Identifying functional and regional differences in chimpanzee stone tool technology

Abstract

The earliest hominin archaeological sites preserve a record of stone tools used for cutting and pounding. Traditionally, sharp-edged flakes were seen as the primary means by which our earliest ancestors interacted with the world. The importance of pounding tools is increasingly apparent. In some cases, they have been compared with stone hammers and anvils used by chimpanzees for nut-cracking. However, there has been little focus on providing a robust descriptive and quantitative characterization of chimpanzee stone tools, allowing for meaningful comparisons between chimpanzee groups and with archaeological artefacts. Here we apply a primate archaeological approach to characterize the range of chimpanzee nut-cracking stone tools from Djouroutou in the Taï National Park. By combining a techno-typological analysis, and two- and three-dimensional measures of damage, we identify clear differences in the location and extent of damage between nut-cracking hammerstones and anvils used at Djouroutou and when compared with other wild chimpanzee populations. Furthermore, we discuss these results in relation to interpretations of Plio-Pleistocene percussive technology. We highlight potential difficulties in identifying the underlying function of percussive artefacts based on morphological or techno-typological attributes alone. The material record from Djouroutou represents an important new datum of chimpanzee regional and material culture.

Keywords: Palaeolithic; archaeology; nut-cracking; percussive technology; primate archaeology; stone tools.

Figure 1.

Examples of chimpanzee nut-cracking in the Taï forest and a nut-cracking site at Djouroutou. (a) Female chimpanzee cracking Panda oleosa nuts using a granodiorite hammerstone on a wooden (panda tree root) anvil (Credit: Liran Samuni, Taï Chimpanzee Project), and (b,c) examples of an active Panda oleosa nut-cracking site at Djouroutou. Note the combination of hammerstone, anvil and fresh nut debris.

Figure 2.

Descriptive statistics for all hammerstones and anvils are included in this study. (a) Maximum dimensions of hammerstones and anvils included in this study; (b) mass of all hammerstones separated by target nut cracked; (c) maximum technological length and width of all flake scars detached from hammerstones; (d) the number of discrete use-wear areas on both hammerstones and anvils; (e) the surface area (cm²) of all discrete use-wear areas on both hammerstones and anvils, and (f) the percentage of surface area (PA) occupied by discrete areas of percussive use-wear.

Figure 3.

Examples of chimpanzee hammerstones (M Gra nite Hammer (a–d); PPQ1003 (e–h); CGG23 (i–l)) from Djouroutou included in this study illustrating their textured surface (a,e,i); three-dimensional surface (b,f,j); surface depth (mm) (c,g,k) and surface gradient (d,h,l) with location of all pits overlain.

Figure 4.

Examples of chimpanzee anvils (PGU3012 (a–d); PrGU3014 (e–h); PGU3013 (i–l)) from Djouroutou included in this study illustrating their textured surface (a,e,i); three-dimensional surface (b,f,j); surface depth (mm) (c,g,k) and surface gradient (d,h,l) with location of all pits overlain. All examples are metamorphosed granite.

Figure 5.

Density plot of depth, gradient and surface roughness values for all use-wear regions between (a) hammerstones and anvils, (b) different raw materials on hammerstones alone and (c) raw materials on anvils alone.

Figure 6.

Comparative analysis between chimpanzee hammerstones and anvils from Djouroutou and anvils reported from Bossou [39]. Boxplots showing differences in (a) the number of discrete use-wear areas, (b) the relative surface area affected by percussive damage (PA), (c) the mean distance from the centre and (d) edge of the active surface to the geometric centre of discrete use-wear areas and (e) the results of a principal component analysis showing differences between the anvils and hammers of Djouroutou and with the anvils from Bossou using all measures listed in table 4.