Environmental influence on mussel (Mytilus edulis) growth - A quantile regression approach

详细信息    查看全文

• 作者：Per Bergströ ; m ; ^{per.bergstrom@marine.gu.se" class="auth_mail" title="E-mail the corresponding author} ; Mats Lindegarth
• 关键词：Growth ; Modelling ; Mytilus edulis ; Quantile regression
• 刊名：Estuarine, Coastal and Shelf Science
• 出版年：2016
• 出版时间：20 March 2016
• 年：2016
• 卷：171
• 期：Complete
• 页码：123-132
• 全文大小：1637 K

文摘

The need for methods for sustainable management and use of coastal ecosystems has increased in the last century. A key aspect for obtaining ecologically and economically sustainable aquaculture in threatened coastal areas is the requirement of geographic information of growth and potential production capacity. Growth varies over time and space and depends on a complex pattern of interactions between the bivalve and a diverse range of environmental factors (e.g. temperature, salinity, food availability). Understanding these processes and modelling the environmental control of bivalve growth has been central in aquaculture. In contrast to the most conventional modelling techniques, quantile regression can handle cases where not all factors are measured and provide the possibility to estimate the effect at different levels of the response distribution and give therefore a more complete picture of the relationship between environmental factors and biological response. Observation of the relationships between environmental factors and growth of the bivalve Mytilus edulis revealed relationships that varied both among level of growth rate and within the range of environmental variables along the Swedish west coast. The strongest patterns were found for water oxygen concentration level which had a negative effect on growth for all oxygen levels and growth levels. However, these patterns coincided with differences in growth among periods and very little of the remaining variability within periods could be explained indicating that interactive processes masked the importance of the individual variables. By using quantile regression and local regression (LOESS) this study was able to provide valuable information on environmental factors influencing the growth of M. edulis and important insight for the development of ecosystem based management tools of aquaculture activities, its use in mitigation efforts and successful management of human use of coastal areas.