An assessment of phylogenetic tools for analyzing the interplay between interspecific interactions and phenotypic evolution

Much ecological and evolutionary theory predicts that interspecific interactions often drive phenotypic diversification and that species phenotypes in turn influence species interactions. Several phylogenetic comparative methods have been developed to assess the importance of such processes in nature; however, the statistical properties of these methods have gone largely untested. Focusing mainly on scenarios of competition between closely-related species, we assess the performance of available comparative approaches for analyzing the interplay between interspecific interactions and species phenotypes. We find that many currently used statistical methods often fail to detect the impact of interspecific interactions on trait evolution, that sister-taxa analyses are particularly unreliable in general, and that recently developed process-based models have more satisfactory statistical properties. Methods for detecting predictors of species interactions are generally more reliable than methods for detecting character displacement. In weighing the strengths and weaknesses of different approaches, we hope to provide a clear guide for empiricists testing hypotheses about the reciprocal effect of interspecific interactions and species phenotypes and to inspire further development of process-based models.

诸多生态学与进化理论均预言：种间相互作用往往驱动表型多样化，而物种表型反过来亦会影响种间相互作用。目前已开发出多种系统发育比较方法（phylogenetic comparative methods），用以评估此类过程在自然系统中的重要性；然而这些方法的统计特性在很大程度上尚未得到充分检验。本研究主要聚焦近缘物种间的竞争场景，评估现有比较方法在分析种间相互作用与物种表型间互作关系时的表现。研究发现，诸多当前常用的统计方法往往无法检出种间相互作用对性状进化的影响；其中姐妹类群（sister-taxa）分析整体上尤为不可靠，而新近开发的基于过程的模型（process-based models）则具备更优异的统计特性。用于检测种间相互作用预测因子的方法，整体上比检测性状替换（character displacement）的方法更为可靠。本研究通过权衡不同方法的优劣，旨在为检验种间相互作用与物种表型间双向互作假说的实证研究者提供清晰的指南，并推动基于过程的模型的进一步开发。