Biomass accumulation after 10-200 years of shifting cultivation in bornean rain forest

I examined the effect of repeated long-fallow shifting cultivation on biomass accumulation in a rain-forest landscape occupied for over 200 years in West Kalimantan, Indonesia, The diameters of stems >5 cm dbh were measured in nine secondary forests using a stratified random design of nested plots (3000 m(2) per site). The stands were 9-12 years old and had experienced from I to 10 or more cycles of shifting cultivation. Accounting for differences in soil fertility, analysis of covariance (ANCOVA) demonstrated that aboveground live biomass increment (ABI, in megagrams per hectare per year) was significantly greater after four cycles than after six or more cycles. ABI was lowest after two cycles, highest after one or four cycles and intermediate after 6-10 cycles. Differences were driven by the density and biomass of trees >10 cm dbh, which followed a similar pattern. Although the diameter increment and contribution to total biomass of trees >10 cm dbh increased significantly with inherent soil fertility, variation in the rate of biomass accumulation depended primarily on the number of prior cultivation cycles. An ANCOVA with soil fertility as the main effect was not significant. One hypothesis to explain observed patterns of ABI associated with changes in tree density is a shift, over many cycles, in the dominance of seed-banking vs. resprouting species. Previous work has shown that total phosphorus did not decline after six or more cycles; however, a shift from available to occluded forms may be related to lower ABI in sites with the longest cultivation history. An 11% decline in ABI could substantially alter the carbon-sequestration value of secondary tropical forests as they enter their second century of persistent human disturbance.