Fages (2020): Horse males became over-represented during the Bronze Age

Open access Horse males became over-represented in archaeological assemblages during the Bronze Age, Antoine Fages, Andaine Seguin-Orlando, Mietje Germonpré, Ludovic Orlando J. Archaeol. Sci. Reports (2020).

Interesting excerpts:

Endogenous DNA levels provided sufficient read numbers to confirm all 19 samples as horses and to determine the molecular sex of each individual. Five new samples from Goyet and three samples from Trou Magritte were identified as females, while all other 11 specimens consisted of males. This provided a 11:12 (~0.92) male:female sex ratio for the Upper Palaeolithic of Belgium. This number indicated a more balanced male:female sex ratio than the 3:7 value obtained when applying the same computational procedure to previously published Upper Paleolithic horse data from the Taymyr peninsula, north-eastern Siberia, Russia (Orlando et al., 2013, Schubert et al., 2014b). Including two more Upper Palaeolithic horse remains from Kokorevo and Merzly Yar, two sites located in the south-western Siberian range, brought the male:female sex ratio to 5:7 (~0.71), on par with the results obtained in the Upper Palaeolithic of Belgium. This indicates no sex-bias amongst the horse bone assemblages pre-dating horse domestication (two-tailed binomial test, p-value = 0.736).

We next applied the same procedure to Neolithic and Eneolithic osseous remains for which sufficient amounts of sequencing reads were previously generated to confirm the specimens investigated as horses. At Botai, where the earliest evidence for horse domestication was reported (Outram et al., 2009), we found no statistical support for unbalanced male:female sex ratios (15:13, ~1.15; two-tailed binomial test, p-value = 0.851). The same was true when including six older Eneolithic/Neolithic specimens from Russia (Lebyazhinka IV, Altata and Derkul) and Iran (Tepe Mehr Ali), which provided a male:female sex-ratio estimate of 19:15 (~1.17, two-tailed binomial test, p-value = 0.608). Likewise, extending the analyses to six horse males and seven females dated to 4500–5200 cal. BP confirmed the presence of statistically balanced sex-ratios between the first half of third millennium and the sixth millennium BCE (25:22, ~1.14; two-tailed binomial test, p-value = 0.771). Binning the remaining horse data that were previously published for 187 horses revealed highly unbalanced sex-ratios for the following time period (i.e. the last 4600 years), which included 146 males and 43 females (ratio ~3.48). This indicates a statistically significant over-representation of males in osseous horse assemblages from approximately 4600 years ago (binomial test, p-value = 2.57e-14).

Geographical distribution of horses included in this study. A. Geographical distribution of female and male horses. Pie charts are proportional to the total number of specimens excavated at each site. Gold refers to females and green to males. B. Temporal and geographical distribution of horses. Pie charts are proportional to the total number of specimens excavated at each site. The colour gradient indicates the average age of samples per site, from 0 (yellow) to 8000 years (red). Sites older than 8000 years are also shown in red.

The time period around ~3900 years ago marked a drastic shift in male:female sex ratios inferred from excavated remains, after which the horse osteological record comprises approximately four males for every female (Fig. 2). This over-representation of horse males was maintained when disregarding those animals excavated from ritual burial sites (77/25 ~ 3.08 males for every female) and even more pronounced in the animal bones found in funerary contexts (66/14 ~ 4.71 males for every female). This indicates that the status of male and female horses dramatically changed during the Bronze Age period. This is in line with archaeozoological evidence from the Late Bronze Age cemeteries of the Volga-Ural region associated with the Sintashta, Potapovka and Petrovka cultures, that suggest a domination of male horses in funerary rates (Kosintsev, 2010). Interestingly, this pattern somehow mirrors that observed in humans, for whom a clear binary gender structure ubiquitous across all funerary practices, clothing, personal ornaments and representations is not observed during the Neolithic but became the norm from the transition between the Neolithic and the Bronze Age onwards (Robb and Harris, 2018). In addition, the prevalence of male horses in funerary contexts throughout the past three millennia is in line with archaeological evidence from burial sites (Bertašius and Daugnora, 2001, Taylor, 2017) and suggests that stallions (or geldings) were more prized for sacrificial rituals. This is possibly due to symbolic attributes then-associated with masculinity, mounted warriors and chariotry, such as power, protection and strength (Frie, 2018). In particular, petroglyph images associated with vehicles, characterized by two wheels with spokes, became typical by the late third – early second millennium BCE (Jacobson-Tepfer, 2012). They are generally associated with male warriors and the emergence of mobile warfare (Anthony, 2007) or ritual needs, in particular the passage to the after-life land (Jones-Bley, 2000). This suggests an essential ideological role of stallions and their use in elite warfare and ritual practices (Drews, 2004, Kelekna, 2009, Novozhenov and Rogozhonskiy, 2019).

Future research should focus on assessing the molecular sex of horses from Early and Middle Bronze Age Pit Grave and Catacomb cultures, which do show evidence for social inequality, but for which sex inequalities remain to be investigated. Regardless, our observations show that the emergence of a gendered vision of the world in the Bronze Age also extended to the domestic animal sphere.