| Summary: | 碩士 === 國立臺灣大學 === 海洋研究所 === 103 === Influenced by growth, maturation, and mortality history, size structure represents a key demographic characteristic of fish populations, and plays an important role in maintaining stability and reproductive output. There are two important external forces that potentially shrink the size structure of a fish population: fishing and warming. Nevertheless, the relative importance of fishing and warming in affecting size structure of fishes has not been quantified. In addition, existing analysis focus on size-based indicators (SBIs) to represent the size structure of fish population, but some of them may not effectively quantify external forces on size structure. To bridge the knowledge gap, we used variation partitioning approach to quantify the relative contribution of temperature, fishing on size structure of exploited stocks, and then test whether these two factors have an interactive effect. We analyze the size structure (length frequency) of 23 exploited stocks in the Northeast Pacific, Northeast Atlantic, Mediterranean Sea, North Sea, Baltic Sea and Arctic Sea. Our results showed that the variance of size structure is affected both by fishing (explaining on average of 9.29%) and temperature (explaining on average of 9.29%). In comparison, the interactive effect on size structure was very subtle (on average of 1.79%). We then examined what determined the relative contribution of fishing and temperature. We found that the relative contribution was related to life history traits, mean fishing mortality, and increasing rate of temperature. Specifically, our result showed that fishing effect on size structure significantly decreased with the growth rate, suggesting the size structure of K-selected species was influenced more by fishing. Furthermore, the temperature effect significantly increased with the intensity of fishing, suggesting that fishing makes the exploited stocks more vulnerable to environmental changes. Finally, by comparing our results of variation partitioning approach versus indicators-based analysis, we found that variation partitioning serving as a size structure-based analysis outperformed analysis using SBIs in terms of the lower p-values, suggesting that the size structure-based analysis featured the adoption of variation partitioning is a more suitable method to investigate the external effects on size structure of fish populations.
|
|---|
