Data from: Vervet monkeys use paths consistent with context-specific spatial movement heuristics

Animal foraging routes are analogous to the computationally demanding "traveling salesman problem" (TSP), where individuals must find the shortest path among several locations before returning to the start. Humans approximate solutions to TSPs using simple heuristics or "rules of thumb," but our knowledge of how other animals solve multidestination routing problems is incomplete. Most nonhuman primate species have shown limited ability to route plan. However, captive vervets were shown to solve a TSP for six sites. These results were consistent with either planning three steps ahead or a risk-avoidance strategy. I investigated how wild vervet monkeys (Chlorocebus pygerythrus) solved a path problem with six, equally rewarding food sites; where site arrangement allowed assessment of whether vervets found the shortest route and/or used paths consistent with one of three simple heuristics to navigate. Single vervets took the shortest possible path in fewer than half of the trials, usually in ways consistent with the most efficient heuristic (the convex hull). When in competition, vervets' paths were consistent with different, more efficient heuristics dependent on their dominance rank (a cluster strategy for dominants and the nearest neighbor rule for subordinates). These results suggest that, like humans, vervets may solve multidestination routing problems by applying simple, adaptive, context-specific "rules of thumb." The heuristics that were consistent with vervet paths in this study are the same as some of those asserted to be used by humans. These spatial movement strategies may have common evolutionary roots and be part of a universal mental navigational toolkit. Alternatively, they may have emerged through convergent evolution as the optimal way to solve multidestination routing problems.

动物觅食路径与计算复杂度极高的旅行商问题（Traveling Salesman Problem, TSP）具有相似性：该问题要求个体在返回起点前，寻找到访多个地点的最短路径。人类可通过简单启发式规则或“经验法则”近似求解旅行商问题，但目前学界对其他动物如何解决多目的地路径规划问题的认知仍不全面。多数非人灵长类物种的路径规划能力较为有限，但此前有研究显示，圈养长尾黑颚猴可完成包含6个站点的旅行商问题求解，该结果既契合“向前规划三步”的策略，也符合风险规避型决策逻辑。本研究针对野生长尾黑颚猴（Chlorocebus pygerythrus）展开探究，旨在明确其如何解决包含6个报酬均等的食物站点的路径规划问题；通过预设站点布局，可判断长尾黑颚猴是否找到了最短路径，以及是否采用了三类简单启发式规则中的某一种完成导航。试验结果显示，单独行动的长尾黑颚猴仅在不足半数的试验中找到了最短路径，且其路径选择通常符合效率最高的启发式规则——凸包（convex hull）策略。当存在竞争场景时，长尾黑颚猴的路径选择会根据其社会支配等级匹配不同的更高效启发式规则：优势个体采用集群策略，从属个体则遵循最近邻规则。上述结果表明，与人类相似，长尾黑颚猴或可通过运用简单、自适应且适配场景的“经验法则”来解决多目的地路径规划问题。本研究中与长尾黑颚猴路径选择相符的启发式规则，与学界提出的人类所使用的部分启发式规则一致。这类空间移动策略或许拥有共同的进化起源，属于通用心理导航工具包的组成部分；也有可能这类策略是通过趋同进化演化而来，成为解决多目的地路径规划问题的最优方式。