Modelling long-term recruitment patterns of blue mussels Mytilus galloprovincialis: a biofouling pest of green-lipped mussel aquaculture in New Zealand

Abstract

The green-lipped mussel Perna canaliculus forms the cornerstone of the New Zealand aquaculture industry. Like shellfish farming globally, P. canaliculus aquaculture is susceptible to the detrimental effects of biofouling. One of the greatest ongoing threats results from the recruitment of blue mussels Mytilus galloprovincialis onto cultured P. canaliculus and ropes used for collecting or on-growing juvenile stock. Using a dataset spanning 40 yr, we modelled spatio-temporal patterns in M. galloprovincialis recruitment in relation to a suite of potential explanatory variables and tested the ability of the best model to forecast recruitment. Our goal was to identify locations (sites and water depths) and times of peak M. galloprovincialis abundances, to enable the industry to implement management practices to minimise operational risks. Despite large inter-annual and spatial variability in recruitment patterns, our analyses revealed an upward trend in the abundance of M. galloprovincialis over the last 2 decades. Generally, seasonal patterns in abundance showed a large recruitment peak in October and a smaller one in April. There was a strong negative effect of depth, and weak effects of sea surface temperature and Southern Oscillation Index (in the previous month) on M. galloprovincialis abundances. The best model fitted the data well (R2 = 0.72); however, it provided only moderate forecasting power (R2= 0.16), highlighting the challenges in forecasting recruitment. These results have been incorporated into a web application that enables aquaculture companies to interactively investigate historic trends in M. galloprovincialis and forecast abundances in the month ahead.