WORK RESEARCH
Changes in the structure, population and composition of 19 Panicum maximum accessions under grazing conditions
R. Machado
Estación Experimental de Pastos y Forrajes "Indio Hatuey", Central España Republicana,
CP 44280, Matanzas, Cuba
E-mail: rey.machado@indio.atenas.inf.cu
ABSTRACT
The objective of this work was to characterize the changes and trend of indicators related to the structure, population and composition in 19 Panicum maximum accessions under simulated grazing conditions. The rotations took place every 32-40 days in the rainy season and 60-70 days in the dry season, with grazing intensity of 80-135 animals/ha/day (the first two years) and 120-160 animals/ha/year (third year). A randomized block design was used with three repetitions. Although in all treatments the live shoots/dead shoots ratio decreased, this quotient was higher than one, for which the substitution rate of the shoots was favorable. In 84,2% of the treatments, the number of tufts increased as compared to the initial status. The weed percentage increased with the rotations, but none of the treatments reached an advanced deterioration status. In this indicator the giant types CIH-13, SIH-697, CIH-3, CC-1146 and SIH-10 stood out, with values that varied between 4,6 and 14,1%, as well as the medium types CIH-15 and Tardío pequeño with 10 and 12,3%, respectively. The evaluated germplasm was concluded to show adequate values, regarding the indicators related to the structure of the tuft, population and percentage of weed invasion; although a high contrast was identified between giant and medium types and within these groups. To analyze the other productive indicators is suggested in order to determine the most outstanding types, based on the results in all the included indicators.
Key words: Panicum maximum, simulated grazing.
INTRODUCTION
Panicum maximum is among the most studied species and it is one of the most used in the tropical world. This is justified by its outstanding qualities in terms of adaptability, even under shade (Pentón, 1999; Pentón, 2000; Obispo et al., 2008), biomass production (Verdeciaet al., 2009), nutritional value (Cáceres and Santana, 1987) and potential for milk (Lamelaet al., 1984) and meat production (Alfonso et al., 1985).
However, in its population it is possible to find a wide morphological variability and environmental adaptability, which determines a differentiated response when managed with productive purposes; aspect which was proven by Machado and Olivera (2004) when evaluating 20 morphologically different accessions in simulated grazing under shade conditions.
Many accessions of this species have not been evaluated yet with animals, for which the objective of this work was to characterize the changes and trend of indicators related to the structure, population, composition of 19 P. maximum accessions when they were subject to simulated grazing conditions.
MATERIALS AND METHODS
Edaphic and climate characteristics
The soil on which the experimental stage was conducted is classified as hydrated Ferralitic Red (Hernández et al., 2003). It has a slightly acid pH (5,60 in CIK), low phosphorus contents (2,43 mg/100g) and moderate contents of total nitrogen (0,18%) and organic matter (3,20%). Among the exchangeable cations calcium prevails (11,84meq/100 g); while the cation exchange capacity (T) is slightly low (19,21 meq/100 g), for which it is considered a moderate-fertility soil.
In the last 15 years the annual average temperature of the zone was 24,3ºC. The maximum temperatures reached 33,4ºC in August and the minimum 14,2C in January. The average addition of the annual rainfall was 1 331,18 mm. Rainfall during the rainy season (May-October) represented, as average, 79,8% of the total annual volume. The evaporation reached maximum values in April (220 mm), while the average annual relative humidity was 82,6%.
The treatments were the accessions: CIH-13, SIH-697, CIH-3, CIH-15, CIH-6, CIH-22, CIH-104, SIH-759, SIH-421, SIH-810, SIH-10, SIH-127 (control), CF-305, CC-1146, Montícola, Serpentinícola, Tardíopequeño, Gramalote Puerto Rico and Likoni (control). A randomized block design with three repetitions was used.
Procedure and measurements
The soil was prepared through the conventional method, consisting in plowing, harrowing, crossing and harrowing. The plots (5,0 x 4,0 m) and the replications were separated by a distance of 2,0 m.
The planting took place in the rainy season. For planting tuft fractions were used, at a distance of 70 cm between rows and plants. Neither irrigation nor fertilization was used. The duration of the study was three years after the plots were established (more than 80% of the plants fully-developed). During the exploitation period eight buffaloes were used, which were managed through a simulated rotational grazing system. The rotations had a frequency of 60-70 days during the dry season and 32-40 days in the rainy season. In the first two years, the grazing intensity was 80 and 120 animals/ha/day for these seasons, respectively. In the third year it increased to 134,3 and 160,2 animals/ha/day, respectively. The stay period fluctuated between 3,0 and 3,5 days during the dry season and 4,0 and 4,5 days in the rainy season.
The measurements carried out at several moments of the experimental period were: live shoots (Sl)/dead shoots (Sd) ratio, by counting these pasture components; population percentage, by counting the existing tufts in the plot area; and botanical composition, for which the area covered by weeds in one square meter (randomly thrown) was estimated, according to the methodology described by Machado (2002).
Data analysis
In order to analyze the data the dynamics of the measured or estimated indicators during the experimental period was described based on descriptive statistics (averages).
RESULTS AND DISCUSSION
Jones et al. (1995) stated that a detailed analysis on the punctual variance that could exist through time provides certain interesting elements to be discussed on the possible contrast among treatments for the population and composition variables. However, they also stated that, when analyzing the occurrences in the plant cover during the first years, it is more important to establish the trend or sense with which this process is developed, than the statistical differences which could be reached between spots, for cover as well as for the area populated by pasture or weeds, because the mean values could hide at a certain moment- the true nature of the process. This could be extended to the structural indicators of pasture, which change substantially through time as a response to the environment (including management).
For such reason, it is more advisable to approach the changes produced in the evaluated indicators under a purely descriptive and analytical point of view, making emphasis on the trend with which they occurred and those arguments that could justify such performance.
In this research, although remarkable differences were observed among the moments in which the measurements of total Sl and Sd were carried out caused by the known effect of seasons and genotype factor, the most important element was to prove that in all treatments there was decrease in the Sl/Sd ratio (table 1); however, 34 months after starting the rotations, this quotient was higher than one in all cases, except in the giant type SIH-759.
The above-explained facts suggest that, practically, all the accessions were capable of maintaining a senescence and shoot substitution rate favorable to the latter, which continuous change ensures the maintenance of tuft and pastureland integrity (Machado, 2002). This is perceived as a positive symptom of adaptation of this germplasm, considering that shoot (stem) survival is an adaptation mechanism used by plants to maintain persistence (Ramírezet al., 2011).
The above-mentioned results are ratified in table 2, in which it is observed that in 16 of the 19 treatments (84,2%) the tuft number was maintained, and even increased, with regards to the initial status.
Although some authors consider cover as a vegetation variable which is stable and slow in response (Ibarra et al., 1999), indeed under these conditions most of the evaluated genotypes were capable of producing generative stems and fertile seed, which germinated at a certain moment and with this it developed a positive self-replanting status. The high seed production capacity of the species P. maximum is acknowledged practically by most researchers who have worked with many of its varieties and cultivars (Buenoet al., 1991); although some varieties, in certain regions, have difficulties in this sense (Torres et al., 2010). The highest values (in percentage terms) in tuft number, at the end of the experiment, were found in the giant types SIH-10 and Serpentinícola, as well as in the medium types Montícola, CIH-15 and CIH-104; the lowest value was found in the giant type CIH-13.
As indicated in table 3, the weed percentage increased as the rotations occurred. In this indicator 12 accessions stood out in which those weeds were below 25%; in this sense the following giant types stood out: CIH-13, SIH-697, CIH-3, CC-1146 (with 7,5 to 14,1%) and particularly SIH-10 with only 4,6%. Among the medium types CIH-15 and Tardíopequeño stood out, with 10,0 and 12,3%, respectively.
In the other treatments more than 25% weeds was detected and the most infested ones were the medium types CIH-6, CIH-22 and Montícola, with 35,8; 44,0 and 41,6%, respectively; as well as the giant types SIH-759, Serpentinícola and Gramalote from Puerto Rico, with 33,3; 27,5 and 24,1%, respectively.
According to the studies conducted by Feldman and Refi (cited by Sardiñas, 2010), weeds develop mechanisms which allow them to colonize the empty spaces that pasture is not able to occupy during the establishment or the exploitation, and can cause losses near 75%, when the weeds are very aggressive, such as Sporolobusindicus, as it occurred in pasturelands of P. maximum cv. Likoni with three years of exploitation which were not subject to any cultural work to reduce infestation (Sardiñas, 2010).
Nevertheless, none of the treatments reached an advanced deterioration status of the pastureland, according to the criteria expressed by Andrade (2006), naming thus this type of degradation when the cover by weeds is between 61 and 100%, for which it may be considered that the most affected ones showed a moderate degradation degree and the others showed a low degree.
An aspect which should be emphasized is the positive influence exerted by the shade of improved pastures on the growth delay of tropical weeds (Dias-Filho, 1998). In this sense, the effect which could have been produced by the shade of the giant types on the emergence and development of weeds was, undoubtedly, much higher than the one which could have been produced by medium types. This explains, to a certain extent, the lower invasion degree undergone by the former, by showing a higher mean height.
The fact that the giant types were the treatments with lower utilization percentage could have also contributed and, for such reason, they were subject to a lower intensity depending on consumption. The results reported by Galloso (2010) confirm these assertions, as he detected that the plots with giant types were visited (as average) on 270 occasions; while the plots of medium types were visited on 307 occasions. When further studying this aspect, the number of visits to the most invaded types (CIH-6, CIH-22 and Montícola) was 59,2% higher, as compared to the least invaded types (CIH-13, SIH-697, CIH-3 and CC-1146).
Nevertheless, among the identified weed species, the highest proportion corresponded to the grasses Dichanthiumcaricosum, Dichanthiumannulatum, Dichanthiumaristatum and common and coastal Cynodondactylon species which are well consumed by buffaloes, according to the results reported by Reyes (2008)-, and the legumes Alysicarpusvaginalis, Desmodiumtriflorum and Indigoferamucronata, all of them with adequate acceptability by the animals.
Only one very little consumed species was identified, Sporolobusindicus, which effect can be very depressing on commercial species (Padilla et al., 2001), and another one which was not consumed was also detected, Mimosa pudica. However, the infestation rate was very low for both, according to the frequency with which they appeared in each of the measurements conducted in the field.
By integrating the results of tables 1, 2 and 3 it is possible to consider that in none of
the treatments there was degradation, because the changes that occurred in the
botanical composition compromised little the stability of the population due to the effect of
competition, the invading species being presumably consumed, which percentage was around 21,7%
(as average) at the end of the experimental period. The cover was maintained, as average,
over 105%, which can be considered an outstanding population level; while the Sl/Sd ratio
was considered adequate.
Starting from the fact that the animals were always introduced to grazing when more
than 80% of the treatments had reached a state of acceptable biomass recovery, it can be
considered that a good adjustment of the rotations was made regarding the necessary rest for
pasture recovery, as well as for reaching the permissible availability of the evaluated
treatments (Milera et al., 2006). For such reasons, it is possible to state that the genotypes were
evaluated under equal conditions, because the differentiation and identification of the outstanding
and worst-performing types was not compromised by this factor, but by the
phenotype-inherent response.
The evaluated germplasm was concluded to show, in general, adequate values regarding the indicators related to the tuft structure, population and percentage of invasion by weeds; although it was possible to identify a high contrast between giant and medium types and within these groups.
To conduct the analysis of the indicators related to availability, utilization percentage, effects produced by pests and diseases, presence of chlorosis, pasture height, percentage of leaves and bromatological composition is suggested, in order to determine the most outstanding types based on the results in all the included indicators.