Maesnant (Wales) 1970s-1990s
- 1 Location and Scale
- 2 Dates
- 3 Climate
- 4 Geology
- 5 Topography
- 6 Vegetation / Land Use
- 7 History
- 8 Hydrological Knowledge Gained
- 8.1 Significance of pipeflow for streamflow response
- 8.2 Area and pattern of contributing areas
- 8.3 Complexity of hillslope processes
- 8.4 Hortonian overland flow
- 8.5 Relation to a/s index
- 8.6 Patterns of hillslope hydrological response
- 8.7 Sources of pipe flow
- 8.8 Implications for infiltration measurements
- 8.9 Comparison of hillslope hydrological processes
- 8.10 Simulation model of pipe flow
- 8.11 Impacts on acidification of streamflow
- 8.12 Solute transport
- 8.13 Hillslope evolution
- 8.14 Stream channel initiation
- 8.15 Sediment yields
- 8.16 Piping impacts on micro-topography, soils and vegetation
- 9 Anecdotes
- 10 Reference Material
- 11 Acknowledgments
Location and Scale
The catchment is located on the western, seaward slope of Plynlimon (Pumlumon Fawr), the highest peak in the Cambrian Mountains of mid-Wales, UK. The catchment covers an area of 0.54 km2. The long axis of the catchment is 2.75 km. The Maesnant stream is a headwater tributary of the River Rheidol, and drains from SE to NW into the Nant y Moch Reservoir approximately 18 km east of Cardigan Bay and the Irish Sea. It falls mainly within Ordnance Survey grid square SN 7887.
1970s - 1990s, run as a research catchment by the Department of Geography and Earth Sciences, Aberystwyth University
Annual rainfall measured during the 1980s averaged 2595 mm, well-distributed throughout the year but with a tendency to a slight peak in early winter. Water losses due to evapotranspiration are relatively low, at around 25%. The climate is Cfb in the Köppen-Geiger classification: temperate, warm summer without a dry season.
The catchment is underlain by Ordovician greywacke, mudstone and grits with partial coverings of soliflucted clayey and stony material forming a 4-5 m high solifluction terrace flanking the stream. Blanket peat up to 1-2 m deep covers some of this material on the lower slopes and extends onto bedrock in mid-slope. Scree slopes occur on parts of the upper slopes.
The catchment falls from 752 m O.D. at the summit of Plynlimon to 465 m O.D. at the Welsh Water sharp-crested weir. Soils are generally poorly developed and belong to the Hiraethog Association (Rudeforth, 1970). These range from rankers on the higher slopes (Powys Series), through peaty and peaty gley podzols of the Hiraethog Series on the lower slopes.
Vegetation / Land Use
The catchment is covered by moorland vegetation and is part of the Crown Estates with grazing rights leased for sheep pasture. Dry grassland predominates. The vegetation grades from bare rock near the summit, through mixed grassland heath and dwarf shrub heath to wet grassland and Juncus and bog mosses on the lower slopes. Vegetation on the lower slopes exhibits marked banding which reflects the drainage lines typically associated with soil pipes. Juncus and Sphagnum dominate the seepage lines and areas of groundwater exfiltration. Blanket bog and wet heath occupy the gentler slopes (c. 4o) on terraces, while acid grassland dominates the steeper slopes (over 12o). Areas drained directly by perennially-flowing soil pipes are marked by linear extension of the acid grassland association running between communities dominated by Vaccinium, Empetrum heath and Eriophorum vaginatum. The zone of ephemerally-flowing pipes is marked by a belt of grassland dominated by Nardus stricta flanked by non-piped areas of wet heath comprising mainly E. vaginatum and Calluna.
The main focus of research has been on the hydrological role of soil pipes, running from 1974 to 1993. This included studies of the sources of pipeflow, the contribution of pipeflow to streamflow, the effects of pipeflow on the throughput of acid rain and its contribution to stream acidification, and the contribution of pipe erosion to sediment movement. This research necessitated the design of new instruments to monitor and to log pipeflow.
Most of this research was funded by NERC Research Grants and studentships, with smaller supplements from University of Wales studentships/awards and the Welsh Acid Rain Project.
The first research at Maesnant was conducted in the early 1970s by a team under Prof John Lewin on sediment sources (Lewin, Cryer and Harrison, 1974; Cryer, 1978, 1980). Pipeflow studies began here in 1974 (Jones, 1975). Focus then shifted briefly to snowfall and snowcover (1975-1978), studying the effects of topography on snow accumulation and snowcover persistence as part of a Wales-wide programme (Waring and Jones, 1980; Waring, 1981; Jones and Taylor, 1984). Work at Maesnant included designing a new type of combined snowfall, rainfall and snowmelt gauge, which was later deployed more widely in the Welsh Acid Rain Project in the 1980s.
After appeals for more research on pipeflow from Newson (1976), working in the adjacent Institute of Hydrology catchments on the east side of Plynlimon, and Jones (1978; 1979), the Natural Environment Research Council awarded a Research Grant for research at Maesnant (1979-1982). This led to the establishment of 17 automatic pipeflow measuring sites arranged in three banks (at the top, middle and outfall of the pipe network) and 3 riparian seepage zones. In 1979 there were no suitable, commercially-available data loggers and we had to design our own, initially on reel-to-reel and later on cassette (Jones, Lawton, Wareing and Crane, 1984). This research was followed in the late 1980s by NERC funding for the study of the effects of pipeflow on stream acidification (Jones and Hyett, 1987; Hyett, 1990). Fieldwork in the early 1990s included a basinwide study of the controls on the development and the location of soil pipes at local and national scales, with collaboration at Maesnant from the late Prof Peter Wathern of the then Department of Biological Sciences (Richardson, 1991; Jones, Wathern, Connelly and Richardson, 1991). The final fieldwork programme focussed on the main pipe system in order to design a process-based model for pipeflow (Connelly, 1993; Jones and Connelly, 2002).
In 1984, the Institute of Hydrology (IH) began to use the Maesnant basin as a control catchment in the Nant-y-Moch grassland improvement study (Roberts, Hudson and Roberts, 1990). IH installed a modern data logging system and an improved compound sharp-crested weir, together with an automatic multiple water quality sampling system. Aberystwyth University collaborated in servicing the loggers and water sampler and shared the data as part of the pipeflow water quality project.
In 1997 the Geological Conservation Review listed Maesnant for SSSI status based on the work done on the pipe networks (Gregory, 1997).
Hydrological Knowledge Gained
The presence of piping has major implications for the application of established theories of streamflow generation and hillslope drainage, erosion and evolution (Jones, 1979; 1987), as well as for hillslope management (Jones, 2004; Jones and Cottrell, 2007).
Significance of pipeflow for streamflow response
The pipe networks on Maesnant make a major contribution to stream runoff, averaging 49% of stream stormflow. The Maesnant pipeflow data are still the most extensive available and are largely corroborated by other monitoring programmes around the world (Jones, 2010b).
Area and pattern of contributing areas
the concept of dynamic contributing areas (DCA) defined by saturation overland flow (SOF) needs modification here as the DCA is defined mostly by subsurface flow through the pipe network (Jones, 1979). The pipe networks nearly double the stormflow contributing area and extend it in ‘fingers’ well beyond the riparian zone: maximum 750m away from the stream. (If all the discharge of the pipes were theoretically converted into a uniform riparian contributing area along the stream, it would extend no more than 70m from the streambank. This has important implications for moorland management, e.g. liming strategy and afforestation (cp. Gee and Stoner, 1989; Jones and Cottrell, 2007).)
Complexity of hillslope processes
Hillslope drainage processes are extremely complex with frequent interchange between overland flow and pipeflow (Jones, 1987a, fig 5) - a feature not yet recognised in standard models of hillslope hydrology.
Hortonian overland flow
Measurements of rainfall intensity, infiltration capacity and hydraulic conductivity indicate that infiltration-excess overland flow should be common, but no Hortonian overland flow was ever seen: (a) the majority of storms had mean rainfall intensities that exceed mean infiltration capacities and all rainstorms had peak 15-minute intensities exceeding them; (b) mapping the infiltration capacities suggests that 50% of storms should generate Hortonian overland flow over 75% of the hillside, but this does not occur; (c) infiltration capacities and hydraulic conductivities are higher above pipes suggesting more drying of the upper soil horizons and macropore development due to pipe drainage.
Relation to a/s index
Pipes do not always follow the a/s index for predicting patterns of soil water drainage. (The index is used in some process-based runoff models and used to guide hillslope management, e.g. TOPMODEL (cp. Gee and Stoner, 1989; Kirkby, 1998).)
Patterns of hillslope hydrological response
Monitoring of the 750m long hillslope shows for the first time the progress of the start and peak of pipeflow as it progresses across the hillslope.
(1) This shows average peak lag times of c. 2 hours in the ephemeral pipes, c. 3h at the head of the perennials and 4-5h when it reaches the streambank (Jones, 2010b, fig. 4: 1997b, fig. 5.5).
(2) Initial stormflow response tends to begin at the base of the hillslope with the perennial pipes and progress uphill.
(3) Response times vary according to the specific combinations of antecedent rainfall and storm rainfall (Jones, 1988, figs. 4 to 6).
(4) The timing of the contributions of individual pipes to the stream hydrograph is highly variable, but most pipes start contributing discharge on the rising limb of the stream hydrograph and peak at or before streamflow peaks (Jones, 1987a).
Sources of pipe flow
(1) Perennial pipes have a baseflow fed by groundwater. Banks of piezometers set across the shallow swales formed by pipe erosion show hydraulic drawdown along the pipes and also indicate a wave of groundwater paralleling the rise and fall of pipe discharge as it progresses downslope during storm runoff (Jones and Connelly, 2002).
(2) Ephemeral pipes on the upper slopes running through shallow soils (c. 150mm deep) do not intersect a permanent phreatic surface and are fed by a combination of direct inflow from surface macropores and a temporary subsurface source fed by infiltration across the hillside around the pipes.
(3) Because of the lack of baseflow in the ephemeral pipes, they respond less frequently: approximately once for every three storm responses in the perennial pipes.
Implications for infiltration measurements
The latter result suggests: (a) rainwater is infiltrating the soil through cracks, blowholes and other macropores which are not sampled by standard double-ring infiltrometers; and (b) this means that a larger “minimum representative area” of hillslope is needed to characterise drainage processes in the presence of piping, paralleling the concept of the “minimum representative volume” proposed by Beven and Germann (1982) (Jones, 1990).
Comparison of hillslope hydrological processes
When peak discharge rates and peak lag times for the pipes were plotted against the area drained by the pipes, the patterns fit neatly in-between diffuse throughflow and saturation overland flow, as plotted by Anderson and Burt (1990). The range of contributing area amongst the 17 Maesnant pipes is naturally far smaller than the USDA data used by the latter, nevertheless, the limited Maesnant data suggest similar trends: high peak lag times with increasing size of contributing area and peak runoff rates decreasing with drainage area (Jones, 1997a; 2010b).
Simulation model of pipe flow
The pipeflow data have been used to develop the first physically-based semi-distributed simulation model of pipeflow generation based on field measurements (Connelly, 1993; Jones and Connelly, 2002). (Nieber and Warner’s (1991) model was based on Darcian theory treating the pipe as a porous medium rather than actual field observations (Jones (2010b)).
Impacts on acidification of streamflow
Pipeflow acidity and yields of dissolved aluminium affect streamflow quality.
(1) Piping exacerbates problems of acid rain runoff by reducing catchment buffering effects as a result of (a) more rapid transmission and reduced residence times, and (b) directing throughflow through the upper, organic horizons and reducing contact between stormwater and weathering mineral surfaces. The drainage and drying out of large sections of hillside also encourages the release of sulphates and organic acids from peaty soils. Although the mean streamwater pH of 5.16 indicates catchment buffering of rainfall at pH 4.84, this buffering is being limited by the pipe system in which the average pH of perennial pipe outfalls is 4.48-4.90, and for the ephemeral outfalls 4.10-4.26 (Hyett, 1990; Jones, 1997d).
(2) Aluminium mobility: Monitoring shows a steady increase in concentrations of dissolved aluminium from the head to the outfall of pipes matching the gradual reduction in pH (Hyett, 1990; Jones, 1997d). (This is potentially more destructive of stream fauna than the acid flush itself.)
Basinwide surveys of electrical conductivity suggest that pipes are redistributing soil water solutes in an organised pattern:
(1) ephemerally-flowing pipes deplete solutes on the upper slopes, transporting solutes downslope and enhancing electrical conductivity towards the bottom of the ephemeral pipe system.
(2) solute levels peak again in the zone of groundwater resurgence at the head of the perennially-flowing pipes and remain higher along the line of pipes than in the surrounding, unpiped areas (Jones et. al., 1991; Jones, 1997d).
Kirkby’s (1978) model of hillslope evolution assumes that subsurface flow is slow and erosion is predominantly solutional. However:
(1) pipeflow velocities measured on Maesnant indicate rates of 0.1 to 1.0 m s-1; higher than overland flow and two orders of magnitude more than diffuse throughflow ((Jones, 1988) and pipeflow sediment traps show c.15% of catchment sediment yield coming from pipes;
(2) Kirkby’s conclusion that erosion by subsurface flow is more solutional than mechanical (loc. cit.) is not born out by Maesnant data (Jones, 1987a; cp. point 12). Cryer (1980) found specific conductivity of flow in perennial pipes was under half that of saturation overland flow;
(3) the pattern of flow down the hillside is more complex, with frequent interchange between surface and subsurface flow;
(4) results suggest that the traditional concept of drainage density needs to be re-evaluated in light of channelised subsurface flows. Jones (1987a) notes that Kirkby’s (1978) formula for calculating drainage density based on rainfall and soil storage capacity grossly overestimates channel density on Maesnant at 5.6 km km-2 against an actual density of just 2, whereas the pipes have densities of 153 (ephemerals) and 63 km km-2 (perennials). This suggests that by diverting surface flows before the critical distance of erosion is reached the pipes may be reducing classical drainage density.
(5) including pipes in the calculation as substitute channels yields a drainage density closer to that predicted by Kirkby’s formula. By considering the relative discharge rates of the pipes and the stream, Jones (1990) calculated an “effective stream density” of 3.1 km km-2.
Stream channel initiation
Further evidence is found of stream channel initiation by unroofing of ephemeral pipes at the head of the northern branch of Maesnant stream (Jones, 1997d). This is still an undervalued source of channel development (cp. Jones, 1971; 1981; 1987b).
Stream sediments are mainly derived from streambank scars (Lewin et al., 1974), but bedload traps on a few of the pipes suggest that pipe erosion accounts for c.15% of the annual sediment yield of the catchment (Jones and Crane, 1984; Jones, 1987a).
Piping impacts on micro-topography, soils and vegetation
Pipe erosion has created linear micro-topography with swales up to 500mm deep, which increase the spatial variability of drainage but in an organised pattern that translates into patterns of soil and vegetation: oligo-amorphous peat soil in the perennial pipe swales running between deeper peat soils; patches of stagnopodzols supporting dry grassland around the ephemeral pipes interrupting the broader expanse of peat supporting mixed grassland heath (Jones, 1997d). Measurements show an annual yield of 200 kg per year in the largest system and it is estimated that the main pipe system is capable of excavating 1 m3 every 7 years (Jones, 1987a).
1. Summer 1982. The 3-year pipeflow study is fully-developed and the basin is hosting the IGU Commission on Field Experiments in Geomorphology.
A coach has been hired to transport the 30 or so international academics, some of advanced years, up to the catchment. The main access to the site lies 2.5 km from the public road up an unmetalled track that leads to Welsh Water’s Llyn Llygad Rheidol water supply reservoir (a corrie lake, literally ‘The Rheidol’s Eye Lake’). We manage to persuade the coach driver to attempt to drive up the track, after giving our assurance that there is room to turn around at the weir, even though he can see that the way is steep and deeply rutted, climbing 50m in the first 500m. What the driver cannot see awaiting round the first bend is moss. Halfway up the slope the rear wheels begin to spin, progress stalls and the vehicle starts slowly sliding sideways off the track. We persuade the driver to allow us to instruct everyone to make for the back of the coach. The extra weight bares fruit, the wheels gain some traction, and seemingly for an age we move very slowly upslope, till with a jolt the tyres find stony ground again, and the passengers return to their seats, a little shaken but still fit and eager to climb the catchment.
A great new use for 30 academics! We hosted numerous subsequent fieldtrips, for groups from Cambridge, Exeter, East Anglia, Southampton and Gröningen as well as our own students, but we never managed to persuade another coach driver to brave the final track.
2. In the early ‘80s we managed to persuade a student hang-glider to jump off the top of Plynlimon with an anti-vibration camera mount. The picture to the right was taken shortly before crash-landing. The lines of pipes are inked in. The weir and black recorder box on the largest perennial pipe stand by the scar left by the excavated outfall. Dye tracing indicated a delay of less than 10 minutes between the recorder and the stream bank.
- Jones, J.A.A. 1997a: Pipeflow contributing areas and runoff response. Hydrological Processes 11(1): 35-41.
- Jones, J.A.A. 1997b: The role of natural pipeflow in dynamic contributing areas and hillslope erosion: extrapolating from the Maesnant data. Physics and Chemistry of the Earth 22(3-4): 303-308.
- Jones, J.A.A. 1997c: Maesnant, Pumlumon (Plynlimon), Wales. Section in Part 3, Fluvial Geomorphology of Wales. In: K.J. Gregory (Ed.) Fluvial Geomorphology of Great Britain, Joint Nature Conservation Committee, Geological Conservation Review Series, Chapman & Hall, London, 165-167.
- Jones, J.A.A. 1987a: The effects of soil piping on contributing areas and erosion patterns, Earth Surface Processes and Landforms 12(3): 229-48.
- Jones, J.A.A. 1986a: Some limitations to the a/s index for predicting basin-wide patterns of soil water drainage. Zeitschrift für Geomorphologie, Supplementband 60: 7-20.
- Jones, J.A.A. and F.G. Crane 1984: Pipeflow and pipe erosion in the Maesnant experimental catchment. In: T.P. Burt and D.E. Walling (eds) Catchment experiments in fluvial geomorphology, GeoBooks, Norwich, 55-72.
- Jones, J.A.A., M. Lawton, D.P. Waring, and F.G. Crane 1984: An Economical Data Logging System for Field Experiments. British Geomorphological Research Group Technical Bulletin No. 31, GeoAbstracts, Norwich, 32pp.
- Jones, J.A.A. 1982a: Experimental studies of pipe hydrology. In: R.B. Bryan and A. Yair (eds) Badlands geomorphology and piping, GeoBooks, Norwich, 355-370.
- Jones, J.A.A. and F.G. Crane 1982b: New evidence for rapid interflow contributions to the streamflow hydrograph. Beiträge zur Hydrologie, Sonderheft 3: 219-232.
- Jones, J.A.A. 1979: Extending the Hewlett model of stream runoff generation. Area 11(2): 110-114.
- Jones, J.A.A. 1978: Soil pipe networks: distribution and discharge. Cambria 5(1): 1-21.
- Jones, J.A.A. 1975: Soil piping and the subsurface initiation of stream channel networks. Unpub. PhD thesis, University of Cambridge, 467pp.
Controls on pipe development
- Jones J.A.A., J.M. Richardson and H.J. Jacobs 1997: Factors controlling the distribution of piping in Britain: a reconnaissance. Geomorphology 20 (3-4): 289-306.
- Richardson, J.M. 1992: Catchment characteristics and the distribution of natural soil piping. Unpub MPil thesis, University of Wales Aberystwyth, 276pp.
- Jones, J.A.A. and L.J. Connelly 2002: A semi-distributed simulation model for natural pipeflow. Journal of Hydrology 262(1-4): 28-49.
- Connelly, L.J. 1993: Modelling natural pipeflow contributions to the streamflow hydrograph. Unpub. PhD thesis, University of Wales Aberystwyth.
- Jones, J.A.A. 1988: Modelling pipeflow contributions to stream runoff. Hydrological Processes 2: 1-17.
Impact of pipe networks on upland environments and stream acidification
- Jones, J.A.A. 2002: Pipeflow, water quality and catchment management in the British uplands. In: C. Cunnane (Ed.) Celtic water in a European framework: pointing the way to quality, National University of Ireland, Galway: 98-114.
- Jones, J.A.A., P. Wathern, L.J. Connelly & J.M. Richardson 1991: Modelling flow in natural soil pipes and its impact on plant ecology in mountain wetlands. In: P. Nachtnebel (Ed.) Hydrological basis of ecologically sound management of soil and groundwater, International Association of Hydrological Sciences Publication No. 202: 131-142.
- Hyett, G.A. 1990: The effects of accelerated throughflow on the water yield chemistry under polluted rainfall. Unpub. PhD thesis, University of Wales Aberystwyth.
- Jones, J.A.A. and G.A. Hyett 1987: The effect of natural pipeflow solutes on the quality of upland streamwater in Wales. Abstract HW8-18, 19th General Assembly of the International Union of Geodesy and Geophysics, Vancouver, Canada, volume 3: 998.
- Jones, J.A.A. 1986b: Acid rain and the Welsh environment. Planet 54: 51-59.
- Cryer, R. 1980: The chemical quality of some pipeflow waters in upland Wales and its implications. Cambria 6(2): 1-19.
- Cryer, R. 1978: A study of the sources and variation of major solutes in selected mid-Wales catchments. Unpub. PhD thesis, University of Wales Aberystwyth.
Colour maps of the impacts on soils and vegetation are published in:
- Jones, J.A.A. 2011: Groundwater in peril. In: J.A.A. Jones (Ed.) Sustaining Groundwater Resources, Springer, Dordrecht, 1-19. ISBN 978-90-481-3425-0.
- Jones, J.A.A. 2010a: Water Sustainability: a global perspective. Routledge, London, 452pp. ISBN 978 1 444 10488 2.
Pipeflow and pipe erosion: the wider context
(including summaries of research from Maesnant)
- Jones, J.A.A. 2010b: Soil piping and catchment response. Hydrological Processes 24: 1548-1566.
- Jones, J.A.A. 2004: Implications of natural soil piping for basin management in the British uplands. Land Degradation and Development 15(3): 325-349.
- Bryan, R.B. and J.A.A. Jones 1997: The significance of soil piping processes: inventory and prospect. Geomorphology 20 (3-4): 209-218.
- Jones, J.A.A. 1997d: Subsurface flow and subsurface erosion: further evidence on forms and controls. In: D.R. Stoddart (Ed.) Process and form in geomorphology, Festschrift for R.J. Chorley, Routledge, London, 74-120.
- Jones, J.A.A. 1994: Soil piping and its hydrogeomorphic function. Cuaternario y Geomorfología 8(3-4): 77-102.
- Jones, J.A.A. 1990: Piping effects in humid lands. In: C.G. Higgins and D.R. Coates (eds) Groundwater geomorphology: the role of subsurface water in earth-surface processes and landforms, Geological Society of America, Special Paper 252: 111-138.
- Jones, J.A.A. 1989: Bank erosion: a review of British research. In: M.A. Ports (Ed.) Hydraulic Engineering, American Society of Civil Engineers, New York, 283-288.
- Jones, J.A.A. 1987b: The initiation of natural drainage networks. Progress in Physical Geography 11(2): 207-45.
- Jones, J.A.A. 1985: Erosion by pipeflow. In: D. Balteanu, S. Dragmirescu and C. Muica (eds) Geomorphological Research for Land and Water Management, Romanian Academy of Science and University of Bucharest, Bucharest, 123-148. (In Romanian.)
- Jones, J.A.A. 1981: The Nature of Soil Piping: a review of research. GeoBooks, Norwich, 301pp. ISBN 0-86094-077-2.
- Lewin, J., R. Cryer and D.I. Harrison 1974: Sources of sediments and solutes in mid-Wales. In: K.J. Gregory and D.E. Walling (eds) Fluvial Processes in Instrumented Catchments, British Geomorphological Research Group Special Publication 6:73-85.
Snowfall and snow cover
- Waring, E.A. and J.A.A. Jones 1980: A snowmelt and water equivalent gauge for British conditions. Hydrological Sciences Bulletin 25(2): 129-134.
- Waring, E.A. 1981: Assessment of snowfall receipts in Wales. Unpub. PhD thesis, University of Wales Aberystwyth.
- Jones, J.A.A. and J.A. Taylor 1984: The climate of Wales. In: H. Carter and H.M. Griffiths (eds) National Atlas of Wales, University of Wales Press, Cardiff, section 1.4.
- Anderson, M.G. and T.P. Burt 1990: Subsurface runoff. In: M.G. Anderson and T.P. Burt (eds) Process studies in hydrology, Wiley, Chichester, 365-400.
- Beven, K.J. and P. Germann 1982: Macropores and water flow in soils. Water Resources Research 18(5): 1311-1325.
- Gee, A.S. and J.H. Stoner 1989: A review of the causes and effects of acidification of surface waters in Wales and potential mitigating techniques. Archives of Environmental Contamination and Toxicology 18: 121-130.
- Gregory, K.J. (Ed.) 1997: Fluvial Geomorphology of Great Britain, Joint Nature Conservation Committee, Geological Conservation Review Series, Chapman & Hall, London, 347pp. ISBN 0-412-78930-2.
- Jones, J.A.A. 1987b: The initiation of natural drainage networks. Progress in Physical Geography 11(2): 207-245.
- Jones, J.A.A. 1971: Soil piping and stream channel initiation. Water Resources Research 7(3): 602-610.
- Jones, J.A.A. and C.I. Cottrell 2007: Long-term changes in streambank soil pipes and the effects of afforestation. Journal of Geophysical Research 112: F01010, doi. F10.1029/2006JF000509, 11pp.
- Kirkby, M.J. 1998: TOPMODEL: a personal view. Hydrological Processes 11(9): 1087-1097.
- Newson, M.D. 1976: Soil piping in upland Wales: a call for more information. Cambria 1: 33-39.
- Nieber, J.L. and G.S. Warner 1991: Soil pipe contribution to steady subsurface stormflow. Hydrological Processes 5(4): 329-344.
- Roberts, A.M., J.A. Hudson and G. Roberts 1990: Nant-y-Moch grassland improvement study. Report No.104, Institute of Hydrology, Wallingford, 85pp (Unpublished).
- Rudeforth, C.C. 1970: Soils of North Cardiganshire. Memoirs of the Soil Survey of England and Wales, Harpenden, 153pp.
We wish to thank the Crown Estates Commissioners and the tenant farmer for permission to instrument the catchment, the Institute of Hydrology and Welsh Water for building and enhancing the weirs, technicians in the Departments of Geography and Physics for designing and building the instruments, Lindsay Collin and Neil Chisholm for supervising the topographic surveying, and the Natural Environment Research Council and University of Wales for funding the research. We also thank our many keen students who worked in the basin, especially our postgrads from Rick Cryer and David Harrison to Anne Waring, Glyn Hyett, Mark Richardson and Liam Connelly, not forgetting postdoc research assistant Francis Crane, who helped set the groundwork for the pipeflow studies.