VARIABILITY OF COWPEA (VIGNA UNGUICULATA L.) NODULATION: WHAT EFFICIENCY IN USE OF THE RHIZOBIAL SYMBIOSIS FOR PLANT GROWTH UNDER ZAÏ CULTIVATION IN SUB

Sorghum-cowpea intercropping is the most common cropping system in the Sub region of Burkina Faso where soil fertility is low. One of its specificity is the so both species in the same zaï holes. This study analyzes the variability of nodulation and growth of cowpea, and their relations according to organic fertilization, farmer’s field, and years. Two treatments with and without organic substrates, respectively, were compared in 12 field trials established during years 2012, 2013 and 16 field tests during 2014, in 3 villages in the north of Burkina. The results showed that the nodulation and shoot growth of cowpea, varied among site under ZF than ZS, whatever the year. The nodulation variability depended upon organic substrate P and N in 2014 under ZF, and upon soil P bio 2014 under ZS. Under ZS, the mean efficiency of indivi weight was 0.6 ± 0.2 and 4.0 ± 1.6 g shoot dry weight per nodule in 2014 and 2012, respectively. In 2013, there was no significant correlation between growth and nodulation, which could be explained by low average rainfall. In 2014 under ZS, the high efficiency in use of the rhizobial symbiosis of cowpea plant was 128.5 ± 35.4 g shoot dry weight per g nodule dry weight and the N derived from nodule N2 fixation could be estimated as 510.9 ± 127.1 mg N per g nodule dry weight. It is concluded that the measurements of nodulation and growth of cowpea at flowering stage may contribute to optimize the organic fertilization of cowpea in zaï intercropping with sorghum and the subsequent improvement of crop yields in Northern-Burkina Faso. et et al. In view of the high proximity of the seedlings of both species of each other, another advantage of the zaï sowing could be the increase in positive interactions between the cereal and the legume, in particular for the processes of N legume and N transfer to the cereal by either rhizo and rhizo-decomposition (Fustec 2011). However, few information exist on the nodulation of cowpea and its efficiency for the crop growth in the farmers’ fields under zaï sowing in Sub Several studies in semi-arid areas have quantified biological nitrogen fixation by cowpea root 2003; Oroko, 2010; de Freitas examined the effects induced by cowpea on the residual nitrogen of soils (Bationo et Ntare 2000; Bationo Bado et al. 2006). In addition, the effects of organic, mineral 6505, Impact Factor: SJIF: 5.995


INTRODUCTION
In Sub-Saharan Africa, sorghum-cowpea intercropping is the most frequent cropping system. Both species are generally sown in the same zaï holes. The zaï sowing is a farmers' innovation to obtain a production from their most degraded fields. The advantages of this practice are considered the concentration in the zaï holes of both rainwater around the seedlings and small amount of organic fertilizer available in the farms (Roose 2004;Zougmoré et al. 2004;Sawadogo al. 2008;Hien et al. 2010;Bayala et al. 2012;Somé 2015).

A B S T R A C T
Sorghum-cowpea intercropping is the most common cropping system in the Sub region of Burkina Faso where soil fertility is low. One of its specificity is the so both species in the same zaï holes. This study analyzes the variability of nodulation and growth of cowpea, and their relations according to organic fertilization, farmer's field, and years. Two treatments with and without organic substrates, respectively, were compared in 12 field trials established during years 2012, 2013 and 16 field tests during 2014, in 3 villages in the north of Burkina. The results showed that the nodulation and shoot growth of cowpea, varied among site under ZF than ZS, whatever the year. The nodulation variability depended upon organic substrate P and N in 2014 under ZF, and upon soil P bio 2014 under ZS. Under ZS, the mean efficiency of indivi weight was 0.6 ± 0.2 and 4.0 ± 1.6 g shoot dry weight per nodule in 2014 and 2012, respectively. In 2013, there was no significant correlation between growth and nodulation, which could be explained by low average rainfall. In 2014 under ZS, the high efficiency in use of the rhizobial symbiosis of cowpea plant was 128.5 ± 35.4 g shoot dry weight per g nodule dry weight and the N derived from nodule N2 fixation could be estimated as 510.9 ± 127.1 mg N per g nodule dry weight. It is concluded that the measurements of nodulation and growth of cowpea at flowering stage may contribute to optimize the organic fertilization of cowpea in zaï intercropping with sorghum and the subsequent improvement of crop yields in Northern-Burkina Faso. cowpea intercropping is the most frequent cropping system. Both species are generally sown in the same zaï holes. The zaï sowing is a farmers' innovation to obtain a production from their most degraded of this practice are considered the concentration in the zaï holes of both rainwater around the seedlings and small amount of organic fertilizer available in . 2004; Sawadogo et . 2012;Somé et al. In view of the high proximity of the seedlings of both species of each other, another advantage of the zaï sowing could be the increase in positive interactions between the cereal and the legume, in particular for the processes of N legume and N transfer to the cereal by either rhizo and rhizo-decomposition (Fustec 2011). However, few information exist on the nodulation of cowpea and its efficiency for the crop growth in the farmers' fields under zaï sowing in Sub Several studies in semi-arid areas have quantified biological nitrogen fixation by cowpea root 2003; Oroko, 2010; de Freitas examined the effects induced by cowpea on the residual nitrogen of soils (Bationo et Ntare 2000;Bationo Bado et al. 2006). In addition, the effects of organic, mineral cowpea intercropping is the most common cropping system in the Sub-Saharan region of Burkina Faso where soil fertility is low. One of its specificity is the sowing of the both species in the same zaï holes. This study analyzes the variability of nodulation and growth of cowpea, and their relations according to organic fertilization, farmer's field, and without organic substrates, namely ZF and ZS, respectively, were compared in 12 field trials established during years 2012, 2013 and 16 field tests during 2014, in 3 villages in the north of Burkina. The results showed that the nodulation and shoot growth of cowpea, varied among sites and years, and were higher under ZF than ZS, whatever the year. The nodulation variability depended upon organic substrate P and N in 2014 under ZF, and upon soil P bio-availability in 2012, 2013 and 2014 under ZS. Under ZS, the mean efficiency of individual nodule to support shoot dry weight was 0.6 ± 0.2 and 4.0 ± 1.6 g shoot dry weight per nodule in 2014 and 2012, respectively. In 2013, there was no significant correlation between growth and nodulation, In 2014 under ZS, the high efficiency in use of the rhizobial symbiosis of cowpea plant was 128.5 ± 35.4 g shoot dry weight per g nodule dry weight and the N derived from nodule N2 fixation could be estimated as 510.9 . It is concluded that the measurements of nodulation and growth of cowpea at flowering stage may contribute to optimize the organic fertilization of cowpea in zaï intercropping with sorghum and the subsequent improvement In view of the high proximity of the seedlings of both species vantage of the zaï sowing could be the increase in positive interactions between the cereal and the legume, in particular for the processes of N 2 fixation by the legume and N transfer to the cereal by either rhizo-deposition decomposition (Fustec et al. 2011;Laberge et al. 2011). However, few information exist on the nodulation of cowpea and its efficiency for the crop growth in the farmers' fields under zaï sowing in Sub-Saharan Africa.
arid areas have quantified biological nitrogen fixation by cowpea root-nodules (Singh and Usha 2003;Oroko, 2010;de Freitas et al. 2012). Other studies have examined the effects induced by cowpea on the residual Bationo et Ntare 2000;Bationo et al. 2002;. 2006). In addition, the effects of organic, mineral

This is an open access article distributed under the Creative Commons Attribution License, which permits
Variability of cowpea (vigna unguiculata l.) nodulation: what efficiency in use of the rhizobial symbiosis for plant growth under zaï cultivation in sub or organo-mineral inputs for increasing nodulation, shoot and root dry weight mineral acquisition of cowpea have been documented (Singh and Usha, 2003;Chemining'wa 2007;Oroko 2010;Santos et al. 2011;Assuming et al. 2013;Kouyaté et al. 2014). However, these studies were generally conducted in research station or a single farmer field, or greenhouse with plastic pots in Burkina (B 2006), in Kenya (Chemining'wa et al., 2007), in Nigeria (Oroko, 2010Egbe et al., 2010), in Ghana (Assuming Brempong et al., 2013), and in Mali (Kouyaté However, in Sub-Saharan Africa, the variability of farmers' fields has not been taken into account in the evaluation of nodulation and biomass production, whereas thus soils diversity should be taken into account in addition to climate variability when suggest technical practices for improving crop yields. Thus, assessment of nitrogen-fixing nodules and their contribution to the growth of cowpea is needed on farmers' fields. Within this framework, this study is aimed at analyzing the variability of nodule numbers and dry weight of cowpea, and the efficiency of nodulation and N d N2 fixation for cowpea plant growth by examining relational models between these parameters.
The three villages have been considered as representative of the diversity of regional soil and climate conditions. The climate is Sudano-Sahelian type, characterized by the alternation of a long dry season from October to May and a short rainy season from May to September when is concentrating 90% of the annual rainfall. Over the period 2000-2014, annual rainfall varied from 500 to 900 mm year with an average of 715 ± 108 mm year -1 mineral inputs for increasing nodulation, shoot and root dry weight mineral acquisition of cowpea have been 003; Chemining'wa et al. . 2011;Assuming-Brempong . 2014). However, these studies were generally conducted in research station or a single farmer field, or greenhouse with plastic pots in Burkina (Bado et al., ., 2007), in Nigeria ., 2010), in Ghana (Assuming-., 2013), and in Mali (Kouyaté et al., 2014).
Saharan Africa, the variability of farmers' t been taken into account in the evaluation of nodulation and biomass production, whereas thus soils diversity should be taken into account in addition to climate variability when suggest technical practices for improving fixing nodules and their contribution to the growth of cowpea is needed on farmers' fields. Within this framework, this study is aimed at analyzing the variability of nodule numbers and dry weight of cowpea, and the efficiency of nodulation and N derived from N2 fixation for cowpea plant growth by examining relational The study was conducted in three villages in the Northern region of Burkina ( Fig. 1), namely Pougyango (12° 58'N, 2° 08'W), Zindiguéssé (13° 16'N, 2° 20'W) and Soumyaga (13° The three villages have been considered as representative of the diversity of regional soil and climate conditions. The Sahelian type, characterized by the season from October to May and a short rainy season from May to September when is concentrating 90% of the annual rainfall. Over the period 2014, annual rainfall varied from 500 to 900 mm year -1 1 ( Fig. 2A). The highest monthly temperatures of 35±5°C occur during April and May and the lowest of 25±4°C during December and January (Fig. 2B).
Soil types are mainly epipetric plinthosols, endo and hypogleyic lixisols (FAO, 2014). They are subject to strong water erosion because of their physiographic position and lack of vegetation cover. Physically, they are mostly gravelly with sandy loam to clay texture in depth. Chemically, nitrogen and phosphorus are deficient, pH is acidic and the concentration of organic matter is often below 1% (FAO, 2000). Natural vegetation is formed by tiger bush, shrub and tree savannas (Guinko, 1998).

Experimental design and crop husbandry
Twelve farmers' fields under zaï cultivation were selected in 2012, 2013 and 16 multi-location test in 2014. In each field two treatments were compared: simple zaï without organic substrate (ZS) and zaï with 0.3 kg dry matter of organic substrate (ZF) per zaï hole, equivalent to 9 t ha substrate was a mixture of livestock dung, crop and household residues stocked in pits during the dry season, from November to May. It was applied in zaï h from May to early June. Its characteristics per village and years are presented in Table 1.
In 2012 and 2013 a randomized block design with 5 replicates was used in each field, whereas in 2014, each field was treated as a block containing the pair of treatments with no intra-field replication designated by multi practice, we designed 5 agronomic blocks in 2012 and 2013, and one of these blocks was maintained in 2014. Elementary plots were 25 m 2 . Each plot contained 7 ro per row, spaced 0.40 m apart on the rows and 0.80 m between rows. The rows were perpendicular to the elementary plots. In 2012, all treatments were sown with sorghum and cowpea intercropping, whereas in 2013 and 2014, a split consisted of 35 zaï holes with sorghum compared to 42 with cowpea and sorghum in intercropping.
In 2012 and 2013, field's trials were named P1, P2, P3, P4 in Pougyango, Z1, Z2, Z3, Z4 in Zindiguéssé and S1, S2, S3, S4 in Soumyaga. These names respectiv the village plus additional number following numerical order. Soil types are mainly epipetric plinthosols, endo-petroplinthic and hypogleyic lixisols (FAO, 2014). They are subject to strong water erosion because of their physiographic position and lack of vegetation cover. Physically, they are mostly gravelly with sandy loam to clay texture in depth. Chemically, nitrogen and phosphorus are deficient, pH is acidic and the concentration of organic matter is often below 1% (FAO, 2000). Natural vegetation is formed by tiger bush, shrub and

Experimental design and crop husbandry
Twelve farmers' fields under zaï cultivation were selected in location test in 2014. In each field two treatments were compared: simple zaï without organic ï with 0.3 kg dry matter of organic substrate (ZF) per zaï hole, equivalent to 9 t ha -1 . The organic substrate was a mixture of livestock dung, crop and household residues stocked in pits during the dry season, from November to May. It was applied in zaï hole before sowing, from May to early June. Its characteristics per village and years are presented in Table 1.
In 2012 and 2013 a randomized block design with 5 replicates was used in each field, whereas in 2014, each field was ng the pair of treatments with no field replication designated by multi-location test. In practice, we designed 5 agronomic blocks in 2012 and 2013, and one of these blocks was maintained in 2014. Elementary . Each plot contained 7 rows of 11 zaï holes per row, spaced 0.40 m apart on the rows and 0.80 m between rows. The rows were perpendicular to the slope of the In 2012, all treatments were sown with sorghum and cowpea intercropping, whereas in 2013 and t consisted of 35 zaï holes with sorghum compared to 42 with cowpea and sorghum in intercropping.
In 2012 and 2013, field's trials were named P1, P2, P3, P4 in Pougyango, Z1, Z2, Z3, Z4 in Zindiguéssé and S1, S2, S3, S4 in Soumyaga. These names respectively use the first letter of the village plus additional number following numerical order.   (Kitson and Mellon, 1944) P contents per plants were calculated by multiplying N and P concentration in the composite sample of dry shoots after grinding to 2 mm, by the corresponding shoot dry weight.

Statistical analysis
Graphics and standard deviation (SD) were performed with Excel 2007. All data collected analysis of variance (ANOVA The mean differences were compared by test at 5% probability. The relationship between parameters was tested by regression analysis with R software (2.14.1). Table 2 Table 2). The means of nodules dry weight per plant (NDW plant increased by ZF by +38% and +46% respectively in 2013 2014 (Table 2).

Data in
For spatial variation in 2012, NN plant significantly from 1.5 ± 1.2 in S4 to 11.7 ± 5.3 in Z1 under ZF, and not significantly from 2.5 ± 2.7 in Z3 under ZS (Fig. 3A). By comparison to ZS, the effec on NN varied from -84% in S3 to +73% in Z1. The increase ) and zaï + organics substrates ( ) . Data are means ± SD of 15 plants of cowpea in 2012 and 2013, and 5 in 2014 by treatment per field trials harvested at flowering stage. *, **, *** indicate that the differences between means per treatment are significantly higher in the same field trial at probability < 0.05, < 0.01 and < 0.001, respectively according to Tukey HSD tests.

2767, March 2017
N concentration of the shoot was determined by the CHN 2000 micro-analyzer. P concentration was determined by the molybdovanadophosphoric acid Kitson and Mellon, 1944). N and P contents per plants were calculated by multiplying N and P concentration in the composite sample of dry shoots after 2 mm, by the corresponding shoot dry weight.
Graphics and standard deviation (SD) were performed with collected were subjected to one-way analysis of variance (ANOVA) using the R software (2.14.1). mean differences were compared by the Tukey's HSD The relationship between parameters was tested by regression analysis with R software (2.14.1). Table 2 show nodules number (NN) and nodules dry (NDW) under simple zaï (ZS) and zaï + organic substrate (ZF) for the field trials in 2012, 2013 and 2014. From the data set, the means of NN per plant (NN plant -1 ) of cowpea intercropped with sorghum in 2012, 2013 and 2014 was significantly increased by ZF compared to ZS (Table 2). For each treatment, overall NN plant -1 of cowpea was almost similar in 2012 and 2013 ( Table 2). The means of nodules dry weight per plant (NDW plant -1 ) of cowpea was highly increased by ZF by +38% and +46% respectively in 2013 and For spatial variation in 2012, NN plant -1 varied, though significantly from 1.5 ± 1.2 in S4 to 11.7 ± 5.3 in Z1 under ZF, and not significantly from 2.5 ± 2.7 in S2 to 5.3 ± 6.9 in Z3 under ZS (Fig. 3A). By comparison to ZS, the effect of ZF 84% in S3 to +73% in Z1. The increase NN was significant only for 2 among 12 field trials P1 and Z1 (Fig. 3A). Comparing the 3 villages, NN plant -1 was significantly higher in Zindiguéssé than in Pougyango and Soumyaga, but only under ZF treatment (Table 2).

Data in
In 2013, NN plant -1 varied significantly from 2.1 ± 2.7 in S3 to 11.6 ± 12.9 in S4 under ZS and from 2.1 ± 2.7 in S3 to 13.9±13.4 in S4 under ZF (Fig. 3B). By comparison to ZS, the effect of ZF on NN plant -1 varied from -49% in Z4 to +67% in S2, though the increase of NN plant -1 was significant only in S2. NN plant -1 was significantly lower in Zindiguéssé than in Pougyango under ZS, and in Pougyango and Soumyaga under ZF ( Table 2).

Plant growth
The means of shoot dry weight (SDW plant -1 ) per plant of cowpea was significantly increased in 2012, 2013 and 2014 by ZF compared to ZS treatment (Table 2). Whatever the treatment, SDW plant -1 was higher in 2012 and 2014 than 2013 (Table 2).

International Journal of Current Advanced Research
SDW plant -1 only significantly increased for 2 on 12 field trials (Z1 and Z2) (Fig. 4C). SDW plant -1 was significantly higher in Pougyango than Zindiguéssé and Soumyaga only under ZS treatment (Table 2).

Relation between nodulation and cowpea plant growth
The relationship of SDW plant -1 as a function of NN was significant under ZS (R 2 = 0.41, P < 0.05) in 2012 with a regression slope, i.e. the efficiency of individual nodule to produce SDW, of 4.0 ± 1.6 g SDW NN -1 (Fig. 5A). The efficiency of individual nodule to produce SDW was the highest in P3 with 6.18 g SDW NN -1 under ZS treatment (Fig.  5A). In 2013 and 2014, correlation between NN and SDW plant -1 were not significant except in 2014 under ZS (R 2 = 0.45, P < 0.01) without P3, with a mean efficiency of 0.6 ± 0.2 g SDW NN -1 (Fig. 5B & 5C).
In order to assess the effectiveness of the symbiosis for the plant growth, the mean values of SDW per treatment for each field trial were plotted as function of NDW in Fig. 5. Whatever the treatment in 2013, no significant correlation could be found between NDW and SDW plant -1 (Fig. 5C). In 2014, NDW and SDW were, globally, significantly correlated under ZS (R 2 = 0.49, P < 0.01) with a value for the regressionslope, i.e. the efficiency in use of the rhizobial symbiosis (EURS) of 128.5 ± 35.4 g SDW g -1 NDW (Fig. 5D). Under ZS, the highest EURS was 308 mg SDW g -1 NDW in P3 field trial (Fig. 5D).

Relation between Soils P and nodulation
In order to assess whether the variations in nodulations among field trials sites were linked with variation in soil P availability, the nodulation parameters were plotted as a function of plant P accumulation in shoot dry weight, as an indicator of P bio-availability (Fig. 7). In 2012, soil P bioavailability and NN plant -1 were correlated (R 2 = 0.44, P < 0.05) only under ZS treatment, with 0.04 ± 0.02 mg P NN -1 , without P3 field trial (Fig. 7A). In 2013, soil P bio-available was not correlated with NN whatever the treatment (Fig. 7B) but with NDW plant -1 (R 2 = 0.39, P < 0.05) under ZS with 1 ± 0.4 mg P g NDW -1 (Fig. 7D). 2014, soil P bio-available and NN plant -1 was not correlated under ZF treatment but high P bio-available was observed in P3 field trial (Fig. 7C). Soil P bio-available was correlated (R 2 = 0.38, P < 0.01) with NN plant -1 under ZS, without P1, with a mean efficiency of 0.09 ± 0.03 mg P NN -1 (Fig. 7C). Soil P bio-available and NDW plant -1 were correlated (R 2 = 0.48, P < 0.01) with a means efficiency of 1 ± 0.3 mg P g NDW -1 (Fig. 7E).

Efficiency of N derived from atmosphere
In 2012 and 2013, any correlation was observed between NN or NDW and N g SDW plant -1 (Fig. 8A, 8B & 8D). NN and mg N g SDW plant -1 were significantly correlated under ZS (R 2 = 0.30, P < 0.05) with efficiency of 1 nodule to produce 2.9 ± 1.2 mg N g SDW -1 only in 2014 (Fig. 8C). During the same year, mg NDW and N accumulated in SDW plant -1 were significantly correlated under ZS (R 2 = 0.53, P < 0.001) with efficiency of Ndfa of 510.9 ± 127.1 mg N fixed g NDW -1 (Fig. 8E). The highest efficiency of 1085.2 mg N g NDW -1 was observed in P3 field trial and the lowest in S3 and S6 field trials with -711.9 and -282.8 mg N g SDW g NDW -1 respectively (Fig 8E).

DISCUSSION
The major result of our work is the large spatial and temporal variation of cowpea nodulation and response to organic substrates in the Sub-Saharan agro-ecosystem of Burkina Faso (Figures 3 & 4). To our knowledge, this spatial and temporal variability has not been described previously using such an extensive data collection in farmers' fields. The interannual variability may be explained by variation of the average rainfall per year. Thus, the low dry weight of nodules and shoots in 2013, whatever the treatment, could be explained by the 20% lower average rainfall in 2013 than in 2012 and 2014 ( Fig. 2A). Similarly, the low rainfall is likely to explain the poor nodulation of soybean in central Kenya (Herrmann et al. (2013). Excess of water may also be detrimental like in Z4 (Fig. 3A) where cowpea could not be harvested in 2012, because of waterlogging in zaï bowl before flowering. The low water infiltration was observed after successive strong rain. It caused the rotting of cowpea plants in the cotylodenary node confirming that cowpea does not tolerate the soil long waterlogging (Hong et al., 1977).
The differences between villages for nodulation and plant growth could be explained by a decreasing rainfall gradient from Pougyango to Soumyaga, and its interactions with soil and organics substrates. Also, the variability in responses of nodulation and cowpea growth to treatments suggests fieldspecific interactions between treatment and rhizosphere conditions. Thus in Brazil, the biomass production of cowpea 37 days after sowing was higher in an Anthrosol than a Ferralsol, both amended with poultry manure (Lehmann et al., 2003).
Within the fertilized treatment, the relationship between N or P content of organic substrates in 2014 and NN of cowpea ( Fig. 6C & 6F) could reflect a beneficial effect of N and P fertilization on rhizobia infectivity under normal water conditions, by contrast with 2013 when the year was less rainy. In 2012 under ZS, NN variability (Fig. 7A) could be explained by the variation of P bio-availability except in P3. In 2013 and 2014 under ZS, the relation between NDW and soil P bio-availability (Fig. 7D & 7E) confirm the later to be an important factor for number and weight of nodules of cowpea. Thus, the use of 20, 40, 80 and 160 kg ha -1 of P 2 O 5 derived from single super phosphate produced higher relative efficiency for nodulation and biological nitrogen fixation of cowpea in greenhouse using pots experiment (Silva et al., 2010). Increasing phosphorus levels from 30 to 120 kg P 2 O 5 ha -1 in sandy clay loam soil of Ghana that initially contained 108 mg P kg -1 and 1.33 g N kg -1 increased the cowpea NN and the N and P contents in the dry matter at harvest (Assuming-Brempong et al., 2013). These may be attributed to phosphorus stimulation of root and plant growths before nodule formation since higher availability of P may stimulate nodulation due to the transfer of energy in form of ATP, and contribute to increase root hair formation and to provide more sites for rhizobia infection of cowpea (Okeleye and Okelana, 1997). The higher NN and NDW, and shoot dry weight (Table  2) of cowpea under ZF than ZS suggests a positive effect of mineralized P from organic fertilizers on the formation and functioning of nodules,whereas mineralized N did not limit nodulation. In this respect, input of 26 kg N ha -1 on humic nitosols that initially had 20% of N, did not affect the NN and NDW of cowpea (Chemining'wa et al., 2007). Also, the capacity of organic substrates to stimulate specifically some populations of arbuscular mycorrhizal fungi of soils could increase nutrient uptake for plants growth (Sharpley and Moyer, 2000;Muthukumar and Udaiyan, 2002) In 2013, the low rainfall could explain the absence of any correlation between NN or NDW and SDW of cowpea ( Fig.  5B & 5D) since water deficit has a negative effect on cowpea nodulation (Zahran, 1999), and its nitrogen fixation can decrease by 75% through nodules detachment (Zablotowicz et al., 1981). In 2014 under ZS, the high EURS of 128.5±35.4 g SDW g -1 NDW (Fig. 5E) suggests that shoot dry weight depends on nodulation and nitrogen fixation activity of nodules in conditions of low availability of soil nitrogen. This high EURS is confirmed by the estimate of 510.9±127.1 mg N g -1 NDW for the N derived from nodule N 2 fixation (Fig.  8E). At low levels of mineral-N, cowpea was recorded to fix from 64 to 134 kg N ha -1 in Northern-Ghana (Dakora et al., 1987) and up to 52 kg N ha -1 in Nigeria as intercropped with maize (Egbe and Egbo, 2011). The high efficiency of individual nodule to produce SDW in 2012 (Fig. 5A), and in 2014 the higher EURS in P3 (Fig. 5E) under ZS, than in other field may be explained by the high initial value of bioavailable P in this field trial. Contrary, the lowest values of EUSR and N accumulation in shoot dry weight in S3 and S6 field-trials in 2014 whatever the treatment (Fig. 5C & 8E) may be explained by the absence of weeding in these fields since competition from the natural weed flora may affect the number of days to 50% flowering and growth of cowpea (Remisona, 2009).

CONCLUSIONS
Nodulation of cowpea under zaï cultivation in Sub-Saharan agro-ecosystem of Burkina Faso varies spatially and temporally. This highlights the interest to conduct studies in several fields over several years. Temporal variation may be due to the variability of average rainfall per year. Whatever the year, nodulation and plant growth in our study were increased by organic fertilization compared to simple zaï. Spatial variability of nodules numbers may be related to N and P contents of fertilizers according to years, whereas under simple zaï, nodule dry weight varied with soil P bioavailability. Soil waterlogging and the absence of weeding also explained some differences of cowpea nodulation between fields whatever the treatment. The efficiency in use of the rhizobial symbiosis was achieved under simple zaï in 2014 with high N derived from nodule N 2 fixation for cowpea plant growth. Our nodular diagnosis should help to draw practical recommendations for nodulation in cowpea-sorghum intercropping and identify specific field-responses in order to improve both crops yields and soils fertility in Northern-Burkina. Thus, more work is needed on soil chemical, physical and biological characteristics, in order to assess their contributions to the observed spatial variations of nodulation and the subsequent growth of cowpea intercropped with sorghum in this region.