MEPS 489:163-175 (2013)  -  DOI: https://doi.org/10.3354/meps10409

ABSTRACT: The Brey model is one of the most frequently used methods to obtain a quick estimate of the secondary production (P) of an area. It is based on an empirical relationship between the production/biomass (P/B) ratio and the (annual) mean weight (W) of the individuals of a population. Estimates of P/B by this model are frequently obtained by using only single measurements of W and B, thus circumventing tedious efforts required by conventional methods. The obtained P values of communities are sums of estimates made for individual species. Any constraints of the model can be fully understood only by evaluating it for single-species populations. Using an extensive data set obtained by monitoring a population of the bivalve Macoma balthica for 33 yr, we evaluated the model by comparing Brey model estimates of P and P/B with direct annual estimates. We corroborate the basis of the model by presenting a significant relationship between observed annual values of W and P/B. The model satisfactorily predicted P when late-winter (but not late-summer) assessments of W and B were used. The model underestimated P/B in the years with high mortality rates (Z), whereas it overestimated P/B in almost all other years. Z values were a better basis for predictions of P/B than W values. The model could predict P/B well on the exclusive basis of W due to the significant correlation between W and Z (low Z values resulted in older and thus heavier individuals). Multi-year averages of model-predicted and observed P/B estimates were similar only when predictions were based on late-winter or annual (not on summer) estimates of W and B. In conclusion, the model cannot be recommended for precise and unbiased P estimates in a single species when no more than a once-only assessment of W and B is available.

KEY WORDS: Biomass · Secondary production · Individual weight · Mortality rate · Macoma balthica · Bivalve · Wadden Sea · Tidal flats · Long-term data series