Busana Michela, Weissing Franz J, Hammers Martijn, Bakker Joke, Dugdale Hannah L, Raj Pant Sara, Richardson David S, Burke Terrence A, Komdeur Jan
Groningen Institute for Evolutionary Life Sciences, University of Groningen, 9700 CC Groningen, The Netherlands.
School of Biological Sciences, University of East-Anglia, Norwich NR4 7TJ, UK.
Behav Ecol. 2021 Dec 24;33(2):352-363. doi: 10.1093/beheco/arab135. eCollection 2022 Mar-Apr.
Even in well-studied organisms, it is often challenging to uncover the social and environmental determinants of fitness. Typically, fitness is determined by a variety of factors that act in concert, thus forming complex networks of causal relationships. Moreover, even strong correlations between social and environmental conditions and fitness components may not be indicative of direct causal links, as the measured variables may be driven by unmeasured (or unmeasurable) causal factors. Standard statistical approaches, like multiple regression analyses, are not suited for disentangling such complex causal relationships. Here, we apply structural equation modeling (SEM), a technique that is specifically designed to reveal causal relationships between variables, and which also allows to include hypothetical causal factors. Therefore, SEM seems ideally suited for comparing alternative hypotheses on how fitness differences arise from differences in social and environmental factors. We apply SEM to a rich data set collected in a long-term study on the Seychelles warbler (), a bird species with facultatively cooperative breeding and a high rate of extra-group paternity. Our analysis reveals that the presence of helpers has a positive effect on the reproductive output of both female and male breeders. In contrast, per capita food availability does not affect reproductive output. Our analysis does not confirm earlier suggestions on other species that the presence of helpers has a negative effect on the reproductive output of male breeders. As such, both female and male breeders should tolerate helpers in their territories, irrespective of food availability.
即使在研究充分的生物体中，要揭示适应性的社会和环境决定因素往往也具有挑战性。通常，适应性由多种协同作用的因素决定，从而形成复杂的因果关系网络。此外，即使社会和环境条件与适应性组成部分之间存在很强的相关性，也可能并不意味着存在直接的因果联系，因为所测量的变量可能受到未测量（或无法测量）的因果因素驱动。标准的统计方法，如多元回归分析，并不适合理清这种复杂的因果关系。在这里，我们应用结构方程模型（SEM），这是一种专门设计用于揭示变量之间因果关系的技术，它还允许纳入假设的因果因素。因此，SEM似乎非常适合比较关于适应性差异如何由社会和环境因素差异产生的替代假设。我们将SEM应用于在一项关于塞舌尔莺的长期研究中收集的丰富数据集，塞舌尔莺是一种具有兼性合作繁殖和高群外父权率的鸟类。我们的分析表明，帮手的存在对雌性和雄性繁殖者的繁殖产出都有积极影响。相比之下，人均食物供应量并不影响繁殖产出。我们的分析没有证实之前关于其他物种的观点，即帮手的存在会对雄性繁殖者的繁殖产出产生负面影响。因此，无论食物供应情况如何，雌性和雄性繁殖者都应该容忍领地内的帮手。
