Nutritional proteins are one of the major metabolism needs, which are rarely satisfied in some developing countries, especially by Africans. Indeed, in Africa, animal proteins are less available for people due to economic constraints.

Therefore, focusing on plant proteins becomes the preferred way to satisfy protein needs in Africa (Mazahib et al., 2013). Among protein-providing crops, legumes take a large part in supplying safe proteins in food systems by increasing proteins levels for human adequate nutrition (Vertès et al., 2015). Moreover, legumes have ability to fix atmospheric nitrogen, and then transfer it into, and fertilize the soil. Like other legume species, Bambara groundnut (Vigna subterranea (L.) Verd.) contains high level of lysine and plays an important nutritional role when it is eaten combined with cereals in rural peoples’ diet (Massawe et al., 2005; Bamshaiye et al., 2011). Moreover, Bambara groundnut takes part in increasing phosphorus bio-availability in the soil (Ndiang et al., 2012; Touré et al., 2013; Gbaguidi et al., 2015).

Despite its attributes, Bambara groundnut always remains a neglected and underutilized species often scoring low yields in Benin (Dansi et al., 2012). In fact, the low crop low yields are caused by low cation exchange capacity, low sum of bases, mineral deficiency (nitrogen and mainly assimilable phosphorus) and low organic matter contents of the soil, which are not favourable for increased growth of many crops in Benin soils (Igue et al., 2013). However, Arbuscular mycorrhizal fungi (AMF) constitute a microorganism group mainly having beneficial effects on plant growth (by mobilization of some nutrients) and tolerance to many biotic and abiotic stresses after initializing symbiosis with plant roots (Smith and Read, 2008; Saïdou et al., 2012; Haougui et al., 2013).

Many researchers have reported use of endomycorrhizal symbiosis to improve productivity of many plants (Aboubacar et al., 2013; Usharani et al., 2014; Haro et al., 2015; Do Rego et al., 2015). In Benin, Houngnandan et al. (2009) evaluated indigenous Glomus species diversity in the clear forest of Isoberlina doka (Craib et Stapf) at Wari-Maro in Benin centre. Tchabi et al. (2008) and Balogoun et al. (2015) also have studied the diversity of endomycorrhizal fungi associated respectively with cotton and cashew tree. These different researches resulted in identification of some species of Glomeromyceta associated with many crops in Benin. To the authors’ knowledge, no research was carried out in West Africa, especially in Benin focusing on AMF species and strains associated to Bambara groundnut. Therefore, the aim of this work was to study AMF community present in Bambara groundnut rhizosphere in the different agro-ecological zones in Benin.

Study areas

Earlier researches showed that Bambara groundnut was grown in Benin with more production areas localized in the northern part of the country (Gbaguidi et al., 2016). Statistical data on Bambara groundnut production also showed that it is cultivated in six agro ecological zones in Benin, namely the Far North Benin (AEZ 1), Cotton region of North Benin (AEZ 2), Food crop region of South Borgou (AEZ 3), West zone of Atacora (AEZ 4), Cotton region of the centre (AEZ 5) and Bar lands (AEZ 6) (Figure 1). However, the production of the crop is not so much important across all sixth AEZ and therefore the study covered only five AEZs.

Survey in the field

Based on Bambara groundnut sowing date in the different areas, the survey was carried out in September 2018 in order to be sure to find Bambara groundnut in the field and especially at flowering phase. Two regions were selected per agro-ecological area. These regions were taken on the base of Bambara groundnut yield, cultural surface affected to this species and its production in these regions the last five years. The online database of FAO (Country Stat) has been used for obtaining this information on Bambara groundnut in each region and it helps for sampling place choice. Each sampling site was georeferenced with GPS.

Laboratory work

Density and diversity assessment of AMF, assessment of mycorrhization level and soil chemical characteristics determination were done at the Laboratory of Soil Microbiology and Microbial Ecology (LMSEM) of the Faculty of Agronomic Sciences (FSA) in the University of Abomey-Calavi.

Soil and root sampling from the fields

Per sampling site, composite samples have been made. Four soil samples were collected (20 soil samples) at four different points using a custom-made corer at 20-cm depth. All of these samples were mixed together and the mixture was used to obtain a composite sample. At each soil sampling site, four plants were also been pulled up for root sampling.

Spore extraction, counting and morphological identification

AMF spores were extracted using the wet sieving and centrifugation method of Gerdmann and Nicholson (1963). One hundred grams of composite soil sample from Bambara groundnut rhizosphere was weighed, mixed in water, stirred thoroughly, let decant before being sieved through serial sieves with different mesh sizes (2 mm, 250 µm, 150 µm, 63 µm and 50 µm). This process was repeated four times. The sediments from the different sieves were collected in tubes and finally centrifuged after adding sucrose solutions (5mL of 20 and 60% w:v) at 4000 rounds/min for 4min at 4ºC (Oehl et al., 2003). After this process, suspended spores were collected and counted under a binocluar magnifying glass using gridded Petri dish (5cm of diameter) to make the counting easier. Based on morphological criteria (colour, presence or absence of suspensor bulb and hyphae, etc.), some AMF morphotypes were identified, counted and grouped into genera by using identification and description keys from the International Collection of Vesicular and Arbuscular Mycorrhizal fungi (INVAM, http//www.invam.caf.wdu.edu). Spore density was expressed as numbers of AMF spores per 100 g of dry soil. AMF spore density in each sample was obtained by summing abundances of all species recorded in the sample.

AMF diversity assessment

Biological diversity indexes; that is, Shannon diversity Index (H’) (Shannon and Weaver, 1962), Simpson diversity Index (1-D) (Simpson, 1949) and Hill diversity index (1-Hill) were calculated for each agro ecological zone to evaluate AMF diversity.

Assessment of mycorrhization

Fresh roots of Bambara groundnut collected during the survey were used. After staining with trypan blue dye using the method described by Phillips and Hayman (1970), 20 fragments of fine Bambara groundnut roots were cut and placed on a glass slide with cover slip, and were observed with an optical microscope (XSP-BM-2CA, AliExpress) to observe different structures of AMF from different samples. Assessment of AMF colonization was performed using intersection method described by Trouvelot et al. (1986). The mycorrhization rates were assessed by two parameters of arbuscular mycorrhizal colonization namely: mycorrhization frequency and mycorrhization intensity.

Soil chemical analysis

From each Bambara groundnut field soil collected during the survey was subsampled (20), air dried and sieved for determination of chemical properties including the pH by potentiometric method in water (soil: water ratio 1:2.5) using a digital pH meter, organic carbon (C) by the Walkley–Black (1934) method, total N by Kjeldahl method and assimilable P by Bray I method.

Statistical analysis

One-way Analysis of Variance (ANOVA) was performed to assess the effects of agro-ecological zone (AEZ) on AMF spore density and diversity parameters, but also soil chemical parameters. When significant differences (P < 0.05) were found, post hoc comparisons of means among AEZ were made using a Student-Newman–Keuls’ test. This ANOVA was performed with non-transformed data after ensuring conformity of these data with ANOVA assumptions. Furthermore, a Factorial Component Analysis (FCA) was performed in order to characterize different zones with identified spore genera. For choosing axes, the two principal’s components must present more than 50% of the total information. Additionally, a Pearson correlation test was performed among all of the studied parameters in order to determine the relationships between them. All statistical analyses were carried out using SAS software version 9.2.

Density of spores associated to Bambara groundnut in different agro-ecological zones

Analysis of variance results showed that all of the recorded mean spore densities were significantly different (p <0.001) among surveyed AEZs (Table 1). In different Bambara rhizosphere soil samples, important spore densities were recorded. The highest spore density (5713 ± 35 spores per 100 g of soil) was recorded in AEZ 2 while AEZ 4 has recorded the lowest spore density (2826 ± 28 spores per 100g of soil). Thus, the ANOVA results showed that the mean spore density recorded was significantly different (p <0.001) among different surveyed AEZs. Additionally, when identification criteria were considered, it appeared that black spores were more frequent compared to other coloured spores; and the AEZ 1 recorded the highest black spore density (3144 ± 56spores per 100g of soil) followed by AEZ 2 (3021 ± 62 spores per 100g of soil) (Table 1). Based on spore size, it generally appeared that spores with small size were most abundant. It was noticed that spore size of 63µm showed the highest spore density compared to other spore sizes (250µm, 150µm and 50µm); and the AEZ 3 had recorded the highest spore density of 63µm (3534 ± 27spores per 100 g of soil) followed by the AEZ 2 (3169 ± 17 spores per 100g of soil (Table 2).

AMF community composition

In the Bambara rhizosphere studied, 14 mycorrhizal fungal morphotypes have been identified according to morphological criteria (colour, presence or absence of suspensor bulb and hyphae, etc.). After identification of these morphotypes based on the INVAM identification key, results indicated that the AMF belonged to five genera (Glomus, Gigaspora, Acaulospora, Scutellospora and Diversispora), four families (Glomeraceae, Acaulosporaceae, Diversisporaceae and Gigasporaceae) and two (02) orders (Diversisporales and Glomerales), all in Glomeromycetes Phylum. Relative abundance of these genera broadly among AEZs showed that Glomus species are the most abundant genera recorded in all surveyed AEZs, and were represented by 66.82% of all recorded genera. Other genera found in surveyed AEZs were Gigaspora, Scutellospora, Acaulospora and Diversispora genera scoring 18.46, 8.25, 4.89 and 1.58%, respectively (Figure 2). Furthermore, analysis of this fungal distribution in the different AEZs indicated that species belonging to Glomus and Diversispora genera are mostly recorded in AEZ 1 and AEZ 4, while species of Gigaspora and Scutellospora genera were most common in AEZ 3 (Figure 3). All identified genera were found in AEZ 2 and AEZ 5.

AMF diversity

Statistical analysis showed that for all diversity indexes, there were significant differences (p<0.0001) among AEZs (Table 3). Shannon index (H’) indicated that the AMF species were most diverse in AEZ 4 based on its highest index value (1.46) as compared to AEZ 3, which recorded the least species diversity (Table 3). When Simpson index (1-D) is considered, it appeared that AEZ 2 showed the most diverse species with the highest Simpson index (0.677) and AEZ 5 showed the least species diversity (0.533). This is true because H’ index considered small populations, whereas the 1-D index considered larger populations. It was indicated earlier that AEZ 4 was a less abundant zone compared to AEZ 2, which was the most abundant zone in term of spores. The Hill index (1-Hill) also showed that all of surveyed zones are mainly diverse, like both previous diversity indexes (Table 3).

Assessment of natural mycorrhization level

As shown in Figure 4, there was a significant (p<0.0001) difference amongst the studied AEZs in terms of the mycorrhization frequency. The highest frequency (45.63%) was recorded in AEZ 1, compared to other AEZs. There was no significant difference amongst other AEZs and the lowest rate of mycorrhization frequency was recorded in AEZ 4 (17.25%). In terms of mycorrhization intensity, there were significant differences amongst AEZs; and the AEZ 3 recorded the significantly highest value (8.06%), while the lowest rate was from AEZ 4 (2.39%).

Chemical characteristics of soil from surveyed agro-ecological zones

The results showed significant differences (p<0.001) amongst the different AEZs based on studied soil parameters (Table 4).

In general, soils from different surveyed zones have relatively low pH (between 5 - 6.5 in all of the surveyed zones), poor in nitrogen (%N < 0.075% in all of surveyed zones) and phosphorus (P < 40 ppm in all of surveyed zones) according to Baize (2000) classification guideline. Furthermore, AEZ 2 had the recorded highest level of studied parameters; that is, Carbon (0.63%), Nitrogen (0.063%) and Phosphorus (29.50 ppm) but was third in pH level as compared to other zones (Table 4). Lowest levels of chemical parameters of soil were mostly recorded in AEZ 1.

Assessment of relationships between all of studied parameters

In AEZ 1, there was a significant positive correlation between Nitrogen and parameters like mycorrhization frequency (r = 0.958), Hill index (r = 0.974) and Simpson index (r = 0.999) (Table 5). This means that increasing of Nitrogen level in soil will induce enhancement of mycorrhization frequency and also AMF diversity. Additionally, it appeared there was a positive significant correlation between all of the diversity indexes (Table 5).

For AEZ 2, there was a significant negative correlation (r = -0.970) between Nitrogen (N) and mycorrhization frequency (F%) as shown in Table 5. So, increasing of N level by one unit led to 97% decrease of F%. Additionally, the Shannon and Hill indexes are positively correlated and both of them increased or decreased in the same way.

In AEZ 3, a significant positive correlation (p = 0.0285) between soil pH and spore density was recorded (Table 5). So increasing or decreasing soil pH could affect spore density in soil. Additionally, a significant negative correlation (r = -0.952) appeared between P and I%. So, a P increase leads to I% decrease in 95.2%. Positive correlation was found among diversity indexes.

For AEZ 5, correlation results were in general not significant except mycorrhization frequency and Shannon index which have shown significant positive correlation (r = 0.986; p = 0.0135) (Table 5). It can be notice that only correlations statistically significant have been taken into account here.

Density of spores associated to Bambara groundnut

Most of the surveyed zones have a high level of spore densities. These spore densities under Bambara groundnut range from 2826 to 5713 spores per 100g of soil. These recorded densities were less than those recorded by Bossou et al. (2019) under maize (6260 spores per 100 g dry weight soil) but higher than those by Johnson et al. (2013) under cowpea (202 ± 42 per 100 g dry weight soil), by Balogoun et al. (2015) in cashew orchards and by Houngnandan et al. (2009) in Isoberlinia doka habitats (237 to 258 spores per 100g of soil). These high spore levels could be due to different factors; and Brundrett (2009) indicated that the presence and natural distribution of glomales were controlled not only by floristic composition but also by environmental conditions. Additionally, legumes had an ability to promote development of fungal propagules by releasing some exudates in their rhizosphere, which favoured development and increase of microorganisms including mycorrhizal fungi (Scheublin et al., 2004). Indeed, a crop species can directly influence mycorrhizal fungal spore’s abundance and composition (Eom et al., 2004; Lovelock et al., 2003). The results in the current study have shown that the largest spore amounts were recorded in AEZ 2, which is a cotton cropping area. In this part of Benin, the land-use system is mostly an agroforestry system in which many forest species such as Vitellaria paradoxa, Parkia biglobosa and Tamarindus indica (Gnangle et al., 2012) were kept in fields during land occupation. Indeed, species such as Parkia biglobosa and Tamarindus indica are also legumes harbouring in their soil and root habitat significant fungal flora that increase in population over many years; and these floras are beneficial for subsequent crops (Guissou et al., 1998). In addition, the cropping fields constitute a continual rotating environment where involved crops promote, with times, a high spore’s abundance (Houngnandan et al., 2009). Furthermore, in the current study, zones where high spore densities were found are mostly characterized by sandy or sandy loamy soils.

This could also explain high densities recorded in these zones because sandy soils are said to favour high glomales populations (Dalpé, 1989). Ferruginous soils on sandstone and the presence of many rocks and concretions (in AEZ 4) as well as heavily degraded soils (in AEZ 5) have limited fertility and may be the cause of low spore densities in these soils compared to others zones. It is also important to note that the samples were taken during the flowering period. Knowing that the crops establishment is done during the raining period, the unimodal climate (a long dry season followed by a long rainy season), and the period from sowing to flowering associated with favourable environmental conditions (moisture and presence of the roots of the plant) could have allowed an activation and multiplication of spores that could explain the high densities recorded. Indeed, Bohrer et al. (2003) reported that spores number is higher in the soil exposed to relatively long water stress conditions (dry season) thanks to the production of spores.

Furthermore, small-sized spores are most abundant compared to those with large size in all surveyed zones. These results are supported by research findings of Bossou et al. (2019) and Johnson et al. (2013) who indicated that AMF spore densities were proportionally inverse with respect to their sizes. These study results are consistent with findings from Bossou et al. (2019) who showed a largest abundance of black spores followed by white spores in all surveyed zones. However, these results are in opposite with findings from Johnson et al. (2013) who reported that black spores became first in terms of density but followed by brown spores. These different results could be explained due to differences in plant species, which were not the same in these different studies areas.

AMF diversity

The fourteen different morphotypes collected and identified based on the morphological characters belong to the genera of Glomus, Gigaspora, Acaulospora, Scutellospora and Diversispora. From these genera, Glomus and Gigaspora were more frequently recorded from all surveyed AEZs. The richness of these genera was higher than those obtained in Benin under different crop habitats including corn (Bossou et al., 2019), cashew (Balogoun et al., 2015) and Isoberlinia doka (Craib and Stapf) (Houngnandan et al., 2009). On the other hand, this genera richness was low as compared to that obtained by Tchabi et al. (2008) in the sub-Saharan savannahs (08 genera), by Johnson et al. (2013) under cowpea in all AEZs in Benin and by Diop et al. (1994) in Senegal (15 species). Results from the present study showed that the spore diversity of fungi varies relatively little from one AEZ to another, with Glomus the most abundant genus in all zones. Indeed, this predominance of Glomus was reported in AMF morphotypes in various tropical soils (Tchabi et al., 2008; Houngnandan et al., 2009) and in agricultural soils from temperate zones (Oehl et al., 2003; Mathimaran et al., 2005). Abundance of this genus in different agro-ecosystems might indicate that it is the most adaptable and available in different environments (Daniell et al., 2001; Brito et al., 2012). Diversity analysis has also shown that AEZ 1 and AEZ 4 were characterized by genera Glomus and Diversispora and AEZ 3 by Gigaspora and Scutellospora. This might indicate that some genera are typical to each zone. This is an important aspect to consider for the determination of the type of inoculum appropriate for each zone.

Mycorrhization frequency and intensity

Mycorrhization frequencies and intensities varied amongst zones. The area recording the highest mycorrhization frequency had neither a good infection level nor the highest levels of soil chemical parameters. At the same time, the area with the highest levels of spore density and different soil chemical parameters recorded low mycorrhization frequency and intensity. This could indicate that low levels of certain soil parameters (nitrogen, carbon, phosphorus and pH) favour the colonization capacity of Bambara groundnut roots by the fungal species present. Indeed, these results corroborate a previous report from Houngnandan et al. (2009) who showed that plots with low phosphorus levels were those with high frequencies of mycorrhization. Conversely, Liu et al. (2012) have shown that high levels of Nitrogen and Phosphorus have reduced Glomeromycetes populations in the soil. There is a very weak correlation between density, frequency and intensity of mycorrhization. It is said that living spores of AMF present in the soil may not function as propagules, but they may be quiescent (inactive because soil conditions are unsuitable) or have an innate period of dormancy-mechanisms, which may help them survive during periods of adverse soil conditions (Brundrett, 2009). In addition, soil parameters might diversely influence establishment of the symbiosis between the mycorrhizal fungi and Bambara groundnut roots.

Correlation between different studied parameters

There is much evidence supporting the hypothesis of a large and diversified influence of soil properties on AM fungi (Sano et al., 2002; Johnson et al., 2005; Mechri et al., 2008; Gryndler et al., 2009). It appeared that AM species may survive and function well within a range of soil and environmental conditions. In the current study, almost no significant correlation was found between soil chemical parameters and spore density, with the exception of soil pH in AEZ 3, where positive correlation with density occurred. Tchabi et al. (2008) reported similar results showing spore production increase with soil pH. In contrast, Houngnandan et al. (2009) found correlation amongst most parameters and indicated a negative correlation not only between Phosphorus, Nitrogen and Soil carbon[MOU1]  with spore density but also between mycorrhizal frequency and spore density. In an earlier report, Subramanian et al. (2006) have already indicated that application of Phosphorus can influence spore production either positively or negatively. 

In the current study, there was no relationship between mycorrhization frequency and intensity. Indeed, the zone with the highest frequencies did not necessarily have strong intensities of mycorrhization and vice versa. However, negative correlation was denoted between mycorrhization frequency or intensity with diversity index. These might stipulate that efficiency of mycorrhization depends on the diversity of species or genera present in the plant rhizosphere. In this study, negative correlations between Nitrogen and Phosphorus with the frequency and intensity of mycorrhization were found. This might show that increase occurrence of some soil parameters may lead to reduction of mycorrhization level. Furthermore, species diversity was positively correlated with some soil parameters (Nitrogen and Phosphorus). This could stipulate that certain species have preferences and tolerance levels in terms of physico-chemical soil parameters; and this could lead to their appearance or disappearance in different soils. These results were opposite to those of Liu et al. (2012) who did not indicate any relationship between soil parameters and diversity. Also, Johnson et al. (2013) found that AMF diversity indexes are negatively correlated with both available and total Phosphorus. All these results may be natural, because AMF are living organisms and have different preference or tolerance levels to some environmental factors. 

The present study has shown that there is an important spore density of AMF associated with Bambara groundnut that differs significantly from one AEZ to another. In addition, it appeared that soil parameters diversely influence both mycorrhization frequency and intensity, but also AMF diversity. Also, it should be noted that having high spore densities does not imply higher levels of symbiosis. Furthermore, it appeared that AEZ 1 and AEZ 4 were characterized by Glomus and Diversispora genera; but Gigaspora and Scutellospora characterized AEZ 3. So it would be appropriate that further studies be carried out in order to confirm these results. Also, achievement of molecular characterization will allow most accuracy in species identification. This will enable an evaluation of effectiveness and efficiency of different collected species in order to develop some ecological technologies in Bambara groundnut fertilization based on AMF.

The authors have not declared any conflicts of interests.