The evolutionary ecology of cognitive ability in the wild

个体认知能力差异的原因以及认知能力的适应意义尚不清楚。造成这种情况的一个主要原因是，很少有人试图了解自然条件下自然选择如何作用于认知变异，或者估计这种变异的遗传性。我们的工作假设是，由于认知功能有成本和收益，由于异质环境条件产生的变量选择，变异得以维持。我们建议在一个多面手雀形目动物的长期研究群体中检验这一假设。大山雀是一年四季的居民，它们使用的食物种类繁多，分布不均，这表明学习在决定觅食效率、最终生存和繁殖成功方面发挥着重要作用。我们的预期是，学习能力在环境条件差的时候被选择，而在环境条件好的时候被选择，因为更高的学习能力是以牺牲其他功能特征为代价的，比如竞争能力。因此，这种变量选择将导致种群的适应性变异。该领域的主要挑战是在确定其适合度的同时测量已知谱系的大量个体的学习能力。我们建议使用一种新颖的系统来自动监测已知身份的个体如何学习将特定颜色的光与食物奖励联系起来。该系统已经在实验室开发，我们目前正在对其进行修改，以便在野外工作。该装置将在整个非繁殖季节放置在研究种群中，不仅能够使用PIT标签（应答器）技术识别每只单独标记的鸟，而且还能记住个体在上次访问时的学习过程。每次访问时，设备还将记录他们的体重，从而使我们能够统计控制身体状况对学习能力的影响（使用翅膀和跗骨长度控制身体大小）。下一阶段将是测量单个鸟类的繁殖成功率，并确定所占栖息地的质量。每年，我们都会定期收集整个种群的这些数据，这样我们就可以估计自然选择对学习能力的影响，在冬季测量，控制栖息地质量（当地种群密度和橡树密度）被个体占据。我们还将直接测试学习能力与竞争能力相权衡的观点，通过询问“细分的领土大小”和该领土内空置的巢箱数量是否与学习能力呈负相关，这在雄性中最有可能被发现。最后，我们将探讨环境和遗传影响在解释学习能力个体差异方面的相对重要性，使用我们研究人群的谱系。所有鸟类的亲缘关系都可以使用现有的威瑟姆种群谱系来估计，这将使我们能够区分环境（例如天气，出生环境条件）和遗传变异来源，使我们能够对这一特征的遗传性进行估计。我们的研究将产生两篇关于认知特征的进化生态学的重要而新颖的论文，以及一篇关于所使用的自动化系统的额外技术论文，并将为在自然条件下对这种变化的近因和选择性后果进行更长，更详细的研究提供概念证明。这些结果不仅会引起进化生物学家和行为生态学家的广泛兴趣，也会引起自然资源保护主义者和生态学家的兴趣，因为它们将有助于理解认知能力如何帮助个体，从而帮助种群适应环境变化。

Why individuals vary in their cognitive ability and the adaptive significance of cognitive ability are poorly understood. One of the main reasons for this is that few attempts have been made to understand how natural selection acts on cognitive variation under natural conditions, or to estimate the heritability of this variation. Our working hypothesis is that, because cognitive functions have costs and benefits, variation is maintained as a result of variable selection which arises because of heterogeneous environmental conditions. We propose to test this hypothesis in a long term study population of a generalist passerine. Great tits are year-round residents that use a wide range of patchily distributed food types, suggesting a role for learning in determining foraging efficiency, and ultimately survival and reproductive success. Our expectation is that learning ability is selected for when environmental conditions are poor, but against when conditions are good because higher learning ability comes at a cost to other functional traits, for example competitive ability. This variable selection would therefore lead to adaptive variation in the population. The major challenge in the field is to measure learning ability in large numbers of individuals of known genealogy at the same time as determining their fitness. We propose to use a novel system to automatically monitor how well individuals of known identity learn to associate a specific coloured light with a food reward. Already developed in the laboratory, we are currently modifying the system to work in the wild. The devices will be placed in the study population throughout the non-breeding season and will not only be able to identify each individually-marked bird using PIT tag (transponder) technology, but will also remember where individuals were in the learning process on their previous visit. Devices will also record their body mass at each visit, thereby allowing us to control statistically for the effects of body condition (controlling for body size using wing and tarsus length) on learning ability. The next stage will be to measure the reproductive success of individual birds and to identify the quality of the habitat occupied. Every year these data are routinely collected for the entire population allowing us to estimate natural selection on learning ability, as measured during the winter, controlling for the habitat quality (local population density and oak tree density) occupied by the individual. We will also directly test the idea that learning ability is traded off against competitive ability by asking whether 'tesselated territory size', and the number of unoccupied nestboxes within that territory, are negatively correlated with learning ability, which is most likely to be detected in males. Finally we will explore the relative importance of environmental and genetic effects in explaining individual variation in learning ability using the pedigree available for our study population. The relatedness of all birds can be estimated using the existing Wytham population pedigree, which will allow us to separate environmental (e.g. weather, natal environmental conditions) and genetic sources of variation, enabling us to generate an estimate of heritability for this trait. Our study will generate two significant and novel papers on the evolutionary ecology of a cognitive trait, and an additional technological paper on the automated system used, and will provide proof of concept for a longer, more detailed study on the proximate causes and selective consequences of this variation under natural conditions. These outcomes will not only be of wide interest to evolutionary biologists and behavioural ecologists, they will also be of interest to conservationists and ecologists because they will help to understand how cognitive ability helps individuals, and hence populations, adapt to environmental variation.