Paper presented at the COST action on Phosphorus meeting. Cordoba, Spain. 13-15 May 1999.
John N. Quinton.
Cranfield University, Silsoe, Bedford MK45 4DT, United Kingdom.
Email: J.Quinton@Cranfield.ac.uk
. Fax: +44 (0)1525 863 300
Paper presented at the COST action on Phosphorus meeting,
Cordoba, Spain. 13-15 May 1999.
Introduction
Given that much of the soil phosphorus is associated with soil surfaces, it is reasonable to presume that soil erosion is likely to be an important mechanism for transporting phosphorus from agricultural fields to the aquatic environment. The work of Catt et al. (1998) supports this. In a comparative study of phosphorus transport routes they showed that in the
Brimstone field experiment, located on a clay soil, losses of phosphorus in soil drainage were between 0.37 and 0.91 kg ha -1 , compared with phosphorus losses of between 0.03 and 1.22 kg ha -1 in surface runoff. Losses in surface runoff reported from the Woburn Erosion
Reference Experiment by the same authors are somewhat higher, at between 0.8 and 18.7 kg ha -1 for a four-year arable rotation.
In this short paper I will give a brief review of soil detachment and transport processes (the term soil erosion is not used to avoid confusion). Then discuss how these processes will influence the transport of P and finally consider the prospects for modelling in this area.
Soil detachment and transport processes
Soil detachment and transport processes are well described in the literature, and reviews can be found in a number of standard texts on the subject e.g. Morgan (1995) and Hudson
(1995).
Soil detachment can be viewed as a number of interacting processes occurring within a landscape. These processes are linked spatially, by flows of energy and material and vary temporally due to changes in rainfall and soil topographical, hydrological and chemical properties. Figure 1 attempts to illustrate these relationships. The key detachment processes of splash (raindrop) detachment and flow detachment are identified and those parameters and variables that influence them are indicated. Figure 1 could be viewed as representing a
- 1 -
Detachment and transport of particle-bound P: processes and prospects for modelling John N. Quinton

Paper presented at the COST action on Phosphorus meeting. Cordoba, Spain. 13-15 May 1999. - 2 - soil erosion model at a point in the landscape, however, such points will be linked to others which they will also influence.
Figure 1, Flow chart demonstrating the influence of of key factors on detachment and transport processes (italic).
The processes of soil detachment and transport have been describe in terms of mathematical equations by a number of authors (e.g. Styczen. and Nielsen, 1989; Poesen,
1985). For splash detachment the most common approach is to relate the amount of splashed material to the kinetic energy of the rainfall. One such relationship is that used by the European Soil Erosion Model (EUROSEM) (Morgan et al., 1998). In EUROSEM raindrop or splash detachment (DR; m3 s-1 m-1 ) is described using the relationship:
DR = k

KE e
 zh
(1) where k is an index of the detachability of the soil (g J  1 ) for which values must be obtained experimentally,  s
is the particle density (kg m -3 ),KE is the total kinetic energy of the net rainfall at the ground surface (J m  2 ),z is an exponent varying between 0.9 and 3.1, depending on soil, and h is the mean depth of the surface water layer (mm).
Equation 1 illustrates the importance of interactions with other components of the erosion process. The kinetic energy of the rainfall is influenced by the rainfall characteristics: larger drops have larger kinetic energies, whilst smaller drops have lower energies. Rainfall characteristics are altered when raindrops hit the plant canopy: some will shatter producing smaller drops, while some will coaless on leaves before falling to the ground as large drops.
Brandt (1989) has shown that, in forests, large drops from leaves may be significant sources of splash detachment. Splash detachment is also affected by the depth of water on the soil surface, which is linked to the soil’s infiltration properties and surface microtopography.
Detachment and transport of particle-bound P: processes and prospects for modelling John N. Quinton

Paper presented at the COST action on Phosphorus meeting. Cordoba, Spain. 13-15 May 1999.
The description of flow detachment and deposition in current process based soil erosion models is more variable. Some models such as WEPP (Nearing et al., 1989) describe both detachment and deposition separately. Others such as EUROSEM (Morgan et al., 1998) and
GUEST (Rose et al., 1983) describe detachment and deposition of material as a continuous interchange of material with the current position of balance giving the ambient sediment concentration. According to the theory used in EUROSEM the flow detachment rate (DF: m 3 s –1 m –1 ) is given by:
DF =

w v s
(C m

C) (2)
- 3 - where  is a flow detachment efficiency coefficient, which becomes equal to unity when material is being deposited, V s
the settling velocity of particles m s -1 , C the actual sediment concentration, w is the width of flow (m) and C m
the equilibrium sediment concentration.
Detachment by flow will be highly dependent upon the velocity of the overland flow, this is in turn dependent upon the roughness of the soil surface over which the water flows, which is affected by the splash detachment. The surface characteristics also influence the amount of flow detachment through influencing the resistance to the shear forces imparted on it by the overland flow. Feed backs also exist here due to changes in the surface properties caused by both splash and flow detachment.
It is worth noting that at present no process-based soil erosion model takes the effect of soil chemical properties into account when describing soil detachment.
Interaction between erosion processes and P transport
Current soil erosion models have concentrated on simulating, bulk soil losses. However, it is well known that amounts of phosphorus transported by surface runoff are often greater then the concentration of P in the bulk soil, because phosphorus is associated primarily with the finer fractions of the soil (Syers & Walker, 1969) and water erosion is a selective process. It is also generally accepted that the selectivity decreases as the magnitude of an erosion event increases (Massey and Jackson, 1952; Quinton and Catt, in prep). This is illustrated in figure
2. Which, using data from the Woburn Erosion Reference Experiment (Catt et al, 1994), shows P concentrations declining as soil loss increases. Interestingly the variability of P concentrations found in the sediment also decreases with increasing P loss. It is hypothesised that this is due to the variability in the intensity of the erosion events for smaller magnitude soil losses: an event may be short duration and intense, or long duration and of low intensity, and give the same soil loss.
Detachment and transport of particle-bound P: processes and prospects for modelling John N. Quinton

Paper presented at the COST action on Phosphorus meeting. Cordoba, Spain. 13-15 May 1999.
Due to their higher selectivity small erosion events contribute a disproportionate amount of P to the total P lost form an area. This is illustrated in figure 3, which shows that, when ranked in order of magnitude, events contributing 50% of the soil loss through relatively small amounts of sediment contribute over 75% of the total P loss.
125
6000
5000
4000
3000
2000
1000
0
0 50 100 150 200 250 300
Soil loss (kg)
Figure 2, Effect of event magnitude on P concentration.
100
75
50
25
0
0 25 50 75 100
Percent of total soil loss
Figure 3, Relationship between ranked soil loss and the percent of total P lost from plot 7 of the Woburn Erosion
Reference Experiment.
- 4 -
Prospects for modelling
Theoretically, selectivity will operate in two components: detachment and transport. There is little doubt that the finer fractions of the soil, once in suspension, will travel the greatest distances. The clay-sized fractions remain in suspension even in standing water due to their colloidal nature. This leads to the enrichment of finer materials as the flow's distance of travel increases. However, whilst in theory selective detachment also seems likely, there has been little experimental work done on the selectivity erosion of different size classes from soils.
Work has instead tended to concentrate on the selective detachment and transport of sediments from non-cohesive sources with a narrow particle size range, for example Poesen and Savat’s (1980) study of particle size separation by splash and runoff, Govers’ (1990) work on shallow flow transport capacities and Poesen’ (1985) study of splash detachment. There are some exceptions, notably the work of Torri and Sfalanga (1986) who considered the selective detachment of material from clay aggregates and found enrichment in the 0.063 to
0.5mm size range.
Present models which simulate P loss by erosion fall into two categories: those which use and empirical equation to calculate a P enrichment ratio or those which attempt to model the selectivity process explicitly.
The empirical approach utilises equations based on the early work of Massey and Jackson
(1952). In their equation P enrichment ratio is related to soil loss: ln(ER) = a + b ln(sed) (3)
Detachment and transport of particle-bound P: processes and prospects for modelling John N. Quinton

Paper presented at the COST action on Phosphorus meeting. Cordoba, Spain. 13-15 May 1999. - 5 - in which ER is the sediment discharge (kg ha -1 ) and a and b are constants which depend upon soil type and cropping practices. Equation 3 is easy to apply, but given the data presented in figure 2, there will be obvious problems for lower magnitude events.
The process-based approach, which is based on theory proposed by Foster et al (1980) has relied on modelling deposition. It is based on three simple rules:
1.
If the energy of the flow is greater than that require to transport all the entrained material then all size classes are transported;
2.
If the energy of the flow is lower than that required to transport the smallest material, then all size classes are deposited;
3.
If the energy is enough to transport some size classes, but not others then is distributed amongst the particle size classes.
This approach, although operational in CREAMS (Knisel et al 1980) and Opus (Smith 1990) is based on limited experimental data, and virtually no validation work has been attempted.
Conclusion
This paper has briefly reviewed the relationship between detachment and transport processes and the loss of P associated with sediment in overland flow. However, at present we are still learning how to model these processes. If we are to better describe the processes of selectivity then experimental and field studies will be required to test existing theory and develop new ones.
References
Brandt, C.J. 1989. The size distribution of throughfall drops under vegetation canopies.
Catena 16, 507-524.
Catt, J.A., Chambers, B.J., Farina, R. Harris, G.L., Hodgkinson, R., Howse, K.R. & Quinton,
J.N. 1998. Phosphorus losses from arable land in England. Soil Use and Management 14, 168-
174.
Catt, J.A., Quinton, J.N. Rickson, R.J. & Styles, P.D.R. 1994. Nutrient losses and crop yields in the Woburn Erosion Reference Experiment. In. Rickson, R.J. (ed), Conserving soil resources:
European perspectives, CAB International, Wallingford.
Foster, G.R. Lane, L.J., Nowlin, J.D., Laflen, J.M. and R.A. Young. 1980. A model to estimate sediment from field sizex areas. In CREAMS: a field scale model for chemicals, runoff and
Detachment and transport of particle-bound P: processes and prospects for modelling John N. Quinton

Paper presented at the COST action on Phosphorus meeting. Cordoba, Spain. 13-15 May 1999. - 6 - erosion from agricultural management systems. (ed. W. G. Knisel) USDA Conservation
Research Report No. 26.
Govers, G. 1990. Empirical relationships for the transporting capacity of overland flow.
International Association of Hydrological Sciences Publication 189, 45-63.
Hudson, N.W. 1995. Soil Conservation. Batsford, London.
Knisel, W.G. (ed), 1980. CREAMS: a field scale model for chemicals, runoff and erosion from agricultural management systems. USDA Conservation Research Report No. 26.
Massey, H.F. and Jackson, M.L. 1952. Selective erosion of soil fertility constituents. Soil
Science Society of America Proceedings 16, 353-356.
Morgan, R.P.C. 1995. Soil erosion and conservation. Longman Scientific and TechnicalHarlow,
UK.
Poesen, J & J. Savat. 1980. Particle-size separation during erosion by splash and runoff.
P427-439. In M. de Boodt & D. Gabriels (eds), Assessment of Erosion, John Wiley & Sons,
Chichester.
Poesen, J.A. 1985. An improved splash transport model. Zeitschrift für Geomorphologie
29:193-211.
Quinton, J.N. and J.A. Catt. in prep. The selective removal of phosphorus from soil.
Rose, C.W., Williams, J.R., Sander G.C. and Barry, D.A. 1983. A mathematical model of soil erosion and deposition processes. I. Theory for a plane element. Soil Science Society of
America Journal 47, 991-995.
Smith, R.E. 1990. Opus, an integrated simulation model for the transport of nonpoint-source pollutants at the field scale. Volume 1, Documentation. USDA-ARS, ARS-98. 120pp.
Styczen, M. and Nielsen, S.A. 1989. A view of soil erosion theory, process-research and model building: possible interactions and future developments. Quaderni di Scienza del Suolo
2, 27-45.
Syers, J.K. & Walker, T.W. 1969. Fractionation of phosphorus in two cultivated soils and particle size separates. Soil Science 108, 283-289.
Torri, D. and Sfalanga, M. 1986. Some aspects of soil erosion modelling. In Agricultural nonpoint source pollution: model selection and application (eds. A. Giorgini & F. Zingales),
Developments in Environmental modelling 10, Elsevier.
Detachment and transport of particle-bound P: processes and prospects for modelling John N. Quinton