Locating unpredictable but essential resources is a task that all mobile animals have to perform in order to survive and reproduce. Research on search strategies has focused largely on independent individuals [1, 2, 3], but many organisms display collective behaviors, including during group search and foraging [4, 5, 6]. One classical experimental search task, informing studies of navigation, memory, and learning, is the location of a reward in a confined, complex maze setting [7, 8]. Rats (Rattus norvegicus) have been paradigmatic in psychological and biological studies [9, 10], but despite rats being highly social [11, 12], their group search behavior has not been investigated. Here, we explore the decision making of rats searching individually, or in groups, for a reward in a complex maze environment. Using automated video tracking, we find that rats exhibit—even when alone—a partially systematic search, leading to a continuous increase in their chance of finding the reward because of increased attraction to unexplored regions. When searching together, however, synergistic group advantages arise through integration of individual exploratory and social behavior. The superior search performances result from a strategy that represents a hierarchy of influential preferences in response to social and asocial cues. Furthermore, we present a computational model to compare the essential factors that influence how collective search operates and to validate that the collective search strategy increases the search efficiency of individuals in groups. This strategy can serve as direct inspiration for designing computational search algorithms and systems, such as autonomous robot groups, to explore areas inaccessible to humans.

Video Abstract

Download : Download video (31MB)

中文翻译：

---

大鼠群体搜索的协同效益。

寻找不可预测但必不可少的资源是所有移动动物为了生存和繁殖而必须执行的任务。搜索策略的研究主要集中在独立个体 [1, 2, 3]，但许多生物体表现出集体行为，包括在群体搜索和觅食期间 [4, 5, 6]。一项经典的实验搜索任务，为导航、记忆和学习的研究提供信息，是在狭窄、复杂的迷宫环境中确定奖励的位置 [7, 8]。大鼠（Rattus norvegicus) 已成为心理学和生物学研究的典范 [9, 10]，但尽管老鼠具有高度的社交性 [11, 12]，但尚未调查它们的群体搜索行为。在这里，我们探讨了老鼠在复杂的迷宫环境中单独或成组搜索以获得奖励的决策。使用自动视频跟踪，我们发现老鼠表现出——即使是单独时——部分系统的搜索，由于对未探索区域的吸引力增加，导致它们找到奖励的机会不断增加。然而，当一起搜索时，协同的群体优势会通过个体探索行为和社会行为的整合而产生。卓越的搜索性能源于一种策略，该策略代表响应社会和非社会线索的有影响力的偏好层次结构。此外，我们提出了一个计算模型来比较影响集体搜索运作方式的基本因素，并验证集体搜索策略提高了群体中个人的搜索效率。这种策略可以作为设计计算搜索算法和系统的直接灵感，例如自主机器人组，以探索人类无法进入的区域。

视频摘要

下载 :下载视频 (31MB)