Analysis of some sources of error in methods used to determine fine root production in forest ecosystems: a simulation approach

Abstract

Fine root production rates are most commonly calculated from periodic measurements of live and dead fine root biomass. The accuracy of production estimates based on this method is very sensitive to violations of the inherent assumptions, particularly the assumption that the processes of fine root production and mortality are temporally separate. A simple model was used to simulate data for a variety of seasonal patterns of live and dead fine root biomass. Fine root production and mortality rates were calculated from these simulated data using two different computational methods. Comparison of the calculated rates with the known rates (the rates used to generate the seasonal patterns) revealed that violations of the above assumptions can result in inaccurate rate estimates. When fine root production and mortality occur simultaneously within a sampling interval, the calculated production rate will greatly underestimate the true value. Additional error in the rate estimates may result from sampling error associated with the fine root biomass data. The model suggested that sampling error can cause either overestimation or underestimation of fine root production.