Abstract
Large areas of tropical forests are affected by anthropogenic disturbances of various intensities. These disturbances alter the structure of the forest ecosystem and consequently its carbon budget. We analysed the role of fine root dynamics in the soil carbon budget of tropical moist forests differing in disturbance intensity. Fine root production, fine root turnover, and the associated carbon fluxes from the fine root system to the soil were estimated in five stands ranging from an old-growth forest with negligible anthropogenic disturbance to a cacao agroforestry system with planted shade trees. Annual fine root production and mortality in the three natural forest sites decreased with increasing forest disturbance, with a reduction of more than 45 % between the undisturbed forest site and the site with large timber extraction. However, the agroforestry sites showed intermediate fine root production and mortality values, due to the presence of crop species or the replacement of natural shade tree species by planted shade tree species. The amount of carbon annually added to the soil carbon stock through fine root litter production was highest in the undisturbed forest and decreased with increasing forest use intensity. Only in the plantation with planted shading trees the fine root C flux again reached an intermediate level. However, the relative importance of root C in the total above- and below-ground C input to the soil increased with increasing forest use intensity. We conclude that moderate to heavy disturbance in these tropical moist forests had a profound impact on fine root turnover and the related carbon sink strength of the below-ground compartment.
Key words: forest disturbance, fine root production, fine root mortality, carbon sequestration, agroforestry, land use change, root litter.
Introduction
Tropical forests are being converted into agricultural land or agroforestry systems at an alarming rate (Nepstad et al. 1999; Achard et al. 2002). In the South-east Asian rainforest region, a variety of low to medium intensity forest use practices are common, including rattan harvest, selective timber extraction and the establishment of small-scale agroforestry systems in the forest margin zone (FWI/GWI, 2002). Forest conversion not only leads to a loss in species diversity (Reiners et al. 1994; Fujisaka et al. 1998; Murdiyarso et al. 2002), but also has a profound effect on the ecosystem‟s carbon budget (Raich 1983; Lal 2005;
Jandl et al. 2006). Changes in forest structure and management practices are likely to alter soil organic carbon content by changed input rates of organic matter, changed decomposability of organic matter and changes in soil moisture and temperature regimes, influencing decomposition rates (Post & Kwon 2000, Lal 2005). Despite the efforts that have been made to investigate the effects of forest disturbance and forest use intensity on soil carbon stocks (e.g. Raich 1983; Jaramillo et al. 2003; Li et al. 2005; Trumbore et al. 2006), there is still an ongoing debate about whether short-rotation plantations, secondary forests or old-growth forests are more effective in the sequestration of atmospheric CO2. For example, Guo & Gifford (2002) concluded from a meta-analysis that conversion of forest to tree plantations with broadleaved trees has little effect on soil carbon stocks, whereas planting of conifer trees reduces the soil carbon stock by about 15 %. In neo-tropical pastures, Silver et al. (2004) found that maintenance of older reforested sites, as opposed to short-rotation systems, is related to a much higher carbon sequestration in long term carbon pools, in particular in recalcitrant soil organic matter. Correspondingly, Schulze et al. (2000) state that, in contrast to the carbon sink management proposed in the Kyoto protocol, preservation of natural old-growth forests may have a larger effect on carbon sequestration than promotion of forest regrowth. These authors argue that, with increasing life-span of the stand, proportionally more carbon should be transferred into the soil carbon pool with the turnover of leaves and roots. The importance of C input to the soil with the turnover of leaves and roots has e.g. been demonstrated in a study by Powers (2004), where total litter input could explain about 50 % of the observed differences in soil organic carbon concentrations (SOC) between various land-use types in Costa Rica. Still, relatively little is known on the role of fine roots in the C budget of converted tropical forests.
Allocation of C to the root system is one of the most important, yet least well quantified fluxes of C in the terrestrial ecosystems (Hendrick & Pregitzer 1992; Davidson et al. 2002;
Matamala et al. 2003). Especially in tropical forests, which are often characterized by rapid fine root turnover, fine rootdynamics can contribute significantly to ecosystem carbon fluxes (Silver et al. 2005). Fine root biomass and turnover may even be more important for the accumulation of carbon in forest soils than above-ground litter input (Lugo & Brown 1993;
Block et al. 2006). Below-ground net primary productivity (BNPP) in forests has been estimated at 30-50 % of the total net primary production (Fogel & Hunt 1979; Keyes & Grier
1981; Vogt 1991; Ruess et al. 1996; Xiao et al. 2003). Yet, data on changes in fine root production with conversion of tropical forests are rare (Vogt et al. 1996; Hertel et al. in press). The aim of the present study is to analyse the role of the fine root system in the soil carbon cycle of tropical moist forests differing in disturbance intensity. Based on an analysis of fine root dynamics, we estimated fine root production, fine root turnover and the associated carbon fluxes from the fine root system to the soil in five stages of a gradient in forest use intensity ranging from a near-natural old-growth forest with negligible anthropogenic disturbance to a cacao agroforestry system with planted shade trees representing a high degree of disturbance. We tested the following hypothesis: (1) An increase in forest use intensity will lead to a decrease in fine root productivity and turnover, as mature late-successional trees are supposed to invest more C in the root system, while younger, early successional species, as well as crop species should invest more C in above-ground structures (Tilman 1985; Yan et al. 2006); (2) Fine root related carbon sink strength decreases along the disturbance gradient due to thinning of the stands and lower fine root mortality rates due to a disturbance-driven release from competition.