Modelling Bronze Age sheepherding and wool production: the case of the Terramara settlement at Montale, Italy

Abstract Arguing for an integrated wool-textile economy in the Bronze Age, this paper assesses characteristics and scale of pastoral economy and sheepherding at the Terramare settlement of Montale (Modena province, Italy). Previous studies argued that Montale was a Bronze Age centre of wool production. The present work enhances the understanding of the local textile economy by investigating the evidence for sheepherding and landscape management at the site. It also proposes an interdisciplinary-based approach to investigate and reconstruct pastoral economy and sheepherding strategies in other prehistoric contexts as well.


Introduction
Arguing for an integrated wool-textile economy in the Bronze Age, this paper assesses characteristics and scale of pastoral economy and sheepherding at the Terramare settlement of Montale on the Po plain in northern Italy. Previous studies argued that Montale was a Bronze Age centre of wool production1. Archaeological evidence from the site in support of this argument includes the astonishing number of textile tools, in particular spindle whorls, but also loom weights2. Large numbers of sheep/goat bones among the faunal remains3 suggest that wool was the primary fibre for textile manufacture. The present work enhances the understanding of the local textile economy by investigating the evidence for sheepherding at the site. This paper also proposes an interdisciplinary-based approach to investigate and reconstruct characteristics and scale of pastoral economy and sheepherding strategies in other prehistoric contexts as well.

Wool in the Bronze Age
From the 2 nd millennium BC, the introduction of wool and of woollen textile production represented in all likelihood a major innovation in Bronze Age Europe. Archaeological and written evidence show that wool became appreciated and used during the 4 th millennium in the Near Eastern region4. Archival documents from Mesopotamia, Anatolia, and the Aegean dated to the 3 rd and 2 nd millennium BC describe that wool and woollen textile production was a complex, but highly rewarding economic activity5. Lacking written sources in Europe, the combined interdisciplinary study of archaeological and faunal remains can help us understand the role of wool and woollen textiles6.
In recent years, the concept of wool economy has emerged in Bronze Age studies7 to underline that engagement with the long chain of essential wool manufacturing processes8, would require -to be economically profitable -also attention to consumption patterns and involvement in trade, thus to be part of an overall political economic design. Raw wool itself, during the Bronze Age, appears to have been a precious and difficult raw material to obtain in large quantities, not to mention qualities9. Thanks to ancient written sources, we have rather precise indications as to quantity and quality of the wool obtainable from Bronze Age sheep in the eastern Mediterranean and Mesopotamia. Archive documents from the Aegean show for instance that the standard weight of a sheep fleece was 700-750 g, and mixed herds of ewes and lambs were expected to produce c. 300 g wool per animal10. The expected wool yield from ewes only would have been around 600 g11. Woollier sheep able to provide a yearly wool yield between 0.7 and 1.12 kg apparently existed in Mesopotamia, suggesting that at some locations, breeding practices improved specific productive traits12. Finally, Eastern Mediterranean wool economies had access to international wool trade and/or managed large numbers of sheep, which necessarily demanded well-organised herding strategies13.

The Terramare settlement at Montale
Montale was one of the many fortified Terramare settlements on the Po plain. It was c. 1 ha in size, and it was surrounded by a ditch c. 40 m wide and 3 m deep, which was filled with water from a nearby river likely to serve several functions including defence15. Terramare populations inhabited the central Po plain from the beginning of the Italian Middle Bronze Age until the end of the local Recent Bronze Age (c. 1700/1650-1200/1150 BC). By the end of the Middle Bronze Age, a site hierarchy emerged with several settlements that reach up to 10-20 ha in exten-sion16. At the end of the Recent Bronze Age, this apparently prosperous system collapses for reasons that are widely debated17. Differently from other prominent sites on the plain, Montale does not change its original size and is probably abandoned before the end of the Recent Bronze Age18. The remains of the Bronze Age village were partly investigated during the nineteenth century providing information in particular about the still preserved parts of the embankment once separating the village from the water-filled ditch19. Thanks to recent stratigraphic excavations (carried out between 1996 and 2001) of a c. 45 m 2 portion of the settlement spared during the manurequarry works in the nineteenth century20, a glimpse of the internal organization at the site and an eleven-phase internal chronology (see Tab. 1) was determined.
What makes Montale exceptional among Terramare settlements, and so far among European Bronze Age settlements in general21, is its extraordinarily high concentration of textile tools, particularly spindle whorls. Nearly 4500 spindle whorls are known from the site suggesting yarn production close to industrial scale22. The prevalence of sheep/goats among the faunal remains from the site has  Brea et al. 1997b;Cardarelli 2009b;Cremaschi et al. 2006. 18 Cardarelli/Labate 2009b19 Boni 18821884;Cardarelli/Labate 2009a. 20 During the nineteenth century, numerous manure quarries were opened in the central Po plain. As it became rapidly evident, those quarries of fertile soil were no more than the remains of Bronze Age settlements (e. g. Bernabò Brea/Mutti 1994;Cardarelli 2009a, 16-18;Saltini 1997). 21 In the rest of Europe, other centres of wool production, such as for instance Százhalombatta-Földvár, in the Hungarian plain (Bergerbrant 2018; Vretemark 2010), certainly existed (see also Belanová Štolcova/Grömer 2010; Kneisel/Schaefer 2019); however more focused research is needed to document them. 22 Sabatini et al. 2018a. been used to support the hypothesis that wool was the main fibre manufactured there23; however, so far issues of pastoral economy at the site have not been investigated in depth.

The Archaeozoology of wool
Like most organic material, wool does not normally preserve in the archaeological record. With the exception of a single textile fragment from Castione dei Marchesi (Parma province), likely dating to the Middle Bronze Age24, no other remains of textiles are known from Terramare settlements. Studies on the local wool economy must, therefore, rely on the archaeozoological evidence for wool-producing animals, and the archaeological evidence for wool technology.
Studies of prehistoric sheep/goats demonstrated that a significant variety existed in animal size and morphological traits across prehistoric Europe25. Although archaeozoology cannot offer information as to the woolliness of their fleeces, it can help understand prehistoric herds' management. The slaughtering age of the animals suggests their primary use26. Adult sheep in general, and wethers in particular, produce the largest amount of wool27; high numbers of such individuals in a flock becomes, therefore, solid evidence of wool production. A high number of slaughtered new-born animals would instead suggest attention to milk products, while a primary interest in meat consumption would give herds with kill-off patterns clustering between approximately 6/9 months and 2/3 years depending on access to winter fodder28.
Ancient written texts, in particular from Bronze Age Crete29, show that local managements were well-aware of the amount of wool obtainable from different categories of sheep. Specialized herds also existed, but it seems that even in the case of distinct Bronze Age wool productions such as those carefully recorded in Aegean archive tablets, a mixed economy strategy, exploiting various products beside wool, was preferred30. When written documents are missing, some critical research questions concerning reliability of inferences from zoo-archaeological evidence must be considered as to the relations between estimated culling patterns, husbandry practices and wool production: 1) To what extent do percentages of different taxa among animal remains in a given place mirror the ancient local domestic fauna? 2) In which ways can finding circumstances bias our understanding of the ancient animal population?
The representativeness of the archaeozoological remains has been discussed in a number of forums31. As to the study of Bronze Age wool economy, two specific aspects appear particularly relevant; the numerical dimension of the ancient faunal population and the relation between sheep and goats. The first problem regards the possibility to estimate the ancient animal population from excavated faunal assemblages. Archaeozoological evidence may be biased by various factors and so not provide a full picture of the relative proportions of ancient domesticated species32. Two major issues shall be addressed in the case of Montale. First, the available faunal remains, although abundant (c. 9000 determined remains)33, and well documented, coming from modern stratigraphic excavations34, docu-27 Barber 1991, 26-27. 28 Payne 1973, 282. 29 E. g. Del Freo et al. 2010Killen 1964. 30 E. g. Halstead 1998Payne 1973, 282. 31 E. g. Albarella 2017; Bartosiewicz et al. 2013; Peres 2010 32 E. g. Barber 1991, 27-28;Grömer/Saliari 2018, 128-129;Peres 2010. 33 De Grossi Mazzorin 2013. 34 Cardarelli 2009a ment behaviour from only a small sample of the ancient Bronze Age village's living area. They may, therefore, be the results of depositional events that cannot be straightforwardly representative for the whole site. Secondly, social, cultural, religious, and practical rules or taboos might have biased the consumption patterns revealed by excavated remains.
To assess the first problem, one should consider that the archaeological evidence from the Terramare area broadly show regular settlement configurations with tightly organized living quarters of similar dimensions, repeatedly rebuilt throughout site sequences35. Montale does not seem to be an exception36; therefore, it is likely that the data acquired for the excavated portion would recur in other parts of the site. On the other hand, animal bones from the settlement ought to be predominantly the remains of meat consumption. Our current understanding of the Terramare societies does not allow to determine whether social, cultural or religious constrains may have influenced eating practices, and the assemblage composition.
Aware of limitations intrinsic to size and circumstances of the available sample, the reconstruction of the ancient pastoral economy at Montale shall be attempted by combining the evidence from archaeological studies with archaeozoological and archaeobotanical investigations, Thiessen polygon models, and strontium isotope tracing analyses. Finally, an attempt to quantify wool production and animal demography is provided. It is proposed that the approach developed by this research has the potential to construct a model useful to investigate other areas and periods as well.

Montale's ancient animal population
Animal remains from the modern excavations at Montale have been preliminarily published37. They suggest that the ancient animal population was largely domestic and that sheep/goats likely prevailed throughout the site sequence. Cattle represented a stable c. 10 % of the bone assemblages and was typically kept until old age, thus likely used in agricultural work. Pigs' bone remains, which can be interpreted as a sign of mixed economy (pigs can be associated to the consumption of leftovers from agricultural produc-35 See Bernabò Brea et al. 1997a;Bernabò Brea et al. 2007;Ferrari et al. 2018. The available data about sheep/goat bone from Montale hint at a considerable animal population. Through out the sequence, their remains dominate the bone assemblages (see Tab. 2). Additionally, their presence seems to increase from approximately 47 % of the local faunal remains during the Middle Bronze Age 2 to c. 62 % of the total during the Recent Bronze Age 1. The sample is not exclusively sheep40; on the exclusive base of traditional archaeozoological methods, it has been estimated that goats represent c. 8 % of the domestic caprine population during most of the sequence (phases I-X, see Tab. 1), while during the last phase (XI) of the Recent Bronze Age 1, the percentage of goats doubles41. A similar pattern has also been observed in other Terramara sites where it has been interpreted as a sign of worsened environmental conditions42, since goats seem to have better chances than sheep to cope with adverse, drier circumstances. Across Eurasia since prehistory, sheep and goats have been kept together, although in varying proportions43. The reasons behind such pastoral strategies might be various44.
In order to propose a convincing picture of Montale's wool economy, only the estimated sheep population will be considered45, because the presence of goats might 38 cf. De Grossi Mazzorin 2013, 258. 39 De Grossi Mazzorin/Ruggini 2009. 40 Sheep and goats are close, but not equivalent species. Thanks to archaeozoology, and today to zooarchaeology by mass-spectrometry (ZooMS), and ancient DNA (aDNA) studies, it is possible to distinguish between their remains (e. g. Buckley et al. 2009;Campana et al. 2013;Sinding et al. 2015). While the ZooMS and the aDNA analyses tend to provide definitive results, archaeozoological analyses are limited in their scope by the fact that just a few bones can be used to distinguish between the two taxa (e.  1980;Greenfield 2005, 22. 44 Both the primary and the secondary products that can be obtained from them have different nutritional values and different harvesting patterns when eatable (like milk and meat) or different characteristics when the main interest is in their skin/fleeces (see Greenfield 2017). Both species can be used to produce wool (e. g. Barber 1991, 20-30;Del Freo et al. 2010, 340;Green 1980, 14), although goat hair has not yet been detected in Bronze Age textiles from continental Europe. Evidence from the Aegean and the Near East (Bender Jørgensen/ Rast-Eicher 2018, 34; Frangipane et al. 2009, 19-20;Michél 2012, 237) suggest that they were in use in neighbouring areas. 45 According to the available data from De Grossi Mazzorin/Ruggini 2009. imply that a percentage of the domestic caprines was not kept for wool.

Thiessen polygons and archaeobotanical records from Montale's territory
A Thiessen polygon model associates settlements with their surrounding territories, and can be applied to the Modena province during the Bronze Age to associate settlements with resource areas. Despite its relatively small size (1 ha), Montale appear to have dominated, during its initial phases I-II (Middle Bronze Age 2), an unusually large territory of c. 2200 ha46. According to the Thiessen polygons' model, the average size of the territories associated with the other Terramare sites of Modena's alta pianura (= part of the Po plain south of Modena, cf. Fig. 1) is a much more modest 260 to 890 ha. Of course, Thiessen polygons represent an abstract model, and other factors invisible to the archaeological record might affect regional landscaped management47; nonetheless, access to an extensive surrounding landscape would doubtlessly represent a crucial factor to the development of intense sheepherding targeted at wool production. Political control over a relatively wide territory would allow the Bronze Age Montale's community to carry out at the same time large-scale animal husbandry integrated with prosperous agriculture.
A substantial confirmation for the interconnected prosperity of Montale seems to come from the pollen diagrams. Pollen analyses from the Terramare area have shown that emphasis on pastoral economy is differen- tial on the plain, and that not all the sites would pursue the same economic strategies. An important indicator for pasture and grazing land has been identified by the presence of cichorieae (one of the six tribes of the sub-family Cichorioidea) in pollen diagrams48. Pollen studies from Montale, as from the neighbouring site of Baggiovara, show a significant presence of chicorieae49, strongly suggesting that they were surrounded by a landscape where animal grazing had a relevant role. Pollen diagrams from other Terramare villages, such as the well-studied site of Poviglio, provide a different picture revealing a surrounding territory, with a prevalent agriculture and just few limited zones probably devoted to pastoral activities50. As mentioned earlier, Montale is situated in the alta pianura (the southern margins of the Po plain) near the potential summer pastures offered by the Appenine Mountains, which were exploited for sheep raising in Roman times51. Earlier studies, based on the archaeological evidence, have suggested interaction between the plain and those mountainous areas throughout the Middle and 48 Florenzano et al. 2015. 49 Accorsi et al. 2009Mercuri et al. 2006, 55. 50 Cremaschi et al. 2016, 168. 51 Corti 2012 Recent Bronze Age52. Although they did not provide conclusive evidence as to Bronze Age sheepherding, Montale's herds could potentially have had access to these mountainous pastures (or perhaps to forage harvested there). Although not proved, the possibility that Montale's community could exploit the mountainous pastures offered by nearby Appenine uplands is not contradicted by the geochemical analyses carried out at the site53.

Strontium isotopes and Bronze Age husbandry practices at Montale
Strontium isotopes have proven to be an interesting means to measure relationships between people, animals and organic material from the past and the landscape in which they were found54. Within the THESP project supporting the present work55, a twofold series of strontium isotope  analyses was carried out to explore the characteristics of the pastoral economy at Montale, and in particular, to understand whether animal seasonal mobility was used to optimise landscape management. A primary goal of the sampling strategy has been to map the biologically available regional strontium baselines around Montale, including both the plain surrounding the site, and the neighbouring northernmost slopes of the Appennine Mountains56. The two largest rivers of the area, with a relatively consistent water flow all year around (Secchia to the west, and Panaro, to the east of Montale) were chosen as eastern and western borders of the studied area. They likely represented considerable landmarks57. The local baseline values have been measured in water and vegetal samples and in soil from different excavation layers at both the site of Montale and at the neighbouring and roughly contemporary necropolis of Casinalbo58.
The second aim of the sampling strategy was to collect sheep teeth from the stratigraphic excavations at Montale to provide evidence to test pastoral mobility. Thirty-five teeth (between two and five per each 11 archaeological phases investigated at the site, Tab, 1) have been sampled for strontium isotope analyses59. Most teeth have been subjected to multiple sampling to investigate mobility through time in the same animal individuals60. The results of the analyses are in course of publication and cannot therefore be provided here. For the scope of this paper, one can say that they are not in contradiction with the hypothesis put forward in this paper that domestic caprines primarily grazed in the territory around Montale, and possibly, if convenient, used Appennine mountain pastures61.

Animal demography and wool economy: attempting quantification Sheep grazing and stocking rates
In terms of pastoral economies, various attempts have been done to calculate the amount of land necessary for successful prehistoric sheepherding62. The extent of grazing land varies greatly depending on climate, environmental factors (i. e. characteristics of the vegetation, access to fresh water, weight and size of the grazing animals) and the degree of human intervention and regulation63. Desert pastures similar to those available for southern Mesopotamian Bronze Age wool economies would need between 6-8 ha to support one sheep in a year, while in the steppe environment of northern Mesopotamia this figure would decrease to 4.5 ha. Probably, about 24.5 ha would be necessary to produce around 3 kg of raw wool in a year or 2 kg of clean/ready to use wool in the south, and 15.75 ha, in the north of Mesopotamia64. On the northern edge of western Eurasia, 1 ha of actively managed heathland could sustain one adult double-coated old Norse sheep, producing 1-2 kg of raw wool per year65. Long-term investigations on the feral Soay population living in the Village Bay sector (c. 175 ha) of the Hirta Island, Scotland, UK, have shown that even when the human intervention is practically non-existent, a harsh environment could sustain between 313 and 423 sheep between 1961 and 1967 and from 211 to 513 animals during the period 1985-1990; thus, between 1.2 and 2.9 animals per hectare66.
Controlled grazing may increase the number of animals in a given landscape and guarantee -i. e. by storing surplus vegetation from the spring/summer seasons -adequate food intake during the winter months67. Controlled grazing and rotational practices, including multispecies grazing, seem to provide ecological and economic advantages and to a certain extent higher resistance to parasites, toxic plants, and predators68. Ethnographical studies show that traditional pastoral societies actively practice forms of rotation and controlled grazing69. Archaeobotan-ical studies from Montale seem to confirm that the immediate benefits of such practices were exploited by local community (Mercuri et al. 2006).
A study from Virginia, USA, shows that a modern supposedly 'safe' rate of 2.5 sheep/acre (c. 5 animals per ha) in favourable environmental conditions can be doubled with controlled grazing70. A survey of modern controlled farming from New Zealand for the period 1987-1990 reports an average eight stock units (each unit consisting of one ewe and one lamb) for summer dry properties and 10.8 stock units for summer moist pastures71. Similar rates are obtained in Sardinia, where irrigated lowlands can apparently be grazed by 10-20 ewe/ha, while the dry hills by 1-3 animal/ha72. Studies on the Appennine mountains of the Parma province, in northern Italy, show a subsistence capacity similar to that estimated for the Sardinian hills, with rates between 0.97 and 3.93 sheep/ha73. As to the scope of this work, one should consider that Modena province's alta pianura during the Bronze Age probably offered relatively good quality pastures all year around, thus high stocking rates (or numbers of animal per units of land). Today, local farmers report that seven to nine sheep per hectare is a convenient stocking rate74. In Roman times, a few kilometres north-east of Montale at the site of Baggiovara, 2 sheep/ha was average for small-scale, single-family farms following a mixed subsistence economic model75. Modern organic farming standards recommend for average quality pastures a sustainable stocking rate at 1.2 livestock units (1 livestock unit = 5 domestic caprines) per ha76. In Italy, current standards for organic farms have set a maximum number of sheep per hectare at 13.3 animals77. According to a study of grazing methods in modern semi-arid Mediterranean conditions, natural pastures appear suitable for a low-medium rate of 1-5 heads/ha; forage crops of grass and legumes can sustain medium-high rates of 5-10 heads/ha, while high-quality forage crops and legumes support generally high rates of 10-15 heads/ha78.
To calculate carrying-capacity rates for the territory of Montale, one should consider that modern rates target bigger and heavier sheep (with adults ewes from 60 up to 130 kg as in the case for instance of the Suffolk breed)  Molle et al. 2004, 197. than more 'primitive' sheep such as the Soays (adult ewes generally between 25-30 kg). The latter are, as mentioned above, able to survive at low-medium stocking rates even in the unmanaged harsh north-Atlantic landscape of Hirta Island, Scotland. Thus, Bronze Age sheep farming could likely use a more convenient coefficient of animal/ha than those applied for modern herding. Additionally, light primitive sheep cause a generally limited degree of damagewhen it comes to trampling of the grazed land -supposedly effecting a faster recovery of the vegetation throughout the year79. A confirmation that Bronze Age sheep from the Po valley would resemble Soay sheep in size comes from archaeozoological studies revealing that the average withers' height is c. 60 cm or below80. Soay sheep's wither height is around 52 cm for ewes and 56 cm for rams81.
The information just summarized show that it is possible, although hypothetical, to attempt to reconstruct stocking rates for the sheep population in given landscapes. In the case of Montale, considering: 1) its vicinity to potential mountainous pastures on the neighbouring Appenine, 2) the small size of the Bronze Age sheep, 3) the likely good quality of the pastures, and 4) the abundance of water, the local territory was probably able to support significant stock rates, between at least 5 to 10 domestic caprines per hectare.

Sheep stocking rates and other domestic fauna
To calculate the stocking rates proposed in the following models (Tab. 3-5), it has been considered that other domestic fauna were also present in the area.
Around 10 % of the local animal remains throughout the site sequence (Tab. 2) belong to cattle, this number paired with the advanced age of the analysed animals suggest that cattle were used primarily as aid in the agricultural work82. The absolute number of cattle at the site was probably limited to a few per household and thus likely well below a stocking rate of one animal/ha. Multispecies grazing experience suggests that one cattle per more than one hectare otherwise used for sheep should not alter the food intake of the prevailing species since they eat different forage. Additionally, the Bronze Age cattle from the plain appear to have been relatively small in size with a withers' height around 110 cm83, thus probably needed a lower food intake than modern cattle.
Pigs represent the second most common domestic taxa at Montale (Tab. 2); their diet is often thought to have been largely human leftovers and temporary browsing in forested areas84, which would be quite consistent (660-436 ha) in the hypothetically calculated Montale's territory. Hence, they too would hardly affect sheep's grazing.
Goats have the same stocking rates as sheep85, and are obviously present in a relatively rather stable 4:1 ratio throughout all phases (I-X), but one (phase XI) when the ratio raise to almost 4:2. All the same, goats normally target different plants than sheep, and in the ratios evident at Montale, they could profitably graze on the same landscape without interfering with each other as to food intake86.
The figures proposed, that feature 2-5 sheep/ha or modern low-medium stocking rates87, are probably modest; however, they should allow excluding the other domestic taxa (cattle, goat and pigs) from the modelling without affecting the potential significance of projected results.

Sheep and wool economy at Montale
About 1 hectare of land per person would secure agricultural subsistence88 and with an estimated population of between 120-130 individuals at Montale, perhaps 130 hectares of land immediately around the settlement should have been dedicated to agricultural production. It has been assessed that the landscape around Terramare settlements was generally managed as following: c. 20 % of the available land was devoted to agriculture, between 20 % and 40 % was covered by forest, while the remaining 40 % to 60 % of the land was used for pasture89. Pollen diagrams from Montale suggests that c. 30 % of the territory surrounding the site was forested90, thus the remaining 50 % could have been grazing land. The Thiessen polygon model suggests that, during the Middle Bronze Age 2, Montale could have controlled a ter-83 De Grossi Mazzorin 2013, 261;Riedel 1996, 54. 84 De Grossi Mazzorin 2013Gregg 1988, 118-123;Randall 2010, 49. 85 Cf. Eurostat LSU. 86 Cf. Elias/Tischew 2016. 87 , 197. 88 Cardarelli 2009b2009c, 44. 89 Cardarelli 2009b, 461. 90 Cremaschi et al. 2016Mercuri et al. 2006, 263. ritory of 2200 ha91, 20 % of which is 440 ha or well beyond that required to support the settlement's small population. Results open up for at least two possibilities; in the first case (scenario A), Montale's inhabitants could have been practising agriculture that produced surplus for export. In the second case (scenario B), a considerably lower percentage (around 6 %) of the territory around the site could have been used for subsistence. Maintaining that 30 % (or 660 ha) was forested and generally used for other purposes than sheep farming, the remaining 50-64 % of the land (c. 1100-1400 ha) could have been devoted to grazing of the local domestic caprines. Quantifications are therefore proposed in two different scenarios (Tab. 3-5): A) Intense agriculture (both agriculture and sheep-farming are export-oriented) B) Subsistence agriculture (only sheep-farming is export-oriented)

Middle Bronze Age 2
The proposed model (Tab. 3) largely confirms the hypothesis that Montale was an early centre of wool production and export92. Despite using a combination of low coefficients (a modest low-medium stocking rate of 2-5 sheep/ ha, and an annual wool yield of c. 300 g of wool per animal)93, modelling show that the hypothetical territory of Montale could support an animal population between 2640 (at 2 heads/ha) and 6600 animals (at 5 heads/ha).
Using as a frame of reference production figures proposed by Eva Andersson Strand and Marie-Louise Nosch94 and considering that a third of the raw wool would have been discarded during cleaning, Montale's territory could easily deliver slightly over 1.3 ton of clean ready to use wool per year.
Since the archaeological evidence suggest yarn production at Montale close to an industrial scale95, it is likely that locally produced wool was cleaned and spun at the site for export. At an average speed of 40 m per hour to obtain a middle range thread96, we get a workload for the hypothetical maximum amount of wool per year (Tab. 3 scenario B), of about 380.000 hours. The latter would make c. 47.000 working days (at 8 hours per day)97, which is a 130 years long occupation for one person or a yeararound labour for 130 craft-workers. 130 people would be the entire settlement population and thus entailing a critical workload for the community, which could be compensated for instance with the help of workers from outside the settlement, or with intense spinning carried out by the whole community during evening hours, or by trading part of the wool as a raw material.
In population, the figures clearly show the necessary hard work, but also the potential economic profits from the local wool economy.

Middle Bronze Age 3
During the Middle Bronze Age 3, the Thiessen polygon model suggests that the Montale's territory shrunk to around 1530 ha99. The archaeological evidence from the stratigraphic excavations show a more limited number of textile tools during phases III-VIII than in the previous Middle Bronze Age 2 phases, however, the limited size of the excavated area might easily affect the reliability of such data100. The relative presence of sheep/goats among the archaeozoological material did not change. One could, therefore, imagine a slightly more intense, in 99 Cardarelli 2009c, Fig. 38. 100 It is relevant to consider that during phase V the excavated area features metallurgical activities and that during phase VII it hosts a granary (Cardarelli/Labate 2009b), thus no textile work could have been profitably going on there during at least these two phases. mations of the landscape around Montale during this period. In the first place, the forested area decreased from c. 30 % to 20 % of the total land surface103. Cereals also decrease in the pollen diagrams, data possibly confirmed by a contemporary sharp diminution of pigs that consumed agriculture leftovers (Tab. 2). On the other hand, already at the end of Phase VIII and more evidently in Phase IX, the cultivation of grape -previously non-existent in the local pollen diagrams -is introduced104. Perhaps grape and its by-products (i. e. wine), might have replaced other activities such as yarn and/or textile production to obtain exchangeable export products. Thanks to an increased diversification of productive activities, raw wool might for instance have become a more profitable export than yarn or clothing105. According to the Thiessen polygons model Montale's landscape during this period would be of c. 2180 ha106. Given the transformations in agricultural practices and land-management, it is difficult to estimate the area of pasturage; however, if one considers that the forested areas is now much smaller than during the Middle Bronze Age, an increased 10 % of the available land could have been used for increased sheep and especially goat raising. Applying the same coefficients of 2-7 sheep/ha proposed for the Middle Bronze Age 3 (Tab. 4), Montale's territory could have produced an expanded annual wool yield (Tab. 5). The hypothetical labour required to clean and spin such amount of wool would -in particular in the case of seven sheep/ha -exceed the workload that the community could reasonably meet, thus theoretically supporting the idea of an economic transformation entailing production and trade in raw wool.

Montale's pastoral economy: concluding remarks
The archaeological evidence for textile tools combined with the archaeozoological indications of a significant sheep/goat population have led to the proposal that the Terramare site of Montale was an early centre of largescale wool production for export107.
Wool economy is a complex endeavour entailing a long chain of working processes and the interplay of several actors108. In the Aegean and the Near East, early wool economies seem to have been the outcome of precise choices managed, and strictly controlled by the central political authorities109. European Bronze Age communities, however, had fundamentally different social and political organizations than those of contemporary Mediterranean and Near Eastern societies110. The European societies were unlikely able to control at once large workforces, wide territories, and extensive trade routes. Therefore, engagement in a profitable surplus-oriented wool economy must have been possible only in the presence of favourable environmental, social, and/or economic preconditions. The small size of Bronze Age sheep and the limited amount of wool yield per head/year must have represented a major problem. Access to wide pastoral landscape and the possibility to organize intense sheepherding near a settlement would have created primary economic factors enabling what would have supported the rise of local wool economies. The aim of this paper has been to assess whether such a wool economy was possible at Montale, and at the same time, to provide a model for comparable studies in other settlement and/or environmental contexts. Different data sets have been combined and discussed including archaeozoological and archaeobotanical studies, Thiessen polygon models, and to a minimum extent strontium isotope tracing analyses. The results show that Montale likely controlled an unusually extensive landscape, a considerable portion of which served for animal grazing. The site is also characterised by a strategic geographical position in close vicinity to possible summer pastures in the Appennine Mountain (cf. Fig. 1). Considering in particular, the economic value of their wool, herds likely grazed close to the settlement, but could potentially have used nearby mountain pastures. Considering the difficulties represented by the relatively homogenous bioavailable strontium isotope of the entire Emilia-Romagna region111, the strontium measured in local sheep teeth112, cannot probably confirm, but certainly makes a case for, the idea that what could be the local environmental comparative advantage was potentially fully exploited by Montale's community.
In order to provide a frame of reference to discuss the nature and size of the local wool economy, it has been attempted to quantify the potential number of sheep managed by the site, and the wool yield that could be obtained each year. Data from the Bronze Age wool economies of the eastern Mediterranean as much as from modern guidelines for organic sheep farming show that husbandry practices are variable and dependent on both local regional environments and socio-political organizations of producing communities. Thus, the proposed quantifications have a strong hypothetical value.
The size of the local herds has been calculated using rather low stocking rates; however, even in the most unfavourable scenarios (see Tab. 3-5), Montale's surroundings could support a considerable number of sheep. In the same conservative scenarios, when applying the minimum wool yield possible (according to relevant Aegean Bronze Age texts), the total wool yield per year would nonetheless reach beyond any reasonable local consumption. The more favourable scenarios, probably still below the local maximum livestock farming capacity, would suggest that most of the community could be fully employed crafting between 0.5 and over 2 tons of wool per year.
The exceptional abundance of spindle whorls recovered at Montale has led to propose a community of practice model to explain the characteristics of the local engagement in wool economy113. Given the lack of clear archaeological indicators for economic inequality, wool economy at the site is envisioned as a corporate investment where the whole community would participate in different ways. This model contrasts sharply with the state regulation known for the Aegean and Mesopotamian Bronze Age, suggesting a group-up line of economic development involving community specialization. Different from other Terramare villages of the Po plain, Montale and the settlements of Modena's alta pianura do not seem to radically change in size and structure with time114. Maybe organized corporate investments and successful economic management could be at the core of such apparent political stability? The question is for future studies.