Reproductive potential : the effects of population structure, condition, egg quality and spawning location of Atlantic cod (Gadus morhua) and haddock (Melanogrammus aeglefinus)
Over the last 40 years fisheries science has atrophied within the straight-jacket of fisheries management dogma. Management models, which insist in predicting future fish populations by using a single index, spawning stock biomass (SSB), have led to a lack of questioning on how the many aspects of flexible reproductive traits can affect reproductive potential. In this thesis I produce and use an individual based modelling approach to link empirical relationships with flexible reproductive parameters to quantify and qualify the effects that individual female size, condition, egg quality, spawning site quality and mortality during spawning can have on stock reproductive potential (SRP) and the temporal distribution of reproductive output. Model construction and sensitivity analysis have highlighted the need for essential data on reproductive traits, such as the connection between female size, condition, batch number and subsequent egg quality. The results of a strategic study, using a wide range of size-selectively harvested populations, indicate that even across populations with the same SSB, changes in age/size structure can reduce SRP up to 74%. The truncation of size structure alone will lead to a shortening of the spawning season by 4 weeks and a 2 week shift in the data of peak spawning. The effect of low condition of individuals can lead to almost total reproductive failure, whereas the effect of increased condition is very dependent upon population structure. Fishing during the spawning season kills season kills serial spawning fish before and while they are still releasing eggs. Model simulations suggest that not fishing or not targeting larger fish during spawning can lead up to a 53% increase in reproductive potential. The model has been parameterised for North Sea cod and haddock and run for the years 1963 to 1999 with an array of different levels of realism/complexity of input data.