BIOMASS AND LITTER ACCUMULATION UNDER MANAGED AND NATURAL TROPICAL FALLOWS

The accumulation of biomass and litter and changes in vegetation struc ture by six managed leguminous fallows, planted in a previously cultiv ated Ultisol (Acrisol) in the Peruvian Amazon, were compared with that of naturally established secondary vegetation. The planted fallows in cluded: Centrosema macrocarpum (CM), Pueraria phaseoloides (PP), Stylo santhes guianensis (SG), Desmodium ovalifolium (DO), Cajanus cajan (CC ), and Inga edulis (IE). The aboveground mass of planted legumes, tree s, herbaceous vegetation, and litter and root mass were measured at 4, 8, 17, 29, 41, and 53 months after fallow initiation; leaf area index (LAI) and leaf area density (LAD) were measured at 4, 8, and 17 month s. After 8 months, total mass in the IE, CC, and DO treatments was sig nificantly greater than in the CM and PP treatments; total mass in the natural fallow (NF) treatment was similar to that in the IE, CC, and DO treatments after 29 months. Aboveground biomass contributed about 7 5-85% of total mass and was greater when trees were present. Abovegrou nd biomass ranged from 8 (CM) to 63 t ha-1 (IE) at 53 months. Planted legume biomass in most treatments increased until 17 or 29 months, but later declined when many of the legumes were replaced by invading tre es. By 53 months, planted legume biomass represented less than 10% of aboveground mass in all treatments except CM and IE. The LAI increased from 4 to 17 months, but there were few differences among treatments at any date; LAD greatly decreased with time in the treatments with tr ees. The formation and maintenance of dense vegetation canopies with h igh LAD in the CM, SC, and PP treatments inhibited the invasion and gr owth of trees. Biomass data at 53 months suggest that the IE and DO tr eatments accumulated similar amounts of biomass and litter as the NF t reatment because of the rapid establishment of planted legumes and the successful invasion of trees and that the CM, SC, and PP treatments p erformed poorly because of the exclusion of trees. Further research co mbining trees and other leguminous or non-leguminous vegetation is req uired to develop management practices that maximize biomass accumulati on during managed fallows.