Contribuição de leguminosas arbóreas em sistemas silvipastoris com Brachiaria decumbens Stapf

Abstract

Understanding the contribution of legume plant fractions in silvopastoral systems is critical to enhance efficiency of ecosystem services in these systems. This research evaluated litter deposition, litter decomposition, biomass accumulation, and chemical composition of gliricidia [Gliricidia sepium (Jacq.) Kunth ex Walp.] and sabi (Mimosa caesalpiniifolia Benth) intercropped with signal grass (Brachiaria decumbens Stapf.). Litter deposition was measured by 0.5 m2 quadrats every 28 days, in a perpendicular transect across the tree rows, from 0.5 to 3.0 m away from the tree trunk, in 2012 and 2013. Other response variables measured included litter N, plant N derived from atmosphere (%Ndda) using the natural abundance technique, lignin, C:N, and ligninN ratios. Two decomposition trials were performed in an exclusion area. In the first trial, leaves were incubated in litter bags. In the second trial, branches with three circunference classes were incubated. In both trials, bags were collected in eight incubation times (0, 4, 8, 16, 32, 64, 128, and 256), in 2011 and 2012. Response variables measured in the decomposition trials included N, lignin, P, C:N, and lignin:N for each incubation period. Tree biomass was determined every six months. Measurements included tree height, stand, number of timbers, biomass, total N, N derived from atmosphere (%Ndfa) using the natural abundance technique, lignin, and C:N ratio of each component (leaves and branches within the three circunferences classes). Density and higher calorific power (HCP) were also determined for the branches. Annual litter deposited by sabi (4540 kg OM ha-1) was greater (P ≤ 0.05) than gliricidia (4200 kg OM ha-1). Gliricidia N concentration was (22,4 g kg-1) 20,4 % greater (P ≤ 0.05) than sabi (18.6 g kg-1), leading to greater litter N input for gliricidia (105 kg ha-1) than sabi (87 kg ha-1). Biological N fixation (BNF) did not differ between tree species ( P > 0.05) with Ndfa ranging from 51 to 70% for gliricidia and 43 to 61% for sabi, equivalent to 64 and 46 kg ha-1 yr-1 of N, respectively. Sabiá C:N ratio (23) was greater (P ≤ 0.05) than gliricidia (19). Lignin concentration ranged from 17 to 30% and lignin:N from 5:1 to 21:1 along the evaluation periods and did not differ (P > 0.05) between species. Gliricidia leaves decomposed faster (k = 0,0038 g.g-1.dia-1) than sabi leaves (k = 0,0012 g.g-1.dia-1), leading to faster nutrient released by gliricidia. Branches of gliricidia also decomposed faster (k = 0,0018 g.g-1.dia-1) than branches of sabi (k = 0,0005 g.g-1.dia-1). Leaves released 73 and 33% of its original N, leading to an annual input of 65 and 42 kg ha-1 N via leaf decomposition, for gliricidia and sabi, respectively. Remaining N was inversely correlated to branch circumference for gliricidia. Branch mineralization of N was 38% for gliricidia and 26% for sabi, with N contributions of 6 and 1 kg ha-1, respectively. Leaf P concentration was 3.1 g kg-1 at day zero, reducing to 1.5 g kg-1 at day 256. Leaf lignin content increased around 165.6 g kg-1 until 184 g kg-1 after 32 d and 210 g kg-1 after 64 d of incubation. Lignin was greater in sabi than in gliricidia, for the three branch circumference classes, reflecting better timber quality and longevity. Leaf Lig:N at time zero was greater for sabi (5:1) than gliricidia (4:1), reflecting the slower decomposition rate for sabi. Gliricidia stand (3070 trees ha-1) was greater than sabi stand (2840 trees ha-1), with plant mortality of 15 and 21% in regard to the initial stand (3600 seedlings ha-1). Branches with greater circumference (class 3) made the greatest contribution for total tree biomass, with 58 and 54% for gliricidia and sabi, respectively. Leaf represented the lowest contribution for total biomass, ranging from 7 to 13% for gliricidia and from 4 to 14% for sabi, and was the fraction with most variation among the months. Leaf and branch nutrient did not vary much along the cycles. Sabi presented greater annual litter deposition and greater litter C:N, reducing decomposition rate. Different decomposition patterns from leaves and branches showed the distinct ability of tree legume to recycle nutrients. The N input and the BNF of gliricidia and sabi represents an opportunity to add N to grassland ecosystems and the other added benefits that come from the tree legume.