Biological Nitrogen Fixation of Cowpea in a No-Till Intercrop under Contrasting Rainfed Agro-Ecological Environments

Abstract

Nitrogen (N) availability under no-till intercropping systems has not been widely investigated in diverse agro-ecological regions in Limpopo Province. Two seasons of rainfed experiments were conducted during 2018/19 and 2020/21 in a 2 × 4 × 2 factorial design to measure the biological nitrogen fixation (BNF) ability of cowpea in an intercropping system with four grain sorghum cultivars at two test locations, Ofcolaco and Syferkuil, of Limpopo Province using the natural abundance technique. The cowpea nitrogen isotope composition (δ ¹⁵N‰) ranged from 0.2 ‰ to 4‰ at Ofcolaco, whereas at Syferkuil, the range was 2 ‰ to 7 ‰. The N derived from air (Ndfa) was from 35% to 92% at Ofcolaco and 4% to 70% at Syferkuil during the two cropping seasons. The amount of N₂ fixed across locations and seasons ranged from 1 kg ha⁻¹ to 71 kg ha⁻¹. In the intercropping system, cowpea fixed more N at higher densities compared with lower densities at the two experimental sites. Biomass was significantly correlated with N accumulated by cowpea (r² > 0.9) at all locations and across seasons. N accumulated in sole cultures was 30% more at Ofcolaco and 36% more at Syferkuil compared with binary cultures. Furthermore, the treatment combination, cowpea density, as well as cropping system, significantly affected N fixation and accumulation. For high productivity, cowpea intercrop with grain sorghum cultivar Enforcer is recommended, as both crops complemented each other when intercropped. The research should investigate further the root distribution and biomass production of sorghum and cowpea, as well as their impact on N intake.

1. Introduction

Food insecurity in Sub-Saharan Africa has become one of the major topics that researchers are paying attention to in recent years. Some of the major staple crops including grain, sorghum, and cowpea have declined in terms of production due to extreme weather events such as drought and heat waves [1]. Modern agricultural production practices such as conventional tillage and excessive application of inorganic fertilizers have resulted in unfavorable growing environments, resulting in low crop productivity [1,2]. A no-tillage system involves minimal soil disturbance which results in the conservation of soil moisture and improved soil structure, compared with a tillage system which disturbs the soil and leads to the emission of greenhouse gases [3]. Nitrogen (N) is one of the essential, yet limiting, nutrients that is required by plants in large quantities for increased growth and productivity [4]. Crop plants access N from different sources including soil, biological nitrogen fixation, and inorganic and organic fertilizers, with the inorganic N fertilizer source being the common practice in Africa [5]. The dependence on inorganic fertilizers has proved to be ineffective due to increasing fertilizer costs which weaken the revenues of smallholder farmers [6] and the potential for long-term soil pollution [7]. Deployment of sustainable agricultural practices is required to improve soil fertility while increasing the N cycle and availability in the soil.

The intercropping system, the practice of planting two or more crops simultaneously on the same piece of land, has been widely adopted as a sustainable practice in Africa. The practice is known for its ability to efficiently utilize nutrients in the soil while enhancing land productivity [8]. The intercropping system is known for its ability to reduce soil erosion, enhance soil moisture, and utilize nutrients efficiently [9]. Enhancing N availability through biological nitrogen fixation (BNF) in the intercropping system is important in areas with poor soil fertility that limits crop productivity. Cowpea is one of the leguminous crops widely grown for its grain and livestock feed in many rural areas of South Africa, particularly, the Limpopo Province. The crop, like other grain legumes, is generally grown by resource-poor farmers on soil with poor soil fertility due to its ability to withstand harsh conditions [10]. The study conducted by [11] explained that legumes such as groundnut and cowpea increase N by up to 52% in the soil as a result of the efficient use of N, which may be beneficial to cereal crops in intercrop or rotation systems. Hence, BNF plays a key role in the provision of N to legumes as well as its non-leguminous component crop in the intercropping system [5]. Biological nitrogen fixation by legumes in an intercropping system increases the availability of N, which is beneficial in areas with poor soil fertility [12]. In the semi-arid regions of Limpopo Province, the knowledge of cowpea–sorghum intercrops is limited due to the method of planting the two crops which growers have adopted. Most cowpea–sorghum grown in the area are broadcasted, which makes it difficult to know how much N is being fixed and transferred to cereal crop. Hence, intercropping studies on N fixation and accumulation such as this one are required under the semi-arid conditions of Limpopo Province.

Most cropping systems studies in the Limpopo Province of South Africa have focused on BNF in conventional sole cropping or maize–legume intercrops and little on grain sorghum as the non-legume component [13,14]. The study conducted in South Africa on cowpea BNF as influenced by sorghum intercropping focused mainly on different cultivars of cowpea under one agro-ecological condition [15]. Furthermore, not much has been done on how N₂ fixation by cowpea is influenced by different cultivars of grain sorghum in a no-till intercropping system. Therefore, studying the extent of benefit in terms of biological nitrogen fixation derived from this intercropping system will assist in managing nitrogen constraints on smallholder farming systems in the Limpopo Province. The information will also contribute to understanding the impacts of BNF of cowpea on component grain sorghum cultivars in the system. For successful crop growth and productivity, growers need to sustain soil fertility and conserve moisture. No-tillage focuses on minimum or little soil disturbance for crop production. There is also limited information on the influence of different populations of cowpea on BNF under different climatic conditions in Limpopo Province. Hence, this paper aimed at (i) comparing the BNF of cowpea in binary and sole cultures under no-tillage, (ii) assessing the effect of different cowpea densities on N₂ fixation, and (iii) measuring the quantity of biological N₂ fixed by cowpea when intercropped with different cultivars of sorghum under contrasting agro-ecological conditions.

2. Materials and Methods

2.1. Experimental Sites

This study was conducted at two distinct agro-ecological zones in Limpopo Province, the northernmost part of South Africa. The first location was the University of Limpopo Experimental Farm, Syferkuil, situated at geographical coordinates of 23°50′02.7″ S and 29°41′25.5″ E. The area received an annual rainfall of about 350 to 500 mm and average maximum and minimum temperatures of 30 °C and 15 °C, respectively. The second location was the Itemeleng Ba-Makhutjwa Primary Cooperative farm at Ofcolaco, located at 24°06′38.3″ S and 30°23′11.8″ E near Tzaneen town, with an annual rainfall of approximately 650 to 700 mm and average maximum and a minimum temperature of 35 °C and 18 °C, respectively. The soil texture at Syferkuil is sandy-clay with a pH of 6.35 [16] and that at Ofcolaco is clay-loam with a pH of 6.06 [17]. The weather parameters during growing seasons were recorded from the two automatic weather stations, situated at the experimental sites at Syferkuil and 27.9 km from the experimental site at Ofcolaco to provide daily weather data during the period of experimentation. A rain gauge was placed at the experimental plot at Ofcolaco to improve seasonal rainfall data collection.

2.2. Experimental Design and Management

Before establishing the experiment, the land at both locations was prepared by first reducing the size of weeds using a motorized slasher, followed by the application of Roundup, a non-selective, systematic, broad-spectrum glyphosate-based post-emergence herbicide one month after slashing. Round-up was applied once at a concentration of 250 mL in 10 L of water. To allow minimal soil disturbance, rows were opened only where the seeds were planted after clearing the land. The rest of the land was left undisturbed. The experiment was laid out in a randomized complete block design in a 2 × 4 × 2 factorial arrangement, replicated four times under rainfed conditions. The factors studied were two cropping systems (intercrop and sole), four sorghum cultivars (Avenger, Enforcer, Titan and NS5511), and two cowpea densities of 37,037 p ha⁻¹ (low) and 74,074 p ha⁻¹ (high) under no-till dryland conditions. Cowpea was planted at an inter-row spacing of 0.9 m × 0.3 m to get a density of 37,037 p ha⁻¹ and 0.9 m × 0.15 m in-row spacing to get a density of 74,074 p ha⁻¹ [18].

Table 1. The code and description of the treatments studied.

Treatments Code	Description
BW2-NSS	High density cowpea intercrop with NS5511
BW1-TIT	Low density cowpea intercrop with Titan
BW1-NS	Low density cowpea intercrop with NS5511
BW1-Sole	Low density cowpea sole
BW2-AVE	High density cowpea intercrop with Avenger
BW2-TIT	High density cowpea intercrop with Titan
BW2-Sole	High density cowpea sole
BW1-AVE	Low density cowpea intercrop with Avenger
BW2-ENF	High density cowpea intercrop with Enforcer
BW1-ENF	Low density cowpea intercrop with Enforcer

shows the treatments codes used in the study as well as their full description names.

Each experimental unit was 3.0 m × 3.6 m, consisting of four rows of sorghum and four rows of cowpea in the intercropped treatment. For grain sorghum, seeds were planted at inter- and intra-row spacings of 0.9 m and 0.3 m, respectively. The cowpea was planted simultaneously between grain sorghum rows at an inter-row spacing of 0.9 m and an intra-row spacing of 0.3 and 0.15 m to obtain the treatment densities of 37,037 and 74,074 plants ha⁻¹, respectively. The spacing between sorghum and cowpea in the intercropped treatment was thus 0.45 m. The details of experimental design and management are also outlined in [17]. Each experimental unit received phosphorus in the form of superphosphate (10.5% P) at 20 kg P ha⁻¹, based on pre-plant soil fertility analysis. Nitrogen was applied as Limestone Ammonium Nitrate (LAN) (28% N) at a rate of 100 kg N ha⁻¹ in a split application of 50 kg N ha⁻¹ each at planting and knee height of grain sorghum. All fertilizers were banded along the row. Standard crop management practices including thinning, weeding, and pest control for both crops were monitored and addressed when necessary throughout the cropping season. Aphids and stalk borer infestation in cowpea and grain sorghum were controlled using Cypermethrin 200 EC at a concentration of 120 mL of Cypermethrin diluted in 64 L of water.

2.3. Sampling Details

The cowpea and sorghum’s aboveground biomass was collected 63 and 83 days after planting, respectively, in intercropping and sole cropping systems from an area of 2.7 m² during the two cropping seasons at both locations. Collected samples were oven-dried at the temperature of 65 °C until they reached constant mass. Biomass was determined by weighing dried plant samples (aboveground) using a weighing balance. Grain yield was determined by harvesting grain from middle rows at 2.4 m². The samples were ground using a Wiley mill to pass through a 2 mm sieve and taken to the laboratory for analysis. Grain yield was used to calculate the land equivalent ratio using the following formula:

$$\mathrm{LER}=\frac{\mathrm{YSbinary}}{\mathrm{YSsole}}+\frac{\mathrm{YCbinary}}{\mathrm{YCsole}}$$

(1)

where: YSbinary is the yield of sorghum in intercropping, YSsole is the yield of sorghum in sole culture, YCbinary is the yield of cowpea in intercropping, and YCsole is the yield of cowpea in sole culture.

2.3.1. ¹⁵N Natural Abundance

The total nitrogen (%) and the natural abundance of ¹⁵N (δ ¹⁵N‰) were analyzed in the laboratory using an isotope ratio mass spectrometer with a N analyzer (IRMS)-N 20-20 ANCA GSL. The following standard formula was used to determine the natural abundance of ¹⁵N as described by [4]:

δ ¹⁵N‰ = [(R_sample / R_standard) − 1] × 1000

(2)

where δ¹⁵N_sample = [(R_sample/R_standard) − 1] × 1000 and where: δ¹⁵N_sample is the value of the δ¹⁵N of the N₂ fixing plant (cowpea), R_sample is the sample isotope ratio (¹⁵N/¹⁴N), and R_standard is ¹⁵N/¹⁴N for atmospheric N₂ (0.0036765).

2.3.2. Biological Nitrogen Fixation (BNF)

BNF was determined using N derived from the atmosphere (%Ndfa) using the ¹⁵N natural abundance method as shown in the formula below:

%NDFA = (δ¹⁵N_{referenceplant} − δ¹⁵N_legume)/(δ¹⁵N_{referenceplant} − β)

(3)

where: δ¹⁵N_{referenceplant} is the value of the δ¹⁵N of the N taken up from the soil, obtained in leaves of the non-fixing reference (sorghum), δ¹⁵N_legume is the value of the δ¹⁵N of the N₂ fixing plant (cowpea), and β is the ¹⁵N value of cowpea leaves.

2.3.3. NDFA Is Nitrogen Derived from the Air

Nitrogen accumulated and fixed were determined using the following formulas:

$$N accumulated=\%N\times Biomass$$

(4)

$$N fixed=\left(Ndfa\times N accumulated\right)/100$$

(5)

2.4. Statistical Analysis

The relevant model assumptions, including normality, independence, and constant variance, were checked before data analysis. The Statistical Analysis System (SAS) version 9.4 was used to analyze data collected using a multivariate analysis of variance (ANOVA) model. Mean separation was done where the means were different, using the least significant difference (LSD) at probability levels of p ≤ 0.05.

3. Results

3.1. Soil Sampling Results Collected after Physiological Maturity

The results in

Table 2. Post-harvest soil organic carbon (mg kg⁻¹), C:N ratio, phosphorus (mg kg⁻¹), and potassium (mg kg⁻¹) collected at the two test locations during season 1 and season 2.

	Ofcolaco
	2018/19				2020/21
Treatment	Org.C	C:N	P	K	Org.C	C:N	P	K
BW2-Sole	1.69a	38.50b	35.97d	213.13a	1.52a	52.23b	42.35	160.82
BW1-AVE	1.60ab	61.86b	115.85a	203.95ab	1.39a–c	70.78ab	65.06	137.01
BW1-TIT	1.58ab	75.03ab	80.9a–c	113.61d	1.34b–d	48.11b	67.47	161.08
BW1-NS	1.55ab	77.20ab	80.95a–c	150.32b–d	1.40a–c	46.33b	48.58	105.15
BW1-Sole	1.54ab	44.50b	34.61d	174.73a–c	1.32cd	79.18ab	46.23	129.60
BW2-NS	1.53ab	61.21b	43.89cd	160.49a-d	1.29cd	99.10ab	48.69	100.62
BW1-ENF	1.52ab	44.83b	57.12b–d	137.35cd	1.51ab	45.75b	37.05	120.58
BW2-AVE	1.46b	59.58b	49.56b–d	121.51cd	1.49ab	39.99b	61.36	130.91
BW2-ENF	1.44b	74.11ab	85.07ab	168.51a–c	1.20d	77.15ab	33.78	94.41
BW2-TIT	1.40b	110.20a	60.49b–d	156.23b–d	1.41a–c	130.20a	43.29	84.45
p ≤ 0.05	*	*	*	*	*	*	ns	ns
Syferkuil
BW2-Sole	0.52	8.99b	22.179	297.65ab	0.89	11.07bc	27.13c	310.64ab
BW1-AVE	0.59	12.39b	17.967	285.29a–c	0.80	16.48a	54.31a	281.39a–c
BW1-TIT	0.76	11.09b	21.428	246.50a–c	0.84	13.08a–c	52.53ab	274.36a–c
BW1-NS	0.79	11.54b	14.754	247.57a–c	0.68	10.35c	32.84a–c	246.92bc
BW1-Sole	0.57	11.94b	35.936	310.66a	0.91	14.57ab	28.63bc	335.3a
BW2-NS	0.56	27.06a	26.109	261.89a–c	0.80	11.79a–c	42.71a–c	264.32bc
BW1-ENF	0.50	14.54b	32.409	260.24a–c	0.67	9.26c	54.86a	220.49c
BW2-AVE	0.72	12.21b	31.442	225.09bc	0.73	15.23ab	31.58a–c	264.34bc
BW2-ENF	0.60	11.58b	20.9	280.53a–c	0.78	13.17a–c	40.09a–c	269.35a–c
BW2-TIT	0.66	11.07b	37.318	199.81c	0.68	12.09a–c	52.71ab	223.61c
p ≤ 0.05	ns	*	ns	*	ns	*	*	*

The means followed by the same letter were not different at p ≤ 0.05. ns = not significant, * = significant at p ≤ 0.05.

indicate soil parameters under different treatments after harvest at Ofcolaco and Syferkuil during each cropping season. There were no significant interactions for cropping system, treatments, and density during the two growing seasons at Ofcolaco and Syferkuil. According to the results, the treatments studied differed in organic carbon, C:N ratio, phosphorus, and potassium at Ofcolaco during the two growing seasons, except for P and K from 2020/21 (Table 2). Treatment BW2-Sole had higher organic carbon and K than treatments BW2-AVE, BW2-ENF, and BW2-TIT, with the means of 1.69% and 213.13 mg kg⁻¹, respectively, in the 2018/19 growing season. In the 2020/21 growing season, organic carbon ranged from 1.20 mg kg⁻¹ to 1.52 mg kg⁻¹, with treatment BW2-Sole being superior to all other treatments. At Syferkuil, organic carbon and P were not significantly different among the treatments in 2018/19, and so was the organic carbon in the 2020/21 growing season. The C:N ratio and K were significantly different among the cowpea treatments at Syferkuil during the two cropping seasons. Treatment BW2-NS had the highest C:N ratio of 27.06, whereas BW2-Sole had the lower C:N ratio of 8.99 in 2018/19. K ranged from 199.81 mg kg⁻¹ to 297.65 mg kg⁻¹. Treatment BW-AVE had the highest C:N ratio in the 2020/21 season with a mean of 16.48. Treatment BW1-ENF obtained higher P with the lowest organic carbon and K.

3.2. Biomass Accumulation

Biomass was highly significant (p < 0.001) among the treatments at Ofcolaco and Syferkuil during the 2018/19 and 2020/21 growing seasons (

Table 3. Cowpea total dry biomass accumulation (kg ha⁻¹) of cowpea treatments from Ofcolaco and Syferkuil during season 1 and season 2.

	Ofcolaco		Syferkuil
	2018/19	2020/21	2018/19	2020/21
Treatment	Biomass
BW2-NS	3145.7ab	1314.2d	2217.3a	2166.0bc
BW1-TIT	1459.3e	1714.8b–d	903.7de	2034.6bc
BW1-NS	1932.9de	1584.0cd	888.9de	1503.7cd
BW1-Sole	1974.1c–e	2496.3ab	1006.8c–e	2042.6bc
BW2-AVE	3022.2a–c	1870.4b–d	1506.2bc	1867.3cd
BW2-TIT	2621.4b–d	2464.2a–c	1454.3b–d	2877.8ab
BW2-Sole	3701.2a	2850.6a	1924.7ab	3566.0a
BW1-AVE	1387.7e	2468.3a–c	615.4e	1166.7d
BW2-ENF	3392.6ab	2050.0a–d	1516.0bc	2204.3bc
BW1-ENF	2004.9c–e	1582.1cd	785.8e	1597.7cd
p < 0.001	***	***	***	***

The means followed by the same letter were not different at p ≤ 0.05. ns = not significant, *** = significant at p < 0.001.

), but the interaction effects of treatment factors were not significant. Treatment BW2-Sole was superior in terms of biomass accumulation at Ofcolaco during the two growing seasons with the means of 3701.2 kg ha⁻¹ and 2850.6 kg ha⁻¹ for season 1 and season 2, respectively. All treatments under intercrop with high cowpea density accumulated a higher biomass compared to the intercrop low density at Ofcolaco during the 2018/19 growing season. However, in 2020/21, the intercropping treatments that accumulated the highest biomasses were BW2-TIT and BW1-AVE, with biomasses of 2468.3 kg ha⁻¹ and 2464.2 kg ha⁻¹.

3.3. Land Equivalent Ratio

The productivity of intercropping (LER) ranged from 1.2 to 1.9 at Ofcolaco (Figure 1). Among the intercrops at Syferkuil, BW2-NS treatment had the highest biomass accumulation of 2217.3 kg ha⁻¹ in the 2018/19 growing season. In the 2020/21 growing season, BW2-Sole and BW2-TIT accumulated highs of 3566.0 kg ha⁻¹ and 2877.8 kg ha⁻¹, respectively. The LER ranged from 1.3 to 1.7 in the 2018/19 and 1.3 to 1.8 in the 2020/21 growing seasons (Figure 1).

3.4. N Fixed and N Accumulated

The cropping system did not affect the amount of nitrogen fixed by cowpea in the 2018/19 growing season at Ofcolaco, but in 2020/21, the effect was significant (Figure 2). Cowpea fixed more N in sole compared with binary cultures with the means of 67.95 kg ha⁻¹ and 48.73 kg ha⁻¹ in the 2020/21 cropping season. Furthermore, the results indicated that at Ofcolaco, cowpea was able to fix 28% more N in sole compared with the binary culture in the same season. At Syferkuil, the amount of N fixed by cowpea was not significantly affected by the cropping system during the two cropping seasons as indicated by Figure 2. However, N fixed was between 19.08 kg ha⁻¹ and 21.73 kg ha⁻¹ at the same location during the two seasons.

The results revealed no significant difference in nitrogen accumulated by cowpea in sole and binary cultures at Ofcolaco and Syferkuil during the 2018/19 cropping season (Figure 3). However, the cropping system had a significant effect on N accumulated by cowpea at the two locations during the 2020/21 cropping season. At Ofcolaco, the amount of N accumulated was higher in sole compared with binary cultures with means of 82.09 kg ha⁻¹ and 57.65 kg ha⁻¹, respectively. At Syferkuil, cowpea accumulated 84.53 kg ha⁻¹ and 54.47 kg ha⁻¹ in the sole and binary cultures, respectively. According to the results, N accumulated was 30% more at Ofcolaco and 36% more at Syferkuil in sole compared with binary cultures.

The cowpea density had a significant effect (p ≤ 0.05) on N fixed in the 2018/19 cropping season at both Ofcolaco and Syferkuil. However, in the 2020/21 cropping season, the density did not significantly affect N fixed at the two locations. The cowpea density of 74,074 plants ha⁻¹ fixed high N at Ofcolaco and Syferkuil during the 2018/19 cropping season compared with 37,037 plants ha⁻¹ (Figure 4A). In the 2018/19 cropping season, N accumulated significantly varied (p ≤ 0.05) between cowpea densities at the two test locations. However, in the 2020/21 cropping season, N accumulated was not significantly different at Ofcolaco and Syferkuil under high and low density. Cowpea accumulated more N when planted at a density of 74,074 plants ha⁻¹ relative to 37,037 plants ha⁻¹. The amount accumulated was more than half in the high density compared with the low density at Ofcolaco and Syferkuil (Figure 4B).

The amount of nitrogen fixed was different among the cowpea treatments in binary and sole cultures at Ofcolaco and Syferkuil during the two growing seasons. At Ofcolaco, N-fixed ranged from 13.48 kg ha⁻¹ to 62.05 kg ha⁻¹ in 2018/19 and 36.19 kg ha⁻¹ and 70.93 kg ha⁻¹ during the 2020/21 growing season (Figure 5). Treatment BW2-ENF fixed the highest N in 2018/19 compared to all other treatments with a mean of 62.05 kg ha⁻¹. The treatment with low N fixation was BW1-TIT with a mean of 13.48 kg ha⁻¹. In the 2020/21 growing season, treatments BW1-Sole had the highest N fixation of 70.93 kg ha⁻¹ followed by BW1-AVE which fixed 66.03 kg ha⁻¹ of N. Treatments BW2-NS and BW1-ENF fixed the lowest N at Ofcolaco during 2018/19 growing season with the means of 39.55 kg ha⁻¹ and 36.19 kg ha⁻¹, respectively. At Syferkuil, N fixation ranged from 10.87 kg ha⁻¹ to 27.34 kg ha⁻¹ in 2018/19 and 1.84 kg ha⁻¹ and 38.19 kg ha⁻¹ in the 2020/21 growing season (Figure 5). Treatment BW2-ENF had the highest N fixation of 30.23 kg ha⁻¹ compared with all other treatments, followed by BW2-NS with 27.34 kg ha⁻¹ at Syferkuil in the 2018/19 growing season. The treatment which fixed low Nwas BW1-ENF with a mean of 10.87 kg ha⁻¹. In 2020/21, treatments BW2-NS and BW2-TIT had the highest N fixed of 38.19 kg ha⁻¹ and 34.98 kg ha⁻¹, respectively, compared with all other treatments (Figure 5). Treatment BW1-AVE had the lowest N fixed of 1.84 kg ha⁻¹ at Syferkuil during the 2020/21 growing season.

There was a significant variation among the cowpea treatments in binary and sole cultures for nitrogen accumulation at Ofcolaco and Syferkuil during the 2018/19 and 2020/21 growing seasons. Treatments BW2-Sole accumulated the highest N of 117.04 kg ha⁻¹, followed by BW2-AVE, BW2-ENF, BW2-NS, and BW2-TIT with the means of 93.55 kg ha⁻¹, 92.98 kg ha⁻¹, 89.84 kg ha⁻¹, and 75.14 kg ha⁻¹ at Ofcolaco in 2018/19 growing season (Figure 6). In 2020/21, treatments BW2-Sole and BW-Sole were superior compared with all other treatments with the means of 82.00 kg ha⁻¹ and 82.29 kg ha⁻¹. At Syferkuil, during the 2018/19 growing season, treatment BW2-NS accumulated higher N compared with all other treatments followed by BW2-Sole with the means of 65.98 kg ha⁻¹ and 56.97 kg ha⁻¹, respectively (Figure 6). In 2020/21, treatment BW2-Sole had a higher N accumulation of 112.83 kg ha⁻¹ followed by BW2-TIT with a mean of 76.61 kg ha⁻¹.

3.5. δ¹⁵N and %Ndfa

There was a significant variation between binary and sole cowpea treatments on δ¹⁵N‰ and %Ndfa during the 2018/19 growing season at Ofcolaco but not in the 2020/21 growing season (

Table 4. Natural abundance of ¹⁵N (‰) and N derived from the atmosphere (%) of cowpea in binary and sole treatments at Ofcolaco and Syferkuil during season 1 and season 2.

	Ofcolaco				Syferkuil
	2018/19		2020/21		2018/19		2020/21
Treatment	δ15N‰	%Ndfa	δ 15N‰	%Ndfa	δ 15N‰	%Ndfa	δ 15N‰	%Ndfa
BW2-NS	4.22a	35.59c	0.25a	92.96a	7.10a	41.68b	4.24a–c	30.28a–c
BW1-TIT	4.19a	36.14c	0.29a	91.69a	4.65ab	61.88ab	3.60bc	40.92ab
BW1-NS	4.15ab	36.75bc	0.71a	77.60a	5.92ab	51.38ab	3.25bc	46.75ab
BW1-Sole	3.94a–c	39.86a–c	0.45a	86.20a	6.27ab	48.55ab	3.72bc	38.88ab
BW2-AVE	3.84a–c	41.39a–c	0.51a	84.35a	4.28ab	64.95ab	4.29a–c	29.41a–c
BW2-TIT	3.28a–c	50.02a–c	0.67a	78.74a	6.74ab	44.64ab	3.04bc	50.21ab
BW2-Sole	3.22a–c	51.01a–c	0.63a	80.09a	6.61ab	45.71ab	4.85ab	20.10bc
BW1-AVE	2.99a–c	54.57a–c	0.38a	88.68a	4.19ab	65.69ab	5.78a	4.61c
BW2-ENF	2.32bc	64.84ab	0.64a	79.72a	3.58b	70.76a	2.57c	57.90a
BW1-ENF	2.30c	65.23a	0.73a	76.97a	6.05ab	50.34ab	2.69c	55.98a
p ≤ 0.05	*	*	ns	ns	*	*	*	*

The means followed by the same letter were not different at p ≤ 0.05. ns = not significant, * = significant at p ≤ 0.05.

). The intercrop treatments BW1-ENF and BW2-ENF had lower δ¹⁵N of 2.30‰ and 2.32‰ values compared with BW2-NS, BW1-TIT, and BW1-NS at this location, but the values did not differ from the other treatments. The percent N derived from atmospheric fixation ranged from 35.59% to 65.23% across the different sole and intercrop treatments. Statistically, the %Ndfa was similar among most of the treatments except for BW1-ENF and BW2-ENF, which accumulated on average 29 percentage points low nitrogen from fixation compared with the average of BW-NS and BW1-TIT. Although in the 2020/21 cropping season the %Ndfa was similar among the treatments, the intercrop treatments BW2-NS and BW1-TIT showed a tendency for superior %Ndfa with the means of 92.96% and 91.69%, respectively, at this location. Unlike Ofcolaco, the treatments significantly influenced δ¹⁵N and %Ndfa at Syferkuil in both the 2018/19 and 2020/21 growing seasons (Table 4). The δ15N values in 2018/19 ranged from 3.58 to 7.10 and were similar among most of the treatments except the BW2-NS intercrop, which produced a higher δ15N value relative to the BW2-ENF intercrop. The %Ndfa ranged from 41.68 to 70.76 in 2018/2019 and, similar to δ¹⁵N, the values were similar among most of the treatments except the BW2-ENF intercrop, which derived approximately 29 percentage points more nitrogen from fixation compared with BW2-NS. The results further indicated that treatment BW2-ENF, with relatively low δ¹⁵N of 3.58‰, derived high amounts of nitrogen from the atmosphere. In the 2020/21 growing season, most treatments had similar letters. However, the treatment BW1-AVE proved to do better in terms of δ¹⁵N, with 4.29‰, whereas treatments BW1-ENF and BW2-ENF proved to have better Ndfa compared with all other treatments, respectively, with the means of 55.98% and 57.90%.

A strong positive relationship between cowpea’s N accumulation and biomass production was observed at both Ofcolaco and Syferkuil during the two cropping seasons, with R² values of more than 0.9 (Figure 7). These reveal that an increase in N accumulation resulted in cowpea attaining more biomass irrespective of the system or the density. According to the results, for every increase 1.0 kg ha⁻¹ in the N accumulation, cowpea produced approximately 30 kg ha⁻¹ of biomass at each location across all the seasons.

4. Discussion

This study revealed the impacts of the cropping system and season on the measured soil parameter, but the interactive effects of the two factors were also observed. For instance, at Ofcolaco the concentrations of organic C, P, and K in 2018/19 were relatively higher among the treatments compared with the amounts recorded in the 2020/21 growing season. However, at Syferkuil, lower concentrations of organic C, C:N ratio, P, and K were recorded in 2018/19 compared with the 2020/21 growing season. Similar results were reported elsewhere [19,20]. Regarding the natural abundance of ¹⁵N, it has been reported that the low amount of soil P increases the δ¹⁵N of legume crops [21]. This agreed with what was observed in this study, where the treatments with low P concentrations ultimately resulted in higher ¹⁵N compared with treatments with high P content. The authors of [22] reported that a high soil P level stimulates growth and mineral N uptake without the Ndfa. It was also observed from our study that the cropping system and the density of cowpea did not show a significant variation for δ¹⁵N and %Ndfa during the two growing seasons. The finding contradicted what other studies [23,24] reported, indicating significant variation in intercropped and sole cowpea.

According to [25], cowpea derived approximately 34% of its N from the air at Syferkuil, implying that the plant had to rely on N from the soil. The study by [4] reported a decrease in N accumulation of biomass due to low N in the soil. Cowpea fixed more N in the sole than in intercrop cultures due to high biomass production as reported in the findings of this study. Similar results were observed in other intercrop studies [26,27]. However, a study conducted in Kenya contradicted what we observed; the authors reported higher N fixation in intercrop compared with the sole system [28]. The higher N fixed was due to the low N available in the soil, which required the crop to exert more effort in increasing BNF than in intercropping with sorghum where inorganic N fertilizer was applied [29,30]. Cowpea fixed more N in the sole system when the density was 74,074 plants ha⁻¹ compared to 37,037 plants ha⁻¹. Thus, more cowpea plants per row imply more competition for soil leading to relatively low available soil N which results in higher atmospheric fixation and accumulation. Furthermore, High cowpea density enhanced biomass production which ultimately resulted in high grain yield accumulation as explained in [17]. The authors [31] reported a similar observation in the millet-cowpea intercrop. In intercropping, the non-legume crop benefits as a result of the legume crop using less of the residual N in the soil [32]. Cowpea fixed more N in high density intercropping in this study due to increased competition for soil N with sorghum. This could provide an alternative N source to inorganic fertilizer while also lowering grower costs [19].

Although high cowpea density increased the N fixation and accumulation in our studies, these N activities were, however, influenced by the grain sorghum cultivar that was intercropped with the cowpea at both locations and in the two seasons. For instance, in 2018/19 at Ofcolaco, treatment BW2-ENF fixed more N, whereas in 2020/21 it was treatment BW2-NS that was superior compared to all other treatments. The results from [33] revealed that in sorghum–cowpea intercrops, N₂ fixed was 24 kg ha⁻¹. However, the treatment combination, such as the type of cultivar intercropped, plays a significant role as it affects N balance and the profitability of the cropping system. According to studies by [23,34], biological nitrogen fixation promotes growth as well as the grain yield of legumes under seasonal variations. This may explain why the N activities reported in this study were different across locations and seasons. Other authors have also reported variations in N fixation and the accumulation of intercrop and sole cropped grain legumes [35]. The amount of N₂ fixed by cowpea in this study was influenced by cropping practices (intercrop system, grain sorghum cultivar, and cowpea density), as well as soil conditions, which ultimately influenced partition into biomass. The authors of [36] reported similar findings that intercropping legumes with cereals increased soil N mineralization, resulting in more transport of N from the soil to the legume plant and less N₂ fixation. Cowpea fixed less N at Syferkuil and was more dependent on the N available in the soil compared with the plants grown at Ofcolaco.

The strong positive relationship observed in this study between cowpea biomass and N accumulated indicated that high N accumulation resulted in higher biomass production. Other authors also reported a strong positive relationship between biomass and N accumulation [37,38]. Cowpea accumulated higher biomass when grown in sole under high density compared with intercropping under low and high densities. The study conducted by [14] similarly reported high biomass in sole culture compared with intercropping. However, other authors have also contradicted this observation, indicating that higher biomass accumulation occurred in an intercropping system involving faba beans and wheat in a strip intercropping system [39]. Hence, the type of intercropping system, the component crops, as well as crop management practices should be considered for future research on nitrogen dynamics and productivity of the system. The LER of the sorghum–cowpea intercrop ranged from 1.2 to 1.9, indicating that there was a high yield advantage when the component crops were grown in an intercropping system compared with sole cultures. The results agreed with what [40] reported in the sorghum–cowpea intercrop and with [24] in the pea–barley intercrop.

5. Conclusions

Biological nitrogen fixation in an intercropping system is complicated, necessitating extensive research into nitrogen supply and uptake by the two crops. When researching BNF by cowpea, factors such as plant density and companion crop cultivar should be considered. The productivity of grain sorghum–cowpea intercrop was greater than 1.0 at the two locations during the two cropping seasons. This implies a higher yield advantage in intercropping compared with the sole system. In general, cowpea fixed on average a higher amount of nitrogen with 47.99 kg ha⁻¹ at Ofcolaco compared with Syferkuil, which had an average of 20.63 kg ha⁻¹. The results revealed higher biomass, N fixation, and accumulation in the sole culture compared with the intercropping system. In this study, high cowpea density consistently increased N fixation and accumulation relative to low density by 50% or more, which resulted in high biomass accumulation. Cowpea intercropping with Enforcer at higher cowpea density under Ofcolaco and Syferkuil conditions can be considered a desirable way to produce sorghum and cowpea simultaneously due to high N fixation. On average, N fixation was 58.12 kg ha⁻¹ and 34.21 kg ha⁻¹ at Ofcolaco and Syferkuil, respectively. Intercropping cowpea with the grain sorghum cultivar, NS5511, can also be considered, as cowpea performed relatively better in terms of N fixation and accumulation when intercropped at high density than when intercropped with Titan and Avenger. The dynamics of N uptake should be further investigated especially under different cultivars of the companion non-leguminous crop for better recommendations. Nevertheless, the intercropping system can be implemented by smallholder farmers in Limpopo Province for the sustainable production of cereal and legume crops, as the LER observed in this study showed high biomass productivity of the system.