Soil carbon dynamics of conventional tillage and no-till agroecosystems atGeorgia Piedmont - HSB-C models

This model was designed to simulate the dynamics of soil organic matter (SO M) and biomass change of detritus soil food webs after the application of c rop residue in agroecosystems. Background data and literature parameters we re used to initialize the model and a short-term field experiment was condu cted to calibrate the model. Our model is a synthesis and integration of ma ny others' previous studies because many others' ideas on food web or SOM m odeling were incorporated into our model. In our model, soil microorganisms were considered to be the major decomposers of organic matter with soil an imals playing an important role in decomposing organic matter by controllin g the population of soil microorganisms through trophic interactions. Thoug h we did not model the soil food web into species or functional groups, we attempted to represent the hierarchical structure and trophic interactions of soil food webs as complete as possible. The whole soil food web was conc eptualized as a 'super organism' to process organic matter in the model dia gram. State variables of our model were categorized into soil organic matte r pools, inorganic carbon pool, and soil organism pools. Flows were modeled based on first order kinetics. A tillage effect was introduced in our mode l by assuming that soil microorganisms, microarthropods and earthworms decr ease temporarily immediately after the tillage operation. We also introduce d a 'minimum population' protection mechanism for each group of soil organi sms by assuming that predation ceases when energy obtained can not offset t he energy expended in prey-searching and feeding. We found that the simulat ion outputs fitted the measured data very well, the simulation outputs were in the range of field measurements in most cases. However, there was a sig nificant discrepancy in respiration between simulation output and field mea surement in both conventional tillage (CT) and no-till (NT), particularly i n NT agroecosystems. We hypothesized that plant roots should be responsible for this discrepancy. The sensitivity analysis of our model showed that th e population of soil microorganisms was controlled by resources rather than predators. However, the populations of soil protozoa, nematodes and microa rthropods were controlled by both resources and predators. (C) 2000 Elsevie r Science B.V. All rights reserved.