Feedbacks between benthic carbon mineralisation and community structure: a simulation-model analysis
Duplisea, D.E. (1998). Feedbacks between benthic carbon mineralisation and community structure: a simulation-model analysis. Ecol. Model. 110: 19-43
In: Ecological Modelling. Elsevier: Amsterdam; Lausanne; New York; Oxford; Shannon; Tokyo. ISSN 0304-3800; e-ISSN 1872-7026, more
|  |
| Keywords |
Aquatic communities > Benthos > Zoobenthos
Chemical compounds > Sulphur compounds > Sulphides
Chemistry > Geochemistry > Sediment chemistry
Composition > Community composition
Cycles > Chemical cycles > Geochemical cycle > Biogeochemical cycle
Ecosystems
Environments > Aquatic environment > Benthic environment
Mineralization
Models
Organic matter > Carbon > Organic carbon
Oxygen consumption
Sediment mixing > Bioturbation
Marine/Coastal |
| Abstract |
A simulation model was constructed as a means to establish and test causal links between organic carbon mineralisation and community structure in a generalised sublittoral, soft-bottom, infaunal, temperate, coastal, marine, benthic community. Two hypothesised positive feedbacks were explicitly included: bioturbation by macrofauna and sulphide accumulation and toxicity to aerobic organisms, that link benthic organisms and the sediment chemical environment. The magnitude and dynamics of bacterial, meiofaunal and macrofaunal biomasses, and benthic O2 consumption simulated over a seasonal regime were reasonable compared with field data. CO2 production was generally underestimated, yet within the range of published, empirical values, and the dynamics compared well with empirical data. Steady state simulations indicated a reasonable and ecologically explicable collective behaviour, such as anaerobic processes dominating total carbon mineralisation and sulphide oxidation accounting for the majority of benthic O2 consumption at high carbon input levels. Sensitivity analysis of selected model parameters indicated that the positive feedbacks of bioturbation and sulphide toxicity were more important than other selected parameters in determining the model outcome. |
|
