The study area is located in the Centre Region of Cameroon, precisely in the Bafia district, between latitudes 4˚40' and 4˚48' North, and longitudes 11˚2' and 11˚8' East ( Figure 1 ). Bafia district covers an area of about 1300 km². The area is located on the southern Cameroonian plateau with an average altitude of 700 m and elevations that can reach 900 m. The climate is the humid sub-equatorial type, with four seasons (a short rainy season, a short dry season, a long rainy season and a long dry season). Average annual rainfall varies between 1400 mm and 1600 mm. Average temperatures vary between 23˚C and 25˚C. Average Relative humidity is 87%. The vegetation in the area is dominated by gallery forest and shrub savannah. It is fairly well watered by rivers and many marshes, including the Mbam river, Sanaga river, Inoubou and Noun [16] . The soils are sesquioxidic, with acrisols and ferralsols dominating on acid rocks found in the forests and indurated ferralitic soils characterised by iron nodules found in the savannah and in some forests. These soils have metamorphic rocks as dominant parent material (gneiss, orthogneiss, mica schists), belonging to the geological group of Yaoundé [17] [18] .

Seven sites ( Table 1 ), representative of the study area, and based on previous survey studies by ORSTOM [19] , were identified. The seven sites were selected based on geomorphological features, soil forming factors, and pedogenic processes as depicted in previous studies by ORSTOM in the study area [19] . For each of the sites, a soil profile was realised under natural vegetation and described according to the FAO soil description guidelines [20] . Soil samples were collected following diagnostic horizons, and stored in polythene bags prior to laboratory analyses. Undisturbed samples for bulk density determination were collected using Kopecky rings of 100 cm³.

Chemical properties were determined according to the procedures described by Pauwels et al. [21] . Soil OC content was determined by the Walkley and Black method, while bulk density was determined as the oven dry mass (105˚C) of each undisturbed core sample by volume. Soil pH was determined using 1:2 soil-H₂O and 1:2 soil-KCl ratios for pH_water and pH_KCl, respectively. Total nitrogen and available phosphorus (P) were determined by the Kjeldahl wet digestion method and the Bray II method, respectively. Exchangeable base cations were determined by the Schollenberger’s method using a 1M ammonium acetate solution buffered at pH 7. The concentrations of exchangeable sodium (Na⁺) and potassium (K⁺) in the solution extract were obtained by flame photometry, while those of calcium (Ca²⁺) and magnesium (Mg²⁺) were obtained by complexometry using a 0.002M Na₂-EDTA solution. The cation exchange capacity (CEC) was determined by direct continuation of the Schollenberger’s method using a 1N KCl solution for the displacement of ammonium ions. The hydrometer method was used for particle size analysis according to the procedures described by Bouyoucos [22] after soil dispersion with a 1N sodium hexametaphosphate solution.

After obtaining daily rainfall for the study area for the past 30 years, the average monthly rainfall over a 30-year period was determined. Potential evapotranspiration (PET) was estimated using the FAO Penman-Monteith method [23] [24] defined by Equation (1).

$\text{ETo}=\frac{0.408\Delta \left(\text{Rn}-\text{G}\right)+\gamma \frac{900}{\text{T}+273}\text{U}2\left(\text{es-ea}\right)}{\Delta +\gamma \left(1+0.34\text{U}2\right)}$ (1)

where:

ETo = Reference evapotranspiration (mm∙day⁻¹),

Rn = net radiation at the crop surface (MJ∙m⁻²∙day⁻¹),

G: soil heat flux density (G = 0 because it is assumed that there is a balance between heat loss during the day and night, MJ∙m⁻²∙day⁻¹),

T: Mean daily air temperature at 2 m height (˚C),

U2: wind speed at 2 m height (m∙s⁻¹),

γ: psychrometric constant (KPa∙˚C⁻¹),

es: saturated vapour pressure (KPa),

ea: actual vapour pressure (mbar),

es-ea: saturation vapour pressure deficit (kPa),

∆: slope of saturation vapour pressure curve (kPa∙˚C⁻¹).

The potential evapotranspiration determined enable the obtention of the climatic diagram ( Figure 2 ) in the study area. Based on the climatic diagram, the humid period begins in early March and ends around mid November, while the growing period starts around the end of the month of March and ends around the end of the month of November.

The land use envisaged is rain-fed cultivation of Egusi melon by smallholders. The suitability of the land was assessed using land evaluation tables proposed by Sys et al. [25] for the needs of watermelon, which belongs to the same family (Cucurbitaceae) and genus (Citrullus) as Egusi melon, given that the specific needs of Egusi melon are unavailable. A parametric method [25] [26] was used to classify soil suitability into four categories: highly suitable (S1), moderately suitable (S2), marginally suitable (S3) and unsuitable (N) ( Table 2 ). In this parametric method (Equation (2)), the soil and climate characteristics are defined using different scores [25]

$\text{I}=\text{A}\times \sqrt{\frac{\text{B}}{100}\times \frac{\text{C}}{100}\times \frac{\text{D}}{100}\cdots \frac{\text{Z}}{100}}$ (2)

where I is the specified index and A, B, C, D, E, etc. are the different scores assigned to each parameter making up the crop requirements, with A corresponding to the lowest score. The climatic characteristics are corrected as follows:

if CI < 25, then $\text{CCI}=\text{1}\text{.6xCI}$; (3)

if 25 < CI < 92.5, then $\text{CCI}=16.67+\text{0}\text{.9xCI}$; (4)

where, CI = climatic index and CCI = corrected climatic index.

If the climatic value is less than 25, for example, the correction is done using Equation (3). If the value is between 25 and 92.5, correction is done using Equation (4). If the value is 92.5, no correction is required. Land indices, on the other hand, are obtained based on Equation (2), and are corrected using Table 3 [27] . The values obtained after correction are used to determine a soil’s capability. These corrections increase the accuracy of evaluation indices and better reflect soil and environmental conditions. The climatic characteristics considered in this study are rainfall, temperature and relative humidity, while the soil characteristics are topography (slope), moisture (flooding and drainage), soil physical characteristics (texture, percentage of volume of gravel content and soil depth), soil fertility characteristics (CEC, base saturation, organic carbon content, soil reaction), salinity (electrical conductivity) and alkalinity (exchangeable sodium percent).

By determining the land index and using the guidelines defined by Sys et al. [28] , the qualitative land suitability classes ( Table 2 ) and the factors limiting plant growth in the different soil series for the crop were determined. Critical values for nutrients and soil fertility parameters are presented in Table 4 [29] - [31] .

Descriptive statistics (mean values, standard error of means, and coefficient of variation) was performed on soil properties. Coefficient of variation (CV %) was used to examine the variability of soil properties and was calculated using Equation (5);

$\text{CV}\left(\text{\%}\right)=\frac{\text{SD}}{\text{X}}\times 100$ (5)

where; SD = standard deviation, x = arithmetic mean of soil property.

Soil properties having CV values < 15% were grouped as least variable, those with CV values between 15% and 35% were grouped as moderately variable and those with CV > 35% indicated high variability [32] . Statistical analysis was facilitated using Microsoft Excel 2016 and SPSS (Version 20).

The morphological and physical characteristics of the studied soils are presented in Table 5 . The soils of profiles 1 and 2 are dominantly reddish, slightly red at the surface and yellowish red in the subsurface horizons of profile 2. Profile 3 is dominantly greyish, while profiles 4 and 6 are brown at the surface and red in the lower horizons. Profile 5 is yellowish brown at the subsurface horizons and greyish black at the surface, while profile 7 is slightly reddish at the surface and pinkish at the bottom of the profile. Surface colour in tropical soils with high rainfall and leaching is strongly linked to organic matter, while colour of subsurface horizons is generally related to the types of minerals present in the soil [33] . The general negative depth function of organic matter in these soils is consistent with the colours. The soils in the study area are dominated by a sub-angular blocky structure, with the exception of unit 2, where the structure is massive in the subsurface. Unit 3 has both granular and massive structures. This reflects the parent material, and could probably be linked to the type of alluvium deposited, and also to the abrupt break in texture [34] . The soils in the study area are predominantly clayey in texture, with the exception of the soils in unit 3, which are predominantly sandy-clay at the surface and linked to the parent material, which consist of fluvial deposits. All the soils in the study area are friable when dry and hard to soft when wet.

SB = Sub angular blocky, GR = Granular, SB → GR = Sub angular blocky parting to Granular, FR = Friable, FI = Firm, VFI = Very firm, ST = Sticky, SST = Slightly sticky, NST = Non-sticky, C = Clear (2 - 5 cm), G = Gradual (5 - 15 cm), D = Diffuse (>15 cm), S = Smooth, W = Wavy, I = Irregular, A = Abrupt. NB: FI = fine (for structure) and firm (for consistence). Source: FAO [20] .

OC = Organic carbon, C/N = carbon to nitrogen ratio, SEB = Sum of exchangeable base cations, CEC = Cation exchange capacity, BS = Base saturation, ESP = Exchangeable sodium percent, EC = Electrical conductivity.

SE: Standard error of means.

The pH values observed in the various study sites are presented in Table 6 . Soils of unit 1 are acidic, while in unit 2, acidity varies from acidic to moderately acidic. In both units, there is a general decrease in acidity with depth. This decrease in pH with depth, especially in tropical soils where heavy rains obtain is mainly due to leaching of bases [35] . Acidity in soils of unit 3 varies from slightly acidic in the surface horizons to acidic in the subsurface horizons, while those in unit 4 are slightly acidic. The decrease in pH in these soils may be related to the accumulation of organic acids in deeper horizons through leaching [36] . In units 5 and 6, the soils are generally slightly acidic throughout the profiles. Unit 7 has a basic character (slightly alkaline) in the upper parts of the profile and slightly acidic in the lower parts of the profile. This degree of alkalinity in soils of unit 7 may be attributed to the recurrent practice of bush burning in the area, a farming practice that releases ash, capable of raising the pH [37] [38] . The coefficient of variation ( Table 7 ) indicates that for all soils, pH is least variable, ranging between 3.17 and 11.61 %. This low variability in pH of the soils is influenced by factors of soil formation, and land management [39] . As pH is a good indicator of soil quality, influencing the availability and uptake of nutrients in crops such as Egusi melon, proper management of soil acidity is necessary despite the low variability.

Unit 1 has a medium organic carbon content and is of very good quality at the surface. The OC content is generally low and of average quality in the lower parts of the profile. In unit 2, OC content generally varies from low to medium with a general decrease with depth. The quality of the OM in this profile has an inverse depth function. The soils in units 3 and 4 have a low to medium OC content of poor quality at the surface horizons and good quality at the subsurface horizons. Units 5, 6 and 7 have medium to high levels of OC with concentrations generally decreasing with depth. In the latter three soils, the organic matter is of very poor quality. Generally, organic matter decreases with depth as would be expected, linked to the presence of dense vegetation at the surface and the accumulation of organic matter from plant litter [40] . The high variability (CV = 17.24% - 45.40%) of the OM quality observed in the soils corroborates with the variety of vegetation and microorganisms observed in the studied soils [41] .

Apart from profiles 1 and 7 where total nitrogen concentrations are relatively higher, all the soils generally have low nitrogen concentrations, which decrease with depth. This low level of nitrogen in the soils may be related to leaching by rainwater [42] , given the heavy rainfalls usually recorded in the study area. The correlation tables ( Table S1 to Table S7 ) show both positive and negative, significant to highly significant correlations between total nitrogen and other soil physical and chemical properties, indicating that several parameters can be explanatory of soil nitrogen availability. In Unit 1, there exists a positive and significant correlation between sand and nitrogen (R = 0.98), indicating that the presence of sand in the soil could affect the availability of nitrogen to plants. In Unit 3, the positive and highly significant correlation between total nitrogen and potential acidity (R = 0.99), indicates that the two are linked by biogeochemical processes, such as nitrogen mineralization, which can be linked to certain types of soil microorganisms available with a pH that favours their development.

Soils in units 1 and 2 have very low concentrations of phosphorus (<7 ppm). Soils in units 3 and 6 have P concentrations that vary from low to very low with a general increase with depth. The increase in phosphorus with depth may be linked to less leaching, as the deeper soil horizons are less subject to leaching than the upper layers, which allows a slightly greater retention of phosphorus at depth [43] . In unit 4, the trend is the reverse of the two previous units, with phosphorus levels decreasing with depth, from low to very low. The soils in units 5 and 7, on the other hand, have low levels throughout the soil profile. The low to very low P levels with negative depth functions in these soils may be linked to phosphorus sorption. Highly weathered soils may contain clay minerals such as iron and aluminium oxides, which have a strong capacity to fix phosphorus and make it unavailable [43] . Average phosphorus levels in Bafia soils ranged from 2.39 ± 0.46 ppm to 10.01 ± 0.63 ppm ( Table 7 ). This shows the low to very low level of phosphorus, which is a constraint for the production of Egusi melon. Low phosphorus levels lead to a delay in flowering, a significant reduction in above-ground biomass (up to 71%) and root biomass (up to 68%), and a reduction in leaf number and plant size [41] . Phosphorus deficiency also limits interaction with other nutrients, essential for crops, which can lead to a drastic drop in yields [44] [45] . Table S1 to Table S7 show a positive and significant correlation between phosphorus and calcium (R = 0.98), and between phosphorus and base saturation (R = 0.98).

The soils in the study area are generally non-saline (<2 dSm⁻¹) ( Table 4 ). This is probably due to the leaching of salts into the lower horizons as a result of the heavy rainfall in the study area.

Cation exchange capacity is high (25 - 40 cmol (+) kg⁻¹) in the soils of units 1, 4, 5, 6 and 7. This may be linked to the accumulation of organic matter in these soils, the clay minerals being essentially low CEC clays [46] . In unit 2, the CEC is average, varying between 15.2 and 28 cmol (+) kg⁻¹, while CEC unit 3 are average at the surface and very high at the subsurface horizons. The nature of the parent materials, which are alluvial deposits and clay content of the soils could be contributing factors for this observation [47] . Leaching of bases down the lower horizons is a possibility [48] .

The presence of cations in the soil is generally linked to parent materials and organic matter. Table 5 indicates that calcium concentration in units 1, 2, 3 and 5 increases slightly with depth, from low (6 - 12 cmol (+) kg⁻¹) high. This can be explained by the accumulation of calcium down the profiles, and by several factors such as pH, organic matter and soil mineralogy [48] . The presence and distribution of these cations in the soil can have an impact on the growth of Egusi melon. It is important to ensure that the soil contains adequate levels of these cations, particularly calcium, which plays a crucial role in plant nutrient uptake and overall growth [49] . Based on the cationic balance ratio Ca/Mg/K, necessary for optimum absorption, which is 76% for calcium, 18% for magnesium and 6% for potassium [50] , there is a pronounced imbalance in calcium in units 1, 3 and 4. In units 2 and 5, there is an imbalance in potassium, magnesium and calcium throughout the soil profiles. Unit 6 is characterized by an imbalance in calcium in all horizons, as well as an imbalance in magnesium in the subsurface horizons. In Unit 7, there is an imbalance in magnesium and potassium.

Based on soil morphological and physico-chemical properties, and taking into account procedures for naming soils using Reference Soil Groups, principal and supplementary qualifiers [51] , three reference soil groups were identified and classified as follows; Profiles 1, 2, 6 and 7 were classified as Haplic Ferralsol (Epic, Ochric, Orthodystric), Haplic Ferralsol (Epic, Ochric, Orthodystric), Gleyic Ferralsol (Dystric, Endic, Profundihumic) and Chromic Ferralsol (Arenic, Endic, Profundihumic) respectively. Profiles 3 and 4 were classified as Eutric Amphifluvic Stagnic Protogleyic Cambisols (Humic) and Rhodic Cambisols (Clayic), respectively. Profile 5, was classified as a Rhodic Acrisol (Clayic, Endic, Humic).

The climatic indices ( Table 8 ) indicate that the suitability class for climatic parameters for the three possible growing seasons in the study area is the same (S2, c, with a moderate limitation due to relative humidity). Based on climatic characteristics, Bafia is moderately suitable for growing Egusi melons. Firstly, the identification of a medium suitability class for climate (class S2) is in line with studies that recognise the importance of climatic parameters such as rainfall and relative humidity for the cultivation of crops [52] [53] . For example, drought has been identified as a major factor affecting agricultural productivity in Africa [53] . In the case of Bafia, moderate limitations due to relative humidity could negatively affect crops during certain periods, but could be managed via irrigation systems or adapted agricultural practices. Soil evaluation shows that the soils in units 1 and 2 are currently unsuitable with very severe but correctable limitations. The soils of units 3, 4, 5 and 6 have marginal suitability, while soils in unit 7 have medium suitability. The different constraints for each unit are presented in Table 9 . It can be observed ( Table 9 ) that 42.9%, 28.6%, 71.4% and 14.3% of the land in the study area is limited by the physical characteristics of the soil (s), topography (t), fertility characteristics (f) and moisture (w), respectively. More specifically, these limitations are related to slope, soil texture, pH, flooding, base saturation, organic matter and gravel. Variations in soil suitability between different units could result in common challenges by farmers in areas with heterogeneity in soils and climate. To cope with such variability, practices need to be adapted to ensure stable yields [54] . The presence of correctable limitations suggests that soil management strategies could be put in place to improve soil suitability, such as fertility enhancement or anti-erosion measures. The specific constraints cited, such as slope, soil texture, flooding, etc., are well documented in studies as factors influencing agricultural production [52] [55] . Soil texture, for example, affects water retention and nutrient availability, while steep slopes can increase the risk of erosion. For this reason, organic matter needs to be added to increase the structural stability of the soil in order to resist erosion and ensure good water retention. Organic matter could also raise the pH, thereby reducing acidity. Biochars derived from animal manure and grass, for example, are used to increase soil N, P, K, Ca, Mg and S content [56] [57] . In addition, the use of mineral fertilisers can help in increasing soil fertility levels for melon crops [58] . As strategies that should be implemented to help small-scale farmers in Bafia cope with the Egusi melon production constraints in future, the following are suggested: Adoption of sustainable practices, such as crop rotation and the use of improved seeds, to optimize yields. Rotating crops helps in breaking pest and disease cycles, improving soil health, and increasing nutrient availability. Associating or rotating Egusi melon with legumes can be particularly beneficial as they fix nitrogen in the soil. Using high quality, disease-resistant seeds can significantly enhance yields. Improved seeds are necessary for better germination rates, growth, and resistance to pests and diseases. Combining biological, cropping, and chemical practices to manage pests can reduce the reliance on chemical pesticides and promote a healthier ecosystem [59] . Fertility trials that can inform on optimal fertilizer rates for Egusi melon production are also to be considered as a perspective.

Table 8. Land suitability evaluation for Egusi melon cultivation in the various study sites.

A = April, J = June, Ju = July, S = September, N = November, V = value, PV = parametric value, GP = growing period, CI = climatic index, C = Clay, SC = sandy clay SCL = sandy clay loam, LI = land index, CEC = cation exchange capacity, SEB = sum of exchangeable bases ESP = exchangeable sodium, EC = electrical conductivity.