摘要: Abstract This paper outlines the different ways in which plant respiration is modelled, with reference to principles set out Cannell and Thornley (Annals of Botany85: 55–67, 2000), first whole-plant ‘toy’ models, then within ecosystem or crop models using growth-maintenance paradigm, finally representing many component processes Hurley Pasture (HPM) Edinburgh Forest Models (EFM), both separate C N substrates from structure. Whole-plant can be formulated so that either maintenance growth take priority for assimilates, difference between total associated resynthesis degraded tissues. All three schemes converted dynamic give similar, reasonable predictions respiration, but all have limiting assumptions scope. Ecosystem use paradigm without separating structure, implicitly assume a fixed cost, uncoupled assimilate supply, rate coefficients chosen range measured values. Separation HPM EFM enables estimates made local growth, phloem loading, ammonium nitrate uptake, reduction, N2fixation other mineral ion leaving ‘residual maintenance’ term. The latter explicitly related substrate supply. Simulated changes grassland over season forest rotation show ratio gross canopy photosynthesis varies expected limited range, residual accounts 46–48% 36–42%, loading 10–12% components small remainder, ratios varying during rotation. It concluded approach extended represent processes, has considerable merit. connected reality at points, it gives more information, examined level as well predictions, open modification knowledge emerges. However, currently, there are still parameters require adjustment model acceptable.