Realised hypothetical phenotypes and the adaptive value of Batesian mimicry

许多动物，如毒蛇或刺痛的黄蜂，向潜在的捕食者宣传，它们通过显示明亮、大胆的颜色图案而得到很好的防御。其他被称为“模仿者”的动物已经进化到可以通过显示与危险的“模型”相似的图案来欺骗捕食者，尽管它们是无害的。研究拟态的进化有可能告诉我们很多关于自然选择如何导致野生动物适应的信息。在拟态中，我们看到了许多突出而美丽的进化可以产生极端适应的例子。然而，一些模仿物，包括许多模仿黄蜂的气垫蝇，只是与它们的模型大致相似。我们预计自然选择会倾向于最令人信服的模仿，因为不太好的模仿应该被捕食者识别和剔除：那么为什么一些模仿是不准确的呢？也许捕食者(如鸟类)不会注意猎物外表的所有方面，所以模仿并不一定要在所有方面都准确。鸟类可能会避开所有与特别令人讨厌的模型大致相似的猎物，因此超过某一点，就没有必要进一步准确了。或者，最好的生存机会可以通过模仿几个令人讨厌的模式物种来获得，而不是那些完美模仿某个特定模式的物种。为了测试这些想法，我们不仅需要知道现有的模仿有多成功，还需要知道具有不同功能组合的替代模仿的性能。更准确的模仿会更好地存活下来吗？如果它是完美的，除了它的颜色呢？这是进化生物学中一个更普遍的问题的一个例子，即预测与我们在自然界中观察到的生物体相比，不存在的生物体可能如何表现，但在理论上是可信的。为了探索模拟物的进化潜力，我们的项目将最近开发的技术结合在一起，以一种生物学家以前从未开发过的方式。首先，我们将从多个角度拍摄苍蝇和黄蜂，并组装3D数字表示。然后，我们将创建新的“变形”图像，这些图像可能是苍蝇和黄蜂之间的中间产物，但却是虚构的，它们与黄蜂的相似性各不相同。第三，我们将使用3D打印来创建这些中间体(以及“真实”昆虫)的模型，这些模型在大小、颜色、图案和形状方面都是逼真的。最后，我们将在各种实验中测量这些3D模型的“存活率”。我们将通过训练鸟类“攻击”飞蝇模型来测试捕食者的反应，但避开黄蜂模型，以获得食物奖励。在了解了这一区别后，这些鸟将接受测试，看看它们对中等外观的3D模型的表现如何。鸟类在攻击前犹豫的时间越长，模仿就越有效，因为在野外，这种延迟会给昆虫一个逃跑的机会。我们将在野外和实验室观察捕食者的行为，在野外，具有不同经验的鸟类自然群落正在觅食，在实验室，我们可以控制每个个体的经历，观察它们是如何学习的。我们还将用螳螂代替鸟类来重复实验，以调查昆虫捕食者对模仿的判断是否有所不同。我们的项目解决了一个长期存在的关于不准确模仿进化的生物学问题，但也将为进化过程提供更广泛的见解。当其他物种要么没有生存下来，要么根本就没有进化的时候，现有物种意味着它们存活了下来，这意味着什么？这反过来又让我们了解进化如何产生现有物种的特征组合，并预测未来进化可能采取的方向。我们的新方法将为其他进化生物学家提供一个框架，在广泛的不同生物体中探索类似的问题。

Many animals, such as venomous snakes or stinging wasps, advertise to potential predators that they are well defended by displaying bright, bold colour patterns. Other animals, known as "mimics", have evolved to deceive predators by displaying patterns similar to the dangerous "models", despite being harmless. Studying the evolution of mimicry has the potential to tell us much about how natural selection leads to adaptation in wild animals.Among mimics, we see many prominent and beautiful examples of the extreme adaptations that evolution can produce. However, some mimics, including many hoverflies that mimic wasps, only approximately resemble their model. We expect natural selection to favour the most convincing mimics, because less good ones ought to be recognised and picked off by predators: so why are some mimics inaccurate? Perhaps predators (such as birds) do not pay attention to all aspects of the prey's appearance, so the mimic does not have to be accurate in every way. Birds might avoid all prey that are roughly similar to a particularly nasty model, so beyond a certain point, further accuracy is unnecessary. Alternatively, the best chances of survival may gained by mimics which approximately resemble several nasty model species, rather than those which are perfect mimics of one particular model. To test these ideas, we need to know not only how successful existing mimics are, but also the performance of alternative mimics with different combinations of features to the existing ones. Would a more accurate mimic survive better? What if it were perfect except for its colour? This is an example of a more general problem in evolutionary biology, of predicting how non-existent, but theoretically plausible, organisms might perform in comparison to those that we observe in nature. To explore the evolutionary potential of mimics, our project brings together recently-developed technologies in a way never exploited before by biologists. First, we will photograph hoverflies and wasps from multiple angles and assemble 3D digital representations. We will then create new "morphed" images of plausible, but imaginary, intermediates between hoverflies and wasps which vary in similarity to the wasps. Thirdly, we will use 3D printing to create models of these intermediates (as well as the "real" insects) that are realistic in terms of their size, colour, pattern and shape. Lastly, we will measure the "survival" of these 3D models in a variety of experiments.We will test the response of predators by training birds to "attack" models of hoverflies, but avoid those of wasps, in order to gain food rewards. Having learnt that distinction, the birds will then be tested for how they behave towards 3D models of intermediate appearance. The longer the birds hesitate before attacking, the more effective is the mimicry, because in the wild such delays would give the insect an opportunity to escape. We will look at predator behaviour both in the wild, where natural communities of birds with varying levels of experience are foraging for food, and in the laboratory, where we can control the experiences of each individual and watch how they learn. We will also repeat the experiments using praying mantises in place of birds, to investigate whether insect predators judge mimicry differently. Our project addresses a long-standing biological question about the evolution of inaccurate mimicry, but will also provide broader insights into evolutionary processes. What is it about existing species that mean they survived when others either did not survive, or never evolved in the first place? This in turn allows us to understand how evolution produced the combinations of characteristics of existing species, and to predict the direction that evolution might take in the future. Our novel method will provide a framework for other evolutionary biologists to explore similar questions in a wide range of different organisms.