Contrasting effects of frugivore assemblages and phylogeny on global variation in fruit length and width

<strong>Aim:</strong> The evolution of fruit size is still poor understood, though the long-standing theories of the dispersal syndrome hypothesis and the correlated evolution hypothesis have highlighted the influences of frugivores and phylogeny on fruit size variations. <strong>Location:</strong> In total, 326 interaction networks were extracted from studies carried out in Asia, Europe, Africa, Hawaii, Australia, New Zealand, and the Neotropics. <strong>Methods:</strong> We compile a global scale dataset on fruit size (i.e., width and length) and plant traits (i.e., leaf area and plant height) of 349 fleshy-fruited plants and the body mass of frugivores from 326 interaction networks. Plant phylogeny was constructed from a megatree containing 74533 vascular plant species. Pagel’s l was calculated to assess the phylogenetic signal in plant traits and fruit size. Phylogenetic generalized linear mixed model (PGLMM) was used to reveal relationships between fruit size and plant traits as well as the body mass of interacting frugivores. We used phylogenetic partial R2s regression model to tease apart the effects of body mass of frugivores, leaf area, plant height and plant phylogeny on the evolution of fruit size. <strong>Results:</strong> We first show that the global variation in fruit width is more explained by frugivorous assemblages than by phylogeny and plant functional traits (leaf area and plant height). Despite the significant evolutionary correlation between fruit length and width, fruit length appears to be more phylogenetically constrained and less sensitive to frugivore-driven selection than fruit width. <strong>Main conclusion:</strong> We found a global phylogenetic constraints on fruit length and frugivore-driven selection on fruit width, providing a novel mechanistic explanation for better understanding the evolution of fruit size across fleshy fruited plant species. <strong>KEYWORDS</strong> Body mass, frugivore, fruit length, fruit size, fruit width, phylogenetic signal

**研究目的**：果实大小的演化机制迄今仍未得到充分阐释，尽管长期存在的传播综合征假说（dispersal syndrome hypothesis）与协同演化假说（correlated evolution hypothesis）已强调了食果动物（frugivore）与系统发育（phylogeny）对果实大小变异的影响。
**研究区域**：本研究从亚洲、欧洲、非洲、夏威夷、澳大利亚、新西兰以及新热带区（Neotropics）的相关研究中，共提取得到326个互作网络。
**研究方法**：我们整合了一套全球尺度的数据集，涵盖326个互作网络中349种肉质果植物的果实大小（即果宽与果长）、植物性状（即叶面积与株高），以及对应食果动物的体质量。植物系统发育树基于包含74533种维管植物的巨型系统发育树构建。我们计算了Pagel’s l以评估植物性状与果实大小的系统发育信号（phylogenetic signal）。采用系统发育广义线性混合模型（phylogenetic generalized linear mixed model, PGLMM）分析果实大小与植物性状、互作食果动物体质量之间的关联。此外，我们通过系统发育偏决定系数回归模型（phylogenetic partial R²s regression model），厘清食果动物体质量、叶面积、株高以及植物系统发育对果实大小演化的相对影响。
**研究结果**：我们首先发现，相较于植物系统发育与植物功能性状（叶面积与株高），食果动物类群能够更好地解释全球范围内的果宽变异。尽管果长与果宽之间存在显著的演化相关性，但果长似乎受到更强的系统发育约束，且对食果动物驱动的选择压力的敏感性低于果宽。
**主要结论**：我们发现全球尺度下果长受系统发育约束、果宽受食果动物驱动的选择作用，这为阐释肉质果植物类群的果实大小演化提供了全新的机制性解释。
**关键词**：体质量、食果动物、果长、果实大小、果宽、系统发育信号