This report is the first installment of a monographic study of mammalian diversity and ethnomammalogy in a sparsely inhabited rainforest region between the Yavarí and Ucayali rivers in northeastern Peru. Our study is based on several large collections of mammals (totaling about 3500 specimens) made at various localities in this region between 1926 and 2003, and on our long-term ethnobiological and linguistic fieldwork with the Matses, a Panoan-speaking group of indigenous Amazonians who still obtain most of their dietary protein by hunting mammals. Our primary objectives are to document the species richness of the regional fauna through taxonomic analysis of collected specimens, and to assess the detail and accuracy of Matses knowledge of mammalian natural history by linguistic analysis of recorded interviews.

The regional primate fauna is definitely known to consist of at least 14 species documented by collected specimens and/or repeated sightings of taxa with visually conspicuous diagnostic traits. This fauna includes three atelids (Alouatta seniculus, Ateles belzebuth, Lagothrix lagothricha), eight cebids (Aotus nancymaae, Callimico goeldii, Callithrix pygmaea, Cebus albifrons, Cebus apella, Saguinus fuscicollis, Saguinus mystax, Saimiri sciureus), and three pitheciids (Cacajao calvus, Callicebus cupreus, Pithecia monachus). All 14 species are known to occur sympatrically at one inventory site, but Goeldi's monkey (Callimico goeldii) is rare and uakaris (Cacajao calvus) seem to be patchily distributed, so some local faunas may have only 12 or even fewer species. This regional fauna is unique because neighboring interfluvial regions lack some species that are present in the Yavarí-Ucayali interfluve, and because some species that are present in neighboring interfluvial regions are not known to occur between the Yavarí and the Ucayali.

Matses knowledge about primate natural history is clearly correlated with size and cultural importance. For example, information obtained from standardized interviews about spider monkeys (Ateles belzebuth, a large game species) can be parsed into 86 observations about its ecology and/or behavior, whereas interviews about pygmy marmosets (Callithrix pygmaea, a small nongame species) contain only nine observations on these topics. Item-by-item comparisons of Matses observations about spider monkeys with the published results of scientific field research suggests that the Matses are generally accurate observers of primate natural history, a conclusion that is additionally supported by comparing community patterns of resource use compiled from our interview data with community-ecological studies of primate faunas in the scientific literature. Most exceptions (discrepancies between Matses observations and the scientific literature) can be explained by cultural inattention to small nongame species.

Although these results suggest that archiving native Amazonian knowledge about mammalian natural history might be a cost-effective alternative to lengthy fieldwork for some research objectives, there are significant linguistic barriers than can inhibit effective cross-cultural communication. Among the Matses, these include a surprisingly large number of zoologically redundant names (synonyms and hyponyms). R