RESEARCH WORK
Agronomic evaluation and selection of Teramnus spp. accessions
Yuseika Olivera
Estación Experimental de Pastos y Forrajes "Indio Hatuey" Central España Republicana, CP 44280, Matanzas, Cuba
E-mail: yuseika.olivera@indio.atenas.inf.cu
ABSTRACT
A Teramnus spp. collection was studied for a year with the objective of evaluating and selecting the most outstanding accessions of this germplasm. The work was done in 3 x 1 m plots, with space between rows of 2 m, which plants had been established since a previous study. In order to obtain the variability and the relationship among variables, a principal component analysis (PCA) was performed; while for grouping the treatments with similar characteristics, regarding the measured and estimated variables, a graphic was made through the result of the PCA. The existence of 76,85% of total variability was observed, based on plant height, leafiness, cover, vigor and yield, and the formation of seven groups was identified. It is concluded that there was an acceptable performance of the accessions and those in group II stood out, mainly T. labialis CIAT-926, CIAT-21197 and CIAT-9128, for which it would be advisable to insert them in the varietal flow and compare them to commercial accessions.
Key words: Evaluation, Teramnus spp.
INTRODUCTION
Among the most used forage plant species are legumes, which stand out not only for their capacity to improve animal production, but also for the great potential they have to contribute to the sustainability of integrated livestock production systems. They have also the capacity of fixing atmospheric dinitrogen by symbiotic fixation through the microorganisms called rhizobia (Oliveira and Manhaes, 2003).
Within the Leguminoseae family, the Teramnus genus has several species, among which Teramnus labialis and Teramnus uncinatus stand out (Machado and Olivera, 2008). The former is very important for the livestock production sector, for which it has been one of the most studied in the country due to its different usages: protein bank (Milera, 1996) and as cover and green manure (Fontes et al., 2009). This legume can nodulate naturally, but it shows specificity for certain Rhizobium strains (López et al., 2002) and positive response in T. labialis seed production has been obtained when using organic fertilization (Gómez et al., 2007).
Because of all the above explained reasons, the decision was made to continue the agronomic evaluation and select the best accessions of a collection that had been studied during the establishment stage (Olivera and Olivares, 2009).
MATERIALS AND METHODS
Soil and climate. The trial was conducted on a lixiviated Ferralitic Red soil of the nodular ferruginous subtype (Hernández et al., 2003), which has a moderately acid pH (5,65), moderate contents of nitrogen and organic matter (3%), low levels of assimilable phosphorus (3,38 mg/100 g), calcium as prevailing element among exchangeable cations (15,90 cmol/kg) and a moderate to low cation exchange capacity.
Table 1 shows the values of some edaphoclimatic indicators during the experimental period. As it is observed, the rainfall volume increased progressively through the months of the rainy season, until it began decreasing in November; then, there was a strong drought, because the values did not exceed the mean of the last five years, and the most critical months were November and January. The air and soil temperature reached seasonal average values and were representative for Cuban conditions.
Experimental procedure:
The plots measured 3 x 1 m with a separation between rows of 2 m; the plants had been previously established (Olivera and Olivares, 2009).
The evaluated species were:
• T. labialis: CIAT-18186, CIAT-8412, CIAT-21199, CIAT-4998, CIAT-490, CIAT-722, CIAT-926, CIAT-9128, CIAT-18185, CIAT-9368, CIAT-21201, CIAT-20072, CIAT-24009, CIAT-4996, CIAT-17858, CIAT-4994, CIAT-17378, CIAT-21197, CIAT-18678, CIAT-928, CIAT-7822, CIAT-7821, CIAT-4989, CIAT-966 and CIAT-18768.
• T. uncinatus: CIAT-508 and CIAT-18176.
Measurements. The measurements and estimations during the experimental period were made through the methodology approved by the National Sub-Commission of Pasture Varieties, proposed by the Experimental Station of Pastures and Forages «Indio Hatuey» (Machado et al., 1997). Plant height was measured, in four spots per plot. For that purpose, a ruler graduated in centimeters was used, which position was perpendicular and always in contact with the soil surface.
In addition, the following indicators were estimated by means of scales: leafiness (1= very bad, 2= low, 3= regular, 4= good and 5= excellent); vigor (1= very bad, 2= low, 3= regular, 4= good and 5= excellent); cover (1= 10-20% area covered (very little covered) 2= 21-40% area covered (little covered), 3= 41-60% area covered (partially covered), 4= 61-80% area covered (covered) and 5= >80% area covered (well covered)); pests [0= 0-1% affected area (immune), 1= 2-10% affected area (resistant), 2= 11-20% affected area (tolerant) and 3= >20% affected area (susceptible)] and diseases (0= 0% plants, their parts or area affected, immune; 1-2= 1-5% plants, their parts or area affected, resistant; 3-4=10-25% plants, their parts or area affected, tolerant and 5-6= 50-100% plants, their parts or area affected, susceptible) and biomass yield.
For the last indicator, a homogenization cut was made, approximately 30 and 50 days (in the rainy and dry season, respectively) before the evaluation cutting; the total of each plot was cut at a height that would not affect plant growth and it was specific for each accession.
Statistical analysis. In order to obtain the variability and relationship among the variables a principal component analysis (PCA) was performed; while for grouping the treatments with similar characteristics, regarding the measured and estimated variables, a graphic was made through the result of the PCA, all this using the statistical pack SPSS version 15.0.
To establish the variables with higher influence on variability the components that had a proper value equal to or higher than 1 were accepted; in addition, the selection criteria taken was that the sum or preponderance factors were over 0,80 and the accumulated variability was higher than 70%.
RESULTS AND DISCUSSION
According to Olivera et al. (2005), Gómez et al. (2006) and Olivera et al. (2010), the principal component analysis is feasible for interpreting a data matrix. Through such analysis (table 2) the total accumulated variance was detected to be high (76,85%) and it was distributed in the first two components. This high variability indicated the differentiation among individuals for these indicators, which constituted the basis for selecting the most outstanding accessions during the study period. Among the indicators that contributed to the high variability is cover. A similar result was reported by Olivera and Olivares (2009) when studying these accessions that formed a collection (50 accessions) during the establishment stage.
In CP1 the indicators that better explained the extracted variance (60,69%) were leafiness, vigor, cover, plant height and yield; they were all positively related among themselves. The second component extracted 15,16% and mainly pest affectation contributed in its formation.
The quantity of accessions per groups formed (7) and their identification are shown in figure 1. In group II, formed by T. labialis accessions, CIAT-926, CIAT-21197 and CIAT-9128 stood out for being the highest (table 3) and having better cover values (value 4 in the scale), which is equivalent to saying that they covered more than 80% of the plot. In addition, they showed the best leafiness values; it means that these accessions can be used as cover plants and green manure, among other usages. Similar results were reported by Fontes et al. (2009) when studying the potential of a group of legumes as cover, among which was T. labialis. These authors indicated that it was the most adequate species to be used as living cover in citrus fruit plantations, in correspondence with the emergence percentage, covered area and height.
The accessions that formed groups I and V had an average performance; the former showed the highest affectations by pests and diseases, and the latter showed low values of vigor and cover. However, this did not affect yield, because in the two groups the value of that indicator was acceptable.
It should be emphasized that in these two groups the two accessions of T. uncinatus were found, which had an acceptable performance, although it is not one of the most widespread species in the different livestock production zones of the country (Olivera et al., 2008; Toral et al., 2009).
The accession CIAT-4994 (group VI) showed similarities with groups I and V, but it did not influence the total biomass yield per cutting; during the evaluation period neither irrigation nor fertilization was used and the climatic conditions were normal for each season (table 1).
The accession belonging to group III and that of group VII showed lower values in the studied indicators (table 3); this means that they were remarkably affected by management and edaphoclimatic conditions, for which future studies of the varietal flow are not necessary.
It is concluded that there was an acceptable performance of the accessions and those in group II stood out, especially T. labialis CIAT-926, CIAT-21197 and CIAT-9128, for which it would be advisable to insert them in the varietal flow and compare them to commercial accessions.
