Legume N₂ fixation is important for sustainable cropping systems in both tropical and subtropical environments, where the soils are naturally deficient in N (Graham and Vance, 2003). The legume-rhizobia symbiosis can provide symbiotic N for meeting the N demand of legumes and, in so doing, spare endogenous soil N for use by cereal crops, in addition to increasing soil N supply through decomposition of legume residues and improving soil structure and organic matter content (Maingi et al., 2000).

Several studies have assessed N nutrition in common bean (Farid and Navabi, 2015; Farid et al., 2016; Jiang et al., 2020), and symbiotic N₂ fixation has generally been found to be lower in common bean when compared to other legumes (Farid et al., 2016). However, applying Rhizobium strain HB-429 to common bean crop increased plant growth, percent N derived from fixation (%Ndfa), the amount of N-fixed, and grain yield by 19, 17, 54, and 48%, respectively, over the uninoculated control in Ethiopia (Samago et al., 2018). These findings show that supplying rhizobial inoculants can replace chemical N fertilizer use and increase bean yields in Ethiopia. An additional benefit following grain harvest in legumes is the release of fixed-N by the legume residues into the soil, which is generally regarded as slow and can match plant uptake from the soil with symbiotic N supply by residues (Chikanai et al., 2018), thus reducing leaching, runoff, and environmental pollution, which are common with added N fertilizer.

In an attempt to increase grain yield and improve household food security, the South African government embarked on a program of providing free inputs, such as improved seeds and the fertilizers NPK, urea, and limestone ammonium nitrate (LAN), for use by resource-poor farmers instead of rhizobial inoculants. Although N fertilizers can be used to overcome the inherently low nutrient levels in African soils and increase crop yields, they are expensive and inaccessible to resource-poor farmers, hence the free N supply to poor farmers by the South African government. Although government supply of fertilizers has increased the production of maize and common bean, in particular, the two major staple foods in South Africa, no study has evaluated the effect of frequent N fertilizer use on plant growth and N₂ fixation of common bean on the fields of farmers. However, the suppressive effects of N fertilizers on legume N₂ fixation is known. For example, high concentrations of NH₃ can inhibit nod-gene expression, nodule formation, and nitrogenase activity in legumes (Vieira et al., 1998; Li et al., 2009), thus reducing their symbiotic N yield. Nitrate is also known to reduce the oxygen diffusion barrier and inhibit nodule functioning via the formation of nitrosylleghemoglobin, which reduces oxygen supply to respiring bacteroids for ATP formation (Dakora and Atkins, 1989).

Several techniques are available for estimating legume N₂ fixation in the field, which include the ¹⁵N natural abundance technique and the ureide assay. Although the ¹⁵N natural abundance technique is robust in quantifying N₂ fixation in legumes (Unkovich et al., 1994; Mohale et al., 2013), it assumes plant uptake of only two sources of N (soil N and atmospheric N₂), which therefore limits its application in situations where external N is added to the soil (Unkovch et al., 2008). Even if there are non-fixing, non-legume plant species growing on the field and sourcing N from the soil and added fertilizer, the ¹⁵N natural abundance technique can still be used to estimate N₂ fixation in the legume species, such as the N-fertilized common bean plants on the fields of farmers. Additionally, tropical legumes belonging to the tribe Phaseoleae (e.g., common bean) export their fixed-N from nodules to shoots in the form of ureides (allantoin and allantoic acid) via the xylem stream (Herridge and People, 1990; Dakora et al., 1992). The concentration of ureides relative to other nitrogenous solutes (nitrate and amino acids) in the xylem sap, stems, or petioles of the tribe Phaseoleae is a reliable measure of N₂ fixation (Herridge and People, 1990; Unkovich et al., 2008) that is not limited by external N inputs.

The application of N fertilizer to symbiotic legumes can decrease nodulation, N₂ fixation, and even grain yield. We hypothesize that the program of providing free N fertilizers (NPK, urea, and LAN) by the South African government to resource-poor farmers could increase grain yield but decrease nodulation and N₂ fixation in beans. This study aimed to assess the effect of N fertilization on plant growth, N₂ fixation, and grain yield of common bean planted without rhizobial inoculation on the fields of farmers in the Eastern Cape Province of South Africa, using the ¹⁵N natural abundance technique and ureide assay.

The South African government policy of providing free N fertilizers to resource-poor farmers has had a positive impact on food security, especially among rural households. Both maize and common bean yields rose with the supply of free fertilizer inputs to emerging small-scale farmers (Figure 1D). In this study, the low endogenous soil N concentrations (0.030–0.033%) found in the unfertilized fields of the Eastern Cape do justify the program of supplying free fertilizer by the government to resource-poor farmers. However, %Ndfa of 84.6, 90.4, and 97.1%, and symbiotic N contribution of 163.3, 161.3, and 140.3 kg ha⁻¹ by unfertilized common bean plants at Lujecweni fields 2, 3, and 4, respectively, when compared to their N-fertilized counterparts, suggest that the use of chemical N fertilizers in common bean production in South Africa is not sustainable due to (i) the high symbiotic performance of unfertilized bean plants relative to their fertilized counterparts, (ii) the high cost of N fertilizers, and (iii) the negative impact of N fertilizers on the environment with frequent use. Exploiting symbiotic N for increased dry bean production in South Africa is therefore a better and sustainable alternative to the use of chemical N fertilizers for increased food and nutritional security.

Additionally, in this study, N-fertilization had a directly negative effect on the symbiotic process of common bean plants. For example, with the unfertilized bean plants, there was a significantly positive correlation between nodule number and nodule fresh weight (r = 0.82^***) and between %Ndfa measured from ¹⁵N natural abundance and % relative ureide-N abundance (r = 0.96^***). Furthermore, the % relative ureide-N was positively correlated with nodule number (r = 0.56^*) and nodule fresh weight (r = 0.82^***), a finding consistent with an uninhibited functional symbiosis. The significantly positive relationship between % relative ureide-N abundance and nodule fresh weight (r = 0.50^***) and between % relative ureide-N abundance and %Ndfa (r = 0.75^***) in N-fertilized common bean plants suggests that the N-fertilized common bean plants were still able to fix N₂ for their growth as they could achieve up to 60% relative ureide-N abundance with a single N-dose application. Tsai et al. (1993) and Tahir et al. (2009) reported that the use of some starter N at planting was beneficial in achieving a synergistic effect on N₂ fixation in common bean by stimulating nodule formation, nitrogenase activity, and plant growth, especially in infertile soils. However, in this study, the % relative ureide-N abundance was negatively correlated with shoot DM (r = −0.31^*), an indication of inadequate symbiotic N supply to shoots. Several studies have shown that plant growth and grain yield of legumes in Africa are directly correlated with symbiotic functioning, and that grain legumes with high levels of N₂ fixation generally elicit greater biomass accumulation, leading to higher grain yield, and vice versa (Belane and Dakora, 2009, 2010; Peoples et al., 2009; Mohale et al., 2013).

The low relative ureide-N (9.2%) observed at Nomhala field 3 that received a double N-dose application (Table 3) shows that, in this study, farmer application of N fertilizers suppressed N₂ fixation in common bean (Vargas et al., 2000; Salvagiotti et al., 2008; Li et al., 2009; Reinprecht et al., 2020); as a result, those bean plants had to complement symbiotic N supply with greater soil N uptake in order to meet their N demand (Van Vessel and Hartley, 2000; Adu-gyamfi et al., 2007). This argument is reinforced by the high nitrate concentration found in stems and petioles of the common bean plants receiving a single and a double dose of N-fertilizer. In fact, Unkovich et al. (2008) have shown that the N solutes in the xylem stream and aqueous stem extracts of members of the tribe Phaseoleae can change from one dominated by ureides in N₂-dependent plants to a system dominated by nitrate and amino acids in plants utilizing soil N. This could therefore also explain the strong inverse relationship found between relative ureide-N abundance and tissue NO3- concentrations of the bean plants in this study. Given that the ureide technique provides an instantaneous measure of legume N₂ fixation at the time of sampling (Alves et al., 2000; Unkovich et al., 2008), the higher ureide-N concentration in stems than petioles could suggest that the stem extract contained xylem ureides that were in transit from root nodules to the leaves, the major sink for fixed-N assimilation.

In this study, although N fertilization markedly suppressed nodulation and nodule symbiotic functioning in bean plants (Tables 3, 4), its application as a single dose generally increased grain yield over either double-dose application or zero-N fertilization (Figure 1D). The fact that double-dose application increased shoot growth at the expense of grain yield is consistent with a recent report by Boddey et al. (2017), who observed higher vegetative growth and lower grain yields in cowpea receiving N fertilizer. The findings of this study therefore suggest that, if N must be applied to bean plants, a single dose supply could help to increase yields while reducing government expenditure to resource-poor farmers. However, the fact that symbiotic dependence of bean plants was markedly decreased by a single dose of N fertilizer applied suggests the need to explore the possibility of achieving greater symbiotic N nutrition and increased grain yield at lower starter N doses in low N soils.

A comparison of the data obtained for symbiotic performance from measuring N₂ fixation using the ¹⁵N natural abundance technique and ureide assay revealed similar trends for plants sampled from the fields of the same farmer, an indication of their precision in measuring N₂ fixation in bean plants on the fields of farmers. For example, %Ndfa by unfertilized bean plants at Lujecweni fields 2, 3, and 4 were 84.6, 90.4, and 97.1%, respectively, using the ¹⁵N natural abundance method, values comparable to the 92.3, 91.6, and 90.2% relative ureide abundance estimated using the ureide technique. Similarly, symbiotic N contribution by unfertilized common bean plants at Lujecweni fields 2, 3, and 4 using the ¹⁵N method was 163.3, 161.3, and 140.3 kg ha⁻¹ respectively, with values similar in magnitude to the total ureides in stems and petioles of 2.0, 1.6, and 1.1 mg g⁻¹ (Tables 3, 4). These trends indicate that the two techniques were similar in their precision and robustness in estimating N₂ fixation in the bean plants on the fields of farmers.

In this study, %Ndfa and ureide levels in stems/petioles were also comparatively similar in trend when the two techniques were used on N-fertilized bean plants from Lujecweni field 1, Mfabantu fields 1 and 2, St. Luthberts field 1, and, to some extent, Tikitiki fields 1 and 2 (Tables 3, 4), and therefore this further confirmed the robustness of the two methods used. However, we also found that the values of %Ndfa at Lujecweni field 1 (51.8%) and Tikitiki field 1 (53.3%) were very high for plants receiving a double dose of N fertilizer and quite high for Tikitiki field 2 (50.1%), Mfabantu fields 1 and 2 (45.5 and 79.9%), and St. Luthberts field 1 (58.9%), which received a single dose of N fertilizer. Some legume symbioses are tolerant of nitrate (Dakora 1998). The only possible explanation for this observation would be that the rhizobia, and therefore bacteroids in root nodules of these N-fertilized bean plants, had a constitutive and/or inducible nitrate reductase activity. Rhizobia with high expression of constitutive nitrate reductase or inducible nitrate reductase genes can form root nodules in the presence of nitrate, but nitrate reductase-minus strains cannot (Serrano and Chamber, 1990). Similarly, bacteroids with a high expression of constitutive nitrate reductase and inducible nitrate reductase genes can fix N₂ in the presence of nitrate, but nitrate reductase-minus strains cannot (Serrano and Chamber, 1990). Therefore, the observed differences in N₂ fixation of bean plants supplied with high doses of mineral N on the fields of farmers could be attributed to the presence of native soil rhizobia with the ability to reduce mineral N and convert N₂ into NH₃ in root nodules, following differentiation into bacteroids.

In conclusion, this study has shown that government supply of free N fertilizers to resource-poor farmers in South Africa increased bean yields for food/nutritional security. Although single- and double-dose N fertilizer application increased plant growth and grain yield compared to unfertilized bean plants, single-dose N fertilizer application produced a much greater grain yield than double dose. This suggests that farmers should stop using a double dose of N fertilizers on bean production, as it decreases yields and can potentially pollute the environment. However, relative to the N-fertilized bean plants, their unfertilized counterparts produced greater nodule fresh weight, higher N derived from fixation, increased amount of symbiotic N, higher percent relative ureide-N abundance, and greater ureide concentration in stems and petioles. Furthermore, contrary to the hypothesis of reduced N₂ fixation with N supply, %Ndfa was very high for some bean plants receiving a double dose of N fertilizer [e.g., Lujecweni field 1 (51.8%) and Tikitiki field 1 (53.3%)] and high for those receiving a single dose of N fertilizer [e.g., Tikitiki field 2 (50.1%), Mfabantu fields 1 and 2 (45.5 and 79.9%), and St. Luthberts field 1 (58.9%)]. Although this was not assessed in our study, it is likely that the rhizobia that effectively nodulated the N-fertilized bean plants and fixed considerable amounts of symbiotic N had constitutive and/or inducible nitrate reductase genes for reducing nitrate in nodules and bacteroids, hence their ability to form root nodules and derived high percent N from fixation in bean with added N.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

SH and SM collected the samples. SH analyzed data and drafted the manuscript. FD conceived the idea, edited and approved the final version of the paper. All authors contributed to the article and approved the submitted version.

This work was supported with grants from the Bill and Melinda Gates Foundation Project on Capacity Building in Legume Sciences in Africa, the South African Department of Science and Technology, the Tshwane University of Technology, the National Research Foundation in Pretoria, and the South African Research Chair in Agrochemurgy and Plant Symbioses. SH is grateful for a competitive masters fellowship from the Bill and Melinda Gates Foundation Project on Capacity Building in Legume Sciences in Africa.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.