Effect of pearl millet–cowpea cropping systems on nitrogen recovery, nitrogen use efficiency and biological fixation using the ¹⁵N tracer technique

Abstract

Nitrogen is one of the most limiting nutrients for crop production in many areas of Africa. One strategy to improve yields is to choose crops with high nitrogen use efficiency (NUE) that can produce economic yields under limited water supply. Little information is available on the comparative performance of pearl millet (Pennisetum glaucum L. R. Br.) and cowpea (Vigna unguiculata L. Walp.) in sole crops and intercrops systems for the NUE of applied fertilizers. This study was conducted under field conditions at the Senegal Agricultural Research Institute from July to October 2001. Two cropping systems were used and included sole crops of pearl millet and cowpea and a millet–cowpea intercrop. ¹⁵N-labeled urea at rates of 20 kg ha⁻¹ (sole and intercrop cowpea) and 41 kg ha⁻¹ (sole millet and intercrop millet) was applied. Sole millet produced 6,041 kg yield ha⁻¹ and accumulated 95.27 kg N ha⁻¹, of which 15.30% was derived from the nitrogen fertilizer and 84.70% from nitrogen mineralized in soil. Therefore, the NUE of the fertilizer was 36.29%. In intercrop millet, sole cowpea and intercrop cowpea, the NUEs were 15.20, 45.33 and 46.00%, respectively, indicating that the use of nitrogen fertilizer significantly decreased in intercrop only for millet. However, no significant difference was observed between sole and intercrop cowpea relative to the nitrogen derived from N₂ fixation. The amount of nitrogen immobilized in the soil was significantly higher in sole millet than in sole cowpea and millet–cowpea intercrop. The Land Equivalent Ratio (LER) of grains and total dry matter showed an advantage of the millet–cowpea cropping system over sole crops.

Key words:

• biological fixation
• ¹⁵N fertilizer
• nitrogen recovery
• nitrogen use efficiency
• pearl millet–cowpea cropping

INTRODUCTION

Pearl millet (Pennisetum glaucum L. R. Br.) is one of the most important cereals in drought-prone areas and is the staple grain for 150 million people in Africa and India (Food and Agriculture Organization 1997). In the semiarid environment of Senegal, pearl millet (millet) is widely associated with cowpea (Vigna unguiculata L. Walp.) for its potential nitrogen fixation. Cowpea grain contains approximately 22% protein and constitutes a major source of protein for resource-poor rural and urban people. The crop residues from cowpea constitute an important source of livestock feed, especially in the dry savannas of the West African Semiarid Tropics. However, growth and productivity of these crops are limited by temporally and spatially erratic rainfall and poor soil fertility (Davis-Carter 1989). The decrease in soil fertility is related to low replenishments. In this regard, Badiane (1993) reported that crop residues were rarely used for soil replenishment in this area of Senegal. According to this author, 60–100% of crop residues are specially used for livestock feed and domestic needs. Furthermore, in the tropical sandy soil of West Africa, inorganic nitrogen (N) represents the main source for plant nutrition. In these soils, nitrogen reserves were very low and impeded plant growth (Sundarapandian and Swamy 1999). Soil fertility in this zone appears to be more limiting to crop and fodder production than rainfall and the use of fertilizers will increase water-use efficiency (Breman and de Wit 1983). Therefore, the use of mineral fertilizer can significantly increase water-use efficiency. The relationship between nitrogen fertilizer uptake and total nitrogen uptake over a growing season depends on timing of the nitrogen fertilizer application (Guindo et al. 1994) and the amount of available nitrogen in the fertilizer (Bufogle et al. 1997). Enhanced nitrogen use efficiency (NUE) on applied nitrogen fertilizer for greater biomass production is essential in systems where nitrogen availability is often low and limiting for plant growth (Muchow et al. 1993). To intensify the cereal production, additional amounts of nitrogen are necessary to maintain the soil fertility in intercropping systems. Leguminous crops are sources of nitrogen and contribute to an increase in the nitrogen uptake of the non-leguminous associated crop (Fujita et al. 1990). To our knowledge, few studies have dealt with the relationship between the applied nitrogen and the nitrogen uptake by crops in this zone of Senegal. This study was carried out to determine: (1) the rates of nitrogen recovered in sole millet, sole cowpea and millet–cowpea intercropping systems under field experimental conditions, (2) the effect of nitrogen fertilizer on the N₂ fixation of sole cowpea and intercrop cowpea. Data from nitrogen recovery, NUE and N₂ fixation enabled the identification of the most suitable cropping system under nitrogen fertilizer application.

MATERIAL AND METHODS

Study site

This study was carried out at the experimental field station of the Senegal Agricultural Research Institute at Bambey (16°30′–16°28′Ν, 15°44′–15°42′W). The climate of this area is referred to as Sahalian–Sudanese and is characterized by a long dry season (November to June) and a short rainy season for 4 months (from July to October). The amount of annual precipitation between 1971 and 2000 was 465 mm year⁻¹ and the average temperature was 28°C (Diop 2002). The soil is described as a ferruginous soil (Maignien 1965) and classified as a Lixisol (Food and Agriculture Organization 1998). Badiane et al. (2000) stated that the clay fraction, which is essentially composed of kaolinite, had low organic matter content ranging from 0.2 to 0.5%, low cation exchange capacity (1–3 cmol kg⁻¹) and low nitrogen content (0.31 g kg⁻¹).

Plant materials

This study was carried out using millet (Pennisetum glaucum L. R. Br.) and cowpea (Vigna unguculata L. Walp.) varieties. Millet cv. Souna3 and cowpea cv. Melakh were used. These two cultivars were developed by the Senegal Agricultural Research Institute. Millet cv. Souna3 has a growth cycle of 90 days and grows with rainfall levels ranging from 380 mm to 460 mm. Cowpea cv. Melakh has a growth cycle of 60 days.

Experimental design

The experiment was carried out in 2001 from July to October (during the rainy season) in three stands corresponding to sole millet, sole cowpea and millet–cowpea intercrop. Each stand was divided into four blocks containing four plots each. Plot size was 8.5 m × 12.3 m with row spacing and plant spacing of 1 m × 1 m in sole millet and 0.5 m × 0.5 m in sole cowpea. In the intercrop system, two millet rows were sown with 150 cm spacing and two cowpea rows were arranged with 50 cm spacing between the millet rows. Plant spacing within a row was 50 cm for both millet and cowpea in the intercrop plots. In the sole millet, three plots were randomly selected from three different blocks and a subplot of 2 m × 3 m was delimited in each plot. After seedling emergence, ¹⁵N-labeled fertilizer (urea: CO [NH₂]₂) solution prepared from 41 kg N ha⁻¹ at 1 atom%¹⁵N was spread along the millet rows. In the sole cowpea, two subplots of 2 m × 2.5 m were delimited in each of three randomly selected plots. One subplot was delimited above the cowpea rows and received 20 kg N ha⁻¹ at 5 atom%¹⁵N. To estimate the biological nitrogen fixation (BNF) of sole cowpea, millet was grown as a reference plant in the other subplot and received 41 kg N ha⁻¹ at 1 atom%¹⁵N. In the millet–cowpea intercrop, a 2 m × 2.5 m subplot was delimited above the millet and cowpea intercropped rows in each of three randomly selected plots and 41 kg N ha⁻¹ at 1 atom%¹⁵N was applied along the millet rows and 20 kg N ha⁻¹ at 5 atom%¹⁵N was applied along the cowpea rows. For the estimation of intercrop cowpea BNF, intercrop millet was used as a reference plant. The other plots in each stand were fertilized with non-labeled urea at the same rate as the labeled areas. KCl was also applied to the whole experimental area at the following recommended rates: 15 kg K ha⁻¹ in sole cowpea, 10.5 kg K ha⁻¹ in sole millet and in the millet–cowpea intercrop. The difference in the applied nitrogen rates between cowpea (20 kg N ha⁻¹) and millet (41 kg N ha⁻¹) was due to the difference in the recommended application doses in Senegal, relatively to their needs in nitrogen. As the level of nitrogen for cowpea was low, 5 atom%¹⁵N was used to detect enough isotopic nitrogen.

Data collection and analyses

¹⁵N labeled plants were harvested and the dry matter (DM; kg ha⁻¹) of all components was recorded: dry matter of grains, straw and ears for millet and dry matter of seeds, herbage and pods for cowpea. Dry matter accumulation (yield) was used to calculate the Land Equivalent Ratio (LER) using the following formula: LER = (Intercrop millet yield/sole millet yield) + (Intercrop cowpea yield/sole cowpea yield). The LER represents the relative land area required for a sole crop to produce the yields achieved in intercropping (International Rice Research Institute 1974). Plant samples were dried, ground, sieved and used for isotopic analysis. Nitrogen (%) and the ¹⁵N atom% excess were determined using the Kjeldahl method (Bremner and Mulvaney 1982) and an elemental nitrogen analyzer (Carlo Erba NA 1500, Milan, Italy) connected to a Finnigan MAT 251 isotope ratio mass spectrometer (Bremen, Germany), respectively. Samples from the different ¹⁵N fertilizer solutions used in this experiment were also analyzed to record the exact ¹⁵N atom% excess. In addition, the total nitrogen (TN; kg ha⁻¹), nitrogen recovery (kg ha⁻¹), NUE (%), nitrogen in the crop derived from the fertilizer (Ndff; %) and nitrogen derived from the atmosphere (Ndfa; %) were determined. The nitrogen recovery, which corresponds to the amount of harvested nitrogen derived from the applied nitrogen was calculated using the equation: Nitrogen recovery (kg ha⁻¹) = TN × Ndff (%)/100 (International Atomic Energy Agency (IAEA) 2001). The Ndff (%) was calculated using the following equation: Ndff (%) = (¹⁵N atom% excess of plant sample/¹⁵N atom% excess of fertilizer) × 100. The TN and NUE were derived from the following equations: TN (kg ha⁻¹) = (DM × N%)/100 and NUE (%) = (Nitrogen recovery/Applied N) × 100. The proportion of cowpea nitrogen derived from the atmosphere (Ndfa; %) was calculated using the ¹⁵N isotope dilution equation (Fried and Middleboe 1977). Ndfa (%) = 1 – (Ndfffix/Ndffref) × 100, where Ndfffix is the Ndff (%) of the N-fixing plant and Ndffref the Ndff (%) of the non-fixing reference plant. The effect of N fertilizer and the intercropping system on the BNF of cowpea was, therefore, deduced from these results. Using the same procedure, soil samples from the labeled subplots of each stand were analyzed after the experiment to estimate the amount of nitrogen (kg ha⁻¹) immobilized in the soil and derived from the nitrogen fertilizer. The percentage of nitrogen lost in each stand was also estimated. All data were subjected to anova using IRRISAT 4.0 (International Rice Research Institute, Manila, Philippines). When appropriate, least significant difference (LSD_0.05) values were used to compare treatment means.

RESULTS AND DISCUSSION

Nitrogen (%) in the different plant components could not be correlated (P > 0.05) to nitrogen fertilizer application. However, as shown in Table 1 the ¹⁵N atom% excess strongly differed (P < 0.001) among the cropping systems. No significant difference was observed between ¹⁵N atom% excess in the sole and intercrop millet, between ¹⁵N atom% excess in the sole and intercrop cowpea; however, a difference was observed between millet and cowpea ¹⁵N atom% excess. This difference resulted from the difference in ¹⁵N atom% excess of the fertilizer. The whole plant DM accumulations of sole millet and intercrop millet were 6,041 and 3,779 kg ha⁻¹, with corresponding TN values of 95.26 and 53.23 kg ha⁻¹, respectively. In sole and intercrop cowpea, 4,656 and 5,118 kg ha⁻¹ DM was accumulated with corresponding TN values of 98.46 and 110.73 kg ha⁻¹, respectively. The DM accumulation was not significantly different at the 5% level. In contrast, TN accumulation (P < 0.05) was significantly higher in the sole millet intercrop millet, whereas no significant deference was observed in intercrop cowpea (Table 2). Results on Ndff (%) and NUE (%) were significantly different at the 5% level. Although no significant difference appeared between sole and intercrop cowpea, sole millet used nitrogen fertilizer more efficiently than intercrop millet. No significant difference was observed between cowpea and sole millet in terms of the NUE (%) of the nitrogen fertilizer. Data from Table 3 show that the applied nitrogen fertilizer was essentially located in the 0–10 cm soil depth. However, considering the 0–20 cm soil depth (Table 4), which is generally exploited by plant roots, nitrogen fertilizer immobilized in sole millet soil (57.88 kg ha⁻¹) was significantly higher than that of sole cowpea soil (32.16 kg ha⁻¹) and intercrop soil (27.09 kg ha⁻¹). These results are in agreement with those of Diekmann et al. (1993) carried out on rice. These authors reported for rice plants a higher recovery of ¹⁵N-labeled fertilizer (% of nitrogen applied) in rice-fallow rather than in rice-legume intercrop over the growing season. In contrast, as shown in Table 2, the cropping system did not have any effect (P > 0.05) on the Ndfa (%). The Ndfa (%) of the sole and intercrop cowpea did not agreed with experiments carried out on cowpea in Nigeria (Eaglesham et al. 1981) and in India (Patra et al. 1986). These authors stated that nitrogen derived from N₂ fixation was enhanced when food legumes were intercropped with cereal crops. However, the same nitrogen fertilizer use efficiency of sole and intercrop cowpea could explain the observed similar N₂ fixation. In sole cowpea, Ssali and Keya (1984) and Eaglesham et al. (1981) found 53% and 79% of Ndfa, respectively, when fertilized with 20 kg ha⁻¹. These authors also estimated that the Ndfa was 59% when cowpea was intercropped with a cereal. Therefore, the relatively high NUE (%) of cowpea in our experiment could explain the low N₂ fixation. These results indicate that in low N₂ fixations, soil mineral nitrogen and nitrogen fertilizers constitute the main sources of nitrogen for legume plants. This contributes to depleting soil fertility rather than improving it as has been demonstrated by Rupela and Saxena (1987) and Blumenthal et al. (1982). The low NUE (%) for intercrop millet compared with that for sole millet could be attributed to nitrogen competition between millet and cowpea in the intercrop system. A probable transfer of N₂ fixed from intercrop cowpea to intercrop millet may also contribute to reduce the use of the applied nitrogen for intercrop millet. A number of researchers (Fujita et al. 1990; Ganry 1990) have already reported a potential transfer of fixed N₂ from leguminous to associated non-leguminous plants in intercrop systems. Based on total ¹⁵N balances (Table 4), nitrogen loss from the soil–plant system in sole cowpea (21.96%) was higher, followed by intercrop (11.71%) and then sole millet. However, the nitrogen loss from sole and intercrop systems was not significantly different. As reported by Vlek and Byrnes (1986), some proportions of nitrogen fertilizer can be lost through ammonia volatilization and denitrification from underground water occurring during the first few days after fertilizer application. Thus, the high tillage and the relatively large root density of millet compared with cowpea could contribute to reduce the nitrogen loss in both sole millet and millet–cowpea intercrops. The LER calculated from the grain yield (1.68) and total DM (1.71) revealed a substantial advantage of the millet–cowpea intercropping system over the sole crops of millet and cowpea (Table 5). These results supported previous studies based on legume–cereal systems (Norman 1974; Shetty et al. 1995), which attributed the advantage of intercropping to an intensification of the crop production, good management of the soil fertility and a more effective exploitation of the environment. In contrast, the large proportion of nitrogen fertilizer immobilized in sole millet soil could be of benefit to the following crop. Immobilization refers to the incorporation of nitrate and exchangeable ammonium into organic forms that are not available to plants. This occurs as microorganisms build news cells during growth. Thus, soil microorganisms compete with plants for available nitrogen, which results in a decrease in the nitrogen availability for the first crop. However, the decomposition of these microorganisms releases substantial available nitrogen for the following crop. The lower plant density in the sole millet plots could result in lower competition for nitrogen fertilizer between microorganisms and millet, resulting in a higher nitrogen immobilization. The higher density in the sole cowpea and millet–cowpea intercrops could explain the lower nitrogen immobilization because of greater competition. Further studies incorporating organic materials and phosphorus fertilizer, which is considered to be limiting in this zone, could be carried out to assess their performance on nitrogen recovery and Ndfa (%) in millet–cowpea cropping systems.

Table 1 Nitrogen (%) and 15N atom% excess of the plant components

Plant components	Cropping system	Dry matter (kg ha^−1)	Nitrogen (%)	^15N atom% excess
Grain/seed	SM	1135	2.31a	0.139a
	IM	570	1.93a	0.112a
	SC	1444	3.72a	0.508b
	IC	1778	3.69a	0.458b
	P	0.087	< 0.001	< 0.001
	LSD0.05	941	0.5	0.06
Straw/herbage	SM	3792a	1.38	0.158a
	IM	2750b	1.28	0.135a
	SC	2117b	1.32	0.405b
	IC	1976b	1.31	0.370b
	P	0.012	0.983	< 0.001
	LSD0.05	945	0.64	0.07
Ears/pods	SM	1115	1.74	0.141a
	IM	459	1.47	0.121a
	SC	1095	1.43	0.527b
	IC	1364	1.48	0.419c
	P	0.184	0.451	< 0.001
	LSD0.05	894	0.49	0.08
Within a column means followed by a common letter are not significantly different at P < 0.05. IC, intercrop cowpea; IM, intercrop millet; LSD, least significant difference; SC, sole cowpea; SM, sole millet. Grain, straw and ears are from millet and seed, herbage and pods are from cowpea.

Table 2 Nitrogen recovery, nitrogen fertilizer use efficiency and N2 fixation (15N dilution method) of cowpea as related to the cropping system

Cropping system	DM	TN	N recovery	Ndff	NUE	Ndfs	Ndfa
	(kg ha^−1)			(%)
SM	6041	95.26a	14.46	15.30a	35.28a	84.71	–
IM	3779	53.23b	6.23	12.56ab	15.20b	87.43	–
SC	4656	98.46a	9.06	9.40bc	45.88a	66.93	23.66
IC	5118	110.73a	9.21	8.46c	46.00a	63.21	28.33
P	0.072	0.018	0.084	0.019	0.003	0.259	0.148
LSD0.05	1644	30.91	3.45	3.91	12.03	32.24	32.52
Within a column means followed by a common letter are not significantly different at P < 0.05. DM, dry matter; IC, intercrop cowpea; IM, intercrop millet; LSD, least significant difference; N recovery, amount of nitrogen in plants derived from the fertilizer; Ndfa, percentage of nitrogen derived from atmosphere (N2 fixation); Ndff, percentage of nitrogen derived from the fertilizer; Ndfs, percentage of nitrogen derived from soil; NUE, nitrogen (fertilizer) use efficiency; SC, sole cowpea; SM, sole millet; TN, total nitrogen. Applied N: 41 kg ha^−1 in millet and 20 kg ha^−1 in cowpea. ^15N atom% excess of fertilizer: 0.984% in sole millet, 1.024% in intercrop millet, 5.107% in sole and intercrop cowpea, 1.024% in reference (a millet crop) sole cowpea. Intercrop millet is the non-fixing reference plant for intercrop cowpea. Sole cowpea Ndffref = 12.46% (calculation not shown) Ndff (%) = (^15N atom% excess of plant sample/^15N atom% excess of fertilizer) × 100. Ndfa (%) = 1 – (Ndfffix/Ndffref) × 100. Ndfffix is the Ndff (%) of the N-fixing plant (cowpea) and Ndffref the Ndff (%) of the non-ixing reference plant (millet). Ndfs (%) = 100 – Ndff (%) for millet crop. Ndfs (%) = 100 – Ndff (%) – Ndfa (%) for cowpea. TN (kg ha^−1) = (DM × N%) × 100. NUE (%) = (N Recovery/ Applied N)  × 100.

Table 3 Nitrogen contents and 15N atom% excess in 0–10 and 10–20 cm soil depths

Soil depth (cm)	Cropping system	Nitrogen (%)	TN (kg ha^−1)	^15N atom% excess	Ndff (%)	Ndff (kg ha^−1)
0–10	SM	0.056	850.00	0.020	2.00a	16.97a
	SC	0.043	650.00	0.029	0.57b	3.77b
	INT	0.045	825.00	0.027	0.55b	4.15b
	P	0.351	0.351	0.751	0.011	0.008
	LSD0.05	0.022	339.79	0.031	0.80	6.74
10–20	SM	0.043	650.00	0.010	1.00	6.77
	SC	0.043	650.00	0.020	0.40	2.67
	INT	0.04	600.00	0.011	1.05	1.20
	P	0.884	0.884	0.072	0.114	0.067
	LSD0.05	0.017	256.86	0.009	0.68	4.44
Within a column means followed by a common letter are not significantly different at P < 0.05. INT, Millet–cowpea intercrop; SC, sole cowpea; SM, sole millet; TN, total nitrogen. Ndff (%) and Ndff (kg ha^−1) are the percentage and the amount of N fertilizer in the total nitrogen of the soil, respectively. Ndff (%) = (^15N atom% excess of soil sample/^15N atom% excess of fertilizer)  × 100. Ndff (kg ha^−1) = (TN × Ndff (%))/100.

Table 4 Nitrogen immobilization in 0–20 cm soil depth and nitrogen loss from nitrogen fertilizer after the experiment

Cropping system	TN	Ndff	Nitrogen immobilized	Nitrogen loss
	(kg ha^−1)		(%)
SM soil	1483.33	23.73a	57.88a	6.84
SC soil	1300.00	6.43a	32.16b	21.96
INT soil	1425.00	5.45a	27.09b	11.71
P	0.644	0.001	0.034	0.078
LSD0.05	485.47	6.17	21.25	20.53
Within a column means followed by a common letter are not significantly different at P < 0.05. INT soil, millet–cowpea intercrop soil; SC soil, sole cowpea soil; SM soil, sole millet soil; TN, total nitrogen; Ndff = amount of nitrogen fertilizer in the total nitrogen of the soil. Nitrogen immobilized = percentage of nitrogen fertilizer recovered in the soil after cultivation. Nitrogen loss = percentage of nitrogen fertilizer lost during the experiment. Nitrogen immobilized = (Ndff (kg ha^−1)/Applied N (kg ha^−1)) × 100. Nitrogen loss = 100 – nitrogen immobilized (%) – NUE (%). NUE (%) is the nitrogen use efficiency of the crops in Table 2.

Table 5 Land Equivalent Ratio (LER) of the experiment based on grain and total dry matter

Cropping system	Grain yield (kg ha^−1)	Total dry matter (kg ha^−1)
Intercrop millet	508	3688
Sole millet	1135	6041
Intercrop cowpea	1364	5118
Sole cowpea	1444	4653
LER	1.68	1.71
LER = (intercrop millet yield/sole millet yield) + (intercrop cowpea yield/sole cowpea yield). The intercrop presents more benefit when LER > 1 and the sole crop is more advantageous when LER < 1.

ACKNOWLEDGMENT

This study was supported by a grant from the IAEA/RAF/5/048/ project, with the collaboration of Senegal, Burkina Faso, Mali and Niger, to fight against desertification in Sahel areas.