The macroevolutionary consequences of trait correlations

地球上生命的多样性是巨大的。数百万年来，进化产生了种类繁多的物种。然而，当我们仔细观察时，我们开始注意到一些令人惊讶的模式。有些类群比其他类群拥有更多的物种，例如甲虫有超过35万种，而大象只有少数几种。其他类群表现出不同的形状和大小，例如夏威夷蜂蜜爬行动物，而其他类群则相当统一，例如小鼠和大鼠。这些模式表明，物种的数量及其形式和功能的多样性存在限制，即生物多样性在生命之树上的分布并不均匀。为什么会这样仍然没有解决，这是进化生物学中的一个基本问题。在这里，我们集中在一个以前未被探索的解释，这种多样性的不均匀分布：物种特征之间的相关性。物种特征几乎可以是外表（如体型或眼睛颜色）、行为或生活史（大小与后代数量）的任何方面。这些特征并不总是相互独立的，例如，投资更多的后代往往导致更小的后代。在这种情况下，我们说相关的特征是相关的。当性状相互关联时，一种性状的进化可能促进或阻止另一种性状的进化。不同物种群体间性状相关性强度的差异可能决定了亲缘物种群体之间的变异。当性状之间的相关性很强时，这可能会阻止非常不同形状和大小的进化。因此，形态和功能的多样性可能很低，近亲物种的数量应该很少。在数百万年的时间里，这些特征的相关性可能会变弱甚至消失，从而使物种以新的方式进化或在数量上繁殖。这些观点表明，性状相关性在我们理解生命多样性方面很重要。我们的建议旨在了解性状相关性如何影响整个生命之树中决定多样性模式的过程（包括物种数量和形式和功能）。我们关注的是鸟类与食物类型和获取相关的性状。鸟喙的大小和形状可以反映鸟类的饮食需求和觅食方式。当特征相关时，一些组合可能永远不会同时发生，例如，可能不可能进化出像杓鹬一样又长又弯，又像鸭子一样又宽又深的喙。为了测试性状相关性如何影响决定生命之树多样性模式的过程，我们需要测量来自许多不同物种的许多个体。要在现场做到这一点几乎是不可能的挑战。相反，我们使用博物馆的标本。世界各地的博物馆收藏了数十亿件标本，这些标本往往来自100多年前的历史收藏。伦敦自然历史博物馆有大约8000万个标本。这些收藏是几乎所有鸟类（超过10,000种）的丰富信息来源。我们将从这些收集中生成数据，包括鸟喙的3D扫描。然后，我们将使用这些数据，以及最先进的分析方法，来评估物种之间和物种内部以及更高的分类分组（如科和目）之间的性状相关性。通过使用博物馆标本和特征相关性重要性的新观点，我们的研究将为生命如何以及为什么多样化提供新的见解。这项研究不仅解决了进化生物学中的一个基本争论，而且还为未来的研究人员创造了有价值的鸟类数据集，并突出了自然历史收藏对前沿研究的重要性。

The diversity of Life on Earth is immense. Over millions of years, evolution has generated a staggering variety of species. When we look more closely, however, we begin to notice some surprising patterns. Some groups have far more species than others, for example there are over 350,000 species of beetle but only a few species of elephant. Other groups exhibit diverse shapes and sizes, for example the Hawaiian honeycreepers, whereas others are fairly uniform, for example mice and rats. These patterns suggest that there are constraints, or limitations, on the numbers of species and their variety of form and function, i.e. biological diversity is not evenly distributed across the tree of life. Why this is so remains unresolved and is a fundamental question in evolutionary biology.Here we focus on a previously unexplored explanation for this uneven distribution of diversity: correlations among species traits. Species traits can be almost any aspect of physical appearance (such as body size or eye colour), behaviour, or life history (size vs. number of offspring). These traits are not always independent of one another, for example investing in more offspring often results in smaller offspring. In such cases we say that the traits involved are correlated. Where traits are correlated, evolution in one trait may promote or prevent evolution in another. Differences in the strength of trait correlations in different groups of species might determine variation between and among groups of related species. Where correlations among traits are strong these may prevent evolution of very different shapes and sizes. Therefore the diversity of form and function might be low, and the number of closely-related species should be few. Over millions of years these trait correlations might become weaker or even disappear, allowing species to evolve in new ways or multiply in number. These ideas suggest that trait correlations are important in our understanding of the diversity of life. Our proposal aims to understand how trait correlations can affect the processes that determine patterns of diversity (both in species number and in form and function) across the tree of life. We focus on traits related to food type and acquisition in birds. The size and shape of bird beaks can reflect dietary requirements and foraging methods. When traits are correlated, some combinations may never occur together, for example, it may not be possible to evolve beaks that are simultaneously long and curved like a curlew, and wide and deep like a duck. To test how trait correlations can affect the processes that determine patterns of diversity across the tree of life we need to measure many individuals from many different species. It would be a near-impossible challenge to do this in the field. Instead we use specimens from museums. Around the world museums house billions of specimens, often from historical collections over 100 years old. The Natural History Museum, London has around 80 million specimens. These collections are a rich source of information on nearly all species of birds (over 10,000 species). We will generate data from these collections, including 3D scans of bird beaks. We will then use these data, along with state-of-the-art analysis methods, to evaluate trait correlations among and within species, and among higher taxonomic groupings such as families and orders. By using museum specimens and a novel view of the importance of trait correlations, our research will provide new insights into how and why life diversifies. Not only will this research address a fundamental debate in evolutionary biology, but it will also create valuable bird datasets for future researchers, and highlight the importance of natural history collections for cutting edge research.