ReMCASA: Constraints on adaptation in social animals: kin recognition mechanisms and the fitness consequences of discrimination rules

进化适应的过程受到作用于表型性状的机械约束的限制。我们的目标是调查亲属识别机制的社会鸟类，以确定在何种程度上限制适应性决策，导致明显的适应不良行为。识别系统在从基因到复杂社会的生物组织的各个层面都无处不在，在这些不同的情况下，歧视往往会产生重大的适应性后果。因此，选择应设计有效的识别机制。然而，识别系统很少是无错误的，那么如何优化错误的频率以最大限度地提高包容性？我们将通过研究社会鸟类长尾山雀Aegithalos caudatus的亲属识别机制来解决这个进化难题，长尾山雀的显着亲属选择合作繁殖系统使它们非常适合调查亲属歧视及其健身后果。长尾山雀歧视近亲作为配偶，并在帮助时支持亲属，但在这两种情况下经常出现明显的适应不良错误。我们将采用一系列最先进的方法来量化和分析社会关系的指标和两个推定的表型线索亲属关系，声学和嗅觉，确定他们如何准确地编码相关信息。我们还将解决持续存在的进化问题，即当遗传识别线索受到正频率依赖性选择（克罗齐尔悖论）时，如何维持遗传识别线索的方差。最后，我们将确定表型线索和社会熟悉度在多大程度上被整合，以优化行为决策，进行实地实验，直接测试这种理解。其结果将是理解为什么会发生适应不良的社会行为，以及识别系统中的错误如何被优化以最大限度地提高包容性适应性的重大进展。

The process of evolutionary adaptation is limited by the mechanistic constraints acting on phenotypic traits. We aim to investigate kin recognition mechanisms in a social bird to determine the extent to which they constrain adaptive decision-making, resulting in apparently maladaptive behaviours. Recognition systems are ubiquitous at every level of biological organisation from genes to complex societies, and discrimination in these diverse situations often has major fitness consequences. Therefore, selection should engineer effective recognition mechanisms. However, recognition systems are rarely error-free, so how is the frequency of errors optimised in order to maximise inclusive fitness? We will address this evolutionary puzzle by studying kin recognition mechanisms in a social bird, the long-tailed tit Aegithalos caudatus, whose remarkable kin-selected cooperative breeding system makes them wonderfully well suited to investigation of kin discrimination and its fitness consequences. Long-tailed tits discriminate against close kin as mates and in favour of kin when helping, but make frequent, apparently maladaptive errors in both contexts. We will employ a range of state-of-the-art methods to quantify and analyse metrics of social affiliation and two putative phenotypic cues to kinship, acoustic and olfactory, determining how accurately they encode relatedness information. We will also address the persistent evolutionary problem of how variance in inherited recognition cues is maintained when they are subjected to positive frequency dependent selection (Crozier's paradox). Finally, we will determine the extent to which phenotypic cues and social familiarity are integrated to optimise behavioural decisions, conducting field experiments to directly test that understanding. The outcome will be a major advance in understanding of why maladaptive social behaviours occur, and how errors in a recognition system are optimised to maximise inclusive fitness.