ArticlePDF Available

Abstract and Figures

Domestication of wild cattle, sheep, and pigs began a process of body size diminution. In most of Western Europe this process continued across prehistory and was not reversed until the Roman period. However, in Italy, an increase in livestock body size occurred during the Iron Age, earlier than the Western provinces. In order to better understand the nature and timing of this early increase in animal size, this paper presents a detailed regional study of taxonomic abundance and biometric data from zooarchaeological assemblages recovered from the Po and Venetian–Friulian Plains in northern Italy. Our results demonstrate a high level of regionality in the choice of species exploited, with husbandry systems focused on different domesticates, as well as regional differences in animal size. However, despite significant variation in species frequencies, settlement structure, and epigraphic tradition, all areas with sufficient data demonstrate similar significant changes in livestock body size. Cattle and sheep increased incrementally in size prior to the Roman conquest in all regions considered; surprisingly, pigs continued to decrease in size throughout later prehistory. The incremental pace and pan-regional character of the size change in cattle and sheep suggests an internally motivated phenomenon rather than herd replacement with a new larger population, as might follow colonisation or conquest. The divergence in size trends for bovids and suids suggests a noteworthy change in cattle and sheep herding practices during the Iron Age or final centuries of the Bronze Age, in contrast with greater continuity in pig management. Our analysis provides a thorough zooarchaeological synthesis for northern Italy and, for the first time, demonstrates that both cattle and sheep increased in size outside of Roman territory well before the conquest of this area. This study offers a basis for future chemical analyses (DNA, isotopes), which will further investigate the cause(s) of livestock size changes in northern Italy.
Content may be subject to copyright.
Pre-Roman improvements to agricultural
production: Evidence from livestock
husbandry in late prehistoric Italy
Angela TrentacosteID
*, Ariadna Nieto-Espinet
, Silvia Valenzuela-Lamas
1Institute of Archaeology, University of Oxford, Oxford, United Kingdom, 2Consejo Superior de
Investigaciones Cientı
´ficas (CISC), Institucio
´Milài Fontanals, Archaeology of Social Dynamics, Barcelona,
These authors contributed equally to this work.
Domestication of wild cattle, sheep, and pigs began a process of body size diminution. In
most of Western Europe this process continued across prehistory and was not reversed
until the Roman period. However, in Italy, an increase in livestock body size occurred during
the Iron Age, earlier than the Western provinces. In order to better understand the nature
and timing of this early increase in animal size, this paper presents a detailed regional study
of taxonomic abundance and biometric data from zooarchaeological assemblages recov-
ered from the Po and Venetian–Friulian Plains in northern Italy. Our results demonstrate a
high level of regionality in the choice of species exploited, with husbandry systems focused
on different domesticates, as well as regional differences in animal size. However, despite
significant variation in species frequencies, settlement structure, and epigraphic tradition, all
areas with sufficient data demonstrate similar significant changes in livestock body size.
Cattle and sheep increased incrementally in size prior to the Roman conquest in all regions
considered; surprisingly, pigs continued to decrease in size throughout later prehistory. The
incremental pace and pan-regional character of the size change in cattle and sheep sug-
gests an internally motivated phenomenon rather than herd replacement with a new larger
population, as might follow colonisation or conquest. The divergence in size trends for
bovids and suids suggests a noteworthy change in cattle and sheep herding practices during
the Iron Age or final centuries of the Bronze Age, in contrast with greater continuity in pig
management. Our analysis provides a thorough zooarchaeological synthesis for northern
Italy and, for the first time, demonstrates that both cattle and sheep increased in size outside
of Roman territory well before the conquest of this area. This study offers a basis for future
chemical analyses (DNA, isotopes), which will further investigate the cause(s) of livestock
size changes in northern Italy.
A reduction in body size is widely recognised consequence of animal domestication [13].
Although the initial timing of morphological changes and of the selective aims that led to their
PLOS ONE | December 31, 2018 1 / 33
Citation: Trentacoste A, Nieto-Espinet A,
Valenzuela-Lamas S (2018) Pre-Roman
improvements to agricultural production: Evidence
from livestock husbandry in late prehistoric Italy.
PLoS ONE 13(12): e0208109.
Editor: Peter F. Biehl, University at Buffalo - The
State University of New York, UNITED STATES
Received: August 22, 2018
Accepted: November 12, 2018
Published: December 31, 2018
Copyright: ©2018 Trentacoste et al. This is an
open access article distributed under the terms of
the Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: Biometric standards
for cattle are available from doi: 10.13140/RG.2.2.
13512.78081. All other relevant data are within the
manuscript and its Supporting Information files.
Funding: This work was financially supported by
the ERC-Starting Grant ZooMWest –
Zooarchaeology and Mobility in the Western
Mediterranean: Husbandry production from the
Late Bronze Age to the Late Antiquity (award
number 716298), funded by the European
Research Council Agency (ERCEA) under the
emergence are still matters of debate [4,5], this decrease in the livestock size proved to be a
persistent and progressive trend. Across Western Europe, domestic animals, and particularly
cattle, continued to decrease in size between the Neolithic and Iron Age [613]. Various expla-
nations for this phenomenon have been put forward, including preference for smaller more
manageable animals [14], and intensification in herding strategies through sub-adult breeding
[15]. Regardless of the origin of this size decrease, the Roman Empire had a significant impact
on its trajectory.
Roman conquest brought about an end to the Iron Age and significant changes to the social
and economic organisation of Western Europe [1618], including to animal farming strategies
[12]. New methods of production are evident in changes to the species exploited and–breaking
with millennia of progressive size diminution–a significant increase in the size of livestock.
Although changes were not uniform [19] and exceptions occur [2021], an increase in animal
size is visible across conquered territories: France [2224], Belgium [25], the Netherlands [26],
Germany [27,28], Switzerland [2931], Britain [32], and Spain [12,33]. This transformation
of livestock husbandry is thought to result from the introduction of different forms of animal
production [21,34].
Uniquely in Western Europe, significant increases in animal size and changes to livestock
frequencies pre-date the Roman conquest of northern and central Italy by centuries [7,35,36].
High percentages of pig remains and growth in animal body size are apparent in Etruscan cit-
ies as well as Rome and its environs during the first millennium BC [35,3739]. These trends
continued to intensify during the Imperial period [40], by which point they had become a hall-
mark of ‘Romanisation’ in other parts of the Empire [19]. Within north/central Italy, these
characteristic developments are linked to the rise of urban Etruscan and Roman culture, as
marginal areas of Italy do not consistently demonstrate similar changes [20,35,41]. Urbanism
is argued to be a primary force in catalysing changes to species frequencies, due to the func-
tional challenges of provisioning cities with protein [37,42,43] and the socio-economic bene-
fits of producing surplus animals [35]. Thus, the ‘Roman’ high-pig pattern became established
during a period of increasing urbanisation during the first millennium BC, but do changes in
animal size have the same origin?
Compared to species representation, the evolution of animal size in late prehistoric Italy is
less understood. Because livestock varied significantly in stature across the Italian peninsula [7,
41,42,44,45], analyses must be conducted on a sub-regional scale in order to establish useful
patterns; a larger frame of analysis will convolute diachronic trends because the regional distri-
bution of sites is not constant through time. A few late Bronze/early Iron Age sites have pro-
duced livestock with withers’ heights taller than those recorded in for earlier phases [45], but
without detailed understanding of regional patterns of size evolution, the significance of these
examples remains unclear: do they represent the origin of an accelerating trend towards larger
animals that culminates in the Roman period? Or do they only appear large because compara-
tive data are drawn from a different geographic area with smaller animals? In order to establish
the origin and potential catalysts for animal size change, higher-resolution understanding of
its development is needed.
This paper examines frequencies of domestic animals and development of livestock size within
northern Italy in order to better understand the evolution of animal husbandry strategies and
the socio-economic and ecological factors that impacted it. Northern Italy provides an ideal
setting to pursue this line of inquiry. The region demonstrates significant and well docu-
mented changes in husbandry regimes across later prehistory [4550], including the
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 2 / 33
direction of SVL. The funders had no role in study
design, data collection and analysis, decision to
publish, or preparation of the manuscript.
Competing interests: The authors have declared
that no competing interests exist.
emergence of the recognised ‘urban Etruscan/Roman’ trend for larger animals and abundant
pigs [7,35,51,52]. These developments took place in an interconnected and climatically com-
parable geographic context. The Po Valley, Italy’s largest alluvial basin, joins the adjacent
Venetian–Friulian Plain to form a unified landscape across most of the region. Within this
area, local conditions would have varied dependant on land use, forest/vegetation cover, and
fluvial networks, but the territory is subject to a similar climate (Ko¨ppen–Geiger climate zone
Cfa), albeit with slightly cooler temperatures and greater precipitation on the higher edges of
the valley (Cfb) [53]. Limiting the study area to a well-connected region with comparable cli-
matic constraints allows for better assessment of the role of socio-economic and cultural fac-
tors in shaping husbandry strategies.
Within this region, we compare five study areas (Fig 1) with distinct archaeological trajecto-
ries. In anticipation of subsequent research on the Roman transition, here we focus on later
prehistory, specifically the Middle Bronze Age to late Iron Age, c. 1650–150 BC [5457]. While
the process was not linear or continuous [56], this time frame encompasses a period of increas-
ing social complexity and regionalisation, and the transition from diffuse villages to cities in
the southern and eastern Po Valley [54,57]. Of particular significance to this study is the devel-
opment of urban Etruscan culture in the South study area [58].
We aim to address three questions: How does livestock representation change through
time? When/where do larger animals emerge? And how do changes in species representation
relate to developments in animal size? These analyses will provide new data on the geographic
Fig 1. Map of sites and study regions. See S1 Table for details. Terrain data available from the U.S. Geological Survey.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 3 / 33
distribution of husbandry regimes and on the nature of diachronic change (e.g. rapid/progres-
sive). This evidence allows us to establish whether Etruscan/Roman animal-exploitation pat-
terns (abundant pigs, large animals) are indeed unique to these cultures, and to what extent
urban settlement structures were perquisite for improvements in livestock. In light of current
evidence for animal management regimes particular to urban Etruscan and Roman culture [7,
35,39,51,52], we hypothesise that changes in species representation and growth in animal
size reflect a new agricultural strategy related to urban settlement networks, and anticipate
greater continuity in other, non-urban parts of the study area.
Archaeological context
Broadly considered, later prehistory saw demographic growth alongside population consolida-
tion, social hierarchisation, and an intensification of production/trade, although these phe-
nomena were neither linear nor homogenous.
The Early Bronze Age (c. 2300–1700 BC) of the Po Valley was characterised by pile-dwell-
ing ‘palafitte’ villages, which represented part of a larger phenomenon of lake-side settlement
in Switzerland and Central-Eastern Europe more generally [59]. During the Middle Bronze
Age (c. 1700–1350 BC) the number of palafitte villages near lake basins declined, while a new
form of settlement–terramare–spread along fluvial courses of the Po Valley [56,59,60]. Terra-
mare villages used structural elements similar to palafitte, but were enclosed by moat and
earthen rampart that functioned as part of the site’s irrigation network, as well as serving a
defensive function [61]. Generally 1–2 ha in size, terramare displayed a high degree of internal
organisation. Their dense distribution and use of complex irrigation and water management
systems points significant population growth [56] and an intense exploitation of the territory
[62,63]. Significant tracks of the region’s forests were cleared, especially in the southern Po
Plain, to supply building material and create space for agriculture and grazing [6369]. Friuli
also developed a more complex settlement pattern, but comparable with Istria, with fortified
castellieri sites on hills [70].
Agriculture was based on cultivation of cereals and, to a lesser extent, legumes [63,71,72].
Millets were introduced as cultivars [73,74]. This diversification of crops allowed the possibil-
ity of more than one annual harvest. Ploughs were used to prepare soil for planting, and mate-
rial evidence for wooden ploughs in northern Italy dates back to Early Bronze Age [75]. Crop
rotation, manuring, and fallowing were probably used to maintain or improve field productiv-
ity [76,77]. The organisation of fields is difficult to reconstruct, but archaeobotanical evidence
suggests a division of land between garden plots and extensive fields for cereals [77]; a similar
land-use strategy has been proposed for Bronze Age agriculture in Switzerland [78,79]. The
diffusion of metal agricultural tools improved farming practices [75,80], and the expansion of
metallurgy may also have encouraged seasonal exploitation of upland pastures near metal
sources [81]. Animals kept within settlements during cold months were foddered with twigs,
grasses, and herbs [82].
The Recent Bronze Age (c. 1350–1150 BC) saw further changes in settlement patterns.
Some sites were abandoned, while others increase in size; certain terramare villages reached an
area of 15–20 ha [56,59]. The development of organised settlements on this scale, alongside
greater specialisation in metal production, and the distribution of status items in necropoli
suggest more hierarchic forms of social organisation than visible in previous centuries [83,84].
At the end of the Recent Bronze Age terramare-palafitte culture collapsed. Settlements were
abandoned, leading to an almost complete depopulation of part of the southern Po Plain [56].
Aridification of environments already subject to deforestation and intensive cultivation appear
to have had a significant role [62,63,85]; movement towards a more hierarchal society also
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 4 / 33
may have caused a social crisis, exacerbated by environmental degradation [56]. While a large
part of the southern plain was depopulated, the lowlands southeast of Verona and northern
hills display greater settlement continuity [86,87], and during the subsequent Final Bronze
Age this area developed into the economic fulcrum of Po Valley. Settlement density in the Fri-
ulian lowlands increased before sharply falling [88].
The Final Bronze Age (c. 1150–950 BC) saw population consolidation along rivers and
communication routes. Settlements developed into the first large clusters with proto-urban
characteristics, the most important of which was Frattesina [89], which demonstrated a high
level of complexity in craft production with the import of raw materials from as far as Egypt
and the Baltic [90]. Shared forms of material culture and funerary practice (cremation) associ-
ated with proto-Villanovian culture spread across the peninsula. From the eighth century BC,
Iron Age material culture developed a strong regional character. Distinct archaeological cul-
tures evolved in areas that would later coincide with those occupied by historical populations:
Etruscans, Venetians, Rhaetians, etc. [57,89]. Population centres in the southern Po Plain, and
later Veneto, grew larger and become more organised, with streets, rectilinear plans, and pub-
lic buildings [58,91]. Social stratification and the emergence an aristocracy visible is in the
funerary record [57,92]. Exchange relationships within the Po Plain intensified, as did trade
with Europe, central Italy, and the Greek world, mediated by the region’s Etruscan cities [93].
Archaeobotanical remains from across northern Italy suggest a degree of continuity in crop
choice between the Bronze and Iron Ages, with a potential increase in the importance of
legumes [94]. The greater diffusions of iron, especially from the sixth century BC, expanded
the use of metal ploughs and agricultural tools, and the potential to work heavier soils [95]. It
has recently been proposed that surplus cereal production in the Iron Age Germany was
achieved through manuring and an expansion in cultivation [96]. However, the applicability
of this conclusion to an Italian lowland context is unclear. Pollen spectra do not indicate signif-
icant clearing during the Iron Age, e.g. [68], although the diffusion of iron tools and ploughs
may have allowed cultivation of heavier soils that were already more or less free of trees. By the
mid first millennium BC vine cultivation had become well established [97,98], and in Etruscan
territory it was sufficiently organised to produce a surplus for export [99101]. This scale of
production suggests established land-holding and long-term strategies in land use.
Urban centres continued to develop in the Veneto into the latter half of the millennium,
but Etruscan influence decreased as a result of Roman expansion in the south and migrations
from the north. Material evidence for La Tène culture in northern Italy expanded from the
sixth century BC, and came to dominate large parts of the region by the fourth century BC
[102,103]. Although described by Roman authors as a violent invasion, the archaeological evi-
dence presents a more complex picture of acculturation following the migration of ‘Celtic’
populations into northern Italy [104]. Etruscan cities that once flourished were abandoned in
the third century BC, and the Roman Republic advanced into the region and gained control
over the area between the third and second centuries BC [55,105].
Materials and methods
Sites and assemblages
Late prehistoric sites from a large portion of the eastern Po Valley and Venetian Plain were
chosen for this study (S1 Table). The study area was limited to these recent valleys and their
borders, in order to compare sites subject to similar climatic conditions. In order to focus on
the economic use of animals, only assemblages from domestic settlement contexts were con-
sidered in analysis of taxon presence/abundance: sanctuaries, cult sites, and ritual deposits
were excluded. Ritual assemblages often have a different distribution of taxa from those found
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 5 / 33
in domestic contexts, e.g. a pronounced focus on a single species [35] (e.g. [106,107]), which
could bias results. Sites are divided into regional groups based on their location within the
study area and regional similarities in material culture. Faunal samples with very long chronol-
ogies that could not be assigned to a useful temporal range were excluded. Only dated contexts
were analysed; surface finds and materials recovered from plowzone were not considered.
Despite efforts for good regional and temporal coverage, faunal assemblages are not evenly dis-
tributed across the study areas and time periods. The assemblages also vary greatly in size,
from a few dozen to tens-of-thousands of bones. Information on some periods is entirely lack-
ing, while others are represented only by a single site. This large range results from differences
in site type, degree of preservation, ancient settlement patterns, and historical interests. Mod-
ern development obscures Iron Age sites in particular, as many major settlements of this
period are buried under modern–as well as medieval and Roman–cities. Although dividing the
study area into smaller sub-regions limited the size of comparative samples, such division was
necessary for investigating differences within the Po Basin. Illustrations and tables have been
organised to clearly demonstrate where gaps are present.
Quantifying taxonomic abundance and animal size
The relative proportions of cattle (Bos taurus), sheep/goat (Ovis aries/Capra hircus), and pigs
(Sus scrofa) were quantified using the number of identified specimens (NISP). Although sub-
ject to several systematic biases, NISP offers a widely employed and easy to calculate method of
assessing the relative frequency of different species [108]. Inter-observer differences in identifi-
cation and recording are a primary concern [109111], as they artificially inflate the impor-
tance of taxa with a greater number of ‘identifiable’ remains. In order to better control of
differences in recording practice and identification skill, rib, and vertebrate fragments were
excluded from NISP counts where possible, as were any specimens not identified as a specific
skeletal element. Complete and partial skeletons were also excluded from these analyses, since
their inclusion would inflate NISP counts. In order to apply statistical tests with a reasonable
degree of confidence, only assemblages with a minimum of 110 specimens identified to the
three main domesticates were considered in NISP analyses. The statistical significance of
inter-regional differences was tested using a chi-squared test in R software [112].
Changes in animal size were assessed using size-index scaled Log Standard Index values
(LSI) [113]. Widely available standards were used for cattle [114], sheep and goats [115], and
pigs [116]. Log ratios offer a useful method for investigating animal size, because they allow
measurements from different elements to be pooled into a larger sample of indices for each
dimension (length and breadth). However, this method of grouping can lead to the overrepre-
sentation of specimens that yield several measurements, a problem that is quickly com-
pounded if a skeleton or articulating limb is included. To avoid over representation of
articulating remains, our LSI analyses included only one measurement per specimen and one
specimen from an articulating group of bones. Bone widths were considered separately from
lengths in order to assess two-dimensional change. Measurements of bone depth were com-
paratively rare and were not included.
Table 1 presents the measurements used in LSI analyses in preferential order. Only one
length or breadth measurement per specimen was considered, e.g. if GLl was available for a
specimen, all subsequent length measurements would be excluded. This list purposefully
excludes elements and measurements particularly sensitive to animal age, although it is also
constrained by measurements included in the published standards. LSI values for cattle, sheep,
and pigs are presented in histograms. Because accurate species distinction can be challenging
for closely related taxa like sheep and goats [117], related species are also included: figures for
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 6 / 33
sheep also display counts and means for goats and undistinguished sheep/goat specimens; fig-
ures for pigs also contain data identified as wild boar. Undistinguished Sus sp. were grouped
with domestic pigs. The Mann-Whitney U Test, conducted in R, was used to identify statisti-
cally significant differences between populations. Only LSI values from cattle, sheep, and pig
were subject to statistical analysis; goats were excluded due to a lack of reliable biometric
Metacarpal measurements were used to investigate diachronic changes in sex ratios for cat-
tle and sheep. The balance of males, females, and castrates within a population can influence
observed variation in animal biometry, especially if the sexual composition of the group
changed over time. Metacarpals are the most sexually dimorphic element in domestic bovids
[119121], and consideration of metacarpal size and shape can provide information on sex
representation [122,123]. To aid interpretation, measurements were compared to data from
animals of known sex: archaeological cattle from Eketorp ringfort [123], identified using
DNA, and modern Shetland sheep [115,124].
Results: Relative abundance of livestock
Regional patterns
Analysis of NISP data for taxon abundance and livestock ratios (Table 2) demonstrated that
livestock representation varied significantly between regions (Fig 2). Inter-regional compari-
sons within each chronological period were statistically significant in nearly all instances, and
these differences persisted through time (S2 Table). Amongst sites of the North group, pigs
were relatively rare compared to cattle and sheep/goat. The proportions of livestock were more
balanced between the three taxa in the Veneto, with the exception of FBA and IA0. Caprines
and pigs were very abundant in the South study area, while cattle were less common. Livestock
ratios from the Friuli and Romagna regions were more variable, which may reflect either a
more heterogeneous approach to husbandry practices or variation resulting from the lower
number of sites from these areas.
Table 1. Measurements used in calculation of LSI values for each taxon.
Dimension Measurement Reference Cattle Sheep/goat Pig
Length GL von den Driesch 1976
Humerus, Radius, Mc, Mt,
Humerus, Radius, Mc, Mt, Femur, Tibia,
Mc III, Mc IV,
GLl von den Driesch 1976
Astragalus Astragalus Astragalus
GLm von den Driesch 1976
Astragalus Astragalus
HTC Davis 1996 [115] Humerus Humerus Humerus
HT Davis 1996 [115] Humerus
Breadth Bd von den Driesch 1976
Humerus, Radius, Mc, Mt, Tibia,
Mc, Mt, Astragalus, Tibia Humerus, Tibia
BT von den Driesch 1976
Humerus Humerus Humerus
Bp von den Driesch 1976
Humerus, Radius, Mc, Mt, Tibia Radius, Mc Radius
BFd von den Driesch 1976
BFp von den Driesch 1976
Radius Radius
Mc = Metacarpal. Mt = Metatarsal.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 7 / 33
Diachronic change
Study regions both north and south of the River Po displayed major diachronic changes in
livestock representation, but each area followed a distinct trajectory. Within individual
regions, diachronic change was statistically significant for most temporal transitions (indicated
in Fig 2;S2 Table). Patterns of animal exploitation evolved in all areas, but the direction and
timing of these changes differed. South of the Po River, a shift in livestock percentages is
apparent in the Iron Age. Its origin is obscured by a lack of data for FBA and IA0, but signifi-
cant changes to livestock ratios occurred in both the South and Romagna regions by IA1. Pigs
replaced sheep/goat as the most abundant taxon in the South group; caprines also decreased in
the Romagna area, but there cattle rose in importance. North of the Po River, changes to NISP
percentages occurred during RBA, earlier than in the southern plain. As in the other regions,
Table 2. NISP and relative percentages of principal livestock by study area.
Study area Period n. of assemblages Mammal NISP NISP 3 dom. Cattle Pig Sheep
Friuli Middle Bronze Age MBA
Middle-Recent Bronze Age MRBA
Recent Bronze Age RBA 1 144 138 42% 35% 23%
Recent-Final Bronze Age RFBA
Final Bronze Age-Early Iron Age FBA-IA0 1 2106 2021 41% 37% 22%
Early Iron Age IA0 1 1153 1100 36% 46% 18%
Mid Iron Age IA1 1 451 437 24% 34% 42%
Late Iron Age IA2 2 813 769 42% 33% 26%
North Middle Bronze Age MBA 4 8714 7872 32% 21% 47%
Middle-Recent Bronze Age MRBA 2 3845 3577 36% 17% 47%
Recent Bronze Age RBA 3 6864 6126 53% 18% 29%
Recent-Final Bronze Age RFBA 2 352 329 38% 28% 34%
Final Bronze Age FBA
Early Iron Age IA0 1 202 182 51% 24% 25%
Mid Iron Age IA1
Late Iron Age IA2 4 7133 6760 48% 24% 28%
Veneto Middle Bronze Age MBA 2 1380 1232 31% 31% 39%
Middle-Recent Bronze Age MRBA
Recent Bronze Age RBA 6 4015 3587 27% 32% 40%
Recent-Final Bronze Age RFBA 2 475 418 41% 30% 30%
Final Bronze Age FBA 1 1426 1145 32% 50% 18%
Early Iron Age IA0 1 781 533 18% 56% 26%
Mid Iron Age IA1 1 1714 1606 26% 26% 48%
Late Iron Age IA2 3 645 568 36% 35% 29%
South Middle Bronze Age MBA 7 6374 5676 22% 33% 45%
Middle-Recent Bronze Age MRBA 2 1051 954 19% 32% 48%
Recent Bronze Age RBA 2 1738 1635 20% 22% 58%
Recent-Final Bronze Age RFBA
Final Bronze Age FBA
Early Iron Age IA0
Mid Iron Age IA1 5 2387 2215 26% 44% 30%
Late Iron Age IA2 11 34168 32082 14% 63% 23%
Romagna Middle Bronze Age MBA 2 2800 2516 19% 30% 51%
Mid Iron Age IA1 1 648 619 49% 34% 16%
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 8 / 33
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 9 / 33
the relative abundance of sheep/goat decreased, with cattle replacing them as the dominant
taxon. This emphasis on cattle exploitation continued in subsequent periods, albeit to a lesser
extent during RFBA–variation that could result from the small sample size. ‘
Unlike regions with a single change in species distribution, Friuli and the Veneto presented
several fluctuations in NISP percentages. After an emphasis on sheep/goat husbandry in MBA
and RBA, NISP data from the Veneto varied across the remainder of the Bronze Age and the
Iron Age. For certain periods few sites were available. This factor probably contributes to the
observed variation: during FBA and IA0, the Veneto is represented by a single site. This settle-
ment, Frattesina, was a unique proto-urban centre with far-reaching trade networks [90]. Its
NISP profile, with a large percentage of pigs, more closely resembles the Etruscan/Roman pat-
tern than that of the Bronze Age sites considered here. Of all the regions, NISP patterns in Fri-
uli are the most variable, potentially as a result of a low number of available assemblages.
Results: Biometry
Analysis of livestock biometry utilised 4886 unique post-cranial measurements. Table 3 pres-
ents summary statistics of LSI values by region and period. As stated above, the statistical sig-
nificance of temporal and inter-regional differences were analysed using the Mann-Whitney U
tests (S3 Table). The significance of diachronic changes is indicated in the relevant histograms;
full results are available in S3 Table.
Diachronic change. LSI values from cattle bone lengths (Fig 3) revealed statistically signif-
icant diachronic changes in the majority of study regions. Except in Friuli, for which there was
too little data to draw conclusions, mean cattle lengths increased in all regions between the
Bronze and Iron Ages. For the North and South study areas, the increase in bone lengths was
most significant when BM-R and IA2 were compared; however, small increases in mean LSI
values during BRF-IA0 and IA1 suggest that these changes were incremental. In the Veneto
study region a significant increase in bone length occurred between BM-R and IA1. The
Romagna region also displays a diachronic increase in average LSI lengths, although this
change was not statistically significant (see S3 Table).
Cattle LSI width values (Fig 4) also increased in size, although the timing of these changes
did not consistently follow that of bone lengths. Width values from Friuli were extremely small
in BM-R, but only three measurements were available. The Mann-Whitney U test demon-
strated a significant size increase in BRF-IA0, but conclusions should be treated tentatively on
account of the small BM-R sample size. Bone widths from the Romagna region increased
between the BM-R and IA1; this increase was significant for widths but not for lengths. In the
North study area, width LSI values behaved in a similar fashion to lengths: there was an incre-
mental increase in the mean between BM-R and BRF-IA0, although this increase was only sta-
tistically significant when BM-R and IA2 were compared. Unlike other regions where lengths
and width values grew in tandem, the two dimensions changed independently in the Veneto
and South study areas. In the Veneto, a statistically significant increase in cattle widths is visi-
ble in the BRF-IA0, earlier than the change in cattle length values during IA1. Southern widths
demonstrate a significant in increase in IA1, while cattle lengths in the region were relatively
stable until IA2. In these two regions, an increase in bone width without comparable alteration
Fig 2. Relative percentages of principal domestic livestock. See Table 2 for sample size. Stars indicate significance of
change from previous period according to Chi square tests: p 0.05, p 0.01, p 0.001(S2 Table).
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 10 / 33
Table 3. Sample size, mean, and standard deviation of LSI values per area and time period.
Length Width
Taxon Region Period n n sites min max mean sd n n sites min max mean sd
Cattle Romagna BM-R 31 2 -0.0640 0.0350 -0.0180 0.0252 35 3 -0.0785 0.0916 0.0064 0.0400
Cattle Romagna IA1 9 1 -0.0481 0.0400 -0.0024 0.0281 37 1 -0.1605 0.1264 0.0309 0.0551
Cattle Friuli BM-R 3 1 -0.1239 -0.0642 -0.0997 0.0314
Cattle Friuli BRF-IA0 51 3 -0.0764 0.0463 0.0008 0.0287 104 4 -0.1239 0.1335 0.0502 0.0502
Cattle Friuli IA1 1 1 0.0052 0.0052 0.0052 NA 4 1 0.0171 0.0693 0.0425 0.0227
Cattle Friuli IA2 1 1 -0.0351 -0.0351 -0.0351 NA 7 2 0.0088 0.0990 0.0466 0.0382
Cattle North BM-R 159 8 -0.1287 0.0600 -0.0176 0.0295 295 9 -0.1777 0.1264 0.0056 0.0533
Cattle North BRF-IA0 3 2 -0.0350 0.0252 -0.0061 0.0302 8 2 -0.0683 0.0760 0.0194 0.0502
Cattle North IA2 79 3 -0.0695 0.0794 -0.0043 0.0326 161 3 -0.0798 0.1601 0.0250 0.0494
Cattle South BM-R 49 8 -0.0778 0.0388 -0.0205 0.0278 107 8 -0.1096 0.1196 -0.0079 0.0526
Cattle South IA1 7 3 -0.0560 0.0144 -0.0195 0.0244 15 5 -0.0168 0.1027 0.0524 0.0364
Cattle South IA2 78 11 -0.0706 0.0869 0.0084 0.0316 134 14 -0.0558 0.1733 0.0508 0.0497
Cattle Veneto BM-R 34 7 -0.1032 0.0283 -0.0303 0.0341 92 7 -0.1581 0.1175 -0.0247 0.0489
Cattle Veneto BRF-IA0 6 1 -0.0600 0.0124 -0.0300 0.0288 15 1 -0.0896 0.0901 0.0015 0.0423
Cattle Veneto IA1 16 2 -0.0756 0.0640 -0.0034 0.0397 33 2 -0.1160 0.1453 0.0175 0.0624
Cattle Veneto IA2 2 2 -0.0282 0.0312 0.0015 0.0420 3 2 -0.0220 0.0751 0.0181 0.0507
Sheep Romagna BM-R 30 2 -0.0705 0.0759 0.0026 0.0331 51 2 -0.1135 0.0589 -0.0133 0.0320
Sheep Romagna IA1 2 1 0.0286 0.0408 0.0347 0.0086 5 1 -0.0445 0.0470 0.0152 0.0367
Sheep Friuli BRF-IA0 22 3 -0.0546 0.1317 0.0543 0.0450 38 3 -0.0188 0.1256 0.0374 0.0351
Sheep Friuli IA1 3 1 0.0745 0.0761 0.0755 0.0008 4 1 0.0454 0.0785 0.0655 0.0144
Sheep Friuli IA2 2 2 0.0477 0.0761 0.0619 0.0201 2 1 0.0069 0.0768 0.0418 0.0494
Sheep North BM-R 131 7 -0.1847 0.0762 0.0038 0.0440 158 6 -0.1657 0.0934 -0.0270 0.0378
Sheep North BRF-IA0 3 2 0.0032 0.0559 0.0298 0.0263 6 2 -0.0417 0.0215 -0.0042 0.0233
Sheep North IA2 52 3 -0.0067 0.1078 0.0370 0.0257 102 3 -0.0410 0.1000 0.0296 0.0284
Sheep South BM-R 26 5 -0.0724 0.0588 -0.0068 0.0366 49 7 -0.0969 0.1268 -0.0279 0.0354
Sheep South IA2 103 3 -0.0369 0.1129 0.0341 0.0297 158 5 -0.0516 0.0931 0.0158 0.0275
Sheep Veneto BM-R 20 5 -0.0566 0.0710 0.0274 0.0339 53 6 -0.1154 0.0802 0.0049 0.0462
Sheep Veneto BRF-IA0 4 1 0.0206 0.0604 0.0322 0.0188 10 1 -0.0231 0.0721 0.0185 0.0306
Sheep Veneto IA1 30 1 0.0175 0.1265 0.0660 0.0270 70 1 -0.0506 0.1021 0.0403 0.0278
Goat Romagna BM-R 2 2 -0.0648 0.0495 -0.0076 0.0808 4 2 -0.0754 0.1430 0.0063 0.0974
Goat Romagna IA1 2 1 -0.0430 0.0220 -0.0105 0.0460
Goat Friuli BRF-IA0 1 1 0.0138 0.0138 0.0138 NA 4 1 0.0680 0.1255 0.0960 0.0323
Goat North BM-R 28 5 -0.1490 0.1699 0.0182 0.0609 41 6 -0.1154 0.1177 0.0093 0.0457
Goat North IA2 11 3 -0.0016 0.0620 0.0368 0.0185 14 3 0.0098 0.0836 0.0343 0.0237
Goat South BM-R 6 2 -0.0200 0.0350 0.0145 0.0201 6 2 -0.0809 0.0296 -0.0172 0.0387
Goat South IA2 21 4 -0.0234 0.0924 0.0315 0.0342 28 4 -0.0075 0.1322 0.0405 0.0307
Goat Veneto BM-R 2 1 -0.0070 0.0049 -0.0011 0.0084 6 3 -0.0479 0.0429 0.0029 0.0424
Goat Veneto IA2 1 1 0.0466 0.0466 0.0466 NA
Sheep/goat Romagna BM-R 1 1 -0.0083 -0.0083 -0.0083 NA 7 1 -0.0798 0.0206 -0.0164 0.0364
Sheep/goat Romagna IA1 11 1 -0.0672 0.0252 -0.0296 0.0266
Sheep/goat Friuli BRF-IA0 6 1 0.0206 0.0506 0.0346 0.0136 6 2 0.0019 0.0969 0.0417 0.0372
Sheep/goat Friuli IA2 1 1 -0.0017 -0.0017 -0.0017 NA
Sheep/goat North BM-R 18 4 -0.1810 0.0256 -0.0519 0.0656 62 5 -0.0986 0.0539 -0.0192 0.0318
Sheep/goat North BRF-IA0 2 2 -0.0035 0.0361 0.0163 0.0280
Sheep/goat North IA2 1 1 0.0289 0.0289 0.0289 NA 2 2 0.0109 0.0506 0.0307 0.0281
Sheep/goat South BM-R 10 4 -0.0749 0.0506 -0.0121 0.0346 56 6 -0.0880 0.0743 -0.0171 0.0327
Sheep/goat South IA1 4 1 -0.0082 0.0771 0.0254 0.0368 13 3 -0.0553 0.0536 0.0096 0.0256
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 11 / 33
in length would have produced a change in the shape of cattle, creating animals which were
more robust, followed in the subsequent period by the emergence of taller and relatively more
slender cattle.
Intra-period differences. LSI results demonstrated significant regional variation in the
size of cattle, even within a single period. The data suggest that cattle in the Veneto were some
of the smallest in northern Italy (in terms of animal dimensions), although data from Friuli
were lacking: lengths from the Veneto were smaller on average than those from other areas
during BM-R, although this difference was only statistically significant when compared to the
North study area (p = 0.042, see S3 Table). The mean of LSI widths from the Veneto was also
lower than that of other BM-R regions, except for the 3 specimens from BM-R Friuli. Again,
these differences were not statistically significant in all cases. Interestingly, after some sparse
evidence for small animals during BM-R, Friuli produced data for large cattle during BRF-IA0.
Lengths from the region during BRF-IA0 were significantly larger than any BM-R study area
(p = 0.003–0.000) as well as other areas during BRF-IA0. Cattle LSI widths from Friuli are
Table 3. (Continued)
Length Width
Taxon Region Period n n sites min max mean sd n n sites min max mean sd
Sheep/goat South IA2 61 7 -0.0417 0.1380 0.0256 0.0322 69 10 -0.0711 0.1361 0.0118 0.0342
Sheep/goat Veneto BM-R 8 1 -0.1064 -0.0034 -0.0555 0.0308 29 5 -0.1316 0.1322 -0.0142 0.0581
Sheep/goat Veneto BRF-IA0 1 1 0.0854 0.0854 0.0854 NA 1 1 0.0726 0.0726 0.0726 NA
Sheep/goat Veneto IA1 1 1 0.0718 0.0718 0.0718 NA 17 1 -0.0693 0.1122 0.0253 0.0499
Sheep/goat Veneto IA2 1 1 0.0222 0.0222 0.0222 NA 1 1 0.0073 0.0073 0.0073 NA
Pig Romagna BM-R 8 1 -0.0390 0.0257 -0.0170 0.0207 14 2 -0.1311 0.0474 -0.0391 0.0451
Pig Romagna IA1 12 1 -0.0615 -0.0044 -0.0283 0.0194 20 1 -0.0888 0.0094 -0.0281 0.0248
Pig Friuli BRF-IA0 7 2 -0.0544 0.0400 -0.0062 0.0290 5 3 -0.0624 0.0371 -0.0114 0.0416
Pig Friuli IA1 1 1 0.0021 0.0021 0.0021 NA 13 1 -0.1367 0.0251 -0.0409 0.0389
Pig Friuli IA2 4 2 -0.0484 -0.0332 -0.0425 0.0067 5 2 -0.0447 0.0326 -0.0029 0.0355
Pig North BM-R 36 8 -0.0678 0.1417 0.0084 0.0501 58 6 -0.1124 0.0894 -0.0348 0.0361
Pig North BRF-IA0 2 1 -0.0544 -0.0229 -0.0387 0.0222 1 1 -0.0600 -0.0600 -0.0600 NA
Pig North IA2 41 3 -0.0741 0.1252 -0.0294 0.0350 43 3 -0.1020 0.0808 -0.0353 0.0377
Pig South BM-R 22 7 -0.0654 0.0348 -0.0031 0.0266 58 6 -0.0892 0.1186 -0.0255 0.0355
Pig South IA1 7 3 -0.1666 -0.0174 -0.0655 0.0474 21 4 -0.0914 0.0588 -0.0391 0.0351
Pig South IA2 676 8 -0.1082 0.1409 -0.0280 0.0352 415 11 -0.1745 0.1022 -0.0408 0.0364
Pig Veneto BM-R 19 5 -0.1184 0.0328 -0.0247 0.0363 41 7 -0.1225 0.0326 -0.0446 0.0406
Pig Veneto BRF-IA0 1 1 -0.0662 -0.0662 -0.0662 NA 5 1 -0.0722 -0.0032 -0.0383 0.0248
Pig Veneto IA1 3 1 -0.0173 0.0624 0.0305 0.0422 12 1 -0.0773 0.0713 -0.0020 0.0568
Pig Veneto IA2 2 1 -0.0459 -0.0445 -0.0452 0.0010
Wild boar Romagna BM-R 3 2 0.0964 0.1208 0.1089 0.0122
Wild Boar Romagna IA1 1 1 0.0869 0.0869 0.0869 NA
Wild boar Friuli BRF-IA0 1 1 0.1165 0.1165 0.1165 NA
Wild Boar Friuli IA1 1 1 0.1029 0.1029 0.1029 NA 1 1 0.1382 0.1382 0.1382 NA
Wild Boar North BM-R 13 2 0.0550 0.1492 0.1097 0.0282 8 3 -0.0331 0.1788 0.0923 0.0809
Wild boar North BRF-IA0 1 1 0.1257 0.1257 0.1257 NA
Wild Boar North IA2 2 1 0.0942 0.1252 0.1097 0.0220 4 1 0.0692 0.1084 0.0925 0.0178
Wild Boar South BM-R 5 2 0.0416 0.1269 0.0931 0.0320 5 2 0.0834 0.2137 0.1451 0.0483
Wild Boar South IA2 2 2 0.0503 0.0900 0.0702 0.0281 3 2 0.0834 0.2017 0.1323 0.0617
Wild Boar Veneto BM-R 1 1 0.1093 0.1093 0.1093 NA 4 3 -0.1225 0.0892 -0.0282 0.0876
Wild boar Veneto IA1 2 1 0.1144 0.1318 0.1231 0.0123
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 12 / 33
also significantly larger than those from other regions during the BM-R (p = 0.000) and
BRF-IA0 Veneto (p = 0.000).
During IA2, the South study area contained the largest cattle, both in terms of individual
measurements as well as for mean size. The Mann-Whitney U test demonstrated significant
differences between Southern bone lengths (p = 0.010) and widths (p = 0.000) and those
from Northern cattle in IA2. They also had greater LSI width values than Venetian cattle, at
least during IA1. Interestingly, the samples from IA2 Southern cattle were not statistically dif-
ferent from BRF-IA0 cattle from Friuli, the largest animals of the Bronze Age.
Sex ratios. Evaluation of the sex ratios within the study regions was difficult due to limited
data. Consideration of cattle metacarpal shape (Fig 5A) did not produce a clear pattern of sex-
ual dimorphism during the Bronze Age (e.g. North, Friuli); however, the large range of values–
comparable with Eketorp ringfort [123]–suggests a mixed population comprised of males,
females, and castrates. Subsequently, the appearance of very robust cattle in IA2 was accompa-
nied by disappearance of the most slender individuals, alongside a clearer separation of slender
and robust shape groups, indicating that both male and female cattle become more robust. Fig
5B suggests that some of the very tall IA2 cattle, particularly in the South, may have been cas-
trates, on account of their long length and slender shape. Because the points lie on the estab-
lished regression line, an internal change in sex ratios is more plausible than the introduction
of another breed or type [122]. Overall, while changes to sex ratios may have impacted the
observed size change, metacarpal data indicate an increase in height and robustness in both
sexes, and thus general change in population size, potentially influenced by–but not only the
result of–changes in sex composition.
Fig 3. Distribution of cattle bone length LSI values by study region and period. Arrows indicate significant diachronic changes according to Mann-Whitney U tests
(S3 Table).
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 13 / 33
Diachronic change. There were fewer measurements available for sheep than cattle, a
result of the difficulty of distinguishing sheep from goats based on the morphology of their
post-cranial remains [117]. Unlike cattle, changes in sheep bone lengths were typically mir-
rored by similar changes in widths. Between the Bronze and Iron Ages, mean LSI values from
sheep increased in size in all regions (Figs 6and 7). In Friuli and Romagna study areas, this
change was not statistically significant (see S3 Table). In the North, the change in mean was
incremental, with statistical differences present in a comparison of BM-R and IA2, as well as
BRF-IA0 to IA2 widths. In the South, sheep measurements were only available for the first and
final time periods, between which a significant change in both dimensions was apparent. An
incremental increase in mean was present in LSI values for both dimension from the Veneto,
but statistical differences were only noted for the change to IA1.
Intra-period differences. Sheep of the same temporal period differed in size across the
study regions. Like for cattle, mean LSI values for sheep from Friuli were larger than contem-
porary animals in other study areas. Compared to animals from the BM-R, LSI values from
BRF-IA0 Friuli were significantly larger in all length (p = 0.015–0.000) and width
(p = 0.001–0.000) comparisons. BM-R sheep from the Veneto were also comparatively
large, with significant differences in length and width values compared to those in the Roma-
gna, North, and South study areas. These regional size differences appear to have persisted
into the Iron Age, although the lack of data from Friuli complicates this assessment. IA2 sheep
in the North and South were of a similar height on average, but LSI width values from the
Fig 4. Distribution of cattle bone width LSI values by study region and period. Arrows indicate significant diachronic changes according to Mann-Whitney U tests
(S3 Table).
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 14 / 33
Fig 5. Cattle metacarpal (a) shape, using slenderness indices (shaft width/length versus distal width/length), and (b) slenderness (shaft width/length) versus
length. Comparative data from Telldahl et al. [123].
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 15 / 33
South were significantly smaller (p = 0.0007). Southern animals were therefore of a compa-
rable height as those in the North, but more slender.
Sex ratios. Data from sheep metacarpals (Fig 8) indicates a shift to taller, more slender
sheep. During the Iron Age, new tall animals emerge and the shortest and most slender drop
out, suggesting development in the size/shape of the whole population and not only a change
in sex ratios. This increase in tall and relatively slender animals could also reflect a shift in sex
ratios to include more castrates; however, such a change would not be consistent with culling
profiles from Iron Age sites south of the Po, which demonstrate an increase in lamb mortality
after the Bronze Age (presumably focused on males). In some cases mortality in the first year
reaches over 50% [35].
Diachronic change. The complex relationship between domestic pigs and wild boar com-
plicated analysis of pig biometry. Histograms of LSI values from pig length (Fig 9) and width
(Fig 10) measurements demonstrate different levels of separation between populations of
domestic pigs and larger wild boar. In some areas/periods, there was a distinct separation
Fig 6. Distribution of sheep bone length LSI values by study region and period. Includes specimens identified as sheep, goat, and sheep/goat. Arrows indicate
significant diachronic changes according to Mann-Whitney U tests (S3 Table) for sheep.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 16 / 33
between the two populations, and specimen identifications conformed to the expected size of
each species. In other study areas, LSI values from the two species were not so easily distin-
guished. This observed overlap between the species resulted from several factors: recording
practices that did not distinguish between wild and domestic suids (grouped here with domes-
tic pigs), misidentifications, and potentially inter-breeding between wild and domestic ani-
mals. For instance, the bi-model distribution of IA2 South LSI widths suggested a large
population of domestic pigs alongside a small population of wild boar. Conversely, for BM-R
North there was not a clear separation between the wild and domestic populations (regardless
of zooarchaeological identification), possibly indicating of a greater level of genetic mixing
between wild and domestic Sus.
In contrast to cattle and sheep, the predominant trend in domestic pigs LSI values was a
reduction in bone lengths. Mean values for pig LSI lengths decreased in all regions between
the Bronze and Iron Ages. This change was only statistically significant in the North and South
study areas (see S3 Table), but the small samples available for regions could have precluded sta-
tistical identification of similar trends. Changes in pig widths were more variable. In the
Romagna study area, the mean of LSI width values increased through time while average
length decreased, suggesting that the pigs became shorter and more robust. In the North,
Fig 7. Distribution of sheep bone width LSI values by study region and period. Includes specimens identified as sheep, goat, and sheep/goat. Arrows indicate
significant diachronic changes according to Mann-Whitney U tests (S3 Table) for sheep.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 17 / 33
Fig 8. Sheep metacarpal slenderness (shaft width/length) versus length. Comparative data from Davis [115,124].
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 18 / 33
width values remained constant while bone lengths decreased; here too pigs became shorter
and relatively more robust. Few length measurements were available for the Veneto, preclud-
ing identification of changes in animal height, but pig bone widths significantly increased in
size between BM-R and IA1. However, misidentification of wild boar may have influenced this
analysis: the largest domestic LSI values for the Veneto are greater than 0.5, an approximate
dividing line between domestic pigs and wild boar in other areas. In contrast to other regions,
pigs in the South became smaller overall, and both width and length LSI values decreased
Intra-period differences. Like the other domestic taxa, pigs varied in size between
regions. For BM-R, pig LSI length values were greatest in the North, while mean widths values
were larger in the South. These differences were only statistically significant for a comparison
of the South (lengths and widths, p = 0.040, 0.016) or North (lengths, p = 0.026) areas with
the Veneto. By IA2, the North and South study areas came to share similar mean length values.
LSI lengths from IA1 in the Romagna study area were comparable to bone lengths from IA2
North and South. Widths from Romagna IA1 pigs were significantly larger than width values
from the South study area (p = 0.023), implying that pigs in Romagna were of a similar height
to Southern animals, but more robust during the Iron Age. Too few data were available to
Fig 9. Distribution of pig bone length LSI values by study region and period. Includes specimens identified as domestic and wild pigs. Arrows indicate significant
diachronic changes according to Mann-Whitney U tests (S3 Table) for domestic pigs.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 19 / 33
draw conclusions about pig bone lengths from Veneto and Friuli. For pig LSI widths in Friuli,
input from misidentified wild boar seems to have been a contributing factor in the larger size
of animals.
Summary of biometric data
With exception of samples containing very few measurements, mean LSI values from cattle
and sheep increased in all regions between the Bronze Age and Iron Age. This increase was
present in both bone widths and lengths, demonstrating an overall increase in animal size.
These changes were statistically demonstrable for study areas with larger data sets: North,
South, and Veneto. Consideration of metacarpal size/shape indicated that both males and
females increased in size; biometric change was not the result of differences in sex ratios. Iden-
tification of the timing of this increase, and particularly of the earliest evidence for change,
proved difficult due to the distribution of the data and the comparatively small quantity of
measurements available for the important transitional period spanning the Final Bronze Age
(BF) and first centuries of the Iron Age (IA0). Nevertheless, statistically significant changes to
length/width LSI values of cattle and sheep were present for comparisons of previous period to
Fig 10. Distribution of pig bone width LSI values by study region and period. Includes specimens identified as domestic and wild pigs. Includes specimens
identified as domestic and wild pigs. Arrows indicate significant diachronic changes according to Mann-Whitney U tests (S3 Table) for domestic pigs.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 20 / 33
IA1 in several study areas. Earlier and statistically significant increases in cattle width values
were also visible in the Veneto in BRF-IA0. Also interesting is that mean LSI values for cattle
and sheep increased between BM-R and BR-IA0 for all periods with data, except for Veneto
cattle lengths. In only one instance was this increase statistically significant (Veneto cattle
widths), but the small samples might have precluded statistical identification of real incremen-
tal change. LSI analysis also revealed differences in the nature of size change for different spe-
cies. While sheep lengths and widths increased in tandem, cattle widths increased prior to
cattle lengths in the South (IA1) and Veneto (BRF-IA0). Pigs displayed changes distinct from
those of bovids. In study areas with larger samples (North, South, and Romagna), mean pig
LSI lengths decreased, indicating that swine became shorter in these regions. In the South, a
corresponding decrease in LSI width values occurred in the IA1. In other regions, width values
were more variable: the mean increased in IA1 Veneto and Romagna study areas, although the
inclusion of misidentified wild boar specimens or hybrids could have contributed to this
Patterns of species exploitation
Previous studies [7,35,4650] have have provided syntheses of regional zooarcheaological
data, but ours is the first to unify considertion of species frequencies and detailed biometric
analysis over later prehistory across such a large part of northern Italy. Our results demon-
strate a high degree of regionality in livestock ratios, which varied significantly between, and
sometimes within, study areas, even during within a single chronological period. Consider-
ation of site location (see Fig 1), precipitation and ground infiltration (Fig 11) demonstrates
that rainfall and elevation did not have a universal impact on species frequencies. Friuli, a
region with a cooler and wetter climate than the central Po Valley, contains a relatively high
proportion of cattle, but the trend is not consistent across all periods, and cattle-dominant
strategies were not constrained to the region. Data from individual sites also suggest diverse
responses to high humidity: Fondo Paviani, which has produced evidence for a local shift to
wetter conditions during the Final Bronze Age [125], saw a decrease in the abundance of cattle
over the same period [126]. Elevation also did not predetermine patterns of exploitation: pre-
Alpine sites preferred cattle, but pigs predominated on settlements in the Apennine range.
Geomorphology may have had a more significant role in shaping differences in animal
management north and south of the River Po. The northern Po Valley contains greater areas
of free-draining gravelly sediments, while the lower plain is characterised by silts and clays.
South of Lake Garda, Pleistocene fluvial and fluvio-glacial alluvial deposits extend southward
much of the way to the River Po, and similar deposits cover the northern Veneto and large
areas of Friuli [129,130]. The surface of the southern Po Plain is younger, and covered with
fine Holocene fluvial deposits that stretch to pede-Apennine alluvial fans. As a result of these
finer sediments and the shallow depth of the water table (currently c. 1.5–3 m from the sur-
face), the lower plain is wetter and less permeable, promoting a damper landscape [131]. The
different hydrology of these areas has modern agricultural implications, which promote or
constrain various types and scales of irrigation and crop cultivation. Areas less suited to certain
forms of plant agriculture may have adopted pastoral economic strategies. Equally, communi-
ties in free-draining areas may also have benefitted from access to a greater range of mobility
routes, without the difficulty of navigating large areas of wet, low-lying territory, where flood-
ing or marshy ground could impede the movement of large herds.
While local environmental conditions and resource availability might promote or preclude
certain forms of production, the diverse diachronic changes in species frequencies
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 21 / 33
demonstrate that cultural preferences and economic strategies also influenced livestock pro-
duction (see also [45]). The significant expansion of pig production on Etruscan sites in the
southern Po Valley is the clearest example of a culturally specific pattern of animal exploitation
[cf. [51]). The increase in pork consumption in Etruscan territory has been related to socio-
economic changes that encouraged surplus meat production, probably used to provision trad-
ers or re-distributed in service of political authority as tax or tribute [35]. By the sixth century
BC, the abundance of pig bones on these sites probably represents a real intensification of
swine husbandry–achieved, for example, through multiple farrowing induced by supplemental
Fig 11. Maps of present-day precipitation and ground infiltration with relative percentages of the three main
domesticates by site. Precipitation data from WorldClim 1.4 (current conditions) by [127].
Infiltration data follows Giuliano et al. [128].
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 22 / 33
feeding–, as well as the movement of animals to central places. Such a strategy has been pro-
posed for the Celtic village of Levroux in France (2nd century BC) [132]. Interestingly, Fratte-
sina in the Veneto also contained a large abundance of pig remains. The identification of an
‘Etruscan’ NISP pattern on this site suggests that the origins of the preference for swine hus-
bandry extend back to the early Iron Age or Final Bronze Age, and the beginning of the proto-
Villanovan culture in northern Italy. Although located in the Veneto, the material culture of
Frattesina also places the site within this cultural phenomenon of southern Po Plain, and links
it directly with later central places like Etruscan Bologna, which succeeded Frattesina as a
major centre of production and exchange [57,133]. As a central place of major inter-regional
importance, similar–if less intense pronounced–socio-economic incentives for surplus meat
production would have existed at Frattesina, which has produced ample evidence for the
acquisition raw materials and craft production, far-reaching trade routes, and social hierarchi-
sation [86,90]. However, in the Veneto this strategy ends with this site; assemblages from later
Venetian settlements do not show the high-pig pattern, despite demonstrating comparable
developments in social stratification and urban settlement organisation, suggesting a strong
cultural preference unique to the Etruscan southern plain.
Divergent patterns of size change
Biometric analyses demonstrated divergent patterns of size change in domestic bovids com-
pared to suids. These trends, in which saw bovids increase in size while pigs decreased, suggest
that biometric change was not the result of a universal, external factor like climatic change, but
rather the product of anthropic modifications to husbandry practices. Compared to regional
diversity in NISP frequencies, size increase was a broad phenomenon which occurred across
Northern Italy, even if regions displayed different degrees of change. The precise origin of
these increases remains unclear due to the limited data available for the Final Bronze Age and
Early Iron Age, but significant transformation in the size of domestic bovids over later prehis-
tory was apparent in the North, South, and Veneto study areas; evaluation of other regions is
constrained by sample size. Where it is possible to assess diachronic trends, change in livestock
size appeared incremental, developing over the Bronze–Iron Age transition, into the late Iron
Age. Metacarpal biometry illustrated size/shape change along the same regression line, condu-
cive with change within a single population. The progressive and pan-regional nature of these
trends suggests that they resulted from internal developments common across northern Italian
communities rather than the introduction of a new population of livestock, as might follow
conquest or colonisation. Body size increase was not limited to urban Etruscan or Venetian
areas, nor were large livestock limited to such sites: in contrast to our initial predictions,
advances in livestock productivity were therefore independent of urban population density or
settlement structures. Broader transformations in animal management must therefore have
played a role.
Advances in sheep and cattle husbandry, as indicated by size increase, occurred without
similar evidence for an improvement in pig body size. Our results demonstrated a diminution
of domestic suids over later prehistory in a continuation of the trend established between the
Neolithic and Bronze Age [134]. Previous studies have identified a disconnect in body-size
change in domestic cattle and sheep versus pigs during later Italian prehistory [7,35,42,44,
45], but our analysis is the first to show a diachronic decrease in carcass size over this period.
Pigs in Italy are thought to have increased in size during the Roman Imperial period [40,135],
although in southern areas diminution may have continued into Roman times [41]. Pig hus-
bandry therefore appears to have been more conservative than sheep or cattle exploitation
over late prehistory, probably as a result of different management styles and feeding strategies.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 23 / 33
Implications for agricultural strategies
In antiquity, most pigs were raised in free-ranging herds, kept for at least part of the year in
local woodlands to take advantage of mast forage [136]. Such systems were common in the
Mediterranean until relatively recently [137,138], and pigs raised in this manner are depen-
dent on natural foods part of the year. Extensively raised animals have the opportunity to
interbreed with wild boar naturally present in the same woodland. In these systems, confine-
ment and habitat fragmentation has a significant impact on access to wild resources and ani-
mals, with consequences for carcass size [139]. Because nutrition is a major element in
determining animal health with significant impact on animal size and fertility [115,140142],
increased population density without food supplementation negatively influences body size.
Enclosure–even in a relatively large enclosure of 290 ha–produced a reduction in body size of
Italian wild boar due to increased competition and reduced food quantity [138]. Fencing or
deforestation during prehistory could have produced a similar effect in domestic pigs. Con-
finement of pig herds did not necessarily equate to keeping animals in stalls, but of limiting
their range of movement. If accompanied by a decrease in breeding age [15], such changes
could produce significant size diminution. A reduction of interbreeding with phenotypically
larger wild boar could also have had a role [134], potentially compounding a size decrease
resulting from changes in territorial range and inter-animal competition. The variability and
lack of separation between wild and domestic populations in pig histograms suggests greater
interbreeding during the Bronze Age than in subsequent periods. Separation of wild and
domestic Sus through enclosure or a reduction in the number of wild animals through hunting
pressure (insinuated by the reduction in wild taxa on archaeological sites, cf. [7]) would reduce
gene flow with larger wild pigs. Habitat reduction through deforestation may have contributed
to the decline of wild boar, but this appears to have been relatively minor, since pollen analyses
from the region do not suggest significant changes in forest cover during the Iron Age [64,69].
Bovid management strategies are more complicated to reconstruct due to the greater num-
ber of products for which they were exploited. A number of factors, alone or in combination,
may have provoked an increase in bovid size: e.g. changes to feeding strategy, access to pasture,
mobility patterns, or gene flow. Broader changes in agricultural production during the Iron
Age suggest greater specialisation (e.g. of viticulture) and surplus output, as an increase in pop-
ulation and non-farming craftsmen would require a concurrent rise in cereal production. In
Iron Age Germany, isotopic studies suggest a surplus was achieved through manuring and an
expansion of arable cultivation, albeit with a significant degree of local variability [96]; how-
ever, similar studies are lacking for northern Italy, and the mechanisms through which greater
agricultural output was achieved are open to debate. While an expansion in cultivated area or
the exploitation of heavier soils may have encouraged the development of larger cattle for trac-
tion, similar pressures would have been present in densely inhabited areas during the Bronze
Age. Furthermore, at least in the Veneto, increases in cattle size around the Bronze–Iron Age
transition pre-date the diffusion of iron agricultural tools. Body size is, however, associated
with milk production [143], an increase in which would have been desirable in both large and
small bovids. A change in vertical mobility patterns in Switzerland during the twelfth century
BC has been argued to result from greater focus on secondary product exploitation [144]. The
impact of this change in strategy on animal biometry was not included in the study. However,
the authors highlight the relationship between changes in animal management and concurrent
cultural change, which presents interesting parallels for our data. In northern Italy, our earliest
evidence for cattle size increase dates to the late Bronze–Iron Age transition–a period of signif-
icant cultural change associated with the spread of proto-Villanovan culture during the Final
Bronze Age and subsequent development of distinct regional cultures during the Early Iron
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 24 / 33
Age. A thorough understanding of the impact of this development on arable agriculture is
lacking, but an intensification of trade and craft specialisation are apparent in the archa-
eological record [145]. As in Switzerland, cultural change may have prompted a change in
management strategy, supported in northern Italy by greater circulation of animals or hus-
bandry knowledge/skills, which moved alongside other materials in better integrated trade
Ultimately, conclusive identification of the cause(s) underlying these developments is not
possible at this juncture. However, this regional synthesis sets a basis to further explore the
subject with targeted analyses (e.g. isotopes, DNA), as well as to identify the areas and periods
that need further research. The results presented here demonstrate that NISP frequencies dis-
played a high degree of regionality, and the lack of correlation between NISP data and mean
body size suggests that changes in animal management were more influential than the focus
on a particular taxon. Quantitative assessment of mortality patterns, which are central to
reconstruction of husbandry regimes, was beyond the scope of this paper due to the diverse
methods used in the literature to record animal age. However, our consideration of metacarpal
biometry indicated that changes in sex ratios, which are closely linked to slaughter strategies,
were not the primary cause of size improvement. Considering differences in taxon preference
and the diverse settlement patterns, material cultures, and languages present across the study
areas, it is interesting to see that the overall tendency in livestock size change–bigger cattle and
sheep, smaller pigs–occurred across the whole region, although with different rhythms. Large
livestock were probably universally desired, but due to various constraints unattainable or too
impractical to sustain. Our results suggest that these limitations changed in later prehistory,
well before the Roman conquest.
An increase in livestock size is a key feature of the Romanisation of Western Europe, and one
believed to result from changes in farming strategies and the importation of larger Roman live-
stock. This change was a reversal of millennia of size diminution beginning with domestica-
tion, and it signalled significant changes in how animals were managed. However, in northern
Italy livestock increased significantly in size prior to the Roman conquest. Etruscan/Roman
civilisation was thought to have had a key role in the development of large cattle and sheep,
but our results demonstrated an earlier size increase in domestic bovids across northern Italy,
regardless of cultural context or degree of urbanisation. This process was visible in all regions
with sufficient samples. Metacarpal biometry and the distribution of size data suggest an incre-
mental, internally mediated process. The origin of the increase is difficult to identify due to a
lacuna in the data, but it is now clear that the late Bronze Age or Early Iron Age was the start-
ing point. Unlike cattle and sheep, pigs decreased in size over later prehistory, a continuation
of the established trend that began in the Neolithic. This diminution implies that herding strat-
egies of domestic suids were significantly different from that of bovids, displaying greater con-
tinuity with established trends. This pattern is present even south of the River Po, where
zooarchaeological evidence for the species exploited suggests a dramatic shift in husbandry
strategy in conjunction with Etruscan culture. In contrast to the widespread trends observed
in biometric analysis, the relative abundance livestock taxa varied significantly between study
regions. In some instances, husbandry patterns are highly culturally specific, such as the emer-
gence of a pig-dominated pattern on Etruscan sites; in other areas assignment of trends to spe-
cific culture was more difficult and may be complicated by local responses to topography.
Nevertheless, the marked regionality visible in NISP frequencies did not impact the overall tra-
jectory of body size change, which was observed across the northern Italy.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 25 / 33
This evolution of livestock size, across regions with diverse taxon frequencies, material
cultures, and settlement patterns, represents a previously undocumented pan-regional
improvement in cattle and sheep during later prehistory. It signals a significant change in the
management of domestic bovids, in which a shift in the aims of husbandry, agricultural knowl-
edge, feeding strategies, or mobility regimes may have played a role. Future DNA and isotopic
analyses will help clarify the role of herd mobility and livestock trade in catalysing these
changes. The socio-economic forces and cultural preferences that shaped these patterns also
deserve further consideration [146148], particularly in relation to how social stratification
and site catchment areas impacted access to resources like pasture, fodder, and desirable
animals. Economic connectivity and complexity were crucial to the expansion of surplus agri-
cultural production and the rise of the Roman market economy [149152], and our zooarch-
aeological data provide evidence that agricultural production began to shift across northern
Italy well before the Roman e
´poque. Advances in livestock management thus follow a similar
trajectory as other forms of archaeological evidence that developed significantly from the Final
Bronze Age/Early Iron Age across the first millennium BC: an expansion of trade, wealth
consolidation, and organised resource exploitation [57,148,149,152154]. Agricultural
improvements in late prehistoric Italy were not the result protohistoric urbanism or Roman
innovation, but part of an earlier incremental phenomenon, likely driven by the same forces
that shaped the broader socio-economic landscape of the region during later prehistory.
Supporting information
S1 Table. List of sites with data on chronology, analysis group, NISP, number of LSI values,
and associated references.
S2 Table. Results of chi-squared tests on livestock NISP.
S3 Table. Results of Mann-Whitney U tests on LSI values.
The authors are grateful to the reviewers and C. Cavazzuti for their comments on previous
drafts of this paper, and to U. Tecchiati and M. Bertolini for their kind help accessing compar-
ative data and references.
Author Contributions
Conceptualization: Angela Trentacoste, Ariadna Nieto-Espinet, Silvia Valenzuela-Lamas.
Investigation: Angela Trentacoste, Ariadna Nieto-Espinet, Silvia Valenzuela-Lamas.
Methodology: Angela Trentacoste, Ariadna Nieto-Espinet, Silvia Valenzuela-Lamas.
Supervision: Silvia Valenzuela-Lamas.
Writing – original draft: Angela Trentacoste.
Writing – review & editing: Angela Trentacoste, Ariadna Nieto-Espinet, Silvia Valenzuela-
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 26 / 33
1. Uerpmann H-P. Metrical analysis of faunal remains from the Middle East. In: Meadow R, Zeder MA,
editors. Approaches to faunal analysis in the Middle East. Cambridge, MA: Peabody Museum; 1978.
p. 41–5.
2. Meadow R. Osteological evidence for the process of animal domestication. In: Clutton-Brock J, editor.
The walking larder: patterns of domestication, pastoralism, and predation. London: Unwin Hyman;
3. Davis SJM. The archaeology of animals. New Haven: Yale University Press; 1987.
4. Zohary D, Tchernov E, Horwitz LK. The role of unconscious selection in the domestication of sheep
and goats. J Zool. 2001; 245(2):129–35.
5. Zeder MA. A metrical analysis of a collection of modern goats (Capra hircus aegargus and C.h.hircus)
from Iran and Iraq: implications for the study of saprine domestication. J Archaeol Sci. 2001; 28(1):61–
6. Matolcsi J. Historische Erforschung der Ko
¨sse des Rindes auf Grund von ungarischem Kno-
chenmaterial. Zeitschrift fu¨r Tierzu¨chtung und Zu¨chtungsbiologie. 1970; 63:155–94.
7. Riedel A. Archeozoological investigations in north-eastern Italy: the exploitation of animals since the
Neolithic. Preistoria Alpina. 1994; 30:43–94.
8. Bo
¨nyi S. History of domestic mammals in Central and Eastern Europe. Budapest: Akade
´; 1974.
9. Altuna J. Historia de la domesticacio
´n en el Paı
´s Vasco desde sus orı
´genes hasta la romanizacio
Munibe. 1980; 32:9–163.
10. Me
´niel P. Contribution àl’histoire de l’e
´levage en Picardie du Ne
´olithique àla fin de l’A
ˆge du Fer.
Amiens: Socie
´de Pre
´histoire du Nord et de la Picardie; 1984.
11. Vigne J-D. Les mammifères post-glaciaires de Corse, e
´tude arche
´ozoologique. Paris: CNRS; 1988.
12. Valenzuela-Lamas S, Albarella U. Animal husbandry across the Western Roman Empire: changes
and continuities. Eur J Archaeol. 2017; 20(3):402–15.
13. Audoin-Rouzeau F. La taille du boeuf domestique en Europe de l’Antiquite
´aux temps modernes.
Fiches d’oste
´ologie animale pour l’arche
´ologie, se
´rie B: mammifères. 1991; 2:3–40.
14. Clutton-Brock J. Domesticated animals from early times. London: British Museum; 1981.
15. Manning K, Timpson A, Shennan S, Crema E. Size reduction in early European domestic cattle relates
to intensification of Neolithic herding strategies. PLOS ONE. 2015; 10(12):e0141873.
10.1371/journal.pone.0141873 PMID: 26630287
16. Keay S, Terrenato N. Italy and the West: comparative issues in Romanization. Oxford: Oxbow
Books; 2001.
17. Mattingly DJ. An Imperial possession: Britain in the Roman Empire, 54 BC–AD 409. London: Pen-
guin; 2007.
18. Stek T. Roman imperialism, globalization and Romanization in early Roman Italy. Research questions
in archaeology and ancient history. Archaeological Dialogues. 2014; 21(1):30–40.
19. King AC. Diet in the Roman world: a regional inter-site comparison of the mammal bones. Journal of
Roman Archaeology. 1999; 12(1):168–202.
20. Trixl S, Steidl B, Peters J. Archaeology and zooarchaeology of the Late Iron Age–Roman transition in
the province of Raetia (100 bc–100 ad). Eur J Archaeol. 2017; 20(3):431–50.
21. Davis S. Faunal remains from Alca
´c¸ova de Santare
´m, Portugal. Lisboa: Instituto Portugues de
Arqueologia; 2006.
22. Fre
´mondeau D, Nuviala P, Duval C. Pigs and cattle in Gaul: the role of Gallic societies in the evolution
of husbandry practices. Eur J Archaeol. 2017; 20(3):494–509.
23. Lepetz S. Effets de la romanisation sur l’e
´levage et la diète carne
´e dans les e
´tablissements ruraux du
nord de la Gaule. L’exemple de l’augmentation de la stature des animaux domestiques. In: Bayard D,
Collart J-L, editors. De la ferme indigène àla villa gallo-romaine. Amiens: Socie
´de Pre
´histoire du
Nord et de la Picardie; 1996. p. 317–24. (Revue arche
´ologique de Picardie; nume
´ro spe
´cial 11).
24. Forest V, Rodet-Belarbi I. A propos de la corpulence des bovins en France durant les pe
´riodes histo-
riques. Gallia. 2002; 59:273–306.
25. Pigière F. The evolution of cattle husbandry practices in the Roman period in Gallia Belgica and west-
ern Germania Inferior. Eur J Archaeol. 2017; 20(3):472–93.
26. Lauwerier R. Animals in Roman times in the Dutch Eastern River area. Amersfoort: Project Oostelijk
Rivierengebied; 1988.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 27 / 33
27. Groot M. Developments in animal husbandry and food supply in Roman Germania Inferior. Eur J
Archaeol. 2017; 20(3):451–71.
28. Teichert M. Size variation in cattle from Germania Romana and Germania Libera. In: Grigson C, Clut-
ton-Brock J, editors. Animals and archaeology 4: husbandry in Europe. Oxford: British Archaeological
Reports; 1984. p. 93–104. (BAR international series; 227).
29. Breuer G, Rehazek A, Stopp B. Gro
¨nderungen des Hausrindes. Osteometrische Untersu-
chungen großer Fundserien aus der Nordschweiz von der Spa
¨tlatènezeit bis ins Fru¨hmittelalter am
Beispiel von Basel, Augst (Augusta Raurica) und Schleitheim-Bru¨el. Jahtesberichte aus August und
Kaiseraugst. 1999; 20:207–28.
30. Groot M, Deschler-Erb S. Carnem et circenses–consumption of animals and their products in Roman
urban and military sites in two regions in the northwestern provinces. Environmental Archaeology.
2017; 22(1):96–112.
31. Groot M, Deschler-Erb S. Market strategies in the Roman provinces: different animal husbandry sys-
tems explored by a comparative regional approach. J Archaeol Sci Rep. 2015; 4:447–60.
32. Albarella U, Johnstone C, Vickers K. The development of animal husbandry from the Late Iron Age to
the end of the Roman period: a case study from South-East Britain. J Archaeol Sci. 2008; 35:1828–48.
33. Colominas L, Ferna
´ndez Rodrı´guez C, Iborra Eres MP. Animal husbandry and hunting practices in
Hispania Tarraconensis: an overview. Eur J Archaeol. 2017; 20(3):510–34.
34. Albarella U. The end of the Sheep Age: people and animals in the Late Iron Age. In: Haselgrove C,
Moore T, editors. The later Iron Age in Britain and beyond. Oxford: Oxbow Books; 2007. p. 393–406.
35. Trentacoste A. Etruscan foodways and demographic demands: contextualizing protohistoric livestock
husbandry in Northern Italy. Eur J Archaeol. 2016; 19(2):279–315.
36. De Grossi Mazzorin J. Economie di allevamento in Italia centrale dalla media etàdel Bronzo alla fine
dell’etàdel Ferro. In: Christie N, editor. Settlement and economy in Italy, 1500 BC to AD 1500: papers
of the 5th conference of Italian archaeology. Oxford: Oxbow Books; 1995. p. 167–77.
37. De Grossi Mazzorin J, Minniti C. L’Utilizzazione degli animal nella documentazione archeozoologica a
Roma e nel Lazio dalla preistoria recente all’etàclassica. In: Drago Troccoli L, editor. Il Lazio dai Colli
Albani ai Monti Lepini tra preistoria ed etàmoderna. Rome: Edizioni Quasar; 2009. p. 39–67.
38. Farello P. L’Emilia dal VI e V secolo aC.: caccia e allevamento. In: Atti del I˚ convegno nazionale di
archeozoologia: Rovigo 5–7 marzo 1993. Rovigo: Centro Polesano di Studi Storici, Archeologici ed
Etnografici; 1995. p. 209–34. (Padusa quaderni; 1).
39. De Grossi Mazzorin J, Minniti C. Changes in lifestyle in ancient Rome (Italy) across the Iron Age/
Roman transition. In: Albarella U, Rizzetto M, Russ H, Vickers K, Viner-Daniels S, editors. The Oxford
handbook of zooarchaeology. Oxford: Oxford University Press; 2017. p. 127–46.
40. MacKinnon M. Production and consumption of animals in Roman Italy: integrating the zooarchaeologi-
cal and textual evidence. Portsmouth: Journal of Roman Archaeology; 2004. (Journal of Roman
Archaeology supplementary series; 54)
41. Love E. The evolution of animal husbandry and society in the backcountry of ancient Italy. M.Sc. The-
sis, University of Sheffield; 2008.
42. Minniti C. Ambiente, sussistenza e l’articolazione sociale nell’Italia centrale tra Bronzo medio e primo
Ferro. Oxford: Archaeopress; 2012. (BAR international series; 2394).
43. De Grossi Mazzorin J. Archaeozoology and habitation models: from a subsistence to a productive
economy in central Italy. In: Brandt JR, Karlsson L, editors. From huts to houses transformations of
ancient societies. Stockholm: Paul Astro
¨ms Fo
¨rlag; 2001. p. 323–30.
44. De Grossi Mazzorin J, Riedel A, Tagliacozzo A. L’Evoluzione delle popolazioni animalie l’economia
nell’etàdel Bronzo recente. In: Cocchi Genick D, editor. L’etàdel Bronzo recente in Italia: atti del con-
gresso nazionale di Lido di Camaiore, 26–29 ottobre 2000. Viareggio: M. Baroni; 2004. p. 227–32.
45. Riedel A, Tecchiati U. Insediamenti ed economia nell’etàdel bronzo e del ferro in Trentino Alto Adige.
Appunti per un modello archeozoologico. In: Atti della XXXIII Riunione Scientifica dell’Istituto Italiano
di Preistoria e Protostoria: preistoria e protostoria del Trentino Alto Adige/Su¨dtirol. Florence: Istituto
Italiano di Preistoria e Protostoria; 2002. p. 117–30.
46. De Grossi Mazzorin J. Considerazioni sullo sfruttamento animale in ambito terramaricolo. In: De
Grossi Mazzorin J, Curci A, Giacobini G, editors. Economia e ambiente nell’Italia padana nell’etàdel
Bronzo: le indagini bioarcheologiche. Bari: Edipuglia; 2013. p. 257–63. (Beni archeologici—conos-
cenza e tecnologie; 11).
47. De Grossi Mazzorin J. Considerazioni sullo sfruttamento animale nell’area delle palafitte. In: De Grossi
Mazzorin J, Curci A, Giacobini G, editors. Economia e ambiente nell’Italia padana nell’etàdel Bronzo:
le indagini bioarcheologiche. Bari: Edipuglia; 2013. p. 155–60. (Beni archeologici—conoscenza e tec-
nologie; 11).
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 28 / 33
48. Maini E, Curci A. Considerazioni sull’economia di allevamento nella Romagna durante l’etàdel
Bronzo. In: De Grossi Mazzorin J, Curci A, Giacobini G, editors. Economia e ambiente nell’Italia
padana nell’etàdel Bronzo: le indagini bioarcheologiche. Bari: Edipuglia; 2013. p. 357–76. (Beni
archeologici—conoscenza e tecnologie; 11).
49. Petrucci G. Lo sfruttamento della fauna del Friuli occidentale tra L’etàdel Ferro e la romanizzazione.
In: Malnati L, Croce Da Villa P, Di Filippo Balestrazzi E, editors. La protostoria tra Sile e Tagliamento:
antiche genti tra Veneto e Friuli. Padova: Esedra editrice; 1996. p. 469–76.
50. Bertolini M, Zanini S, Thun Hohenstein U. Nuovi dati sullo sfruttamento e gestione delle risorse animali
tra il Bronzo antico ed il Bronzo recente nei territori del medio-basso Veronese e il basso Polesine. In:
Leonardi G, Tine
´V, editors. Preistoria e protostoria del Veneto. Florence: Istituto Italiano di Preistoria
e Protostoria; 2015. p. 321–6. (Studi di preistoria e protostoria; 2).
51. Curci A. I dati archeozoologici. In: Govi E, Sassatelli G, editors. Marzabotto: La casa della regio IV—
insula 2. Vol 2, i materiali. Bologna: Ante Quem; 2010. p. 397–420.
52. Farello P. Caccia, pesca e allevamento nell’Etruria padana dall’VIII al IV secolo a.C. In: Curci A, Vitali
D, editors. Animali tra uomini e dei: archeozoologia del mondo preromano: atti del convegno interna-
zionale 8–9 novembre 2002. Bologna: Ante Quem; 2006. p. 97–109. (Studi e scavi, nuova serie; 14).
53. Rubel F, Brugger K, Haslinger KIA. The climate of the European Alps: shift of very high resolution Ko
pen–Geiger climate zones 1800–2100. Meteorologische Zeitschrift. 2017; 26(2):115–25.
54. Pesando F. L’Italia antica: culture e forme del popolamento nel I millennio a.C. Rome: Carocci; 2005.
55. Bandelli G. Considerazioni storiche sull’urbanizzazione cisalpina di etàrepubblicana (283–89 a.C.).
In: Brecciaroli Taborelli L, editor. Forme e tempi dell’urbanizzazione nella Cisalpina (II secolo aC–I
secolo dC): atti delle giornate di studio, Torino, 4–6 maggio 2006. Borgo San Lorenzo: All’Insegna del
Giglio; 2007. p. 15–28.
56. Cardarelli A. The collapse of the terramare culture and growth of new economic and social system dur-
ing the late Bronze Age in Italy. Scienze dell’antichità. 2009; 15:449–520.
57. Bietti Sestieri AM. L’Italia nell’etàdel bronzo e del ferro: dalle palafitte a Romolo (2200–700 a.C.).
Rome: Carocci editore; 2010.
58. Govi E. Etruscan urbanism at Bologna, Marzabotto and in the Po Valley. In: Robinson EC, editor.
Papers on Italian urbanism in the first millennium BC. Portsmouth, Rhode Island: Journal of Roman
Archeology; 2014. p. 81–111. (Journal of Roman Archeology Supplementary Series; 97).
59. Balista C, Leonardi G. Le strategie d’insediamento tra II e inizio I millennio a.C. in Italia settentrionale e
centro-orientale. In: Bianchi P, editor. Atti della XXXV Riunione Scientifica: le comunitàdella preistoria
italiana studi e richerche sul neolitico e le etàdei metalli. Florence: Istituto Italiano di Preistoria e Pro-
tostoria; 2003. p. 159–72.
60. Pearce M. New research on the terramare of northern Italy. Antiquity. 1998; 72(278):743–6.
61. Cremaschi M, Pizzi C. Water resources in the Bronze Age villages (terramare) of the north Italian Po
plain: recent investigation at Terramara Santa Rosa di Poviglio. Antiquity Project Gallery. 2011; 85
62. Cremaschi M, Pizzi C, Valsecchi V. Water management and land use in the terramare and a possible
climatic co-factor in their abandonment: the case study of the terramara of Poviglio Santa Rosa (north-
ern Italy). Quat Int. 2006; 151:87–98.
63. Mercuri AM, Accorsi CA, Mazzanti MB, Bosi G, Cardarelli A, Labate D, et al. Economy and environ-
ment of Bronze Age settlements–terramaras–on the Po Plain (Northern Italy): first results from the
archaeobotanical research at the Terramara di Montale. Veg Hist Archaeobot. 2006; 16:23–60.
64. Ravazzi C, Marchetti M, Zanon M, Perego R, Quirino T, Deaddis M, et al. Lake evolution and land-
scape history in the lower Mincio River valley, unravelling drainage changes in the central Po Plain (N-
Italy) since the Bronze Age. Quat Int. 2013; 288:195–205.
65. Mercuri AM, Mazzanti MB, Torri P, Vigliotti L, Bosi G, Florenzano A, et al. A marine/terrestrial integra-
tion for mid-late Holocene vegetation history and the development of the cultural landscape in the Po
valley as a result of human impact and climate change. Veg Hist Archaeobot. 2012; 21(4):353–72.
66. Vescovi E, Kaltenrieder P, Tinner W. Late-Glacial and Holocene vegetation history of Pavullo nel
Frignano (Northern Apennines, Italy). Rev Palaeobot Palynol. 2010; 160(1):32–45.
67. Valsecchi V, Tinner W, Finsinger W, Ammann B. Human impact during the Bronze Age on the vegeta-
tion at Lago Lucone (northern Italy). Veg Hist Archaeobot. 2006; 15(2):99–113.
68. Cremaschi M, Nicosia C, Salvioni M. L’uso del suolo nell’Eneolitico e nel Bronzo antico, nuovi dati
dalla Pianura Padana centrale. In: Atti della XLIII Riunione Scientifica: l’etàdel rame in Italia. Florence:
Istituto italiano di preistoria e protostoria; 2011. p. 225–31.
69. Mercuri AM, Montecchi MC, Pellacani G, Florenzano A, Rattighieri E, Cardarelli A. Environment,
human impact and the role of trees on the Po plain during the Middle and Recent Bronze Age: pollen
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 29 / 33
evidence from the local influence of the terramare of Baggiovara and Casinalbo. Rev Palaeobot Paly-
nol. 2015; 218:231–49.
70. MihovilićK. Castellieri-Gradine of the Northern Adriatic In: Harding A, Fokkens H, editors. The Oxford
handbook of the European Bronze Age. Oxford: Oxford University Press; 2013. p. 863–76.
71. Fiorentino G, Castiglioni E, Rottoli M, Nisbet R. Le colture agricole in Italia nel corso dell’etàdel
Bronzo: sintesi dei dati e linee di tendenza. In: Cocchi Genick D, editor. L’etàdel bronzo recente in Ita-
lia: atti del congresso nazionale di Lido di Camaiore, 26–29 ottobre 2000. Viareggio: M. Baroni; 2004.
p. 219–26.
72. Carra M. Le indagini archeobotaniche nell’area dell’Emilia orientale e della Romagna. In: De Grossi
Mazzorin J, Curci A, Giacobini G, editors. Economia e ambiente nell’Italia padana nell’etàdel Bronzo:
le indagini bioarcheologiche. Bari: Edipuglia; 2013. p. 239–356. (Beni archeologici—conoscenza e
tecnologie; 11).
73. Tafuri MA, Craig OE, Canci A. Stable isotope evidence for the consumption of millet and other plants
in Bronze Age Italy. Am J Phys Anthropol. 2009; 139(2):146–53.
PMID: 19051259
74. Varalli A, Moggi-Cecchi J, Dori I, Boccone S, Bortoluzzi S, Salzani P, et al. Dietary continuity vs. dis-
continuity in Bronze Age Italy. The isotopic evidence from Arano di Cellore (Illasi, Verona, Italy). J
Archaeol Sci Rep. 2016; 7:104–13.
75. Peroni R. Protostoria dell’Italia continentale: la penisola italiana nelle etàdel bronzo e del ferro. Rome:
Biblioteca di storia patria; 1989.
76. Mercuri AM, Mazzanti M, Bosi G, Cardarelli A, Olmi L, Torri P. Cereal fields from the Middle-Recent
Bronze Age, as found in the Terramara di Montale, in the Po Plain (Emilia Romagna, Northern Italy),
based on pollen, seeds/fruits and microcharchoals. In: Morel J-P, Tresserras JJ, Matamala JC, edi-
tors. The archaeology of crop fields and gardens. Bari: Edipuglia; 2006. p. 251–70.
77. Perego R. Contribution to the development of the Bronze Age plant economy in the surrounding of the
Alps: an archaeobotanical case study of two Early and Middle Bronze Age sites in northern Italy (Lake
Garda region). Ph.D. Thesis, Universit of Basel; 2017.
78. Jacomet S, Brombacher C. Geschichte der Flora in der Regio Basiliensis seit 7500 Jahren: Ergeb-
nisse von Untersuchungen pflanzlicher Makroreste aus archa
¨ologischen Ausgrabungen. Mitteilungen
der Naturforschenden Gesellschaften beider Basel. 2009; 11:27–106.
79. Bogaard A. Farming practice and society in the central European Neolithic and Bronze Age: an
archaeobotanical response to the secondary products revolution model. In: Hadjikoumis A, Robinson
E, Viner S, editors. The dynamics of neolithisation in Europe: studies in honour of Andrew Sherratt.
Oxford: Oxbow Books; 2011. p. 266–83.
80. Harding A. Bronze agricultural implements in Bronze Age Europe. In: de G. Sieveking G, Longworth
IH, Wilson KE, editors. Problems in economic and social archaeology. London: Duckworth; 1976. p.
81. Pearce M. Hard cheese: upland pastoralism in the Italian Bronze and Iron Ages. In: Collis J, Pearce M,
Nicolis F, editors. Summer farms: seasonal exploitation of the uplands from prehistory to the present.
Sheffield: JR Collis Publications; 2016. p. 47–56. (Sheffield archaeological monographs; 16).
82. Karg S. Winter- and spring-foddering of sheep/goat in the Bronze Age site of Fiavè-Carera, Northern
Italy. Environmental Archaeology. 1998; 1(1):87–94.
83. Cardarelli A. Terramare: l’organizzazione sociale e politica delle comunità. In: BernabòBrea M, Car-
darelli A, Cremaschi M, editors. Le terramare: la piu
`antica civiltàpadana. Milan: Electa; 1997. p.
84. CupitòM. Tipocronologia del bronzo medio e recente tra l’Adige e il Mincio sulla base delle evidenze
funerarie. Padova: Imprimatur; 2006.
85. Cremaschi M, Mercuri AM, Torri P, Florenzano A, Pizzi C, Marchesini M, et al. Climate change versus
land management in the Po Plain (Northern Italy) during the Bronze Age: new insights from the VP/VG
sequence of the terramara Santa Rosa di Poviglio. Quat Sci Rev. 2016; 136:153–72.
86. Leonardi G. Premesse sociali e culturali alla formazione dei centri protourbani del Veneto. Bollettino di
Archeologia Online, Volume speciale F/F1/4, 2010:23–35.
87. CupitòM, Dalla Longa E, Donadel V, Leonardi G. Resistances to the 12th century BC crisis in the
Veneto region: the case studies of Fondo Paviani and Montebello Vicentino. In: Kneisel J, Kirleis W,
Dal Corso M, Taylor N, Tiedtke V, editors. Collapse or continuity? Environment and development of
Bronze Age human landscape. Bonn: Verlag Dr. Rudolf Habelt GmbH; 2012. p. 55–70.
88. Vicenzutto D. L’etàdel bronzo in Italia settentrionale. In: Tasca G, Putzolu C, Vicenzutto D, editors. Un
castelliere nel medio Friuli: Gradiscje di Codroipo, 2004–2014. Pasian di Prato: IPAC; 2015. p. 16–
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 30 / 33
89. Bonetto J. Veneto. Rome: Libreria dello Stato, Istituto poligrafico e Zecca dello Stato; 2009.
90. Bietti Sestieri AM, Bellintani P, Salzani L, Ivana A, Chiaffon B, De Grossi Mazzorin J, et al. Frattesina:
un centro internazionale di produzione e di scambio nell’Etàdel bronzo del Veneto. In: Leonardi G,
´V, editors. Preistoria e protostoria del Veneto. Florence: Istituto Italiano di Preistoria e Proto-
storia; 2015. p. 389–400. (Studi di preistoria e protostoria; 2).
91. Balista C, Gambacurta G, Ruta Serafini A. Sviluppi di urbanistica atestina. In: Ruta Serafini A, editor.
Este preromana: una cittàe i suoi santuari. Treviso: Canova; 2002. p. 105–21.
92. Biba T, de Marinis RC. The Northern Adriatic. In: Colin H, Katharina R-S, Peter SW, editors. The
Oxford handbook of the European Iron Age [Internet]. Oxford: Oxford University Press; 2018. Avail-
able from:
93. Sassatelli G. Le cittàdell’Etruria padana. In: Bruni S, editor. Gli Etruschi delle città: Fonti, ricerche e
scavi. Milano: Silvana Editoriale; 2010. p. 204–15.
94. Rottoli M, Fanetti D, Bosi G, Castiglioni E, editors. The agriculture in Northern Italy during Iron Age: a
review. IWGP 2016 Paris: International Work-Group for Palaeoethnobotany 2016; 2016; Paris: Inter-
national Work Group for Palaeoethnobotany. Available online: https://iwgp2016paris.sciencesconf.
95. Giulierini P. Ereditàdel mondo rurale etrusco nell’agricoltura toscana. In: Il mondo rurale etrusco. Flor-
ence: Consiglio regionale della Toscana; 2009. p. 73–104. (Edizioni dell’Assemblea; 22).
96. Styring A, Ro
¨sch M, Stephan E, Stika H-P, Fischer E, Sillmann M, et al. Centralisation and long-term
change in farming regimes: comparing agricultural practices in Neolithic and Iron Age south-west Ger-
many. Proceedings of the Prehistoric Society. 2017; 83:357–81.
97. Delpino F. Viticoltura, produzione e consumo del vino nell’Etruria protostorica. In: Ciacci A, Rendini P,
Zifferero A, editors. Archeologia della vite e del vino in Toscana e nel Lazio: dalle tecniche dell’indagine
archeologica alle prospettive della biologia molecolare. Florence: All’Insegna del Giglio; 2012. p.
189–99. (Quaderni del Dipartimento di archeologia e storia delle arti, Sezione archeologica, Università
di Siena; 65).
98. de Grummond NT. Grape pips from Etruscan and Roman Cetamura del Chianti: on stratigraphy, liter-
ary sources and pruning hooks. Etruscan Studies. 2018; 21(1–2):
99. Gras M. Trafics tyrrheniens archaïques. Rome: Ecole franc¸aise de Rome; 1985.
100. Perkins P. Production and commercialization of Etruscan wine in the Albegna Valley. In: Ciacci A,
Rendini P, Zifferero A, editors. Archeologia della vite e del vino in Toscana e nel Lazio: dalle tecniche
dell’indagine archeologia alle prospettive della biologia molecolare. Borgo San Lorenzo: All’Insegna
del Giglio; 2012.
101. McGovern PE, Luley BP, Rovira N, Mirzoian A, Callahan MP, Smith KE, et al. Beginning of viniculture
in France. Proceedings of the National Academy of Sciences. 2013; 110(25): 10147–10152.
102. Frey O-H. The Celts in Italy. In: Green MJ, editor. The Celtic world. London: Routledge; 1995. p. 515–
103. Williams JHC. Beyond the Rubicon: Romans and Gauls in Republican Italy. Oxford: Oxford University
Press; 2001.
104. Bourdin S. Les peuples de l’Italie pre
´romaine: identite
´s, territoires et relations inter-ethniques en Italie
centrale et septentrionale (VIIIe-Ier s. av. J.-C.) Rome: E
´cole franc¸aise de Rome; 2012.
105. Gabba E. Rome and Italy in the second century B.C. In: Astin AE, Walbank FW, Frederiksen MW,
Ogilvie RM, editors. The Cambridge ancient history. Vol. 8, Rome and the Mediterranean to 133 BC.
Cambridge: Cambridge University Press; 1989. p. 197–243.
106. Barker G. Animals, ritual and power in ancient Samnium. In: Meniel P, editor. Animal et pratiques reli-
gieuses: les manifestations mate
´rielles. Actes du colloque international de Compiègne, 11–13 novem-
bre 1988. 1989. p. 111–7. (Anthropozoologica, troisième nume
´ro spe
107. De Grossi Mazzorin J, Minniti C. Testimonianze di pratiche cultuali nella Grotta 10 di Sorgenti della
Nova: recenti analisi sul materiale osteologico. In: Negroni Catacchio N, editor. Preistoria e protostoria
in Etruria: atti del quinto incontro di studi: paesaggi d’acque, ricerche e scavi; vol. 2. Milano: Centro
Studi di Preistoria e Archeologia; 2002. p. 627–36.
108. Lyman RL. Quantitative palaeozoology. Cambridge: Cambridge Univeristy Press; 2008.
109. Morin E, Ready E, Boileau A, Beauval C, Coumont M-P. Problems of identification and quantification
in archaeozoological analysis, part II: presentation of an alternative counting method. J Archaeol
Method Theory. 2016:1–36.
110. Morin E, Ready E, Boileau A, Beauval C, Coumont M-P. Problems of identification and quantification
in archaeozoological analysis, part I: insights from a blind test. J Archaeol Method Theory. 2016:1–52.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 31 / 33
111. Grayson DK. On the quantification of vertebrate archaeofaunas. Advances in archaeological method
and theory. 1979; 2:199–237.
112. R Core Team. R: a language and environment for statistical computing. Version 3.5.0 [software]. R
Foundation for Statistical Computing; 2018 23 April. Available from:
113. Meadow R. The use of size index scaling techniques for research on archaeozoological collections
from the Middle East. In: Becker C, Manhart H, Peters J, Schibler J, editors. Historia Animalium ex
Ossibus Festschrift fu¨r Angela von den Driesch. Rahden/Westf.: Verlag Marie Leidorf GmbH; 1999.
p. 285–300.
114. Nieto-Espinet A. Element measure standard biometrical data from a cow dated to the Early Bronze
Age (Minferri, Catalonia) [digital resource]. 2018. Available from:
115. Davis S. Measurements of a group of adult female shetland sheep skeletons from a single flock: a
baseline for zooarchaeologists. J Archaeol Sci. 1996; 23:593–612.
116. Albarella U, Payne S. Neolithic pigs from Durrington Walls, Wiltshire, England: a biometrical database.
J Archaeol Sci. 2005; 32(4):589–99.
117. Salvagno L, Albarella U. A morphometric system to distinguish sheep and goat postcranial bones.
PLOS ONE. 2017; 12(6):e0178543. PMID: 28594831
118. von den Driesch A. A guide to the measurement of animal bones from archaeological sites. Harvard:
Peabody Museum Bulletin 1; 1976.
119. Bartosiewicz L. Cattle metapodials revisited: a brief review. Archaeozoologia 1987; 1:47–51.
120. Higham C. The metrical attributes of two samples of bovine limb bones. J Zool. 1969; 157:63–74.
121. Thomas R. A statistical evaluation of criteria used in sexing cattle metapodia. Archaeozoologia 1986;
122. Albarella U. Shape variation of cattle metapodials: age, sex or breed? Some examples from medieval
and post medieval sites. Anthropozoologica 1997; 25–26:37–4
123. Telldahl Y, Svensson EM, Go
¨m A, StoråJ. Osteometric and molecular sexing of cattle metapo-
dia. J Archaeol Sci. 2012; 39(1):121–7.
124. Davis S. The effect of castration and age on the development of the Shetland sheep skeleton and a
metric comparison between bones of males, females and castrates. J Archaeol Sci. 2000; 27(5):373–
125. Nicosia C, Balista C, CupitòM, Ertani A, Leonardi G, Nardi S, et al. Anthropogenic deposits from the
Bronze Age site of Fondo Paviani (Verona, Italy): Pedochemical and micropedological characteristics.
Quat Int. 2011; 243(2):280–92.
126. De Grossi Mazzorin J. Fondo Paviani e Frattesina: economia animale di due central places della tarda
etàdel bronzo veneta. In: Leonardi G, Tine
´V, editors. Preistoria e protostoria del Veneto. Florence:
Istituto Italiano di Preistoria e Protostoria; 2015. p. 389–400. (Studi di preistoria e protostoria; 2).
127. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A. Very high resolution interpolated climate sur-
faces for global land areas. Int J Climatol. 2005; 25(15):1965–78.
128. Giuliano G, Mari GM, Cavallin A, De Amicis M. Ricerca sulla vulnerabilitànaturale e sul rischio di inqui-
namento delle acque sotterranee nella pianura padana e veneto friulana: carta della infiltrabilitàregio-
nale, carta idrogeologica regionale, carta della vulnerabilitàregionale (scala 1:500.000). Rome:
Istituto Poligrafico e Zecca dello Stato; 1998.
129. Fontana A, Mozzi P, Bondesan A. Alluvial megafans in the Venetian–Friulian Plain (north-eastern
Italy): evidence of sedimentary and erosive phases during Late Pleistocene and Holocene. Quat Int.
2008; 189(1):71–90.
130. Castiglioni GB. Geomorphology of the Po plain. Geografia Fisica e Dinamica Quaternaria. 1999;
Suppl. III(3):7–20.
131. Marchetti M. Environmental changes in the central Po Plain (northern Italy) due to fluvial modifications
and anthropogenic activities. Geomorphology. 2002; 44(3):361–73.
132. Fre
´mondeau D, Horard-Herbin M-P, Buchsenschutz O, Ughetto-Monfrin J, Balasse M. Standardized
pork production at the Celtic village of Levroux Les Arènes (France, 2nd c. BC): Evidence from kill-off
patterns and birth seasonality inferred from enamel δ18O analysis. J Archaeol Sci Rep. 2015; 2:215–
133. Bietti Sestieri AM. L’etàdel bronzo finale nella penisola italiana. Padusa. 2008; 44:7–54.
134. Albarella U, Tagliacozzo A, Dobney K, Rowley-Conwy P. Pig hunting and husbandry in prehistoric
Italy: a contribution to the domestication debate. Proceedings of the Prehistoric Society. 2006;
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 32 / 33
135. MacKinnon M. High on the hog: linking zooarchaeological, literary, and artistic data for pig breeds in
Roman Italy. American Journal of Archaeology. 2001; 105(4):649–73.
136. Wealleans AL. Such as pigs eat: the rise and fall of the pannage pig in the UK. J Sci Food Agric. 2013;
93(9):2076–83. PMID: 23553313
137. Albarella U, Manconi F, Trentacoste A. A week on the plateau: pig husbandry, mobility and resource
exploitation in central Sardinia In: Albarella U, Trentacoste A, editors. Ethnozooarchaeology: the pres-
ent and past of human–animal relationships. Oxford: Oxbow Books; 2011. p. 143–59.
138. Halstead P, Isaakidou V. A pig fed by hand is worth two in the bush: ethnoarchaeology of pig hus-
bandry in Greece and its archaeological implications. In: Albarella U, Trentacoste A, editors. Ethno-
zooarchaeology: the present and past of human–animal relationships. Oxford: Oxbow Books; 2011.
p. 161–74.
139. Mattioli S, Pedone P. Body growth in a confined wild boar population. Journal of Mountain Ecology.
1995; 3:64–5.
140. Dickerson JWT, McCance RA. Severe undernutrition in growing and adult animals. 8. The dimensions
and chemistry of the long bones. Brit J Nutrition. 1961; 15:567–76.
141. Weinstock J. Late Pleistocene reindeer populations in Middle and Western Europe: an osteometrical
study of Rangifer tarandus. Tubingen: Mo Vince Verlag; 2000.
142. Pa
´lsson H, Verge
´s JB. Effects of the plane of nutrition on growth and the development of carcass qual-
ity in lambs. Part I. The effects of high and low planes of nutrition at different ages. J Agric Sci. 2009;
143. Ahlborn G, Dempfle L. Genetic parameters for milk production and body size in New Zealand Holstein-
Friesian and Jersey. Livestock Production Science. 1992; 31(3):205–19.
144. Reitmaier T, Doppler T, Pike AWG, Deschler-Erb S, Hajdas I, Walser C, et al. Alpine cattle manage-
ment during the Bronze Age at Ramosch-Mottata, Switzerland. Quat Int. 2018; 484:19–31.
145. Bietti Sestieri AM. Italy in Europe in the Early Iron Age. Proceedings of the Prehistoric Society. 1997;
146. Redding RW. The pig and the chicken in the Middle East: modeling human subsistence behavior in
the archaeological record using historical and animal husbandry data. Journal of Archaeological
Research. 2015; 23(4):325–68.
147. deFrance SD. Zooarchaeology in complex societies: political economy, status, and ideology. Journal
of Archaeological Research. 2009; 17(2):105–68.
148. Crabtree P. Zooarchaeology and complex societies: some uses of faunal analysis for the study of
trade, social status, and ethnicity. In: Schiffer MB, editor. Archaeological method and theory. Tucson:
University of Arizona Press; 1990. p. 155–205.
149. Nijboer AJ. From household production to workshops: archaeological evidence for economic transfor-
mations, pre-monetary exchange and urbanisation in central Italy from 800 to 400 BC. Groningen:
Groningen Institute of Archaeology; 1998.
150. Economy Nijboer A., 10th cent.-730 BCE. In: Naso A, editor. Etruscology. Berlin: De Gruyter; 2017.
p. 795–810.
151. Mattingly DJ. Oil for export? A comparison of Libyan, Spanish and Tunisian olive oil production in the
Roman empire. Journal of Roman Archaeology. 2015; 1:33–56.
152. Ward-Perkins B. The fall of Rome and the end of civilization. Oxford: Oxford University Press; 2006.
153. Giardino C, Belardelli C, Malizia A. Power and the individual in funerary ideology: the emergence of
the aristocracy in the Villanovan period in the Bologna region. In: Herring E, Whitehouse R, Wilkins J,
editors. Papers of the fourth conference of Italian archaeology. Vol. 2, The archaeology of power; part
2. London: Accordia Research Centre; 1991. p. 9–19.
154. Pacciarelli M. Dal villaggio alla città: la svolta protourbana del 1000 a.C. nell’Italia tirrenica. Florence:
All’Insegna del Giglio; 2001.
Evidence for pre-Roman improvements to livestock husbandry in late prehistoric Italy
PLOS ONE | December 31, 2018 33 / 33
... Measurements of bone depth were comparatively rare and not included. The available reference standards were used by default in the Zoolog package for sheep and goats (Clutton-Brock et al., 1990), cattle (Trentacoste et al., 2018), and pigs (Stopp and Deschler-Erb, 2018). The standard used for sheep (Ovis aries) is the mean measurement of a group of male Soay sheep of known age (Clutton-Brock et al., 1990). ...
... The standard for goats (Capra hircus) consists of the mean measurements of a group of a feral flock of goats of unknown age and sex collected on the Holly Island of Scotland (Clutton-Brock et al., 1990). The cattle (Bos taurus) standard is a female cow dated to the Early Bronze Age from Minferri in Catalonia (Trentacoste et al., 2018). Finally, the standard used for pigs (Sus scrofa) is a museum reference specimen from Basel (Inv. ...
... In 2016, two comprehensive studies with a collection of archaeozoological data for two Italian regions were published: a study of meat consumption in the early Middle Ages Central Italy (Rome and Latium) (Minniti, 2016) and a study of dietary habits in northern Italy . Northern Italy was also the focus of a comparative analysis by a European Research Council funded project led by Silvia Valenzuela-Lamas (Trentacoste et al., 2018;. This explored changing farming strategies from the Bronze Age to Late Antiquity, and was underpinned by a study of biometrical aspects of faunal remains. ...
The RomAniDat data community is a network of experts from diverse fields including archaeozoology, archaeology, and history, who collaborate on the collection of faunal data from the Roman period. Compiled data is made available via Pandora, a new technological platform that allows for the building of data communities where its members directly manage and curate data. RomAniDat members can submit data compilations with different structural metadata and assign to these a DOI. It is also possible to list external data compilations with previously assigned DOIs. Under RomAniDat, generalized metadata schemas are being developed to offer users standardized data views which combine separate compilations. These include references to both primary publications and data compilations. Here, we present a first metadata schema to visualize combined data from publications reporting on the distribution of faunal assemblages. We also present an archaeofaunal compilation for Roman Italy and briefly illustrate its research potential. RomAniDat can be used to identify data gaps and set future research agendas, as a tool for cultural heritage protection, and for multiple research purposes at various temporal and spatial scales.
... In addition, it has been proposed to use only one measurement per bone and one bone per articulating group of bones to avoid including multiple measurements from a single specimen (Trentacoste et al., 2018(Trentacoste et al., , 2021Pozo et al., 2022). However, because in archaeozoology it is often impossible to identify all the bones that belong to single animals, and because we have demonstrated that LSI values obtained from measurements of a single articulated skeleton can greatly vary (Jeanjean et al., 2022), and that no rationale exists (to our knowledge) for selecting the measurement to retain and its influence on the LSI analyses, our analyses are based on all available measurements. ...
Log Size Index (LSI) has been developed and used in archaeozoology to perform biometric studies of often small datasets, a recurrent limitation of archaeological faunal assemblages. This approach consists of aggregating measurements taken on different anatomical parts by standardisation, using a reference to which the studied measurements are scaled. No consensus exists in the choice of the reference to be used for LSI analyses and several datasets are now available, differing in their specimens and variables composition, and geographic and temporal origin. Using a biometric dataset of 5533 measurements of cattle, suids, sheep and goat from southern France and dated from the Roman times we compared the use of several references, provided by the R package zoolog, with a ‘mean’ reference specifically built for our studied sample and that corresponds to an averaged specimen. We demonstrated that references differ both in their raw measurements and their allometries, and that the choice of the reference influences the number of variables and measurements that can be analysed, but also diachronic LSI analyses and potentially their interpretations. While the main patterns are similar among references, local variation exist and it could not be excluded that the choice of the reference will more strongly influence other studies. We advocate to use the mean of the studied archaeological assemblage as reference, with the double benefit of exploiting to the maximum the number of available measurements, and of minimising, on average, the possible allometric differences between the studied population and the reference.
Durant les âges du Fer, la vallée du Rhône se trouve à la jonction entre le monde celtique du nord et du centre de l’Europe et le monde méditerranéen. Elle constitue donc un espace de circulation soumis à différentes influences culturelles. L’étude des faunes archéologiques issues des sites de la région permet donc d’approcher les spécificités de l’élevage, de la chasse et de la consommation carnée des populations protohistoriques, en un mot, de la relation entre elles et les populations animales. Par le prisme de l’archéozoologie, la vallée du Rhône apparaît comme un espace parfaitement intégré au reste de la Gaule. Ainsi, au premier âge du Fer, les pratiques alimentaires et d’élevage montrent une partition de la région entre l’influence des mondes nord-alpin et celtique méditerranéen. Au second, les pratiques impliquant des animaux et des produits d’origine animale présentent des évolutions, parmi lesquelles une augmentation plus ou moins forte de la part des porcs dans l’alimentation. Ces changements correspondent à ceux qui s’observent dans l’ensemble de la Gaule et qui sont sans doute en lien avec une réorganisation de la société. La morphologie des espèces domestiques et son évolution, de même que la pratique de la chasse, sont également plus ou moins similaires à celles du reste de la Gaule, du VIIIe au Ier siècle av. J.-C. L’influence des cultures grecque et italique, en matière alimentaire, semble être anecdotique, du point de vue des produits d’origine animale, les évolutions semblant davantage liées à l’organisation et à la démographie des populations. L’alimentation rhodanienne nous offre donc l’image d’une région pleinement intégrée aux cultures alimentaires celtiques et présente des évolutions moins liées à celles de l’environnement qu’à des problématiques sociales et économiques.
Full-text available
Log Size Indexes (LSI) allow the increase of the number of data and have been used in a number of zooarchaeological studies since 1950. However, some standards to calculate the log ratios remain unpublished, the calculation of the indexes can be tedious, and it is further hindered by the diversity of data recording practices. The R package ‘zoolog’ enables calculation of thousands of log-ratios in seconds, with the advantage that the users can choose between different public references, which increases the repeatability and comparability of the results, allowing the smooth integration of references and databases with heterogeneous nomenclatures. Alternatively, the users may use their own references. This paper presents the main functionalities and procedures enabled by the package ‘zoolog’, together with some examples of use. A real dataset and several examples with R code are provided with the aim of facilitating osteometrical analyses in zooarchaeology.
Animal management is shaped by its environmental and landscape context, but these factors are rarely investigated quantitatively in zooarchaeological studies. Here we aim to examine the relationship between trends in zooarchaeological data and environmental and climatic dynamics between the Middle Bronze Age and Late Antiquity in lowland northern Italy (Po–Friulian Plain). This region provides an ideal test case to investigate the impact of landscape variables (precipitation, solar irradiance, elevation, soil characteristics) and climate evolution due to the area's relatively homogenous topography and climatic conditions. This study presents a new elaboration and visualisation of zooarchaeological data from northern Italy, investigates correlations between these data and landscape variables, and contextualises trends in relation to regional environmental and climate proxies at two scales. This analysis reveals a shift towards more heterogeneous livestock on a regional level during the Late Iron Age and Roman period, and strong evidence for a correlation between cattle representation and local soil characteristics at a site-level during the Bronze Age. Consideration of climate data shows little relationship between species representation, livestock body size, and climate proxies, indicating that human social dynamics rather than climate change were the primary driver to changes in animal management on the regional macro-scale.
Anatolia witnessed an increase in mobility and interaction during the Iron Age. Peoples from Southeastern Europe and Southwest Asia migrated into and across Anatolia. At the same time, under the influence of expanding and competing polities, trade and military mobility reinforced supra-regional networks across land and sea. Little is known about how animal husbandry practices in Iron Age Anatolia changed in the face of these large-scale movements of people and goods. Especially little understood remains how past mobility and connectivity influenced pig husbandry, as pigs are generally considered rather immobile animals, and Iron Age animal mobility studies tend to focus on pastoralist practices (e.g. Hammer and Arbuckle, 2017). Yet, pig husbandry practices are highly plastic, and may have changed dynamically along with evolving economic and socio-cultural circumstances during the first millennium BC. In addition, palaeogenetic studies provide evidence that durign the Iron Age pigs with European lineages appeared and subsequently spread over Anatolia (Ottoni et al., 2013) suggesting pigs may have been actively incorporated in trade and mobility. Building on these two observations, this paper explores pig husbandry practices over the course of the Anatolian Iron Age (1200–600 BCE) by (1) discussing diachronic change in relative abundance of pigs and mortality patterns over different sites in first millennium BC Anatolia; (2) investigating whether the introduction of European pigs in Anatolia coincided with noticeable phenotypic changes in pig populations by looking at pig biometry using the R package zoolog. Primary data from various key sites are presented (Troy, Klazomenai, Gordion, Kerkenes, Kinet Höyük) alongside a meta-analysis of published zooarchaeological data.
Full-text available
Il volume è strutturato in otto capitoli che intendono presentare al lettore entrambe le realtà culturali presenti nel territorio di Codroipo, lo scavo e il museo, sottolineando così l’importanza dell’osmosi tra i due ambiti per la creazione di un percorso che contempli non solo la ricerca archeologica, spesso appannaggio di pochi, ma anche la possibilità di divulgazione dei risultati a più ampio raggio. Un inquadramento generale dell’età del Bronzo nell’Italia settentrionale e in Friuli e una presentazione della metodologia dello scavo archeologico e dei vari strumenti informatici e digitali, adottati per la documentazione, fanno da premessa ai risultati delle dieci campagne di scavo e permettono anche ai lettori meno esperti una più agevole lettura e comprensione delle evidenze archeologiche. Il capitolo relativo allo scavo e all’illustrazione delle fasi cronologiche e tipologiche della ceramica del sito della Gradiscje rappresenta il cuore del volume e, sebbene si tratti di una sintesi destinata ad una prima divulgazione dei risultati, è molto denso e corredato di numerose planimetrie, sezioni, ricostruzioni tridimensionali, fotografie e disegni della ceramica che ne rendono la lettura molto interessante e utile anche per gli specialisti di settore.
Full-text available
Archaeological evidence for economic transformations, pre-monetary exchange and urbanization in central Italy from 800 to 400 BC The main topic of research are the technological and social-economic developments associated with the advance of the workshop mode of production. This may lead to the conclusion that too much emphasis is placed on functional facets of society. Though religious and political aspects play a major role in more complex societies, these topics have not been fundamentally examined during the investigations. Whenever appropriate they have been included in order to illustrate a shift in demand or to explain specific circumstances but they are not discussed per se. It must be emphasized that a pre-industrial society is being investigated and that agriculture remained the solid base on which the development presented could ensue. The changes in craft specialization should be considered as gradual changes with eventually major consequences. I will argue that a redirection of the production facilities is an intrinsic component of the centralization processes occurring in central Italy from 800 to 400 BC. They are embedded in the transitions from village to towns, from communal to private property and from tribal to state formation. The advance of the workshop mode of production was an active component of these cultural transformations. It was both the result of past developments as well as agent of conditions to come. The increase in craft specialization was, however, minor in terms of the percentage of the population who shifted their activities towards the production of commodities. By far most of the labor employed in central Italy during these four centuries, was still engaged in agriculture. This cannot be stressed enough since I do not want to impart the impression that I overestimate the extent to which the population of the proto-urban and urban centers in central Italy became engaged in industry and trade rather than agricultural activities.
Full-text available
This study uses two novel archaeobotanical techniques – crop carbon and nitrogen isotope analysis and functional weed ecology – to determine directly how the intensity of agricultural practice changed from the Neolithic to the Early Iron Age in south-west Germany, with the emergence of fortified hilltop settlements (Fürstensitze or chiefly seats) regarded as the first urban centres in central Europe. The crop isotope and functional weed ecological evidence suggest that surplus cereal production in the Early Iron Age was achieved through sustained use of manure combined with expansion in arable cultivation, both developments that are connected with more widespread use of animal traction. The increased scale of cultivation is broadly apparent across rural as well as fortified hilltop centres in the Early Iron Age, and considerable variability in manuring intensity is consistent with agricultural decision-making at a local level rather than centralised control. Additionally, the more intensive manuring of hulled six-row barley, used in beer production, demonstrates that the political importance of drinking and feasting in Early Iron Age society was reflected in crop husbandry practices. In terms of animal husbandry, faunal isotope data reveal a radical decrease in forest cover, potentially reflecting an expansion in the scale of herding accompanying that of arable cultivation. Site-specific patterning points to a range of herding strategies, from specialised herding of cattle at the Heuneburg to generalised patterns of livestock management at rural sites.
Recently an article appeared raising some issues about the interpretation of grape pips that were excavated at Cetamura del Chianti by the present writer (2012-14). This commentary makes suggestions concerning the arguments in that article with reference to 1) stratigraphy at the site; 2) literary sources on Etruscan viticulture; and 3) the use of the pruning hook by the Etruscans. The present article makes a contribution to the study of Etruscan viticulture by assembling an appendix on actual pruning hooks that have been discovered in Italy dating from the Late Bronze Age down to the second century B. C. E., as well as an appendix on representations of a youth holding the pruning hook in Etruscan art, mainly from the fourth and third centuries B. C. E.
This special issue of the European Journal of Archaeology discusses aspects of animal husbandry in a number of provinces of the Western Roman Empire. In this introduction, we describe the general characteristics of animal husbandry in pre-Roman and Roman times to assess any changes that may have occurred after the Roman conquest. The results suggest that the territoriality typifying the first millennium bc had a significant impact on production, resulting in a decrease in cattle size and frequencies across Europe. Nevertheless, not all the regions reacted in the same way, and regional communities that focused their animal production on pigs implemented more sustainable husbandry practices over time. By bringing together studies carried out across Europe, this journal issue highlights the existence of cases of both change and continuity across the Empire, and the (uneven) impact of the market economy on animal husbandry and dietary practices in climatically different regions.