Per Bergström

Modeling and predicting the growth of the mussel, Mytilus edulis: implications for planning of aquaculture and eutrophication mitigation

Abstract The increased pressure on the marine ecosystems highlights the need for policies and integrated approaches for sustainable management of coastal areas. Spatial planning based on geographic information of human activities, ecological structures and functions, and their associated goods and services is a fundamental component in this context. Here, we evaluate the potential of predictive modeling to provide spatial data on one ecosystem function, mussel growth for use in such processes. We developed a methodology based on statistical modeling, spatial prediction, and mapping for the relative growth of the blue mussel, Mytilus edulis. We evaluated the performance of different modeling techniques and classification schemes using empirical measurements of growth from 144 sampling sites and data on biological, chemical, and physical predictors. Following comparisons of the different techniques and schemes, we developed random forest models to predict growth along the Swedish west coast. Implemented into GIS the best model produced in this study predicts that low, intermediate, and high growth rates can be expected in 53%, 32%, and 15% of modeled area, respectively. The results of this study also suggest that the nature and quality of predictor data hold the key to improving the predictive power of models. On a more general note, this study exemplifies a feasible approach based on measuring, modeling, and mapping for obtaining scientifically based spatial information on ecosystem functions and services affected by a complex set of factors. Such information is fundamental for maritime spatial planning and ecosystem‐based management and its importance is likely to increase in the future. Because of its close link to nutrient assimilation and production yield, site‐specific information of soft tissue growth such as the map of predicted growth rate developed in this study can be used as a tool for optimizing actions aimed at mitigating eutrophication and aquaculture operations and in maritime spatial planning processes of coastal areas.

Monitoring Recruitment Patterns of Mussels and Fouling Tunicates in Mariculture

ABSTRACT Methods to increase the precision of spat collection and strategies to mitigate fouling are greatly needed in aquaculture production. As such, larval recruitment of mussels and a common tunicate species was investigated. Recruitment was measured in shallow (1–2 m) and deeper (4–5 m) water at three sites during the summer of 2012. In addition, to evaluate the importance of timing in deployment of mussel ropes, differences in mussel yield were examined. The settlement plates provided a good description of the settling community with high temporal resolution. Peaks in recruitment were observed for both mussels and tunicates but recruitment rates and the timing of peaks differed among sites. Although mussel larvae preferred shaded substrates at some sites and times, these substrates were consistently preferred by tunicates. Mussels preferred to settle at shallow depths, whereas tunicates were consistently more abundant deeper. In contrast to predictions, there was no positive relationship between the yield of mussels on ropes and settlement rates on corresponding weeks. Somewhat surprisingly, the final abundance of mussels and tunicates were not related to the length of the recruitment and growth period. These results indicate that not only initial recruitment, but also mortality and repeated recruitment events are important processes shaping these dynamic assemblages. Combining the results, a minimum recommendation for monitoring larval settlement is to use, at two depths, one monitoring unit with several dark-surfaced sampling plates. Considering the fact that timing of deployment of mussel ropes in relation to mussel and tunicate settlement has been identified as a problem, it is believed that such methods can be used to optimize production of mussels. Thus, studies like these can also contribute to optimize farming techniques and practices in a broader context.