Cuckoo host interactions: co-evolution and learning

在自然界中，物种之间的冲突经常涉及进化军备竞赛，例如。捕食者/寄生虫攻击策略的改进导致猎物/宿主防御的改进，从而导致捕食者/寄生虫的进一步改进，等等。杜鹃与寄主之间的相互作用为研究这一过程提供了一个很好的课题。杜鹃是一种著名的骗子：它依靠其他鸟类(宿主)来抚养它的幼崽。它每个寄主巢只产一个蛋。如果主人接受了这一点，11天后杜鹃幼鸟首先孵化出来。仅仅几个小时大的时候，它就会排出宿主的卵，一个接一个地让它们在背上保持平衡，然后把它们抛到巢的边缘。不足为奇的是，东道主已经进化出防御系统。一种是排出与它们自己不同的卵子。作为回应，杜鹃进化成了遗传上截然不同的种族，每个种族都专注于一个宿主物种，并像宿主蛋一样产卵，而宿主不太可能注意到这一点。另一种是攻击成年杜鹃，如果它们在巢中看到杜鹃(因此更有可能是寄生虫)，就会增加拒绝卵子的可能性。作为回应，杜鹃进化出了快速产卵(约10秒)，以减少它向宿主发出警报的机会。虽然军备竞赛导致了杜鹃和宿主的基因变化，但个体也可能通过学习来提高自己的成功。其目的是研究双方的学习情况。(A)主人的学习：主人很容易接近和攻击成年杜鹃，与之形成鲜明对比的是，它们对危险的敌人，如麻雀，保持警惕。主人如何认识到杜鹃是巢中的敌人呢？之前的研究表明，从杜鹃身上分离出来的寄主种群仍然拒绝与它们自己的种群不同的卵，但对布谷鸟栖息在巢中的反应要弱得多，这不会像寄生种群那样刺激更多的卵子排斥。这些结果表明，未被寄生的宿主种群可能只是没有认识到成年杜鹃是敌人，因为它们没有机会了解这一点。我们将做实验，用各种敌人的坐骑呈现给有颜色环的、已知年龄的芦苇鸣禽(最受欢迎的杜鹃宿主)，以测试下列问题：潜在筑巢敌人的哪些特征会引发围攻；围攻的叫声是否会吸引邻居；这种经历是否会导致幼稚的鸟类在自己的领土上遇到相同的敌人时，增加围捕和拒绝产卵；宿主是否倾向于对杜鹃做出攻击性反应，或者是否可以欺骗它们了解到，一种新颖但无害的鸟类是巢中的敌人；社交学习是否可以沿着宿主领土的链条传递敌人的识别。(B)通过杜鹃学习：我们发现杜鹃不仅为其特定的寄主物种进化了一个“匹配良好的蛋”，而且还进化出了一个“良好的”雏鸟。由芦苇莺饲养的雏鸟的乞讨叫声与由矮种鸟饲养的杜鹃不同。实验表明，当幼小的杜鹃在不同的寄主巢之间交换时，它们会适当地改变它们的语调，所以它们必须通过经验来学习如何最好地乞讨。我们将研究从孵化开始的乞讨叫声，以测试杜鹃是否会尝试不同的叫声，并通过反复试验学习使用哪种叫声。我们还将进行实验，向不同的宿主物种广播各种杜鹃乞讨的叫声，以测试杜鹃选择的叫声是否最能刺激它们的特定宿主喂食它们。这项关于宿主和杜鹃如何通过学习微调它们的反应的研究，将帮助我们了解敌人适应环境变化的速度有多快。未来，我们希望将学习的影响融入到杜鹃-宿主共同进化的过程中。

In Nature, conflicts between species often involve evolutionary arms races, eg. improvements in strategies of attack by predators/parasites lead to improvements in defence by prey/hosts, leading to further predator/parasite improvements, and so on. Interactions between the cuckoo Cuculus canorus and its hosts provide a good subject for studying this process. The cuckoo is a famous cheat: it relies on other bird species (hosts) to raise its young. It lays just one egg per host nest. If the hosts accept this, eleven days later the cuckoo chick hatches first. Just a few hours old, it then ejects the host's eggs, balancing them on its back, one by one, and heaving them over the nest rim. Not surprisingly, hosts have evolved defences. One is to eject eggs unlike their own. In response, the cuckoo has evolved into genetically-distinct races, each specialising on one host species and laying eggs just like its host eggs, which the hosts are less likely to notice. Another is to attack adult cuckoos and increase the likelihood of rejecting eggs if they see a cuckoo on their nest (and so are more certain of parasitism). In response, the cuckoo has evolved rapid laying (c. 10 seconds) to decrease the chance it alerts the host. Although the arms race has led to genetic changes in cuckoos and hosts, individuals are also likely to improve their success through learning. The aim is to examine a case of learning on each side. (a) Learning by hosts: Hosts readily approach and attack adult cuckoos, in marked contrast to their wariness of dangerous enemies, such as sparrowhawks. How do hosts come to recognise the cuckoo as a nest enemy? Previous studies show that host populations isolated from cuckoos still reject eggs unlike their own, but have a much weaker response to cuckoo mounts on their nest and this does not stimulate increased egg rejection as it does in parasitised populations. These results suggest that unparasitised host populations may simply not recognise the adult cuckoo as an enemy because they haven't had the chance to learn this. We will do experiments with presentations of mounts of various enemies to colour-ringed, known-aged, reed warblers (a favourite cuckoo host) to test the following: what features of a potential nest enemy induce mobbing; whether mobbing calls attract neighbours; whether this experience leads naive birds to increase their mobbing and egg rejection when they encounter the same enemy on their own territory; whether hosts are predisposed to react aggressively to cuckoos, or whether they can be tricked into learning that a novel, but harmless, bird is a nest enemy; whether social learning can transmit enemy recognition along a chain of host territories. (b) Learning by cuckoos: We have discovered that not only have cuckoos evolved a 'well-matched egg' for their particular host species, but also a 'well-tuned' chick. Nestling cuckoos raised by reed warblers have different begging calls from those raised by dunnocks. Experiments show that young cuckoos change their tune appropriately when swapped between host nests, so they must somehow learn, through experience, how best to beg. We will study the begging calls from hatching to test whether cuckoos try out different calls and learn, by trial and error, which to use. We will also do experiments in which we broadcast various cuckoo begging calls to different host species to test whether the calls chosen by the cuckoos are the ones which are most effective in stimulating their particular hosts to feed them. This study of how both hosts and cuckoos fine-tune their responses by learning will help us to understand how rapidly enemies can adapt to environmental changes. In future, we hope to incorporate the effects of learning on the course of cuckoo-host co-evolution.