Evolution and management of humid landscapes in northern Dauphiné (Rhone valley,France): Contribution of charcoal and wood studies

Abstract

Using various archaeological and geoarchaeological operations, charcoal and waterlogged wood assemblages have been sampled in the marshy areas from the lower Dauphiné (Rhone valley, France). Their identification allows reconstructing the evolution of the woody vegetation in relation to climatohydrological changes and with human practices in the plain since the mid Holocene. It appears that humid-land forests have experienced a shift from ash formations (dominating during Pre- and Protohistory) toward alder formations between the Bronze Age and Roman Period. That vegetation change seems to be linked with pastoral practices in which fire is used as a clearing and regeneration tool. The intense pastoral use of the plain, together with the humidity of the soils when not artificially drained, may also have prevented the development of dense and mature forests. Finally, we show that beech, which is currently absent from the plain, probably grew in the marshlands during the past.

Keywords

charcoal humid lowlands mid/late Holocene pastoralism Rhone valley vegetation dynamic waterlogged wood

Introduction

The humid plains of the lower Dauphiné areas (200 m a.s.l.) are located in the Rhone river catchment, at the foothills of the northern French pre-Alps, between Grenoble and Lyon, around the limestone plateau of ‘l’Isle Crémieu’ (400 m a.s.l.). Two important marsh areas, the whole surface of the Bourgoin-la Verpillière basin and a large part of the Basses Terres basin (Figure 1), occupied by large peaty formations and intensely drained since the beginning of the 19th century have been identified and studied by aerial photograph analysis (Bernigaud, 2012; Gaucher, 2011). In spite of its being occupied by marshes during the second part of the Holocene, the area has been intensively occupied and used by human societies since at least the late-Neolithic period, thus it is covered by drainage and irrigation ditch networks, some of them dating back to the very beginning of the Iron Age (2800–2500 bc), and the rivers have been canalised in several locations since the end of the Protohistoric period (Bernigaud, 2012; Gaucher, 2011). During the final Neolithic, the human occupation is indeed asserted by various archaeological evidence (Bocquet, 1967), among which the palafittic site discovered in the Paladru lake, in the southern part of the ‘Terres Froides’ area, halfway between Lyon and Grenoble, as well as the first evidence of forest clearing in the pollen record (Clerc, 1988; Doyen et al., 2013; Richard and Gauthier, 2007). Since then, human occupation in the region has been continuous during Pre- and Protohistory and reached a high density during the Gallo-Roman period (1st. century bc–ad 5th century) according to the abundance of archaeological sites, the development of hydraulic systems and the rise of anthropic markers in pollen diagrams (Berger et al., 2003; Bernigaud, 2012; Gaucher, 2011)

Figure 1.

Location of the study area and sites of investigation.

The evolution of the vegetal cover (cf. infra) since the Lateglacial at a regional scale is documented since the 1980s by several pollen records from long cores sampled in lakes and peats (Clerc, 1988; Doyen et al., 2013). Nevertheless, that regional framework given by pollen data probably hides local discrepancies linked to the variety of landscapes and human occupations. The lower Dauphiné area is indeed made of various landscape units inherited from a complex geomorphological and fluvial history, from the pre-Alps hills to the alluvial plain and the marshy lowlands. Soils, vegetation and human management of the landscape can be very different and the regional pattern cannot highlight the local history of the relationships between human societies and their environment. Thus, a focus on each kind of landscape unit and the use of palaeoecological methods that are able to trace the multiscalar diversity of the vegetation mosaic are needed to better understand that history. In this paper, we propose to look at the evolution and management of humid lands of the lower Dauphiné that were precociously colonized by Neolithic farmers, and increasingly during the Protohistoric and Historic period. We will use mainly botanical macroremains found on archaeological sites, as well as in natural sequences, which is the only way to reconstruct wet landscapes at a local scale and to identify the impact of the contemporaneous human activities on humid vegetation.

Because of numerous archaeological and geoarchaeological operations, we were able to study the management by human societies of these peculiar landscapes since the Neolithic. A dozen sites, mainly natural sequences and palaeohydrological structures (palaeochannel and drainage/irrigation ditches/canals), but also archaeological structures, have provided wood charcoal and/or waterlogged wood remains. Five localities: Saint-Romain de Jalionas-le Vernai, Frontonas-les Sétives, Bourgoin-Jallieu-les Vers, Le Bouchage-le Mollard and Les Avenières-le Pré de la Cour, gave enough data to document the evolution of the vegetation in these humid landscapes and to better understand the relationships between societies and their environment in contexts where the control of water fluxes is the main challenge for land exploitation.

Sampling sites

Two large humid areas were the subject of palaeoenvironmental research: the lower Delphinian Rhone river basin, namely ‘les Basses Terres’ basin to the eastern part, and the lower zone of the Bourbre river basin to the southwestern part, namely ‘Bourgoin-la Verpillière’ basin (Figure 1). These two basins, created by tectonic events, were scoured by glaciers during the Pleniglacial period (before 45,000 BP) (Coutterand et al., 2009; Forat, 1954; Mandier et al., 2003), in the Basses Terres basin. These basins were progressively filled by sandy to gravelly fluvioglacial deposits of alpine origin and/or by fine lacustrine laminated sedimentation during the Lateglacial and the first part of the Holocene. Their terminal accretion by sandy-gravel alluvium and, finally, peats and silts occurred since the mid Holocene. Today, and since a major avulsion phase resulting from the torrential activity of the Rhone river and its tributaries (Berger et al., 2008; Bravard, 2010; Salvador et al., 2004), which probably occurred during the 4th or 3rd centuries bc, the Rhone river flows in the Basses Terres basin on the east side of the Avenières molassic outcrop; however, its former course, on the western side, is still marked in the landscape by meander scars (cf. ‘palaeochannel’, Figure 1).

Present-day vegetation

According to the vegetation maps (Dobremez et al., 1986; Service de la carte de végétation du laboratoire de botanique et de biologie végétale de l’université de Grenoble, 1978), the vegetation mainly belongs to the Quercus/Carpinus series. Riparian and humid lands vegetation is dominated by alder (Alnus glutinosa), while evolved forests correspond to oak/hornbeam associations. Mountain influences from the pre-Alps determine the presence of beech (Fagus sylvatica) formations at height, above 800 m a.s.l. Fir (Abies alba) grows in the mountain vegetation belt, in the pre-Alps. Botanical surveys of the Serves and Grand Plan marshes (Villaret, 1999) describe small patches of oak-beech association growing in the humid plain, on small alluvial reliefs. Mediterranean influences determine the occurrence of vegetation belonging to the Delphino-Jurassian Quercus pubescens series, and even Mediterranean Quercus pubescens associations can be found in the Rhone valley around Lyon or Vienne. South of Vienne, evergreen oaks (Quercus ilex) can locally be found on south-oriented slopes (Ozenda, 1985). Forests actually only appear as patches, as this area is deeply anthropo-constructed nowadays.

Regional vegetation history

The colonisation of the lower Dauphiné alpine foreland by vegetation began after the definitive glacier retreat, around 18,000 years ago (Mandier et al., 2003). Postglacial climatic fluctuations first triggered the shift from Dryas steppic environment to birch and pine formations, before Holocene warming allowed the settlement of a more thermophilic vegetation. The large spread of hazel during the Boreal is followed by the establishment of forests. Fir regionally appeared early (Boreal/Atlantic transition) and spread during the Atlantic, while beech rather characterises the Sub-Boreal. From the Atlantic to the Sub-Boreal, the altitudinal vegetation stages settle, leading to the dominance of deciduous oak forests in the plains and on the hills and to that of fir formations not only at the mountain stage but also on the top of the hills, at lower altitude (below 500 m) than its present repartition (Clerc, 1988). It must be noticed that during the Holocene, according to pollen data, fir and beech evolutions are not synchronous in the area. If they form mixed forests during most of the Atlantic and Sub-Boreal stages, fir spreads earlier, even at middle altitude, and is supplanted by beech during the Sub-Boreal. Beech expansion is halted later by anthropic deforestation (Clerc, 1988). Tangible evidence of forest clearing first appears in pollen diagrams at the beginning of the 3rd millennium bc (end of the Neolithic period).

Sites from the Bourgoin-la Verpillière marsh

The first pool of sites is located in the marsh of ‘Bourgoin-la Verpillière’, which was drained up to the beginning of the 19th century ad. Several small rivers, now canalised, used to converge inside that basin.

The first site was surveyed prior to the construction of a hospital in the district of Bourgoin-Jallieu (Bleu, 2007), in an area known as ‘Les Vers’ (Figure 2). This toponym refers to a land used for the production of ‘green’ (as opposed to ‘dry’) fodder. A late-Neolithic settlement was discovered, as well as a Roman hydraulic network and a Roman mill. Charcoal fragments were recovered in the Neolithic site, in the ditch and the underlying peat level, in the mill canal and in the peaty layer covering the roman structures. The Roman ditch and the underlying peat also provided waterlogged wood.

Figure 2.

Sites from the Bourgoin-la Verpillière marsh (see Table 1 for calibrated dates): Bourgoin-Jallieu ‘Les Vers’: Location of the archaeological structures and of geaoarchaeological trenches, aerial view of the sampled Neolithic site and photographs sections through the hydraulic structures. Frontonas ‘les Sétives’: location of the trenches, sampled trench and log, and photograph of the uprooted trunks layer.

Table 1.

Radiocarbon dates.

Site	Sample	Date	Calib intcal 09	Lab. code	Dated sample
Bourgoin-Jallieu ‘les Vers’	Neolithic pit	4160±30 BP	2880–2631 bc	Poz-21182	Fraxinus sp., charcoal
Bourgoin-Jallieu ‘les Vers’	S15ter-1.52/1.60 m	3155±30 BP	1501–1322 bc	Poz-21123	Alnus sp., charcoal
Bourgoin-Jallieu ‘les Vers’	base roman mill channel-S142-T1	1855±30 BP	ad 82–234	Poz-21130	undeterminate twig
Bourgoin-Jallieu ‘les Vers’	S15 ter/T3	1790±30 BP	ad 132–331	Poz-21125	Fraxinus sp., Alnus sp. twigs, wood
Bourgoin-Jallieu ‘les Vers’	S15ter/T2	1615±30 BP	ad 387–539	Poz-21126	Alnus sp., Fagus sp., charcoal
Bourgoin-Jallieu ‘les Vers’	roman mill channel post-use plugging-S142-T24	1570±30 BP	ad 420–557	Poz-21129	Corylus-Betulacea, charcoal
Bourgoin-Jallieu ‘les Vers’	S59 US 8b	460±30 BP	ad 1412–1469	Poz-21121	Alnus sp., charcoal
Frontonas ‘les Sétives’	Trench 1, uprooted tree	2835±35 BP	1115–808 bc	Ly-13889	Fraxinus sp., waterlogged wood
Frontonas ‘les Sétives’	Trench 1, T22	1525±30 BP	ad 432–604	Ly-3960	wood
Frontonas ‘les Sétives’	Trench 1, T17	1840±40 BP	ad 75–313	Ly-3753	undeterminate wood
Saint-Romain de Jallionas ‘pont romain’	T5 US11, sandy Girondan channel	3670±35 BP	2191–1947 bc	Ly-4117 (SacA-6946)	charcoal
Saint-Romain de Jallionas ‘pont romain’	T5-T6, peaty layer	3190±35 BP	1523–1418 bc	Ly-4119 (SacA-6948)	charcoal
Saint-Romain de Jallionas ‘pont romain’	T5-T6, peaty layer	3045±45 BP	1418–1132 bc	Ly-4120 (SacA-6949)	seed
Saint-Romain de Jallionas ‘pont romain’	T5, alluvial sandy gravel layer (US 13–14)	4170±30 BP	2881–2633 bc	Ly-4830 SacA-10371	seed
Saint-Romain de Jallionas ‘pont romain’	T6, Silty filling of a ditch (US 27, 28, 32)	120±30 BP	ad 1680–1939	Ly-2273	Populus sp., charcoal
Saint Romain de Jallionas-T4	T4-m15, US 16: last pedogeneised filling of a ditch	1870±50 BP	ad 20–320	GifA-102201	charcoal
Saint Romain de Jallionas-T5	T4-m392, US 9	2490±60 BP	790–410 bc	GifA-102202	Quercus sp., charcoal
Les Avenières ‘Pré de la Cour’	S3, −3.00 m, phase I	2665±30 BP	894–794 bc	Saca 7486	wood
Les Avenières ‘Pré de la Cour’	S2, US2, phase I	2075±30 BP	180–1 bc	SacA-9744	Fagus sylvatica, charcoal
Les Avenières ‘Pré de la Cour’	S7-T4, phase II	1910±30 BP	ad 21–210	SacA-9743	Populus sp./Salix sp. wood
Les Avenières ‘Pré de la Cour’	S8-phase II	1935±30 BP	ad 1–130	SacA-7491	wood
Les Avenières ‘Pré de la Cour’	S7-US1, phase II	1850±30 BP	ad 85–235	SacA-7490	charcoal
Les Avenières ‘Pré de la Cour’	S7-US3, phase II	1740±30 BP	ad 234–389	SacA-7488	charcoal
Les Avenières ‘Pré de la Cour’	S9-US3, phase III	1600±30 BP	ad 404–540	SacA-7494	undeterminate twig
Les Avenières ‘Pré de la Cour’	S9-US5, phase III	1595±30 BP	ad 409–541	SacA-7493	undeterminate twig
Les Avenières ‘Pré de la Cour’	S9-US5, phase III	1580±30 BP	ad 415–547	SacA-7496	undeterminate twig
Les Avenières ‘Pré de la Cour’	S9-US6, phase III	1550±30 BP	ad 426–578	SacA-7495	undeterminate twig
Les Avenières ‘Pré de la Cour’	S9-US8, phase IV	715±30 BP	ad 1240–1385	SacA-7492	undeterminate twig
Les Avenières ‘Pré de la Cour’	S9-US3-4, phase II–III	995±30 BP	ad 980–1160	SacA-21425	Maloideae, wood (bow net)
Les Avenières ‘Pré de la Cour’	S9-US8, phase IV	695±35 BP	ad 1259–1390	SacA-9745	wood
Le Bouchage ‘Le Mollard’	S4, 119–123 cm	2815±35 BP	1107–847 bc	Gif. 11862	peat
Le Bouchage ‘Le Mollard’	S4, 195 cm	2975±35 BP	1263–1055 bc	Ly-3754	Poaceae, charcoal
Le Bouchage ‘Le Mollard’	S4, 258 cm	3050±70 BP	1488–1057 bc	Gif.A-102036	wood
Le Bouchage ‘Le Mollard’	S4, 290–310 cm	3105±30 BP	1440–1260 bc	Lyon-7228 (SacA 20254)	charcoal
Le Bouchage ‘Le Mollard’	S4, 310–320 cm	3345±30 BP	1729–1530 bc	Lyon-7229 (SacA 20255)	charcoal
Le Bouchage ‘Le Mollard’	S4, 537 cm	3310±70 BP	1744–1437 bc	Gif.A-102037	wood

The second site from the Bourgoin-La Verpillière marsh is located only a few kilometres northwest from ‘les Vers’, in the district of Frontonas, and is called ‘les Sétives’, which is an ancient toponym for humid meadows used for fodder production (Figure 2). Nowadays, the waters are canalized, but the area used to be at the confluence of two small rivers, the Catelan/Chéruy and the Loudon. The river Bourbre has been diverted to the Loudon bed since the 17th century ad, according to ancient map studies (Bernigaud, 2012).

The geoarchaeological operation, supervised by N Bernigaud and J-F Berger in 2006, aimed at revealing the fossil river beds and the irrigation canals, but it also allowed the study of sedimentary infilling of the basin and the discovery of a layer of fossil trunks dated from the Bronze Age (Figure 2). Besides the study of these waterlogged woods, charcoal assemblages from undated Pre- or Protohistoric levels of log 21 (Figure 2) were analysed, as well as punctual samples from Iron Age and Gallo-Roman hydraulic structures.

The third site, in the district of Saint-Romain-de-Jalionas, is located close to the western edge of the Isle Crémieu plateau, on the western side of the small ‘le Grand Plan’ marsh formed by the partial blocking of the natural drainage by a moraine belt (Figures 1 and 3). An important roman villa, continuously occupied from the second Iron Age to the central Middle Age, is excavated on the site called ‘Le Vernai’ (verne is a vernacular name of Alnus glutinosa) by R Royet (Royet et al., 2006). In order to better understand the exploitation of the humid lands, in particular by the study of the drainage/irrigation networks since the second Iron Age, mechanical geoarchaological surveys were carried out in the surrounding marsh and on the bank of the Girondan river (Berger et al., 2003; Royet et al., 2004). The results of wood charcoal analysis from a 6th century ad latrine studied in the framework of the Masters work of M Ploton (2001) will be compared with those obtained from charcoal assemblages dated from the Bronze Age to the Modern period recovered from the marsh.

Figure 3.

Sites from the Grand Plan marsh at Saint Romain de Jalionas: Location of the mechanical trenches, Roman and modern hydraulic networks and ‘Le Vernai’ roman villa in the marsh, topographical and geomorphical transect from the Girondan river alluvial plain to the moraine hills on the northeast side of the Grand Plan marsh and photograph of a section through a gaulish ditch.

Sites from the Basses Terres basin

In the marsh of ‘Les Avenières’ and further downstream in the Basses Terres basin, the investigations concern Rhone river palaeochannels.

At ‘Les Avenières-le Pré de la Cour’ (Figure 4), several trenches made with a mechanical digger allowed the restoration of the past fluvial dynamic of the courses of the Rhone river and its local tributaries from the 6th century bc to the 14th century ad. Fluvial geomorphological studies have hypothesised that if the Rhone river flowed in its northern bed since the 4th century bc, some water continued to follow the previous channel for a few centuries (phase I). When the the water supply in this channel stops, an oxbow lake forms (phase II), and finally sediment progressively fills the channel. This dead channel is later followed by the main southern tributary of the Rhone river (probably the Guiers river), which flows from the Chartreuse massif mountains (source at 1800 m a.s.l.) (phase III). A small local river coming from the southern molassic hills flows into the fossil Rhone channel during the 5th and the 6th centuries ad (phase IV). A wooden installation, as well as a bow net (Figure 4), have been found in the upper part of this alluvial formation and provide evidence of the exploitation of halieutical resources at the beginning of the Middle Ages (ad 986–1154). In the central part of the Middle Ages (ad 986–1390), a peaty layer covers the hydrological structures and the main part of the river bed (phase V). Finally, a canal is dug during the 14th century ad in the fossil palaeochannel and across the upper peaty layer (phase VI). All these phases provided charcoal, in variable amounts, and sometimes waterlogged wood.

Figure 4.

Sites from the Basses Terres basin (see Table 1 for calibrated dates): Les Avenières ‘Pré de la cour’: map showing the location of the palaeomeander and of the trenches, correlation of the trenches and geomorphological/palaeohydrological phases and photograph of the wood bow net found in trench S9. Le Bouchage ‘Le Mollard’: map showing the location of the palaeomeander and of the drilling, section through the palaeochannel and studied drilling core.

The last sampling site, ‘Le Bouchage-Le Mollard’, is a paleomeander of the Rhone river which stopped being supplied with water during the mid-Holocene period (around 1600 cal. bc) (Salvador et al., 2004), contemporaneous with the Bronze Age culture in the region. Charcoal assemblages come from a core drilling and were taken every 5 or 10 cm (Figure 4).

Methods

Taxonomic identification

Wood and charcoal were recovered by wet sieving (2 mm and 0.5 mm) of sediment samples, except in a few cases when very big wood fragments (trunks, posts) were found. In that case, only a piece was taken in order to perform the taxonomical identification. Charcoal fragments were air-dried before storage. On the contrary, waterlogged wood was kept in water to avoid the development of aerobic microorganisms and the collapse of anatomical structures that occurs when uncharred wood dries.

The anatomical identification of the charcoal fragments was carried out with a reflected light bright field/dark field microscope (50× to 500×). Charcoal fragments were hand-broken in order to allow the observation of the three anatomical sections of wood (transverse, longitudinal-tangential and longitudinal-radial). For waterlogged wood fragments, thin sections (ideally three per sample, along the three anatomical planes) were taken using a razor blade and observed in a drop of water on a microscope slide, using a transmitted-light microscope, at magnification 50× to 400×. In both cases, the observed structures were compared with those described in the atlases of wood anatomy (Schweingruber, 1978, 1990) and with an extensive reference collection of present-day burnt woods. Most of the time, the whole sample was studied, as each sample only provided a low number of fragments.

Representation of the results and chronological frame

Despite the variations in the number of fragments from one sample to another, we propose a schematic representation of the results in the form of bar diagrams based on the relative abundance of the taxa. The samples made of a small (<30) or very small (<10) number of fragments are indicated. In case of rapid sedimentation and nearly contemporaneous assemblages, the wood or charcoal results from several sediment layers have been summed (in order to improve the statistical accuracy of the assemblages and to allow calculating relative abundances), but the two kinds of remains (charred versus waterlogged wood) were never mixed. The chronological framework was established on the basis of numerous ¹⁴C dates (Table 1), on the chrono-cultural attribution of the archaeological material and on stratigraphic correlations.

Deposition contexts and macroremains transportation

The interpretation of charcoal and wood assemblages must take into account their context, as the spatial scale involved may vary considerably (Delhon et al., 2003; Schroedter et al., 2012). On archaeological sites, charcoal mostly corresponds to ligneous fuel. When it is sampled in structures where it progressively accumulated over a long period and over several uses of domestic hearths, the charcoal spectra reflects well the composition of the woody vegetation around the site (Chabal, 1994). In hydraulic structures, charcoal and waterlogged wood correspond either to in situ vegetation or to the alluviation of branches that fell from the trees growing along the water or to charcoal from fires that occurred higher in the catchment (Delhon, 2005). In the Bas Dauphiné, the watershed of each hydraulic system is well known and controlled by photo-interpretation, and geomorphic and topographic studies (Berger, 2001), which gives a reliable scale of analysis. In palaeosoils, peat layers, marsh deposits or sometimes in oxbow lakes, waterlogged wood and charcoal mainly come from local vegetation (local run off from neighbouring banks, falling branches from in situ vegetation or of charcoal from in situ fires) and transportation is reduced or absent (except during extreme flood events). In oxbow lakes, both situations can occur: ecofacts can be transported from upper watersheds by floods and/or be produced by the riparian vegetation during periods of low fluvial activity and authigenic peat formation. Thus, an exogenous input of mountain species is possible in the Basse Terre area whenever the Rhone sedimentation is concerned, as this river has a large upstream catchment (more than 4000 km²), but also when the Guiers river, a Rhone tributary, is involved, as its catchment exceeds 1800 m a.s.l. in the Chartreuse pre-Alp massif. Nevertheless, whenever neither the Rhône nor the Guiers are involved in the sedimentation, the source area remains restricted by the topography. The basin of Bourgoin-la Verpillière and the Grand Plan marsh form flat landscapes of several square kilometres, disconnected from the Rhone and Guiers rivers catchments. They are delimited by the slopes of the Crémieu plateau (400 m a.s.l.) in the north and the Terres Froides hills (600 m a.s.l.) in the south. Long distance input is thus unlikely in that zone most of the time.

The origin of forest fire (natural versus anthropogenic) can be discussed at the regional scale, in relation to the climatic context in which the fire occurs or at the local scale by taking into account the anthropic features and contexts associated with the charcoal assemblages (Berger and Thiébault, 2002). Nevertheless, when dealing with highly anthropized contexts (e.g. the Grand Plan marsh, under direct influence of a villa, drained, irrigated and used for agropastoral purposes during the second Iron Age and the Roman period), an anthropic origin is often the most probable, all the more when the vegetation is not particularly fire-prone (e.g. wet contexts). When the hydraulic networks are working, charcoal fluxes are often associated with open landscapes and pollen assemblages characteristic of farming activities (rise of apophytes and remains of parasitic worms from mammals’ intestines) and thus could be mainly linked to the maintenance of the ditches (destruction of the vegetation growing on the banks using fire) or to agro-pastoral practices (clearings, pastoral fires) (Berger et al., 2003).

Results and individual interpretation of the sites

Sites from the Bourgoin-la Verpillière marsh

Bourgoin-Jallieu-Les Vers

At Bourgoin-Jallieu-Les Vers, 621 charcoal fragments and 72 pieces of wood have been identified (Table 2, Figure 5). They represent 13 taxa (13 in charcoal assemblages, 5 in wood assemblages).

Table 2.

Results of wood and charcoal analyses from Bourgoin-Jallieu ‘les Vers’.

Samples
Context	Neolithic pit	Trench S15 ter			Several samples	Trench S15 ter		Trench S59
Context (precision)			T9		Roman mill	14C2	14C1	US8C
Sample precision		−1.52 m to −1.60 m	−2.25 m to 4.10 m
¹⁴C dates	4160±30 BP	3155±30 BP			1855±30 BP/1570±30 BP	1615±30 BP		460±30 BP
Calib. ¹⁴C dates (bc/ad)	2880–2631 bc	1501–1322 bc				ad 387–539		ad 1412–1469
Stratigraphic correlation dates			Between 1501 and 1322 bc and ad 132–331	Before ad 132–331	Between ad 82–234 and ad 420–557
Chronocultural attribution	Late Neolithic		Early Roman Empire		Roman (Drag 33 ceram.: ad 80–240 at the base of the sequence)		Medieval to modern

Charcoal

Alnus sp.		141			207	22	1
Corylus avellana	18				13		6
Cornus sp.					2			1
Fagus sylvatica				2		1
Frangula alnus
Fraxinus sp.	152				3		2
Pinus t. sylvestris	2
Maloideae	1							8
Populus sp./Salix sp.				9
Quercus deciduous	7				2		1	3
Sambucus sp.					4
Ulmus sp.								11
Viburnum t. opulus				2
Total	180	141		13	231	23	10	23	621

Wood

Alnus sp.		45		2
Fagus sylvatica			4
Fraxinus sp.				1
Quercus deciduous
Populus sp.			20
Total	0	45	24	3					72

Figure 5.

Charcoal and wood diagram of Bourgoin-Jallieu ‘Les Vers’. Black: charcoal, > 30 fragments; dark grey: charcoal, < 30 fragments; light grey: waterlogged wood; dots: very small assemblages or abundance < 1%.

The late-Neolithic period (2880–2631 bc) is characterised by the dominance of ash (Fraxinus sp., 84%), accompanied by forest taxa (deciduous Quercus, 3%) and moreover light-demanding taxa (Corylus avellana, 10%). The Prehistoric settlement bears evidence of rearing activities, ash leaves and branches may have been gathered to be used as fodder, as is the case in various Neolithic pastoral sites (Argant et al., 1991; Delhon et al., 2008; Laederich and Thiébault, 2004; Thiébault, 2005). Anyway, the local edaphic and climatic conditions are favourable for the spread of that species, i.e. sufficient soil and atmospheric humidity, but absence of continuously wet soils (Rameau et al., 1989). Together with hazel, it may form pioneer/postpioneer formations before the development of more mature and shade-loving oak forests. A bit further south, in the middle Rhone valley, the final Neolithic appears to be a period of great development of ash, probably linked with an increase in humidity and a shift in pastoral practices (Delhon, 2005). The inhabitants of the site may have also gathered fuel in oak forests or on their edges, and the exploited territory could also involve drier lands, as shown by the occurrence of pine (Pinus t. sylvestris), the presence of which in the marsh itself is unlikely. Later, at 1501–1322 bc, the presence of Alnus in the form of waterlogged wood is evidence of its development in situ, which fits with the settling of a peaty level during the late Neolithic and Bronze Age in different sites of the basin. Alder can grow on very wet and poor soils (Rameau et al., 1989); its settlement to the detriment of ash may be linked to the shift in edaphic conditions. From the formation of this layer and until we sampled the botanical remains, the humidity was sufficient to allow the preservation of uncharred wood. In such a wet context, where soils are continuously wet, the occurrence of an alder charcoal layer (100% of the 141 identified fragments) could be linked with pastoral practices, as there is little chance that a natural fire could destroy such a humid vegetation. As alder is avoided by cattle (Darinot and Morand, 2001), browsing favours its development to the detriment of more palatable trees. Thus, fire following cutting and seasoning on the field, is a good way to get rid of Alnus thickets during the dry season and to restore productive pastures (Bernigaud, 2012).

The Roman hydraulic structures (ditch, irrigation and mill canal) also provided a lot of alder, mostly charcoal, showing that the drainage was not optimzal and also suggesting a clearing of the pastures and/or the bank of the ditches using fire. Several drier periods are documented in the area during the Roman period (Arnaud et al., 2012; Berger and Bravard, 2012), but even during more humid phases the seasonality and the artificial drainage probably allowed summer fire clearing of dead but also perhaps standing biomass. More mature forests are also recorded with few fragments of oak and beech (Fagus sylvatica), but their in situ location in the marsh is not obvious. These remains may have been transported by water from the molassic hills of the Bourbre river catchment, and thus may suggest that Fagus occurred on its slopes. Nevertheless, these reliefs do not exceed 600 m a.s.l., while nowadays mesophilous beech forests develop over 800 m a.s.l., but the species can be found in pockets from 400 m a.s.l. on north-oriented slopes in the lower Dauphiné region (Wywial and Bravard, 1997), and has even been described on the plain, as small patches, in fresh and humid environments such as marshes (Villaret, 1999). If beech can grow on wet sites, its flat root system makes it very susceptible to falling (Lebourgeois and Jabiol, 2002) which reduces the chances of it reaching certain age and size and of gaining dominance on these sites.

The peaty soil above the hydraulic structures gave only few charcoal fragments, but the qualitative interpretation of the taxa list suggests an open but quite evolved forest assemblage (Ulmus sp., Maloideae, Cornus sp., deciduous Quercus). The natural improvement of the drainage during the Medieval Climatic Optimum and the decline of anthropogenic activities in the lower humid areas of the Isle Crémieu between the 7th and the 10th centuries ad (Berger et al., 2008; Bernigaud, 2012) could have favoured the evolution of the vegetation toward more mature formations. A text from the end of the 13th century (Charte d’affranchissement de Bourgoin, 1298) mentioned a forest, ‘le bois du Vert’, in the area (Comte, 1984).

Frontonas-Les Sétives

The charcoal assemblages (Table 3 and Figure 6) come from log 21, in which charcoal layers alternate with the peat loam infilling of the marsh. The chronostratigraphic context of the sequence indicates an early- to mid-Holocene age for the main charcoal assemblages from log 21. The bottom of the sequence provided pollen assemblages in which non-arboreal and pine pollen dominates (Argant, unpublished data, 2007), which led us to attribute it to the Lateglacial period, and the top corresponds to the lateral deposits of a Roman canal system, radiocarbon dated. One sample (T13) provided 407 of the 525 identified fragments of the sequence, and showed a great biodiversity (at least 15 taxa). Pine and oak, which are absent from this sample, have been identified in great quantity for the first and maybe occasionally for the second one in older levels of the same site, thus the whole biodiversity of the sequence attains 17 taxa. Waterlogged wood has been sampled in a level of lying trunks, which was discovered in several trenches. This accumulation of uprooted trees is dated to 2835 ± 35 BP (1115–808 bc), which corresponds to the late Bronze Age period. Twenty-six of them have been sampled and identified, six taxa were recognized, including beech (absent from the charcoal spectra), which brings the total biodiversity to 18 taxa.

Table 3.

Results of wood and charcoal analyses from Frontonas “les Sétives”.

Samples
Context	Trunk level	Log 21
Context (precision)
Sample precision		T1	T2	T3	T9	T11	T13	T14	T15
¹⁴C dates	2835 ± 35
Calib. ¹⁴C dates (bc/ad)	1115–808 bc
Stratigraphic correlation dates		Before 1840 ± 40 (ad 75–313)

Charcoal

Acer campestre							4
Alnus sp.							25	2
Alnus sp./Betula sp.							1
Betula sp.							1	23
Carpinus betulus							4
Cornus sp.							3	1
cf. Cornus sp.							1	1
Corylus avellana				1			40
cf. Corylus avellana							4
cf. Ericaceae							1
Ericaceae t. Ledum palustre							4
Euonymus europaeus							2
Frangula alnus							19		3
cf. Frangula alnus							2
Fraxinus sp.					4	3	259		7
cf. Fraxinus sp.							2
Pinus t. sylvestris		13	1	45
Maloideae							16		10
cf. Maloideae								1
Prunus t. avium/padus							7
Prunus large rays							2
cf. Quercus sp.		1			1
cf. Viburnum sp.							2
Viscum album					1		6
cf. Viscum album							2
Total		14	1	46	6	3	407	28	20	525
Unidentifiable angiosperm dicotyledon			1		2		13
Unidentifiable gymnosperm					1
Unidentifiable		2	5		4	3	3

Wood

Alnus sp.	4
Corylus avellana	2
Fagus sylvatica	1
Fraxinus sp.	16
Quercus deciduous	3
Total	26									26
Unidentifiable	1

Figure 6.

Charcoal and wood diagrams of Frontonas ‘Les Sétives’. For the charcoal bar diagram: black: > 30 fragments; dark grey: < 30 fragments; dots: very small assemblages or abundance < 1%; stars: sole or very dominant taxon in small assemblages.

Most of the charcoal fragments from the base of the peat layer were pine fragments, which fits with the pollen data (Argant, unpublished data, 2007) and could refer to Lateglacial or early-Holocene vegetation (samples T1, T2, T3). The upper part (T9 to T15) of the sequence records the settling of a humid forest, quite diverse (particularly in sample T13), and dominated by ash and other riparian taxa (buckthorn: Frangula alnus, alder, birch: Betula sp., probably Labrador tea: cf. Ledum palustre). That ecosystem indicates that the ash forest was exploited by the late-Neolithic population of Bourgoin-Jallieu-les Vers (supra), but no ¹⁴C dating is available yet, so the contemporaneousness of both records is only hypothetical. The layer of trunks, dated from 1115 to 808 bc, probably witnesses an acute destruction of the paludal forest growing in the marsh, the cause of which is unknown (storm?). It also records nearly the same vegetation as the late-Neolithic site of Bourgoin-Jallieu-les Vers (dated from 2880 to 2631 bc), characterized by the dominance of ash (16 among the 26 identified trunks), in spite of its being much more recent. It must be noted that one of the trunks (about 80 cm diameter) was identified as beech (Fagus sylvatica), which is currently absent or very rare at low altitudes in the area. As that level corresponds to in situ uprooted trees, the presence of beech growing in the marsh during the Bronze Age must be considered.

Saint-Romain-de-Jalionas-le Vernai

At Saint-Romain-de-Jalionas, archaeological charcoal was obtained from one latrine of the 5th century ad occupation (Ploton, 2001) as well as charcoal from the surrounding marsh, which was exploited by the inhabitants of the site at least since the second Iron Age (Berger et al., 2003; Royet et al., 2004, 2006). These botanical assemblages (Table 4) came from trenches through the marsh. They were sampled in the sediment infillings of the ditches or of the marsh formations themselves (T4, T5, T6), in a palaeochannel and in the floodplain of the Girondan river (T57, Ploton, 2001) that flows in the western part of the site (Figure 3). They are dated from the late Neolithic/Bronze Age transition (3670 ± 35 BP, 2191–1947 bc) to the Modern or Contemporaneous periods (ditches). The amount of identified charcoal fragments is very variable from one period to another, and the total biodiversity of the marsh assemblages reaches at least 24 taxa (Table 4). In the latrine, at least 15 taxa were identified among the 254 identified charcoal fragments (Table 4, Figure 7).

Table 4.

Results of charcoal analyses from Saint-Romain de Jallionas ‘le Vernai’.

Samples
Context	Marsh									Villa
Context (precision)	T5/T6	T5/T6	T4/T57	T57 (ditch near the villa)		T4	T5/T6	T5	T5/T6	Latrine
Sample precision	Several u.s.	M15 u.s. 16	M392 u.s. 10	Several u.s.	Several u.s.	M392 u.s.9	u.s. 11, 12, 13	u.s. 11	Several u.s.	c1094
¹⁴C dates		1870±50						3670±35
Calib. ¹⁴C dates (bc/ad)		ad 20–320						2191–1947 bc
Stratigraphic correlation dates			2nd–3rd centuries ad				Between 2190–1940 and 1418–1132 bc		Before 3670±35 BP (2190–1940 bc)
Chronocultural attribution	Mod. to Cont.			2nd century ad (ceramic)	40 bc–ad 100 (ceramic)	8th–5th centuries bc				5th century ad

Charcoal

Abies alba				10	2					8
Acer campestre							24
Acer platanoides/pseudoplatanus							1
Acer sp.				35	2	1				2
Alnus sp.				3	2		38	6	6	7
Betula sp.				3					4
Betulaceae/Corylaceae				46	2					17
Carpinus betulus				161	40					33
cf. Carpinus										9
Cornus sp.				2
Corylus avellana				210	19					3
cf. Corylus avellana										1
Fagus sylvatica				146	47		4	1		12
cf. Fagaceae										1
Fraxinus sp.							33	5	4
Fraxinus sp./Juglans regia		1		18	5				1	2
Hedera helix				2
Ilex aquifolium					1
Juniperus sp.				6
Oleaceae										1
Pinus t. sylvestris					8				3	1
Poaceae									3
Maloideae		1		37	7					37
cf. Pomoideae										14
Populus sp.	2
Populus sp./Salix sp.										5
Prunus sp.				5	5			1		8
Quercus deciduous	5	11		355	82	3	25	4	1	71
Quercus cf. sclerophyllous				2						1
cf. Quercus										7
Rhamnaceae				9	1					4
Sambucus							1
Ulmus sp.			11	12	12					8
cf. Ulmus										2
Vitis vinifera					5		1
Total	7	13	11	1062	240	4	127	17	14	254	1749
Source			Partly from Ploton (2001)							Ploton (2001)

Figure 7.

Charcoal diagrams from the marsh (bar diagram) and the villa (circle graph) of Saint Romain de Jallionas ‘le Vernai’. For the bar diagram: black: charcoal, > 30 fragments; dark grey: charcoal, < 30 fragments; dots: very small assemblages or abundance < 1%.

In the marsh, during Prehistoric times, the vegetation is quite diverse and is mainly made up of species that grow on humid soils: alder, ash, birch, and grasses (Poaceae). A more precise identification of grass charcoal is usually not possible, but the preserved fragments showed diameters of several millimetres, which favours large grasses, and most probably reeds (Phragmites communis) that grow on wet soils and around humid zones. More mature forests also exist, represented by oak, beech and associated species such as maples (Acer spp.) or Prunus. They could develop on the fluvioglacial microtopographies (0.5–1 m above the mean level of the marsh), formed during the Lateglacial or the early Holocene, which are still visible in the landscape and are favourable to the local development of drier soils and associated vegetation (Villaret, 1999).

The occurrence of pine (Pinus t. sylvestris) can be an echo of the vegetation located on the drier stations on the western slopes of the Crémieu plateau. The Girondan river is likely to have brought charcoal from the foot and the slopes of the plateau. Nevertheless, a reworking of older charcoal, due to the erosion of Lateglacial levels, is also possible.

During the Roman period, the villa was a large establishment, especially during the 1st and 2nd centuries ad (Royet et al., 2006), and the surrounding humid lands were extensively exploited (Berger et al., 2003; Royet et al., 2006). During the end of the 1st century bc until the 3rd century ad, the analysis of charcoal from the marsh reveals a forest-vegetation dominated by oak and hornbeam, as today. If the oak/hornbeam association seems to be the most common one, more light-demanding and cooler facies exist, with Rosaceae (Maloideae and Prunus) for the former and beech, holly and maybe fir for the latter. Compared with Pre-/Protohistoric levels, humid vegetations seems to be decreasing and dryland species are observed. It could be the sign of an improvement in the drainage of the marsh, which is documented by geoarchaeological data (Berger et al., 2003) and fits with a regional drier and warmer period (Frisia et al., 2005; Holzhauser et al., 2005; Magny, 2004) with the exception of a short humid period around the change of era (Berger and Bravard, 2012). It can also be an indication that pastoral land clearing reached the fringes of the marsh land, as shown by the abundance and diversity of oak forest species. The drainage of humid soils and a probable active fight against these pervasive species may have caused the decrease of hydrophilous taxa in charcoal assemblages.

From the end of the Roman period until the 6th century ad, the latrine of the villa provided abundant data. The charcoal from other areas of the villa has not yet been analysed. The charcoal assemblage shows that different trees were exploited for fuel by the inhabitants of the site, but most of the species can be found in the immediate surroundings, in a deciduous Quercus and Carpinus betulus open forest and in riparian formations. Both kinds of vegetation could have formed a continuum from dense to more open and from very wet to better drained areas. Drier stands may have existed, perhaps in the close morainic hills located in the western and the northern parts of the marsh, as shown by the occurrence in very low amounts of xerophytic taxa, but these may also have been brought from distant areas by the inhabitants. Beech has been identified several times (12 fragments, 4.7% of the identified fragments), together with fir (Abies alba) (8 fragments, 3.1%). Both species are currently growing out of the studied area, in the hill/mountain vegetation belt. As these charcoal fragments have been found on an archaeological site, they could be the remains of allochtonous trees, whose wood has been brought on site by the inhabitants, to be used as fuel or for a more specialized purpose (tool, furniture, construction).

After the 6th century, the drainage network and the agro-pastoral exploitation of the marsh were probably abandoned and the aquifers increased in the marshes at a regional scale, and thus peat development was favoured (Berger, 2003). Charcoal in sediment becomes very rare, probably because of the cessation of agro-pastoral fires and also the increase of the aquifers in the marshes at a regional scale which is not favourable to vegetation fire (Berger, 2003)

Sites from the Basses Terres basin

Les Avenières-Pré de la Cour

At Pré de la Cour, all the phases of evolution of the hydrosystem (Figure 4) from the shift of the Rhone channel from the south to the north of the Avenières outcrop to the canalisation of the Huert/Bièvre waters have provided charcoal, in variable amounts, and sometimes waterlogged wood. The 639 identified fragments of charcoal refer at least to 21 taxa, among which at least 14 are also represented in the waterlogged wood assemblage, composed of 337 identified fragments (Table 5, Figure 8).

Table 5.

Results of wood and charcoal analyses from les Avenières ‘Pré de la Cour’.

Samples
Context	Palaeo-Rhone channel	2nd channel	3rd channel	Peat	Canal
Context (precision)	Rhone (or Guiers) active channel	Rhone (or Guiers) oxbow lake	Local river (Huert) channel		‘Canal de l’Huert’
Sample precision	S3/−300; S2/80 to 55; S8/T7T8	S7/T5 to G1	S8/T5,T2; S9/several samples	S9/T10 to T12	S3/T1, T2
¹⁴C dates	2075±30 BP	1910±30 BP to 1740±30 BP	1600±30 BP to 1550±30 BP	715±30 BP, 695±35 BP
Calib. ¹⁴C dates (bc/ad)	180–1 bc	ad 21–210 to ad 234–389	ad 404–540 to ad 426–578	ad 1240–1385, ad 1259–1390
Stratigraphic correlation dates
Dates from texts					14th century ad?
Chronocultural attribution	2nd century bc/2nd century ad	2nd century ad/4th century ad	5th century ad/6th century ad	Medieval	Historic

Charcoal

Abies alba	10
Abies alba/Juniperus sp.	1
Acer sp.	2			2
Alnus sp./cf. Alnus sp.	20		17	26
Betula sp.	1
Cornus sp.	1		1
Corylus avellana	24		19	3
cf. Euonimus			1
Fagus sylvatica	86		4	1
Fraxinus sp./cf. Fraxinus	14		1
Juniperus sp./cf. Juniperus	3	1
Pinus sylvestris	6		5		1
Poaceae including cf. Phragmites communis			17	182
Maloideae	6		8
Populus sp./Salix sp.	6		5	35
Prunus sp. (including P. dulcis)	1		5
Quercus deciduous	23		68	11	2
Quercus cf. sclerophyllous			2
Rhamnaceae (including Frangula and Rhamnus)	2		3	5
Taxus baccata	1
Ulmus sp.	1		5
cf. Viburnum	1
Total	209	1	161	265	3	639
Sample precision	S3/ −300; S8/T7T8	S7/T7, T4, G3	S9/several samples	S9/T10 to T12	S3/T1

Charcoal

Abies alba	7
Acer sp.	4
Alnus sp./cf. Alnus sp.	13	8	198	12
Cornus sp.			2
Corylus avellana			2
Fagus sylvatica	2
Fraxinus sp./cf. Fraxinus	1
Pinus sylvestris	2
Maloideae			6	2
Populus sp./Salix sp.	28	2	3	2
Prunus sp.			1
Quercus deciduous	2		10		2
Rhamnaceae (including Frangula and Rhamnus)			3	1
Ulmus sp.			24
Total	59	10	249	17	2	337

Figure 8.

Les Avenières ‘Pré de la cour’, charcoal and wood diagram. black: charcoal, > 30 fragments; dark grey: charcoal, < 30 fragments; light grey: waterlogged wood; dots: very small assemblages or abundance < 1%.

Between the 2nd century bc and the 2nd century ad, while the channel was still supplied by the Rhone or Guiers rivers flood deposits, waterlogged wood records a riparian formation with alder and poplar/willow as well as forests at least partly located on the pre-alpine slopes and including beech, fir and yew (Taxus baccata). Pine is also present in the assemblage, and probably grew out of the humid alluvial plain. Waterlogged posts have been observed further in the palaeochannel; it is possible that some of the identified wood fragments are pieces of these human constructions. The charcoal spectrum from the flood layers located on the convex bank is more diverse, riparian formations are less conspicuous but more species are involved (Alnus sp., Populus sp./Salix sp., Rhamnaceae, Fraxinus sp., Ulmus sp., Betula sp.). Forest and edges/undergrowth trees and shrubs are numerous, and beech/fir formations are represented. Beech reaches nearly 40% of the identified charcoal fragments. Such numbers of taxa usually growing at higher altitude confirm that, despite the settling of a new Rhone riverbed north from the Avenières outcrop, the channel is still supplied by flood water from an extra-local river, whose catchment reaches the mountain vegetation belt (Rhone river – split in two courses, or the Guiers river). That exogenous input of macroremains from upstream vegetation is confirmed by the change in the spectra when the meander is disconnected from the fluvial network and evolves as an oxbow-lake (2nd/4th centuries ad). Only a riparian formation, likely very local, is recorded in the form of waterlogged wood.

During a new detrital phase linked to the deterioration of hydrosedimentary conditions (5th/6th centuries ad), charcoal and wood accumulate again in the part of the channel then occupied by the Huert/Bièvre river. In spite of a probably very local catchment, the wood spectrum is diverse, and records a riparian formation in which alder dominates, accompanied by elm (Ulmus sp.), deciduous oak and various taxa characteristic of humid soils and/or of forest edges. Charcoal shows that fire, which is most likely an agro-pastoral tool, mostly affects oak forests (deciduous oak, accompanied by various edges and undergrowth species). The biodiversity is quite high and beech is observed. In this context of a microlocal origin of the sedimentation (Huert and Bièvre rivers have small local watersheds), it must have grown in the nearby molassic delphinian ‘Terres froides’ hills, which culminate around 300/400m a.s.l. Among various riparian taxa (alder, poplar/willow, Rhamnaceae, ash, elm), monocots charcoal (likely Phragmites sp.) has been identified and reaches 10% of the spectra. It could indicate a clearing of the bank of the channel. Fire clearing of reed formations is confirmed by the charcoal sampled in the peat layer formed during the central Middle Ages which covers the Huert-Bièvre alluvial formations: 70% of the 265 identified charcoal fragments belong to the Poaceae/Phragmites sp. group, accompanied by humid lands trees (alder, poplar/willow). Natural fire is unlikely in this humid context, and human practices are more probable. They can be linked with the use of the marsh as grazing land and the regeneration of the pastures using fire. Reeds formation is not recorded by waterlogged wood. We assume that it is due to the poor preservation of uncharred grasses. On the other hand, it records well the importance of alder, whose development on humid soils can be linked with pastoral activities (Bernigaud, 2012; Darinot and Morand, 2001).

Le Bouchage-Le Mollard

Those samples come from a 550 cm core in a paleomeander of the Rhone river which changed its course by an avulsion process around 1600 bc. The abandoned channel progressively filled from that date to 1000 bc, which corresponds to the middle to late Bronze Age period. Charcoal is rare in the lower part of the sequence, and very abundant between 150 and 200 cm deep, which corresponds to the period 1200–1000 bc (Berger and Salvador, 2010), but the assemblage appeared to bear a very low biodiversity (Table 6, Figure 9).

Table 6.

Results of charcoal analyses from le Bouchage ‘le Mollard’.

Samples
Context	S4 – Rhone palaeochannel
Depth (cm)	150–160	160–170	170–182	182–192	192–201	201–209	229–239	268–280	304–309	460–470	470–480	480–490	490–500	500–510	520–530
¹⁴C dates					2975±35 BP				3105±30 BP
Calib. ¹⁴C dates (bc/ad)					1263–1055 bc				1440–1260 bc
Stratigraphic correlation dates	Before 2815±35 BP (1107–847 bc)					Before 3050±70 BP (17,488–1057 bc)				Between 3345±30 BP (1729–1530 bc) and 3310±70 BP (1744–1437 bc)

Charcoal

cf. Acer sp													1
Fraxinus sp.													2
Dicotyledon angiosperm													2		1
cf. Phragmites communis	8	5	26	8	8	13			1
Poaceae	10	35	266	49	132	25	2	1		3	2	2	1	9
Monocotyledon angiosperm								2		3
Total	18	40	292	57	140	38	2	3	1	6	2	2	6	9	1	617
Unidentifiable	1						1								2

Figure 9.

Le Bouchage ‘Le Mollard’, charcoal diagram. black: > 30 fragments; dark grey: < 30 fragments; dots: very small assemblages or abundance < 1%.

In the lower part of the sequence, ash and maybe maple (Acer sp.) have occasionally been identified. Broadleaved tree charcoal fragments were not only rare in the core but also small and poorly preserved, which made their identification very difficult. Monocots, probably Phragmites sp, are present since the beginning of the charcoal record and are rapidly (since 490 cm) the only identified taxon. In the upper part of the sequence (209–150 cm), it is found in great quantities. Fire clearing and regeneration of the reed formation growing in the humid plain around the dead meander is most probably linked with pastoral practices at the end of the Bronze Age. A pollen analysis has been performed (Richard and Gauthier, 2007) but unfortunately pollen was not preserved above 2.30 m deep, which is the depth where reed charcoal begins to be abundant. In the lower levels, human impact pollen indicators are not conspicuous.

Discussion

In spite of the heterogeneity of the data, these results allow a first pattern of the evolution of humid lands under human pressure in the two studied marshes to be drawn (Figure 10). The data chronologically spread from the late Neolithic to the Modern period, but Iron and Middle Age periods are lacking and the accuracy of the dates is highly variable (direct ¹⁴C dates, stratigraphic correlation, archaeological attribution). Nevertheless, in all the sites two main types of vegetation are recorded: marsh or riparian vegetation and more or less evolved oak forests. The general pattern shows, after the probable settlement of pine at the beginning of the Holocene, that ash-alder forests characterized these humid lowlands at the end of the Neolithic and during the Bronze Age. The establishment of these ash-dominated formations is not documented by our data. Later, after a second gap (corresponding to the Iron Age) in the documentation, alder spread to the detriment of ash. Besides these paludal formations, more mature forests sometimes get to settle, but they always remain more or less open and hygrophilous.

Figure 10.

Holocene evolution of the vegetation in the Bas Dauphiné marshes according to charcoal and waterlogged wood data.

Comparison with the vegetation history in other humid areas is not easy, as our investigations are based on a novel methodology. The analysis of macrobotanical remains from natural sequences or from fossil hydraulic structures is not widely used and the sedimentary context – which indicates the catchment area – is a major factor of variation of botanical assemblages (Schroedter et al., 2012). Moreover, we showed that the observed vegetation changes are strongly linked with local variation of the soils (moisture, linked with the efficiency of natural or artificial drainage) and of the anthropic pressure (pastoral use of the marshes). At the considered scale, global changes, such as climate variation, seem to only have a secondary impact on vegetation trends.

Humid lands evolution

It seems that all the humid-land forests have experienced a shift from ash formations (dominating during Pre-Protohistory) toward alder formations, between the Bronze Age and Roman period. Alder is better adapted than ash to waterlogged soils (Rameau et al., 1989), thus one can suppose that its development is linked with an increase of the edaphic humidity in the lower lands. Nevertheless, the replacement of ash by alder corresponds with the Roman period, when drainage networks were the most developed in the two studied basins (before the 19th century drainage of the marshes). In the Grand Plan marsh as well as in the Avenières death Rhone valley, marsh hydromorphic and fluvial soils are characterized during the Roman period by a subangular to prismatic structure and the strong increase of oxidizing features. These features are clearly distinct from that of the grey massive hydromorphic or peaty soils that are observed before and after this period in the same areas (Berger, 2003; Berger et al., 2003). This change in soil features reveals the seasonal fluctuation of the aquifers and the improvement of the drainage. Thus, it is unlikely that alder benefited from an increase of soil humidity. Its spread may have been favoured by the pastoral exploitation of the marshes, as it is generally avoided by cattle (Darinot and Morand, 2001). The abundance of alder charcoal would thus witness the clearing and regeneration of pasture using fire. We also saw twice (at Le Bouchage-le Mollard during Bronze Age and at Les Avenières-Pré de la Cour during Middle Age) the burning of reeds formations which also must be linked with pasture management in a humid context.

Oak forest maturation

In these humid lands, it seems that mature forest has difficulty establishing. Oak forest appears at Saint-Romain de Jalionas-le Vernai and at Bourgoin-Jallieu-Les Vers, and to a lesser extent at Les Avenières-Pré de la Cour. At Saint-Romain de Jalionas-le Vernai, it corresponds first to a fluvial context (before 1418–1132 bc) in which the wood macroremains may have been brought from the slopes by alluvial phenomena, and later to a period of intense exploitation of the landscape (Roman period) linked with the presence of an important villa. The present-day vegetation (Villaret, 1999) shows that the oak-hornbeam forest recorded in the latter case may have developed inside the marshy area. Nevertheless, it is more likely, in a context of intense human occupation, that the intense exploitation of the landscape reached the fringe of the marsh, and thus resulted in the deforestation of drier and more mature forests. At Les Avenières-Pré de la Cour, long-distance transportation of wood and charcoal by the river floods (Rhone river or Guiers river) is obvious when fir/beech formations are recorded (2nd century bc/2nd century ad). The record is more local to microregional when oak open formations appear (5th century ad/6th century ad), but in both cases alder largely dominates waterlogged wood assemblages, showing that it remains the main tree which grows in situ. At Bourgoin-Jallieu-Les Vers, the forest is less mature, with a lot of elm, and it appears in the available record at a relatively recent period (not before the end of the Middle Ages, but data are lacking for most of the Medieval period because of a low sedimentation rate in all sectors, and a long period of vegetal reconquest without any fire). That forest may be the one described in 13th century texts and its settlement to the detriment of alder brushwood could be due to a change in the management of the area, including a regulation of pastoral practices that allows forest species to grow (Comte, 1984).

Past presence of beech and fir

The occurrence of beech in charcoal or wood assemblages is not a sufficient argument to prove inputs from distant areas. It is recurrent in the reconstructed past vegetations, and even though its current repartition in the region mainly concerns the pre-Alps heights, it does exist nowadays here and there in the lower Dauphiné plains. The presence of a large uprooted beech trunk at Frontonas-Les Sétives (1115–808 bc) is evidence that it did grow in the marsh area in the past, as previously suggested by Wywial and Bravard (1997). It was probably included in the most mature facies of ash and oak formations. Thus, its presence in charcoal assemblages with local significance could be linked to the burning of old and/or better-drained forest stands (such as at Saint-Romain de Jalionas-le Vernai, supra) and does not bear any climatic significance. A regression of beech forests in the lower plains linked to the disappearance of primary forests has been evidenced in the middle Rhone valley (Delhon and Thiébault, 2005).

Fir has been identified in three different contexts, always during the Roman period (Saint-Romain de Jalionas-le Vernai: in the marsh and in the latrine, les Avenières-Pré de la Cour). It has been shown that past expansion of fir in Europe was much larger than its present days repartition. Evidence for its growth at low altitude not only in the lower Dauphine (Clerc, 1988) but also in other southern Alps lowlands (Tinner et al., 1999) and even on the Mediterranean littoral during the first half of the Holocene rose from palynological records from the border of the southern Alps in northern Italy (Wick and Möhl, 2006), and it seems that the species reached its current repartition after a mid-Holocene decline triggered by anthropic factors (in particular fire) and climatic change (reduced moisture availability) (Tinner and Lotter, 2006). Nevertheless, it probably never grew in the marsh area: in all three cases, it may have been brought from higher altitudes, probably by humans, perhaps also by rivers.

Conclusion

Since the end of the Neolithic, landscape management by human societies of the lower Dauphiné resulted in a shift from a forest vegetation cover, probably more or less continuous and showing wet/open facies dominated by ash and alder and drier/closed facies dominated by oak, to a discontinuous vegetation, made of alder thickets in the humid pastured lands and of remnants of oak forests in marginal areas. The only example of forest reconquest recorded by our data only reaches the stage of a pioneer formation dominated by elm, at Bourgoin-Jallieu-Les Vers, in the second half of the Middle Ages (the forest was well developed around the 15th century ad according to radiocarbon date).

From a methodological point of view, charcoal and wood assemblages from geoarchaeological contexts are not easy to interpret as data are often disparate in terms of quantity, quality, chronological accuracy, etc., but they are an efficient way to document in situ landscape uses and evolution, especially when compared with charcoal data from the archaeological sites and with geoarchaeological and archaeological data at a microregional scale.

Footnotes

Acknowledgements

We thank the Centre pour le radiocarbon of Lyon 1 University and Laboratoire de Mesure du Carbone 14, UMS 2572, ARTEMIS in Saclay for ¹⁴C measurements by SMA in the frame of the National Service to CEA, CNRS, IRD, IRSN and Ministère de la Culture et de la Communication. Many thanks to Ch Oberlin (Centre de datation par le radiocarbone, Lyon) and Ch Moreau (Laboratoire de mesure du carbone 14, Saclay). The authors are greatful to the Service Régional d’Archéologie Rhône-Alpes (A Lebot-Helly) and the Conseil Régional de l’Isère (A Cayol-Gerin and A Clavier) for making this research possible.

Funding

This study was carried out in the framework of the program Peuplement et milieu en Bas Dauphiné (Isle Crémieu) de l’apparition de l’agriculture à l’époque moderne directed by J-F Berger and funded by the French culture ministry and of the PYGMALION program: PaléohYdroloGy and huMAn–cLimate-envIronment interactiONs in the Alps directed by F Arnaud and funded by the Agence Nationale de la Recherche. Field operations were partly carried out under the framework of the PhDs of N Bernigaud (2012) and , and partly in the framework of rescue excavations carried by the INRAP, under the supervision of S Bleu, O Franc and S Saintot. The excavation of the ‘Vernai’ villa is directed by R Royet (Service Regional d’Archéologie Rhône-Alpes).

References

Argant

Heinz

Brochier

J-L

(1991) Pollens, charbons de bois et sédiments: l’action humaine et la végétation, le cas de la grotte d’Antonnaire (Montmaur-en-Diois, Drôme). Revue d’Archéométrie 15: 29–40.

Arnaud

Révillon

Debret

. (2012) Lake Bourget regional erosion patterns reconstruction reveals Holocene NW European Alps soil evolution and paleohydrology. Quaternary Science Reviews 51: 81–92.

Berger

J-F

(2001) Les fossés bordiers historiques et l’histoire agraire rhodanienne. Etudes Rurales 153/154: 59–90.

Berger

J-F

(2003) Les étapes de la morphogenèse holocène dans le sud de la France. In: Van der Leeuw

Favory

Fiches

J-L

(dir.) Archéologie et systèmes socio-environnementaux. Études multiscalaires sur la vallée du Rhône dans le programme Archaeomedes. Collection CRA-Monographies, 27, Paris: CNRS Éditions, pp. 87–167.

Berger

J-F

Bravard

J-P

(2012) Le développement économique romain face aux crises environnementales. Le cas de la Gaule Narbonnaise. In: Berger

J-F

(ed.) Colloque international: Des climats et des Hommes: Glaciologie, climatologie, archéologie, histoire. Editions la Découverte, pp. 269–289.

Berger

J-F

Salvador

P-G

(2010) Les principaux acquis de l’étude paléoenvironnementale sur l’évolution holocène de la plaine alluviale des Basses Terres. In: Berger

J-F

(dir.) PCR « Peuplement et milieu en Bas-Dauphiné (Isle Crémieu), de l’apparition de l’agriculture à l’époque moderne ». pp. 507–530.

Berger

J-F

Thiébault

(2002) The study and significance of charcoal as an indicator of ancient fire: An application to the middle Rhone valley (France). In: Thiébault

(ed.) Charcoal Analysis, Methodological Approaches, Palaeoecological Results and Wood-uses. British Archaeological Report, International series, pp. 25–41.

Berger

J-F

Royet

Argant

. (2003) Une villa gallo-romaine en milieu humide: Le Vernai à Saint Romain-de-Jalionas (Isère). In: Favory

Vignot

(dir.) Actualités de la recherche en histoire et archéologie agraires, Actes du colloque international AGER V de Besançon, septembre 2000. Besançon: Presses Universitaires de Franche-Comté (Série Environnement, sociétés et archéologie), pp. 157–172.

Berger

J-F

Salvador

P-G

Franc

. (2008) La chronologie fluviale postglaciaire du haut bassin rhodanien. Collection EDYTEM, Cahiers de Paléoenvironnement 6: 117–144.

10.

Bernigaud

(2012) Les anthroposystèmes des marais de Bourgoin-La Verpillière (Isère) du Néolithique final à l’Antiquité tardive (3000 av. J.-C. –600 ap. J.-C.), Archéologie du paysage et de l’environnement. PhD thesis, Université de Nice Sophia-Antipolis.

11.

Bleu

(2007) Rapport INRAP de l’opération de Bourgoin-Jallieu « Zac de la Maladière ». SRA Rhône-Alpes.

12.

Bocquet

(1967) Le problème de l’occupation palafittique au nord du Bas-Dauphiné du Néolithique final au Bronze moyen. Bulletin S.P.F. Etudes et Travaux 64(2): 501–516.

13.

Bravard

J-P

(2010) Discontinuities in braided patterns: The River Rhône from Geneva to the Camargue delta before river training. Geomorphology 117(3–4): 219–233.

14.

Chabal

(1994) Apports de l’anthracologie à la connaissance des paysages passés: Performances et limites. Histoire et Mesure IX(3/4): 317–338.

15.

Clerc

(1988) Recherches pollenanalytiques sur la paléocologie tardiglaciaire et holocène du Bas-Dauphiné. PhD Thesis, Université d’Aix-Marseille.

16.

Comte

(1984) Histoire de Bourgoin des Origines à la Révolution. Lyon: Bellier.

17.

Coutterand

Schoeneich

Nicoud

(2009) Le lobe glaciaire lyonnais au maximum würmien glacier du Rhône ou/et glaciers savoyards? Cahiers de Géographie Collection EDYTEM 8: 9–20.

18.

Darinot

Morand

(2001) Management of Wet Meadows in the Lavours Marsh, Implementing Grazing. Tools in Preserving Biodiversity in Memoral and Boreonemoral Biomes of Europe. NACONEX, pp. 86–93.

19.

Delhon

(2005) Anthropisation et paléoclimats du Tardiglaciaire à l’Holocène en moyenne vallée du Rhône: Études pluridisciplinaires des spectres phytolithiques et pédo-anthracologiques de séquences naturelles et de sites archéologiques. PhD Thesis, Université Paris I.

20.

Delhon

Thiébault

(2005) The migration of beech (Fagus sylvatica L.) up the Rhone: The Mediterranean history of a ‘mountain’ species. Journal of Vegetation History and Archaeobotany 14: 119–132.

21.

Delhon

Martin

Argant

. (2008) Shepherds and plants in the Alps: Multi-proxy archaeobotanical analysis of neolithic dung from ‘La Grande Rivoire’ (Isère, France). Journal of Archaeological Science 35(11): 2937–2952.

22.

Delhon

Moutarde

Tengberg

. (2003) Les perceptions et les représentations de l’espace à travers les analyses archéobotaniques. Etudes Rurales 167/168: 285–294.

23.

Dobremez

J-F

Lacassin

Mazars

(1986) Carte de végétation de la France, feuille de Lyon. Institut Géographique National.

24.

Doyen

Vannière

Berger

J-F

. (2013) Land-use changes and environmental dynamics in the upper Rhone valley since Neolithic times inferred from sediments in Lac Moras. The Holocene DOI: 10.1177/0959683612475142.

25.

Forat

(1954) Aux confins septentrionaux du Bas-Dauphiné: Les Basses Terres. Etude morphologique. Revue de Géographie Alpine 4: 675–712.

26.

Frisia

Borsato

Spotl

. (2005) Climate variability in the SE Alps of Italy over the past 17,000 years reconstructed from a stalagmite record. Boreas 34: 445–455.

27.

Gaucher

(2011) L’occupation du sol et l’évolution du paysage du Haut-Rhône (Ain, Isère) depuis le Néolithique jusqu’aux Temps Modernes. PhD thesis, Université de Nice Sophia-Antipolis.

28.

Holzhauser

Magny

Zumbuhl

(2005) Glacier and lake-level variations in west-central Europe over the last 3500 years. The Holocene 15: 789–801.

29.

Laederich

Thiébault

(2004) L’apport des végétaux par l’homme pour la nourriture du troupeau au Néolithique. In: Boetsch

(ed.) Actes du colloque « Plantes qui guérissent, plantes qui nourrissent». Gap: Librairie des Hautes-Alpes, pp. 29–43.

30.

Lebourgeois

Jabiol

(2002) Enracinements comparés du Chêne sessile, du Chêne pédonculé et du Hêtre. Réflexions sur l’autécologie des essences. Revue Forestière Française 1: 17–42.

31.

Magny

(2004) Holocene climatic variability as reflected by mid-European lake-level fluctuations and its probable impact on prehistoric human settlements. Quaternary International 113: 65–79.

32.

Mandier

Evin

Argant

. (2003) Chronostratigraphie des accumulations würmiennes dans la moyenne vallée du Rhône: l’apport des dates radiocarbones. Quaternaire 14: 113–127.

33.

Ozenda

(1985) La végétation de la chaîne alpine dans l’espace montagnard européen. Paris: Masson.

34.

Ploton

(2001) Le Vernai à Saint-Romain de Jalionas: Un site archéologique et son terroir: Approche par l’anthracologie de relations entre l’habitat et son environnement. D.E.A. Environnement et Archéologie, Université Paris X.

35.

Rameau

J-C

Mansion

Dumé

(1989) Flore forestière française, guide écologique illustré. Tome 1: Plaines et collines. Paris: Institut pour le développement forestier.

36.

Richard

Gauthier

(2007) Bilan des données polliniques concernant l’âge du Bronze dans le Jura et le Nord des Alpes. In: Richard

Magny

Mordant

(eds) Environnement et Cultures à l’âge du Bronze en Europe occidentale. Documents préhistoriques 21, Editions du CTHS, pp. 71–87.

37.

Royet

Berger

J-F

Bernigaud

. (2004) La gestion d’un milieu humide: le site du Vernai et le marais du Grand-Plan à Saint-Romain-de-Jalionas (Isère), de La Tène au haut Moyen Age. Fleuves et marais, une histoire au croisement de la nature et de la culture. Paris: CTHS, pp. 253–281.

38.

Royet

Berger

J-F

Laroche

. (2006) Les mutations d’un domaine de La Tène au haut Moyen Âge: Le Vernai à Saint-Romain-de-Jalionas (Isère), Gallia 63: 283–325.

39.

Salvador

P-G

Berger

J-F

Gauthier

. (2004) Holocene fluctuations of the Rhone river in the alluvial plain of the Basses-Terres (Isère, Ain, France). Quaternaire 15(1–2): 177–186.

40.

Schroedter

Dreibrodt

Hofmann

. (2012) Interdisciplinary interpretation of challenging archives: Charcoal assemblages in Drina Valley alluvial and colluvial sediments (Jagnilo, Bosnia and Herzegovina). Quaternary International doi: 10.1016/j.quaint.2112.02.030.

41.

Schweingruber

(1978) Mikroskopische Holzanatomie / Anatomie microscopique du bois. Birmensdorf: Swiss Federal Institute of Forestry Research.

42.

Schweingruber

(1990) Anatomy of European Woods. Bern/Stuttgart: P. Haupt.

43.

Service de la carte de végétation du laboratoire de botanique et de biologievégétale de l’université de Grenoble (1978) Carte de végétation de la France, feuille de Grenoble. Institut Géographique National.

44.

Thiébault

(2005) L’apport du fourrage d’arbre dans l’élevage depuis le Néolithique. Anthropozoologica 40(1): 95–108.

45.

Tinner

Lotter

(2006) Holocene expansions of Fagus silvatica and Abies alba in Central Europe: Where are we after eight decades of debate? Quaternary Science Reviews 25: 526–549.

46.

Tinner

Hubschmid

Wehrli

. (1999) Long-term forest fire ecology and dynamics in southern Switzerland. Journal of Ecology 87: 273–289.

47.

Villaret

J-C

(1999) Inventaire botaniques des marais de Serves et du Grand Plan (communes de Saint-Romain-de-Jalionas et de Leyrieu). Conservatoire botanique national alpin.

48.

Wick

Möhl

(2006) The mid-Holocene extinction of silver fir (Abies alba) in the Southern Alps: A consequence of forest fires? Palaeobotanical records and forest simulations. Vegetation History and Archaeobotany 15: 435–444.

49.

Wywial

Bravard

J-P

(1997) Eléments pour une écologie historique du hêtre dans le bas Dauphiné et la Dombes (France). In: Burnouf

Bravard

J-P

Chouquer

(eds) La dynamique des paysages protohistoriques, antiques, médiévaux et modernes. Actes des XVIIe rencontres internationales d’archéologie et d’histoire d’Antibes. APDCA, Sophia Antipolis, pp. 475–491.