Part I: Evolution, Taxonomy, and Domestication44
As mentioned previously, during domestication, the most
significant morphological change observable in archaeological
remains is a decrease in size. This has been observed not only in
pigs, but also in other domestic mammals (dogs, cattle, goats,
sheep) (e.g. Davis 1981; Peters et al. 1999; Albarella et al. 2005;
Zeder 2006; Hongo et al. 2009). As a consequence, bones and
teeth of small size are generally attributed to domestic pigs, and
large values to wild boar (see reviews in Albarella et al. 2006;
Rowley-Conwy et al. 2012).
Modern domestic pigs have significantly smaller teeth
(Cucchi et al. 2011b; Evin et al. 2013, 2015a), and crania (Owen
et al. 2014), but an important overlap exists with wild boar.
However, analyses of molar and cranial shape consistently pro-
vide better discrimination between modern wild and domestic
specimens than size alone (Table 4.2), even when the cranium
is fragmented, confirming the higher discriminating power of
GMM over more traditional morphometric approaches.Archaeological Applications of GMM in Identifying
Wild and Domestic Pigs
Identification of archaeological specimens has focused on three
different geographic contexts: Western Eurasia including sites
from Anatolia (Bartosiewicz et al. 2013), Northern Germany
(Krause-Kyora et al. 2013), and Romania (Evin et al. 2015b;
Balasse et al. 2016); China (Cucchi et al. 2011b, 2016); and Island
South East Asia (Larson et al. 2007b; Cucchi et al. 2009).Western Eurasia
The study of Chalcolithic pig remains from Çamlibel Tarlasi in
Central Anatolia (Bartosiewicz et al. 2013) was based on a mul-
tidisciplinary approach combining traditional morphometry,
ancient DNA (aDNA) analysis, and GMM. Altogether the results
indicate a small-scale husbandry practice in which domestic pigs
dominate. In this study, the domestic pigs identified presented
teeth all of small size and domestic shape.
Another study combining aDNA (both mitochondrial and
nuclear DNA) and GMM allowed the identification of specimenswith domestic characteristics in Ertebølle (Mesolithic hunter-
gatherer) and agricultural sites in northern Germany. Mixed
characteristics correspond to a Near Eastern (Y1) haplotype, an
MC1R allele corresponding to a Black spotted coat colour and
a molar shape more similar to modern domestic pigs than to
modern wild boar. One particular specimen from an Ertebølle
context possessed all these features, in addition to showing a
particularly large molar size. This combination of traits appears
to demonstrate the presence of pigs with domestic ancestry in a
hunter-gatherer context some 500 years earlier than previously
thought, calling into question the use of size alone to identify the
status of past Sus specimens. While these conclusions have been
challenged (Evin et al. 2014; Rowley-Conwy & Zeder 2014), a
lack of specific terminology to describe specimens intermediate
between the wild and domestic extremes of the process contin-
ues to confuse interpretation.
Things became clearer with the study of archaeological
specimens from Romania (Evin et al. 2015b). In this study, a
new protocol for identification was proposed, using the large
pre-domestic specimens from the Mesolithic as a reference for a
wild molar shape and the small (smaller than past and modern
wild boar) archaeological specimens as a reference for domestic
molar shape. In doing so, all remaining specimens with a large
molar size were identified according to their shape. A signifi-
cant proportion of those large specimens possessing a domestic
tooth shape were found to be present from the beginning of the
Neolithic through to the Iron Age. This result potentially reflects
the concept of long-term gene flow (previously described)
between wild or domestic forms. Whether these specimens were
actually living in the wild or in close proximity to humans can-
not be assessed by GMM alone and requires the use of additional
(e.g. isotopic) approaches to better understand their true bio-
logical and cultural status (Balasse et al. 2016).Pigs in the Chinese Archaeological Record
It is in China that GMM approaches were first implemented to
track phenotypic divergence induced by the domestication pro-
cess within the archaeological record (Cucchi et al. 2011b). Jing
and Flad (2002) initially challenged the long-standing claim for
a very ancient centre of pig domestication in Southern China
circa 10, 000 BP (Nelson 1998), which drove Cucchi et al. to set
about testing the multiple origin hypothesis. They used Middle
Neolithic specimens from the Yangshao culture as baselines
for the ancient domestic phenotype, and combined these with
modern wild and domestic populations in order to re-evaluate
the presence of domestic pigs at: Zengpiyan, a 10,000-year-
old occupation in a rock shelter where small specimens of Sus
scrofa were claimed to be domestic; Xinglongwa, an 8000-year-
old early Neolithic site in northern China considered one of
the other possible Chinese domestication centres where pigs
are associated with human burials (Shelach 2000); and Jiahu, a
9000-year-old early Neolithic site of the Peiligang culture, where
the earliest evidence of rice cultivation has been claimed (Chen
et al. 1995). The phenotypic relationships seen among these
modern and ancient pig samples clearly refuted the claims for
very early domestic pig at Zengpyian and instead indicated the
occurrence of hunted wild boar of the small southern ChineseTable 4.2 Cross-validation percentages obtained for linear discriminant
analyses between west Palaearctic modern wild and domestic pigs for
molars (Evin et al. 2013) and cranial (Owen et al. 2014) size and shape.Element studied Size Shape
M 2 89.4 93.0
M 3 82.2 92.3
M^2 85.6 93.0
M^3 77.3 92.1
Whole crania 72.2 100
Parietal 43.3 97.9
Nasal 50.5 97.9
Orbit 52.6 88.7
Zygomatic 62.9 97.9
Tooth row 48.5 87.6
Basicranium 72.2 100.00612:31:36