Background

Nitrogen (N) and phosphorus (P) are two important soil nutrients affecting the N2-fixing process [1]. The majority of the soils in sub-Saharan Africa are deficient in both N and P [2]. Nitrogen is considered to be the most limiting nutrient for plant growth, at least in most soils of Ethiopia [3]. To alleviate N deficiency, Rhizobium inoculation is not alone sufficient to boost crop yield increase [4, 5]. Soil conditions, and especially the nutrient content of the soil, should be at optimum level.

Peanut is one of the most importance oil crops cultivated in eastern part of Ethiopia. Peanut plants grow best in well-drained sandy soils and sunny warm temperatures with moderate rainfall. Cultivated fields with a peanut cultivation history contained (as estimated by most probable numbers) an average of 1.6 × 103 rhizobia g−1 of soil [6]. Field studies conducted by Nambiar et al. [7] showed 87% N2 fixation which is 183–190 kg N ha−1 by N difference method by using non-nodulating groundnut and nodulating isolines. This study found N2 fixation ranging from 53 to 63% when higher Ν rates (100–150 kg ha−1) were applied. The ability of symbiotic N2 fixation and nodulation (nodule weight and numbers) has been detected among cultivars of peanut [8]. Differences were also found among peanut cultivars in terms of response to Bradyrhizobium inoculation and N fertilizer application [9].

A small amount of combined N at the beginning of growth is needed for N2-fixing plants [10, 11]. A low concentration of nitrate favors the initial establishment of the nodules, perhaps as an additional source of nitrogen [12]. Application of small amounts of fertilizer Ν has been shown to increase seedling vigor [13], through correcting possible N deficiency during early growth stage [14]. In chickpea, Namvar et al. [15] found the significant increase in plant growth, N2 fixation, and yield of was found when N was applied with Rhizobium. A significant increase in chickpea grain yield production as a result of 100 kg N ha−1 application was obtained [16], despite chickpea is able to fix up to 80% of its N requirement from symbiotic N2 fixation [17]. In common bean, Hungria et al. [18] found a significant increase in nodulation and grain yield when a small amount of inorganic N was applied with Rhizobium.

At high concentrations of nitrate in the growing medium, both nodulation and N2 fixation in almost all legume species is inhibited [19,20,21,22,23]. In common bean, there was no further increase in grain yield and reduction of nodulation and N2 fixation with N rates above 12.5 mg N kg−1 soil [24]. Reduction in plant dry weight of soybean has also been noted when plants were treated with the high amount of N (128 kg N ha−1) at planting [22].

Legume plants require a high amount of P for nodulation and N2 fixation, and N2-fixing legumes will require more P than those supplied with combined nitrogen [25,26,27]. An increase in the total plant dry matter (DM), nodule number and the N2-fixation rate was found with increased P, for example, cowpea [28], soybean [29] and chickpea [30]. High P levels counteracted the negative effect of high N on nodulation [31]. Phosphorus application increased the effect of N application and rhizobia inoculation on the grain yield of common bean [32]. This could be because of the fact that nitrogen fixation is a very expensive process in energy terms, exceeding 16–18 mol ATP per mol N2 fixed [33]. Therefore, the hypothesis of this work has been stated that whether the crop is inoculated with suitable Rhizobium or not, N fertilizer is required for maximum growth and yield of peanut in the study site. Therefore, the aim of this study was to evaluate the effect of inorganic fertilizer (NP fertilizer) application and effective Bradyrhizobium inoculation on nodulation and yield of peanut in major growing areas of eastern Ethiopia.

Methods

Description of the study sites

The field experiment was carried out at Babile [09°13.234′N and 042°19.407′E at 5478 ft above sea level (asl)] and Fedis (09°06.941′N and 042°04.835′E at an altitude of 5476 ft asl) experimental sites during the rainy season of May to October of 2014 cropping season. The experimental site has a bimodal rainfall pattern with the major season occurring between July and October and the minor season between March and May. Climatically, Fedis site is the semiarid area with mean annual maximum and a minimum temperature of about 27.8 and 8.8 °C, respectively, and mean annual rainfall of 714.3 mm. While the mean maximum and minimum temperature at Babile site are 20 and 9 °C with mean rainfall of 540 mm. Agriculture in Fedis and Babile sites is characterized by smallholder mixed farming activities, which include cash crops (Khat and fruits), food crops (peanut solely and intercropped with sorghum) and livestock (dairy and beef cattle, goats and poultry).

Initial soil properties

The soil of the Babile site is reddish brown gravelly sandy loam with sandy clay loam texture (18% clay, 6% silt and 79% sand), near neutral in reaction (pH = 6.6), 0.04 mS/cm electrical conductivity, 0.56% organic carbon, 0.06% total N, 2.22 mg/kg available P, 4.18 cmol(+)kg exchangeable Ca, 3.5 cmol(+)kg exchangeable Mg, 0.15 cmol(+)kg exchangeable Na, 0.34 cmol(+)kg exchangeable K and 6.59 cmol(+)kg cation exchange capacity.

The top layer (0–20 cm) soil fertility at Fedis site is considered silty clay loam (36% clay, 45% silt, 19% sand) with an organic C content of 1.32%, pH of 7.76, electric conductivity of 0.06 mS/cm, total N of 0.12%, available P of 1.78, cation exchange capacity of 32.22 cmol(+)kg and exchangeable Ca, Mg, Na and K of 23.12, 12.87, 0.12 and 1.09 cmol(+)kg, respectively.

Source of peanut variety and inocula

The peanut variety used in the experiments was BaHa-gudo, kindly supplied by Peanut Improvement Project, Haramaya University. Botanically, this variety is erect growth habit and large seeded [34]. They have found up to 2.1 and 1.1 ton ha−1 productivity at Fedis and Babile, respectively. Native and effective isolate of Bradyrhizobium was obtained in the form of filter cake-based inoculants from Haramaya University, Biofertilizer Research and Development Laboratory.

Experimental design and treatments

A factorial combination of the following treatments was used in triplicate: N fertilizer applied at rates of 0, 20 and 40 kg ha−1; three doses of P fertilizer application: 0, 30 and 60 kg P2O5 ha−1 and two levels of Bradyrhizobium inoculation: inoculated and uninoculated. The experiment was laid out in a randomized complete block design with the size of each plot being 6 m × 6 m, while the sample plot was 3 m × 2.2 m. Peanut variety BaHa-gudo was planted at a spacing of 0.40 and 0.1 m inter- and intra-row, respectively. Urea and tri-superphosphate fertilizer were used as source of N and P, respectively.

Agronomic data

At the R2 stage of peanut, five plants from the central three rows were uprooted. The nodulation status (nodule number dry eight) and shoot dry weight were recorded. At harvest, the pod weight, the total biomass yield and kernel weight were measured. Shelling % was calculated based on the pod and kernel weight following the formula indicated in Pattee et al. [35].

Statistical data analysis

Data were tested for normality using the Shapiro–Wilk test, and all non-normal data were log transformed. A mixed-model three-way analysis of variance was performed based on a randomized complete block design with N rates, P rates and inoculation as fixed effects. Treatment means in interactions were compared using Tukey-adjusted least significant (LS) means, and treatment means in simple main effects were compared by performing post hoc Tukey honestly significant difference (HSD) test. The data were analyzed with analysis of variance (ANOVA) using SAS version 9.2 statistical software.

Result

Analysis of variance showed that Bradyrhizobium inoculation (I), nitrogen (N) and phosphorus (P) application and their two and three ways interaction, except the main effect of inoculation, had significant effect on the nodule number (NN) at Babile site (Table 1). At Fedis site, the NN of peanut was significantly influenced by the main effect of N and I, and N × P and I × N × P (Table 2). The NN increased with increasing N and P application rates, but inoculation did not have a significant effect on the NN at Babile site (Table 3). At Fedis site, only Bradyrhizobium inoculation significantly increased the NN when compared to the uninoculated plants, but the reduction of NN was found due to inorganic N application. Higher mean NN was recorded at Babile site than Fedis site.

Table 1 Analysis of variance of measures traits of peanut as affected by the main effect of Bradyrhizobium inoculation, nitrogen and phosphorus application and their two and three ways interaction at Babile experimental site
Full size table
Table 2 Analysis of variance of measures traits of peanut as affected by the main effect of Bradyrhizobium inoculation, nitrogen and phosphorus application and their two and three ways interaction at Fedis experimental site
Full size table
Table 3 Effect of nitrogen, phosphorus and Bradyrhizobium application on nodulation, total pegs weight and total biomass yield of peanut at Babile and Fedis experimental sites
Full size table

Inoculation along with 60 kg P2O5/ha and N unfertilized plants resulted in significantly higher NN than the same rates of N and P without inoculation at Babile site (Fig. 1a, b). At Fedis site, inoculation without P application was found to increase significantly the NN over P unfertilized without inoculation (Fig. 2a). In addition, inoculation along with inorganic N at 0 and 20 kg N ha−1 gave significantly higher NN than the same rates of N without inoculation (Fig. 2b).

Fig. 1
figure 1

Nodule number of peanut as affected by a phosphorus and b nitrogen rates of application at Babile sites

Full size image
Fig. 2
figure 2

Nodule number of peanut as affected by a phosphorus and b nitrogen rates of application at Fedis sites

Full size image

The effect of P, I × N and N × P was significant on nodule dry weight at Babile site (Table 1), while the main effect of I, N, P, and two and three ways interaction except I × P had significant influence on the NDW at Fedis site (Table 2). As it has been found in NN, inorganic N application significantly decreased the NDW at Fedis, but a slight increase in NDW was found at Babile (Table 3). However, P application significantly improved the NDW and the highest NDW was found to have at 60 kg P2O5 ha−1. Inoculation of Bradyrhizobium significantly increased the NDW at Fedis but not observed at Babile.

At Babile, Bradyrhizobium inoculation did not influence the NDW with P application rates when compared to the same rates without inoculation (Fig. 3a). Among the N application rates, the positive effect of inoculation found only in N unfertilized plant (Fig. 3b). At Fedis, the higher NDW by inoculation was recorded at 30 kg P2O5 ha−1 than the same rate without inoculation (Fig. 4a). However, the better NDW with inoculation was recorded in plants N control when compared to N control without inoculation (Fig. 4b).

Fig. 3
figure 3

Nodule dry weight of peanut as affected by a phosphorus and b nitrogen rates of application at Babile sites

Full size image
Fig. 4
figure 4

Nodule dry weight of peanut as affected by a phosphorus and b nitrogen rates of application at Fedis sites

Full size image

The main effect of I and P significantly affected the total pod’s weight (TPW) at Babile (Table 1). At Fedis, the two and three ways interaction, and the main effect of P significantly (P < 0.05) influenced the TPW (Table 2). The effect of inorganic N application on TPW was not significant in both experimental sites (Table 3). A significant increase in TPW due to P application was obtained only at Babile. Inoculation did not improve significantly the TPW in both locations.

ANOVA showed the significant effect of N and N × P on total biomass yield (TBY) at Babile site (Table 1). At Fedis site, except for the main effect of I and I × P, the main effect of N and P and their two and three ways interaction significantly influenced the TBY. The highest TBY at Babile and Fedis sites was recorded at 20 and 40 kg N ha−1, respectively (Table 2). However, there was no significant effect of P application and Bradyrhizobium inoculation on TBY at both locations.

ANOVA revealed that only the main effect of Bradyrhizobium inoculation significantly influenced shelling % at Babile site (Table 1). Inoculation caused a significant increase in the shelling % over the uninoculated plant (Table 4). At Fedis site, there was a significant effect of two-way interaction of N × P on shelling % (Table 2). Relatively, higher shelling % was found at Fedis than Babile site.

Table 4 Effect of nitrogen, phosphorus and Bradyrhizobium application on grain yield, plant tissue N accumulation and shelling percentage of peanut at Babile and Fedis experimental sites
Full size table

The grain yield of peanut was significantly influenced by the main effects and two and three ways interaction of I, N and P at both locations, except the main effect of I and I × P at Babile site (Table 1). The main effect of P and I × P, N × P and I × N × P was found to be significant on GY at Fedis site (Table 2). The highest GY at Babile and Fedis sites was found to have at 20 and 40 kg N ha−1, respectively, but the effect was not significant when compared to N control at both sites (Table 4). At 60 and 30 kg P2O5 application, GY was increased significantly over the P control at Babile and Fedis sites, respectively. However, the effect of inoculation did not influence the GY at both locations. Higher GY was recorded at Fedis than Babile site.

At Babile site, Bradyrhizobium inoculation did not improve the GY of peanut regardless of P and N rates of application (Fig. 5a, b). The better effect of P at 30 kg P2O5 ha−1 on GY was found at the uninoculated plant when compared to the same rate with inoculation (Fig. 6a). Inoculation, however, increased the GY when applied with 20 kg N ha−1 (Fig. 6b).

Fig. 5
figure 5

Grain yield of peanut as affected by a phosphorus and b nitrogen rates of application at Babile sites

Full size image
Fig. 6
figure 6

Grain yield of peanut as affected by a phosphorus and b nitrogen rates of application at Fedis sites

Full size image

Except for I × P, the main effect of I, N and P and their interaction was significant on plant N accumulation at both locations (Table 1). Significantly higher plant N accumulation was recorded at 40 kg N ha−1 and N control plants at Babile and Fedis site, respectively (Table 4). The highest plant N accumulation was found at 60 and 30 kg P2O5 ha−1 in Babile and Fedis site, respectively. A significant increase in plant N accumulation at Babile site was found to record by Bradyrhizobium inoculation, but an opposite effect was observed at Fedis site.

Discussion

The addition of starter N and inorganic P is vital for improving the food legumes production in degraded soils in sub-Saharan Africa. However, further study to know appropriate rates of N and P in combination with Bradyrhizobium inoculation is needed to improve the productivity of peanut. Hence, the two and three ways interaction effect of N, P and Bradyrhizobium inoculation on nodule number (NN) and nodule dry weight (NDW) was significant at P < 0.05. This suggests the need for different rates of N and P, and Bradyrhizobium inoculation requirement at Babile and Fedis site. Although the Babile and Fedis soil had the native rhizobia population nodulating peanut > 103, Bradyrhizobium inoculation significantly increased the NDW at both locations. This might be because of the less competitiveness of native rhizobia as compared to the inoculated isolates [36]. Kishinevsky et al. [37] found appreciable nodulation increase in peanut by inoculation of Bradyrhizobium. However, Bradyrhizobium inoculation failed to show a significant increase in plant N accumulation at Babile site.

The present findings show that inorganic N application at Fedis site significantly suppressed the NN and NDW and reduced the plant accumulated N. However, N application increased significantly the total biomass yield of peanut with no effect on grain yield and shelling %. These results are in agreement with those obtained in previous studies [24, 38] in which it was shown that significant reduction of nodulation and nitrogenase activity by N application did not decrease the biomass yield of alfalfa and common bean. However, reduction of grain yield by N application without inoculation was found as compared to inoculated soybean [39]. The present study shows a significant increase in NN, NDW and grain yield of peanut by 20 kg N ha−1 application in Babile indicates the need of starter N for peanut production at this site [40].

The NN and NDW at 60 kg P2O5 ha−1 and N control with Bradyrhizobium inoculation were found to be higher than the same rates without inoculation. This indicates the need of P application for enhancing the competitiveness and infectiveness of inoculation at Babile site. It has been reported on bean by Leidi and Rodriguez-Navarro [41] that an increase in nodulation and symbiotic N2 fixation with increasing P application occurs only at lower N concentration in the soil. Given the low available P status of the Fedis soil, better effectiveness of Bradyrhizobium on NN and NDW was found with NP control checks. The non-response of peanut to P at this site when Bradyrhizobium was inoculated could be associated with less effectiveness of the inoculated isolate against the indigenous isolate. On top of this, the deficiency of other nutrients rather than NP might be found in Fedis site which may reduce the responsiveness of peanut to NP application.

When the rates of NP application increased, the Bradyrhizobium inoculation effect on NN and NDW was decreased. The data would rather have showed that the NN and NDW in uninoculated plants improved better than that of inoculated plants. Less effectiveness and competitiveness of inoculated isolate when compared to the native rhizobia probably became more obvious at higher NP application rates. It is not clear how the nodules formation suppresses in inoculated plants. Nitrogen application may promote or suppress the nodulation. When nitrate was applied above starter amount, the nodules formation had reduced [37, 42] besides high N applied did not improve the productivity of food legume such as chickpea [15].

Although inoculation increased the NN and NDW at Fedis site, inoculation had no discernible effect on total pods weight, total biomass yield and grain yield of peanut at both locations. This unpredictable result has been reported on peanut by Kishinevsky et al. [37]. Van Rossum [43] found that an increase in nodule weight was not always associated with increased yield of peanut, but it was determined by peanut cultivar. Several studies showed that peanut has a low response to P fertilizer because it is able to utilize P from non-labile P sources such as Ca–phytate [44], Fe–P [45] and Al–P [46].

In this study, it is shown that usage of 60 and 30 kg P2O5 ha−1 increased significantly the grain yield of peanut at Babile and Fedis sites, respectively. In contrast to this, peanut has been considered as less responsive for P applications [47, 48] due to the fact that critical soil P levels for peanut may be lower than for other crops [49]. It is also evident from the present results that N application did not increase significantly the yield of peanut at both locations, though N is relevant for enhancing photosynthesis leading to higher dry matter production and partitioning of the assimilated products [50, 51]. These results compare well with those reported in the literature for peanut [36]. This less effect of N application on yield might be due to high N2 fixation potential of peanut-rhizobia symbiosis [52].

In this study, it has been demonstrated that Bradyrhizobium inoculation did not increase the grain yield when compared to the indigenous rhizobia without inoculation at both sites. The result also revealed that the grain yield of inoculated plants showed the decreasing trends as the rates of N and P increased. This implies that the NP application increases the effectiveness of indigenous rhizobia rather than the inoculated isolate. It is not clear how high rates of NP application reduce the effectiveness of inoculated isolate. In contrast to this, Namvar et al. [15] found that grain yield of inoculated plants in all N rates was higher than uninoculated plants at the same rate of N application. Other studies reported that the effect of P application on nodulation and yield of lentil has been more pronounced when it was applied with Rhizobium [53]. The present research output could be due to the presence of competent and effective rhizobia of peanut in soils of the study site [54].

A significant reduction in plant N concentration of peanut with increasing N and P application and Bradyrhizobium inoculation was recorded at Fedis site. These results support the findings of several authors concerning inhibition of N2 fixation reduction at high NP rate of application [41, 55]. Overall, P increased the nodulation and grain yield of peanut in both study sites more than Bradyrhizobium and N did; by inference, P must have limited peanut growth more than N. Low N and inoculation responses were obtained because there were enough effective indigenous peanut rhizobia, as shown by the good nodules and grain yield production.

Conclusion

This study shows the need of different rate of N and P fertilizer, and Bradyrhizobium inoculation requirement for peanut production in different locations in eastern Ethiopia. Native rhizobia are better effective when compared to the inoculated isolate and Bradyrhizobium inoculation effect can be determined by the native rhizobia population. Phosphorus application at the rate of 60 and 30 kg P2O5 ha−1 can be used to increase the grain yield and effectiveness of the native rhizobia of peanut at Babile and Fedis soils, respectively. Long- and short-term research work to enhance the effectiveness of native rhizobia and NP application and insure food security using balanced nutrient managements in sandy and degraded soil of eastern Ethiopia is needed.