No abstract

increasing their access to key resources, such as food or mates.1-5 Alternatively, it has 5 been argued to be a non-adaptive result of human impacts, such as habitat destruction 6 or provisioning of food. [6][7][8][9] To discriminate between these hypotheses we compiled long-7 term information from 18 chimpanzee communities and 4 bonobo communities. Our 8 data include 152 killings (N=58 observed, 41 inferred, and 53 suspected killings) by 9 chimpanzees in 15 communities and one suspected killing by bonobos. We found that 10 males had the greatest involvement as attackers (92% of participants) and victims 11 (73%); most killings (66%) involved intercommunity attacks; and attackers greatly 12 outnumbered their victims (median 8:1 ratio). Variation in rates of killing among 13communities depended on demographic variables but was unrelated to measures of 14 human impacts. These results from all major study populations over the last five 15 decades are consistent with previously proposed adaptive explanations for killing by 16 chimpanzees but not with the human impact hypothesis. 17 18Conspecific killing has been documented at multiple chimpanzee study sites, 2-5,10-12 but rates 19 vary greatly among sites. The human impact hypothesis and the adaptive strategies 20 hypothesis yield contrasting predictions, which we test here (Tables 1, 2). The human impact 21 hypothesis states that killing occurs mainly as an incidental outcome of aggression, 22 exacerbated by human activities such as providing a concentrated food resource, 23 deforestation-induced crowding, anthropogenic diseases or hunting. Accordingly, lethal 24 aggression should be high where human disturbance is high. In contrast, the adaptive strategies hypothesis views aggression as an evolved strategic 27 response by which aggressors tend to increase their fitness through increased access to 28 territory, food, mates or other benefits. [1][2][3][4][5][10][11][12][13][14][15][16][17] 45Intracommunity infanticide by females may result from intense competition among females 46 for the best feeding areas.17 Population differences in rates of killing are accordingly 47 expected to result from socioecological factors such as differences in grouping patterns 2,11 48 and/or demography.14 Lethal aggression thus occurs within a diverse set of circumstances, 49 but is expected to be most commonly committed by males; directed towards males; directed 50 6 towards non-kin, particularly members of other groups; and committed when overwhelming 51 numerical superiority reduces the costs of killing. 52 53Previous studies have developed and tested these specific hypotheses 2,5,[11][12][13][14][15][16][17] ; the present study 54 represents the first effort to test multiple hypotheses simultaneously with a comprehensive 55 dataset. To do so, we assembled data from 18 chimpanzee communities from both eastern 56 (N=12) and western (N=6) clades 24 of chimpanzees studied over 426 years (median = 21 57 years; range: 4-53) and from 4 bonobo communities studied for 92 years (media...

Phylogenetic trees for the four extant species of African hominoids are presented, based on mtDNA control region-1 sequences from 1,158 unique haplotypes. We include 83 new haplotypes of western chimpanzees and bonobos. Phylogenetic analysis of this enlarged database, which takes intraspecific geographic variability into account, reveals different patterns of evolution among species and great heterogeneity in species-level variation. Several chimpanzee and bonobo clades (and even single social groups) have retained substantially more mitochondrial variation than is seen in the entire human species. Among the 811 human haplotypes, those that branch off early are predominantly but not exclusively African. Neighbor joining trees provide strong evidence that eastern chimpanzee and human clades have experienced reduced effective population sizes, the latter apparently since the Homo sapiens-neanderthalensis split. Application of topiary pruning resolves ambiguities in the phylogenetic tree that are attributable to homoplasies in the data set. The diverse patterns of mtDNA sequence variation seen in today's hominoid taxa probably ref lect historical differences in ecological plasticity, female-biased dispersal, range fragmentation over differing periods of time, and competition among social groups. These results are relevant to the origin of zoonotic diseases, including HIV-1, and call into question some aspects of the current taxonomic treatment and conservation management of gorillas and chimpanzees.In spite of the absence of relevant fossils, studies of genetic variation have settled the question of whether humans are more closely related to gorillas or chimpanzees in favor of the latter (1). However, most such comparative studies have been based on no more than six individuals of each taxon with the exception of humans (2-9). Furthermore, the individual apes used in these comparisons are typically of unknown geographic origin, and there has been a tendency to treat all chimpanzees as a homogeneous group. Advances in DNA sequencing technology and noninvasive genotyping (10-13) now permit a more thorough analysis of the geographic variation within and among mitochondrial DNA (mtDNA) sequences in each of the living African hominoids. The noncoding control region 1 (CR1) was selected for analysis because it is hypervariable and was thought to be relatively free from direct natural selection and therefore was thought to provide a less ambiguous record of mutational change and phylogenetic relationships (9).Here we present a comparison of genetic variation in all nine recognized taxa of African or African-derived hominoid: (i) western lowland gorilla (Gorilla g. gorilla); (ii) eastern moun- (14). We also include three other groups in our analyses: a Neandertal (Homo s. neanderthalensis) and Bornean and Sumatran orangutans (Pongo p. pygmaeus and Pongo p. abelii, respectively). Using common names except when ambiguous, we show that these taxa have very different amounts and patterns of genetic variation, with hum...