Collaborative Proposal: Linking process to pattern through an experimental network approach to identify the behavioral mechanisms of reproductive isolation

动物经常以非随机的方式与它们自己的成员和密切相关的物种相互作用，这对种群内部和种群之间的进化有着深远的影响，包括遗传差异的积累或破坏。密切相关的亚种经常接触并杂交，或杂交。本研究将通过实验研究社会互动（网络）、形态特征、基因组祖先和迁徙行为如何解释两个接触区（每个接触区由两对杂交程度不同的小燕子亚种组成）的杂交数量。将野外观察和实验与基因组分析相结合，可以直接测试个体行为如何影响杂交。该研究使用小型接近记录仪和网络分析方法来测量单个燕子之间相互作用的频率。数据收集将通过广泛的学生培训和国际合作来完成。社会网络的使用、可视化和解释将通过以下方式广泛传播：1)会议报告、研讨会和出版物；2)整合到课程中；3)公开发布新的网络分析方法代码；4)使用练习和评估工具对研究数据进行在线动态交互式可视化。这项工作代表了一种高度新颖和综合的方法来理解密切相关的种群如何演变成不同的物种，对生物多样性和保护研究具有重要意义。分类行为（非随机的社会互动）被认为是最近分化的动物类群之间基因流动的关键障碍。然而，直接测量分类行为在自然界是很少可能的。本研究利用小燕子（Hirundo rustica）亚种复合体，实验研究了在进化历史上不同的亚种对之间的两个杂交区，分类行为对生殖隔离的贡献。以下将从每个混合区收集：1)通过近距离社会互动构建社会网络；2)精细尺度基因组信息测量个体祖先；3)越冬地点、迁徙路线和到达时间数据来自地理定位器和稳定同位素；4)形态变化测量；5)分子父权分析数据构建基于内外配对的受精网络，作为亚种间杂交的直接测量。受精网络的构建将有助于分析杂交在多大程度上是由社会互动、亚特异性祖先、迁徙行为或形态介导的。这些特征可能彼此相关，它们在生殖隔离中的综合和相对重要性将使用矩阵回归方法进行分析。在从每个杂交区收集相关数据后，将进行表型和育种时间的实验操作，以测试这些性状与受精之间的因果关系，为实验测试和操作杂交预测因子提供极其难得的机会。这些比较提供了以前难以处理的机会，可以通过实验分析个体水平的变异如何预测繁殖决策以及种群杂交的程度。该奖项由环境生物学综合有机体系统部门的行为系统和环境生物学部门的进化过程共同资助。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Animals frequently interact with members of their own and closely related species in non-random ways that have profound implications for evolution within and between populations, including the buildup or breakdown of genetic differences. Closely related subspecies often come into contact and interbreed, or, hybridize. This research will experimentally study how social interactions (networks), morphological traits, genomic ancestry, and migratory behavior explain the amount of interbreeding in two contact zones each consisting of two pairs of barn swallow subspecies that differ in degree of hybridization. The integration of observation and experiments in the wild with genomic analyses will allow for a direct test of how individual behaviors impact hybridization. The research uses miniaturized proximity loggers and network analytical methods to measure the frequency of interactions between individual swallows. Data collection will be accomplished through extensive student training and international collaboration. The use, visualization, and interpretation of social networks will be broadly disseminated through 1) conference presentations, workshops, and publications, 2) integrated into courses, 3) public release of code for new network analysis methods and 4) online dynamic interactive visualizations of research data with exercises and assessment tools. This work represents a highly novel and integrative approach to understanding how closely related populations evolve into different species, with important implications for biodiversity and conservation studies.Assortative behaviors (non-random social interactions) are thought to represent critical barriers to gene flow among recently diverged animal taxa. However, direct measurement of assortative behavior in nature is rarely possible. This research leverages the barn swallow (Hirundo rustica) subspecies complex to experimentally examine the contribution of assortative behavior to reproductive isolation in two hybrid zones between subspecies pairs that differ in evolutionary history. The following will be collected from each hybrid zone: 1) close-range social interactions from proximity tags to construct social networks, 2) fine-scale genomic information to measure individual ancestry, 3) wintering locations, migratory routes and timing of arrival data from geolocators and stable isotopes, 4) measures of morphological variation, and 5) data for molecular paternity analyses to construct fertilization networks based on within- and extra-pair matings as a direct measure of hybridization between subspecies. The construction of fertilization networks will enable analyses about the degree to which hybridization is mediated by social interactions, subspecific ancestry, migratory behavior, or morphology. These features are likely correlated with one another and their combined and relative importance in reproductive isolation will be analyzed using matrix regression methods. After collecting correlative data from each hybrid zone, experimental manipulations of phenotype and timing of breeding will be conducted to test for causal relationships between these traits and fertilizations, providing an extremely rare opportunity to experimentally test and manipulate predictors of hybridization. These comparisons offer the previously intractable opportunity to experimentally analyze how individual-level variation predicts reproductive decisions and the degree to which populations hybridize.This award was co-funded by Behavioral Systems in the Division of Integrative Organismal Systems and Evolutionary Processes in the Division of Environmental Biology.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.