Phenotypic covariance structure and its divergence for acoustic mate attraction signals among four cricket species
Phenotypic covariance structure and its divergence for acoustic mate attraction signals among four cricket species
复制标题
DOI：
10.1002/ece3.76
复制
发表时间：
2012-01-01
影响因子：
2.6
通讯作者：
Gorelick, Root
中科院分区：
文献类型：
--
作者：
Bertram, Susan M.;Fitzsimmons, Lauren P.;Gorelick, Root
表型方差-协方差矩阵(P)描述了种群在表型空间中的多变量分布，提供了在多重共线性背景下测量性状的适当性的直接见解（即，它们是否描述了任何独立于其他性状的显著方差），以及性状协方差是否限制了可用于选择的表型组合。考虑到P的重要性，表型协方差很少被联合分析，P的维度很少在严格的统计框架中被研究，这是令人惊讶的。在这里，我们使用重复测量方法分别量化了四种蟋蟀种群的P，使用了七种被认为可以增强配偶吸引力的声学信号特征。在所有四个物种中，P都是全维或几乎全维的，这表明所有性状都传递了一些独立于其他性状的信息，并且表型性状协方差并不限制可用于选择的信号性状的组合。P在种间也存在显著差异，尽管表型变异的优势轴（P (max)）在3个种（Acheta domesticus、Gryllus assimilis和G. texensis）之间基本相同，但在4个种（G. veletis）之间存在差异。在棉铃虫和家蝇中，p（max）与体型相关，而在棉铃虫和棉铃虫中，p（max）与剩余质量（一种条件测量）无关。本研究揭示了联合分析表型性状的重要性。
The phenotypic variance-covariance matrix (P) describes the multivariate distribution of a population in phenotypic space, providing direct insight into the appropriateness of measured traits within the context of multicollinearity (i.e., do they describe any significant variance that is independent of other traits), and whether trait covariances restrict the combinations of phenotypes available to selection. Given the importance of P, it is therefore surprising that phenotypic covariances are seldom jointly analyzed and that the dimensionality of P has rarely been investigated in a rigorous statistical framework. Here, we used a repeated measures approach to quantify P separately for populations of four cricket species using seven acoustic signaling traits thought to enhance mate attraction. P was of full or almost full dimensionality in all four species, indicating that all traits conveyed some information that was independent of the other traits, and that phenotypic trait covariances do not constrain the combinations of signaling traits available to selection. P also differed significantly among species, although the dominant axis of phenotypic variation (p(max)) was largely shared among three of the species (Acheta domesticus, Gryllus assimilis, G. texensis), but different in the fourth (G. veletis). In G. veletis and A. domesticus, but not G. assimilis and G. texensis, p(max) was correlated with body size, while p(max) was not correlated with residual mass (a condition measure) in any of the species. This study reveals the importance of jointly analyzing phenotypic traits.