Geographic information systems in rainforest ecosystem research: the Sarawak pilot model

Abstract

In 1963, a 20 ha permanent ecological research plot, now Forest Department Research Plot (RP) 146, was established in an area of Kerangas and Mixed Dipterocarp forest in Sabal Forest Reserve, West Sarawak. All trees over 10 cm dbh were enumerated and the topography sketched in 20 × 20 m recording units (exactly: chain-squares). The trees were not mapped because we did not realize at the time the great heterogeneity of soils and species associations. The conventional traditional textbook analyses of species richness, tree species diversity (evenness of mixture), and general growing stock patterns and densities in the sixties were easily determined (Weiscke, 1982). However, we encountered problems with determining associations between tree species, between species and soil features, and between patterns and assumed dynamic trends. The chain-square grid was too coarse to catch the fine fabric of soil and vegetation heterogeneity which varies at a scale of metres and even less. Another disadvantage was that plant locations within the 20 × 20 m square were not recorded. This made it impossible to analyse species associations and cohort dynamics among the smaller trees. Unless dynamic cohorts and neighbourhood conditions can be determined and observed in time series of remeasurements, any measurements of growth, nutrient stocks and cycles and hydrological conditions will be difficult to interpret. The shifting-mosaic dynamics have proved to be so strong and the time-series of patterns of the shifting mosaics among the smaller trees so turbulent (Poker 1992) that research into processes which are influenced by the autecological properties and physiological state of the trees at the point of measurement is a shot in the dark unless the fine-grained spatial information is available.