Long-term consequences of poaching on African elephant social structure, genetic relatedness, physiology and reproduction

We assessed the long-term impacts of poaching by investigating genetic, physiological and reproductive correlates of a disturbed social structure resulting from historically heavy poaching of an African elephant ( Loxodonta africana) population in Mikumi National Park, Tanzania. We collected fecal DNA, hormone metabolites and behavioral data from 218 adult female elephants from 109 groups differing in size, age structure and family structure over 25 months from 2003 to 2005. Using 10 microsatellite loci, DNA was genotyped in 203 elephants and pair-wise relatedness was calculated among adult females within and between groups. We examined the demographic character of this population with relatedness measures and observations of female bonding, and we examined individual physiology with fecal glucocorticoid levels and reproductive output measures. The distribution in group size Mikumi has changed little since 1989 with a high proportion of solitary females. However, some recovery has occurred as evidenced by a 14.2% increase in the number of families with tusked old matriarchs. The population is characterized by small group size, considerable variation in group relatedness, females with no first-order adult relatives and females that form only weak social bonds. We used gene-drop analysis and a model of an unpoached pedigree to compare our observed Mikumi group relatedness to a simulated unpoached expectation. The majority of groups in Mikumi contain 2 to 3 adults; of these, 45% were classified as genetically disrupted. Bonding, quantified with a pair-wise association index, was significantly correlated with relatedness; however only half of the females formed strong bonds with other females, and relatedness was substantially lower for a given bond strength as compared to an unpoached population. Female elephants lacking kin demonstrated considerable behavioral plasticity in this disturbed environment, grouping with other females lacking kin, with established groups, or remaining alone, unable to form any stable adult female-bonds. Disrupted groups were less cohesive, as quantified with an index of proximity, and engaged in more cooperative and competitive interactions with other groups than genetically intact groups. Females from groups that lacked an old matriarch, first-order adult relatives and strong social bonds had significantly higher fecal glucocorticoid values than those from groups with these features (all females R2 = 0.31; females in multi-adult groups R2 = 0.46). Females that frequented isolated areas with historically high poaching risk had higher fecal glucocorticoid values than those in low-poaching risk areas. Females with weak bonds and low group relatedness had significantly lower reproductive output (R 2 (U) = 0.21); those from disrupted groups had significantly lowered reproductive output, despite many being in their reproductive prime. These results suggest that long-term negative impacts from poaching of old, related matriarchs have persisted among adult female elephants 1.5 decades after the 1989 ivory ban was implemented.