[论文解读] Sparse movement data can reveal social influences on individual travel decisions

Tyler R. Bonnell, S. Peter Henzi+1 more|arXiv (Cornell University)|Nov 4, 2015

Primate Behavior and Ecology参考文献 20被引用 3

一句话总结

本研究表明，即使在低时间分辨率下采集的稀疏运动数据，也能有效揭示狒狒个体在旅行决策中受到的社会影响。通过对南非德胡普自然保护区一只狒狒群体的数据应用改进的力匹配方法，作者发现个体受特定群体成员位置的影响，远超过对整个群体的影响，表明尽管数据分辨率有限，有针对性的社会纽带仍是群体凝聚力的主要驱动力。

ABSTRACT

The monitoring of animal movement patterns provides insights into animals decision-making behaviour. It is generally assumed that high-resolution data are needed to extract meaningful behavioural patterns, which potentially limits the application of this approach. Obtaining high-resolution movement data continues to be an economic and technical challenge, particularly for animals that live in social groups. Here, we test whether accurate movement behaviour can be extracted from data that possesses increasingly lower temporal resolution. To do so, we use a modified version of force matching, in which simulated forces acting on a focal animal are compared to observed movement data. We show that useful information can be extracted from sparse data. We apply this approach to a sparse movement dataset collected on the adult members of a troop of baboons in the DeHoop Nature Reserve, South Africa. We use these data to test the hypothesis that individuals are sensitive to isolation from the group as a whole or, alternatively, whether they are sensitive to the location of specific individuals within the group. Using data from a focal animal, our data provide support for both hypothesis, with stronger support for the latter. Although the focal animal was found to be sensitive to the group, this occurred only on a small number of occasions when the group as a whole was highly clustered as a single entity away from the focal animal. We suggest that specific social interactions may thus drive overall group cohesion. Given that sparse movement data is informative about individual movement behaviour, we suggest that both high (~seconds) and relatively low (~minutes) resolution datasets are valuable for the study of how individuals react to and manipulate their local social and ecological environments.

研究动机与目标

• 探究低分辨率运动数据是否能可靠揭示个体对社会与生态因素的行为反应。
• 检验狒狒在移动决策中是响应整个群体还是群体内特定个体。
• 评估改进的力匹配方法从稀疏数据中提取行为模式的有效性。
• 确定群体层面互动与二元社会互动对个体移动决策的相对影响。

提出的方法

• 应用改进的力匹配技术，将模拟作用于焦点个体的力与实际移动轨迹进行比较。
• 该方法利用空间和时间数据，基于与其他个体或群体质心的接近程度，推断社会吸引力。
• 使用南非德胡普自然保护区一只焦点狒狒的群体数据，运动记录时间分辨率约为分钟级。
• 通过比较模型对实际移动模式的拟合程度，评估模型对整个群体与特定个体的敏感性。
• 通过统计验证评估模型在不同数据稀疏程度下对运动行为的解释能力。
• 在多个时间分辨率下测试该方法，以确定有意义行为模式仍可检测的临界阈值。

实验结果

研究问题

• RQ1 时间分辨率较低（如分钟级）的运动数据是否仍能揭示社会动物中个体有意义的行为反应？
• RQ2 狒狒在移动决策中主要响应整个群体的空间位置，还是群体内特定个体的位置？
• RQ3 随着数据分辨率降低，行为推断的准确性如何变化？
• RQ4 与整体群体移动相比，特定二元关系在维系群体凝聚力方面发挥何种作用？
• RQ5 力匹配方法能否有效从现实动物群体的稀疏运动数据中推断社会影响？

主要发现

• 改进的力匹配方法成功从时间分辨率在分钟量级的运动数据中提取出有意义的行为模式。
• 焦点狒狒对特定个体位置的响应强度，明显高于对整个群体的响应。
• 仅在群体高度聚集且与焦点个体在空间上明显分离的罕见情况下，才观察到群体层面的凝聚力效应。
• 结果表明，个体移动决策主要受二元社会纽带驱动，而非一般性的群体凝聚力。
• 稀疏数据可提供关于运动中社会影响的可靠洞察，挑战了高分辨率数据为必需的假设。
• 高分辨率（约秒级）和低分辨率（约分钟级）数据集对研究个体对社会与生态环境的响应均具有重要价值。