AGRONOMIC PERFORMANCE OF ERECT AND SEMI-ERECT COWPEA GENOTYPES IN THE NORTH OF MINAS GERAIS , BRAZIL 1

The objective of this work was to evaluate the agronomic performance of elite-strains of cowpea cultivars under the conditions of the north of Minas Gerais (MG), Brazil. Fifteen elite-strains and five erect and semi-erect cultivars of cowpea were evaluated in the 2014 and 2015 summer-fall crop seasons in Janaúba MG. A randomized block experimental design was used with four replications. Plant size, lodging and value of cultivation, pod length and grain yield were evaluated. The data of the two crops were subjected to individual and joint analysis of variance. The genotype-crop interactions significantly affected all studied variables. The strains MNC04-792F-146 and MNC04-792F-144 stood out as the most productive genotypes in both crops. The cultivars BRS Itaim and BRS Novaera presented adequate size and lodging characteristics for cultivation, and values of cultivation that meet commercial standards, thus representing good alternatives for the genotypes cultivated in the north of Minas Gerais.


INTRODUCTION
Cowpea (Vigna unguiculata (L.) Walp.) is a widely-cultivated legume in the semi-arid regions of Africa, Brazil and United States (ROCHA et al., 2009).The great genetic variability of this species allows to use it for various purposes and cultivate it in various production systems (ANDRADE et al., 2010).Cowpea has great socioeconomic importance in the Northeast region of Brazil, since it is one of the staple foods for the low-income population of this region (LIMA et al., 2011).However, the cultivation of this legume has expanding to more technified areas of the country due to the increased market demand.
The north of Minas Gerais, Brazil, is a traditional region of cowpea production and consumption, however, the production in this region is limited by the low level of technology employed and cultivars used, which are recommended based on studies in other regions of the country.
The estimated Brazilian production of cowpea for the 2015-2016 crop is 390,100 Mg, with average yield of 373 kg ha -1 (CONAB, 2016).However, the current yields are not consistent with the crop productive potential, since grain yields higher than 4,000 kg ha -1 were found under experimental conditions (SILVA, 2014).Therefore, improvements of morphoagronomic characteristics of cowpea, especially those related to grain yield, must be emphasized to reach higher yields (TEIXEIRA et al., 2007).
In addition, the improvement of characteristics related to the plant architecture, aiming to develop more erect cultivars (MATOS FILHO et al., 2009) with good commercial quality is essential, due to the trend of use of more technified production systems, which generate the need for cultivars with suitable size and architecture characteristics to a greater densification and mechanization of the crop, including harvesting (BEZERRA et al., 2008).
Cowpea cultivars are cropped in different regions of the country, thus, cultivars adapted to a certain cultivation condition may not perform satisfactorily in other environmental conditions (TEODORO et al., 2015).These different performances are due to the cowpea intrinsic genetic, physiological and morphological characteristics, which cause different responses of cultivars to local edaphoclimatic conditions (SANTOS et al., 2009).
The different responses of genotypes to growing environments are usually evaluated through the interaction genotype x environment (TEODORO et al., 2015).According to Santos, Araújo and Menezes (2000), the effects of this interaction may be due to environmental conditions, soil fertility, producers' technological knowledge, management system used, which change the genotype performance.
In this context, it is important to carry out regional studies, aiming to select superior cowpea genotypes and to recommend cultivars appropriate to local growing conditions and consumer markets.Thus, the objective of this work was to evaluate the agronomic performance of elite-strains of erect and semi-erect cowpea cultivars under the conditions of the north of Minas Gerais, Brazil.

MATERIAL AND METHODS
The experiments were conducted in the Experimental Farm of the Universidade Estadual de Montes Claros, Janaúba, Minas Gerais (MG) (15º47'50"S, 43º18'31"W and altitude of 516 m), in the 2014 and 2015 summer-fall crop seasons.The treatments consisted of 20 erect and semi-erect cowpea genotypes, 15 elite-strains and 5 cultivars, all developed by the Embrapa Meio-Norte Cowpea Breeding Program in the evaluation of value of cultivation and use (VCU) of the group erect and semi-erect.The genotypes evaluated, their origins/parents and commercial subclasses are described in Table 1.
A randomized block experimental design was used, with four replications.The plots consisted of four 5-m rows spaced 0.5 m apart, with about ten plants per meter.The two central rows of each plot were considered for evaluation, discarding 0.5 m from each end of the rows, resulting in an evaluation area of 4.0 m² per plot.
The soil of the experimental area was prepared with one plowing and two harrowing at pre-planting.The soil was then furrowed and fertilized using a fertilizer-sowing machine.The seeds were sowed in March of each year, using manual seeders.The soil was fertilized following the recommendations of Melo, Cardoso and Salviano (2005) for cowpea, based on the soil chemical analysis, consisting of 20 kg ha -1 of P 2 O 5 and K 2 O at planting and 20 kg ha -1 of N as sidedressing, 25 days after planting.The weeds were manually removed at 20 days after emergence.A complementary irrigation was used from planting to grain filling through a conventional spraying system with a total water depth of approximately 350 mm and three days of applications.The temperature and precipitation during the experimental period are presented in Figure 1.
The plant size (PS), lodging degree (LD) and value of cultivation (VC) of the studied genotypes were evaluated at physiologic maturity of the plants, according to a visual scale of grades, as described in Table 2.Then, the plots were harvested to evaluate the average pod length (PL), measuring the length of 20 pods, randomly selected at harvest; and the average grain yield (GY), weighing all the grains harvested in the evaluation area of each plot and correcting the values to a 13% of humidity and transforming them to kg ha -1 .
Table 1.Origins/parents and commercial subclasses of the erect and semi-erect cowpea genotypes used for the evaluations of value of cultivation and use.BL = Branco Liso; BR = Branco Rugoso; FR = Fradinho; ML = Mulato; SV = Sempre Verde.Source: Embrapa Meio-Norte (Experiment worksheet, 2016).The data of the two crops were subjected to individual and joint analysis of variance.The differences between the means of the genotypes were evaluated by the Scott-Knott test, at 5% of significance, while the crops were compared by the F test, at 5% of significance.

RESULTS AND DISCUSSION
According to the joint analysis of variance, the interaction of genotypes with crops significantly affected all variables evaluated.The plant size of the genotypes in the 2014 crop were similar.
The cultivars BRS Cauamé and BRS Itaim, followed by the strains MNC04-762F-9, MNC04-795F-159, MNC04-795F-155 and cultivar BRS Novaera had the lowest grades of plant size in the 2015 crop, therefore, these genotypes were the most erect.All genotypes in the 2014 crop had lower or equal grades to those in the 2015 crop (Table 3).This result may be related to the climatic conditions of each crop, since the 2014 crop had slightly lower temperatures and precipitation than the 2015 crop (Figure 1).However, the differences were little and all genotypes presented plant size grades lower than 2.0 in both crops (Table 3), denoting their classification as semi-erect or erect (Table 2).Table 3.Average plant size, lodging degree and pod length of 20 erect and semi-erect cowpea genotypes, evaluated in the summer-fall crop season of 2014 and 2015.
Means followed by the same uppercase letter in the column comparing the genotypes and lowercase letters in the row comparing crops were similar by the Scott-Knott test and F test at 5% of significance, respectively.The lodging degree of the genotypes was significantly different in both crops.The lodging degree grades ranged from 1.25 to 4.33 in the 2014 crop, with the genotypes MNC04-792F-146, MNC04-792F-143, BRS Guariba, BRS Novaera and BRS Itaim presenting the lowest values; and from 1.00 to 5.00 in the 2015 crop, with the cultivars BRS Novaera, BRS Itaim and BRS Cauamé presenting the lowest values.Thus, the cultivars BRS Novaera and BRS Itaim stood out among the genotypes with the lowest lodging rates in both crops (Table 3).Machado et al. (2008) characterize 22 erect cowpea genotypes regarding their precocity, yield and architecture, under irrigated conditions in Teresina, Piauí, Brazil, and found similar lodging degrees, with values ranging from 1.0 to 4.75.A low lodging degree of the plants is desired for erect cowpea genotypes, so that they can be cultivated in more technified production systems, with intense use of mechanization.
Similar to the results of plant size, the highest lodging grades were found in the 2015 crop.This result is related to the climatic conditions during the experimental period, since during the 2015 crop occurred higher temperatures and precipitation (Figure 1), which may have resulted in a greater growth of the plants, modifying their architecture and making them more susceptible to lodging.According to Rocha et al. (2009), characteristics of the cowpea plant architecture, such as length of hypocotyl, internodes, main and secondary branches and peduncle, and their growth habit, can result in more or less lodging of plants.
According to Silva and Neves (2011), plants with smaller pods and smaller number of grains and thus, lighter pods, are preferred in semi-mechanized and mechanized crops, since they reduce the stalk bending and breaking.According to these authors, these characteristics reduce the pod contact with the soil and losses by pod and grain rotting.On the other hand, small pods can reduce the plant yield, since they produce fewer grains and, depending on the magnitude of this reduction, the plant could not compensate by increasing the number of pods.This characteristic must be evaluated together with other variables, especially grain yield, since the use of plants with smaller pods may not reduce the yield due to the compensatory capacity of the plants (phenotypic plasticity), as reported by Freire Filho, Lima and Ribeiro (2005) and Santos et al. (2011).
The value of cultivation is determined based on evaluations of the plant's general appearance, vigor, architecture, load, pod characteristics, grain quality and phytosanitary aspect (SILVA; NEVES, 2011).New cowpea cultivars must have value of cultivation of at least 3.00 to meet the requirements of the markets, thus meaning that the material has most of the characteristics suitable for commercial cultivation (Table 2).Therefore, the strain MNC04-762F-9 and the cultivars BRS Novaera, BRS Tumucumaque and BRS Itaim stood out in this study, presenting values of cultivation equal to or greater than 3.00 in both crops (Table 4).and MNC04-792F-144 also stood out, presenting good yields in both crops.