Iron production, settlement and environment:
a regional approach

Iron smelting belongs to a relatively widely studied field within Iron Age production activities in La Tène Europe, even if the attention it has received has varied across time and space. There have been recent general surveys by R. Pleiner¹, for France by M. Leroy, and, M. Mangin et al.² and by C. Domergue et al.³, for northern Europe by H. Jöns⁴, for Oberlausitz by V. Hirsekorn⁵ and for Poland by K. Bielenin, M. Mangin and S. Orzechowski⁶. Whilst until recently there has not been reliable archaeological evidence for iron production of the 6th-5th centuries BC in the Late Hallstatt to Early La Tène periods⁷, the record relating to the 4th-3rd and especially 2nd-1st centuries BC (or later La Tène period) is richer. For the lastmentioned phase of the Iron Age, the existence of a rather limited number of specialised bloomery sites seems to be typical. This pattern changes in the succeeding Roman Period (1st-4th centuries AD) with its large number of bloomeries, recovered either in groups or dispersed.

Iron smelting was carried out throughout La Tène Europe, although the dating of some of the assumed bloomery locations to this period has been doubted⁸. In Bohemia, only a few sites are considered to be of La Tène date, and a much higher number of bloomeries has been dated to the Roman period. Since it is assumed that some types of installations such as furnaces continued to be used, and local iron ore sources continued to be exploited, it can be difficult or even impossible to distinguish the evidence for La Tène iron smelting from that of the Roman period. It is true that bloomery sites needed an ecological setting which offered the key resources: iron ore, wood and water. Where appropriate locations were resettled and reused at different periods within the Iron Age, it may be assumed that the main reason was the local occurrence of iron ore. Prospectors had their task made much easier when there were visible signs of bloomery activity – such as slag – from earlier periods of iron production on site, and it was sufficient to ascertain whether or not local iron ore sources had already been exhausted. When a La Tène component (such as settlement evidence) is recorded on iron production sites attributed to the Roman period, it is reasonable to ask whether iron smelting was already carried out there in the La Tène period. Traces of La Tène iron production may be less pronounced, which is particularly true in the case of shallow bowl furnaces without slag pits – supposed to be the prevailing La Tène type; these could have been covered, disturbed or destroyed by later iron smelting activities. The difficulty of identifying Iron Age iron smelting in areas of later iron production has been observed for example in the course of the Dietz-Hölzetal Project in Hessen⁹.

Field survey data indicates that iron was smelted in the Iron Age on a much larger scale in Bohemia than can be demonstrated from the results of destructive archaeological investigations (excavations); this seems to be valid for Iron Age (and also medieval) Europe in general. The identification of surface or only slightly sunken smelting installations (furnaces) is difficult and their existence is often reflected only by production waste, such as bloomery slag and pieces of furnace lining which were removed from furnaces and dumped on the surface nearby. Today, this waste can mostly be found deposited in the ploughsoil or in cultural layers close to the modern surface. This is also true of slag blocks from slag pit furnaces which were often removed to allow further use of the furnaces¹⁰. This is the reason why surface survey is essential in research into Iron Age as well as medieval iron production. The ready recognition of bloomery slag and the fact that it belongs to the relatively indestructable category of archaeological finds facilitate the identification of iron smelting sites.

Nevertheless, large-scale iron production that would have created a product of supra-regional significance cannot be assumed in Iron Age Bohemia. Substantial areas with evidence of bloomeries, indicating intensive activity and a large volume of iron production – and therefore with a large volume of production waste – would certainly not have escaped the attention of archaeologists.

Research into iron smelting has progressed in recent years from the excavation of individual workshops and sites to the prospection and investigation of whole regions. Projects focused on iron production evidence, some accompanied by environmental studies, have been carried out¹¹, but also projects – less frequent to date – investigating the mutual relationships between settlement, production and environment. Projects conducted recently in Bohemia belong to the latter type and some of their methodological approaches and results concerning the La Tène and Roman periods will be considered in the following pages. They are the Loděnice and Říčansko Projects in Central Bohemia (fig. 1).

The Loděnice Project

Since the earlier twentieth century, finds from the Loděnice region in the north-western part of Central Bohemia have reflected the industrial character of the territory where at least two crafts, sapropelite working and iron smelting, were carried out during the La Tène period (LT B2-C1, 3rd-2nd centuries BC). The raw materials required were sourced from outcrops of the Kounov coal seam found within the study area. Iron ore sources, previously unknown, have been identified in the region as the result of both archaeological and geological surveys. The area is located in undulating terrain covered by kambisoils at the watershed of two streams; ecologically, prevailing conditions may be described as below average for Bohemia. During this project, which took place in 1991-1995, an area of 253 ha was investigated by a combination of excavation and fieldwalking¹². A high density of production waste was found and dated to the La Tène period by diagnostic artefacts and the spatial relationship to nearby settlement. The distribution and quantity of bloomery slag, as well as comparison with sapropelite production waste and settlement finds such as pottery permitted the recognition of over thirty bloomery sites and their correlation with settlement areas¹³. Production features such as iron smelting furnaces were only recorded in a few cases. The probable reason is that for the most part shallow furnaces were used in the La Tène period, so that their remains can only be anticipated in the ploughsoil or on the present ground surface in this agricultural landscape. Later iron production is attested in the study area by one workshop dated to the late Roman period (4th century AD).

The study of the archaeological finds using GIS and vector synthesis¹⁴ has enabled the modelling of the shifts in the settlement areas as well as iron smelting areas in the La Tène period¹⁵. If the assumption is correct that the need for wood supplies was the reason for the displacement of the iron smelting areas outside the settlement zones, it could then be postulated that the deforestation of large areas may have taken place¹⁶. Theoretically, the Loděnice region could have formed the industrial (iron-producing) sector from which the Stradonice oppidum, located nearby, was provisioned (fig. 1).

Archaeological analysis of La Tène settlement evidence in the study area resulted in the development of a model of an industrial zone: this can be understood as a group of communities linked by production activities requiring the exploitation of the same raw material source or using the same materials, and possibly enjoying a monopoly of access to them. The zones of Kounov-type sapropelite working, and that of local iron smelting, are typical examples of industrial zones¹⁷. The Loděnice project has clearly demonstrated that surface survey has a key role to play in the study of prehistoric iron production as well as in the modelling of production distribution and of course also related residential areas.

The Říčansko project

The region of Říčansko, located in the eastern neighbourhood of Prague offers, in contrast with Loděnice, finds of a larger number of iron smelting sites including slag-pit furnaces, usually dated archaeologically to the Roman period where dating is possible. The more thoroughly studied parts of the study area have provided evidence of relatively dense settlement, not only in the Roman period, but also in the La Tène period. This region is crossed by several small streams and covered by orthic luvisoils and kambisoils, with climatic conditions that are not very favourable; the south-east part exhibits considerable slopes and a granitic bedrock indicative of less favourable conditions for farming.

An area extending to some 230 hectares¹⁸ was investigated by archaeological methods including fieldwalking, and by geology and geochemistry. Poor conditions for fieldwalking, combined perhaps with the locally prevalent type of slag-pit iron smelting furnace, from which relatively less slag than in the case of shallow furnaces reaches the present-day surface, resulted in fewer surface finds of bloomery slag than in the other study area. Nevertheless, the results obtained enabled the identification of at least twenty-five iron smelting areas/workshops here; these were modelled as circles, each with a radius of 500 m. Multiple sources of iron ore were recognised, possibly including a local type of bog iron ore¹⁹; finds of bloomery slag have been ascribed to the different ores²⁰. It is remarkable that the La Tène iron smelting component here shows a closer spatial relationship to stable iron ore sources, mostly of Ordovician origin, in the northern part of the region, than to the presumed locations where bog iron ore (or ochre) could have been formed²¹, although this observation is rather tentative as yet (fig. 2). The chronological fixing of surface finds of bloomery slag by reference to accompanying finds is a problem, as the La Tène and Roman activity areas often overlapped. When the spatial distributions iron smelting and settlement here were compared, the slag finds are correlated more closely with the La Tène period settlement sites (fig. 3). Study of the spatial distribution of La Tène and Roman finds in the landscape, and their relationships with natural conditions and mineral resources did not reveal any specific characteristics for the siting of iron smelting areas compared with those of farming settlements. It seems that iron production was carried out within the areas used by the farming communities, or, more generally, within the farming zones.

Ecological demands of iron smelting

Accessible iron ore and wood are essential for iron production. Whilst it is usually assumed that suitable iron ores were widely available –although this cannot be stated categorically as not all iron ores are of sufficient quality– this is not wholly true for wood, whether used as a fuel or transformed into the charcoal necessary for reduction in the furnace. It has been suggested that a shortage of wood could have represented a limiting factor in Iron Age iron production. As has already been stated, Neustupný²², arguing from the high numbers of bloomery areas at Loděnice, postulated substantial deforestation here. Investigations carried out by one of the present writers at Říčansko however suggest²³ that the consumption of wood for local iron production need not have caused significant deforestation provided that some form of forest management was carried out –a topic discussed further below.

Estimations of the ecological demands of iron production require: 1. the creation of a model of iron consumption by a single rural community and the estimation of the number of smelts undertaken and the volume of wood necessary to produce a given quantity of iron; 2. the reconstruction of the number and distribution of communities and their settlement areas within the region, and calculation of the volume of iron and wood needed for all these communities; and 3. comparison of the quantity of wood considered necessary with the possible wood production of the region, allowing for the sustainable development of the forest. From this, the possible level of iron production in the region, allowing for ecological sustainability, can be deduced and the degree of self-sufficiency of individual regions in terms of iron production considered.

Iron consumption of a rural community

The inventory of iron equipment in a typical rural site was hypothesised from estimates made of the basis of the property of the inhabitants of an average non-enclosed rural hamlet or small village of the LT B-D period. It is assumed that such a community included four families of two adults and two children, or some sixteen to twenty persons²⁴. As is indicated by the common occurrence of smithing slag in La Tène settlements, the existence of one smith in the model village was assumed; the smith may have been a part-time specialist.

The equipment inventory was derived from the assemblages of iron artefacts known from La Tène settlement and production areas including burials and hoards. Such estimates tend to be rather low, corresponding to minimal needs. The basic equipment can be reconstructed based on categories which occur in the Iron Age²⁵, but also in later contexts, for instance in medieval times²⁶. These categories are: agricultural tools, craft tools (in the La Tène period mostly smithing equipment), domestic and construction components, wagon fittings and horse harness, and warrior equipment. Personal adornment and exceptional tools were not considered in the model developed here.

The minimum estimate²⁷ comprises two groups of iron equipment, respectively communal and household. The communal group consists of artefacts which are present as a single set in the community (and ignores the issue of their ownership). These artefacts could have been shared by the community, or used during ceremonies, or perhaps they were components of the restricted equipment of the local elite (e.g. chariot and horse harness, cauldrons, and perhaps heavy agricultural tools). Also, to be considered within this group are the specialist tools of individual craft workers in the village – for example the equipment of a smithy. The equipment of two warriors who could have been provided by the village is also included here²⁸. The other group is represented by artefacts used in a household. This equipment has to be multiplied by the presumed number of households in the village.

The weight estimate of individual iron artefacts is based on recorded data²⁹. Summing the calculations amounts to 9 kg of communal iron and 8 kg of household iron (represented by 4 households with 2 kg each), in total about 20 kg of iron for one rural community³⁰.

The pattern of use of iron artefacts was reconstructed as follows: 70 % can be recycled or repaired = 14 kg, 25 % was lost to the community through transfer elsewhere (due to the actions of war, as dowries or gravegoods) = 5 kg, and 5 % was lost = 1 kg. Therefore, 30 % by weight of the iron inventory, representing 6 kg, must have been reproduced, possibly once every five years, or five times per generation of twenty-five years; during this span, the requirement for new iron would have totalled c. 30 kg. Each generation, each community would thus have needed 20 kg (recyclable iron) + 30 kg (loss) = 50 kg of iron in total.

La Tène iron production has been calculated based on the yield of iron from a slag-pit furnace³¹. This amounts to c. 2 kg on average, but further losses during the post-reduction processing would have reduced this figure to one kilogramme of iron incorporated in artefacts. 50 kg of iron would therefore correspond to 50 smelts, and the total iron consumption of a village over twenty-five years could have been furnished by two smelts a year.

The volume of the iron inventory of a single rural village could have been, of course, larger. A more diversified assortment (of agricultural tools, etc.) may be postulated as well as some artefacts being of heavier weights or present in larger numbers than suggested above (for example weaponry for more than two men). Therefore multiples of the minimum estimates laid out above are employed in the following model of the impact of iron smelting on the landscape.

It can be assumed that the basic iron inventory of the Roman period communities living in barbaricum did not differ much from that of the La Tène period. This can be deduced from calculation already made by others. According to Lund³² one farmstead in North Germany in the first centuries AD could have had iron equipment weighing some 5-10 kg and suffered a yearly loss of 1-2 kg. These numbers are of the same order of magnitude as our estimates.

Modelling hypothetical settlement areas for the La Tène period

The model of La Tène settlement areas (fig. 4) was derived from the economic model suggested for the early Iron Age or Hallstatt period³³. It assumed that the agricultural system practiced in the Iron Age was at subsistence level³⁴ and founded on sustainable development, because rural settlements seem to survive at the same locations for centuries. Current opinion maintains that rural settlements were small, consisting of several, or perhaps only a single, farmstead. As stated above, the average community size was estimated at four families with four to five members each, or some sixteen to twenty people in total. It is assumed that the subsistence strategy of such a community was an optimally balanced mixture of arable and pastoral farming, where cattle were fed on leaf fodder – and during the La Tène period possibly by hay – in winter; and forest grazing was practiced only exceptionally, for instance in extremely dry years. According to the above model supported by ethnographic research, it was concluded that the minimum quantity of land necessary for cultivation, by ploughing, grazing and winter fodder production is one hectare per person³⁵. If there are twenty people in the village, then the minimal land required for these purposes should amount to twenty hectares. Using the GIS (IDRISI programme), these areas were modelled around the settlement areas, again according to ethnographic observations, based on the availability of good quality soils with slopes of less than 5^o.

Calculation of the area of forest that would have been exploited has been derived from the estimated amount of timber required for fuel and construction, from the amount of leaf fodder used in the farming system, and according to the demands that permanent recovery of the forest would have made, in other words what was taken out of the system should not have impeded regeneration. The area of forest necessary for the acquisition of leaf fodder by one community is estimated as the product of 4.7 hectares of managed open forest of almost park-like character. The estimated forest area required for fuel and construction wood (calculated again with regard to sustainable development, that is the yearly natural accretion of wood will replace what has been taken³⁶) will have increased the overall forest area required by one community to c. 10 hectares, but this estimate –as will be discussed below– is probably too high.

The alternative method of feeding cattle is forest grazing. The area required for the pasture of domestic animals in one community, without depleting the forest, is about 450 hectares which exceeds the potential of our presumed settlement areas. Even in the Říčansko region, a demand of this magnitude could not have been met without certain destruction of the forest, or without regulating access to pasture as is also known from ethnographic sources. These figures indicate why forest grazing is considered only to have been an occasional source of fodder in times of shortage of hay or leaf fodder.

The modelled Iron Age village populated by twenty persons would thus need about 20 ha of cultivated land and 40 ha of forest, probably located in the immediate vicinity of the settlement. These calculations are fundamental to the model of hypothetical settlement catchment and dwelling areas, as shown in fig. 4. There are apparently on this basis many more places suitable for settling than are recorded as containing settlement sites, even restricting potential sites to those located at a distance of up to 300 m from a water source, with acceptable soils and with less than 5^o slopes. Theoretically, there could have been many more communities (up to a hundred times more) living in the region then are known today to archaeologists³⁷.

Forest as a source of wood for charcoal production

The character and economic yield of prehistoric forests are still a subject for debate. Since the first pollen analyses³⁸ which are the most important source of data for past vegetation reconstruction, most researchers have considered the idea that, at least from the Boreal, temperate Europe was covered by closed forest. This idea was attacked by Vera³⁹, and the ongoing discussion following this has given rise to the hypothesis of rather more open forest cover produced either by large herbivores or, more likely, as the result of anthropogenic impacts⁴⁰. Since the Boreal/Atlantic transition (c. 8000 BP), a complex deciduous forest characterised mostly by oak prevailed in Bohemian lowlands. A mixed vegetation cover containing varied proportions of trees, at different growth stages from young individuals to the final stages of decay, interspersed with open glades, may be presumed⁴¹.

The Říčansko region is nowadays covered by mixed forests in which deciduous trees dominate. To calculate wood production, it is necessary to estimate the density of trees within the forest. In the climax phase in the temperate zone, a maximum of 100 large trees per hectare of deciduous forest is suggested, complemented by hundreds or even thousands of trees of different ages and trunk sizes⁴².

The estimate of wood production in the Iron Age forest of the Říčansko area requires data (a) on the total currrent annual increment and (b) on the average wood production of “natural forest” (e.g. not the plantation). This depends on wood species, age, quality, stand density. The estimation of the approximate representation of wood species and stand density in such type of mixed forest (based on the data from existing European and North American “primeval” forests) was attempted by a forestry specialist, V. Zatloukal (pers. comm.) as follows: oak 60 %, pine 20 %, beech and hornbeam together 20 %. Average stand density was estimated at 70 % of the optimal state (index 0.7). The average age of oak in forest of this type was calculated to be 100 years, with corresponding figures for pine of 80 years and for beech of 90 years, assuming the quality of the land concerned was middle-ranking for Bohemia. The estimations of increments are according to Černý, Pařez and Malík⁴³. The year wood production (= year increment) of one hectare of forest is represented by the volume in cubic metres of the tree including branches and bark. The year average increment is 4,6 m³ (tab. 1).

The average wood production of temperate European mixed forest is 605 kg/m³ (e.g. pine 520, beech 720, oak 760, spruce 470). The average wood production of lower quality wood, e.g. from thin trees, branches or by coppicing is 330 kg m^{3 44}. Similar numbers can be estimated also or for the “natural” forest in the model.

The choice of wood for charcoal production need not have followed strict rules and it woud have depended on the prevailing vegetation in any given area. Pleiner⁴⁵ has stated that almost all wood species and even bushes such as box and juniper were used for charcoal production in European La Tène and Roman bloomeries. Pott⁴⁶, drawing on palynological and geobotanical data, described the controlled exploitation of forest for charcoal production as required in the Iron Age for iron ore smelting in the Hauberg area of Siegerland in northern Rheinland-Westfalen; similar quantities would have been required in the medieval and postmedieval periods. The system was based on a rotation of fields, pastures and coppiced forest composed of quickly regenerating wood species offering a permanent yield of wood as poles.    

As a result of an experiment on the production of charcoal, Pleiner⁴⁷ observed that the ratio of charcoal produced to fresh wood employed was 1:5.7. The early medieval (8th-11th century AD) ironworks of Sauerland in West Germany produced charcoal from oak at a ratio of 1:5. The iron-charcoal ratio by weight in Iron Age bloomeries would have been approximately 1:10-15⁴⁸. Assuming that the production of 1 kg of iron indeed required 10-15 kg of charcoal, then the quantity of wood required for charcoal production would amount to 57-85 kg. Using the higher figure and doubling the required amount of charcoal to cover the demands of ore roasting, post-reduction processing as well as wood utilised in the construction of the bloomery workshop, increases the requirement to c. 170 kg of wood. Total amount of wood needed for initial 20 kg of iron (e.g. iron for one rural community) would be 3400 kg. The year consumption of one La Tène community (assuming an annual output of 6 kg of iron compensating loss) would be 1020 kg of wood. One tonne (1000 kg) of wood represents c. 1.6 m³ of pine, 1.1 m³ of beech or oak or 1.7 m³ of spruce. The required 1000 kg of wood corresponds to the average volume of c. 1.6 m³. Approximately 16 m³ of wood would be needed to get iron which each community had to produce per generation (e.g. within 25 years) on condition that 6 kg loss would be reproduced every five years. If the reproduction of iron occured every year than total amount of 48 m³ would be needed.

Taking into consideration the yearly average increment in such woodland of 4.6 m³/ha, the total sum of wood got through natural growth within 25 years would be 115 m³/ha. The volume of timber necessary for the production of the required amount of charcoal would be many times less than this. If iron smelting during the La Tène and Roman periods did not exceed the minimal needs of a community, it would thus not have left any trace in the forest cover, other than the changing species composition in the case of the selective extraction of a particular species.

It is hardly possible to envisage under what conditions iron smelting would have caused a wood shortage. This is due to the fact that it is not known whether charcoal was produced in the immediate vicinity of the settlements or the iron smelting areas, although according to ethnographic parallels it might be supposed that the latter was the case. Alternatively it could be brought from more distant places (this is less likely from an economic perspective but more consumer friendly in that the fumes from charcoal piles would not have caused inconvenience to people working in the bloomeries).

Hypothetical production of iron in the study region during the La Tène period

Taking into consideration the known (minimum) number of settlement sites in the 230 km² study region, the settlement and production areas (including fields, pastures and forest exploited for leaf fodder, wood for fuel and construction) would have occupied c. 5300 ha. There would have been some 17 655 ha left for forests⁴⁹. A forest of this size would allow the production of 212 tonnes of iron just from the exploitable yearly increase through natural growth. The yield from one hectare of clear-felled forest would amount to 450 m³ of wood (according to calculations by V. Zatloukal for forests in the Říčansko region), and this would have sufficed for the production of c. 1600 kg of iron, which equates to 260 times the yearly production required for one community, or to the functional iron inventories of forty communities.

About 1000 La Tène rural settlements are presently known in Bohemia (according to the Archaeological Database of Bohemia in the Institute of Archaeology in Prague). According to our model, this would correspond to a population of c. 16 000. If this were increased tenfold (which would correspond to the estimation of the population of both villages and oppida⁵⁰), then c. 500-1700 tonnes of iron would be produced during each generation (25 years), requiring wood from max. 1030 hectares of clear-felled forest. In two hundred years, 8560 ha of forest would have been cut down in the whole of Bohemia (this is an area smaller than that calculated for the unexploited forest in the Říčansko region). At the same time, taking into account the natural growth cycle, forest would have reached its climax, that is the fully productive stage again.      

The possibility that iron was produced in the Říčansko region for the nearby oppidum of Závist should be briefly considered. According to Drda and Rybová⁵¹, three specialised smithies could have been working within Závist producing at least 115 different types of formally and functionally differentiated artefacts. It was assumed that iron smelting did not take place within the oppidum, but that iron was brought there from the vicinity. An iron ore lode is located nearby, but no traces of iron production from it have as yet been found. A possible provider of iron for Závist could thus have been the Říčansko region.

Although population estimates for individual oppida in Bohemia are usually lower – reaching hundreds, at most several thousands, of people⁵² – a rather higher assumption of 5000 of people per oppidum was used in our model. If the inhabitants of the oppida produced iron artefacts for their own consumption only, then in year 0, at the beginning of production, they would need to make 6,25 t of iron, requiring some 1062.5 t of wood. This could have been obtained from four hectares of fully exploited, that is, clear-felled forest or from the exploitable increase through natural growth of 500 hectares of forest. Yearly iron production would have required 1,2 hectares of clear-felled forest or the gathering of the increment from 150 hectares of forest. Each generation living at the Závist oppidum would have produced 54 tonnes of iron requiring 34 hectares of clear-felled forest or the annual incremental growth from 4250 hectares of forest.    

It follows that, even if the Říčansko region did represent the iron producing sector for Závist, the necessary iron production would not have caused the forest vegetation of the region to diminish, and iron production there would not have been limited by the over-consumption of woodland (fig. 4). Maximum iron production is more likely to have been limited by the exhaustion of local iron ore sources.

Conclusions

The complex analysis of settlement activities and settlement areas, production components and environmental parameters offers the opportunity to model the density of settlement and volumes of production, and to hypothesise their impact on the environment. A key part of such models should be the consideration of the extent and nature of forest exploitation, taking into account the probability of planned forest management. This can substantially change some currently-held opinions on economic independence or about the mutual dependence of communities and regions. As far as research on iron production is concerned, surface survey should be considered as one of its indispensable tools. Field survey can identify the last remnants of iron smelting, which often cannot be found other than in the ploughsoil or within cultural layers or indeed on the ground surface of today.

Acknowledgements

Our thanks are due to Čeněk Čišecký for GIS mapping and to Eva Čepeláková for graphics. The research concerned was carried out within the project AVOZ80020508 of the Academy of Sciences of the Czech Republic.

References

• Bataille, G. and J.-P. Guillaumet, dir. (2006): Les dépôts métalliques au second âge du Fer en Europe tempérée, Bibracte 11, Glux-en-Glenne.
• Behre, K.-E., dir. (1986): Anthropogenic indicators in pollen diagrams, Rotterdam, Boston.
• Bielenin, K., M. Mangin and S. Orzechowski (1995): “La sidérurgie ancienne et l’exploitation minière dans les Montagnes-Sainte-Croix (Petite Pologne) I. Bilan des recherches 1955-1990“, Dialogues d’Histoire Ancienne, 21/1, 203-224.
• Bintliff, J., M. Kuna and N. Venclová, dir. (2000): Future of surface artefact survey in Europe, Sheffield.
• Birks, H. J. B. (2005): “Mind the gap: how open were European primeval forests?”, Trends in Ecology and Evolution, 20, 4, 154-156.
• Brinkkemper, O. (1991): Wetland farming in the area to the south of the Meuse estuary during the Iron Age and Roman period. An environmental and palaeo-economic reconstruction, Annalecta Praehistorica Leidensia 24.
• Buko, A. and P. Urbanczyk, dir. (2000): Archaeologia w teorii i praktyce, Warszawa.
• Černý, M., J. Pařez and Z. Malík (1996): “Růstové a taxační tabulky hlavních dřevin České republiky”, Příloha č. 3 Vyhlášky MZe č. 84/1996 Sb. o lesním hospodářském plánování.
• Dobiat, C., S. Sievers and T. Stöllner, dir. (2002): Dürrnberg und Manching. Wirtschaftsarchäologie im ostkeltischen Raum, Bonn.
• Domergue, C., V. Serneels, B. Cauuet, J.-M. Paillier and S. Orzechowski (2006): “Mines et métallurgies en Gaule à la fin de l’âge du Fer et à l’époque romaine”, in: Paunier, dir. 2006, 131-162.
• Drda, P. and A. Rybová (1995): Les Celtes de Bohême, Paris.
• Drda, P. and A. Rybová (1997): “Keltská oppida v centru Boiohaema (Die keltischen Oppida im Zentrum Boiohaemums)”, Památky archeologické, 88, 65-123.
• Dreslerová, D. (1995): “A settlement-economic model for a prehistoric microregion: settlement activities in the Vinoř-stream basin during the Hallstatt period”, in: Kuna & Venclová, dir. 1995, 145-160.
• Dreslerová, D. (2001): “Využití GIS při zkoumání struktury mikroregionů (Using GIS for the study of microregional structures)”, in: Kozlowski & Neustupný, dir. 2001, 55-68.
• Dreslerová, D. (2008): “Ekonomický potenciál regionu Říčanska z hlediska pravěkého hutnictví (Economic potential of the region of Říčansko from the point of view of prehistoric iron smelting) “, in: Venclová et al., dir. 2008, 266-279.
• Dreslerová, D. and J. Sádlo (2000): “Les jako součást pravěké kulturní krajiny (The forest as a component of the prehistoric cultural landscape)”, Archeologické rozhledy 52, 330-346.
• Firbas, F. (1949): Spät- und nacheiszeitliche Waldgeschichte Mitteleuropas nördlich der Alpen. Erster Band: Algemeine Waldgeschichte, Jena.
• Firbas, F. (1952): Spät- und nacheiszeitliche Waldgeschichte Mitteleuropas nördlich der Alpen. Zweiter Band: Waldgeschichte der einzelnen Landschaften, Jena.
• Gassmann, G., A. Hauptmann and C. Hübner, dir. (2005): Forschungen zur keltischen Eisenerzverhüttung in Südwestdeutschland, Stuttgart.
• Gassmann, G., M. Rösch and G. Wieland (2006): “Das Neuenbürger Erzrevier im Nordschwarzwald als Wirtschaftsraum während der Späthallstatt- und Frühlatènezeit”, Germania, 84, 273-306.
• Hajnalová, M. and D. Dreslerová (2010): “Ethnobotany of einkorn and emmer in Romania and Slovakia: towards interpretation of archaeological evidence”, Památky archeologické, 101, 169-202.
• Haselgrove, C. and R. Hingley (2006): “Iron deposition and its significance in pre-Roman Britain”, in: Bataille & Guillaumet, dir. 2006, 147-163.
• Hirsekorn, V. (2000): “Zur Archäometallurgie des Eisens in der Oberlausitz von der älteren Eisenzeit (Billendorfer Gruppe) bis zum Hochmittelalter”, Arbeits- und Forschungsberichte der Sächsischen Bodendenkmalpflege, 42, 13-84.
• Iversen, J. (1941): “Landnam i Danmarks Stenalder: En pollenanalytisk Undersøgelse over det første Landbrugs Indvirkning paa Vegetationsudviklingen (Land occupation in Denmark’s stone age)”, Danmarks Geologiske Undersøgelse II. række, 66, 1-68.
• Jacobi, G. (1974): Werkzeug und Gerät aus dem Oppidum von Manching, Die Ausgrabungen in Manching Bd. 5, Wiesbaden.
• Jockenhövel, A. and Ch. Willms (2005): Das Dietzhölzetal-Projekt. Archäometallurgische Untersuchungen zur Geschichte und Struktur der mittelalterlichen Eisengewinnung im Lahn-Dill-Gebiet (Hessen), Rahden, Westf.
• Jöns, H. (1997): Frühe Eisengewinnung in Joldelund, Kr. Nordfriesland. Ein Beitrag zur Siedlungs- und Technikgeschichte Schleswig-Holsteins 1, Bonn.
• Jöns, H. (2007): “Zur ältesten Eisenverhüttung in Norddeutschland und im südlichen Skandinavien”, in: Möllers et al., dir. 2007, 53-71.
• Kozlowski, J. K. and E. Neustupný, dir. (2001): Archeologia przestrzeni, Kraków.
• Kuna, M. and N. Venclová, dir. (1995): Whither Archaeology. Papers in honour of Evžen Neustupný, Praha.
• Leroy, M., M. Mangin, B. Raïssouni, M. Boukezzoula and H. Laurent (2000): “La sidérurgie dans l’est de la France”, Gallia, 57, 11-21.
• Malkovský, M. (2008): “Výskyty železných rud v geologických útvarech jihovýchodní části Prahy a jejího okolí (Occurrence of iron ores in the geological formations of southeastern Prague and its vicinity)”, in: Venclová et al., dir. 2008, 203-228.
• Michálek, J. (1999): Keltský poklad z Bezdědovic na Blatensku (Eisengerätenhortfund der Latènezeit aus Bezdědovice bei Blatná, Südböhmen), Blatná – Strakonice.
• Mitchell, F. J. G. (2005): “How open were European primeval forests? Hypothesis testing using palaeoecological data”, Journal of Ecology, 93, 168-177.
• Möllers, S., W. Schlüter and S. Sievers, dir. (2007): Keltische Einflüsse im nördlichen Mitteleuropa während der mittleren und jüngeren vorrömischen Eisenzeit, Bonn.
• Nekuda, V. (1975): Pfaffenschlag. Zaniklá středověká ves u Slavonic, Brno.
• Neustupný, E. (1983): Demografie pravěkých pohřebišť (The demography of prehistoric cemeteries), Praha.
• Neustupný, E. (2000): “The measure of deforestation in prehistoric Central Europe”, in: Buko & Urbanczyk, dir. 2000, 345-352.
• Neustupný, E., dir. (1998): Space in prehistoric Bohemia, Praha.
• Neustupný, E. and N. Venclová (1998): “The Loděnice region in prehistoric times”, in: Neustupný, dir. 1998, 84-105.
• Neustupný, E. and N. Venclová (2000): “Surveying prehistoric industrial activities: the case of iron production”, in: Bintliff et al., dir. 2000, 93-104.
• Neustupný, E. and N. Venclová (2001): “Posouzení analytických sběrů matematickými metodami a geografickými informačními systémy (Results achieved by means of formal methods)”, in: Venclová, dir. 2001, 187-204.
• Paunier, D., dir. (2006): Celtes et Gaulois. L’Archéologie face à l’Histoire. La romanisation et la question de l’héritage celtique, Bibracte 12/5, Glux-en-Glenne.
• Pleiner, R. (1993): The Celtic sword, Oxford.
• Pleiner, R. (1995): “Výroba železa ve starokeltském světě”, Študijné zvesti Archeologického ústavu SAV, 31, 43-49.
• Pleiner, R. (2000): Iron in Archaeology. The European bloomery smelters, Praha.
• Pleiner, R. and M. Princ (1984): “Die latènezeitliche Eisenverhüttung und die Untersuchung einer Rennschmelze in Mšec, Böhmen”, Památky archeologické, 75, 133-180.
• Post, L. von (1916): “Omskokstradspollen i sydsvenska torfmosselagerfoljder (foredragsreferat)”, Geologisk Forening, 46, Stockholm.
• Pott, R. (1986): “Der pollenanalytische Nachweis extensiver Waldbewirtschaftungen in den Haubergen des Siegerlandes”, in: Behre, dir. 1986, 125-134.
• Rustoiu, A. (2006): “A journey to Mediterranean. Peregrinations of a Celtic warrior from Transylvania”, Focusing on Iron Age élites, Studia Universitatis Babeş-Bolyai, Historia 51, 1 Special Issue, 42-85.
• Schäfer, A. (2002): “Manching –Kelheim– Berching-Pollanten. Eisen als Wirtschaftsfaktor”, in: Dobiat et al., dir. 2002, 219-241.
• Venclová, N. (1995): “Settlement area, production area and settlement zone“, in: Kuna & Venclová, dir. 1995, 161-169.
• Venclová, N. (1999): “Katalog”, in: Michálek 1999, 8-28.
• Venclová, N. (2002): “Wirtschaftsmodelle der Latènezeit”, in: Dobiat et al., dir. 2002, 33-48.
• Venclová, N., dir. (2001): Výroba a sídla v době laténské. Projekt Loděnice (Production and settlement in the La Tène period. The Loděnice Project), Praha.
• Venclová, N. et al., dir. (2008): Hutnický region Říčansko (The iron smelting region of Říčansko), Praha.
• Vera, F. W. M. (2000): Grazing Ecology and Forest History, Wallingford.
• Waldhauser, J. (2001): Encyklopedie Keltů v Čechách, Praha.
• Wischenbarth, P. (2001): “Eisenschlackenplätze im Rothtal (Bayerisch-Schwaben)”, in: Wischenbarth et al., dir. 2001, 9-54.
• Wischenbarth, P., R. Ambs and G. Gassmann, dir. (2001): Keltische Stahl- und Eisenproduktion im Rothtal (Bayerisch-Schwaben), Berichte zur Archäologie im Landkreis Neu-Ulm und in den angrenzenden Gebieten Bd. 2, Neu-Ulm.
• Zavřel, J. (2008): “Geochemický rozbor železářských strusek a železných rud z Říčanska (Geochemical analyses of iron smelting slags and iron ores from the Říčansko region)”, in: Venclová et al., dir. 2008, 242-258.
• Zelená zpráva (2002): Green report 2002: www.uhul.cz/zelenazprava/2002

Footnotes

1. Pleiner 2000.
2. Leroy et al. 2000.
3. Domergue et al. 2006.
4. Jöns 2007.
5. Hirsekorn 2000.
6. Bielenin et al. 1995.
7. Neuenbürg-Waldrennach, Baden-Württemberg: Gassmann et al., dir. 2005, 61-71; Gassmann et al. 2006.
8. e.g. at Kelheim and around Manching: Wischenbarth 2001, 37, 42; Schäfer 2002, 225-227.
9. Jockenhövel & Willms 2005, 377.
10. Neustupný & Venclová 2000, 94-97.
11. Neustupný & Venclová 2000, 94-97.
12. Venclová, dir. 2001.
13. Venclová, dir. 2001, fig. II.
14. Neustupný & Venclová 1998; Neustupný & Venclová, dir. 2001.
15. Venclová dir. 2001, 183, fig. IV.
16. Neustupný & Venclová 2000.
17. Venclová 1995; Venclová, dir. 2001, 8; Venclová 2002.
18. Venclová et al., dir. 2008.
19. Or rather ochre: Malkovský 2008.
20. Zavřel 2008.
21. Venclová et al., dir. 2008, 262.
22. Neustupný 2000.
23. Dreslerová 2008.
24. Neustupný 1983.
25. Jacobi 1974; Haselgrove & Hingley 2006, 156, fig. 7; Michálek 1999, 36-45.
26. Nekuda 1975, 136.
27. Venclová et al., dir. 2008.
28. Cf. Rustoiu 2006, 61-69 for the presumed percentage of warriors in the population.
29. Pleiner 1993; Pleiner 1995; Venclová 1999.
30. Venclová et al., dir. 2008, tab. 48.
31. Cf. Pleiner & Princ 1984, 161-162; Pleiner 2000, 246.
32. Cited by Jöns 1997, 185.
33. Dreslerová 1995; Dreslerová 2001.
34. Brinkkemper 1991.
35. Hajnalová & Dreslerová 2010.
36. Dreslerová 1995.
37. Dreslerová 2008, fig. 170.
38. von Post 1916; Firbas 1949; Firbas 1952; Iversen 1941.
39. Vera 2000.
40. Mitchell 2005; Birks 2005.
41. Dreslerová & Sádlo 2000.
42. Dreslerova 2008.
43. Černý et al. 1996.
44. Dreslerová 2008, 275-276.
45. Pleiner 2000, 116-117.
46. Pott 1986.
47. Pleiner 2000, 118.
48. Pleiner 2000, 126.
49. Dreslerová 2008, fig. 167.
50. Waldhauser 2001, 23.
51. Drda & Rybová 1997, 96, 98-100, tab. 1.
52. Drda & Rybová 1995, 148.