Dataset S4 from Foraging history of individual elephants using DNA metabarcoding

Individual animals should adjust diets according to food availability. We used DNA metabarcoding to construct individual-level dietary timeseries for elephants from two family groups in Kenya varying in habitat use, social position and reproductive status. We detected at least 367 dietary plant taxa, with up to 137 unique plant sequences in one fecal sample. Results matched well-established trends: elephants tended to eat more grass when it rained and other plants when dry. Nested within these switches from ‘grazing’ to ‘browsing’ strategies, dietary DNA revealed seasonal shifts in food richness, composition and overlap between individuals. Elephants of both families converged on relatively cohesive diets in dry seasons but varied in their maintenance of cohesion during wet seasons. Dietary cohesion throughout the timeseries of the subdominant ‘Artists’ family was stronger and more consistently positive compared to the dominant ‘Royals’ family. The greater degree of individuality within the dominant family's timeseries could reflect more divergent nutritional requirements associated with calf-dependency and/or priority access to preferred habitats. Whereas theory predicts that individuals should specialize on different foods under resource scarcity, our data suggest family bonds may promote cohesion and foster the emergence of diverse feeding cultures reflecting links between social behaviour and nutrition.动物个体应依据食物可获取性调整饮食结构。本研究采用DNA宏条形码（DNA metabarcoding）技术，为肯尼亚两个象群的大象构建了个体水平的饮食时间序列数据集，这两个象群在栖息地利用模式、社会地位以及繁殖状态上均存在差异。本次检测共检出至少367种食用植物类群，单份粪便样本中最多可获得137条独特植物序列。研究结果与学界已确立的规律高度吻合：大象在雨季倾向于取食更多草本植物，旱季则更多取食木本及其他类型植物。在从‘草本取食（grazing）’到‘木本取食（browsing）’的取食策略转换中，饮食DNA分析揭示了个体间食物丰富度、群落组成以及食物重叠度的季节动态变化。两个象群的大象在旱季均会形成相对统一的饮食结构，但在雨季维持饮食一致性的程度存在显著差异。相较于优势象群‘Royals’，从属象群‘Artists’在整个时间序列中的饮食一致性更强，且该一致性始终保持正向。优势象群‘Royals’的时间序列中个体间差异化程度更高，这可能反映了与幼崽抚育相关的营养需求更具分化性，或是该象群拥有优先获取优质栖息地的权利。尽管理论预测在资源匮乏条件下，个体应特化取食不同食物以适应竞争，但本研究数据表明，家族纽带或许能促进群体饮食的一致性，并催生多样化的取食文化，这一发现反映了社会行为与营养获取之间的内在关联。