Soybean performance in succession to the intercropping of corn with marandu grass and pigeonpea in an integrated agricultural production system

ABSTRACT Integrated agricultural production systems have been improved with the introduction of Urochloa grasses aiming at obtaining straw for a no-tillage system (NTS) and the increase of legumes, which promote physicochemical improvements in soils and guarantee intensification and sustainability of production. This study aimed to evaluate the agronomic traits and yield of soybean in succession to the intercropping of corn with marandu grass and/or pigeonpea in five populations of plants harvested for silage production. The experiment was carried out in the Cerrado region of Selvíria, MS, Brazil, during the 2016/17 and 2017/18 agricultural years, in an Oxisol. The experimental design consisted of randomized blocks, with four replications. The treatments were arranged in a 2×5 factorial scheme, consisting of off-season corn intercropped with pigeonpea (Cajanus cajan (L.) Millsp.) and off-season corn intercropped with marandu grass (Urochloa brizantha) and pigeonpea, in addition to five sowing densities (0, 6, 12, 18, and 24 seeds m-1). Soybean was planted in succession aiming to evaluate the effect of remaining straw on crop performance. Soybean grown in succession to the triple intercropping of corn with marandu and pigeon pea shows increased productivity.


INTRODUCTION
Integrated crop livestock systems (ICLS) have gained prominence in agricultural global scenario, in which the synergistic interaction of components provides sustainable intensification in crops and livestock production systems.
ICLS contributes for overall efficiency of farms avoiding negatives effects of agricultural production, in which many effects such as soil fertility improvement, greater crop yield, labor efficiency, increase soil biological activity and produce a diversity of foods are listed, enrusing food security around the world (Sekaran et al., 2021).
The versatility of IAPSs allows for several arrangements that are suitable according to the characteristics of the region and production objectives.In the Cerrado, the most attractive modality is autumnal cultivations (off-season), as the intercropped forage generates straw after harvesting grains or plant material for silage for the continuity of the no-tillage system (NTS).NTS is currently an alternative for the sustainability of tropical agricultural systems, promoting protection and improvements in soil physical, chemical, and microbiological quality, contributing to Soybean performance in succession to the intercropping of corn with marandu grass and pigeonpea in an integrated agricultural production system increasing crop yield, includind soybean crop (Laroca et al., 2018), as it is the main oilseed cultivated in the world and essential to the Brazilian economy.
Corn stands out among the crops used in ICLS due to its various applications, whether in animal feed, such as grains or silage, or commercialization of dry grains.Intercropping of maize with forage plants presents positive results for both grain yield and silage production (Bessa et al., 2018), such as noticed by Pariz et al. (2017) in which the intercropping of maize and marandu grass led to an increase in silage production and the forage itself when compared with monocrops.Corn intercropped with forages in the autumn is an alternative to increasing the amount of straw and nutrient cycling in NTS (Mendonça et al., 2015).
Grasses of the genus Urochloa present good performance in integrated systems, as they are an excellent source of forage in animal feed and provide an excellent ground surface cover (Kluthcouski & Yokoyama, 2003;Nascente & Crusciol, 2012) remaining the straw in the area for a long period compared to the straw of other plant species.
Moreover, the intercropping of legume forage species with corn stands out in the increase of silage protein contents, N cycling, and the diversification of soil cover straw for NTS, as the use of species with a high C/N ratio may lead to the temporary immobilization of nitrogen in the soil (Oliveira et al., 2010).
However, in addition to the low availability of cultivars of tropical forage legumes on the market, not all of them have the potential for use in intercropping due to the competition effect.Pigeonpea (Cajanus cajan L. Millsp.)stands out among the forages used in the Cerrado region due to its high potential for forage production for cutting or protein bank.This species improves the physicochemical conditions of soils and has the potential for intercropping with annual crops due to its erect growth habit and type of taproot system.Furthermore, this legume has good adaptation in clayey soils and can be used in intercropping with grasses, as it fixes atmospheric N 2 (Balbino et al., 2011).
The ensilage practice using crops that produce grains intercropped with tropical forage species, both grasses and legumes, have functions that go beyond animal feeding, as they also contribute to improving the production environment.Regrowth process of forage plants provides straw for NTS so that the plant residues present on the soil surface benefit crops sown in succession, increasing grain yield, and providing improvements in soil physical, chemical, and biological properties (Garcia et al., 2004).
Thus, this research aimed to evaluate the effect of the straw from the regrowth of five sowing densities of pigeonpea intercropped with corn and/or marandu grass harvested for silage in the second crop season on the development and yield of soybean in succession.

MATERIAL AND METHODS
The experiment was carried out in the municipality of Pigeonpea 'Aratá' was sown in the inter-row of corn, also at a spacing of 0.45 m, at densities of 0, 6, 12, 18, and 24 seeds m −1 , using the same seed drill used for corn and on the same day of corn sowing with or without grass.

Corn intercropping was harvested in September 2016
and September 2017 aiming at silage production, with a cutting height of 0.30 m.The average productivity of dry matter for silage was 38,7 t ha -1 in 2016 and 33,5 t ha -1 in 2017.After ensiling, the post-harvest residue remained under rest for grass and pigeonpea regrowth.The area was desiccated after regrowth (around 60 days) to form straw, using the herbicides glyphosate (1560 g ai ha −1 ) and Select (1.0 L ha −1 ), with subsequent reap of plant residues using a horizontal shredder (Triton).Average regrowth residue was 5,0 t ha -1 for corn with pigeonpea intercropping, and 8,5 t ha -1 for intercropping of corn with Marandu and pigeonpea.
Then, soybean was sown using the cultivar BMX Potência RR on November 15, 2016, and November 10, 2017, at a density of approximately 16 seeds per meter of furrow, with an inter-row spacing of 0.45 m, aiming at a stand of 300,000 plants ha −1 , using a seed-cum-fertilizer drill a rod-type furrow opener (machete) for NTS.
Soybean seeds were treated with Vitavax-Thiram and inoculated with bacteria of the genus Bradyrhizobium japonicum (600,000 cells/gram of seeds), and the base fertilization consisted of 370 kg ha −1 of the NPK formulation 04-20-20 (14.8 kg N ha −1 , 74.0 kg P 2 O 5 ha −1 , and 74.0 kg K 2 O ha −1 ) in the first growing season and 350 kg ha −1 of the NPK formulation 04-30-10 (14.0 kg N ha −1 , 105.0 kg P 2 O 5 ha −1 , and 35.0 kg K 2 O ha −1 ) in the second growing season, as recommended by Boletim Técnico 100 (Raij et al., 1997).No further fertilization was carried out.Phytosanitary management was carried out according to the crop requirements.Soybean was harvested in late February 2017 and early March 2018.
The evaluations of morphological components and production were carried out by collecting 10 soybean plants in the useful area to obtain the plant height, first-pod height, the number of pods per plant, and the number of grains per pod.
All plants from 6 m of each experimental unit (two 3-m rows) were counted, manually harvested, and threshed on a mechanical threshing machine to determine the final stand and grain yield.Subsequently, the moisture was corrected to 13% (wet basis) and grain yield was calculated.A sample of grains was taken to the laboratory to determine the 100-grain weight.The grains were counted on an electronic counter and weighed on a precision electronic balance, with means of four samples.
The results were subjected to the Shapiro-Wilk normality test and the analysis of variance using the F-test (P ≤ 0.05).The effect of residues with or without grass was compared by Tukey's test (P ≤ 0.05).The effect of pigeonpea sowing density, when significant, was evaluated by regression analysis (P ≤ 0.05).The analysis was performed using the SISVAR ® software (Ferreira, 2008).

RESULTS AND DISCUSSION
The 2016/17 agricultural year presented a higher plant final stand (PFS) of soybean on the residue from the intercropping with marandu grass (Table 1).Krutzmann et al. (2013)  In this context, less water retention may have occurred on the soil surface at germination due to the absence of straw on the soil of the plot without grass and the rapid pigeonpea straw decomposition, which may have influenced the initial growth of soybean seedlings (Figure 1).Soybean PFS was not influenced by pigeonpea sowing density or the interaction between density and intercropping with marandu grass (Table 1).Plant height (PH) presented no difference as a function of the isolated effect of intercropping with grass, sowing density of the predecessor pigeonpea, and the interaction between them in both agricultural years (Table 1).Plant height is influenced by the length of the vegetative period, in which the range from 0.80 to 0.96 m is the minimum height considered normal for the crop (Pereira Júnior et al., 2010), as these values reduce the losses of pods not harvested in mechanized harvesting (Chioderoli et al., 2012).
Therefore, the mean value observed in the present study is above the minimum required.
Intercropping with marandu grass provided higher first-pod height (FPH) (Table 2) in the 2016/17 growing season although the plants showed the same PH, which may be related to the straw effect, which provides higher soil moisture and supply of N, P, and Ca with residual grass.Garcia et al. (2014) observed no differences when evaluating the effect of straw decomposition in the corn/ Urochloa and corn/Megathyrsus maximus intercropping and residual nitrogen fertilization for soybean FPH in an area close to that of the present study.
The 2017/18 agricultural year showed no difference for FPH according to the inclusion or not of grass, as observed for PH, due to favorable climate conditions, which provided uniform cultivation.No difference was observed for FPH as a function of the isolated effect of the sowing density of the predecessor pigeonpea.Also, no interaction was observed between factors for this soybean morphological attribute (Table 2) in both agricultural years.
The number of pods per plant (NPP) was higher for soybean cultivation on plant residues without intercropping with marandu grass (Table 2) in the 2016/17 growing season.Table 1 shows that the lowest plant stand was found in this treatment, which indicates that the highest NPP for this treatment can be attributed to the compensation capacity that soybean plants have when there is a reduction in plant Soybean performance in succession to the intercropping of corn with marandu grass and pigeonpea in an integrated agricultural production system density, increasing the individual legume production, mainly in cultivars with an indeterminate growth habit.
According to Procópio et al. (2013), soybean has a high ability to compensate for lower plant densities, mainly forming a higher number of legumes per individual when the population is reduced.Krutzmann et al. (2013) also found similar NPP values for soybean grown on U. brizantha 'Marandu' straw.Means followed by different letters in the column differ from each other by Tukey's test at a 5% probability.CV: coefficient of variation.
The 2017/18 agricultural year showed no difference between treatments for NPP, which is related to the higher crop uniformity due to a better rainfall distribution (Figure 2).
As observed for NPP, the number of grains per pod (NGP) was higher in soybean cultivation without marandu grass residues (Table 3) in the first growing season (2016/17) due to the compensation ability of soybean plants, resulting in higher numbers of grains derived from the higher number of pods in smaller stands.However, the water input was not sufficiently adequate and distributed to promote grain filling, necessary to evidence this effect, result in larger grains, and have an effect on the 100-grain weight (100W), which was similar between treatments.
No significant effect was observed for pigeonpea sowing densities nor for the interaction between intercropping with grass and pigeonpea sowing density for NGP (Table 3).
NPP and 100W showed no significant differences in intercropping with grass and pigeonpea at different densities in the 2017/18 agricultural year due to the effect of plant population homogeneity and better rainfall distribution (Figure 2).
Grain yield in the 2016/17 growing season showed no significant difference for the isolated effect of intercropping with marandu grass, pigeonpea sowing density, and the interaction between intercropping with marandu grass and pigeonpea sowing density (Table 4).The regression analysis presented no adjustment for grain yield (R 2 = 0,3447), what could be associated with a large variation in the first year, mainly due to water stress in the early stages of plant development, which contributed to negative effects on plots formation.The presence of grass in the predecessor cultivation did not affect grain yield, as observed by Krutzmann et al. (2013) for soybean planted on different vegetation cover, which showed that the vegetation cover from U. ruziziensis and U. brizantha 'Marandu' pastures intercropped with Tanzania grass did not influence soybean yield.Pacheco et al. (2011) also observed that soybean yield was not influenced when grown on residues of Urochloa ruziziensis, Urochloa brizantha 'Marandu', millet (Pennisetum glaucum), intercropping Urochloa ruziziensis + pigeonpea, and fallow.
On the other hand, Chioderoli et al. (2012) observed that the production of soybean grown on residues of Urochloa brizantha 'Marandu' reached values above the regional mean in high technology areas, with an overall mean of 4.1 Mg ha −1 , but in an area irrigated by of center pivot.
The presence of grass in the predecessor intercropping in the 2016/17 agricultural year did not change soybean yield due to the compensatory effect between the other components of production because, even with a higher number of pods per plant and grains per pod in the treatment without grass, the highest stand occurred in its presence.Moreover, the 100-grain weight showed no effect, equaling the yield between treatments.
The following agricultural year (2017/18) presented differences in soybean grain yield, which was higher in the successor cultivation to the intercropping with marandu grass (Table 4).Although not significant, NPP (Table 2), NGP, and 100W (Table 3), associated with the compensatory effect of the lower PFS (Table 1), were numerically higher in soybean cultivation with marandu grass residues when compared to cultivation without grass (Table 2).It may have led to the interaction between these variables, reflecting a higher grain yield in the 2017/2018 growing season.The regression analysis presented increasing linear adjustment (R 2 = 0,8142), according to inclusion level of pigeonpea.
Soybean performance in succession to the intercropping of corn with marandu grass and pigeonpea in an integrated agricultural production system  The interaction of the isolated effects of crops shows positive results despite the higher interspecific competition between crops and the higher complexity of the system in these triple intercrops.The soil exploration by the root system in the intercropping of corn with Urochloa, together with the produced straw, favors water infiltration, the reduction in the erosive process, and the use of water and nutrients, contributing to higher values of soybean grain yield and maintaining system stability (Chioderoli et al., 2012).Likewise, the intercropping between corn and pigeonpea is promising in promoting improvements in soil structure (Chieza et al., 2013), in addition to presenting the ability to biologically fix nitrogen and make it available for the successor crop, as it is a legume.
Added to the effects on soil physicochemical attributes, this intercropping is very beneficial, reflecting in positive responses of the crop in succession.According to Pariz et al. (2020), this type of intercropped cultivation presents better cycling and nutrient use efficiency.Thus, the triple intercropping with corn, pigeonpea, and marandu grass is a productive and highly sustainable system.

CONCLUSIONS
Soybean grown in succession to the triple intercropping of corn with marandu and pigeon pea shows increased productivity.

ACKNOWLEDGEMENTS, FINANCIAL SUPPORT AND FULL DISCLOSURE
This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior -Brasil (CAPES) -Finance Code 001.
There are no conflicts of interest in carrying out the research and publishing the manuscript.
Selvíria, Mato Grosso do Sul State, Brazil (20°20′05″ S and 51°24′26″ W, altitude of 335 m) from November 2016 to March 2017 and November 2017 to February 2018.The soil in the area was classified as a clayey-textured typic dystrophic Red Latosol (Oxisol) (Santos et al., 2018).The climate is classified as humid tropical, Aw according to the Köppen classification, with a rainy season in the summer and a dry season in the winter.The climate data of the experimental period, referring to the maximum, mean, and minimum temperature and rainfall were collected at the meteorological station located at the Teaching, Research, and Extension Farm (FEPE) of Unesp in Selvíria, MS, Brazil.The experiment was carried out under low altitude Cerrado conditions, in an experimental area cultivated under NTS for six years, with corn being the predecessor crop intercropped with marandu grass and/or pigeonpea (March 2016) intended for silage production.Soil analysis at a depth of 0.0 to 0.20 m, carried out according to the methodology of Raij et al. (2001) before the experiment was set up, showed the following chemical attributes in 2016/17: P (resin) = 35 mg dm −3 ; organic matter (OM) = 27 g dm −3 ; pH (CaCl 2 ) = 5.5; K + = 2.9; Ca 2+ = 37; Mg 2+ = 29; H+Al = 29; Al 3+ = 0; sum of bases (SB) = 68.9;cation exchange capacity (CEC) = 98.3 mmol c dm −3 ; and base saturation (V) = 69%.In 2017/2018 the chemical atributes were: P (resin) = 20 mg dm −3 ; organic matter (OM) = 29 g dm −3 ; pH (CaCl 2 ) = 5.0; K + = 10.6;Ca 2+ = 32, Mg 2+ = 24; H+Al = 38; Al 3+ = 0, sum of bases (SB) = 66.6; cation exchange capacity (CEC) = 104.6 mmol c dm −3 ; and base saturation (V) = 64%.The experimental design was randomized blocks in a 2×5 factorial arrangement, with four replications.The treatmens consisted in two types of intercropping (1.Corn+ Marandu + Pigeonpea; 2. Corn + Pigeonpea) at five sowing densities of inter-row pigeonpea, as follows: a) off-season corn intercropped with five sowing densities of pigeonpea (0, 6, 12, 18, and 24 seeds m −1 ) and b) off-season corn intercropped with marandu grass and five sowing densities Roseana Ramos Pereira et al. of pigeonpea (0, 6, 12, 18, and 24 seeds m −1 ).The experiment consisted of 40 plots of 20 m in length and 3.6 m in width, and the useful evaluation area consisted of three central rows of 18 m in length of each plot set up in the same areas in both seasons.Corn was sown in May 2016 and May 2017 intercropped or not with marandu grass (60% PLS with 7 kg ha −1 of seeds) in the row under NTS at a depth of 0.06 m, using a seed-cum-fertilizer drill provided with a third box where the grass seeds were conditioned and a rod-type furrow opener mechanism (machete).The corn hybrid was DKB 350 YR, sown at a density of 3.3 seeds per meter, spaced at 0.45 m between rows, targeting a population of around 66,000 plants ha −1 .
found similar results in the intercropping of U. brizantha with Tanzania grass, predecessors to the soybean crop.However, different results from those observed in the present study are reported in the literature (Nascente & Crusciol, 2012; Chioderoli et al., 2012), which are attributed to abiotic factors, mainly the climate and soil of each region.

Figure 1 :
Figure 1: Maximum, mean, and minimum temperatures and rainfall presented in fortnights, during the experimental period in the 2016/17 agricultural year.

Figure 2 :
Figure 2: Maximum, mean, and minimum temperatures and rainfall, presented in fortnights, during the experimental period in the 2017/18 agricultural year.

Pariz
et al. (2020)  analyzed a triple intercropping of corn + pigeonpea + marandu grass for silage associated with black oat in overseeding for sheep grazing during three seasons and observed that this type of intercropping was a highly effective option for silage production and the improvement of other system yield elements, with higher straw coverage and soybean yield in succession in the fourth growing season.

Table 1 :
Plant final stand (PFS) and plant height (PH) of soybean grown on the residual straw of corn intercropped with pigeonpea at five sowing densities with or without Urochloa brizantha 'Marandu' in the 2016/17 and 2017/18 agricultural years Means followed by different letters in the column differ from each other by Tukey's test a 5% probability.CV: coefficient of variation.

Table 2 :
First-pod height (FPH)and number of pods per plant (NPP) of soybean grown on the residual straw of corn intercropped with pigeonpea at five sowing densities with or without Urochloa brizantha 'Marandu' in the 2016/17 and 2017/18 agricultural years

Table 3 :
Number of grains per pod (NGP) and 100-grain weight (100W) of soybean grown on the residual straw of corn intercropped with pigeonpea at five sowing densities with or without Urochloa brizantha 'Marandu' in the 2016/17 and 2017/18 agricultural years Means followed by different letters in the column differ from each other by Tukey's test at a 5% probability.CV: coefficient of variation.

Table 4 :
Yield of soybean grown on the residual straw of corn intercropped with pigeonpea at five sowing densities with or without Urochloa brizantha 'Marandu' in the 2016/17 and 2017/18 agricultural years Means followed by different letters in the column differ from each other by Tukey's test at a 5% probability.CV: coefficient of variation.