Quantification of termite attack on lying dead wood by a line intersection method in the Kabili-Sepilok Forest Reserve, Sabah, Malaysia

1. A line intersection method was used to estimate abundance (technically linear abundance: m(1) m(-2)), biovolume (m(3) ha(-1)) and size class distribution (defined by diameter) of lying dead wood in tropical forest. Additional semi-quantitative protocols assessed decay state (4 classes), termite attack (5 classes) and live termite occupancy (3 classes).

2. Three forest types (kerangas, alluvial and sandstone) were sampled in the Kabili-Sepilok Forest Reserve of Eastern Sabah, using plots of 30 x 30 m. Approximately 50 man-hours were required per site, at a replication of three plots per site and three well-separated sites per forest type.

3. Mean biovolume of lying dead wood exceeded 8 x 103 m3 ha(-1) in kerangas (= heath) forest, with lower values in other types. Large items (> 19 cm diameter) were less than 10% of total abundance, but represented the largest biovolume, exceeding (alluvial) or equalling (kerangas) the total biovolumes of smaller categories combined. Most items (not less than 75%) were present as small wood (< 10 cm diameter). Items in the highest decay class had the highest biovolume.

4. Termite attack was greater in the kerangas, where nearly 90% of items showed evidence of consumption, compared with 58% in the alluvial and 40% in the sandstone forests. Over 40% of items in the kerangas contained live termites compared with 25% in the alluvial and 15% in the sandstone. Items in the highest attack class (= almost total internal destruction) represented about one-half of the total biovolume available in the alluvial and kerangas forest types, and about one-third in the sandstone.