® .Geomorphology 39 2001 69­82
www.elsevier.nlrlocatergeomorph
Late Holocene channel changes of the Middle Trent: channel
response to a thousand-year flood record
A.G. Brown a,)
, L. Cooper b
, C.R. Salisbury c
, D.N. Smith d
a
Department of Geography, School of Geography and Archaeology, Uniersity of Exeter, Amory Building, Rennes Drie,
Exeter EX4 4RJ, UK
b
Uniersity of Leicester Archaeological Serices, Leicester LE1 7RH, UK
c
165 Tollerton Rd., Tollerton, Nr Nottingham NG12 4FT, UK
d
Department of Ancient History and Archaeology, Uniersity of Birmingham, Edgbaston, Birmingham B15 2TT, UK
Received 16 December 1999; received in revised form 1 July 2000; accepted 11 November 2000
Abstract
This paper presents recent work on the floodplain sedimentology of the Middle Trent using data from gravel pits,
archaeological sites and documentary sources. The Middle Trent has been unusually active during the Holocene in
comparison with other large lowland rivers in the British Isles. The Holocene floodplain fill is dominated by sands and
gravels with abundant structural evidence of changes in channel pattern and channel type. A thousand-year record of channel
change has been reconstructed from palaeochannels and gravel units with radiocarbon dating of brushwood, palaeomagnetic
dating of fine channel fills, dendrochronological dating of timber structures and dating via archaeological typologies. The
Trent also has a reasonably well-recorded flood history at least since the 11th century AD. A comparison of the flood record
and channel change indicates that the same degree of morphological and sedimentary response is not necessarily associated
with floods of similar magnitudes, i.e. there is no constant relationship between event magnitude and landform change.
Instead, the response seems dependent on the existing state of the channel and medium-term trajectory of channel change.
There is evidence at both the Hemington and Colwick reaches of a cycle of channel change involving a change in channel
typology, and dating evidence that this may have migrated downstream.
The results of this work provide a medium-term perspective on channel change, which may be more appropriate for large
British rivers than short-term monitoring for both model validation and planning purposes. q 2001 Elsevier Science B.V. All
rights reserved.
Keywords: Channel change; Floods; Late Holocene; Medieval bridges; Beetles; Trent
)
Corresponding author. Tel.: q44-1392-263-331; fax: q44-
1392-263-342.
® .E-mail addresses: a.g.brown@exeter.ac.uk A.G. Brown ,
® .SMITHDN@hhs.bham.ac.uk D.N. Smith .
1. Introduction
In contrast to most other lowland rivers in Britain,
the Trent has been particularly active over the last
0169-555Xr01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.
® .PII: S0169-555X 01 00052-6
( )A.G. Brown et al.rGeomorphology 39 2001 69­8270
1000 years. This paper presents evidence of this
activity and attempts to relate it to the flood history
of the Trent basin. Large lowland rivers in the
British Isles are generally characterised by channel
stability over the last 1000 years. The lower Severn
has suffered from channel siltation in the perimarine
® .zone Brown, 1991 , but otherwise there is little
evidence of significant meander migration or avul-
sion over this period. Likewise, the middle and lower
Thames remained largely stable since major changes
®in planform in Prehistory ca. 3500­2000 years BP,
.e.g. Needham, 1992; Allan et al., 1997 . This is not
®the case with upland rivers Howard and Macklin,
.1999; Merrett and Macklin, 1999 , or many smaller
®piedmont and lowland rivers Hooke, 1977; Hooke et
. ® .al., 1990 . Passmore et al. 1993 have shown how
aggradation and transformation to a braided pattern
occurred in reaches of the Upper Severn and the
Tyne both in the late Roman period and the 13th and
14th centuries. The apparent Late Holocene stability
® .Fig. 1. Site location map inset and a map of palaeochannels in the Hemington area.
( )A.G. Brown et al.rGeomorphology 39 2001 69­82 71
of lowland rivers is probably due to a combination of
® .low stream power Ferguson, 1981 , the creation of
highly cohesive banks by overbank siltation, and the
management of watercourses through embanking,
bank protection and similar measures. The Middle
Trent, however, is anomalous as despite bisecting the
® .English Midlands Fig. 1 and having a moderate to
® y3 .low slope average 0.5=10 , it has undergone
major channel changes during the Late Holocene.
The excavation of a series of medieval bridges, fish
weirs and other structures at Hemington and earlier
archaeological work at Colwick has provided a wealth
of information on channel changes of the Trent,
® .which extend back over 1000 years Salisbury, 1992 .
This allows assessment of the response of the Trent
to the combination of changing flood regime and
intrinsic variables, which control the reaction and
®relaxation times of rivers Brunsden and Thornes,
.1979; Bull, 1991 .
2. Methods
At all the sites referred to in this paper, the quarry
® .faces exposures produced by aggregate extraction
have been observed as part of archaeological investi-
gations. Faces have been monitored since 1987 as
® .part of planning consent mostly by C. Salisbury
and archaeological remains and associated natural
features are recorded and surveyed using an EDM
Total Station and section drawing. As part of the
excavation of the buried Medieval bridges at Hem-
®ington Cooper and Ripper, 1994; Cooper et al.,
.1994; Cooper, 1998, 1999 , the sediments were drawn
and recorded using standard archaeological practise
® .i.e. recording by context by the Leicester Archaeo-
logical Unit. The archaeological recording has also
covered natural features including channels; around
40 sections and 80 EDM files over the last 2 years.
This archaeological stratigraphy has been translated
into an alluvial stratigraphy by applying standard
® .sedimentological rules and procedures Miall, 1978
largely involving the recognition of erosional bound-
® .ing surfaces EBS and distinctions between clast-
®and matrix-supported gravels Brown and Salisbury,
.in press . Multi-method dating has included den-
drochronology, radiocarbon assay, palaeomagnetic
® .dating Ellis and Brown, 1998 and archaeological
® .dating artefact typology . An approximate estimate
of the channel-forming discharge associated with the
Medieval channels used both the slope­area method
and HECII step backwater modelling. More detail of
®this is given elsewhere Brown and Salisbury, in
.press . Coleoptera from organic layers within the
gravels and associated with Bridge I were processed
from bulk samples using the standard method of
® .paraffin flotation as outlined in Kenward et al. 1980 .
The Coleoptera were identified using a range of
entomological keys and by direct comparison to the
Gorham Collection of British Coleoptera at the De-
partment of Ancient History and Archaeology, Uni-
versity of Birmingham. The taxonomy follows Lucht
® .1987 .
3. The record of channel change from sedimen-
tary and archaeological data
The paper uses data from two sites within the
Burton-on-Trent to Newark section of the valley
® .Hemington and Colwick, Fig. 1 , a section in which
the main channel is joined by two tributaries drain-
®ing the northern upland half of the catchment the
.Dove and the Derwent and one draining the south-
® .ern lowland catchment zone the Soar . Both sites
are the result of aggregate quarrying but stratigraphic
investigations have been undertaken in support of
archaeological investigations; at Colwick in the 1970s
and at Hemington in the 1990s. The Hemington and
Colwick reaches have a large number of still visible
palaeochannels dating from most periods of the
® .Holocene Fig. 1 . These are incised into the lower
terrace, which is now known to be of Younger Dryas
®age on the basis of recent radiocarbon dating Brown
.and Salisbury, in press . Extensive and systematic
quarrying of aggregate in the Hemington reach has
exposed a number of archaeological features over the
last 10 years and from the Colwick­Holme Pierre-
point reach archaeological finds have included a
®dug-out canoes and fish weirs Cummins and Run-
.dle, 1969; Losco-Bradley and Salisbury, 1988 .
The sedimentology of both Colwick and Heming-
ton is both subtle and complicated as all the Holocene
gravels are reworked Late Devensian gravels with
®only a minor reduction in mean grain size Deven-
.sian d 13 mm, Medieval d 8 mm . The Deven-50 50
( )A.G. Brown et al.rGeomorphology 39 2001 69­8272
®Fig. 2. Selected stratigraphic sections around Bridge I recorded by Cooper and Ripper and generalised into bounding surfaces sediment
.discontinuities .
( )A.G. Brown et al.rGeomorphology 39 2001 69­82 73
sian gravels contain many cryoturbation features,
especially ice-wedge casts, while acknowledging the
problems associated with the identification of such
® .features Worsley, 1996 , the form and linearity
hence polygonal planform of these features suggest
they did originate from freezing of the gravels under
mean annual temperatures below 08C. Large intra
formational masses of compact clay have also re-
vealed, under SEM analysis, a lenticular microfabric
® .typical of freezing Van Vleit-Lanoe, 1986 . The¨
distinguishing criteria of the Holocene reworked
gravels are lack of cryogenic features and colour, as
the Devensian gravels have a red hue from the
hematite-rich matrix. The presence of archaeological
remains also distinguishes the Holocene from the
Devensian gravels. At Hemington, over 50 sections
surrounding the Medieval bridges have been recorded
® .by Cooper and Ripper in press . Fifteen of these
have been redrawn and sedimentologically inter-
preted, a selection of which is presented in Fig. 2 in
order to show the scour and fill nature of the stratig-
raphy. This suggests that there are at least three
separate flood units recorded at the site of Bridge I.
There is a general decrease in the complexity of the
® .channel fill from the oldest Bridge I to the youngest
® .bridge Bridge III as evidenced by the stratigraphy;
however, this is probably also due to the changing
cross-sectional shape of the channel that widens over
time.
By combining the archaeological, sedimentologi-
cal and dating evidence, a tentative sequence of
channel changes over the last 1500 years can be
postulated, based upon both the archaeological and
® .geomorphological evidence Fig. 3 . The channel
pattern is reconstructed from the bridge locations and
dimensions, the sedimentology associated with the
Fig. 3. Hypothetical model of channel changes at Hemington from archaeological, sedimentological and dating evidence.
( )A.G. Brown et al.rGeomorphology 39 2001 69­8274
bridges, the pattern of palaeochannels, derived from
the monitoring and plotting of artefacts such as dated
fishweirs and anchor stones. For an unknown period
prior to the 6th century BC, the Trent had a single
® . ® .Fig. 4. Major Late Holocene channels of the river Trent at Colwick a and channel change 1609­1840 b . Both adapted from Salisbury et
® .al. 1984 .
( )A.G. Brown et al.rGeomorphology 39 2001 69­82 75
sinuous channel forming a large meander upstream
of Sawley and across Hemington Fields. Evidence
for a single meandering channel comes from several
reaches including the dated meander at Hemington
®which is of similar proportions to the Sawley
.palaeomeander, see Fig. 1 , and downstream at Col-
® .wick Salisbury et al., 1984 and Holme Pierrepoint
® .Cummins and Rundle, 1969 . By the 10th century,
sinuosity had decreased and there is some evidence
of two channels. The primary evidence of this are
fish weirs discovered in a western extension of the
® .aggregate workings FW I and FW II , at least one of
®which dates to the late 10th­mid 12th century FW
. ®I , and a Norman Mill discovered in 1985 Clay and
.Salisbury, 1990 . Together, these indicate a WNW­
ESE flow over the mill dam, which considering the
®evidence of a functioning channel at Bridge I but
one too small to accommodate the full Trent dis-
.charge implies a bifurcation into eastern and west-
ern channels at or upstream of the Mill Dam. The
Mill Dam was destroyed in a flood around 1140 AD
® .Clay and Salisbury, 1990 . Between the 11th and
15th centuries flow in the easterly channel increases
and is presumably balanced by a reduction in the
discharge of the westerly channel. The evidence for
this is primarily from the increased length and ro-
bustness of bridges built across the easterly channel
and the results of slope­area and HECII step-back-
water modelling, which suggest an increase from a
bankful discharge of 50­140 m3
sy1
for Bridge I to
somewhere between 220 and 350 m3
sy1
for Bridge
III. The large ranges of the estimates reflect the
many sources of error inherent in such calculations.
One possible source of error arises because the input
channel dimensions are based upon the structures
®rather than the channel edges except in the case of
.the S bank of Bridge III and therefore include an
estimate for the maximum width of the outer casson
or pier to the channel bank. However, the present
bankful discharge of the Trent is around 200 m3
sy1
suggesting that the majority or all the flow of the
Trent had shifted into the channel by the estimated
®date of Bridge III construction in ca. AD 1239 last
.felling date from bridge pile . Consistent paleomag-
netic dates from a section adjacent to Bridge III
show that fine sediments were filling the channel
®from 1375"50 to 1400"50 Ellis and Brown,
.1998, Section 10 onwards, implying that the main
channel had avulsed back into the westerly channel
by this date. However, it seems likely that the east-
ern channel remained open, taking flood flows be-
tween the 15th and 16th centuries, eventually leaving
only an oxbow lake named the Old Trent along the
line of the old county boundary. These dates provide
a minimum date for the destruction of Bridge III by
floods.
® .At Colwick studies by Salisbury et al. 1984
revealed a similar cycle of channel change during
® .this period Fig. 4 with a well-established southern
Roman course, a northerly Saxon course with two
® . ®meanders 9th century and a divided section at
. ®Colwick Hall as recorded in 1270 Salisbury, 1984;
.Salisbury et al., 1984 . There is evidence that a
phase of more rapid channel change occurred be-
tween 1300 and 1416 with limited braiding, avulsion
and channel abandonment. In the late 14th century,
considerable channel engineering was undertaken by
® .Richard Byron Lord of Colwick and weirs were
removed, which had obstructed the channels. By the
® .16th century map of 1602; Salisbury, 1983 , avul-
sion to the present meandering planform had oc-
curred. Indeed, in the Colwick reach since 1609
channel change has been limited to abandonment of
the Hesgang meander loop and limited meander ex-
pansion.
4. Coleoptera and channel conditions
Many issues concerning the response of the River
Trent to the flood record and channel change are
evident in the insect faunas recovered during archae-
ological investigations. This is particularly the case
with those associated with the bridges at Hemington.
The insect faunas are from organic sandy silts de-
®posited below a hurdle panel made of interwoven
.woody stems and under dislodged timbers, associ-
®ated with the second caisson bridge pier of box-type
. ® .construction of the 11th Century bridge Bridge I .
The close association between the bridge structure
and the insect faunas enables direct reconstruction of
not only the landscape immediately surrounding the
bridge, but also the nature of fluvial conditions,
riversides and channel bottom at this time.
The continued existence of reworked gravel and
sands in the bed of the channel is one of the key
( )A.G. Brown et al.rGeomorphology 39 2001 69­8276
features that allowed the Trent at Hemington to
maintain a multi-channel form through the accumula-
tion of in-channel bars. The presence of these condi-
tions in the past is clearly indicated by the insect
fauna recovered. Many of the species of water beetle
present are today more characteristic of upland
gravel-bed rivers rather than those of the silt laden
lowlands. Stictotatsus duodecempustulatus, Pota-
monetectes depressus, Orectochilus illous and the
Ariffle beetlesB or elmids Elmis aenea, Esolus paral-
lelepipedus, Oulimmnius species, Limnius olck-
mari, Riolus cupreus and R. subiolaceus, are nor-
mally distributed outside of the lowlands in Britain
® .today Holland, 1972; Friday, 1988 . In addition to
requiring oxygenated waters, elmids are limited in
their distribution to those rivers with sands and
®gravels in their channel base Nilsson and Holmen
.1995; Holland 1972 . This may be particularly true
of two of the elmids present in the archaeological
material at Hemington. Stenelmis caniculata and
Macronychus quadrituberculatus are considered to
®be very restricted or rare in Britain today the former
® .has a Red Data Book RDB 3 status and the latter
® .RDB 2 status in Shirt 1987 . The literature suggests
that these two species are often associated with scour
holes and plunge pools in larger water channels, such
® .as in rivers Steffan, 1979 . The subsequent decline
of these species between the 12th century AD and
today probably results from changes in the sediment
input to the Trent or river modification and engineer-
ing, resulting in the formation of cohesive gravel
channel bases and banks.
A number of terrestrial species also suggests that
the sands and gravels extended to the bank sides. In
particular, the small ground beetle Benbidion punc-
tatum is usually found on barren areas of stones and
® .gravel beside fast flowing waters Lindroth, 1974 . It
is clear from this insect fauna that though the bridge
may have been undermined as the result of the 1141
® .flood events Table 1 unstable gravel and sand bed
conditions must have existed throughout the entire
period of the use of the bridges. Such channel bed
conditions can be expected to exacerbate the effects
of high discharge events and floods. The implied bed
instability is supported by calculations, which at-
tempted to estimate the maximum scour depth from
the structure of Bridge I. The relationship between
actual scour depth and estimated scour depth only
converges if the assumption is made that the bed was
® .not armoured Brown and Salisbury, in press .
One of the most important implications of the
recent sedimentological work at the bridge site is the
implication that the Trent, almost alone amongst
Britain's lowland rivers, experiences a period of
large scale channel change and migration in the
medieval period. This is also largely confirmed in
Table 1
® .Floods that occurred in the 11th­14th centuries on the Trent largely from Potter, 1964 with the floods most likely for the destruction of the
bridges at Hemington indicated. Potter also notes increase in river freezing and summer droughts in the 13th Century. Pontage is a tax raised
for the repair of bridges
Date Comments
Bridge I ® .1141 2 Feb First recorded flood on the Trent, breached the Spalford Bank. A rain on snow flood similar to 1795,
1946 and 1947
1205 Severe winter, snow and river frozen
1216r17 Severe winter, snow and river frozen
Bridge 2?
1255 Jul A summer flood exacerbated by flood debris
Bridge 2? ® .1279 Caused channel change in the lower Trent Riley et al., 1995
1305­1306 Dec­Jan Severe winter freeze ended by 3 days of rain
Bridge 3
1309r1310 Severe winter floods destroyed several bridges and damaged Hethbeth Bridge, Nottingham
® .pontage issued for its repair and may have caused channel change at Shardlow and Wilne Pasture
® .Lang Holme, Courtney forthcoming
® .1315r16 Rainfall floods pontage again issued for Hethbeth bridge
1322 Severe winter floods
1310­1330 A period of severe winters with damage to bridges almost every winter
1403 A severe flood, which breached the Spalford Bank and caused channel change at Sawley
® .and Lockington Sand Holme, Courtney forthcoming
( )A.G. Brown et al.rGeomorphology 39 2001 69­82 77
the 11th century fauna of insects recovered at Hem-
ington. The paleoentomological record of several of
Britain's rivers, including the Trent, contains sub-
stantial numbers of elmid species in Neolithic and
®Bronze Age deposits Dinnin 1997; Howard et al.,
1999; Osborne, 1988; Robinson, 1991, 1993; Smith,
.in press but, it is suggested, few after this time
® .Osborne, 1988 . As suggested above, it is thought
the decline of elmids in riverine insect faunas is
linked to the onset of the deposition of fine grain
® .AalluvialB sediment Osborne 1988 . The initiation of
sedimentological change is normally thought to be
®either late Prehistoric or Iron age in date Robinson
.and Lambrick, 1984 , and so it is assumed that this is
also the date of the disappearance of the elmids in
most lowland rivers. At present, this is a persuasive,
but to some extent a circular, argument due to a lack
of research into the paleoentomology of post Iron
®Age channels outside of the Trent catchment Smith,
.in press . In spite of this, it seems fair to assert that
the Trent, both in terms of the sediments it produces
and its insect fauna, is not following the normal
pattern of development we observe in the other
lowland rivers elsewhere in Britain.
5. The flood records of the Trent basin
The Trent basin has an unusually comprehensive,
although incomplete, flood history for the last 900
®years. This has been compiled by Potter 1964 and
.personal communication for the period prior to the
instrumented record. Although the record is essen-
tially qualitative prior to the 18th century, there are
two ways in which relative magnitude can be gauged.
Firstly, only the largest floods breached the Spalford
Bank, which was a flood defense protecting the
Witham valley and the city of Lincoln. From the
gauged record, it is known that it is breached at
discharges over approximately 1000 m3
sy1
. Sec-
ondly, damage to bridges and `pontage' a right to
borrow money for bridge repairs, can indicate the
largest floods. However, the timing of the flood must
be taken into consideration here as a major cause of
bridge damage was late summer floods, which swept
hay-ricks downstream causing bridge obstruction and
damage greater than that which might be expected
for the discharge. The flood record shows several
®large floods during the 12th­13th centuries Table
.1 , which can be partly matched with the bridge
record using the last felling date derived from den-
drochronology and the date of the succeeding bridge.
® .The dendrochronology shows that the first casson
phase of Bridge I was destroyed in c. 111 AD, prior
to the documented flood record, but that the 1141
flood was most likely responsible for the destruction
® .of the second pile phase of Bridge I. The destruc-
tion of Bridge II is more problematic but the last
timber date was 1224 so it could have been de-
stroyed by recorded floods in 1255 or 1279, or
unrecorded floods. The chronology of the third bridge
is more difficult as it was partly of stone construc-
tion, with evidence of maintenance from an inter-pier
®timber with a felling date of 1270­1305 Howard in
.Cooper and Ripper, in press , and so the 1309r10,
1316r16, 1322 and 1402 floods are all possible
® .candidates. However, the last structural bridge tim-
ber recorded in the whole quarry is dated to 1270­
® .1305, and historical Courtney, in press and topo-
graphic work has revealed that the inception of
Wilne Ferry was in 1310r1311. Since the ferry
would have charged and the bridge would not have
charged, as it served the King's Road, this implies
that the bridge had been lost. The date of 1305r6, or
more likely 1309r1310, is also supported by an
associated agreement between land owners, regard-
ing access to cut-off land refereed to as Lang Holme
and now believed to be Shardlow and Wilne pasture.
The 1402 flood caused major avulsion between
Hemington and Colwick at Sand Holme near the
Sawley meander, leading to land exchange between
two manors and agreements to try and stabilise
channel location. It is noticeable that for at least 5 of
the 11 floods or flood periods listed in Table 1
freezing conditions, rain-on-snow or snowmelt was a
major factor, probably resulting from continental
®blocking of westerly depressions e.g. a Scandina-
.vian high pressure system . This is in contrast to
only three similar events recorded out of the 10
®largest recorded floods since the 15th century Table
.2 .
The record for the 17th­19th centuries shows a
®lower frequency of large approaching or over 1000
3 y1.m s floods, with a return interval of 66 years for
such floods between AD 1402 and 1792 as opposed
to 24 years between AD 1100 and 1402. However,
( )A.G. Brown et al.rGeomorphology 39 2001 69­8278
Table 2
The 10 largest floods for which Q is known or has been reliably
estimated on the Middle Trent
Rank Date Synoptic conditions
1 1795 Feb Largely snowmelt with some rain,
3 y1® .blocking conditions 1400 m s
2 1875 Oct Depressional rain
3 1947 Mar Snowmelt with heavy rain, blocking
conditions--high over Scandinavia
4 1952 Nov Depressional rain
5 1960 Dec Depressional rain
6 1946 Feb Snowmelt and rain, blocking conditions
7 1932 May Depressional rain
8 1855 Mar Depressional rain
9 1901 Jan Depressional rain
10 1886 May Depressional rain
this difference may partly reflect less information on
particular floods and so a less complete series. The
largest recorded flood in Britain occurred on the
® .Trent in 1795 Acreman, 1989 . It had an estimated
3 y1 ® .discharge 1416 m s Archer, 1993 and was
typically a rain on snowmelt flood associated with
blocking of westerlies by high pressure over main-
® .land Europe Table 2 . A period of increased flood
magnitudes during the 12th­15th centuries AD is
consistent with data indicating an increased fre-
quency of wet and severe winters at the end of the
®Little Climatic Optimum ca. AD 900­1300 e.g.
Britton, 1937; Hughes and Diaz, 1994; Brown, 1996,
.1998 .
Given the evidence of the flood record, the cause
of the phase of channel change recorded at Heming-
ton would appear to be fundamentally climatically
controlled reflecting the sensitivity of the reach to an
increase in flood magnituderfrequency and bedload
transport rates. High bedload transport rates are sug-
gested by the abundant evidence of in-channel gravel
deposition and little evidence of overbank siltation.
In this respect, the delay of ca. 100­200 years in the
response of the Colwick reach may reflect the time-
lag associated with the movement of the bedload
® .pulse or slug sensu Nicholas et al., 1995 over a
® y1 .distance of 21 km 100­200 m year . The passage
of sediment through a reach typically induces aggra-
dation, which may result in the conversion of a
single-thread channel to a multi-channel system
® .Nicholas et al., 1995 . An alternative hypothesis is
that the intervening input of the Soar altered channel
response. A combination of more accurate dating and
clast provenancing would be required to test these
hypotheses. Fluvial studies over the last decade have
shown that climatic instability at the end of the Little
Climatic Optimum was responsible for increased flu-
® .vial activity Rumsby and Macklin, 1996 and allu-
® .vial fan sedimentation Harvey and Renwick, 1987
in the uplands and pulses of topsoil erosion from
®lowland predominantly arable catchments Borck,
.1989; Foster et al., 2000 .
6. Discussion
The question immediately raised is why did the
floods of the 12th­15th centuries cause a complete
cyclical metamorphosis of the channel at Hemington
and similar changes at Colwick and yet later floods
have had relatively little geomorphological effect.
The first possibility is that the earlier floods were of
far greater magnitude. It is extremely difficult to
estimate the magnitude of single floods in alluvial
reaches with a long flood history. The documentary
record, while supporting the observation of a run of
® .large floods over a relatively short period 300 years ,
does not support the hypothesis that they were of
significantly greater magnitude than the 1795 or
1947 floods. A second possibility is that since the
15th century channel engineering has effectively
constrained the channel. A common form of channel
training on the Trent was the `kidweir'; a line of
posts, brushwood and wattle. At Colwick, a kidweir
has been radiocarbon dated to cal. AD 1458­1640
® .Lord and Salisbury, 1997 . There was also channel
improvements for coal barges during the 16th cen-
tury including lines of posts and piling designed to
increase siltation of the banks and secondary chan-
nels, a process known as warping. Downstream of
Nottingham, there was shoal removal the construc-
tion of rubble walls and the planting of willow trees
® .osiers ; a progressive channelisation of the river.
This may explain the lack of response to the 1795
flood at Colwick but not at Hemington where the
river was far less channelised. An important indirect
human impact has almost certainly been an increase
in suspended load supplied from arable fields. The
agricultural revolution in the 18th century facilitated
the cultivation of the heavier and less fertile soils of
( )A.G. Brown et al.rGeomorphology 39 2001 69­82 79
the North Midlands and it is generally accepted that
extensive land drainage has increased sediment con-
veyance and increased peak discharges from small
® .tributaries Green, 1979; Higgs, 1987 . An increase
in suspended sediment concentrations would have
led to increased overbank sedimentation, helping to
accumulate silt and fix channel banks and to in-
creased silt and clay deposition on the channel bed
on the falling limbs of floods. The cyclic form of
channel change at Hemington also suggests there
may be an autogenic component to channel response
to floods whereby the channel is in transition, and so
sensitive to moderate floods in between more stable
states, the single-thread meandering pattern. On a
Quaternary time scale, the Trent record would reveal
a catastrophic change from a meandering to a multi-
channel state, through avulsion and an increase in
stream power associated with a period of frequent
® .large floods Graff, 1988 , but a more gradual change
back from the multi-channel to the meandering state.
Depending upon bank stability, and height, the me-
andering state may take many decades or centuries to
become susceptible to avulsion as this increases in
probability as sinuosity increases. This element,
® .which is in essence Lane and Richards' 1997 `con-
figurational state', requires testing using numerical
modelling and observations of systems over appro-
priate time scales, probably 1­2 ka years in
medium-sized rivers.
7. Conclusions
If it is accepted that the response time of geomor-
® .phic systems is scale-related i.e. size-time related
then medium-sized to larger channels will only show
explainable changes at the 102
­103
time scale. If
this is the case then historical and archaeological
information is invaluable in allowing a reconstruc-
tion of fluvial trends, which may then be related to
forcing factors such as climate and land use change.
The broadening of alluvial palaeoecology has also
allowed a new range of techniques to be used at
®indicators of fluvial conditions Amoros and Van
.Urk, 1989 . In particular, Coleoptera have the poten-
tial to add process detail to reconstructions of past
fluvial regime including bed and bank conditions and
floodplain land-use. In this case, they show how the
channel had a mobile, un-armoured bed and rela-
tively lower suspended sediment load than it carries
today.
At both Hemington and Colwick, there is a cyclic
phase of channel change from single channel mean-
dering to active braided to fixed multi-channel state
® .anastomosing and finally back to a single channel
meandering state. This cycle rakes place over 300­
400 years between the 9th and 14th centuries, or a
® .little 100­200 years later at Colwick, and is driven
® .by a series of large floods Fig. 5 . During this
period, the recurrence time of floods was probably
far shorter than the relaxation time of the channel
causing channel conditions to be predominantly tran-
® .sient TF)1 sensu Brunsden and Thornes 1979 .
The climatic context of these floods has been dis-
® .cussed elsewhere Brown, 1998 ; however, they do
coincide with the Late Medieval Climatic Deteriora-
tion, which is generally accepted to have been a time
of climatic instability. This channel response is
unique for a larger lowland river in England and is
almost certainly the result of the unusual hydrometry
of the basin. Within 50 km of the Middle Trent, four
major tributaries join the main channel and, as
Fig. 5. Late Holocene cycle of channel change at Hemington, the
driving mechanism is believed to be a flood-induced increase in
bedload transport rates and low bank resistance.
( )A.G. Brown et al.rGeomorphology 39 2001 69­8280
® .Knighton 1987 has shown, the estimated mean
annual flood increases from around under 200 m3
sy1
to nearly 450 m3
sy1
. Two of these tributaries
drain adjacent uplands and they alone increase the
3 y1 ®estimated mean annual flood by 200 m s Knigh-
. ® .ton, 1987 . As Knighton 1998, p. 61 remarks such
Adiscontinuity present in the fluvial system, . . . has
implications for downstream channel adjustmentB.
The result is that the Trent is very sensitive to rain
on snow events and these can cause a switch in the
ratio of contributing discharge from a Trent-to a
® .Derwent-dominance Brown, 1998 . The difference
in hydrology is illustrated by the flood multiplier
® .ratio of mean Q to maximum recorded Q , which is
1.6 for the Trent at Shardlow and 3.2 for the Der-
® .went at Longbridge based on NERC 1975 figures .
As Hemington is at the junction of the Trent and
Derwent, this may have been particularly important
in driving channel change in this reach. From the
14th century onwards, there is little relationship be-
tween the rates or relative magnitudes of channel
change and flood history. Several possible reasons
include a change in channel form causing an auto-
genic change in channel sensitivity, and increase in
suspended sediment load and direct channel modifi-
cations encouraging a fixed single-channel form. The
remarkable Late Holocene history of the Middle
Trent illustrates how the response to floods may
depend on the state of the channel and this is deter-
®mined not only by the past flood history sensu
.Erskine and Warner, 1999 but any other change in
channel or floodplain conditions including both di-
rect and indirect human impacts.
Acknowledgements
The authors must acknowledge all involved in the
Hemington project especially S. Ripper, P. Clay, M.
Beamish and other members of University of Leices-
® .ter Archaeological Services ULAS . The work was
begun with support from J. Mellor and the Leicester-
shire Archaeological Unit, English Heritage with the
agreement and help of Ennemix Construction materi-
als. Work has continued supported by Lafarge Red-
land. Invaluable assistance was also given by the late
H. Potter and his work on the flood records of the
Trent.
References
Acreman, M.C., 1989. Extreme historical UK floods and maxi-
mum flood estimation. Journal of the Institute of Water Engi-
neers and Managers 3, 404­412.
Allan, T., Hey, G., Miles, D., 1997. A line of time: approaches to
archaeology in the upper and middle Thames valley. World
Archaeology 29, 114­129.
Amoros, C., Van Urk, G., 1989. Palaeoecological analysis of large
rivers: some principles and methods. In: Petts, G.E., Moller,¨
® .H., Roux, A.L. Eds. , Historical Change of Large Alluvial
Rivers: Western Europe. Wiley, Chichester, pp. 167­182.
Archer, N., 1993. Discharge estimate for Britain's greates flood:
River Tyne, 17th November 1771. British Hydrological Soci-
ety 4th National Hydrology, 4.1­4.6.
Borck, H-E., 1989. Soil erosion during the past millennium in
Central Europe and its significance within the geomorphody-
® .namics of the Holocene. In: Ahnert, A. Ed. , Landforms and
Landform Evolution in Wester Germany. Catena Supplement,
vol. 15, pp. 121­131.
Britton C.E., 1937. A meteorological chronology to AD 1450.
Meteorological Office Geophysical memoirs 70, London
HMSO, 177 pp.
Brown, A.G., 1991. Hydrogeomorphology and palaeoecology of
the Severn basin during the last 18,000 years: orders of change
in a maritime catchment. In: Gregory, K.J., Starkel, L.,
® .Thornes, J.B. Eds. , Fluvial Processes in the Temperate Zone
During the Last 15,000 Years. Wiley, Chichester, pp. 147­169.
Brown, A.G., 1996. Human dimensions of palaeohydrological
® .change. In: Branson, J., Brown, A.G., Gregory, K.J. Eds. ,
Palaeohydrology and Global Changes: The Context of Palaeo-
hydrology, Geological Society Special Publication, vol. 115,
Geological Society, London, pp. 57­72.
Brown, A.G., 1998. Fluvial evidence of the Medieval Warm
Period and the Late Medieval Climatic Deterioration. In:
® .Benito, G., Baker, G., Gregory, V.R. Eds. , Palaeohydrology
and Environmental Change. Wiley, Chichester, pp. 43­52.
Brown, A.G., Salisbury, C.R., in press. The geomorphology of the
® .Hemington reach. In: Cooper, L., Ripper, S. Eds. , The
Hemington Bridges: The Excavation of Three Medieval
Bridges at Hemington Quarry, Castle Donington, Leicester-
shire. English HeritagerULAS, Leicester.
Brunsden, D., Thornes, J.B., 1979. Landscape sensitivity and
change. Transactions of the Institute of British Geographers
New Series 4, 463­484.
Bull, W.B., 1991. Geomorphic Response to Climatic Change.
Oxford Univ. Press, New York.
Clay, P., Salisbury, C.R., 1990. A Norman mill dam and other
sites at Hemington fields, Castle Donington, Leicestershire.
The Archaeological Journal 147, 276­307.
Cooper, L., 1998. An evaluation across the projected line of the
Hemingyon Bridges. Transactions of the Leicestershire Ar-
chaeological and Historical Society 72, 164.
Cooper, L., 1999. Castle DoningtonrLockington­Hemington,
Hemington Quarry. Transactions of the Leicestershire Archae-
ological and Historical Society 73, 91­97.
Cooper, L., Ripper, S., 1994. The Medieval Trent Bridges at
( )A.G. Brown et al.rGeomorphology 39 2001 69­82 81
Hemington Fields, Castle Donington. Transactions of the Le-
icestershire Archaeological and Historical Society 68, 153­
161.
Cooper, L., Ripper, S., in press. The Hemington Bridges. Univer-
sity of Leicester Archaeological ServicesrEnglish Heritage,
Leicester.
Cooper, L., Ripper, S., Clay, P., 1994. The Hemington Bridges.
Current Archaeology 30, 197.
Courtney, P., in press. Hemington Bridge in historical context. In:
® .Cooper, L., Ripper, S. Eds. , The Hemington Bridges: The
Excavationm of Three Medieval Bridges at Hemington Quarry,
Castle Donington, Leicestershire. English HeritagerULAS,
Leicester.
Cummins, W.A., Rundle, A.J., 1969. The geological environment
of the dug-out canoes from Holme Pierrepoint, Notting-
hamshire. Mercian Geologist 3, 177­188.
Dinnin, M., 1997. Holocene Beetle Assemblages from the Lower
Trent Floodplain at Bole Ings, Nottinghamshire, UK. Quater-
nary Proceedings 5, 83­104.
Ellis, C., Brown, A.G., 1998. The archaeomagnetic dating of
palaeochannel sediments: data from the Medieval channel fills
at Hemington, Leicestershire. Journal of Archaeological Sci-
ence 25, 149­163.
Erskine, W.D., Warner, R.F., 1999. Significance of riverbank
erosion as a sediment source in the alternating flood regimes
of south-eastern Australia. In: Brown, A.G., Quine, T.A.
® .Eds. , Fluvial Processes and Environmental Change. Wiley,
Chichester, pp. 139­164.
Ferguson, R.I., 1981. Channel form and channel changes. In:
® .Lewin, J. Ed. , British Rivers. Allen and Unwin, London, pp.
90­125.
Foster, I.D.L., Mighall, T.M., Wotton, C., Owens, P.N., Walling,
D.E., 2000. Evidence of medieval soil erosion in the South
Hams region of Devon, UK. Holocene 10, 261­271.
Friday, L.E., 1988. A key to the adults of the British water
beetles. Field Studies 7, 1­151.
Graff, W.L., 1988. Applications of catastrophe theory in fluvial
® .geomorphology. In: Anderson, M.G. Ed. , Modelling Geo-
morphological Systems. Wiley, Chichester, pp. 3­47.
Green, F.H.W., 1979. Field under-drainage and the hydrological
® .cycle. In: Hollis, G.H. Ed. , Man's Impact on the Hydrologi-
cal Cycle in the United Kingdom. Geo Abstracts, Norwich, pp.
9­18.
Harvey, A.M., Renwick, W.H., 1987. Holocene alluvial fan and
terrace formation in the Bowland Fells, northwest England.
Earth Surface Processes and Landforms 12, 249­257.
Higgs, G., 1987. Environmental change and hydrological re-
sponse: flooding in the Upper Severn catchment. In: Gregory,
® .K.J., Lewin, J., Thornes, J.B. Eds. , Palaeohydrology in
Practice: A River Basin Analysis. Wiley, Chichester, pp.
131­160.
Holland, D.G., 1972. A key to the larvae, pupae and adults of the
British species of Elminthidae. Freshwater Biological Associa-
tion Scientific Publication, Windermere, p. 26.
Hooke, J.M., 1977. The distribution and nature of changes in river
® .channel pattern. In: Gregory, K.J. Ed. , River Channel
Changes. Wiley, Chichester, pp. 280­285.
Hooke, J.M., Harvey, A.M., Miller, S.Y., Redmond, C.E., 1990.
The chronology and stratigraphy of the alluvial terraces of the
river Dane valley, Cheshire. Earth Surface Processes and
Landforms 15, 717­737.
Howard, A., Macklin, M.G., 1999. A generic geomorphological
approach to archaeological interpretation and prospection in
British river valleys: a guide for archaeologists investigating
Holocene landscapes. Antiquity 73, 527­541.
Howard, A.J., Smith, D., Garton, D., Hillam, J., Pearce, M., 1999.
Middle to Late Holocene environmental change in the Middle
® .and Lower Trent Valley. In: Brown, A.G., Quine, T.A. Eds. ,
Fluvial Processes and Environmental Change. Wiley, London,
pp. 165­178.
® .Hughes, M.K., Diaz, H.F. Eds. , 1994. The Medieval Warm
Period. Kluwer Academic Publishing, Dordrecht, p. 342.
Kenward, H.K., Hall, A.R., Jones, A.K.G., 1980. A tested set of
techniques for the extraction of plant and animal macrofossils
from waterlogged archaeological deposits. Scientific Archaeol-
ogy 22, 3­15.
Knighton, A.D., 1987. Estimating the mean annual flood of the
Trent basin. East Midland Geographer 10, 1­6.
Knighton, A.D., 1998. Fluvial Forms and Processes: A New
Perspective. Arnold, London.
Lane, S.N., Richards, K.S., 1997. Linking river channel form and
process: time, space and causality revisited. Earth Surface
Processes and Landforms 22, 249­260.
Lindroth, C.H., 1974. Coleoptera: carabidae. Handbooks for the
Identification of British Insects: IV. Part 2. Royal Entomologi-
cal Society of London, London.
Lord, P., Salisbury, C.R., 1997. Brush-piling: eighteenth century
engineering in an American wilderness. Industrial Archaeol-
ogy Review 19, 49­60.
Losco-Bradley, P.M., Salisbury, C., 1988. A Saxon and Norman
® .fishweir at Colwick, Nottinghamshire. In: Aston, M. Ed. ,
Medieval Fish, Fisheries and Fishponds in England. British
Archaeological Reports International Series, vol. 182, pp.
329­351.
Lucht, W.H., 1987. Die Kafer Mitteleuropas. Katalog. Goecke and¨
Evers, Krefeld.
Merrett, S.P., Macklin, M.G., 1999. Historic river response to
extreme flooding in the Yorkshire Dales, Northern England.
® .In: Brown, A.G., Quine, T.A. Eds. , Fluvial Processes and
Environmental Change. Wiley, Chichester, pp. 345­360.
Miall, A.D., 1978. Lithofacies types and vertical profile models in
® .braided river deposits: a summary. In: Miall, A.D. Ed. ,
Fluvial Sedimentology. Canadian Society of Petroleum Geolo-
gists, Memoir, vol. 5, pp. 597­604.
Needham, S., 1992. Holocene alluviation and interstratified settle-
ment evidence in the Thames valley at Runnymede Bridge. In:
® .Needham, S., Macklin, M.G. Eds. , Alluvial Archaeology in
Britain. Oxbow Books, Oxford, pp. 249­260.
Nicholas, A.P., Ashworth, P.J., Kirkby, M.J., Macklin, M.G.,
Murray, T., 1995. Sediment slugs: large-scale fluctuations in
fluvial sediment transport rates and storage volumes. Progress
in Physical Geography 19, 501­519.
®Nilsson, A.N., Holmen, M., 1995. The aquatic Adephaga Cole-
.optera of Fennoscandia and Denmark: II. Dytiscidae. Fauna
Entomologica Scandinavica 32, 1­192.
Osborne, P.J., 1988. A late bronze age insect fauna from the river
( )A.G. Brown et al.rGeomorphology 39 2001 69­8282
avon, Warwickshire, England: its implications for the terres-
trial and fluvial environment and for climate. Journal of
Archaeological Science 15, 715­727.
Passmore, D.G., Macklin, M.G., Brewer, P.A., Lewin, J., Rumsby,
B.T., Newson, M.D., 1993. Variability of late Holocene braid-
® .ing in Britain. In: Best, J., Bristow, C. Eds. , Braided Rivers:
Form, Process and Economic Applications. Geological Society
Special Publication, London, pp. 205­230.
Potter, H.R., 1964. Introduction to the history of the floods and
droughts of the trent basin. Unpublished manuscript of a
Conference held at Loughborough University, Loughborough,
October.
Riley, D.N., Buckland, P.C., Wade, J.S., 1995. Aerial reconnais-
sance and excavation at Littleborough-on-Trent, Notts. Britan-
nia 36, 253­284.
Robinson, M.A., 1991. Neolithic and late bronze age insect
® .assemblages. In: Needham, S. Ed. , Excavation and Salvage
at Runnymede Bridge 1978. British Museum Press and En-
glish Heritage, London, pp. 277­326.
Robinson, M.A., 1993. The insect remains. In: Allen, T.G.,
Robinson, M.A., The Prehistoric Landscape and Iron Age
Enclosed Settlement at Mingies Ditch. Hardwick-with-Yel-
ford, Oxon. Thames Valley Landscapes: The Windrush Valley
Volume 2. The Oxford Archaeological Unit, pp. 7­16 and
101­141.
Robinson, M., Lambrick, G.H., 1984. Holocene alluviation and
hydrology in the Upper Thames basin. Nature 308, 809­814.
Rumsby, B.T., Macklin, M.G., 1996. River response to the last
® .neoglacial the `Little Ice Age' in northern, central and
western Europe. In: Branson, J., Brown, A.G., Gregory, K.J.
® .Eds. , Global Continental Changes: The Context of Palaeohy-
drology Geological Society Special Publications, vol. 115, pp.
217­233.
Salisbury, C.R., 1983. An early Tudor map of the river Trent in
Nottinghamshire. Transactions of the Thoreton Society 87,
54­59.
Salisbury, C.R., 1984. Flandrian courses of the River Trent at
Colwick, Nottinghamshire. Mercian Geologist 9, 198­207.
Salisbury, C.R., 1992. The archaeological evidence for
palaeochannels in the Trent valley. In: Needham, S., Macklin,
® .M.G. Eds. , Alluvial Archaeology in Britain. Oxbow Books,
Oxford, pp. 155­162.
Salisbury, C.R., Whitley, P.J., Litton, C.D., Fox, J.L., 1984.
Flandrian courses of the river Trent at Colwick Nottingham.
Mercian Geologist 9, 189­207.
Shirt, D.B., 1987. British Red Data Books: 2. Insects. Nature
Conservancy Council.
Smith, D.N., in press. Disappearance of Elmid Ariffle beetlesB
from lowland river systems--the impact of alluviation. In:
® .O'Connor, T., Nicholson, R. Eds. , People as an Agent of
Environmental Change. Oxford: Oxbow Books.
Steffan, A.W., 1979. Familie: dryopidae. In: Freude, H., Harde,
® .K.W., Lohse, G.A. Eds. , Die Kafer Mitteleuropas. Band, vol.¨
6, Goecke and Evers, Krefeld, pp. 265­293.
Van Vliet-Lanoe, B., 1986. Micromorphology. In: Calloe, P.,¨
® .Cornford, J.M. Eds. , La Cotte de St. Brelade 1961­1978:
Excavations by C.B.M. McBurney. Geo Books, Norwich, pp.
91­96.
Worsley, P., 1996. On ice-wedge casts and soft sediment deforma-
tions. Quaternary Newsletter 78, 1­7.