生活史特徵與漁撈壓力對魚類族群空間分布的變異度-平均值關係之影響

Abstract

前人的研究指出，族群在空間中的豐度平均值(M)與變異度(V)呈現一冪次關係（泰勒冪次定律），意即V=aMb。許多因子可能影響指數b，如聚集程度、生長速率與繁殖率等；然而，係數a 所代表的意義卻尚無定論。本篇研究中，我們估算了一九五一年至二〇〇七年，南加州洋流生態系中二十九種海洋魚類族群豐度的空間平均值與變異度關係。我們並檢驗各物種的泰勒指數（b）是否會受到該物種的生活史特性影響。另外，在考量生活史特性的差異後，我們也藉由一般線性回歸（General Linear Model）比較商業目標魚種與非目標魚種的泰勒指數，檢驗漁業對魚類空間分布的影響。結果顯示，排除平均豐度的影響後，所有的生活史特徵都會與泰勒指數成顯著線性相關。以非目標魚種來說，具有r 生殖策略相關特徵的物種具有較高的泰勒指數; 然而，於漁業目標魚種中此現象卻不顯著。這可能是由於漁業壓力改變了目標魚種的族群平均生活史特徵（如最大體長），使得目標魚種有較高的泰勒指數，亦即分布變得較為密集。我們也建立了一個體基準模型（individual-based model），探討泰勒冪次定律的形成過程，以及改變族群年齡結構後，是否會改變泰勒指數。其結果顯示，透過簡單的族群增減及移動過程即可產生泰勒冪次定律。另外，發生年齡截斷效應（age-truncation）的物種，雖然並未於模型中產生較大的泰勒指數，其平均空間變異度與平均值的比值仍較控制組為高。

A power-law relationship between population variance and mean abundance, V=aMb, is commonly observed in ecology. Many factors have been proposed to influence ‘b’, such as aggregation degree, growth rate, and reproduction, although the interpretation of the intercept ‘a’ remains elusive. In this study, we estimated the spatial variance-mean relationship of 29 fish species collected from the southern California Current Ecosystem spanning from 1951-2007. We investigated whether Taylor’s exponent ‘b’ is related to life history traits of fishes. In addition, we examined the fishing impacts by comparing exploited versus unexploited species, accounting for life history variation using a general linear model. We found that after removing the influence of mean abundance, all life history traits play a significant role in determining the exponent. Unexploited species with traits related to r-strategy tend to have higher Taylor’s exponent. However, the relationship between the exponents and life history traits of exploited species is much weaker than that of the unexploited species. Our results suggest that fishing may change the exponent of a species through changing their life history traits, such as maximum length and maturation age. Thus, the exploited species exhibited a higher variance in spatial distribution than an unexploited species with the similar abundance and recorded life traits. We also develop an individual-based model to investigate the generating processes of Taylor’s Power Law, and whether age-truncated species have higher Taylor’s exponents. Our model shows that Taylor’s Power Law can be produced through demographic processes. Species with higher directional moving ability behave more aggregated. When reproduction rate increases, our model shows that species have higher average spatial variance but reflecting in Taylor’s intercept instead of exponent. Furthermore, the average spatial variance-mean ratio of exploited species is also reproduced in our model, reflecting the increasing aggregation. Our model suggests species with r-selective traits and suffering age-truncation will have higher spatial variance, but the different mechanisms of change in a and b still needed further studies.