The Science of the Total Environment 266 Ž2001.

113᎐123

Controls on the nutrient content of suspended

sediment transported by British rivers

D.E. WallingU , M.A. Russell, B.W. Webb

Department of Geography, Uni¨ ersity of Exeter, Exeter EX4 4RJ, UK

Received 17 September 1999; accepted 28 June 2000

Abstract

Recent studies of nutrient fluxes from river basins have emphasised the importance of sediment-associated
transport. In order to develop an improved understanding of sediment-associated nutrient transport in UK rivers and
of contrasts between individual rivers, attention has been directed to spatial and temporal variations in the nutrient
ŽN and P. content of suspended sediment transported by four rivers, which embrace a representative range of British
conditions. Bulk samples of suspended sediment were collected during storm events and these were analysed for both
total N and P content and for individual fractions of these nutrients. Significant temporal and spatial variability in
these various measures of the nutrient content of suspended sediment has been documented. Spatial variability has
been linked to catchment characteristics, which in turn influence sediment sources. Patterns of temporal variability,
which are reasonably consistent among the different rivers, have been related both to variations in other sediment
properties and to hydrometeorological conditions. 䊚 2001 Elsevier Science B.V. All rights reserved.

Keywords: Nutrients; Suspended sediment; Phosphorus; Nitrogen

U
Corresponding author. Tel.: q44-1392-263345; fax: q44-1392-263342.
E-mail address: d.e.walling@exeter.ac.uk ŽD.E. Walling..

0048-9697r01r$ - see front matter 䊚 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 8 - 9 6 9 7 Ž 0 0 . 0 0 7 4 6 - 4
114 D.E. Walling et al. r The Science of the Total En¨ ironment 266 (2001) 113᎐123
1. Introduction
Recent studies of nutrient fluxes from river
basins and investigations of catchment nutrient
budgets have emphasised the importance of sedi-
ment-associated transport Žcf. Kronvang, 1992;
Dorioz, 1996; Foster et al., 1996.. In a review of
the role of suspended sediment in the global
transport of major elements from the land to the
oceans by rivers, Meybeck Ž1984. estimated that
sediment-associated transport contributes approx-
imately 57 and 95%, respectively, of the total N
and P flux. The precise magnitude of the sedi-
ment-associated component of the nutrient flux
will, however, vary from river to river in response
to local conditions and in a detailed investigation
of nutrient fluxes in four British rivers. Walling et
al. Ž1997. reported that sediment-associated
transport accounted for 26᎐75% of the annual
total-P load and 3᎐8% of the annual total-N load.
Differences among the individual rivers were
shown to reflect the influence of catchment char-
acteristics such as relief, geology and land use,
the hydrometeorological conditions and the rela-
tive importance of point source inputs.
In simple terms, the magnitude of the sedi-
ment-associated component of the specific nutri-
ent flux Žt kmy2 yeary1 . will reflect the magni-
tude of the specific suspended sediment yield and
the nutrient content of the suspended sediment
transported by the river. The latter will in turn
reflect the catchment soil types and land use, the
selectivity of the erosion and sediment delivery
processes responsible for mobilising the sediment
and transporting it to and through the river sys-
tem and the potential for nutrient uptake by the
sediment from the water column. Information on
the nutrient content of suspended sediment and
an appreciation of the factors influencing the
magnitude of this content must, therefore, be
seen as an important requirement for developing
an improved understanding of the sediment-asso-
ciated component of catchment nutrient budgets.
Furthermore, coupling information on the nutri-
ent content of suspended sediment with sus-
pended sediment load data could afford an effec-
tive means of estimating the magnitude of the
sediment-associated nutrient flux in river basins
where detailed monitoring has not been under-
taken. Any attempt to combine sediment load
data with information on the average nutrient
content of suspended sediment would, neverthe-
less, need to be based on a sound assessment of
the temporal variability of this content. To begin
to meet these requirements in relation to British
rivers, attention has been directed to document-
ing the spatial and temporal variability of the
nutrient content of suspended sediment trans-
ported by four rivers, which embrace a represen-
tative range of catchment characteristics and hy-
drological regimes, and to identifying the key
factors influencing such spatial and temporal vari-
ability.
2. The study rivers
The study rivers ŽFig. 1., which comprise the
River Severn, the Warwickshire Avon, the River
Exe and its tributary the River Dart were selected
to embrace a range of catchment characteristics,
including size and effluent inputs Žcf. Table 1..
The River Severn is the largest single drainage
basin in England and Wales and the river drains
an area of 6850 km2 above the monitoring station
at Upton on Severn. The catchment embraces a
great variety of terrain types, ranging from moun-
tainous uplands in the western headwaters, pre-
dominantly underlain by Ordovician and Silurian
mudstones, shales and sandstones; through rolling
hills in the central areas, underlain primarily by
Carboniferous and Devonian sandstones and
limestones; to lowlands in the south and north-
east developed on Triassic sandstones and mud-
stones. Mean annual precipitation ranges from
approximately 2300 mm in the west to less than
800 mm in the south. The land use in the western
and central regions is dominated by pastoral
farming, but arable farming is of major impor-
tance in the lowlands. Several large urban centres
discharge effluent to the River Severn along its
course.
The River Avon is major tributary of the River
Severn, with a catchment area of 2674 km2 above
the gauging station at Bredon. It ranks as the
tenth largest river in Britain in terms of drainage
 D.E. Walling et al. r The Science of the Total En¨ ironment 266 (2001) 113᎐123 115

Fig. 1. The study catchments.

area. The catchment is relatively flat and over mm in the south-west. A large proportion of the
80% of its area is underlain by impermeable clays catchment is devoted to arable cultivation, but
and mudstones. Annual precipitation varies from more than 7% of the area is classified as urban.
720 mm in the north-east to approximately 600 Nearly 1 million people live within the catchment,
116 D.E. Walling et al. r The Science of the Total En¨ ironment 266 (2001) 113᎐123
Table 1
Summary catchment characteristics for the four study rivers
River Monitoring Area
and Žkm2 .
sampling site
Severn Upton 6850
Avon Bredon 2674
Exe Thorverton 601
Dart Bickleigh 46
a
Population equivalent served by sewage treatment works.
and discharges of sewage effluent to the river
system total approximately 200 Ml dayy1 and
account for approximately half of the dry weather
flow at the catchment outlet.
The catchment of the River Exe, which drains
an area of 601 km2 above the gauging station at
Thorverton, is underlain by resistant Devonian
grits and slates in the north and Carboniferous
and Permian shales and sandstones towards the
south. The headwaters rise on Exmoor and are
characterised by moorland and upland pasture,
whilst further south the river drains an area of
undulating terrain dominated by pasture, but with
some arable cultivation. The mean annual precip-
itation ranges from 1800 mm in the north to 850
mm at the basin outlet. There are few urban
areas within the catchment, and effluent dis-
charges contribute only a small proportion of the
flow.
The River Dart Ž46 km2 . is the smallest of the
study catchments and is a tributary of the River
Exe. It is characterised by steeply sloping incised
terrain developed on Upper Carboniferous sand-
stones and shales. Land use is dominated by
permanent pasture and dairy farming, but some
cultivation of cereals and fodder crops occurs on
the interfluves and other flatter areas. The mean
annual precipitation varies between 900 and 1200
mm and there are no significant effluent dis-
charges within the catchment.
3. Methods
The sampling programme employed for the
study involved the collection of bulk suspended
Mean % Arable Population
annual Q equiv.a
Žm3 sy1 . Žkmy2 .
84 32 99
16 49 261
16 11 59
1 12 1
sediment samples Ž100᎐200 l. at the gauging sta-
tions associated with the four river basins, during
selected storm events over the period January
1995 to January 1996. The events were selected to
provide a representative coverage of event magni-
tude Žpeak discharge. and time of year. Samples
were collected only when suspended sediment
concentrations exceeded 100 mg ly1 ; sampled
concentrations were generally in the range
100᎐1000 mg ly1 . The sediment was recovered
from these samples by continuous flow centrifu-
gation and a proportion of this was freeze dried.
Analysis of the nutrient content of the sus-
pended sediment samples involved measurement
of the following forms: Non-apatite inorganic P
ŽNAIP., Apatite P ŽAP., Algal available P ŽAAP.,
Organic P ŽOP., Total P ŽTP., Inorganic N ŽIN.
and Total Nitrogen ŽTN.. NAIP and AP were
determined by the method of Williams et al.
Ž1976. and TP was measured following the proce-
dure of Olsen and Sommers Ž1982.. Organic P
was calculated by difference as TP-ŽNAIPq AP..
AAP and IN were determined using fresh sedi-
ment within 24 h of collection using the methods
of Dorich et al. Ž1984. and Keeney and Nelson
Ž1982., respectively. TN was measured using a
Carlo Erba 1400 Automatic Carbon-Nitrogen
Analyser.
These measurements of the nutrient content of
the sediment were complemented by analyses of
several other sediment properties listed in Table
2 including, pH, total organic carbon ŽTOC., ab-
solute particle size composition Ž% - 63 ␮m, %
- 2 ␮m and d50 ., specific surface area ŽSSA.,
cation exchange capacity ŽCEC., calcium carbo-
 D.E. Walling et al. r The Science of the Total En¨ ironment 266 (2001) 113᎐123 117

Table 2 is in the range 3.7᎐7.2%. Suspended sediment

Mean values for selected chemical and physical properties of
from the Rivers Exe and Dart has a higher TOC
suspended sediment from the four study rivers a
content than that from the Severn and Avon and
Severn Avon Exe Dart this can be attributed to contrasts in landuse, with
the pasture soils of the Exe and Dart catchments
TOC 37 207 39 787 72 301 65 501 being characterised by a higher organic matter
CEC 27.4 42.5 32.2 28.9
content than the arable soils of the Avon catch-
CaCO3 1788 6861 674 649
Feob 5950 5960 10 168 10 276 ment, and to the increased importance of channel
Mnob 751 611 1415 1458 and gully erosion and subsoil sediment sources
% - 63 ␮m 97.1 98.8 96.8 93.7 within the mountainous and upland headwaters
% - 2 ␮m 20.2 31.8 23.2 23.8 of the Severn catchment, which can be expected
d50 7.69 4.31 5.7 5.65
to produce sediment with a lower organic matter
SSA 1.07 1.70 1.24 1.20
n 17 11 22 19 content. Contrasts in the calcite ŽCaCO 3 . content
of the suspended sediment among the four study
a
All values in ␮g gy1 except CEC, mol c kgy1 ; d50 , ␮m; rivers largely reflect geological control and the
specific surface area, m2 gy1 .
b
Measured using the method of Schwertmann Ž1964.. presence of limestone outcrops in the Avon
catchment, where the CaCO 3 content of sus-
pended sediment is an order of magnitude or
nate ŽCaC0 3 . content, and the oxalate-extractable
more greater than that in the other rivers. A
iron ŽFe o . and manganese ŽMn o . content.
more detailed analysis of contrasts in suspended
sediment properties among the study rivers is
provided by Table 3 which represents the results
4. Results of an analysis of variance for selected properties.
Although the properties of suspended sediment
4.1. Suspended sediment properties from the Rivers Dart and Exe are, as might be
expected, generally similar, there are clear con-
Mean values for a number of physical and trasts between these and the other two rivers for
chemical properties of the suspended sediment many of the properties. As indicated above, these
transported by the four rivers are represented in contrasts reflect the influence of the varying
Table 2. The suspended sediment transported by catchment characteristics Že.g. geology, topogra-
all four rivers can be seen to be relatively fine, phy and land use., and the primary suspended
with the silt q clay Ž- 63 ␮m. fraction accounting sediment sources.
on average for approximately 95% of the total
mass and the clay Ž- 2 ␮m. fraction 20᎐30%. 4.2. Inter-catchment ¨ ariability in the nutrient
The suspended sediment transported by the River content of suspended sediment
Avon, with an average clay fraction content of
32%, is significantly finer than that transported by Mean values for the seven measures of nutrient
the Rivers Severn, Exe and Dart and this contrast content of suspended sediment analysed in the
is reflected in the increased SSA and CEC values. study are represented in Table 4. In all the study
The finer suspended sediment transported by the rivers, the TP content of the suspended sediment
River Avon can be accounted for in terms of the is dominated by NAIP Ž60᎐70%. which in turn
‘softer’ rocks Žclays and mudstones. found in this accounts for AAP representing a substantial pro-
catchment and the low relief, which is likely to be portion Ž48᎐59%. of the TP content. The TN
reflected by reduced efficiency of sediment deliv- content of the suspended sediment is predomi-
ery through the slope-channel system, and thus nantly organic in form and this fraction accounts
greater opportunity for deposition of the coarser for 94᎐98% of the TN. The contrasts in sus-
fractions. The average organic carbon ŽTOC. con- pended sediment properties among the four study
tent of suspended sediment from the study rivers catchments outlined above are paralleled by sev-
118 D.E. Walling et al. r The Science of the Total En¨ ironment 266 (2001) 113᎐123

Table 3
The results of an analysis of variance used to identify significant differences in the suspended sediment properties among the four
study rivers a

TOC CEC CaCO3 Feo

S A E D S A E D S A E D S A E D

S
A ᎐ = = ᎐
E = = ᎐ = = = ᎐ =
D = = ᎐ ᎐ = ᎐ = = ᎐ ᎐ = =

- 2 ␮m - 63 ␮m d50 SSA
S A E D S A E D S A E D S A E D

S
A = ᎐ = =
E = = ᎐ = = = = =
D = = ᎐ = = = = = ᎐ = = ᎐
a
The study rivers have been designated S ŽSevern., A ŽAvon., E ŽExe., D ŽDart.. Significant differences at the 5% level are
shown by crosses.

eral differences in the nutrient content of sus-
pended sediment. Table 5 represents the results
of an analysis of variance aimed at testing the
significance of inter-catchment contrasts. As
shown by Table 5, the AAP, OP and IN contents
of suspended sediment are not significantly dif-
ferent among the catchments, but significant con-
trasts are found among some of the rivers for TP,
NAIP, AP and TN. In the case of the TP, NAIP
and TN content of suspended sediment, the key
contrast is between the River Severn and the
other three rivers, with the Severn producing
lower values. These lower values are likely to
reflect both the coarser nature of the suspended
sediment from the River Severn Žcf. Table 2. and
the greater proportion of upland area within the
Severn catchment, which will be characterised by
soils with a lower nutrient status. The AP content
of suspended sediment shows significant differ-
ences among all catchments. These contrasts re-
flect the varying underlying geology of the
catchments, since the AP content of suspended
sediment can be expected to reflect the miner-
alogy of the rocks outcropping in a drainage
basin. Although the River Avon catchment ex-
hibits an important contrast with the other study
basins in terms of the greater importance of ef-
fluent discharges from STWs ŽTable 1., this con-

Table 4
Mean values for the nutrient content of suspended sediment from the four study rivers a

Severn Avon Exe Dart

TP 1105 Ž261. 1637 Ž429. 1633 Ž362. 1585 Ž404.

AAP 648 Ž326. 934 Ž415. 798 Ž223. 760 Ž333.
NAIP 661 Ž307. 1102 Ž229. 1082 Ž260. 1136 Ž170.
AP 149 Ž26. 199 Ž62. 88 Ž30. 54 Ž30.
OP 295 Ž130. 336 Ž247. 463 Ž227. 395 Ž299.
TN 2344 Ž828. 5521 Ž792. 4672 Ž1217. 5323 Ž1911.
IN 150 Ž133. 148 Ž101. 284 Ž198. 228 Ž135.
a
All values in ␮g gy1 . Standard deviation given in parentheses.
 D.E. Walling et al. r The Science of the Total En¨ ironment 266 (2001) 113᎐123 119

Table 5
The results of an analysis of variance used to identify significant differences in the nutrient content of suspended sediment between
the four study rivers a

TP NAIP AP OP
S A E D S A E D S A E D S A E D

S
A = = = ᎐
E = ᎐ = ᎐ = = ᎐ ᎐
D = ᎐ ᎐ = ᎐ ᎐ = = = ᎐ ᎐ ᎐

AAP TN IN
S A E D S A E D S A E D

S
A ᎐ ᎐ ᎐
E ᎐ ᎐ ᎐ ᎐ ᎐ ᎐
D ᎐ ᎐ ᎐ = = = ᎐ ᎐ ᎐
a
The study rivers have been designated S ŽSevern., A ŽAvon., E ŽExe., D ŽDart.. Significant differences at the 5% level are
shown by crosses.

trast is not clearly reflected in the nutrient con-
tent of its suspended sediment. Russell et al.
Ž1998. have, however, identified important con-
trasts in the magnitude of the dissolved nutrient
export from this basin which are closely linked to
the importance of effluent discharges.
4.3. Temporal ¨ ariation in the nutrient content of
suspended sediment
As indicated by the standard deviation values
for the mean nutrient concentrations listed in
Table 4, the nutrient content of suspended sedi-
ment collected from the four study rivers evi-
dences appreciable temporal variability. Further
analysis of this variability has focused on assess-
ing the influence of both temporal variability in
the physical and chemical properties of the sus-
pended sediment and more general hydrometeo-
rological controls, such as season and river dis-
charge, on the nutrient content of suspended
sediment. This has been achieved by recording
the Pearson correlation coefficients between the
values of nutrient concentration associated with
the suspended sediment samples collected from
the individual rivers and four independent vari-
ables representing suspended sediment properties
and four representing hydrometeorological con-
trols Žcf. Table 6.. These independent variables
were selected on the basis of a correlation analy-
sis undertaken on a larger set of potential inde-
pendent variables and aimed at identifying impor-
tant and independent controls.
The four suspended sediment properties in-
cluded in Table 6 are TOC, CEC, Fe o and % - 2
␮m. Only the correlation coefficients for correla-
tions significant at ) 95% have been listed. The
results indicate that temporal variations in the
four suspended sediment properties exert a sig-
nificant control over nutrient content, although
the precise extent of this influence varies from
river to river and among the individual nutrient
parameters. It is noteworthy that the largest set
of significant relationships has been recorded for
the River Dart. This is a reflection of the small
size of this catchment. With only a single rock
type and little variation in terrain characteristics
within the catchment, there is little opportunity
for the relationship between nutrient content and
sediment properties to be complicated by varia-
tions in the source area of the sediment. In con-
trast, the River Severn, which has the largest
drainage basin and the greatest diversity of ge-
ology and terrain characteristics, provides the
smallest number of significant correlations. In a
large catchment, the precise relationship between,
120 D.E. Walling et al. r The Science of the Total En¨ ironment 266 (2001) 113᎐123

Table 6
A correlation matrix showing significant Žat 95% level. relationships between the nutrient content of suspended sediment for the
four study rivers and a selection of independent variables representing suspended sediment properties and hydrometeorological
conditions

AAP NAIP AP OP TP IN TN

Se¨ ern
Sediment properties
TOC 0.90 0.94 0.60 y0.66 0.84 0.86 0.96
CEC ᎐ ᎐ ᎐ ᎐ ᎐ ᎐ ᎐
Fe0 0.71 0.79 ᎐ ᎐ 0.74 0.63 0.68
% - 2 ␮m ᎐ ᎐ ᎐ ᎐ ᎐ ᎐ ᎐
Hydrometeorological conditions
log Q y0.89 y0.82 y0.73 0.56 y0.76 y0.74 y0.92
log Q7 y0.86 y0.93 y0.58 0.68 y0.81 y0.86 y0.90
SSC y ᎐ ᎐ ᎐ ᎐ ᎐ ᎐
Season 0.57 0.59 0.55 y0.59 ᎐ 0.72 0.70

A¨ on
Sediment properties
TOC 0.95 0.90 0.87 ᎐ 0.81 0.90 0.96
CEC ᎐ ᎐ ᎐ ᎐ ᎐ y0.78 y0.71
Fe0 0.82 0.90 0.73 ᎐ 0.72 0.94 0.95
% - 2␮m ᎐ ᎐ ᎐ ᎐ ᎐ y0.65 ᎐
Hydrometeorological conditions
log Q ᎐ ᎐ ᎐ 0.72 0.69 ᎐ ᎐
log Q7 y0.68 ᎐ ᎐ ᎐ ᎐ ᎐ ᎐
SSC 0.80 ᎐ 0.70 0.86 0.89 ᎐ ᎐
Season 0.93 0.83 0.75 ᎐ 0.83 0.93 0.90

Exe
Sediment properties
TOC 0.83 0.70 ᎐ ᎐ 0.62 0.59 0.88
CEC 0.62 0.65 ᎐ ᎐ 0.69 ᎐ ᎐
Fe0 0.78 0.49 ᎐ ᎐ ᎐ 0.72 0.72
% - 2␮m ᎐ ᎐ ᎐ ᎐ ᎐ ᎐ ᎐
Hydrometeorological conditions
log Q y0.84 y0.74 ᎐ ᎐ y0.79 y0.59 y0.85
log Q7 y0.86 y0.71 y0.44 ᎐ y0.63 y0.57 y0.84
SSC y0.45 ᎐ ᎐ ᎐ y0.44 ᎐ ᎐
Season 0.89 0.54 0.57 ᎐ 0.60 0.69 0.89

Dart
Sediment properties
TOC 0.75 0.62 ᎐ 0.76 0.82 0.51 0.60
CEC 0.53 0.58 ᎐ 0.64 0.71 ᎐ 0.46
Fe0 0.70 0.66 ᎐ ᎐ 0.59 0.70 0.58
% - 2␮m ᎐ 0.61 ᎐ 0.46 0.56 ᎐ ᎐
Hydrometeorological conditions
log Q y0.52 ᎐ ᎐ y0.48 y0.53 y0.59 ᎐
log Q7 y0.73 y0.55 ᎐ y0.62 y0.69 y0.74 y0.48
SSC ᎐ ᎐ ᎐ ᎐ ᎐ ᎐ ᎐
Season 0.68 ᎐ 0.48 0.61 0.62 0.73 0.50
 D.E. Walling et al. r The Science of the Total En¨ ironment 266 (2001) 113᎐123
for example, the grain size composition of the
suspended sediment and its nutrient content may
vary among subcatchments underlain by different
rock types and soil types. Thus, the relationship
established for the catchment outlet will be com-
plicated by the spatial origin of the suspended
sediment associated with individual samples. Of
the sediment properties, TOC appears to exert
the strongest control over the nutrient content of
the suspended sediment samples collected from
the four rivers, since significant relationships exist
for at least five of the measures of nutrient con-
tent for all rivers. With the exception of the OP
content of suspended sediment from the River
Severn, all the significant relationships are posi-
tive, indicating that the nutrient content of the
suspended sediment is directly related to the
magnitude of its organic fraction. In the case of
inorganic nutrient forms, this direct relationship
must reflect the influence of factors linked more
indirectly to the magnitude of the organic frac-
tion. As might be expected, the CEC and Fe o
content of the suspended sediment samples also
can be seen to exhibit a significant positive rela-
tionship with several of the nutrient parameters,
with significant inverse relationships being
recorded only in the case of IN and TN for the
River Avon. Perhaps, surprisingly, the grain size
composition of suspended sediment, as indexed
by the magnitude of the - 2-␮m fraction, ap-
pears to exert only a limited influence on the
nutrient content of the suspended sediment sam-
ples collected from the four study rivers. There
are no significant relationships with this variable
for the Rivers Severn and Exe and a total of only
four significant relationships for the remaining
two rivers.
The four independent variables representing
hydrometeorological controls include the dis-
charge and suspended sediment concentration at
the time of sampling ŽlogQ and SSC., a measure
of antecedent flow conditions provided by the
average flow in the previous 7 days ŽlogQ 7 ., and a
seasonal index calculated as wŽsineŽjulian day,
where day 213 s 90⬚. q 1.r2. ŽSeason.x. For these
variables, the River Avon stands out as showing
the lowest number of significant correlations. This
situation can be explained by the increased im-
121
Fig. 2. Seasonal variation of the TP content of suspended
sediment samples collected from the study rivers.
portance of point sources and effluent discharges
to the nutrient loads transported by this river.
This component of the nutrient flux is likely to be
less strongly influenced by catchment-wide hydro-
logical processes, the variation of which will be
reflected by the four independent variables. Thus
whereas logQ and logQ 7 can be seen to exert a
significant influence on many of the nutrient
parameters for the Rivers Severn, Exe and Dart,
they are much less important for the River Avon.
Where significant relationships exist, these two
variables generally show negative relationships,
which conform to physical principles. The seaso-
nal index shows a significant positive relationship
with six out of seven nutrient parameters for each
river. An example of this relationship is provided
by Fig. 2, which plots the variation through the
year of the relative TP concentration of sus-
pended sediment for all samples collected from
the four rivers. In this example, and in the case of
all other relationships with this variable except
that for OP for the River Severn, the trend re-
flected by the Season variable involves a mini-
mum at the end of February and a maximum 6
months later at the end of August, indicating that
nutrient concentrations in the sediment are great-
est in the summer and lowest in the winter.
Nutrient content shows no significant correlation
with suspended sediment concentration for the
Rivers Severn and Dart and only a very limited
122 D.E. Walling et al. r The Science of the Total En¨ ironment 266 (2001) 113᎐123
number of significant relationships for the re-
maining two rivers. However, in the case of the
River Exe these relationships are negative, whilst
for the River Avon, they are positive, again indi-
cating the distinctive behaviour of the River Avon.
5. Perspective
The results presented above demonstrate a
number of important features relating to the con-
trols of the nutrient content of suspended sedi-
ment transported by British rivers. These are as
follows:
1. Despite substantial contrasts in terms of
catchment characteristics and hydrological
regime, there is only limited variation in both
suspended sediment properties and the nutri-
ent content of suspended sediment among
the four study rivers. The greatest contrasts
are associated with the River Severn, the
catchment of which extends into the Welsh
mountains and contains a large area of up-
land.
2. The nutrient content of suspended sediment
demonstrates appreciable temporal variation
in all four study rivers. The influence of sus-
pended sediment properties on this temporal
variability is most clearly shown for smaller
river basins, where such relationships are un-
complicated by temporal variations in the
spatial origin of sediment. Several suspended
sediment properties have been shown to exert
a significant influence on the nutrient content
of the sediment. Particle size composition was,
however, shown to be of limited importance
in influencing nutrient content. The influence
of hydrometeorological conditions on tem-
poral variations in the nutrient content of
suspended sediment is seen most clearly for
more natural catchments where point source
effluent discharges are of limited importance.
Discharge at the time of sampling, antecedent
flow conditions and time of year were all
shown to exert a significant influence on the
nutrient content of suspended sediment.
Further investigation of these controls on the
spatial and temporal variation of the nutrient
content of suspended sediment could provide a
useful basis for developing procedures for pre-
dicting the nutrient content of suspended sedi-
ment which could be used in estimating
sediment-associated nutrient fluxes on the basis
of a record of suspended sediment discharge and
limited measurements of the nutrient content of
the sediment. Future work could usefully extend
this study to include other UK rivers and more
extensive sampling to provide information on
variations in the nutrient content of suspended
sediment during individual storm events.
Acknowledgements
The results presented above represent part of a
larger study of nutrient transport by river systems
undertaken within the framework of the multina-
tional BINOCULARS research project which was
funded by the EC Environment Programme ŽPro-
ject EV5VCT940495.. The authors gratefully ac-
knowledge this financial support, the generosity
and help of local landowners and the Environ-
ment Agency in permitting access to river sam-
pling and monitoring sites, the co-operation of
the Environment Agency in providing flow data
for the monitoring sites on the Rivers Exe, Severn
and Avon, and the assistance of Rhonda Bearne
with laboratory analysis of water and sediment
samples.
References
Dorich RA, Nelson DW, Sommers LE. Algal availability of
phosphorus in suspended stream sediments of varying par-
ticle size. J Environ Qual 1984;13:82᎐86.
Dorioz JM. Phosphorus storage, transport and transformation
in river systems of the Lake Leman basin. In: Kronvang B,
Svendsen LM, Sibbesen E, editors. Sediment and Phospho-
rus. NERI Tech Rep 1996;178:81᎐88.
Foster IDL, Baban SMJ, Wade SD, Charlesworth SM, Buck-
land PJ, Wagstaff, K. Sediment-associated phosphorus
transport in the Warwickshire River Avon, UK. In: Walling
DE, Webb BW, editors. Erosion and sediment yield: global
and regional perspectives ŽProc Exeter Symp, July 1996.
1996; IAHS Publ no 236: 303᎐312.
Keeney DR, Nelson DW. Nitrogen ᎏ inorganic forms. In:
 D.E. Walling et al. r The Science of the Total En¨ ironment 266 (2001) 113᎐123
Page AL, Miller RH, Keeney DR, editors. Methods of soil
analysis, part 2. Chemical and microbiological properties
ᎏ agronomy, vol 9. 2nd ed, 1982:1982.
Kronvang B. The export of particulate matter, particulate
phosphorus and dissolved phosphorus from two agricultural
river basins: implications on estimating the non-point phos-
phorus load. Water Res 1992;26:1347᎐1358.
Meybeck M. Les fleuves et le cycle geochimique des ´ ´
elements.
` d’Etat, Universite´ Pierre et Marie Curie, Paris, 1984.
These
Olsen SR, Sommers LE. Phosphorus. In: Page AL, Miller RH,
Keeney DR, editors. Methods of soil analysis, part 2.
Chemical and microbiological properties ᎏ agronomy, vol.
9. 2nd ed, 1982.
123
Russell MA, Walling DE, Webb BW, Bearne R. The composi-
tion of nutrient fluxes from contrasting UK river basins.
Hydrol Process 1998;12:1461᎐1482.
Schwertmann U. Differenzierung der eisenoxide des bodens
durch photochemische extraction mit sauer ammoniumox-
¨
alat-losung. ¨ Bodenkd 1964;105:197᎐202.
Z Planzenernahr
Walling DE, Webb BW, Russell MA. Sediment-associated
nutrient transport in UK rivers. In Freshwater contamina-
tion ŽProc Rabat Symp, April᎐May 1997. Int Assoc Hydrol
Sci 1997; IAHS Publ no 243: 69᎐81.
Williams JDH, Jaquet JM, Thomas RL. Forms of phosphorus
in the surficial sediments of Lake Erie. J Fish Res Board
Can 1976;33:413᎐429.