四种鸣禽个体间曲目的差异与领域维持时间的Meta分析

Meta-analysis of the Relationship between Repertoire Sizes and Territory Tenure in Four Songbird Species

鸟类鸣唱曲目的大小被认为是雄鸟质量的指标,其与领域维持时间的关系在多个鸟种中进行过检验。从单项研究的结果看,曲目与领域维持时间的关系在种间甚至同一物种的种群间均有较大的差异。为了克服单项研究功效低的缺陷,本文采用Meta分析,检验鸣禽曲目大小与领域维持时间的关系。共收集到10篇文献,涉及大山雀（Parus major）、大苇莺（Acrocephalus arundinaceus）、歌带鹀（Melospiza melodia）和强脚树莺（Horornis fortipes）共12个种群的研究。Meta分析发现,曲目大的个体领域维持时间显著长于曲目小的个体（Cohen′s d效应量为0.60,95%置信区间为0.38-0.82）。通过Meta回归,发现曲目大小与领域维持时间的关联不随繁殖地纬度而变化（斜率为0.00,95%置信区间为﹣0.02-0.02）,但在平均曲目大的种群中曲目大小与领域维持时间的关联较弱（斜率为﹣0.14,95%置信区间为﹣0.30-0.03）。本研究可为了解鸟类曲目的功能及其演化提供资料。

Song repertoire size is often cited as an honest index of male quality. The relationship between repertoire size and survival, for which often using territory tenure as a proxy, has been tested in some species. However, this relationship is still unclear. Some studies have demonstrated that males with large repertoires keep territory longer, whereas other studies have failed to support this prediction（Table 1）. To make general inferences from this mixed evidence, we quantitatively reviewed the relevant literature based on a meta-analytic approach. We performed key word searches using the terms ＂song repertoire＂, ＂song complexity＂, ＂song versatility＂ in combination with the terms ＂longevity＂, ＂survival＂, ＂life span＂, ＂viability＂, ＂territory tenure＂ from Web of Science（http：//apps.webofknowledge.com [2015-4-15]）. We also conducted cross-reference searches by consulting previous reviews on sexually selected traits. Finally, we collected the data for 4 species from 12 populations, including great tit（Parus major）, great reed warbler（Acrocephalus arundinaceus）, song sparrow（Melospiza melodia）, brownish-flanked bush warbler（Horornis fortipes）. We calculated Fisher′s z or log odds ratio effect size from each primitive paper, and translated into Cohen′s d effect size for Meta-analysis. Based on Random-effects Meta-analysis, mean effect size（Cohen′s d） for the association between song repertoires and territory tenure was 0.60（95% confidence interval range from 0.38 to 0.82）, which was statistically significant（random-effects meta-analysis model, n = 12, P〈0.001）（Fig. 1）. Based on Meta-regression analysis, the degree of association between song repertoires and territory tenure was independent of latitude of study area（Meta-Regression, estimate = 0.00, 95% CI from﹣0.02 to 0.02, P = 0.953）（Fig. 2）, but generally weak in populations with large song repertoires（Meta-Regression, estimate = ﹣0.14, 95% CI from﹣0.30 to 0.03, P = 0.110）（Fig. 3）. Our results provide that repertoire size can serve as a potential signal of territory tenure capability.