thereby increasing the likelihood of water stress conditions since there is less readily available water in the soil for plant roots [2]. This detrimental effect is accentuated in species with a reduced ability to control stomatal opening. Similarly, some research reported from a field experiment on rice that low relative humidity during the night severely hampered the growth of individual leaves and the expansion of the total leaf area [66].

On the contrary, high relative humidity has two different effects on growth. First, plants may exhibit increased growth due to greater stomatal opening, thus increasing CO₂ intake. However, on the other hand, they might exhibit slower development since there is less transpiration, which results in less nutrient transfer [67].

High relative humidity restricts plants’ transpiration and nutrient absorption, relieving the water stress on the soil [2]. In addition, although plants regulate the temperature of their tissues by transpiring water, which transforms its state from liquid to vapor, high relative humidity in the presence of high solar radiation values can induce excess thermal problems [68]. For instance, the average relative humidity during the maize growth period is around 80%, which is high. When relative humidity levels are too high, a plant cannot make water evaporate (part of the transpiration process) or draw nutrients from the soil, and when this occurs for a prolonged period, the plant eventually rots or cause diseases and insect pests, leading to a decrement in the total output. Likewise, high humidity coupled with high temperature can cause fungal disease outbreaks [69], thus affecting the final yield. Figure 4 summarizes the main findings of our study.

This study was conducted in three regions and only on two staple crops. Future studies should be conducted on the two others regions and staple crops for a more accurate outlook of the impact of weather on crop yield in Togo. Furthermore, field experiments, such as those conducted in the country’s northern region to assess the crop yield response to different irrigation management strategies using crop simulation models, are vividly encouraged. The author [70] used the OCCASION (Optimal Climate Change Adaption Strategies in Irrigation) framework. Five irrigation management techniques were assessed with regard to their effects on the inter-seasonal variability of the predicted yields and enhancements of the yield potential, ranging from no irrigation (NI) to controlled deficit irrigation (CDI) and full irrigation (FI). They discovered significant rainfall variability throughout the wet season, which causes substantial uncertainty in the anticipated yield under rainfed circumstances (NI). When supplemental irrigation management systems (CDI or FI) requiring a relatively low water demand of only 150 mm were implemented, this variability decreased considerably. It was demonstrated that both irrigation management options (CDI and FI) during the dry season would boost the local variety TZEE-W potential yield up to 4.84 Mg/ha while simultaneously reducing the variability of the predicted yield. However, even with CDI management, more than 400 mm of water would be needed to start irrigation in Northern Togo during the dry season.

5. Summary of Results

Weather plays an epic role in agriculture. Good weather factors improve crop yields; on the other hand, bad weather conditions have a downward effect on yields, thus, worsening worldwide food insecurity. In a unified framework, this paper combined the studies on the impact of the main meteorological factors (rainfall, temperature, and sunshine), the additional meteorological factors (wind speed and relative air humidity) on maize and sorghum yields through their (meteorological factors) direct impacts on the five different growth stages of maize and sorghum in three administrative regions of Togo (Plateau, Central, and Savannah). To achieve our objective, we used maize and sorghum yields and meteorological data from 1990 to 2019. Results showed rainfall’s beneficial and detrimental effects on maize and sorghum yield across stages and regions. The same trend was observed for temperature and was more significant across all growth stages of maize and sorghum in the Savannah and Plateau regions. The correlation between sunshine and maize yield was more accentuated in the Central and Savannah regions, while a negative effect of sunshine on sorghum yield was recorded in all the growth stages in the Central region. Furthermore, the wind speed was also beneficial and detrimental to maize and sorghum yields, especially during all growth stages in the three studied regions, although it was more notable for sorghum in Plateau and Savannah regions.

Conclusively, all five meteorological factors work together for yield improvement, ensuring food security. Adding variables such as wind speed and relative humidity improved the fit of the models better, as the R square was higher than the models without those variables (Table 6). Three main policy implications can arise from the study’s findings. First, there is an urgent need to promote and facilitate the acquisition of irrigation equipment to tackle the natural uncertainty of rainfall. Second, the adverse effect of climate change leading to heat and drought extreme events is becoming worrisome; therefore, the Togolese agricultural research agencies should direct their research toward creating new breeds that are drought and heat-resistant and shortcycle varieties. Third, the meteorological institution needs to be well equipped with upgraded technologies and machinery that will enable them to produce more accurate meteorological data, an essential tool for a reliable early warning system aiming to increase farmers’ awareness before the beginning of every growing season.