Detritus decomposition and nutrient dynamics in a forested headwater stream

• We modeled detritus decomposition in a stream ecosystem.
• The models are stoichiometrically explicit.
• The models include immobilization, nutrient mining of detritus, and different leaf materials.
• The most inclusive model best predicted inorganic nutrient concentrations in the water column.

Field experiment and modeling studies have shown that microbial processes during leaf decomposition can modify detritus nutrients and inorganic nutrients in the system. In this study, we developed three models (Models I, II, and III) for leaf decomposition in a shaded headwater stream with little nutrient input, and we predicted the nutrient patterns in the stream. We used data from Hugh White Creek, NC, and synthesized several important ecological concepts from previous studies, i.e., ecological stoichiometry, microbial nutrient mining, and microbial–substrate interaction. In Model I, we calculated decomposition based on microbial metabolism rather than using first-order decay; in Model II, we further added an alternative microbial nutrient acquisition mechanism in which microorganisms (miners) mine detritus nutrients rather than using nutrients from stream water; and in Model III, we additionally broke detritus into three substrate groups (labile, intermediate, and recalcitrant material). Results showed that Model I fit the nutrient data the poorest; for the other two models, Model III performed better in fitting nitrate and Model II did slightly better in fitting ammonium; Models II and III predicted similar nutrient patterns, which were different from the patterns by Model I, suggesting that microbial mining may play an important role in decomposition.