SIZE-BASED BIOMASS AND PRODUCTION MODELS IN THE ST-LAWRENCE GREAT-LAKES

Recent theoretical advances indicate that the size distribution of nor malized biomass in aquatic food webs can be modelled by a primary line ar scaling and a series of repeated quadratic curves corresponding to component trophic groups. These secondary parabolas all have the same curvature and are shifted fixed distances from one another. Hence the parabolic normalized biomass distribution for any trophic group can be predicted from a given trophic group. We use mean annual zooplankton data from Lakes Ontario and Erie to predict fish biomass spectra, and then convert fish biomass to annual production using established body size-dependent production relationships. For Lake Ontario we estimate fish production to be 19.2 +/- 5.0 fresh g m(-2) yr(-1), which compare s favourably with independent growth model estimates. Lake Erie total fish production is estimated to be 22.1 +/- 5.2 g m(-2) yr(-1), of whi ch combined annual sports and commercial harvest constitute 4-6%. Thes e size-based production techniques are less costly and time-consuming than traditional species-based approaches. (C) 1995 International Coun cil for the Exploration of the Sea