Data from: Translocation experiments show dialects are socially learned in marmoset monkeys

The acoustic properties of vocalizations in common marmosets differ between populations. These differences may be the result of social vocal learning, but they can also result from environmental or genetic differences between populations. We performed translocation experiments to separately quantify the influence of a change in the physical environment (experiment 1), and a change in the social environment (experiment 2) on the acoustic properties of calls from individual captive marmosets. If population differences were due to genetic differences, we expected no change in the vocalizations of the translocated marmosets. If differences were due to environmental factors, we expected vocalizations to permanently change contingent with environmental changes. If social learning was involved, we expected that the vocalizations of animals translocated to a new population with a different dialect would become more similar to the new population. In experiment 1, we translocated marmosets to a different physical environment without changing the social composition of the groups or their neighbours. Immediately after the translocation to the new facility, one out of three call types showed a significant change in call structure, but 5-6 weeks later, the calls were no longer different from before the translocation. Thus, the novel physical environment did not induce long lasting changes in the vocalizations of the marmosets. In experiment 2, we translocated marmosets to a new population with a different dialect. Importantly, our previous work had shown that these two populations differed significantly in vocalization structure. The translocated marmosets were still housed in their original social group, but after translocation they were surrounded by the vocalizations from neighbouring groups of the new population. The vocal distance between the translocated individuals and the new population decreased for two out of three call types over 16 weeks. Thus, even without direct social contact or interaction, the vocalizations of the translocated animals converged towards the new population, indicating that common marmosets can modify their calls due to acoustic input from conspecifics alone, via crowd vocal learning. To our knowledge, this is the first study able to distinguish between different explanations for vocal dialects as well as to show crowd vocal learning in a primate species.普通狨猴（common marmosets）的鸣叫声声学特征存在种群间差异。这类差异可能源于社会声学学习，也可能由种群间的环境或遗传差异所导致。我们开展了易位实验，以分别量化物理环境改变（实验1）与社会环境改变（实验2）对圈养个体狨猴鸣叫声声学特征的影响。若种群差异由遗传差异引起，我们预期易位狨猴的鸣叫声不会发生改变；若差异源于环境因素，预期鸣叫声会随环境变化发生永久性改变；若涉及社会学习，则预期被迁移至方言不同的新种群的个体，其鸣叫声会愈发接近新种群。在实验1中，我们将狨猴易位至不同物理环境，同时未改变群体的社会组成或周边社群环境。在易位至新设施后即刻，三类鸣叫声中有一类的结构出现显著变化，但在5至6周后，其鸣叫声与易位前已无显著差异。由此可见，新型物理环境并未对狨猴的鸣叫声产生长期持久的改变。在实验2中，我们将狨猴易位至拥有不同方言的新种群。值得注意的是，我们此前的研究已证实，这两个种群的鸣叫声结构存在显著差异。易位的狨猴仍保留原社会群体，但在易位后，其周边环绕的是新种群邻群的鸣叫声。在16周的观测周期内，三类鸣叫声中有两类的易位个体与新种群间的声学距离逐渐缩小。由此可见，即便无直接社会接触或互动，易位个体的鸣叫声仍会向新种群趋同，这表明普通狨猴仅通过同类个体的声学输入，即可通过群体声学学习调整自身的鸣叫声。据我们所知，本研究首次实现了对鸣唱方言不同成因的区分，同时首次在灵长类动物中证实了群体声学学习现象。