Clay content affected soil fertility at all depths within these highly
weathered Acrisol soils, supporting our first hypothesis. The impacts
of land-use change on soil biochemical characteristics and soil nutrient
stocks at all depths were limited, and thus only partly support our second
hypothesis. Only soil pH was altered by land-use change across all
depths in each landscape largely due to the legacy of ashes from biomass
burning during conversion and to liming practices employed in
our oil palm plantations. Since the majority of the overall variation in
soil biochemical characteristics was among replicate plots, except for
SOC stocks, which exhibited larger proportions of variation between
landscapes than variation among plots, this did not support our third
hypothesis. Our results indicate that in order to detect significant effects
of land-use change on soil biochemical characteristics in our nested
sampling experimental design, more replicate plots per land-use type
within each landscape should be sampled. The high spatial variability
of soil biochemical properties at the landscape scale illustrates that for
extrapolation over large spatial scales, stratification must be based on
the main driver (e.g., soil texture and/or land-use age within the same
soil group and climatic condition) of the soil properties of interest.
Therefore, future studies should focus on a range of soil parameters,
test variability among plots or sites and determine the number of replications
at the landscape level in order to determine overarching effects
of land-use change on soil properties.