INTRODUCTION
One of the factors that most influences on yield and quality of grasses and forages is the soil where they develop, which can be altered by natural causes or caused by human activity, where salinization processes plays an important role. Worldwide, this scourge affects millions of hectares and it is estimated that 1.5 million hectares with irrigation systems are lost due to this cause, which means the reduction of eleven billion dollars in agricultural productivity (FAO 2018).
In Cuba, a million hectares of soil are affected (by different factors); around 15 % of the agricultural area and with a serious risk of continue increasing. In the eastern region of the country the greatest damage occurs with around 650 000 hectares of soil dedicated to livestock and agriculture, and is one of the main causes of the low economic and productive efficiency of the territory, due to the decrease in the grasses and forages production, up to 25 % in many of the enterprises in the region. In Valle del Cauto, with an extension of 9 540 km2, the salinization process reaches 38 % of its area (INRH 2017). However, as a strategy to solve this problem, new varieties of Cenchrus purpureus were obtained at the Instituto de Ciencia Animal through in vitro tissue culture, with possible salinity tolerance (Herrera et al. 2003). When considering the previous explained, the objective of this research was to evaluate the new varieties under edaphoclimatic conditions of an area from Valle del Cauto.
MATERIALS AND METHODS
Location, climate and soil. The experiment was carried out in the Estación Experimental de Pastos y Forrajes (EEPF) from the Instituto de Investigaciones Agropecuarias (IIA) "Jorge Dimitrov" located 10 ½ km from Bayamo city, Cuba, and located at 20º ]18'13 " north latitude and 76º 39'48 "west longitude . The climate of the region and where the station is located, is classified as tropical humid (Barranco and Díaz 1989), with maximum temperatures that oscillate between 30-34 ºC. The recipitations during the evaluation exceeded the historical average of the region of 769 mm and the relative humidity registered values that oscillated between 51 and 98 %.
The soil of the experimental area is Fluvisol type, with little differentiation, according to the new genetic classification of Cuba soils (Hernández et al. 2015). It has medium fertility, its texture is of light clay, with generally good drainage and flat topography. Table 1 shows the chemical composition of the soil in the experimental area.
Table 1.
Chemical composition of the soil from the experimental area
Experimental design and treatments. A completely random block design with five replications was used and the experimental unit was the five-meter long furrow (Herrera et al. 2013a b). The treatments were eight new Cenchrus purpureus varieties tolerant to salinity (501, 502, 503, 504, 505, 506, 507 and 508) obtained by in vitro tissue culture at the Instituto de Ciencia Animal (Herrera et al.2003). In addition, C. purpureus cv. Cuba CT-115 (progenitor) was used as control. From the establishment cut, samplings from 30 to 105 days of regrowth were carried out every 15 days and it was determined: length and width (cm) of the fourth leaf completely open from the apex, length and width of the fourth internode (cm) from the soil level, green and dry weight of the bunch (g/bunch), height (cm), content of leaves and stems (%) and leaf area (cm2) according to the methodology described by Herrera (2006). No irrigation or fertilization was used. The rainy season (May to August) and dry season (November to February) were evaluated.
Statistical analysis. To the information obtained in both experimental stages multivariate analysis of main components was carried out. The main components were determined with eigenvalues higher than the unit and preponderance factors higher than 0.75 (Torres et al. 2008). Then, a cluster analysis was carried out and for each group the mean value, standard deviation and variation coefficient were determined for each indicator.
RESULTS AND DISCUSSION
In Valle del Cauto, are specific conditions very different from the rest of Cuba regions, especially the climatic characteristics (higher temperatures and lower rainfalls), topography and soil types (Ledea et al. 2016 and 2017). This determines the need to evaluate varieties that show higher adaptive specificity, so that the grasses structure can be established for these particularities and the productive demands that have not been solved can be cover. In this study and based on previous results (Martínez et al.1988 and Herrera et al.1990), the variables height, leaves content, length and width of leaves, length and width of the internode, yield and dry matter percentage were considered since they had showed their effectiveness to demonstrate the variability between clones obtained by tissue cultures and chemical mutagenesis. In addition, the leaf area was included since it was an outstanding indicator when explaining the variability between Pennisetum varieties tolerant to drought (Ray et al. 2016).
Table 2 shows the results of the main component analysis for the rainy period. Four main components were established with an eigenvalue higher than the unit, and to select the variables, the preponderance factor was taken equal to or higher than 0.75, which explained 93 % of the variability among the varieties.
Table 2.
Results of the main component analysis in the rainy season
The first component explains the highest percent of total variability, characterized by the variables yield, leaves width and leaf area, which are positively correlated with this component. In the second component, explains the higher variability the leaves percent and the height, the first negatively correlated with the component and establishing a negative biological relation with height, which explains that, at higher height, the varieties show lower leaves percent.
The third and fourth component expresses the highest variability through the variables length and width of the internode, respectively. The two negatively correlated with its components. These variables were the most important, helping to explain 93.14 % of the variability among the varieties. However, it was interesting to include the leaf area that could be determined by the climatic and salinity conditions (from low to medium) of the experimental area soil. In addition, this performance can be attributed to the way of obtaining these varieties, through in vitro tissue cultures and stress situation caused by the saline environment where they were developed (Herrera et al.2003).
The agronomic variables yield and mor-phoagronomic variables leaves width, height, width and length of the internode, and leaves percent are those that most influence on the variability of these varieties, which coincide with those reported by Martínez et al. (1988), when they evaluated Cenchrus somaclones in experimental areas of the Instituto de Ciencia Animal without irrigation or fertilization and was reaffirmed with what was found by Herrera et al. (1990) when using these same variables in the evaluation of King grass mutants obtained by physical mutagenesis. Ledea et al. (2017) when studying in similar conditions the Cenchrus purpureus cv. Cuba CT-115 and the CT-500 variety found that the dry matter yield and the absolute growth rate were the variables that most influenced.
The same analysis was carried out for the evaluation in the dry season and three main components were established (table3).
In the first component, the variables internode width, leaf length, yield and leaves percent predominated. The second component was characterized by leaves width and leaf area and the internode length influenced the third. These components explained 86.12 % of the variability.
These variables are the most important to explain the variability among the varieties and, if it is compared to that obtained in the rainy season, there is a difference between the cumulative total variances, which can be attributed to climatic conditions that are not favorable for the development of tropical grasses in this season when there is no irrigation and the variability between varieties is lower.
Table 3.
Results of the main component analysis in the dry season
A similar response to those of this study was found by Díaz (2007) in this period where he was able to group the variables into two components that explained 76.63 % of the variability in drought-tolerant Cenchrus varieties, highlighting the length, number and width of leaves, yield, absolute growth rate and leaf area.
Martínez (2007) in the dry season found 84.08 %, explaining the variability between clones, which does not coincide with the results of this study, difference that may be attributable to the different edaphoclimatic conditions in which these researches were carried out and especially the soil, since those of Valle del Cauto, despite the presence of degradation processes, are low in N and P. In this sense, Cruz et al. (2017) when evaluating mixed varieties (tolerant to drought and salinity) found that the variables height and stem thickness were the most influential in their results.
The cluster analysis showed five groups for the rainy season and four for the dry season (table 4).
Table 4.
Groups formed in the cluster analysis
For the rainy season group one was formed with the variety that was characterized by higher height, yield and leaf area than the rest, indicators in which much attention is needed (table 5).
Group two, allowed to group the varieties of lower height, medium yield and characteristics in the leaf region very peculiar; they stand out in an indicator that is the internode width, which was only higher in group one and group four, showing its probable high storage capacity of water, carbohydrates and other nutrients necessary for the synthesis of other compounds after each cutting or grazing (Duarte et al. 2018).
These results coincide with those reported by Ray et al. (2018), when evaluating drought-tolerant varieties found a higher leaves percentage and leaf area than the control (Cuba CT-115), which allows generating a higher amount of biomass and providing the animals with elements with higher nutritional quality. This research shows the adaptability of these varieties to the conditions of degraded ecosystems.
Table 5.
Mean values, standard deviation and variation coefficient of the indicators in the groups formed for the rainy season
H (height, cm); LP (leaves percentage); Y (yield, gDM/bunch); DM (dry matter, %); LW (leaves width, cm); LL (leaves lenght, cm); LA (leaf area, cm2); IW (internode width, cm) and IL (internode lenght, cm)
In the dry season, four groups were defined and the mean value, the standard deviation and the variation coefficient for the indicators were quantified (table 6).
Table 6.
Mean values, standard deviation and variation coefficient of the indicators in the groups formed for thedry season
H (height, cm); LP (leaves percentage); Y (yield, gDM/bunch); DM (dry matter, %); LW (leaves width, cm); LL (leaves lenght, cm); LA (leaf area, cm2); IW (internode width, cm) and IL (internode lenght, cm)
Group one was formed by the variety 508 with results analogous to those found in the rainy season, and stands out as promising in the edaphoclimatic conditions of the Valle del Cauto, in both seasons of the year. However, the leaves percentage was lower than the others, conditioned its high yield by the stems content, which relates it to cut varieties and not those that can be grazed within this genus. A negative aspect in this period was the low dry matter content of all varieties, an indicator that correlates with the nutrients contribution.
Group two was formed by those that showed lower height and higher leaves percentage, with a shortening of the internode length. These characteristics give it a certain defense structure against the dry season (Herrera 2006) and ideal conditions to be used by grazing animals because of the position shown by meristems close to the soil (Leal et al. 2018). The low growth is an adaptive characteristic to survive salinity stress, as it can be affected up to 25 % without appreciable symptoms (Martín et al. 2012). In addition, it decreases leaves width, which could influence on the minor loss of water by transpiration.
Group three was characterized by having acceptable leaf area values and the fourth showed lower yield. The study has identified a series of general trends of this species in the study area, which coincides with that reported by Ray et al. (2016) when grouping tolerant drought varieties, where the leaves width, total yield and leaves yield, and absolute growth rate were the determining variables for the grouping.
It is concluded that the analysis of main components showed that the indicators yield, leaves percentage, leaf width, leaf area and internode length, explain the higher variability among the varieties, where the varieties with similar characteristics were grouped in both evaluated seasons. It is suggested to deepen the bromatological indicators to draw up management strategies for this region.