The Fluvial History of the Tavy – a Medieval Overview

Rivers feel like such a natural part of our environment. In a townscape, where everything is tarmac, concrete and stone, we may tarry by an urban river; its flow and riparian banks giving pleasure and a contemplative nature-space in the midst of the built environment. For big-city rivers like the Thames, we probably understand that bank-side development has constrained the channel and that what is observed is sub-natural. But, for a river like the Tavy, in my hometown of Tavistock, it is hard to imagine the river as a man-made construct, altered over time by how we have engineered the banks and used the catchment.

If we could time-shift, and re-visit the town through the centuries, I am sure most of us would expect to see buildings and streets coming and going, and architecture altering, but how many of us would imagine the river changing?  What the Tavy was like in the Bronze Age differs to the Saxon river-scape, as does the medieval river to the Victorian one.  It is the purpose of this blog – the second on the theme of river transformation – to suggest that our rivers are not as natural as we like to think, and are more altered than we may appreciate. This is important because we can’t fully understand the historical development of our riverside settlements, without understanding the synergy between rivers and place. An environmental history of rivers also helps inform how we might judge future river and floodplain use; not to mention future risk.

“Medieval conditions have been physically obscured and overlain by later developments, and perhaps conceptually because of a limited awareness of the variability and dynamics of the physical environment.”  

Lewin 2010

With an academic background in Physical Geography I knew a bit about the geomorphology of rivers, and had an awareness that they have changed over time, but I had never explored the topic in depth, nor used it to try and interpret my home environment. In my first blog [A Fluvial History of the Tavy from the Late Glacial to the Medieval] I used academic literature on British rivers to give a feel for how the Tavy may have developed after the last ice-age up to the medieval. This blog picks up the story from this point, because it is from the medieval (Late-Saxon onwards) that significant human modifications to rivers began in earnest. While I am draping the cloth of this story on the Tavy to help me understand my local environment, much of what I say is generalisable to other British rivers. Whatever your local river, I hope you find some information that helps you re-imagine your river-scapes, and prompts you to think about the fluvio-history of your favourite places.

But first a recap ….

In my first blog on this subject, I wrote about how the Tavy might have developed after the ice-age and through the many subsequent millennia. I suggested (frequently citing the work of the eminent river geomorphologist John Lewin) that, after the last glacial, the river system of the Tavy would likely have been unstable and braided (that means it has multiple channels), choked with the coarse sediments created in glacial and periglacial conditions.

Forests spread across the landscape as temperatures warmed, and so catchment soils developed and were bound by the stabilising effects of vegetation. Catchment erosion was therefore limited. Human activity gradually deforested the land, particularly from the Bronze Age, and pastoral farming spread. This would have increased flooding, because there was less vegetation to impede rainwater as vegetation helps store rain and so is useful for softening peak flood discharge.

Agriculture would have started to cause soil erosion. Initially this would have been at a slow rate because pasture-land is still vegetated (unlike ploughed land which is very vulnerable to erosion). Gradually, due to the shallow build-up of fine sediments on the riverbanks and floodplain, a river such as the Tavy, may have begun to transition from a braided river into an anabranching form (perhaps similar to that shown in the image of Glen Feshie above). This term is used to describe a river with more than one channel, but with semi-permanent islands in between channels. It is more stable than a braided river which has multiple, unstable and mobile channels.

“Multichannel or divided channel systems are rarely found in present-day British rivers, although this has not always been the case. … Drainage of wetlands and channelization of lowland rivers has contributed to the contraction and the loss of multiple-channel systems that were once much more prevalent during the mid- to late Holocene prior to human modification.”

Taylor et al , 2000

The Tavy, with its thin floodplain deposits, would have been a much wider and shallower river than the one we see today. With lower riverbanks, it would have been an easier river to ford. Bridges would have therefore been less necessary. In the first millennia AD, bridging the Tavy in Tavistock would have been possible but would have required a longer span than the bridges we see today because of the wider channel. The juvenile floodplain would not have been blanketed in thick, rich and fertile alluvial soils, but a much thinner layer of sediments. As a result, redundant former channels across the early floodplain would have been evident across its surface. The floodplain before the medieval would therefore have been lower in elevation, rockier, with an uneven surface, and more prone to flooding. Such a habitat would not be suitable for settlement, nor particularly useful for farming.

With this early river history fresh in our heads, let us now move on to explore what happened from the medieval onwards. I will use the literature to strip away the current river landscape, to imagine what the Tavy might have been like 1000 years ago or more – what Brown et al (2018) describe as a ‘millennia-long record of human interference [that] has been interwoven into what we might perceive as classic river landscapes’.  I have drawn predominantly on the work of Lewin (2010) and recommend it to anyone looking to understand this theme in more detail.

One of the main reasons why British river systems have changed so much from the medieval onwards is because of the impact of soil erosion, which mobilizes sediments from catchments and deposits them in rivers, floodplains, estuaries and the sea. Early farming, from the Bronze Age onwards, did cause small-scale soil erosion, but throughout Britain, the national picture is one of accelerated floodplain deposition from about years ago. Rates of floodplain sedimentation in the 11^th to 14^th C were ten times higher than in the early Holocene (Macklin, Jones, and Lewin, 2010), with rates dropping off after the Black Death. This was caused by population pressure and agricultural change, specifically the introduction of the mould-board plough, which facilitated deeper ploughing. Farming practices using ridge and furrow also effectively manufactured a microtopography of gullies which helped feed sediments into river systems. Pressure was so great that even on Dartmoor, agriculture expanded onto this marginal land in the 13^th C to extend the area under production. Reduced arable cultivation post-plague, and the rise of land enclosure, both led to some improvement in soil conservation (Lewin, 2010). This would have been more acutely felt in the British Midlands where open field systems had dominated medieval agricultural practice; less so for a Devon river like the Tavy. Adding to the erosivity, it is thought that increased rainfall in these medieval centuries also contributed to washing away more soil (e.g. Brown, 2009).

“evidence suggests that soil erosion has flooded systems with fine sediment, creating tabular floodplains on top of a former, more varied topography of channel bars, levees, cutoffs, and wetlands”  

Lewin, 2010

The national picture hides more nuanced regional and catchment-specific differences, meaning that phases and timings of accelerated sedimentation vary from river to river. Mining for example, is another activity which causes significant soil erosion and is therefore a very important consideration in the depositional history of Dartmoor rivers. Thorndycraft et al (2004), in a paper on alluvial records of tin mining, found that for the Erme, there was accelerated sedimentation in the Middle Ages, but also from the 4^th to the 7^th centuries; a phase of tinning which precedes historical records and of which little is known, but which the depositional record of rivers suggests was a very active time of mining. Across Britain, studies have also found increased sedimentation associated with Roman era mining (Macklin, Jones, and Lewin, 2010).

The depth of sediments is highly variable across floodplains (Foulds and Macklin, 2006). It varies laterally from the river, down river, and also from catchment to catchment. However, to give an idea of the quantity of sediments blanketing valley bottoms, studies suggest depth is in the realm of an average of 2-3 m (Lewin, 2010; O’Shea and Lewin, 2020). I did some probing into the available data for the Tavy and was able to find borehole data from the site investigation for the Morrisons supermarket site (BGS Environment Viewer). This showed alluvial clay deposits up to 3.4 m deep. I also know from building work at my own property, which lies at the edge of the floodplain, that my house sits atop several metres of, presumably medieval, river clay river.

So why are all of these sediments getting into the river system so important? The way rivers look and behave isn’t just a function of the water in the river channel but also the interplay of the water with the sediments. Changes to the sedimentary regime have a large impact on the form and functioning of a river system. With such great quantities of sediments getting into British rivers in the medieval,  these greatly altered the rivers themselves, with consequences for how people used their river-scapes.

When the amount of fine sediment getting into a river increases, it has a number of impacts. Aside from the polluting effects of the turbid water, high sediment loads begin to clog up river basins. Sediment-rich water from overbanking flood flows increases the sedimentation rate on floodplains and so these floodplains gradually develop deeper alluvial soils. The once stony valley bottoms, with their thin soils and rough micro-topographic undulations of redundant palaeo-channels and storm channels, become blanketed in sediments. This flattens the topography to give a fertile, nutrient rich and flat valley floor which makes ‘medieval floodplain relief at specific times difficult to assess’ (Lewin, 2010). This sedimentation tends to go mostly un-noticed by us, but we may be conscious of it, evidenced by the half-buried arches of some medieval bridges, and stories of the silting up of former ports like Helston in Cornwall or Chester in Cheshire.

With the build-up of sediments, rivers became narrower and more deeply channelised between vegetated banks (Lewin, 2010, Brown and Keough, 1992). The Tavy would no longer be braided but would be moving towards an anastomosing form, with a main channel punctuated by islands of relatively stable sediment [see the image – above, right – of the River Drau Austria as an exemplar]. We know this is true for the Tavy because, as late as 1741, ‘the island’ is recorded in the middle of the Tavy near the location of the Great Bridge in the Delafontaine engraving of the town [see image below].

Although medieval river channels could still migrate over time, such as the way meander bends gradually alter their positions over time, and catastrophic flood events could cause channel patterns to change overnight, for the most part, channels developed more stability than their pre-medieval incarnations. Also, because the ‘new’ medieval rivers were increasingly confined within clearly defined steep banks, this meant that river water depth increased and channel width narrowed compared to the unconfined, wide, shallow and braided rivers of pre-history.  This altered the way people crossed rivers. The fording of deeper rivers with steeper banks, for example, became more difficult. Conversely, the bridging of rivers became more favourable as a consequence of their narrowed channels and stable banks, and may in itself be causal in the story of the increase in bridge building in the medieval.  

“extra impetus was added to commercial expansion for both bridge building and raised causeway construction. This came partly from a floodplain topography no longer visible to us, and partly from the transformations previously brought about a narrowing of channels and the floodplain deposition of eroded soil material. The positive advantage of this for later generations was the increased potential of flatter floodplains with deep soils for summer hay pasture and watermeadows, particularly if erosion intensity and inundation frequency eased.”

Lewin, 2010

Floodplain alluvium, derived from the eroded soils, was rich in nutrients, and so these flat fertile plains developed into agriculturally valuable water meadows for grazing cattle, considerably extending the area of medieval meadowland onto land not previously suited for agriculture (Lewin, 2010). The fine-grained floodplain alluvium had a high impermeable clay content which promoted water-logging. These rapidly thickening clay-rich floodplain soils, rather than transmitting water rapidly through the floodplain via the palaeo-channel and storm-channel network, developed a propensity for greater flooding due to the retention of the inundating flood water (Lewin, 2010).

What overall effect these changes had on flooding is unclear. On the one hand, medieval rivers, bound by steep banks, may have ‘over-banked’ water less frequently (Lewin, 2010). Wider and shallower pre-medieval rivers, with multiple channels, and scope to spill excess water into palaeo-channels at high flows, may have flooded more often and more easily. However, such a system was probably more effective at dealing with flood flows with flood water being disgorged through valleys rapidly. To this we must also factor in continued vegetation and soil loss from catchments, meaning that water storage capacity in the catchment was reduced. A consequence of this would be that peak flood discharge would have increased and been more rapidly transmitted into river systems. Finally, we must add to this the impact of reduced drainage by the impediment of clay-lined valley floors, meaning that flood waters hung around for a long time. Overall, these changes possibly meant that flood frequency reduced. With the flat fertile floodplains beginning to be more frequently used for settlement, industry and agriculture, it seems fair to suggest that, if flooding overall was less frequent, on the occasions when it did occur, it would probably have been more extreme and problematic in scale, duration and impact.

What does all this mean for the Tavy and Tavistock; the focus of my attention? If we go back to the first records of the foundation of the town c 961 AD then what these early medieval people would have seen of the Tavy would have been a wide river with shallow banks. They would not have recognised the deep single channel of the river today. They would also have seen a different floodplain to us. Today’s floodplain is occupied by the town centre, by housing, roads, industrial and retail units, schools and recreational areas. At the foundation of the town, the floodplain would have been unoccupied by settlement. It would have been lower in relation to the river than it is today and would have had a thinner coating of sediment. Former glacial and post glacial coarse sediments, palaeo-channels and river bars, with watery back-swamps may well have still been evident under a thin drape of sediment. Over time, the townsfolk may have noticed the floodplain thickening and the river altering. Within the town itself, this floodplain accretion would have been further supplemented by the building up of the ground surface by phases of building development and the ongoing waste disposal generated by the townsfolk, channelising the Tavy still further. The effects of this are clear today – compare the height of the river banks within the town compared to the height of the banks in similar unconstrained floodplain stretches in rural locations [see the comparative images below].

As the medieval decades ticked by, the changing form and behaviour of the river and its floodplain meant that new ways of using and working alongside the Tavy became possible, a theme I will explore in my next blog on this fluvial theme.  I could have made my next fluvial post a continuation of this chronological summary of how British rivers have changed, led once more by John Lewin, who brings the geomorphological history closer to the present with his paper on British floodplains in the period of enlightenment (Lewin, 2013). I will almost certainly come back to this at a later date, but for now I want to start to take a more in-depth look at the various ways that people responded to their changing river environment in the medieval. I am going to start with the evidence for the spread of civic and monastic settlements onto floodplains. Using some of the published evidence, I hope to be able to interpret and speculate about what this might have meant for how Tavistock developed. It is my intention in subsequent blogs to then move on to consider other themes – river crossings; the agricultural use of floodplains, the engineering of weirs and watermills, and mining and other contaminants. Through all of these topics I want to use my local environment to think more deeply about the archaeology and history of a specific place, in this case Tavistock, but through a hydrological lens.

“Today, floodplains are densely cultivated, and the modern European countryside is principally derived from the human landscape modification that occurred in medieval times.”

Brandolini and Cremaschi, 2018

To our modern perception, this focus on water may not seem that important. Our lives are not shaped and dictated by water to anywhere near the same degree as people in the past. We do not need to give water too much of our attention. Our drinking water is literally ‘on tap’. We are so disassociated from food production that we do not need to think about how water impacts agriculture. Streams through towns are mostly culverted so we do not see the extent to which water flows through our streets. Our own waste is sanitised by flushing it away, out of sight and mind (and nose!). The most we ever really need to think about water is a decision about what clothes and shoes to put on if it is raining, or if the inclemency will impinge on our recreational plans. Sometimes local flooding causes very real problems, but for most of us, most of the time, this is not an issue.

In the past water was a daily concern, be this drinking, cooking, washing, defecating and other waste removal, food production, power generation, travelling over (on bridges and in boats), and enabling industrial activity. In all of these realms, before modern systems of pipes brought potable water to our taps, and then swished it away in sewers, the provision of water was a perpetual priority and problem. Pre-modern water management dictated and constrained choices about how and where we lived, and required engineering, within the financial and technological constraints of the time. It is therefore my hope that in the next few blogs on these aspects of the nexus of rivers and development that I hope to be able to further develop this point – that to understand our landscapes and townscapes; to understand our history and archaeology – it is important and necessary to appreciate how this happened in relation to our changing rivers, because they have had an especially significant, yet poorly understood and under-appreciated hand, in shaping how our places have come to be the way they are.

References

Brandolini, F. and Cremaschi, M., 2018. The impact of late Holocene flood management on the Central Po plain (Northern Italy). Sustainability, 10(11), p.3968.

Brown, A.G., 2009. Colluvial and alluvial response to land use change in Midland England: An integrated geoarchaeological approach. Geomorphology, 108(1-2), pp.92-106.

Brown, A.G., Keough, M.K., 1992. Palaeochannels and palaeolandsurfaces: the geoarchaeological potential of some Midland (U.K.) floodplains. In: Needham, S., Macklin, M.G. (Eds.), Archaeology Under Alluvium. Oxbow Books, Oxford, pp. 185–196.

Brown, A.G., Lespez, L., Sear, D.A., Macaire, J.J., Houben, P., Klimek, K., Brazier, R.E., Van Oost, K. and Pears, B., 2018. Natural vs anthropogenic streams in Europe: history, ecology and implications for restoration, river-rewilding and riverine ecosystem services. Earth-Science Reviews, 180, pp.185-205.

Foulds, S.A. and Macklin, M.G., 2006. Holocene land-use change and its impact on river basin dynamics in Great Britain and Ireland. Progress in Physical Geography, 30(5), pp.589-604.

Lewin, J., 2010. Medieval environmental impacts and feedbacks: the lowland floodplains of England and Wales. Geoarchaeology, 25(3), pp.267-311.

Lewin, J., 2013. Enlightenment and the GM floodplain. Earth Surface Processes and Landforms, 38(1), pp.17-29.

Macklin, M.G., Jones, A.F. and Lewin, J., 2010. River response to rapid Holocene environmental change: evidence and explanation in British catchments. Quaternary Science Reviews, 29(13-14), pp.1555-1576.

O’Shea, T.E. and Lewin, J., 2020. Urban flooding in Britain: an approach to comparing ancient and contemporary flood exposure. Natural Hazards, 104(1), pp.581-591.

Taylor, M.P., Macklin, M.G. and Hudson-Edwards, K., 2000. River sedimentation and fluvial response to Holocene environmental change in the Yorkshire Ouse Basin, northern England. The Holocene, 10(2), pp.201-212