Rain forest on Maraca Island, Roraima, Brazil: Soil and litter process response to artificial gaps

Artificial canopy gaps were made by felling trees in rain forest on Maraca Island. There were three replicates of each gap size: small (c. 40 m2), medium (c. 150 m2), large (c. 350 m2), very large (c. 2500 m2), and unfelled forest controls (2500 m2). Changes in microclimate, soil chemistry, and soil biological processes including decomposition were measured for up to 12 months. Higher maximum and lower minimum air temperatures occurred with increasing gap size but there was little consistent effect of gap size on soil chemistry. Microbial biomass, soil respiration, and nitrogen mineralization and nitrification were investigated in medium gaps. No differences were found between the gaps and the unfelled forest, and all had their lowest values in the dry season. Within the medium gaps higher values for these variables occurred in the 'under crown' position (where the crowns of the felled trees had fallen under the surrounding canopy). Three litter-bag decomposition experiments were made over 217 d. No relationship was found between decomposition and nutrient release rates and size of gap or position within it and no clear trend was found with position along a central trunk within a gap, but a higher weight loss and nutrient release occurred under the crowns of fallen trees. It is concluded that decomposition and nutrient release rates are not influenced in a consistent way by gap size. It may be that small areas of high nutrient release can be caused by the local addition of large quantities of substrate, even though process rates remain unaltered. Because large additions of readily decomposable substrate are most likely to occur under a closed canopy, there is no support from the Maraca data that growth responses to gaps will involve an interaction between enhanced nutrient supply and increased light