ZooMS confirms geometric royalsocietypublishing.org/journal/rsos morphometrics species identification of ancient Research sheep and goat Cite this article: Jeanjean M et al. 2023 ZooMS Marine Jeanjean1, Krista McGrath2, Silvia Valenzuela- Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 confirms geometric morphometrics species identification of ancient sheep and goat. R. Soc. Lamas3, Ariadna Nieto-Espinet4, Renate Schafberg5, Open Sci. 10: 230672. Pere Miquel Parés-Casanova6, Sergio Jiménez- https://doi.org/10.1098/rsos.230672 Manchón3, Claude Guintard7,8, Faiza Tekkouk9, Rania Ridouh9, Cyprien Mureau1 and Allowen Evin1 Received: 17 May 2023 1 Accepted: 25 August 2023 Institute of Evolutionary Science-Montpellier (ISEM), University of Montpellier, CNRS, EPHE, IRD, Montpellier, France 2 Department of Prehistory & Institute of Environmental Science and Technology (ICTA), Universitat Autònoma de Barcelona, 08193, Barcelona, Spain 3 Archaeology of Social Dynamics (ASD), Institució Milà i Fontanals de Recerca en Humanitats, Subject Category: Consejo Superior de Investigaciones Científicas (IMF-CSIC), C/ Egipcíaques 15, 08001 Barcelona, Spain Earth and environmental science 4 Grup d’Investigació Prehistòrica (GIP), Departament d’Història, Universidad de Lleida, 25005 Lleida, Spain 5 Subject Areas: Central Natural Science Collections, Martin Luther University Halle-Wittenberg, Domplatz 4, taxonomy and systematics/environmental science 06108 Halle (Saale), Germany 6 Institució Catalana d’Història Natural, Barcelona, Catalonia, Spain 7 Laboratoire d’Anatomie comparée, Ecole Nationale Vétérinaire, de l’Agroalimentaire et de Keywords: l’Alimentation, Nantes Atlantique – ONIRIS, Nantes Cedex 03, France 8 caprines identifications, GMM, Ovis aries, Capra GEROM, UPRES EA 4658, LABCOM ANR NEXTBONE, Faculté de santé de l’Université d’Angers, hircus, paleoproteins, third lower molar Angers, France 9 Institut des Sciences Vétérinaires, Laboratoire « Gestion de la santé et productions animales », Université des frères Mentouri, El Khroub, Algérie MJ, 0000-0001-5407-509X; KM, 0000-0003-0692-1217; Author for correspondence: AE, 0000-0003-4515-1649 Allowen Evin e-mail: allowen.evin@umontpellier.fr Geometric morphometrics can effectively distinguish isolated third lower molars of present-day sheep and goat, but its applicability to archaeological specimens has yet to be established. Using a modern reference collection of 743 sheep and goats and a two-dimensional landmark-based geometric morphometric (GMM) protocol, this study aimed to morphometrically identify 109 archaeological specimens, used as case studies, dating from the Late Neolithic to the modern period/era. These morphometric identifications were then compared to molecular identifications via collagen peptide Electronic supplementary material is available mass fingerprinting, known as Zooarcheology by Mass online https://doi.org/10.6084/m9.figshare.c. 6824155. © 2023 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited. Spectrometry (ZooMS). ZooMS confirmed the morphometric identifications for 104 specimens, with 2 the five misidentified specimens all morphometrically identified as goat. Modern sheep and goats royalsocietypublishing.org/journal/rsos have larger teeth and distinct shapes compared to their archaeological counterparts, suggesting strong differences between archaeological and modern specimens potentially linked with recent breed improvement or geographical origin of the specimens. In addition, for both species, some of the archaeological dental morphologies do not match with any of our modern references. This study validates the applicability of geometric morphometrics for identifying isolated archaeological sheep and goat teeth. It represents a stepping stone for future, non-destructive, bioarchaeological studies of the two species. 1. Introduction Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 R. Soc. Open Sci. 10: 230672 Zooarchaeology by mass spectrometry (ZooMS) is increasingly used in archaeology for reliable taxonomic identification of organic materials [1] such as bones [2], leather [3] or parchment [4]. ZooMS, by using peptide mass fingerprinting of collagen, provides an inexpensive tool for archaeological studies and has been used to distinguish domestic caprine species and to detect their arrival in Africa [5–8], Asia [9,10] and Europe [11,12], and to study distinct farming practices [13–20]. Sheep (Ovis aries) and goat (Capra hircus) are two important species of agro-pastoral systems in the Mediterranean basin since the Neolithic and were often herded together [21–29]. However, they possess distinct ecological [30,31] and economical properties [32–36], making it of prime interest to establish secure identifications of ancient specimens before studying the two species separately. Criteria for distinguishing sheep and goat among archaeological assemblages exist, notably from post- cranial bones [37–39] and have greatly improved the ability to differentiate the two species. On teeth, discrete criteria have been proposed [31,40–45], but none has been universally used for identifying isolated lower third molars that are often found among archaeological assemblages, since, unlike bones, they are better preserved [46], but also often left unidentified because of the lack of secure identification criteria. Geometric morphometrics (GMM) is more and more used in bioarchaeology [47]. This series of analytical tools is based on the multivariate statistical analysis of set of Cartesian coordinate data allowing us to explore in-depth morphometric variation [48]. Recently, such approaches have been used to quantify sheep and goat morphometric variation based on the size and shape of their lower third molar [49]. This study demonstrated the ability to identify adult modern specimens of the two species with a probability of 93.3%. In addition, the identification of Middle Ages specimens from the site of Missignac (Aimargues, southern France), used as a first case study, appeared congruent with the initial zooarchaeological faunal analyses and interpretations [50,51]. However, in order to assess whether the same geometric morphometrics (GMM) protocol can be used to identify ancient sheep and goat teeth originating from a broader spatio-temporal scale, it is necessary to validate the approach using securely identified archaeological specimens. Archaeological teeth can be found either enclosed in mandibular bone that possess distinct features between the two species [40,43], or isolated (i.e. without the bone), which make their identification particularly challenging even if some discrete criteria have been proposed for both mandible and teeth [31,40–45]. Being able to provide reliable identification is therefore of prime interest to bioarchaeological studies. Here we aim to (1) identify archaeological specimens based on their third molar shape using geometric morphometrics (GMM), (2) compare those morphometric identifications with newly generated molecular Mass Spectrometry (ZooMS) identifications, and (3) compare the size and shape morphometric variation of modern sheep and goat with their securely identified archaeological counterparts. 2. Material and methods Third lower molars of 743 modern specimens, corresponding to 521 sheep and 222 goats of various breeds and geographical origins from Western Eurasia, including western Europe and north Africa, were used as a reference for the geometric morphometrics analysis (electronic supplementary material, table S1). Part of these specimens were already included in a previous study [49]. Only one tooth per specimen was considered. All studied specimens were adults, with fully erupted third lower molars. 3 royalsocietypublishing.org/journal/rsos Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 R. Soc. Open Sci. 10: 230672 0 100 200 km Figure 1. Location of the 13 archaeological sites of the sampled specimens. The numbers refer to the column ID of table 1. Tooth wear stages were used to estimate the age at death of the specimens following the classification established by Payne (1973, 1985) [42,52] and they were ranging from stages 0 to 17 G (i.e. from 1–2 years to more than 8 years). Sheep and goat display distinct sexual dimorphism and changes through age based on their third molar morphometry, but both factors were found to have little impact on the species differentiation [49]. In addition, 109 individuals (electronic supplementary material, table S2) from 13 archaeological sites, selected for their chronological and spatial diversity, dated from the Neolithic to Modern period and located in the north-western part of the Mediterranean basin ( present-day Catalonia region in Spain and Occitanie region in Southern France), were used to compare GMM and ZooMS identifications (figure 1, table 1). The selected archaeological specimens consisted of third lower molars with at least a small portion of adherent mandibular bone. For each specimen, the size and shape of the third lower molar were quantified through geometric morphometrics and used for morphometric identification, and a fragment of mandibular bone (normally between 10 to 50 mg) was sampled for molecular ZooMS identification. 2.1. Geometric morphometrics The occlusal view of the third lower molars was photographed using a Nikon d90 LSR camera paired with a 60 mm macro lens (AF-S Micro NIKKOR) attached to a photographic arm (manfrotto 244RC) following a previously published protocol [49]. Right molars were preferentially photographed, and when necessary, photographs of left molars were mirrored prior to GMM data acquisition. Third molar size and shape were quantified using two-dimensional landmarks and sliding landmarks-based geometric morphometrics approaches following the protocol developed in our previous study [49] (figure 2). The coordinates of seven landmarks and of 48 sliding semi-landmarks distributed along six curves (eight equidistant points in each) were acquired using TPSdig2 [68] (for a formal description of the landmark positions see electronic supplementary material, table S2 of [49]) (figure 2). This protocol does not include points along the mesial part of the teeth that is often affected by lateral wear (Jeanjean et al., [49]). All data are available in electronic supplementary material, table S3. Table 1. Description of the 13 archaeological sites sampled, indicating geographical origin, chrono-cultural period of occupation, 4 and number of specimens studied by geometric morphometrics and ZooMS. royalsocietypublishing.org/journal/rsos chrono-cultural number of ID Country city name of the site occupation specimens references 1 Spain Sant Quirze del Bòbila Madurell-Mas Neolithic 1 [53] Vallès Duran 2 Spain Maldà Cantorella Neolithic 1 [54] 2 Spain Maldà Cantorella Bronze Age 1 [54] 3 Spain Sabadell Bòbila Madurell - Can Bronze Age 2 [55] Gambús 1 Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 1 Spain Sant Quirze del Bòbila Madurell-Mas Iron Age 2 [53] R. Soc. Open Sci. 10: 230672 Vallès Duran 4 France Sigean Pech Maho Iron Age 15 [56] 5 Spain Arbeca Vilars Iron Age 12 [57] 6 Spain Calafell Alorda Park Iron Age 1 [58] 7 Spain L’Escala Empúries Iron Age 1 [59] 8 Spain Sabadell Can Feu Iron Age 1 [60] 4 France Sigean Pech Maho Iron Age/Antiquity 1 [56] transition 3 Spain Sabadell Bòbila Madurell - Antiquity 21 [55] Can Gambús 1 9 Spain Lleida L’Antic Portal de Antiquity 1 [61] Magdalena 10 France Aimargues Missignac Middle Age 18 [62,63] 11 France Montpellier CNR Middle Age 20 [64] 12 Spain La Llacuna Vilademàger Middle Age 1 [65] 13 Spain Lleida La Cuirassa Middle Age 4 [66,67] 13 Spain Lleida La Cuirassa Modern period 1 [66,67] 5 Spain Arbeca Vilars Out of stratigraphy 4 [57] 2.2. Morphometrics identification All modern and archaeological specimens were superimposed together using a Generalized Procrustes Analysis (GPA) during which the sliding semi-landmarks were allowed to slide by minimizing the sum of the Procrustes distances between each individual and the mean conformation [69–71]. The Procrustes residuals were first analysed using a Principal Component Analysis (PCA) [72–74] (electronic supplementary material, figure S1). A predictive linear discriminant analysis ( pLDA), computed on the 20 first PCA scores that maximize the differences between groups [75], was then used to predict to which species the archaeological specimens belonged. Because unbalanced sample size can have a profound effect on discriminant analyses [76], and the fact that sheep greatly outnumber goat in the reference dataset, we used a resampling procedure to down sample the number of sheep to the smaller number of goats, as recommended by previous analysis [77]. The morphometrics identification was therefore based on 100 predictive discriminant analyses ( pLDA) performed on two equal-size samples (here 222 in each group) whose results were summarized by the percentage of times a specimen was identified to its correct group [78]. All discriminant analyses were retained and no threshold was fixed for posterior probabilities (i.e. all identifications were based on the 100 pLDA). Only LDA with a cross validation percentage (CVP) above the 3rd quartile of all CVP were retained in order to select the most discriminant analyses. Morphometric identifications were only based on molar shape data, and not size, since it has been previously identified as the most discriminative criteria between sheep and goat [49]. buccal 5 royalsocietypublishing.org/journal/rsos mesial distal lingual 1 3 5 7 Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 R. Soc. Open Sci. 10: 230672 6 4 2 Figure 2. Geometric morphometrics protocol: position of the seven landmarks (in blue) and 48 sliding semi-landmarks (in red) whose coordinates were acquired on a right third lower molar of a sheep. Landmark positions are described in electronic supplementary material, table S2 of [49]. 2.3. ZooMS Between 10 and 50 mg of bone was weighed out for each specimen. 250 µl of 0.6 M hydrochloric acid (HCl) was added to each sample then stored at 4°C until being demineralized. Once demineralized, the acid was discarded and three washes of 200 µl each of ammonium bicarbonate (AmBic, NH2HCO2, pH8) were performed to remove any remaining acid. 100 µl of AmBic was added and the samples were gelatinized at 65°C for 1 h. Samples were centrifuged for 1 minute and 50 µl of the gelatinized supernatant was transferred to a new tube and 0.4 µg µl−1 of trypsin was added. The trypsinated samples were digested at 37°C overnight, after which the samples were acidified with 1 µl of 5% trifluoroacetic acid to stop the trypsin. C18 zip tips (Pierce, Thermo Scientific) were used to purify the samples which were eluted in a 50 µl volume. 1 µl of sample topped with 1 µl of matrix (α- cyano-4-hydroxycinnamic acid) was spotted in triplicate along with calibration standards on either a Bruker MTP 384 ground steel MADLI plate and analysed on a Bruker Ultraflex III MALDI-ToF-MS, or onto a SCIEX Opti ToF stainless steel MALDI plate and analysed with an AB SCIEX 4700 MALDI- ToF- MS in the Department of Prehistory & Institute of Environmental Science and Technology (ICTA) at Universitat Autònoma de Barcelona (Barcelona, Spain). Triplicate spectra for each sample were averaged and analysed using mMass [79] and compared to a database of known reference species [2,80–82]. 2.4. Comparison of ancient and modern sheep and goat The overall percentage of correct identification was computed comparing the geometric morphometrics and ZooMS identifications. Securely ZooMS-identified archaeological sheep and goat were then compared with their modern counterparts using Wilcox-tests for size and Procrustes Anova ( procD.lm function from geomorph package [83]) for shape. Size and shape variation were visualized using boxplot and PCA, respectively. Homogeneity of size and shape difference between species and between archaeological versus modern specimens were tested using 2-way Anova and 2-way Procrustes Anova, respectively. Morphometrics proximities between modern sheep and goat and the archaeological specimens identified via ZooMS were visualized by a neighbour-joining network computed on Mahalanobis distances. Graphics were created in R using the ‘ggplot2’ package [84] Change in molar shape and size variance between archaeological and modern specimens of sheep and goat was assessed, respectively, using the disparity test of the geomorph R package [85,86] and Fligner test (stats package [87]). All the analysis were performed in R v. 6.1.524 [87] through Rstudio v. 4.2.2 [88]. n = 45 n = 64 n = 222 n = 521 6 royalsocietypublishing.org/journal/rsos 1.9 log(CS) 1.7 1.5 archaeological goat archaeological sheep modern goat modern sheep Figure 3. Third lower molar size variation: boxplot showing variation in log centroid size between modern (dark) and archaeological Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 (light) sheep (green) and goat (blue) identified by ZooMS. For differences between sites and chronologies, see electronic R. Soc. Open Sci. 10: 230672 supplementary material, figure S2. 3. Results 3.1. Identification for archaeological specimens Based on their third molar shape, 59 archaeological specimens were identified as sheep (54.1%) and 50 as goat (45.9%) (figure 3). The 109 identified were based on 31 LDA with cross-validation percentages above 94.22% [90% Confidence Interval: 93.01–95.50]. For the same 109 specimens, ZooMS identified 64 (58.7%) specimens as sheep and 45 (41.3%) as goat (all spectra supporting the results are available in the Dryad Digital Repository https://doi.org/10.5061/ dryad.fj6q5741m [89]). The two approaches provided the same species identification for 104 (95.4%) of the archaeological specimens corresponding to 59 sheep and 45 goats. The five (4.6%) remaining specimens are from the sites of Can Gambús (Catalonia, Late Antiquity) (n = 3) and Missignac (France, Late Antiquity and Middle Age) (n = 2). These misidentified specimens are all morphometrically identified as goat but molecularly identified as sheep. All of these misidentified specimens fell into the most common age category of 4–8 years old (i.e. 11 G) (electronic supplementary material, table S2). Looking more closely at these misidentifications, it appears that three out of the five specimens were identified 100% of the time to goat via GMM (i.e. all LDA give ‘goat’ as identification, no matter the probability of each of the 100 LDAs), while the two last ones were attributed respectively to goat only in 74.2% and 81.8% of the cases (see electronic supplementary material, figure S1 for probability of each LDA). 3.2. Morphometrics comparison with modern specimens The ancient samples studied were selected to represent a diversity of geographical origins (Catalonia and the South of France) and chrono-cultural periods (Late Neolithic to modern period) but certainly do not cover the entire past sheep and goat diversity in the North Western Mediterranean basin. Sample sizes per region and chrono-cultural period were too small to envision detailed morphometrics comparison and all archaeological specimens were analysed jointly (see electronic supplementary material, figure S2 for a visualization of size differences between sites and chrono-cultural periods). Overall, the archaeological sheep and goat possessed smaller third lower molars than their modern relatives (figure 3, table 2). Modern sheep also appear larger than goat, contrary to the pattern observed for archaeological specimens (figure 3, table 2). A PCA of all archaeological and modern sheep and goat shows that the modern specimens studied show new morphologies that didn’t exist in the past (figure 4). Shape differences between sheep and goat are present for both modern and archaeological specimens (respectively p-value = 0.013 and p-value = 0.001). Mean shape visualization and dissimilarity network on shape confirm that there is a higher proximity within species than within periods, that is to say the archaeological goat (identified through ZooMS) reassemble more to modern goat than to archaeological sheep (and conversely) (figure 5), even if there are shape differences between archaeological and modern sheep ( p-value = 0.001) and goat ( p-value = 0.004). Table 2. Differences in size between modern sheep and goat, and the archaeological specimens identified by ZooMS. Pairwise 7 Wilcoxon test were done between each group. The p-values in bold are significant ( p-value < 0.05) after Bonferroni correction royalsocietypublishing.org/journal/rsos for multiple comparisons. archaeological goat archaeological sheep modern goat archaeological sheep 1 modern goat 3.9 × 10−6 1.4 × 10−7 −11 modern sheep 9.6 × 10 2.9 × 10−14 9.1 × 10−8 In addition, differences between sheep and goat are homogeneous between modern and archaeological specimens for both shape (interaction term of a 2-way Procrustes ANOVA, p-value = Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 0.463) and size ( p-value = 0.138). R. Soc. Open Sci. 10: 230672 The only difference in terms of size or shape variance is seen between modern sheep and goat, with modern sheep showing more variation than modern goat (table 3). 4. Discussion We demonstrate, for the first time, that geometric morphometrics can be an efficient alternative to molecular identification with 95.4% of the morphometric identifications being congruent with those provided by the molecular ZooMS identification. Identification of paleontological or archaeological remains is often done by actualism, defined as ‘the methodology of inferring the nature of past events by analogy with processes observable and in action in the present’ [90]. In order to fully explain the morphological diversity in the past, a better understanding of current diversity and its incorporation in the analyses is needed [91,92]. The hypothesis based on these identifications assumed that modern specimens are relevant for studying remains from the past, with negligible influence of extrinsic (e.g. temporal or geographical) variation. Modern specimens are often the only available reference of securely identified species, however, it has been demonstrated that modern references can lead to misidentifications. For example, modern dog breeds are not a good proxy to identify archaeological specimens [93]. This is also the case for other domestic species for which it is well known that direct selection over the last two centuries was intensified, followed by standardization of morphology and performance with the emergence of the breed concept [94–97]. This also applies to sheep and goat, whose breeds became genetically more and more uniformized [95], even if our morphometric results do not detect increase in diversity between our ancient and modern datasets. However, we detected morphologies that exist today that were not present in the studied archaeological dataset. This result will have to be explored further using more archaeological specimens. We found that a large majority (95.4%) of the specimens were correctly identified based on their third molar shape, suggesting that the morphometric integrity, and thus the distinction of the species, are maintained through time, at least for the third lower molar. This means that, based on their third molar shape, the differences between sheep and goat remain the same through time. On the other hand, modern specimens of both species possess larger teeth than their archaeological counterparts with a strong overlap between sheep and goat, confirming the impossibility of using size alone for species identification. This size difference between archaeological and modern specimens, although little studied, has also been suggested for cattle [98] and pigs [99]. The five misaligned specimens, those identified as sheep by ZooMS and goat by geometric morphometrics, suggest that some ancient sheep dental morphologies were more similar to the here- captured present-day goat dental morphologies. It seems important to note that the percentage of correct identification of 95.4% observed here is similar to the 93.3% percentage of correct cross- validation between modern sheep and goat obtained in a previous study [49]. At least four hypotheses could be envisaged to explain the misidentified specimens. First, even if our modern reference collection includes 521 sheep and 222 goats belonging to a range of breeds originating from Europe and North Africa, the entire existing modern diversity is certainly not represented. Furthermore, this reference collection was purposely built to study the north western Mediterranean region, and so may not be directly suitable for other geographical areas. Similarly, two waves of exotic foods are known during Iron Age [100], the Roman and medieval periods [101,102] that may have influenced the diversity we observe in our archaeological dataset. Consequently, we archaeological goat 8 archaeological sheep royalsocietypublishing.org/journal/rsos modern goat 0.10 modern sheep 0.05 PC 2 (11.61%) Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 R. Soc. Open Sci. 10: 230672 0 –0.05 –0.10 –0.2 –0.1 0 0.1 PC 1 (50.94%) Figure 4. Two first axes of the PCA showing variation between modern (dark) and archaeological (light) sheep (green) and goat (blue) identified by ZooMS. PCA for each species are available in electronic supplementary material, figure S3. (a) (b) modern sheep archaeological sheep modern goat archaeological goat archaeological goat archaeological sheep modern goat modern sheep Figure 5. Visualization of the mean shape (a) and dissimilarity network (b) between archaeological (light) and modern (dark) sheep (green) and goat (blue). Archaeological specimens are identified by ZooMS. cannot exclude that these five individuals may reflect animal translocation across regions, from areas where sheep dental morphology was similar to that of present-day goats. Second, these animals could correspond to taxa not included in our analyses—e.g. the mouflon (Ovis orientalis musimon, a feral form of sheep), the alpine (Capra ibex) or Iberian ibex (Capra pyrenaica) or the Barbary sheep (Ammotragus lervia)—whose dental morphology and geographical distributions in the past are largely unknown, and whose distinction based on collagen peptide sequences is limited. Due to the highly conserved nature of the amino acid sequence of collagen, ZooMS cannot be used to distinguish between domestic and wild specimens of the same species, and sometimes cannot distinguish between closely related taxa, as their collagen sequences are identical or near identical Table 3. Comparison of size and shape variance between modern and archaeological sheep and goat. Size variance was tested 9 with the Fligner test, with a Khi2 test statistic, and shape variance with a disparity approach (pairwise test on Procrustes royalsocietypublishing.org/journal/rsos variance, distance between each Procrustes variance). size shape Khi2 p-value distance p-value archaeological sheep/archaeological goat 2.34 0.12 1,71 × 10−3 0.26 modern sheep/modern goat 15.41 8.63 × 10−5 2,07 × 10−3 0.002 archaeological sheep/modern sheep 3.64 × 10−4 0.98 1,69 × 10−3 0.11 archaeological goat/modern goat 0.11 0.73 1,34 × 10−3 0.31 Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 R. Soc. Open Sci. 10: 230672 [80]. For example, domestic sheep, mouflon and Barbary sheep cannot be distinguished from another using ZooMS as their collagen peptide markers are the same. It is possible that these five samples belong to taxa that are closely related to domestic sheep however have dentition more similar to that of domesticated goats. This hypothesis could be tested in the future through e.g. palaeogenomics. Third, age and sex are known to have a significant impact on sheep and goat third molar morphology, but that both factors have limited impact on the ability to discriminate the two species [49]. Indeed, even if age (estimated through wear stages) affects the third lower molar shape of both species, a closer proximity between species than between age categories has been found [49]. In addition, it was demonstrated that the impact of age on the between species differentiation was mostly due to the oldest specimens (more than 8 years) [49]. Because the mis-indentitied specimens had an estimated age of 4–8 years, it seems unlikely that age caused the mismatch between the morphometric and molecular identifications. Sexual dimorphism in teeth size and shape has been identified in both species; but as for age, a greater proximity has been found between species than sex, and the sexual dimorphism has no impact on modern caprine identification [49]. Finally, even if very rare, sheep-goat hybrids exist (e.g. [103,104]). Such hybrids would likely be difficult, if not impossible, to identify using either ZooMS or GMM, as presumably the results for both analyses would simply indicate either sheep or goat, rather than a mix. Nevertheless, with a approximately 95% success rate, this newly generated data allows us to build a secure reference collection of ancient morphologies for the two species from the Neolithic to Modern periods in the North Western Mediterranean basin. This will enable non-molecular, and thus non- destructive, methods of analysis of dentition for larger zooarchaeological assemblages to improve our understanding of the husbandry history of these two species in space and time. Because this study aims only at validating the methodological approach, the archaeological interpretations require further consideration. While the number of specimens studied here did not allow detailed analyses, adding more specimens will allow further exploration of the environmental and socio-cultural factors that shaped the past agrobiodiversity and its evolution. While it is known that modern [31] and ancient goat and sheep have different diets [105], the impact of diet in tooth morphology in those species remains to be explored, but in any case did not have a major impact on the taxonomic identification. The observed size and shape differences between archaeological and modern specimens of both sheep and goats are in agreement with previous zooarchaeological results. Here we observed that archaeological specimens have on average smaller molars than their modern counterparts. Sheep and goat are known to have increased in size during Modern period [106–108]. Caprines were particularly improved upon over the last centuries [108–110], which could explain the differences noticed between archaeological and modern groups. However, despite the presence of both primitive and standardized breeds in our referential, we did not find that modern specimens were more or less diverse than the archaeological groups. The only observed variance difference is between modern sheep and goat, where modern sheep are, for both size and shape, more variable than modern goat. This result could be due to the greater number of sheep compared to goat in the modern referential. Moreover, today officially recognized sheep breeds are also more numerous than goat, with for example around 50 different French sheep breeds compared to only 15 breeds of goat [97,111]. Moreover, goats were mainly selected for the purpose of milk, while sheep have been selected for meat, milk and wool [97,111], which could explain this greater morphological diversity, and the greater number of sheep breeds compared to goat. Archaeological sheep and goat dental remains can prove difficult to identify to the species level [38,40,42,43,112,113]. This study is a new contribution toward analyses dedicated to understanding the evolution of both sheep and goat in archaeological records using non-destructive morphometric 10 identification of dental specimens. Being able to identify isolated third molars, that often remain royalsocietypublishing.org/journal/rsos unidentified in archaeozoology, allows new perspectives for further research. It is now possible to use the powerful and non-invasive quantitative tools of geometric morphometrics, via archaeophenomics [47], to explore in more depth the separate evolution of sheep and goat through time, and follow their socio-economic and cultural roles in past animal husbandry. Such studies require secure identification and quantification which, particularly for large scale assemblages, is best accomplished through integrating molecular and non-molecular methodologies. Integrative approaches that combine GMM and biomolecular methods, such as ZooMS, are in high demand, with such integration leading to clearer perceptions of domestic species evolution through time and space. Ethics. This work did not require ethical approval from a human subject or animal welfare committee. Data accessibility. The data are provided in electronic supplementary material [114]. Declaration of AI use. We have not used AI-assisted technologies in creating this article. Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 Authors’ contributions. M.J.: data curation, formal analysis, investigation, methodology, software, validation, R. Soc. Open Sci. 10: 230672 visualization, writing—original draft, writing—review and editing; K.M.M.: methodology, resources, writing— review and editing; S.V.: resources, writing—review and editing; A.N.: resources, writing—review and editing; R.S.: resources; PM.P.: resources, writing—review and editing; S.J.: resources, writing—review and editing; C.G.: resources, writing—review and editing; F.T.: resources; R.R.: resources; C.M.: visualization, writing—review and editing; A.E.: conceptualization, formal analysis, funding acquisition, methodology, project administration, software, supervision, validation, visualization, writing—original draft, writing—review and editing. All authors gave final approval for publication and agreed to be held accountable for the work performed therein. Conflict of interest declaration. We declare we have no competing interests. Funding. This project received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (Grant Agreement No. 852573). We would like to thank the Laboratori de Proteòmica CSIC-UAB at the Universitat Autònoma de Barcelona for access to the Bruker Ultraflex III MALDI-TOF-MS, and the Centres Cientifics i Tecnologics of the Universitat de Barcelona for access to the AB SCIEX 4700 MALDI-TOF-MS. We acknowledge the Service Regional de l’Archéologie of the Occitanie Region for permission for studying and destructive sampling of the archaeological material, as well as the directors of the different excavations included in this study. Acknowledgements. We also acknowledge Museum National d’Histoire Naturelle (Paris), Museum für Haustierkunde, Oniris Nantes and the Musée d’Histoire Naturelle de Nîmes for allowing to consult their collections. References 1. Richter KK, Codlin MC, Seabrook M, Warinner C. 6. Culley C et al. 2021 Collagen fingerprinting results, new problems. J. Mediterr. Archaeol. 28, 2022 A primer for ZooMS applications in traces the introduction of caprines to island 105–131. (doi:10.1558/jmea.v28i1.27503) archaeology. Proc. Natl Acad. Sci. USA Eastern Africa. R. Soc. Open Sci. 8, 202341. 12. Sjögren KG, Buckley M, Vretemark M, Axelsson 119, e2109323119. (doi:10.1073/pnas. (doi:10.1098/rsos.202341) T. 2023 Evaluating caprine remains of the 2109323119) 7. Janzen A et al. 2021 Distinguishing African Swedish Funnel Beaker culture through ZooMS. 2. Buckley M, Collins M, Thomas-Oates J, Wilson bovids using Zooarchaeology by Mass Archaeol. Anthropol. Sci. 15, 47. (doi:10.1007/ JC. 2009 Species identification by analysis of Spectrometry (ZooMS): New peptide markers s12520-023-01743-2) bone collagen using matrix-assisted laser and insights into Iron Age economies in 13. Buckley M, Kansa SW. 2011 Collagen desorption/ionisation time-of-flight mass Zambia. PLoS ONE 16, 1–36. (doi:10.1371/ fingerprinting of archaeological bone and teeth spectrometry. Rapid Commun. Mass Spectrom. journal.pone.0251061) remains from Domuztepe, South Eastern Turkey. 23, 3843–3854. (doi:10.1002/rcm) 8. Culley C, Janzen A, Brown S, Prendergast ME, Archaeol. Anthropol. Sci. 3, 271–280. (doi:10. 3. Ebsen JA, Haase K, Larsen R, Sommer DVP, Brandt Shipton C, Ndiema E, Petraglia MD, Boivin N, 1007/s12520-011-0066-z) LØ. 2019 Identifying archaeological leather – Crowther A. 2021 Iron Age hunting and herding 14. Price MD, Buckley M, Kersel MM, Rowan YM. discussing the potential of grain pattern analysis in coastal eastern Africa: ZooMS identification of 2013 Animal Management Strategies during the and zooarchaeology by mass spectrometry domesticates and wild bovids at Panga ya Saidi, Chalcolithic in the Lower Galilee: New Data from (ZooMS) through a case study involving medieval Kenya. J. Archaeol. Sci. 130, 105368. (doi:10. Marj Rabba (Israel). Paléorient 39, 183–200. shoe parts from Denmark. J. Cult. Herit. 39, 1016/j.jas.2021.105368) (doi:10.3406/paleo.2013.5527) 21–31. (doi:10.1016/j.culher.2019.04.008) 9. Taylor W et al. 2018 Early pastoral economies 15. Vaiglova P, Bogaard A, Collins M, Cavanagh W, 4. Fiddyment S et al. 2015 Animal origin of 13th- along the Ancient Silk Road: biomolecular evidence Mee C, Renard J, Lamb A, Gardeisen A, Fraser R. century uterine vellum revealed using from the Alay Valley, Kyrgyzstan. PLoS ONE 13, 2014 An integrated stable isotope study of noninvasive peptide fingerprinting. Proc. Natl. 1–19. (doi:10.1371/journal.pone.0205646) plants and animals from Kouphovouno, Acad. Sci. U. S. A. 112, 15 066–15 071. (doi:10. 10. Taylor WTT et al. 2021 Evidence for early southern Greece: A new look at Neolithic 1073/pnas.1512264112) dispersal of domestic sheep into Central Asia. farming. J. Archaeol. Sci. 42, 201–215. (doi:10. 5. Coutu AN, Taurozzi AJ, Mackie M, Jensen TZT, Nat. Hum. Behav. 5, 1169–1179. (doi:10.1038/ 1016/j.jas.2013.10.023) Collins MJ, Sealy J. 2021 Palaeoproteomics s41562-021-01083-y) 16. Pilaar Birch SE, Scheu A, Buckley M, Çakırlar C. confirm earliest domesticated sheep in southern 11. Martins H, Xavier Oms F, Pereira L, Pike AWG, 2019 Combined osteomorphological, isotopic, Africa ca. 2000 BP. Sci. Rep. 11, 1–11. (doi:10. Rowsell K, Zilhão J. 2015 Radiocarbon dating aDNA, and ZooMS analyses of sheep and goat 1038/s41598-021-85756-8) the beginning of the Neolithic in Iberia: New remains from Neolithic Ulucak, Turkey. Archaeol. Anthropol. Sci. 11, 1669–1681. (doi:10.1007/ 31. Balasse M, Ambrose SH. 2005 Distinguishing Capra hircus dental remains. J. Archaeol. Sci. 13, 11 s12520-018-0624-8) sheep and goats using dental morphology and 579–595. (doi:10.1016/0305-4403(86)90042-7) 17. Prendergast ME, Janzen A, Buckley M, Grillo KM. stable carbon isotopes in C4 grassland 46. Lyman RL. 1984 Bone density and differential royalsocietypublishing.org/journal/rsos 2019 Sorting the sheep from the goats in the environments. J. Archaeol. Sci. 32, 691–702. survivorship of fossil classes. J. Anthropol. Pastoral Neolithic: morphological and (doi:10.1016/j.jas.2004.11.013) Archaeol. 3, 259–299. (doi:10.1016/0278- biomolecular approaches at Luxmanda, 32. De Serre O. 1600 Le théatre d’agriculture et 4165(84)90004-7) Tanzania. Archaeol. Anthropol. Sci. 11, ménage des champs. Arles: Actes Sud. 47. Evin A, Bouby L, Bonhomme V, Jeanty A, 3047–3062. (doi:10.1007/s12520-018-0737-0) 33. Bourrier T. 1897 Les industries des abattoirs. Jeanjean M, Terral J-F. 2022 Archaeophenomics 18. Gron KJ, Rowley-Conwy P, Jensen TZT, Taurozzi Paris: Librairie J.-B. Baillère et Fils. of ancient domestic plants and animals using AJ, Marciniak A. 2020 Separating caprine 34. Helmer D, Vigne J-D. 2004 La gestion des geometric morphometrics : a review. Peer (Capra/Ovis) distal tibiae: A case study from the cheptels de caprinés au Néolithique dans le Community J. 2, e27. (doi:10.24072/pcjournal. Polish Neolithic. Int. J. Osteoarchaeol. 30, Midi de la France. In Approches fonctionnelles en 126) 170–179. (doi:10.1002/oa.2844) préhistoire (eds P Bodu, C Constantin), pp. 48. Rohlf FJ, Marcus LF. 1993 A Revolution in 19. Ananyevskaya E, Buckley M, Pal Chowdhury M, 397–407. Paris: Société Préhistorique Française Morphometrics. Trends Ecol. Evol. 8, 129–132. Tabaldiev K, Motuzaite Matuzeviciute G. 2021 Édition. (doi:10.2307/20047487) Specialized wool production economy of 35. Helmer D, Gourichon L, Sidi Maamar H, Vigne J. 49. Jeanjean M et al. 2022 Sorting the flock: Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 prehistoric farmstead of Chap I in the highlands 2005 L’élevage des caprinés néolithiques dans Quantitative identification of sheep and goat R. Soc. Open Sci. 10: 230672 of Central Tian Shan (Kyrgyzstan). le sud-est de la France : saisonnalité des from isolated third lower molars and mandibles Int. J. Osteoarchaeol. 31, 18–28. (doi:10.1002/ abattages, relations entre grottes-bergeries et through geometric morphometrics. J. Archaeol. oa.2921) sites de plein air. Anthropozoologica 40, Sci. 141, 105580. (doi:10.1016/j.jas.2022. 20. Bläuer A, Solala H, Kinnunen J, Rannamäe E. 167–189. 105580) 2022 Multidisciplinary History of Goats in 36. Blaise E. 2009 Economie animale et gestion des 50. Maufras O, Hernandez J, Rochette M, Thomas B. Finland: A Comparative Approach. Heritage 5, troupeaux au Néolithique final en Provence, 2015 Genèse, évolution et désertion de 1947–1959. (doi:10.3390/heritage5030101) Université de Provence - Aix-Marseille I. Missignac (Aimargues, Gard), villa des Ve-XIIIe 21. Boyazoglu J. 2002 Livestock research and 37. Boessneck J. 1971 Osteological Differences siècles. Archéologie du Midi Médiéval. environmental sustainability with special between Sheep (Ovis aries Linné) and Goat 51. Bardot-Cambot A, Fabre M, Forest V. 2017 Étude reference to the Mediterranean basin. Small (Capra hircus Linné). In Science in archaeology: a archéozoologique : Ostéologie, conchyliologie. In Ruminant Research 45, 193–200. (doi:10.1016/ survey of progress and research (eds D Lattes, castelle, pahon, pinède. Se déplacer, S0921-4488(02)00100-1)) Brotherweel, E Higgs), pp. 331–358. London: cultiver et mourir dans la campagne lattoise de 22. Galaty J, Johnson D. 1990 Pastoral systems. Thames and Hudson. l’Antiquité à nos jours (ed. M Compan), pp. New York, NY: Guilford Press. 38. Zeder M, Lapham HA. 2010 Assessing the 256–274. Montpellier: Rapport final d’opération 23. Enne G, Zucca C, Montoldi A, Noe L. 2004 The reliability of criteria used to identify postcranial – Fouille préventive, Service Régional de role of grazing in agropastoral systems in the bones in sheep, Ovis, and goats, Capra. l’Archéologie Occitanie. Mediterranean region and their environmental J. Archaeol. Sci. 37, 2887–2905. (doi:10.1016/j. 52. Payne S. 1973 Kill-off patterns in sheep and sustainability. Adv. Geoecol. 37, 29–46. jas.2010.06.032) goats: the mandibles from Aşvan Kale. 24. Durmus M, Agossou DJ, Koluman N. 2019 39. Prummel W, Frisch HJ. 1986 A guide for the Anatolian studies 23, 281–303. Sustainability of Small Ruminant Production in distinction of species, sex and body side in 53. Bordas A, Pou R, Parpal A, Martín A. 1994 Mediterranean Region. J. Environ. Sci. Eng. B 8, bones of sheep and goat. J. Archaeol. Sci. 13, Excavacions arqueològiques 1991–1992 a la 241–248. (doi:10.17265/2162-5263/2019.06. 567–577. (doi:10.1016/0305-4403(86)90041-5) Bòbila Madurell - Mas Duran (Sant Quirze del 005) 40. Halstead P, Collins P, Isaakidou V. 2002 Sorting Vallès, Vallès Occidental). Trib. d’Arqueologia 25. Marandet MC. 2021 L’homme et l’animal dans the sheep from the goats: Morphological 1992–1993, 31–47. les sociétés méditerranéennes. Perpignan, distinctions between the mandibles and 54. Abad ÒE, Garra AM, Bieto ET, Aixalà AV, Oller France: Presses Universitaires De Perpignan. mandibular teeth of adult Ovis and Capra. NA. 2012 Cantorella (Maldà, Urgell), un nou 26. Nieto-Espinet A, Huet T, Trentacoste A, J. Archaeol. Sci. 29, 545–553. (doi:10.1006/jasc. assentament a l’aire lliure del neolític final - Guimarães S, Orengo H, Valenzuela-Lamas S. 2001.0777) calcolític i del bronze ple a la vall del Corb. Trib. 2021 Resilience and livestock adaptations to 41. Helmer D. 2000 Discrimination des genres Ovis d’arqueologia 2011-2012, 129–172. demographic growth and technological change: et Capra a l’aide des premolaires inferieures 3 et 55. Roig J, Coll JM, Gibaja JF, Chambon P, Villar V, A diachronic perspective from the Late Bronze 4 et interpretation des ages d’abattage: Ruiz J, Terradas X, Subirà ME. 2010 La Age to Late Antiquity in NE Iberia. PLoS ONE l’example de dikili tash (Grece). necrópolis de Can Gambús-1 (Sabadell, 16, 1–47. (doi:10.1371/journal.pone.0246201) Anthropozoologica 31, 29–38. Barcelona) nuevos conocimientos sobre las 27. Valenzuela-Lamas S, Albarella U. 2017 Animal 42. Payne S. 1985 Morphological distinctions prácticas funerarias durante el Neolítico husbandry across the Western Roman Empire: between the mandibular teeth of young sheep, medio en el noreste de la Península Ibérica. Changes and continuities. Eur. J. Archaeol. 20, Ovis, and goats, Capra. J. Archaeol. Sci. Trab. Prehist. 67, 59–84. (doi:10.3989/tp. 402–415. (doi:10.1017/eaa.2017.22) 12, 139–147. (doi:10.1016/0305-4403(85) 2010.10031) 28. Halstead P. 1996 Pastoralism or household 90058-5) 56. Gailledrat E et al. 2012 Rapport de fouille Pech herding? problems of scale and specialization in 43. Zeder M, Pilaar SE. 2010 Assessing the reliability Maho (Sigean, Aude). early Greek animal husbandry. World Archaeol. 28, of criteria used to identify mandibles and 57. Alonso N, Junyent E, Lafuente Á, López JB, 20–42. (doi:10.1080/00438243.1996.9980329) mandibular teeth in sheep, Ovis, and goats, Tartera E. 2000 La fortaleza de Arbeca. El 29. Trentacoste A, Nieto-Espinet A, Guimarães Capra. J. Archaeol. Sci. 37, 225–242. (doi:10. proyecto Vilars 2000. Investigación, recuperación Chiarelli S, Valenzuela-Lamas S. 2022 Systems 1016/j.jas.2009.10.002) y socialización del conocimiento y del change: Investigating climatic and 44. Gillis R, Chaix L, Vigne JD. 2011 An assessment patrimonio. Trab. Prehist. 57, 161–173. (doi:10. environmental impacts on livestock production of morphological criteria for discriminating 3989/tp.2000.v57.i2.255) in lowland Italy between the Bronze Age and sheep and goat mandibles on a large prehistoric 58. Pou J, Sanmartí J, Santacana J. 1995 La Late Antiquity (c. 1700 BC – AD 700). Quat. Int. archaeological assemblage (Kerma, Sudan). reconstrucció del poblat ibèric d’Alorda Park o 662, 26–36. (doi:10.1016/j.quaint.2022.11.005) J. Archaeol. Sci. 38, 2324–2339. (doi:10.1016/j. de les Toixoneres (Calafell, Baix Penedès). Trib. 30. Shackleton DM, Shank CC. 1984 A Review of the jas.2011.04.012) d’Arqueologia 1993–1994, 51–62. Social Behavior of Feral and Wild Sheep and 45. Grine FE, Fosse G, Krause DW, Jungers WL. 1986 59. Aquilué X, Castanyer P, Santos M, Tremoleda J. Goats. J. Anim. Sci. 58, 500–509. (doi:10.2527/ Analysis of enamel ultrastructure in 2009 Resultats de les darreres intervencions jas1984.582500x) archaeology: The identification of Ovis aries and arqueològiques a la Neàpolis de la ciutat grega d’Empòrion (Empúries, l’Escala, Alt Empordà). superimposition of landmarks. Syst. Zool. 39, sheep and goat. Dryad Digital Repository. 12 Trib. d’Arqueologia 2009, 121–147. 40–59. (doi:10.2307/2992207) (doi:10.5061/dryad.fj6q5741m) 60. Martin A, Miret J, Blanch RM, Aliaga S, Enrich R, 74. Goodall CR. 1995 Procrustes methods in the 90. Rudwick MJS. 1971 Uniformity and progression: royalsocietypublishing.org/journal/rsos Colomer S, Albizuri S, Bosch J. 1988 Campanya statistical analysis of shape revisited. In Current reflections on the structure of geological theory d’excavacions arqueològiques 1987–88 al issues in statistical shape analysis (eds KV in the age of Lyell. In Perspectives in the history jaciment de la Bòbila Madurell-Can Feu (Sant Mardia, CA Gill), pp. 18–33. Leeds: University of of science and technology (ed. D Roller), pp. Quirze del Vallès, Vallès occidental). Arraona Leeds Press. 209–227. Norman: University of Oklahoma Rev. d’història 3, 9–23. 75. Baylac M, Friess M. 2005 Fourier descriptors, Press. 61. Julià M, Lázaro P, Lorencio C, Loriente A, Oliver Procrustes superimposition, and data 91. Kitts DB. 1977 The structure of geology. Dallas: A, Plens M, Puig F. 1989 Excavacions d’urgència dimensionality: an example of cranial shape Southern Methodist University Press. a l’antic Portal de Magdalena. Lleida (Segrià). analysis in modern human populations. In 92. Pobiner BL, Braun DR. 2005 Applying Actualism: Excavacions Arqueol. d’urgència a les Comarq. Modern morphometrics in physical anthropology Considerations for Future. J. Taphon. 3, 57–66. Lleida 9, 203–226. (ed. DE Slice), pp. 145–165. Boston: Springer. 93. Brassard C et al. 2022 Unexpected 62. Maufras O, Mercier C. 2002 Habitat et terroir du 76. Kovarovic K, Aiello L, Cardini A, Lockwood CA. morphological diversity in ancient dogs IVe au XIIe s. à Saint- Gilles-le-Vieux 2011 Discriminant function analyses in compared to modern relatives. Proc. R. Soc. B (Aimargues, Gard). In Archéologie du TGV archaeology: Are classification rates too good to Biol. Sci. 289, 20220147. (doi:10.1098/rspb. Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 méditerranée. Fiche de syntèse, tome be true? J. Archaeol. Sci. 38, 3006–3018. 2022.0147) R. Soc. Open Sci. 10: 230672 3. Antiquité, moyen âge, epoque moderne, pp. (doi:10.1016/j.jas.2011.06.028) 94. Taberlet P, Valentini A, Rezaei HR, Naderi S, 945–972. Lattes: Association pour la Recherche 77. Evin A, Cucchi T, Cardini A, Strand Vidarsdottir Pompanon F, Negrini R, Ajmone-Marsan P. 2008 Archéologique en Languedoc Oriental. U, Larson G, Dobney K, Strand U, Larson G, Are cattle, sheep, and goats endangered 63. Maufras O, Hernandez J, Rochette M, Thomas B. Dobney K. 2013 The long and winding road : species? Mol. Ecol. 17, 275–284. (doi:10.1111/j. 2018 Aimargues - Madame - Saint-Gilles le identifying pig domestication through molar 1365-294X.2007.03475.x) Vieux: Missignac, villa médiévale et ses abords size and shape. J. Archaeol. Sci. 40, 735–743. 95. Taberlet P, Coissac E, Pansu J, Pompanon F. (Ve - XIIIe s.). (doi:10.1016/j.jas.2012.08.005) 2011 Conservation genetics of cattle, sheep, and 64. Henry E. 2016 Montpellier, CNR rue de la Vieille 78. Evin A et al. 2015 Unravelling the complexity of goats. Comptes Rendus - Biol. 334, 247–254. Coraterie, Rapport final d’opération. domestication: A case study using (doi:10.1016/j.crvi.2010.12.007) 65. Romeu PG, Masachs JM. 1968 Hallazgos morphometrics and ancient DNA analyses of 96. Denis B. 2010 Races bovins: histoire, aptitudes, prehistóricos en los alrededores de La Lacuna. archaeological pigs from Romania. Philos. situation actuelle. Chaumont: Castor & Pollux. Castillo de Vilademager. Empúries Rev. món Trans. R. Soc. B Biol. Sci. 370, 1–8. (doi:10. 97. Denis B, Vaissaire J-P. 2021 Les races d’animaux clàssic i Antig. Tardana 30, 207–213. 1098/rstb.2013.0616) domestiques en France. Etude générale et 66. Morán Álvarez M, Payà Mercè FX. 2021 Fonts 79. Strohalm M, Kavan D, Novák P, Volný M, inventaire. Paris: Chartes Delachaux et Niestlé. documentals i arqueològuiques per al Havlíček V. 2010 MMass 3: A cross-platform 98. Albarella U. 1997 Shape variation of cattle coneixement de la Cuirassa jueva de Lleida. software environment for precise analysis of metapodials: age, sex or breed? Some examples Urtx. Rev. d’Humanitats l’Urgell 35, 133–177. mass spectrometric data. Anal. Chem. 82, from Medieval and Postmedieval sites. 67. Nieto-Espinet A, Valenzuela-Lamas S. 2019 4648–4651. (doi:10.1021/ac100818g) Anthropozoologica 25–26, 37–48. Muerte o conversión. Del bienestar a la 80. Buckley M, Whitcher Kansa S, Howard S, 99. White S. 2011 From globalized pig breeds to hambruna en el barrio judío de la Cuirassa de Campbell S, Thomas-Oates J, Collins M. 2010 capitalist pigs: A study in animal cultures and Lleida (siglos xii-xiv). In comunicació presentada Distinguishing between archaeological sheep evolutionary history. Environ. Hist. Durh. N. C. en workshop. Minorías y transiciones en la edad and goat bones using a single collagen peptide. 16, 94–120. (doi:10.1093/envhis/emq143) media a través de la zooarqueologia (s. x–xv). J. Archaeol. Sci. 37, 13–20. (doi:10.1016/j.jas. 100. Gaastra JS. 2014 Shipping sheep or creating 28 i 29 de novembre del 2019. Universitat de 2009.08.020) cattle: Domesticate size changes with Greek Granada, Universidad de Granada. 81. Kirby DP, Buckley M, Promise E, Trauger SA, colonisation in Magna Graecia. J. Archaeol. Sci. 68. Rohlf FJ. 2006 Tps Dig v2.32, Ecology and Holdcraft TR. 2013 Identification of collagen- 52, 483–496. (doi:10.1016/j.jas.2014.08.030) Evolution. based materials in cultural heritage. Analyst 101. Livarda A. 2011 Spicing up life in northwestern 69. Perez SI, Bernal V, Gonzalez PN. 2006 138, 4849–4858. (doi:10.1039/c3an00925d) Europe: Exotic food plant imports in the Roman Differences between sliding semi-landmark 82. Welker F, Soressi M, Rendu W, Hublin JJ, Collins and medieval world. Veg. Hist. Archaeobot. 20, methods in geometric morphometrics, with an M. 2015 Using ZooMS to identify fragmentary 143–164. (doi:10.1007/s00334-010-0273-z) application to human craniofacial and dental bone from the Late Middle/Early Upper 102. Aubaile F. 2012 Pathways of diffusion of some variation. J. Anat. 208, 769–784. (doi:10.1111/j. Palaeolithic sequence of Les Cottés, France. plants and animals between Asia and the 1469-7580.2006.00576.x) J. Archaeol. Sci. 54, 279–286. (doi:10.1016/j.jas. Mediterranean region. Rev. d’ethnoécologie. 1. 70. Sampson P, Bookstein F, Sheenan F, Bolson E. 2014.12.010) 1–26. (doi:10.4000/ethnoecologie.714) 1996 Eigenshape analysis of left ventricular 83. Adams DC, Collyer M, Kaliontzopoulou A. 2020 103. Bunch TD, Foote WC, Spillett JJ. 1976 Sheep- outlines from contrast ventriculograms. In Geomorph: Software for geometric goat hybrid karyotypes. Theriogenology Advances in morphometrics (eds L Marcus, M morphometric analyses. 6, 379–385. (doi:10.1016/0093-691X(76) Corti, L Loy, G Naylar, D Slice), pp. 211–233. 84. Wickham H. 2016 ggplot2: Elegant Graphics for 90104-7) New-York: NATO ASI Series A: Life Sciences, Data Analysis. 104. Mine OM, Kedikilwe K, Ndebele RT, Nsoso SJ. Plenum. 85. Collyer M, Adams DC. 2018 RRPP: An R package 2000 Sheep-goat hybrid born under natural 71. Sheets HD, Kim K, Mitchell CE, Sheets DH, Kim for fitting linear models to high-dimensional conditions. Small Rumin. Res. 37, 141–145. K, Mitchell CE. 2004 A combined landmark and data using residual randomization. (doi:10.1016/S0921-4488(99)00146-7) outline-based approach to ontogenetic shape 86. Collyer M, Adams DC. 2019 RRPP: Linear Model 105. Jiménez-Manchón S. 2020 Pratiques d’élevage change in the Ordovician trilobite Triarthrus Evaluation with Randomized Residuals in a entre l’Empordà et le languedoc à l’âge du Fer: becki. In Morphometrics (ed. AM Elewa), pp. Permutation Procedure. R package version 0.4.0. archéozoologie, alimentation animale et 67–81. New-York: Springer. (doi:10.1007/978-3- 87. R Core Team. 2022 R: a language and saisonnalité, Université Paul Valéry - Montpellier III. 662-08865-4) environment for statistical computing. 106. Robin O, Clavel B. 2018 The diversity evolution 72. Gower JC. 1975 Generalized procrustes analysis. 88. Posit team. 2023 RStudio: Integrated of sheep morphology in French Psychometrika 40, 33–51. (doi:10.1007/ Development Environment for R. zooarchaeological remains from the 9th to the BF02291478) 89. Jeanjean M et al. 2023 Data from: ZooMS 19th century: Analysis of pastoral strategy. 73. Rohlf FJ, Slice D. 1990 Extensions of the Spectra from: ZooMS confirms geometric J. Archaeol. Sci. 99, 55–65. (doi:10.1016/j.jas. procrustes method for the optimal morphometrics species identification of ancient 2018.08.017) 107. Clavel B. 2001 L’animal dans l’alimentation Fiches d’ostéologie Anim. pour l’archéologie. Conwy, D Serjeantson, P Halstead), pp. 50–82. 13 médiévale et moderne en France du Nord (XIIIe CRA-CNRS 3, 36. Oxford: Oxbow Books. (doi:10.2307/j. - XVIIe siècles). Rev. archéologique Picardie. 110. Davis SJM, Gonçalves MJ, Gabriel S. 2008 ctt1s4751q.14) royalsocietypublishing.org/journal/rsos Numéro spécial 19, 9–204. (doi:10.3406/pica. Animal remains from a Moslem period (12 th/ 113. Salvagno L, Albarella U. 2017 A morphometric 2001.3065) 13 th century AD) lixeira (garbage dump) in system to distinguish sheep and goat 108. Grau-Sologestoa I. 2015 Livestock management Silves, Algarve, Portugal. Rev. Port. Arqueol. 11, postcranial bones. PLoS One 12, 1–37. (doi:10. in Spain from Roman to post-medieval times: A 183–258. 1371/journal.pone.0178543) biometrical analysis of cattle, sheep/goat and 111. Babo D. 2000 Races ovines et caprines 114. Jeanjean M et al. 2023 Supplementary material pig. J. Archaeol. Sci. 54, 123–134. (doi:10.1016/ françaises. 1ère. Paris: Editions France Agricole. from: ZooMS confirms geometric morphometrics j.jas.2014.11.038) 112. Davis SJM. 2017 Towards a metrical species identification of ancient sheep and 109. Audoin-Rouzeau F. 1991 La taille du mouton en distinction between sheep and goat astragali. goat. Figshare. (doi:10.6084/m9.figshare.c. Europe de l’Antiquité aux Temps Modernes. In Economic zooarchaeology (eds P Rowley- 6824155) Downloaded from https://royalsocietypublishing.org/ on 08 November 2023 R. Soc. Open Sci. 10: 230672