Agronomic performance and chemical composition of genotypes and cultivars of Megathyrsus maximus in Roraima’s savannas

The agronomic performance of 23 genotypes and six commercial cultivars of Megathyrsus maximus was evaluated under natural field conditions at the Roraima´s savannas, during the period of June 2015 to September 2019. The evaluated parameters were green dry matter (GDM) yield, leaf, stem and dead DM yields, crude protein (CP) contents, concentrations of neutral detergent fiber and acid detergent fiber. The genotypes and cultivars evaluated affected the yields and the chemical composition of the forage. The most promising genotypes showed superior agronomic performance than the commercial cultivars evaluated. The genotypes B16; PM10; PM40 and PM14 showed the highest green dry matter yields and forage with better nutritional quality, higher CP and lower fiber contents. These showed average yields of 5,103 and 3,925 kg ha -1 , for green DM and leaf DM, respectively, which represented increments of 20.7 and 12.8%, compared with those registered for the commercial cultivars. The genotype PM15 evidenced the highest productive performance during the dry season, constituting an option for regions with climatic restrictions, notably reduced rainfall. All genotypes and cultivars showed seasonal growth. The genotypes that showed the best seasonal distribution of forage production were PM11, PM36, PM30 and PM33, which provided 47.9; 46.8; 41.0 and 40.7% of GDM production during the dry season. In general, higher crude protein and lower fiber contents were directly correlated with the percentage of leaves in the available forage. The evaluated genotypes presented great genetic variability for the accumulation of forage and its chemical composition that can subsidize the generation of new commercial cultivars.


Introduction
In Roraima, livestock is one of the activities that have the greatest economic, environmental and social relevance.
Cultivated pastures represent the main forage resource for feeding herds. In less intensive production systems, forage grasses of the genus Brachiaria predominate, as they are more adaptable to soils with low and medium natural fertility and better tolerance to the dry season (Braga, 1998;Gianluppi et al., 2001;Costa et al., 2017). However, with the gradual intensification of production systems, mainly through the implementation of crop-livestock-forest integration systems, which have greater economic and environmental sustainability, species of the genus Megathyrsus are gradually replacing those of the genus Brachiaria, considering their greater forage productivity with high nutritional value (Silva, 2019;Costa et al., 2020).
The introduction of new forage grass genotypes allow identification and selection of the most promising accessions to be evaluated in the more advanced stages of evaluation programs, in order to overcome the biotic pressures and constitute more efficient animal production systems by releasing new cultivars (Costa et al., 2019;Cruz et al., 2021). In tropical regions, the economic and environmental sustainability of livestock has been a direct consequence of the programs of evaluation and selection of forage germplasm, mainly grasses, with positive effects on livestock resilience, productivity and economy, in addition to considerably reducing their pressure on the environment. Basically, there are two ways to increase the productivity of forage plants: a) through the improvement of the environment where the plant develops and, b) by replacing forage grasses those selected as promising and resulting from selection, adaptation or generation processes of new genotypes (Silva & Nascimento Júnior, 2007;Lemaire et al., 2011;Barbero et al, 2015). Thus, the pre-selected forage grass genotypes must be submitted to different ecological conditions in order to obtain greater reliability in future recommendations (Costa et al., 2017;Souza, 2018;Tesk et al., 2020).
In this work, the effects of genotypes and cultivars on forage accumulation and chemical composition of Megathyrsus maximus were evaluated in the Roraima´s savannas.

Methodology
The research was performed under field conditions using quantitative method. As there are still gaps about evaluation and selection of M. maximus cultivars and genotypes on the productivity of cultivated pastures, the hypothetical-deductive method was chosen to be used .
During the experimental period, 11 and 9 cuts were executed, respectively in the rainy and dry season. The maintenance fertilization consisted of the annual application, at the beginning of the rainy season, of 80 kg of N ha -1 , 60 kg of P2O5 ha -1 and 60 kg of K2O ha -1 , in the form of urea, triple superphosphate and potassium chloride, respectively.
Forage yields were estimate through mechanical cuts, performed at average intervals of 42 to 49 days during the rainy season and 56 to 63 days during the dry season, at a height of 20 cm above the ground. In each evaluation, after separating the components (leaves, stems and dead material), these were placed in paper bags and weighed to estimate the production of green biomass of each part, and later, dried in an air-forced drier at 65ºC for 72 hours to determine the production of dry matter (DM) of the plant parts. They were later ground in a sieve with a mesh of 5.0 mm for chemical composition determination. Nitrogen (N) contents were analyzed according to procedures described by Silva & Queiroz (2002) and Silva (2009). Crude protein contents were obtained by multiplying the N content by the factor 6.25. The neutral detergent fiber (NDF) and acid detergent fiber (ADF) contents were determined by the methodology proposed by Van Soest et al. (1991).
The data were subjected to analysis of variance considering the significance level of 5% probability. In order to estimate the response of the parameters evaluated of the grass genotypes and cultivars, means were compared by using the Scott-Knott test, at the level of 5% probability.

Genotypes/ Cultivars
Rainy season 1 Dry The productive performance of the genotypes and commercial cultivars of the grass were satisfactory and allowed the selection of promising germplasm for the establishment or recovery of pastures in the environmental conditions of Roraima.
The GDM yields recorded were higher than those reported by Beber (2018), evaluating 21 genotypes and two cultivars of M.
maximus (Mombaça and Tanzânia) in an forest environment in the State of Acre, Brazil, which reported GDM average yields Research, Society and Development, v. 11, n. 9, e55011932285, 2022 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v11i9.32285 of 1,928 kg ha -1 during the rainy season and 1,417 kg ha -1 during the dry season. The average production of leaf DM was 2,078 kg ha -1 and 1,874 kg ha -1 , respectively, for the rainy and dry seasons. In the Tocantins´s Amazon Biome, Santos (2021), evaluating nine cultivars and one genotype of Panicum maximum (PM31), estimated 10,760 kg ha -1 ; 9,910 kg ha -1 and 610 kg ha -1 the average annual productivity of GDM, leaf DM and stem DM, respectively. The genotype PM31 performance (11,380 kg ha -1 of GDM) was superior to that observed in the Paredão (9,350 kg ha -1 ), Tanzânia (9,130 kg ha -1 ) and Tamani (8,060 kg ha -1 ) cultivars. In the State of Rondônia, Costa et al. (2016) reported average annual yields of 5.171 kg ha -1 of GDM and 4,186 kg ha -1 of leaf DM for a collection of 18 P. maximum genotypes, values similar to those observed in the present study.
Considering the agronomic aspects (plant height, density and regrowth speed), productivity, distribution and percentage of leaves in the forage, the most promising genotypes were PM41, PM46, PM38, PM42, PM32 and PM30, which showed an average increase of 18% in forage production, compared to commercial cultivars evaluated (Aruana, Massai, Mombaça and Tanzânia).
In general, dead DM yields were relatively low, since most evaluations to determine forage availability were performed when plants intercepted 95% of the photosynthetically active radiation, which contributes to a significant reduction in leaf senescence rates. Santos (2021), evaluating several cultivars of P. maximum, submitted to defoliation frequencies of 41 days, estimated dead DM yields lower than those obtained in this work. The cultivars Mombaça (180 kg ha -1 ) and Zuri (100 kg ha -1 ) showed lowest yields, while the highest values were obtained with BRS Quênia (330 kg ha -1 ) and Tanzania (320 kg ha -1 ) cultivars. Likewise, Almeida (2015) found lower dead DM for genotypes PM39 (110 kg ha -1 ), PM14 (100 kg ha -1 ) and PM30 (70 kg ha -1 ), compared to genotypes PM34 (160 kg ha -1 ) and PM40 (270 kg ha -1 ), subjected to defoliation frequencies of 21 days.
Senescence reflects the natural physiological process that characterizes the last stage of leaf development. This starts after its complete expansion and progressively accentuates with increase in leaf area due to the shading of the leaves inserted in the lower portion and the low supply of photosynthetically active radiation, characterized by intense competition for light, nutrients and water between the different strata of the plant (Lemaire et al., 2011;Pereira, 2018;Martuscello et al., 2019). The leaf senescence reduces the quality of the forage. However it represents an important physiological process in the dynamics of the grass tissue flow, since about 35; 68; 86 and 42% of N, P, K and Mg, respectively, can be recycled from senescent leaves and used for the production of new leaf tissues (Sarmiento et al., 2016;Andrade, 2019;Costa et al., 2020).
The genotypes PM40, PM14, B16 and PM10 showed average yields of 5,103 and 3,925 kg ha -1 , respectively for GDM and leaf DM, which represented increments of 20.7 and 12.8%, respectively, compared with the average yields provided by commercial cultivars (4,228 and 3,480 kg ha -1 , respectively) ( Table 1). This behavior demonstrates the positive genetic gain obtained from the selection of the most promising grass genotypes that can be recommended for pastures establishment in the Roraima´s savannas.
All genotypes and cultivars showed seasonal growth. This behavior was more pronounced in cultivars that provided 30.0% of GDM during the dry season, while for the genotypes the availability of GDM was 33.4%. The genotypes that showed the best seasonal distribution of forage production were PM11, PM36, PM30 and PM33, which provided 47.9; 46.8; 41.0 and 40.7% of GDM production during the dry season (Table 1). In the Amazon region, photoperiod and temperature are the factors that less interfere with forage productivity, as they are available throughout the year in a range that satisfactorily meets the need required by forage grasses for their good development. However, precipitation is the factor that most affects forage availability, making it necessary to select more efficient grass genotypes in the uptake and use of water. Similarly, in the Rondônia´s savannas, Costa et al. (2016), evaluating a collection of 18 Megathyrsus genotypes, reported that the average forage productivity recorded during the dry season corresponded to about 35.8% of that obtained in the rainy season.

Final Considerations
The genotypes and cultivars evaluated affected the yields and the chemical composition of the grass forage.
The most promising genotypes were B16; PM10; PM40 and PM14, which showed superior agronomic performance than the commercial cultivars evaluated providing the highest green dry matter yields and forage with better nutritional quality, higher crude protein and lower fiber contents.
The genotype PM15 evidenced the highest productive performance during the dry season, constituting an option for regions with climatic restrictions, notably reduced rainfall.
In general, higher crude protein and lower fiber contents were directly correlated with the percentage of leaves in the available forage.
The genotypes of Megathyrsus maximus evaluated present great genetic variability for the accumulation of forage and its chemical composition that can subsidize the generation of new commercial cultivars.