Soil chemical properties
In general, Skhwahlane and Dikgwale sites appeared to have intermediate to high levels of C, N, Ca, Mg, B, and Fe relative to the other sites (Table 1).
δ15N values of reference plants
Several plant species, including grasses, herbaceous weeds, and cereals, were collected from farmers’ fields in the six villages (namely, Machipe, Majakaneng, Dikgwale, Malekutu, Phameni, and Skhwahlane) in Mpumalanga Province. Where more than one reference plant was sampled per farm, the average value of the combined δ15N values was used to estimate %Ndfa of Bambara groundnut on that farm. A recent study has shown that the combined mean δ15N value of herbaceous weeds can be used to estimate a legume’s dependence on symbiotic N2 fixation for its N nutrition (Nyemba and Dakora 2010).
A comparison of Bambara groundnut plant growth between and among the 26 fields surveyed in Mpumalanga revealed marked differences in DM yield (Table 3). Across the board, DM yield was highest at Dikgwale (field 9, 162 g plant−1), followed by Skhwahlane (field 25, 141.9 g plant−1), Malekutu (field 18, 109.1 g plant−1), Skhwahlane (field 23, 89.2 g plant−1), and Skhwahlane (field 24, 86.5 g plant−1). The lowest DM (6.0 g plant−1) was obtained at Malekutu (field 12).
Dry matter yield of shoots plus pods differed significantly (p ≤ 0.05) between fields around each village surveyed. For example, at the Machipe site, DM production was highest in field 4 (64 g plant−1) and lowest in field 5 (17 g plant−1), while at Majakaneng, DM yield was 81 g plant−1 in field 6 compared to 20 g plant−1 in field 8 (Table 3). Similarly, Bambara groundnut showed better growth at Dikgwale in field 9 (163 g dry mass plant−1) compared to field 8 (68.6 g plant−1).
Field 10 at Dikgwale similarly showed greater plant growth (163.3 g DM plant−1) than field 9 (68.6 g DM plant−1). Of the 9 farms sampled at Malekutu, farm 12 produced 5.9 g DM plant−1 in contrast to farm 18, which yielded 109.1 g DM plant−1 (Table 3). Significant differences were similarly found in the DM yield of shoots plus pods at Phameni. For example, shoot plus pod biomass production ranged from 7.8 g DM plant−1 (field 22) to 53.4 g plant−1 (field 19) at the Phameni site, while it ranged from 73 g DM plant−1 (field 26) to 142 g DM plant−1 (farm 25) at Skhwahlane (Table 3).
N concentration and N content
N concentration and content of Bambara groundnut plants sampled across Mpumalanga Province also differed significantly between fields. The N concentration of shoots plus pods ranged from 1.7 to 3.9 % across the 26 farmers’ fields studied (Table 3). As a result of those differences in N concentration, N content of shoots plus pods also differed between and among fields. For example, the N content of Bambara groundnut shoots plus pods was much greater at Skhwahlane field 25 (4.4 g plant−1), followed by Dikgwale field 9 (3.9 g plant−1), Malekutu field 18 (3.6 g plant−1), and Skhwahlane field 24 (3.2 g plant−1). In contrast, the lowest DM yield was produced at Malekutu (see fields 12, 14, and 13).
δ15N values and %Ndfa
The δ15N of Bambara groundnut plants sampled in Mpumalanga Province showed marked differences across the 26 farmers’ fields surveyed and varied from −1.77 to +1.92 ‰ for shoots and −1.29 to +2.61 ‰ for shoots plus pods (Table 3). The highest δ15N value was obtained at Malekutu (field 10), followed by Skhwahlane (fields 26 and 25) and Dikgwale (field 8). In contrast, the lowest δ15N values for shoots plus pods were recorded at Malekutu (field 12) and Phameni (field 20). The %Ndfa of Bambara groundnut shoots plus pods also differed markedly between farmers’ fields and ranged from 33 % (field 10) to 98 % at Malekutu (field 12). Except for Malekutu (field 10) and Skhwahlane (field 25), where the legume derived below 50 % N (i.e., 33 and 43 %, respectively) from symbiotic fixation, Bambara groundnut was able to obtain between 52 and 98 % of its N nutrition on all the other fields studied (Table 4).
Amounts of N-fixed
The amounts of N-fixed by Bambara groundnut plants on farmers’ fields differed significantly (p ≤ 0.05) between and among the 26 farms surveyed. As shown in Table 4, symbiotic N contribution by Bambara groundnut across the 26 fields ranged from 4 to 200 kg ha−1, with both minimum and maximum values being recorded at Malekutu fields 12 and 18, respectively. The sites where Bambara groundnut plants made a greater N contribution (116–144 kg N-fixed ha−1) to the N economy of the cropping system included Malekutu (field 16), Skhwahlane (fields 23, 24, and 25), and Machipe (field 4). By contrast, the areas with very low N contribution were Malekutu (field 14) and Phameni (field 21).
Net N returns to soil
The net N returns to soil from Bambara groundnut shoots ranged from 7 to 177 kg N ha−1 (Table 4). Despite the low plant density, the estimated N returns to soil were quite substantial at Skhwahlane (fields 23 and 24), Malekutu (field 16), Machipe (field 4), and Malekutu (field 18) (i.e., 177, 170, 137, 125, and 124 kg N ha−1, respectively; see Table 4), indicating the potential of Bambara groundnut to contribute to the N economy of cropping systems.
Soil mineral N uptake
Soil N uptake differed markedly between and among Bambara groundnut plants sampled from the 26 farmers’ fields (Table 4) and ranged from 0.01 kg N ha−1 at Malekutu (field 12) to 173 kg N ha−1 at Skhwahlane (field 25). Other locations where soil N uptake was high included Skhwahlane (116 kg N ha−1 at field 24), Malekutu (89 kg N ha−1 at field 4), and Skhwahlane (63 kg N ha−1 at field 26). Soil N uptake by Bambara groundnut plants was however below 10 kg N ha−1 in nine other farmers’ fields (Table 4).
Plant density and its relationship with biomass production and N2 fixation
There were large differences in Bambara groundnut plant density across the 26 farmers’ fields surveyed (Table 3). Relative to other sites, plant density was highest at Malekutu (e.g., fields 13 and 10 with 16 and 12 plants m−2, respectively). In fact, 15 out of the 26 fields surveyed had very low plant densities, ranging between 3 and 6 plants m−2 (Table 3).
As shown in Fig. 1, there was evidence to show that low plant density affected DM yield and hence N2 fixation. Thus, plant biomass increased where legume plant density was high (Fig. 1a). The amount of N-fixed was however also much greater where Bambara groundnut density was higher in the cropping system (Fig. 1b).
C concentration in Bambara groundnut shoots plus pods differed markedly across the 26 farmers’ fields studied and ranged from 40 to 49 % (Table 5). Plants at Skhwahlane (field 23) exhibited the highest C concentration (48.9 %), followed by Phameni (46 % in each of fields 20 and 22), Skhwahlane (46 % in field 26), and Majakaneng (46 % in field 6). The lowest C concentration was recorded at Malekutu (40 % at field 10), followed by Machipe (43 % in each of fields 2 and 4), Phameni (43 % in field 19), Dikgwale (43 % in field 9), Majakaneng (43 % in field 7), and Malekutu (43 % in field 18).
As a result, the C content of Bambara groundnut shoots plus pods also differed significantly across all fields and ranged from 2.6 to 70.9 g plant−1 (Table 5). The highest C content was found at Dikgwale (70.9 g plant−1 in field 9), followed by Skhwahlane (62.5 g plant−1 in field 25) and Malekutu (46.8 g plant−1 in field 18). In contrast, the lowest C content was recorded in plants harvested from Malekutu (2.6 g plant−1 in field 12), Phameni (3.6 g plant−1 in field 21), and Malekutu (4.3 g plant−1 in each of fields 13 and 14).
C/N ratio and photosynthetic fixed N use efficiency
The C/N ratio is a measure of the C and N relationship in plants and is therefore an estimate of photosynthetic N use efficiency. As with other parameters, there were significant differences in C/N ratios across all 26 fields where Bambara groundnut was sampled. Overall, the C/N ratio ranged from 10.7 g g−1 in field 13 to 26.6 g g−1 in field 12 at Malekutu (Table 5).
The photosynthetic N-fixed use efficiency (measured as shoot C per unit N-fixed in shoots) also varied with the amounts of symbiotic N and C accumulated by field-grown Bambara groundnut plants. As shown in Table 5, the Bambara groundnut material sampled from Malekutu (field 10) showed the highest photosynthetic fixed N use efficiency (154 g C g−1 N-fixed), followed by Malekutu (83 g C g−1 N-fixed in field 12), Majakaneng (58 g C g−1 N-fixed in field 6), and Phameni (54 g C g−1 N-fixed in field 22). The Bambara groundnut shoots with the lowest photosynthetic N-fixed use efficiency were sampled from Skhwahlane (field 24), Dikgwale (fields 8 and 9), and Skhwahlane (field 23), and they respectively recorded 5.0, 5.9, 6.5, and 6.1 g C g−1 N-fixed (Table 5).
A one-way ANOVA of δ13C from Bambara groundnut shoots revealed significantly marked variation between and among farmers’ fields at the different study sites. The δ13C values of Bambara groundnut shoots were much greater (i.e., less negative) at Skhwahlane (fields 23 and 24, −26.20 and −26.68 ‰, respectively) when compared to the other sites/fields (Table 5). However, the fields that showed much lower, or more negative, δ13C values in Bambara groundnut shoots include Phameni (−28.24 ‰ in field 21), Malekutu (−28.06 ‰ in field 13), and Dikgwale (−28.01 ‰ in field 9). Thus, the δ13C values of Bambara groundnut shoots ranged from −28.24 to −26.20 ‰ across the 26 farmers’ fields studied (Table 5).
Bambara groundnut plant density in farmers’ fields was found to correlate positively with N-fixed (r = 0.15*), with %N (r = 0.31***) and with δ15N (r = 0.126***). The δ13C of Bambara groundnut shoots plus pods also correlated positively with dry matter (r = 0.14*), with C content (r = 0.15*), with N content (r = 0.14*), and with N-fixed (r = 0.15*). Furthermore, N-fixed correlated positively with DM yield (r = 0.76**) and with C content (r = 0.76**), just as shoot N content correlated positively with DM yield (r = 0.86*), with C content (r = 0.0.95**), and with δ13C (r = 0.0.14*).