SoilWaterCharacteristics ofGleysols in theBamenda (Cameroon) Wetlands and Implications for Agricultural Management Strategies

Department of Geology, Higher Teacher Training College, e University of Bamenda, P.O. Box 39, Bambili, Cameroon Department of Soil Science, University of Dschang, P.O. Box 222, Dschang, Cameroon Department of Mining and Mineral Engineering, National Higher Polytechnic Institute, University of Bamenda, P.O. Box 39, Bambili, Cameroon Department of Earth Sciences, University of Ngaoundéré, P.O. Box 454, Ngaoundéré, Cameroon


Introduction
Soil water information is necessary for rainfall partition, establishment of irrigation schedules, and partitioning of net radiation [1,2]. e response of soil water properties is a key indicator of the impact of agricultural management on the movement of water and chemicals through the soil [3].
Wetlands are temporarily or permanently flooded areas where the soils are water-saturated and waterlogging is common [4].
Gleysols, typical of wetlands, are characterized by reduction or localized segregation of iron, due to temporary or permanent waterlogging causing anaerobic conditions [5].
ese soils occur where hydrological conditions controlled by abundant rainfall, gentle slope, low landscape position, or impermeable soils lead to soil surface saturation by water for a sufficiently long period enough for waterlogging to occur [6]. Water fluctuations at depths control the chemical and/or biological processes within both the water column and soil particles [7,8]. Wetland soils are unique among soils and possess morphological, physical, and chemical properties that readily distinguish them from upland soils [9,10]. Considerable studies have been carried out to understand the moisture characteristics of tropical soils [11][12][13]. However, little effort has been dedicated to the understanding of Gleysols water characteristics in tropical wetland ecosystems.
Wetlands cover approximately 6% of the Earth's surface in all climates [12]. Although these wetland soils may be minor inclusions in terms of their spatial distribution, they are important contributors to agricultural productivity [5].
ese areas contain about 35% of global terrestrial carbon critical to agricultural production and climate change mitigation [14]. e Ramsar Convention on Wetland Protection [4] highlights key information on wetlands and climate change mitigation and adaptation such as the assessment of carbon uptake and storage. is has revealed that continuous loss and degradation of wetlands results in significant losses of soil organic carbon stock (SOCS) to the atmosphere [14].
According to the Ramsar Sites Information Service [15], there are seven Ramsar sites in Cameroon ( e Barombi-Mbo Crater Lake, the Rio Del Rey Estuaries, Cameroonian section of the Ntem River, Cameroonian Section of the Sangha River, Cameroonian section of the Lake Chad, Waza-Logone floodplain, and the Ebogo Humid Zone) covering a total surface area of 8270.6 km 2 . e positions of the sites are shown in Figure 1. Many Cameroonian vast wetlands have not yet been studied in detail and thus remain unselected as Ramsar sites such as the Limbe and Wouri estuaries (Littoral Cameroon), the Nun River valley, the Menchum/Mezam Valley (Northwest Cameroon), the Santchou Floodplains (West Cameroon), amongst others. e study and selection of these wetlands as Ramsar Sites in Cameroon could contribute to one of the goals of the Ramsar Convention's strategic plan [4], which was to attain a protected area of 250 million hectares since 2015.
In Bamenda Town (Northwest Cameroon), major wetlands occur along the Mezam River floodplains and its tributaries especially in Ngomgham, Mulang, Ntenefor, Below Foncha, and Mile 4 Nkwen [15][16][17]. e lawless occupation of wetlands has intensified recently due to demographic pressure as soils in these areas are very fertile and support year-round market gardening [18,19]. Some human activities in these wetlands (land reclamation, waste disposal, deforestation, agriculture, industrialization, etc.) have led to the degradation of most of these ecosystems [20]. e main agricultural practice in Bamenda wetlands is market gardening and dry season maize farming [21]. In these wetland soils, Asongwe et al. [21] reported the spatial variability of physicochemical properties of soils under vegetable. Mofor et al. [22] focused on trace element status and environmental implications on soils and Zea mays near dumpsites. Land preparation for crop production is conducted under submerged conditions generating a massive plough layer which affects management strategies and crop performance. Works that concern the moisture characteristics of those Gleysols are inexistent despite the fact that soil water is the major factor affecting agriculture in these areas. Some question thus require answers: what are the moisture characteristics of the Bamenda wetland soils; is there a relationship between the moisture characteristics and the other soil properties; what are the best management strategies of the soils for sustainable agriculture? e aim of the present study was to determine the moisture characteristics of the Bamenda Gleysols and to establish a link between moisture characteristics and selected soil characteristics, as well as to recommend best management strategies of the Gleysols for crop production.
e results obtained will enable to support the conservation, restoration, and management of wetland soils and to create social awareness on the importance of wetlands in line with one of the goals of the Ramsar Convention's strategic plan.

Study Site Description.
e Bamenda municipality is located on the northwest flank of the Bamenda Mountain which is a stratovolcano situated along the Cameroon Volcanic Line [23,24]. It extends from longitude 10°08′ to 10°12′ E and latitude 5°55′ to 6°00′ N and at an altitude of 2621 m above sea level ( Figure 2). e town covers a surface area of 71.23 km 2 . e climate is the Cameroonian-type equatorial climate characterized by two seasons: a long rainy season of 7 months (April to October) and a short dry season of 5 months (November to March). e mean annual rainfall is 2670 mm and the average annual air temperature is 25°C. e Mezam River draining the town is a second-order perennial stream fed by several other small streams, most of which originate from the Bamenda escarpments. ey form a dense dendritic pattern. e major winds affecting this area are Harmattan (that bring the dry season) and the monsoon (that brings rain). Rainfall is heavy and often destructive. e vegetation is the Guinea Savannah type with short stunted trees (Bamenda Grassfields) and short deciduous trees; meanwhile Raffia palms grow in the swampy valleys. e town is located along the Cameroon Volcanic Line and exhibits two very distinct relief environments: the high lava plateau or Up Station (1400 m above sea level) and the low plateau or Downtown (1200 m above sea level), separated by an escarpment of about 150 m. Geologically, Bamenda is underlain by Precambrian granite-gneiss basement, and overlain by volcanic rocks like basalts, trachyte, dolerite, and ignimbrites [25], sedimentary silty clays of the Mezam River floodplains [26].
e dominant soils are Ferralsols with minor Lithosols on the hillslopes and Gleysols in the swampy valleys [9,21]. Farming is the main activity of the inhabitants and it is mostly crop-based farming, pastoral nomadism, mixed crop-livestock, and secondary and tertiary activities [21]. (20) sampling points, representative of the Bamenda Wetlands, were selected within five neighborhoods (Mulang, Ngomgham, Below Foncha, and Mile 4 Nkwen) in the Bamenda municipality wetlands ( Figure 2). One sample was collected per point at 0-50 cm depth (effective rooting zone for a majority of crops cultivated in this area: green beans, onion, licks, huckleberry, tomatoes, carrot, etc.). Disturbed and undisturbed samples were taken for three repetitions. e total area sampled was about 30 km 2 . Sampling was conducted in the dry season in November 2019 when the water level in the area was low. Soil samples collected were packed in air-tied plastic bags and taken to the laboratory for analysis.

Methodology and Analytics. Twenty
In the laboratory, the bulk density (D b ) was determined by paraffin coating method and particle density (D p ) was measured by pycnometer method. Porosity (p) was deduced from the particle and bulk densities. e particle size distribution was measured by Robinson's pipette method and    SOC was analyzed by Walkley-Black method. All these analyses were conducted according to the procedures described in [27]. e soil organic carbon stock (SOCS) was calculated according to the following equation [14]: where SOC is the organic carbon content (g kg −1 ), d is the soil layer thickness (m), δ 2mm is the coarse fragments (>2 mm diameter) content (%), and BD is the soil bulk density (Mg·m −3 ). e hygroscopic water content was calculated by using the weight loss of an air-dry sample, after subjecting it to an oven temperature of 105°C for 24 hours [28]. To determine the soil water content at field capacity (FC), the soil cores were first saturated and their soil water content at 33 kPa was measured using the tension tables [29]. e permanent wilting point (WP) water content at 1500 kPa suction was measured on the subsample using a pressure plate apparatus [30]. e capillary water (CW) was calculated as the difference between the hygroscopic water and the field capacity water content [31]. e unavailable water content (UW) was obtained as the difference between the capillary water and water content permanent wilting point [32]. Gravity water (GW) was obtained as the difference between total porosity and capillary water. e air content (AC) was calculated as the difference between the total porosity and the moisture content at field capacity [5]. e available water reserve (AWR) was calculated as the difference between the FC and WP [33]. e available water content (AWC) was calculated according to the following equation [5]: where FC is the moisture content at field capacity (%), WP is the moisture content at permanent wilting point (%), E is the depth of the soil layer (in dm), and D b is the soil bulk density in g·cm −3 . e readily available water content (RAW) was estimated as the product of the management allowed deficiency (approximately 2/3 for wetland soils) and the AWC [32]: e water-holding capacity (WHC) was calculated as the AWC multiplied by 2 according to GEPPA [34].
Before calculating the infiltration depth (in dm·m −1 ) in each of the soil layers, the volumetric water content was calculated according to the following equation [35]: where V w is the volumetric water content (at field capacity) in, FC is the gravimetric water content at field capacity, D b is the soil bulk density, and D w is the density of water (1 g·cm −3 ). e infiltration depth of the soil layer was then obtained as the product of V w and the soil depth (in dm) [35].

Statistical Analyses.
Pearson linear correlation test and principal component analysis (PCA) enabled establishing the relationship between soil characteristics.

Soil
Characteristics. e characteristics of the Bamenda Gleysols sampled in the five different neighborhoods are compiled in Table 1; meanwhile, results of summary statistical analysis of these soil characteristics are presented in Table 2. Soil characteristics such as particle density, bulk density, total porosity, and particle distribution did not vary much amongst the different sites. e SOC and SOCS were also very high for all the studied sites, but the ranges of values amongst the different sites differed very lightly (Table 1). e particle density of the soils ranged between 2.3 and 2.5 g·cm −3 , with a mean of 2.36 g·cm −3 , and a very low coefficient of variation (CV) of 2.86% (Table 2). About 45% of the samples showed a low particle density of 2.3 g/cm 3 while the rest ranged from 2.4 to 2.5 g·cm −3 . e bulk density ranged from 1.3 to 1.6 g·cm −3 , with a mean of 1.43 g cm −3 and a very low CV. e total porosity varied from 40 to 43.48%, with a mean of 41.67% and changed very little from one locality to another (CV � 3.0%). e soils showed only two grain size classes: clayey to clayey loam texture (Figure 3). e clayey fraction was moderately variable (CV � 21.0%), and 90.0% of the values fell between 30 and 45.0% clay content. e fine silt contents ranged from 8.0 to 29.0%, with a high CV. e coarse silt contents ranged from 10.0 to 22.0%, with an average of 17.90% and a moderate CV. e fine sand fraction varied from 6.0 to 22.0%, with a mean of 13.6% and a CV of 32.0%. e coarse sand content followed almost the same trend as fine sand, ranging from 9 to 22%, giving a mean of 15.1% and CV of 29%. e contents in coarse fragment varied from 3 to 25%, with an average value of 12% and a high CV. e sand/silt ratio fluctuated between 0.55 and 1.96, portraying a moderate coefficient of variation of 34%. e silt/clay ratio varied between 0.55 and 1.90, with a low CV of 10.5%. e (silt + sand)-to-total earth ratios of the Gleysols ranged between 52.29 and 75.73%. e SOC fluctuated from 5.67 to 19.45%, with an average of 13.66% and a CV of 30.32%. e SOCS varied between 251.63 and 804.93 Mg·ha −1 with a mean of 553.30 Mg·ha −1 and a moderate CV. e soil moisture characteristics are compiled in Table 3 and the results of summary statistical analysis of moisture retention characteristics are presented in Table 4. e hygroscopic water content varied between 7.20 and 23.03%, with a mean of 13.72% and a CV of 36.0%. e WP (pF3 � 1500 kPa) ranged between 22.14 and 58.86%, with a mean of 44.8% and CV of 25.0%. e moisture content at field capacity FC (pF3 � 33 kPa) changed between 48. 16

Correlation and Principal Component Analysis of the Soil Characteristics.
ere was a positive correlation between sand content and D p and a negative correlation between D P and clay content (Table 5). D b was correlating negatively with SOC while sand and clay contents showed a significant negative correlation. Clay and SOC contents showed a significant positive correlation. e coarse fragments contents correlated negatively with SOC and D b , but positively with D b . e silt/clay ratio showed a significant positive correlation with D b , D p as well as sand, silt, and rock fragment contents, but correlated negatively with SOC and clay contents. Among the soil properties, SOCS was positively correlated with D b , clay, and silt-to-clay ratio, but negatively correlated with D p , sand, silt, coarse fragments, and SOC contents. No significant correlation was observed between the soil moisture characteristics and silt/clay ratio. e SOCS showed a positive correlation with FC, PW, CW, and WHC and a negative correlation with HW, AWR, and UW. e analysis of linear correlation of soil moisture characteristics revealed that WP showed a significant positive correlation with AWC and RAW and a significant negative correlation with AWR and UW (Table 6). Also, FC showed a positive correlation with AWC, RAW, WHC, and infiltration depth (D) of the soil layer. e AWR showed a very significant negative correlation with AWC, RAW, and WHC. e AWC displayed a strong negative correlation with UW, while RAW displayed a strong negative correlation with UW. e CW revealed a strong positive correlation with D. Also, UW showed a significant negative correlation with WHC. e PCA enabled observing a reduction of the 17 original variables described to only three principal components (PC1, PC2, and PC3) explaining 73.73% of the total variance explained (Table 7). us, PC1 explained 45.12% of the total variance and revealed significant loadings on WP, AWR, AWC, RAW, UW, WHC, clay, and SOC, while PC2 expressed 18.93% of the total variance and showed significant loadings at FC, CW, and D (Figure 4). PC3 expressed significant loadings with D b and sand with a total variance of 9.68%. e PC1 was named the soil water retention factor because of a high factor loading of moisture characteristics with soil colloids (clay and organic carbon). e PC2 was referred to as the available water factor because of high factor loading between available water, capillary water, and infiltration depth. e PC3 was denoted the soil compaction factor as sand and bulk density affect soil porosity.

Soil Water Characteristic Curves.
Soil moisture characteristic curves enabled establishing the relationship between the volumetric water content and soil matric potential of the studied soil ( Figure 5), using water retention data (Table 8). e curves revealed a sigmoid shape which sloped gently from the HW to UW, steepened more from UW to AW, and became gentle from gravity water to complete pore saturation. e Gleysols of Mile 4 Nkwen showed the lowest water retention curve while those of Ngomgham displayed the highest ones ( Figure 5). e soils attained full water saturation (HW + AWC + GW + AC) at a range of 57.68% of dry soil (Mile 4 Nkwen) to 91.70% of dry soil (Ngomgham).

Soil Moisture Storage Capacity of Gleysols as Affected by Physical Properties and SOC.
e Gleysols of the Bamenda wetlands were reddish brown, wet, and waterlogged, with reddish yellow and blue patches in the more waterlogged areas. e surface aspect of the studied Gleysols is shown in Figure 6. ese properties are common to all wetland soils [28]. e massive structure and high density of these soils can be attributed to high clay content. e high clay content in turn made the soils stick to farm tools. e sand/silt ratio is an index of weathering intensity according to Sharma et al. [36]. All the ratio values were greater than 0.45%, typical of moderate weather. e silt/clay ratio of the Gleysols was >0.15, indicating that the soils are relatively young or >0.20 portraying a high degree of weathering potential [37]. ese indices agreed with values of alluvial floodplain Gleysols [13]. e plant-available water content of the Bamenda Gleysols was very high according to GEPPA [34]. Such high   values have already been reported in some wetland soils of North Cameroon [13]. Also, comparatively lower available water contents have also been reported elsewhere, such as 110.00 mm/m in Australia [38], 125.00 mm/m in Sudan [39], and 230.00 mm/m in India [40]. e difference is often attributed to soil texture, climate, and topographic positions [9,10].
e WHC values for soils from 95% of the sampling points of the Bamenda Gleysols were very high (>600 mm/m) and comparable with values of some wetland soil in North Cameroon [13]. us, Gleysols dominated by silt and sand have low WHC values while clay-rich ones show high WHC values [38].
e high values of RAW observed in most of the studied soils are a function of the soil texture and the organic matter contents of the Gleysols. e air capacity of the studied Gleysols was high compared to the optimum root aeration value of 10.00% and the critical value of 5.00% required for optimum plant performance [40]. Such high air contents indicate high free water potential. So, most of the water that infiltrates into pores is retained in the profiles and induces soil impermeability, surface ponding, poor aeration, and floods in the rainy season [41].  e soil moisture characteristics increased considerably from upstream at Mile 4 Nkwen to downstream at Ngomgham, Mulang, below Foncha, and Ntamulung. is might be related to transport and deposition of fine earth material from upstream to downstream and to the geomorphological configuration of the studied site or even to specific soil characteristics like the mineralogical composition of the clayey fraction. Also, several soil internal factors may be responsible for the variation in the soil moisture retention curve per site such as nonuniformity of pores, contact angle difference, and entrapped air in pore space [42]. e capillary water of the wetland soils was high as this water fraction is held to soil particles by weak surface tension forces of soil particles and is thus available to plants. During conditions of water stress (particularly during the dry season), capillarity plays a vital role in maintaining crop performance. e soil water loss is faster at the surface than at depth causing a higher water tension at the surface than at depth [43,44]. It causes a higher water tension at the surface than at depth and this creates a water capillary current which moves from base to surface and, hence, water rises to the surface to be available to plants [45,46]. e FC and WP are primarily influenced by irregular pore geometry and discontinuity, and variations in texture and mineralogy [47][48][49][50].
In this study, clay content of the soils correlated best with the moisture characteristics, more than any other property. e clay and SOC contents, probably linked together as revealed by their significant positive correlation, favor the occurrence of micropores and menisci that generate capillary forces. us, they increase the specific surface area of the soil matrix and, consequently, water adsorption [48]. Although silt/clay ratio displays no significant correlation with soil moisture characteristics, it negatively strongly correlates with SOC. e size and structure of clay particles were more suitable for association with SOM molecules than the size and the structure of bigger soil particles [49]. e clay and SOC contents are indispensable in the water and nutrient retention for plants based on their high specific surface area [50]. Sand and coarse fragments contents tend to increase soil particle density thereby increasing macroporosity and reducing microporosity, hence lowering the soil capacity to retain water [51]. us, low correlation coefficients between sand and coarse fragments contents and WHC are documented [52,53]; meanwhile, the negative correlation between SOC and silt content is not in line with some results [44]. e statistical analysis of the studied soil properties followed a log-normal distribution as revealed by the coefficient of kurtosis and skewness indicating possible anthropogenic disturbances. Under natural grasslands, many soil variables tend to conform to normal distribution due to the homogenization effect of biotic factors [54]. Following irrigation or rainfall, that saturates the soil, making its water content higher than FC, there should normally be a downward movement of water by gravity action. Infiltration is related to hydraulic conductivity that leads to decrease in soil moisture content as infiltration leads to soil water loss.

Implications of the Gleysols Characteristics on Agricultural
Management Strategies. In the Bamenda wetlands, water losses often attain WP in the dry season making plantavailable water contents low. is condition of water stress has been reported to cause drastic decrease in yields in wetlands [55]. erefore, for irrigation, the "plant readily available water" constitutes the ideal operating range of soil moisture for irrigation management [29,56]. us, irrigation scheduling should involve the replacement of depleted water not higher than the FC to avoid waterlogging in these soils as drainage and/or deep percolation is poor. But since the soils are flooded in the rainy season, it implies that their water content is above the FC and thus necessitating the drainage of excess water during farming in the rainy season. e soil water characteristic curves become an indispensable to design irrigation schedules, especially in the dry season during moisture deficiency or to drain excess water in the rainy season [57][58][59].
To achieve high yields without creating excess drainage, it is necessary to know the crop readily available to be cultivated. After readily available water has been used, plant roots cannot easily extract water and nutrients from the soil and this is the time to irrigate or refill point. e drier the soil is, the more water it needs to return to field capacity. us, in irrigation, FC and refill point are critical values for the correct use of many of the soil water monitoring. e high clay content of the Gleysols in the Bamenda wetlands makes tillage and weed control in these areas difficult due to the presence of water and sticky nature of soils to farm tools. e high bulk density of the Gleysols is very high (1.6 g·cm −3 ) compared to the optimum value (1.33 g·cm −3 ) for root penetration but also close to the upper limit (1.75 g·cm −3 ) for root growth, and is therefore a limiting factor to agriculture [60]. High bulk density implies that the soils are not well aerated as most of the pores are saturated with water during wet conditions. e highly significant correlation between clay and organic matter contents reveals that both are linked together to form organo-mineral complexes, and thus, the SOC is being protected by soil matrix. ese two soil colloids play a vital role in the retention of water and nutrients for crop needs as revealed by their high correlation with most of the moisture characteristics. e high SOC implies a high cation exchange capacity (CEC). e overall CEC of SOC and clay contents favors the retention and availability of nutrients to maintain crop production in those Gleysols [60][61][62].
Particle size distribution also conditions chemical nutrient supply [63]. Most of the studied Gleysols samples have clay plus silt content greater than 35%; this indicates poor nutrient assimilation by plants. A correlation has also been established between particle size distribution and yields [63]. e (silt + sand)-to-total earth ratios of the Gleysols were very high (>>35%) relative to the ratio range of 15 to 35% necessary for optimum yields and indicate potentially poor yields for a majority of tropical crops [63]. e SOCS of the Bamenda wetlands Gleysols is very high (>200 Mg·ha −1 ) as such wetlands, especially floodplains, are often very productive [4]. Such ecosystems have the ability of sequestering and storing carbon through photosynthesis and organic matter accumulation in soils and plant biomass thereby offering the opportunity for regaining lost productivity especially under agricultural systems. e properties of these studied soils could be similar to those of some Ramsar sites in Cameroon, although with some site specific features. Table 9 shows different Ramsar sites in Cameroon.

Conclusion
In Bamenda City (Northwest Cameroon), population increase has led to considerable lawless occupation of wetlands especially as these areas are very fertile and support yearround agriculture. Land use is often done without necessary technical know-how of sustainable land management techniques. is work was focused on the determination of moisture characteristics of Gleysols in the Bamenda wetlands and to establish a link between them and selected soil characteristics affecting crop production. e major findings revealed that the Gleysols showed very high organic carbon contents and very high soil organic carbon stocks implying a high level of carbon sequestration. Major constraints of the soils to farming were clay plus silt contents, massive structure, and waterlogging. e moisture properties revealed very high water-holding capacity and very high plant-available water. e particle density and coarse fragments and sand contents correlated negatively with the soil moisture retention characteristics and could be reducing the soil's water-holding capacity. e soil moisture characteristics curves were sigmoid-shaped and their heights decreased from upstream to downstream probably portraying increase of finer material with fluvial transport distant.
e present work might have been limited in terms of the inadequate number of samples studied. Further research is necessary on these wetland soils of the northwest region of Cameroon to evaluate their potential risk of degradation based on a higher number of samples. Also, the results presented in this work might have been biased based on the fact that the tests were conducted solely in the laboratory. In the future, measurements of water storage capacity of these soils should be done directly in the field using more accurate methods such as electrical resistance and tensiometric and heat-diffusion methods. Irrigation schedules on this these soils should involve the replacement of depleted water with water amounts not exceeding the FC to avoid waterlogging in these soils as drainage and/or deep percolation is poor. Adapted crops like lowland rice could also be introduced to extend the surfaces of rice cultivation in the zone.

Data Availability
Data can be made available on request.

Conflicts of Interest
e authors declare no conflicts of interest.