Introduction

In Mexico, the largest population of cattle is concentrated in Veracruz, state that has an inventory of 3.77 million heads, equivalent to 11.8% of national herd; in 2012, in Veracruz produced 1.8 million tons of carcass meat and 10 880.9 million liters of milk, placing it in the first and fifth national place as producer of these goods, respectively (^{SIAP, 2013}). In this state, predominate the production system dual-purpose cattle, in which, grasslands are the main food supply (^{Calderón et al., 2007}). The sustainability of animal production systems based on grasslands, depends both on fodder production and the persistence of these (^{Ramírez et al., 2011}). In the state of Veracruz 1.04 million hectares are devoted to pasture (^{INEGI, 2007}), mainly: Estrella de Africa (Cynodon plectostachyus Vanderyst); Privilegio (Panicum maximum Jacquin), cvs Tanzania and Mombaza; Jaragua [Hyperrhenia rufa (Ness) Stapf]; Pangola (Digitaria decumbens Stent); Elefante (Pennisetum purpureum Schumach); Llanero (Andropogon gayanus Kunth) and species from the genus Urochloa, as Señal (U. decumbens Stapf) and Insurgente [U. brizantha A. (Hochst. ex A. Rich.) R. D. Webster], which stand out for their adaptation to different environmental conditions and high productivity (^{Enriquez et al., 1999}).

Urochloa brizantha (Hochst. Ex A. Rich.) R. Webster cv Insurgente (known as palisade grass), identified previously as Brachiaria brizantha (Hochst. Ex A. Rich.) Stapf., has gained acceptance among producers for their outstanding characteristics such as high growth rates, good nutritional quality and resistance to spotted spittlebug of pastures (Aeneolamia spp.) (^{Enríquez et al., 1999}; ^{Enríquez and Romero, 1999}).

The productivity of a meadow can decrease by the use of unsuitable forage species for environmental conditions, overgrazing, especially during periods of low rainfall, incidence of pests and diseases, establishment in areas with fragile soils, depletion of soil nutrients especially improved species, low or no use of fertilizers, high infestation of herbaceous and woody weeds, and indiscriminate burning (^{González and Meléndez, 1980}; ^{Spain and Gualdrón, 1991}; ^{Modesto and Mascarenhas, 2001}; ^{Boddey et al., 2004}; ^{Silva et al., 2004}; ^{Calderón et al., 2007}). Degradation begins with loss of plant vigor and reflects in decreased coverage and loss of desirable plants species; the latter allows the development of weeds or leaves bare soil, which favors compaction by trampling of animals, and finally low production of forage. In this situation, the meadow loses productivity and stands less than 0.5 animal unit per hectare, while a newly established pasture can stand two or more animal unit per hectare (^{Oliveira et al., 2004}; ^{Silva et al., 2004}; ^{Padilla and Sardinas, 2005}; ^{Rincón, 2006}; ^{Padilla et al., 2009}). Recovery or rehabilitation of a meadow to its production capacity per unit area and per animal, to acceptable ecological and economic levels, implies the presence of one or more desirable forage species, which are likely to be preserved (^{Spain and Gualdrón, 1991}; ^{Padilla and Sardinas, 2005}).

In livestock basins from Central America, 50 to 80% of grasslands have some degree of degradation, its stocking rate is less than 40% compared to pastures receiving proper management (^{Holmann et al., 2004}). In tropical areas from central Veracruz it is common to detect degraded grasslands in different degrees; although there is no precise quantification of the magnitude of this problem, it is estimated that the percentage of grassland with some degree of degradation could be similar to that presented in Central America. Rehabilitation of degraded grasslands could increase the productivity of livestock in the region by increasing the number of animals that can sustain the meadow. For this reason, an experiment was conducted in order to determine the effect of different mechanical work, fertilization and chemical weed control in the recovery of palisade grass meadow in a tropical area from the center of the state of Veracruz.

Materials and methods

Location of the experiment and study period: the experiment was conducted in a degraded meadow of U. brizantha cv. Insurgente (palisade grass) established in 2003, in the Ranch El Carpintero, located in La Calentura, municipality of Medellin, Veracruz. The experimental site is located at 18° 57’ 47’’ north latitude, 96° 12’ 20.7’’ west longitude at 10 masl. The climate of the region is Aw’’₁ (w) e.g., corresponding to hot humid, average humidity, with summer rainfall, annual rainfall of 1,300 mm and average temperature of 25 °C (^{García, 2004}). The soil is classified as Vertisol with sandy clay loam tecture. The experiment began on August 2011 and ended in December 2012 and was conducted under rainfed conditions.

Experimental design: the treatments were distributed at random in a randomized block design with four replications, in 14 m wide * 15 m long plot.

Meadow characterization: for the analysis of soil four tours were made in zig-zag in the experimental field with a Dutch auger collecting 30 samples at a depth of 0-20 cm, with which a sample of 3 kg was formed, and was analyzed in Laboratory of Soil Analysis, Plant and Water of the Experimental Field Cotaxtla from INIFAP. To characterize the initial vegetation in the meadow, 192 samples of 1 m² randomly distributed in each block were taken. In each sample, visual coverage estimation was made by determining the area occupied by the palisade grass and weeds and the number of both were counted to determine density; with these data, the absolute and relative frequency by species, and the importance value (^{Cox, 1980}) was determined.

Treatments: 11 treatments were evaluated, of which four corresponded to agricultural work, six to chemical weed control and the control, in which no activity was performed in the plots (Table 1). Weeding was performed at ground level with a rotating brush cutter with tractor PTO; while harrowing was performed using a 20 disc harrow. Fertilization was performed manually with the 46-23-00 formula of N, P and K, using urea and triple calcium superphosphate as nitrogen and phosphorus source, respectively. One day before the application of herbicides, the weed from plots corresponding to chemical control was weeded to 60 cm. Herbicides application was made on September 15, 2011, using a motorized backpack sprayer, equipped with a four flat fan nozzles 8003, which provided an expenditure of 466 L ha^-1. The chemical control treatments were added a nonionic surfactant, at dose of 250 mL per 100 L of water.

Meadow management: since its establishment in 2003, the meadow has been handled with rotational grazing using dairy cows and their calves, in the dry season was overgrazed and was not applied fertilizer. After the application of treatments until the completion of the first cut, the meadow was not grazed (September 15 to December 15, 2011); in turn, in 2012, the meadow was used as traditionally done by the producer.

Weather conditions: rainfall, evaporation and maximum and minimum temperature data during the experiment, were obtained of the Meteorological Station from CONAGUA, located in Paso del Toro, Medellin, Veracruz, which is closest to the experimental field.

Results

Weather conditions: in the period in which the first stage of the experiment was carried out (September to November 2011), rainfall was 240 mm and evaporation of 225.8 mm while the maximum and minimum temperature for the period was 30.4 and 19.9 ^oC, respectively, condition that favored plant growth. December 2011 to May 2012 (period corresponding to the dry season), evaporation was greater than rainfall in 181 mm which was reflected in little grass growth and from June to December 2012 (corresponding to the second evaluation period of the experiment), the rainfall was greater than evaporation and temperature increased (Figure 1).

Figure 1 Weather conditions during the experimental period. Meteorological station from CONAGUA, located in Paso del Toro, Medellin, Veracruz. 

Initial state of the meadow: the soil of the meadow had the following characteristic: pH= 6.2, organic matter= 2.7%, inorganic nitrogen= 21 ppm, P= 14 ppm, K= 382 ppm, Ca= 2 303 ppm, Mg= 291 ppm, Fe= 47 ppm, Cu= 3.3 ppm and bulk density= 1.2 g cm^-3. The initial coverage of palisade grass was 27%, other grasses 15%, broadleaf 56% and bare soil 2%. Broadleaf density 50.3 plants m^-2 and palisade grass 4.8 plants m^-2. 20 weed species belonging to nine botanical families were identified. Broadleaf weeds with higher value of importance by greater coverage, density and frequency were: Desmodium sp., Calopogonium mucunoides Divert, Aeschynomene american L. and Euphorbia hirta L. Plants from these four species accounted for 43% coverage (Table 2).

Palisade grass coverage: it was modified (p≤ 0.05) by effect of the treatments and time (Table 3). Since treatment application to 75 DDAT, the highest palisade grass coverage was in the chemical weed control treatment; coverage values ranged from 96 to 99%. In weeding treatments and weeding + fertilization, coverage at 30 and 60 DDAT was 63 and 43% lower than the control and at 75 DDAT was similar to this. In treatments that included harrowing, at 75 DDAT, obtained the lowest coverage of palisade grass, which was 72% lower than control.

In evaluations made at 320, 390 and 440 DDAT on chemical weed control plots, showed higher coverage of palisade grass, with values above 90% in the three samplings, except for the treatment with 2,4-D whose coverage was 81, 83, and 92% at 320, 390 and 440 DDAT. Weeding had coverage statistically similar to the control, while with weeding + fertilizer, palisade grass coverage at 320 and 390 DDAT was lower by 30 and 22% to control and at 440 DDAT was statistically equal to control. In harrowing treatments palisade grass showed the lowest coverage throughout the evaluation period; thus at 440 DDAT with weeding + harrowing and with weeding + fertilization + harrowing coverage was 25 and 19% lower than control, respectively (Table 3).

Weed coverage: weed coverage was affected (p≤ 0.05) by the treatments and time (Table 4). In all evaluations, chemical control treatments had the lowest weed coverage at 30 DDAT, values ranged from 1 to 9% at 60 DDAT, from 0 to 6% and at 75 DDAT from 0 to 3%. Weeding + fertilization + harrowing and weeding + harrowing treatments showed the highest coverage; at 75 DDAT, were higher than control in 64 and 29%, respectively; while with weeding and weeding + fertilization, weed coverage was statistically similar to control (Table 4). In the rainy season from 2012, the lowest weed coverage was observed in the chemical weed control treatment, with values ranging from 1 to 14%; while in control and treatments with mechanical work, weed coverage decreased gradually, presenting the lowest coverage at 440 DDAT. Treatments in which harrowing was applied weed coverage was higher; in weeding + harrowing, was 53, 44 and 26% and in weeding + fertilization + harrowing, 61, 35, and 27%; which are higher values than those from control at 320, 390 and 440 DDAT (Table 4).

Plant density of palisade grass: plant density of palisade grass showed significant changes (p≤ 0.05) by effect of treatments (Table 5). During the experimental period, the lowest plant density was for the treatments that included harrowing, whose values ranged from 2.4 plants m^-2 at 30 DDAT to 4.5 plants m^-2at 440 DDAT. Moreover, chemical weed control treatment showed the highest density of palisade grass; at 30 DDAT density ranged from 4.3 to 5.8 plants m^-2 and at 440 DDAT, was 7 to 8.5 plants m^-2.

Weed plant density: a significant effect was detected (p≤ 0.05) in weed population by effect of treatments. At 60 DDAT in chemical control treatments, showed the lowest density of weeds (less than 6 plants m^-2); while in treatments that included weeding, harrowing and fertilization, it was over between 290 and 350% regarding control (Table 6).

Palisade grass biomass: palisade grass biomass showed significant changes (p≤ 0.05) by effect of treatment (Table 7). At 75 DDAT, the highest biomass production was obtained in chemical weed control treatments, exceeding control by 38%; treatment 2,4-D was the one that produced less biomass, exceeding control by 22% and in agricultural works treatments produced biomass was lower than control.

At the beginning of the rainy season from 2012, at 320 DDAT, in chemical control treatments biomass production of palisade grass increased from 1122 to 1370 kg MS ha^-1. Biomass in weeding and weeding + fertilization treatments were statistically equal to control; whereas in the treatments with harrowing, biomass was lower than control. As the rainy season advances at 390 and 440 DDAT, biomass production of palisade grass from control increased and was statistically similar to that of some of the chemical control treatments. In weeding and weeding + fertilization treatments, biomass was statistically equal to control; whereas, in the two harrowing treatments, it was lower than the control at 320, 390 and 440 DDAT.

Discussion

The degradation degree of a meadow can be determined by the vegetation cover of desirable species, as well as that of the undesirable species or weeds, bare areas of vegetation and biomass production (^{Padilla et al., 2009}; ^{Oliveira et al., 2004}; ^{Silva et al., 2004}). The experimental meadow had an initial coverage of palisade grass of 27.2%, weeds 55.7% and other grasses 14%, indicating that it is a meadow with a moderate level of degradation (^{Padilla et al., 2009}). With chemical weed control, palisade grass coverage increased to values close to 90% or higher, indicating that the meadow became productive.

The largest increase in coverage of palisade grass with herbicide treatments was due to efficient weed control without damaging the grass, because these have a systemic action, destroying the whole plant, preventing re-growth, which reflected in less coverage and weed density (Table 4; Table 6). These results are consistent with those reported by ^{Esqueda and Tosquy (2007)} Pangola grass. Because weeds compete for space and nutrients with the species of interest and also immobilize them in their tissues (^{Moreno et al., 2000}; ^{Dias-Filho, 2007}), when controlled, left nutrients and space available for growth of palisade grass, favoring its development, manifesting itself in increased coverage and greater accumulation of biomass (Tables 3; Table 7), similar to that determined by ^{Esqueda and Tosquy (2007)}, in Pangola grass; ^{Esqueda et al. (2009)}, in Estrella de Africa grass and ^{Esqueda et al. (2010)}, in Llanero grass.

Weeding and weeding + fertilization treatments were not efficient for the rehabilitation of the meadow, having a similar condition to that of control. This was because only provided temporary weed control, allowing the formation of new growths and the emergence of new plants from seeds in the soil (^{Silva Dias-Filho, 2001}, ^{Valbuena and Acosta, 2006}; ^{Pellegrini et al., 2007}) which is consistent with the results of other experiments (^{Valbuena and Acosta, 2006}; ^{Esqueda and Tosquy, 2007}; ^{Pellegrini et al., 2007}; ^{Esqueda et al., 2009}; ^{Esqueda et al., 2010}), which reached coverage and density similar to control. The presence of weeds had a negative effect on the growth of desirable species (^{Moreno et al., 2000}; ^{Dias-Filho, 2007}), which reflected in low biomass accumulation of palisade grass, as has been reported in Estrella de Africa, Llanero and Pangola (^{Esqueda and Tosquy, 2007}; ^{Pellegrini et al., 2007}; ^{Esqueda et al., 2009}; ^{Esqueda et al., 2010}).

Weeding + harrowing and weeding + fertilization + harrowing had negative effect on the meadow, because harrowing damaged both weeds and palisade grass, causing reduction in the number of plant, coverage and biomass of this (Table 3, Table 5 and Table 7) and therefore nulling a

possible response of fertilizer application, consistent with the results of ^{Silva et al. (2004)} in Brachiaria humidicola. Due to the damage to plants, palisade grass density in these treatments were lower than the optimal set number, four to eight plants m^-2 (^{Ara et al., 2004}); the rest of the treatments and control maintained plant density within the optimum range. The spaces left by the death of palisade grass plants were occupied by weeds generated from new growths and seed reserve in soil (^{Silva Dias-Filho, 2001}; ^{Valbuena and Acosta 2006}; ^{Pellegrini et al., 2007}) resulting in higher coverage and weed biomass.

Conclusions

Chemical weed control with 2,4-D, picloram + 2,4-D, metsulfuron methyl and aminopyralid + metsulfuron methyl, improves the condition of the meadow: increasing the coverage of palisade grass, reduces weed coverage and density, resulting in greater biomass accumulation of palisade grass.

Weeding + harrowing and weeding + fertilization + harrowing deteriorated the condition of the meadow: reducing coverage, density and biomass production of palisade grass and increase of weed cover.

Weeding and weeding + fertilization do not improve the condition of the meadow.