Introduction

As a major route for exchange and communications, the Loire River has attracted human populations for millennia. This is attested to by the numerous archaeological remains found in the Loire valley (dugout canoes, evidence of harbours, standing stones, dolmens, etc.). The oldest of these show signs of human presence during the Weichselian period, such as the Roc-en-Pail rock shelter at Chalonnes-sur-Loire (Visset 1980) or the Pierre Meslière site (Ancenis/Saint-Géréon), which has been dated to Palaeolithic times. The two main streams of neolithisation in Western Europe (Danubian and Cardial) appear to have met at a later stage along the banks of the Loire (Visset et al. 2002).

Human communities have succeeded each other for millennia along the Loire Valley. At first nomadic, over time they settled, thus gradually helping to shape the landscape as they cleared forests and planted crops.

This paper describes how the regional vegetation evolved at nine sites (thirteen surveys) in the central and downstream areas of the River Loire, and provides a chronology of interaction between human populations and their environment along the river valley (Fig. 1).

Fig. 1
figure 1

Location of the study area

Full size image

Investigations conducted during the last few years in various regions (Marambat 1995; Leroyer 1997; Puertas 1997; Galop 1998; Trément 1999; Barbier 1999; Ouguerram 2002; Cyprien 2002) have demonstrated the importance and fruitfulness of an interdisciplinary approach that promotes an overall understanding of the way environments function. By combining palynology, sedimentology, geography and archaeology in our research, we have gained an overall view of the impact of human societies on the environment (cultivation, deforestation, habitats, herding, ponds, mills, etc.) from the preboreal period to the present.

Other reference studies have been carried out in the wider region of Western France, along the Armorican coast (Morzadec-Kerfourn 1974), the lower Loire (Visset 1979), in the Loire estuary (Voeltzel 1987), the Breton-Vendée marshes (Bernard 1996), the Paris Basin (Leroyer 1997), the northern Mayenne area in Normandy (Barbier 1999) and, recently, in the Erdre river valley (Ouguerram 2002).

Our relatively large study area (about 180 km in length) has proved to be highly varied, since it covers the Loire estuary (at Cordemais), a palaeo-channel (at Rezé), the marshlands adjacent to the Loire (Oudon, Ancenis, Méron, Champtocé) and the wetlands of Carquefou, Changeon and Distré, these latter being more distant from the river’s influence (Fig. 1).

Materials and methods

Sediment samples were collected by using a motor-driven GIK drill (10 cm diameter) (Visset and Hauray 1980, 1988) or a percussion drill. The majority of sediments were mud. Sampling for pollen analysis was performed every 5 cm (sometimes at an interval of 4 or 3 cm). For pollen extraction mud samples were treated with Thoulet’s solution and bog sediments by the classic method for peat. Pollen diagrams were constructed using G3PAL software. For each sample, about 300 to 400 pollen grains were counted. Alnus (local vegetation) and Pinus were excluded from calculation for general pollen diagrams to avoid masking the pollen representation of other species (the calculation sum is also about 300 grains). Portions of the most representative diagrams are selected for each period. In these diagrams pollen groups are formed as follows according to their ecological, biological or anthropic characteristics (Behre 1981):

Cropland plants::

Centaurea cyanus, Papaver

Apophytes::

Rumex, Mercurialis annua

Ruderals::

Plantago, Convolvulus arvensis, Polygonum aviculare, Cuscuta, Brassicaceae, Chenopodiaceae, Caryophyllaceae, Artemisia, Ranunculaceae, Potentilla, Trifolium, Solanaceae, Urticaceae, Lamiaceae, Alchemilla, Valerianella, Fabaceae, Compositae (Asteraceae, Cichorioideae)

Heathlands::

Ulex, Calluna (Erica), Campanulaceae, Asphodelus, Polygala, Gentiana pneumonanthe, Helianthemum

Aquatic plants::

Myriophyllum, Nymphaea, Nuphar, Potamogeton, Lemna, Nymphoïdes, Utricularia, Sagittaria, Hydrocharis, Ranunculus aquatilis, Menyanthes, Callitriche, Hottonia

Paludal plants::

Typha, Sparganium, Iris, Ludwigia, Alismataceae, Menyanthes, Anagallis, Lysimachia, Lythrum, Filipendula, Valeriana, Polygonum persicaria, Rumex hydrolapathum, Hydrocotyle, Equisetum, Apiaceae (Apium), Rubiaceae, Scrophularia aquatica, Sanguisorba minor, Thalictrum, Oenotheraceae (Epilobium)

Bushes::

Hedera, Lonicera, Cornus, Evonymus, Viburnum, Sambucus, Ligustrum, Prunus, Viscum, Rubus, Ilex

Shrubs of humid zones::

Salix, Populus, Rhamnus, Myrica

Only the most significant pollen curves are shown.

Radiocarbon dating was performed at the Laboratory of Isotope Geochemistry in Tucson, Arizona. The results are summarised in Table 1.

Table 1 Radiocarbon dates, calibration after Stuiver and Reimer (1998)
Full size table

Results and discussion

Dynamics of post-glacial vegetation

The preboreal period (10000–8900 B.P.)

The preboreal period is represented only at the Munet site (Fig. 1, site 8). The valley began to fill in during the early Holocene period, a process characterised by the deposition of loamy sediment as far as the Atlantic coast. During the preboreal period alluvial sedimentation of the flood loam type occurred in calm waters. The hillsides were initially covered with Pinus (pine) (Fig. 2) with some Betula (birch) and Juniperus (juniper) that subsequently gave way to Corylus (hazel) (9755 B.P.). Mesothermophilous species such as Ulmus (elm) and Quercus (oak) heralded an improvement in climatic conditions (rise of temperatures). Locally temporarily flooded mud flats were covered with Cyperaceae, as evidenced by a small number of aquatic and marshland species.

Fig. 2
figure 2

The Preboreal and Boreal period, Munet site

Full size image

The Boreal period (8900–7800 B.P.)

Forests with Corylus, Quercus, Ulmus and Tilia (lime) became established, effectively protecting hill slopes, thus probably accounting for the low rate of sedimentation and the fine grain of the minerals washed down to the marshes as far as the Atlantic ocean. These marshes, which were flooded in the early boreal period (presence of aquatic and marsh species such as Typha and Sparganium), subsequently began to dry out (Fig. 2).

The Atlantic period (7800–5700 B.P.)

The hillsides were densely wooded until the end of the Atlantic period, with a mixture of Quercus and Tilia taking over more strongly from Corylus towards the end of the period. Fagus (beech) is in evidence at Oudon (Fig. 1, site 4; Fig. 3a) and at Cordemais (Fig. 1, site 1; Fig. 3b) in the Loire estuary. Quercus ilex (holm oak) appears only at Oudon towards 6650 B.P. (Fig. 3a); the Hâvre valley seems to have offered the right ecological conditions for its establishment, although the species remained infrequent until the early subatlantic period. In the valley of the Erdre, a tributary of the Loire, Ouguerram (2002) points out the station-dependent character of Quercus ilex during the Atlantic period. Planchais (1967) had already referred to evidence of its occurrence during the earlier boreal period. Today, according to many authors, this thermophilous species frequently occurs spontaneously along the Atlantic coast. According to Dupont (2001), its northern limit of spontaneous growth is on the island of Noirmoutier (Vendée, France), and it has almost certainly been introduced further north. The data collected in the Erdre and Loire valleys provide further confirmation that the northern limit of spontaneous growth of holm oak during the Atlantic period reached at least as far as the Erdre valley. Only the marshy areas around Oudon (Fig. 3a) were covered by Alnus (alder), with Vitis (wild vines). The valley bottom was filled by mainly loamy deposits in the Munet marsh and at Oudon, where sedimentation occurred in stagnant waters.

Fig. 3
figure 3

The Atlantic period. Oudon site (a), Cordemais site (b)

Full size image
Two exceptions: Cordemais and Oudon

At Cordemais, not far from the freshwater marshes, a form of herbaceous vegetation (chenopodiaceae and Poaceae) was found which is typical of halophytic environments (Fig. 3b) and bears witness to the formation of a salt marsh where dinoflagellate cysts (such as Spiniferites bentori or Lingulodinium machaerophorum) were discovered. These cysts are good markers of salinity, Morzadec-Kerfourn (1976, 1992) noting that Spiniferites bentori is found in abundance in the maritime domain while Lingulodinium machaerophorum is associated with brackish environments.

At Oudon, about 80 km from the present-day mouth of the Loire, this phenomenon is less evident and more limited in time, with green silt deposited in calm, slow-moving waters. The presence of sub-halophytic vegetation made up of chenopodiaceae, also associated with dinoflagellate cysts (though less numerous than at Cordemais), suggests that a mixture of brackish waters may have penetrated the Hâvre valley from (at least) 6740 B.P. to 5000 B.P. (Fig. 3a). This phenomenon appears to have occurred at the same time as the marine transgression which is known to have taken place from 7000 to 5000 B.P. (Mozadec-Kerfourn 1974; Visset 1979; Bernard 1996).

The Subboreal period (5700–3600 B.P.)

The landscape began to change radically during the Subboreal period. Variations in sea level, transgressive around 5700 B.P., regressive around 4500 B.P. (Visset et al. 1994, 1995, 1996), had repercussions in the Loire valley where the river appears to have been indirectly affected by these oscillations, with evidence of changes in its hydrological regime.

Morzadec-Kerfourn (1974) considers that the Subboreal period along the Armorican coastline was a relatively humid time during which the marshes became established. This was the case in numerous marshlands where the heightening of the water regime is reflected in a clear increase in paludal taxa (Lysimachia, Sparganium, Typha, etc.), magnocaricae and alders (except when the water level was too high). These observations appear to be contemporary with the rise in sea level known to have occurred at about 5000 B.P.

From the beginning of the regressive phase, around 4500 B.P., peat deposits appeared at most sites, despite the observed increase in waterway flow, reflected in the sedimentology by an increase in fine and coarse-grained sand, both in the Douet river (munet marsh) and the Hâvre (at Oudon).

In the sites near the estuary (Cordemais), marine regression caused partial erosion of sediments and, further upstream (at Oudon), a complete withdrawal of brackish waters (disappearance of Chenopodiaceae) (Fig. 4a).

Fig. 4
figure 4

The Subboreal period. Oudon site (a), Munet site (b), Cordemais site (c)

Full size image

In the Munet marsh, a hiatus in sedimentation is apparent before 4500 B.P. This seems to have occurred at the same time as the marine regression phase, whose impact, according to some authors (Gruet 1978; Cassen et al. 1999), is still visible today about 150 km away from the coast, especially in the rate of burial of standing stones and dolmens (e.g. at Champtocé or Ancenis) (Ménanteau 1973). However, is it really likely that the withdrawal of the sea could have had an impact on a place some 180 km away from the present-day shoreline, while peat deposits were building up in other sites further downstream? Could this not rather be explained by local phenomena occurring upstream? At present few examples have been found in the region, although there is evidence in some places along the Loire and Authion valleys (not far from the Munet marshes) of highly contrasting hydrodynamics during the Atlantic and Subboreal periods (Visset et al. 1999). It was also between 4700 and 4300 B.P. that silty coastal ridges began to form near river mouths, permanently halting the penetration of seawater (Oudon) and gradually cutting off inland environments from direct marine influences (Cordemais).

The hillsides remained covered with Quercus that was gradually replaced by Corylus (Fig. 4a) except around Munet (Fig. 4b). Fagus also remained, but Ulmus and Tilia declined throughout the period.

There is evidence of Buxus (box wood) around 5600 B.P. in the Loire estuary at Cordemais (Fig. 4c). Buxus has already been recorded in the western region by Ouguerram (2002) at around 7350 B.P. (about 6200 cal B.C.), by Voeltzel (1987) around 6500 B.P. and by Visset (1979) around 5200 B.P., which may indicate the spontaneous presence of this species in a few ecological niches in the region (Bernard 1996), the Cordemais soils and the schist soils around Ancenis (Fig. 1, site 5) appearing to be favourable for its establishment (either very temporary or very localised?).

From the subboreal period onwards, the environment became subject not only to climatic variations but also to the activity of human societies. Some forest clearing, probably associated with the beginnings of livestock farming, (ruderal species and grasses) and cropping (wheat, rye, buckwheat and flax) appeared in the Loire valley.

The Subatlantic period (3600 B.P. to the present)

The influence of marine movements during the Subatlantic period (starting in Western France at 3600 B.P., Morzadec-Kerfourn 1974) seems to have been restricted to the estuarine zone, i. e. at Cordemais. Here, where the sediments dating back to the late Subboreal period have been eroded, we also find a hiatus in sedimentation relating to the Bronze and Iron Ages. This could have been the consequence of the succession of regressive movements associated with the lowering of the sea level during the Bronze Age, around 3000 B.P. (Morzadec-Kerfourn 1974; Prigent et al. 1983; Shennan 1986; Pirazzoli 1991; Visset et al. 1994; Marambat 1995; Bernard 1996) and the Iron Age (Prigent et al. 1983). As we saw previously the continental future of the marsh was associated with the establishment of the silty coastal ridges during the second half of the Subboreal period. It should be noted that as a result of these variations, the current level of the sea seems to have been reached at some point between the late Gallo-Roman period and A.D. 1000 (Prigent et al. 1983; Visset 1988).

The continuous presence of Fagus pollen (Fig. 5) is a characteristic of the Subatlantic period around 3600 B.P. in the Armorican Massif in France, along with the disappearance of Tilia (Morzadec-Kerfourn 1974) noted in the Loire valley mainly during the Bronze and Iron Ages around 2350 B.P. at Oudon and 2650 B.P. at Ancenis.

Fig. 5
figure 5

Composite diagram from the region. Evolution of the plant landscapes from the Preboreal period in the Loire valley

Full size image

Quercus ilex remained (although very infrequent) in the Méron marshes (Fig. 1, site 6) until the Middle Ages, and still exists today at Ancenis, where the schist soils appear to be favourable to its growth. According to Bernard (1996), holm oak could also be encouraged by human activity. The other areas do not seem to have had the right conditions for the establishment of this particular forest species.

Interactions between human activity and the environment (Fig. 5, LOI.III, IV, V)

The Neolithic period 7100–3900 B.P.

Early Neolithic

The impacts of human activity probably began at the beginning of the early Neolithic period. Only the site at Champtocé (Fig. 1 site 7; Fig. 6a) reveals early human activity, in the decline of forest cover around 6600 B.P. and in the clear progression of grasses and ruderal species. Livestock farming became established, but it is surprising that there is no evidence of crop farming during this time.

Fig. 6
figure 6

The Neolithic period. Champtocé site (a), Ancenis site (b), Méron site (c)

Full size image
Mid and Late Neolithic

Forest clearing on a small scale increased in frequency during the mid to late Neolithic period. In addition to livestock farming, which already existed in the early Neolithic period, human societies began to diversify their activities through the adoption of crop farming (Fig. 5).

Secale type (rye) (pollen size 70 µm, elongated form) is present from about 5700 B.P. in the Ancenis area (Fig. 6b). Rye is not indigenous in the Loire valley (Fournier 1946). Rye was domesticated and brought to Europe together with other cereals from the Near East, but was not used as a cultivated plant until the Late Iron Age. Secale grew in the fields together with cereals as a weed (with a ca. 1000× higher pollen production than other cereals) and was acculturated in Europe much later (unlike Triticum and Hordeum). Behre (1992) notes that the few pollen grains present indicate a wild variety (already domesticated but not acculturated), with rye cultivation appearing during the Gallo-Roman period in Germany, but only beginning to flourish in the Middle Ages.

Van Zeist (1963) has found Secale in West of France during the Subboreal period and Marguerie (1992) notes the presence of this cereal in late Neolithic in a site of Isère.

Fagopyrum (buckwheat), Linum (flax) and Cerealia (cereals) (around 5600 B.P.) are present in the Loire estuary at Cordemais (Fig. 4c). This again supports the findings of various studies that demonstrate the early introduction from the Middle East, especially of Fagopyrum, in Western France. Buckwheat occurred sporadically from the late Mesolithic period (Visset et al. 2002) and through the Neolithic and Bronze Ages (Visset 1979), but did not really flourish until the start of the Middle Ages (Voeltzel 1987; Nassiet 1998; Barbier 1999). Some buckwheat pollen was discovered in a neolithic vase from the Somme (Girard 1973). However its cultivation by Mesolithic people is not confirmed as pollen grains are not always good evidence of cultivation and could also have accumulated from the wild.

These plants are found again towards 5400 B.P. at Champtocé (Carcaud et al. 2000) and towards 4600 B.P. at Ancenis, Oudon and in the Méron marshes. Numerous archaeological remains dating back to the Neolithic period (standing stones and dolmen at Anetz, the Pierre Blanche standing stone at Oudon, la Boire standing stone at Champtocé, etc.) confirm the existence of human populations in the valley. The exception is the Munet marsh where there seems to be no evidence of human occupation (Cyprien et al. 2001a).

Juglans and Castanea pollen appears at Méron (Fig. 6c) towards 4700 B.P. It should be noted that their introduction had usually been dated to the Gallo-Roman period (Visset 1979). However, walnut has been recorded during the Neolithic period in the Erdre valley (Ouguerram and Visset 2001a), in the Fourneau peat bog in Mayenne (Barbier 1999), in the Morbihan area (Visset et al. 1996), where occurrences are noted at 5425 B.P. (4437–4045 B.C.), then during the early Bronze Age at la Trinité-sur-mer (Visset et al. 1995) in the Nantes region (Barbier and Visset 1997) and finally on the Breton peninsula as from the Iron Age (Marguerie 1992).

Renewed development of hazel and oak woods can be seen towards the end of the Neolithic period (Fig. 6b), associated in some sites with the virtual disappearance of cereals, regression of meadow lands and the establishment of heaths in certain spots.

The Bronze Age: 3900–2800 B.P.

Forest cover on hillsides remained relatively dense although a slow decline began around 3400 B.P. (Fig. 7a), except at the Carquefou (Fig. 1, site 3) and Munet sites. This appears to have encouraged the growth in many areas of Fagus and Corylus, with the more open environment encouraging hazel pollination (Bégeot 1998).

Fig. 7
figure 7

The Bronze Age. Méron site (a), Oudon site (b)

Full size image

Pastures and crop fields began to spread. The wild vine pollen occurring in the Atlantic region now forms a continuous curve (in parallel with the presence of cereal and hemp) and may have been cultivated. Ouguerram (2002) notes that Vitis could probably be cultivated during the Bronze Age in the Erdre valley. Cannabis-Humulus (hemp), which is much in evidence on the Oudon site around 3500 B.P. (Fig. 7b), may also be considered to have been cultivated from about this time as in the Erdre valley (Ouguerram 2002), although the crop did not reach its peak until Mediaeval times.

Towards 3100 B.P., forest clearance intensified and the environment gradually opened up.

In many marshlands (Méron, Ancenis, Oudon and Munet), increasing humidity is reflected more or less clearly. Towards 3355 B.P., in the Méron marsh (Fig. 7a), there was considerable development of aquatic plants (such as Myriophyllum alterniflorum, aquatic Ranunculaceae, etc.) along with numerous paludal species, attesting to a sea level rise in sinks that gradually asphyxiated the previously extensive alder woods and wetland shrubs.

Occurring at the peak of a transgressive episode (Morzadec-Kerfourn 1974; Visset et al. 1995), successive variations in the height of the water column in marshlands encouraged alternating deposits of peat and silt.

The Iron Age: 2800–2000 B.P.

There is evidence of the beginnings of the Iron Age in many places, in the continuity of agro-sylvo-pastoral practices that opened up the environment. Livestock farming predominated on the hillsides (tremendous increase in grasses and ruderal species). Crops such as cereals, vines and hemp, probably grown in the clay lowlands and well represented at Carquefou and Oudon (Fig. 8), as well as walnut, spread to plots cleared by large-scale deforestation.

Fig. 8
figure 8

The Iron Age. Carquefou fossé site (a), Oudon site (b)

Full size image

During the late Celtic period, around 2350 B.P. (at Oudon) (Fig. 8b) and towards 2100 B.P. (in the Méron marsh), the entire forest system began to flourish once more, especially Quercus and Corylus woods but also Fraxinus (ash), Fagus and Ulmus (elm) in the Hâvre valley, at Ancenis and in the Méron marsh. This forest renewal was paralleled by a decline in the pasture and crop field system, the sites in the Loire valley seeming to have been abandoned for reasons as yet unknown.

Forest clearance began again at the end of the Iron Age, as well as livestock and crop farming with rye (Fig. 8b) and buckwheat. Sweet chestnut, already in evidence during Neolithic times, also occurred.

Sediment deposits in stagnant waters, mainly on clay beds, became more peaty and sandy towards the end of this period. This change may be related to a renewal of slow-moving water flows (Oudon).

The Gallo-Roman period: 2000 B.P.–A.D. 500

As a result of the marked decline of the forest system (especially of oak woods) that began in the Bronze Age and continued in part during the Iron Age, the environment became more open than ever before (except at Cordemais where the forest cover remained), encouraging in places the growth of hazel woods (Fig. 9a), especially during the early imperial period. In Gaul Quercus was the most widely used timber for boat building, construction (roof timbers and foundations), tools, statuary, etc.

Fig. 9
figure 9

The gallo-Roman period. Vallée du Seil site (a), Méron site (b)

Full size image

At the Méron site (Fig. 9b), where forests had already virtually disappeared by the beginning of the Gallo-Roman period, Quercus woods with Corylus, Fagus and Betula temporarily appeared once more, along with crop diversification. According to Barbier (1999) this development, also observed in the Mayenne area in Western France (Fig. 1), could signify the establishment of a system of forest management (which has continued until the sub-present) after the intensive deforestation that occurred during the Iron Age, rather than being a sign of abandonment.

With the exception of the Munet marsh, where there is little evidence of human activity, the Gallo-Roman period is characterised by further intensification and diversification (already much in evidence during the Iron Age) of crop farming in the Armorican region (Marguerie 1992). Cereal growing, wheat and rye (Fig. 9b), clearly increased, together with apophytes. Hemp was grown in the Loire estuary around A.D. 200, as shown by Voeltzel (1987), with flax, buckwheat, walnut, sweet chestnut and vines all grown in the region. Wine growing became established in the Loire estuary (Visset 1979) near Nantes (Deschamps and Pirault 1999) and around Angers (Provost 1993) (Fig. 1).

Some marsh woodlands were intensively cleared, reflecting efforts to extend hay meadows. Pastures became larger, livestock farming intensified and became predominant. According to Marquet and Poulain (in Provost 1993), pigs were the most common domestic animals in the Loire valley during the Gallo-Roman period.

Events during the later imperial period caused profound changes in agro-pastoral practices (Fig. 9a) that were reflected in a severe decline in crop and livestock farming and pastures. Abandoned fields reverted to woodland (especially oak). A similar situation has been shown to have developed in several Armorican peat bog areas (Marguerie 1992; Barbier 1999; Cyprien and Visset 2001) and in the Paris Basin (Leroyer 1997) and this could be the reflection of an economic and demographic decline resulting from the troubles that occurred in France in the late 3rd century A.D. (Giot 1977). Many settlements were abandoned, although in some sites that escaped the recession, Munet marshes and Cordemais for example, human activity continued to flourish.

Locally, variations in water levels resulted in changes in marshland vegetation and in deposits of mainly clay sediment.

The Middle Ages: A.D. 500–1500

The Early Middle Ages

Forest clearing on hillsides began anew (Fig. 10a). Alder woods were also felled, except at the Changeon site (Fig. 1, site 9) and at Carquefou, where felling did not start until the middle of the period.

Fig. 10
figure 10

The Middle Ages. Oudon site (a), Ancenis site (b)

Full size image

The surrounding landscape was essentially the same as during the early Gallo-Roman Empire, though with more cereal growing (except in the Munet marsh). Hemp growing flourished (Fig. 10a) although pastoral activities remained predominant. This confirms that in western France the early Middle Ages were characterised by an increase in crop and livestock farming (Barbier 1999; Ouguerram and Visset 2001a, 2001b).

The central Middle Ages

Except in the Munet marshes, where crop farming developed to a moderate extent, the central Middle Ages were associated everywhere with the spread of agro-sylvo-pastoralism.

These numerous human activities resulted in a partition of local land use with hemp growing reached a peak (975 B.P.) (Fig. 10b) that reflected the activities of the textile industry along the River Loire (Carcaud et al. 2000). The open hillside landscapes prefigured the hedge-and-field systems of today.

Moreover, management of the waters of the numerous marshes began to develop as societies established water mills (Champtocé, Munet) and dams, while fisheries began to proliferate. Temporary flooding of sinks encouraged alternating peat and silt sedimentation during the entire mediaeval period. Other marshes that dried out completely during the summer were used as pasture areas for livestock herds. In the Changeon site, active hydrological phases with sand deposits were followed by very short low-activity phases with deposits of clay and peat sediments.

The late Middle Ages

Human encroachment on landscapes remained significant, woodlands continued to regress, livestock farming intensified along with crop farming (vine, walnut and buckwheat) although wheat seems to have declined in favour of rye. According to Ruas (1990), the Armorican region of Western France is likely to have been a rye growing area since the 10th century, and as we saw earlier, the plant had occurred since Neolithic times in the Loire valley.

It seems that the two fibre plants, flax (Fig. 10b), also noted by Voeltzel (1987) towards A.D. 1000 and hemp, were in competition within the textile industry which had gained considerable importance in this region during the Middle Ages (Marguerie 1992).

The modern period and the present: A.D. 1500 to the present

The major characteristic of these periods is the persistence of open landscapes, although Scots pine was planted along coastlines around 1900.

Livestock farming was (and still is) the main activity along the River Loire, with numerous meadows and pasturelands extending into wetlands during the summer.

At Champtocé, development activities took place before 1850, after the signing of the Drainage Acts. All uses of the marshlands influenced sedimentation patterns.

Even though no cropping practices are recorded in the Loire estuary (Cordemais), they were highly varied everywhere else. Cereals (wheat) and hemp seem to have been abandoned at Oudon and Champtocé, while a hemp textile mill was established at Ancenis (Fig. 11a).

Fig. 11
figure 11

The modern period and the present. Ancenis site (a), Carquefou tranchée site (b)

Full size image

The growing of buckwheat, which continued until the 1950 s at Carquefou, was extended along with legumes (Fabaceae) (Fig. 11b) in fields that were surrounded by hazel and oak hedges (Lebrun 1972).

Liming, introduced during the French Revolution, was a major technical advance that helped to improve poor soils and transform them into arable land. Lime kilns were built on calcareous outcrops.

Maize, which was known as “Italian wheat” or “Turkish wheat”, has been grown in the Loire valley near Chinon (Fig. 1) since the second half of the 18th century (Maillard 1992). It was rarely grown outside this area and only appeared in the western region of France in the 1950s.

Conclusions

This research led to further knowledge of the evolution of the Loire river area around Cordemais, near the river mouth to Saint-Nicolas-de-Bourgueil, since the Preboreal period up to the present and takes into account its bioclimatic and subsequently its anthropic context. Human communities appear to have settled on a persistent basis in the region from the Bronze Age, although crop farming appeared as early as the middle and late Neolithic period (probably rye, buckwheat, flax and cereals) in the downstream reaches of the Loire and subsequently along the entire river. However, cereal growing was not extensive in comparison to livestock farming, which was predominant around all the sites studied. In time, crop farming intensified and became more diversified (hemp, flax, vines and later, maize) until a true system of environmental management became established. “The environment and human societies have thus been involved in a pattern of interaction that has resulted in the construction of a landscape” (Carcaud et al. 2003).

Further research, synoptic reports and comparisons with upstream sites may help to establish the chronology of human impacts on the environment along the whole length of the Loire river. They will aid the understanding of how downstream movements of alluvial material may be linked to the destabilisation of hillsides in the Massif Central. The research is entirely in line with the studies undertaken both in the upstream reaches (Cubizolle et al. 2001; Ballut 2001) and the central part of the Loire (Garcin et al. 2001) “where the research issues are raised in an integrated way from the outset, rather than by disciplinary field” (Burnouf et al. 2001).