Note on the Movement of Waste on Screes in the Orongo-rongo District, near Wellington.
[Read before Wellington Branch, September 27, 1944: received by the Editor, September 28, 1944; issued separately, March, 1945.]
Observations and measurements of banded screes in this district indicate that stone stripes are produced by the removal of fine waste by parallel, subsurface streams. A comparison is drawn between erosion on the screes and erosion in arid and semi-arid climates on the basis that in both cases there is a great excess of loose waste material, so that the flowing water acts as a transporting agent only, and does not directly shape the landscape. A further example of a land form shaped by weathering of the rock and the transportation of the resulting waste by water is pointed in the case of convex hills.
On the hills forming the eastern side of the Orongorongo River Valley numerous landslides have occurred. The scars of many are still visible from Wellington. Wherever the slump scarp has exposed the shattered greywacke of which a large part of the top of the hills is composed, long and wide screes have developed. Some are about 300 m. in vertical height and about 0.75 km. in width. In other places, landslides have exposed fault pug and brecciated rock in which extensive gullying has taken place, with the consequent development beneath the scarp of fans of rubble and clay.
The screes in the Wellington waterworks area have been studied by the writer during the years 1941–44. It has been found that the screes typically show marked, longitudinal striping. Bands of coarse boulders, averaging roughly 0.3 m. in diameter, alternate with bands of smaller stones, averaging roughly 0.1 m. in diameter. The bands on any one scree may be approximately the same width or the bands of larger stones may be narrower than the bands of smaller stones. The bands are from 1 to 3 m. wide. The smaller screes may present a flat surface from side to side but the larger screes tend to be domed slightly between a point on one side and a point of the same altitude on the opposite side. In places the bands of larger stones lie in perceptible dips, forming very shallow, parallel gullies.
Between the mouth of the Orongorongo River and the mouth of the Mukumuku there are a number of small shingle slips on the slopes rising from the coast. These slips probably owe their origin, in part, to the weakening or destruction of the vegetative cover by burning and overgrazing and the subsequent exposure of the shattered greywacke, which very readily forms screes. Here also is shown the same banding of large and small stones. One small, shallow scree which runs through a grove of isolated trees was seen in October, 1941, with driftwood scattered down the stripes of large stones.
In the same district are a number of boulder fans, some about 0.5 km. wide, formed on the narrow level strip of marine terrace between sea and hill, by small streams that flow in steep, narrow gullies down the hill. These fans show a network made of zones of large stones surrounding roughly polygonal centres of smaller stones. The fans are roughly half-cone shaped, but in most places the larger stones are slightly below the surrounding surface, when they appear similar to braided water channels. One fan is conspicuous in that it has become covered with sward except for the anastomosing lines of large stones.
It has been suggested that similar segregation of stones observed in other parts of the world (Sharpe, 1938) and in New Zealand (Zotov, 1940) is due to the action of frost. However, the following observations on the Orongorongo screes do not agree with this frost theory.
Very low temperatures such as would be necessary to freeze the ground to a depth have not been recorded in this area. Mr. Eric Riddiford, resident at the mouth of the Orongorongo River for the past 60 years, informs the writer that he has not seen the ground in his garden frozen, and that lemons and tender plants grow well there. Rainfall, however, is heavy. At the waterworks rainfall station the average annual precipitation for the last 15 years is 127.8 inches.
On the shingle slips examined, the stones were found to be moving rapidly down hill. Horizontal lines were marked on three slips by spilling roadmarking paint at intervals between two points of the same altitude on opposite sides of the scree. After three to four months most of the painted stones had rolled down the slope distances of from a few decimetres to 10 m. After nine to ten months the marks were lost. There was no indication that stones on any part of the scree moved faster than on any other part for the short period over which the marks remained visible.
On May 2, 1941, a line was marked across a scree in the waterfall branch of Ryan's Creek, a tributary of the Orongorongo River. When the site was revisited on June 8 of the same year a new landslide had occurred. The mountain above the original line presented an unstable face of shattered greywacke and argillite. From this face about 100 m. above the original line, a slump block roughly 20 m. across had fallen, leaving an arcuate scarp in the shattered rock. The block had largely crumbled to form fresh scree which had overwhelmed most of the painted line. Already there was incipient banding on this new scree. One month later there were well defined bands of larger and smaller stones. The stripes of larger stones widened and coalesced at the base where they joined together in a steep transverse gully full of large stones.
It is suggested by the writer that the bands of larger stones mark the position of subsurface watercourses. When rain falls, all the water that reaches the scree through its surface or by seepage or runoff from higher ground, is absorbed immediately into the porous shingle. It flows downhill through the interstices between the coarse waste which extends about 1 to 2 m. below the surface. It may be assumed that small fragments will be moved down hill
by the subsurface flow. In places at the edges of some screes there is much fine material of the size of peagravel and sand completely filling all the larger void spaces. This is taken to indicate that in the deepest parts of the screes there are no large void spaces, and ground water probably is present. It might be expected that the water which flows within the layer of loose waste material would move as a sheet, but in some manner it becomes concentrated in equal, parallel strips from which so much fine waste is removed that only coarse boulders are left on the surface. Similarly, on the more gently sloping boulder fans the water flowing within the loose waste collects in more or less equal anastomosing streams. The supply of fine waste on the scree is in excess of the carrying power of the water, so that usually enough of the smaller waste finds its way into the stripes of large stones to prevent these from developing into superficial gullies. The rock beneath the scree remains protected from the flowing water by a layer of closely packed fragments.
The parallel channels developed on limestone blocks and rarely on other rocks and described as lapiés (Cotton, 1941, p. 280) appear to result from a similar concentration of flowing water into more or less equal, parallel streams. The closely spaced, parallel gullies which are formed by the initial erosion of a soft, porous terrain on a steep slope—e.g., the drainage pattern on volcanic “shower” material described by Cotton (1941, p. 57) appear also to be analogous. In the loose waste of the scree slopes where gully forms do not result, the watercourses remain closely spaced and parallel, while, on the other hand, the pattern of the V-shaped gullies of bad-land erosion rapidly changes as some are enlarged by chance at the expense of others.
At the toe of the scree the bands of smaller stones disappear and a zone composed of only large boulders extends over a gradual and diminishing slope. On all the screes observed in the Orongorongo, a painted line across the middle of the slope showed that at this point the stones were moving rapidly downhill. The scarps at the head of the screes consisted of shattered greywacke cliffs up to 15 m. high, from which the supply of waste was continuous. At the foot of each scree was a tongue of large boulders up to 1 m. in diameter, many of which carried crustaceous lichen colonies several centimetres across. This indicated a more stable zone where the stones had not been disturbed for a period of perhaps ten years or more. The smaller sized waste material on the scree appears to be moved by the agency of running water through the interstices beneath the large boulders. If this is in fact the case, it may be anticipated that if a large scree were cut into by a river and a cliff formed across the toe of the slope, temporary streams would emerge during heavy rain from a layer below the surface boulders. Fans of fine waste would be expected to form at the outlet of these streams.
Certain points arising from this study of screes appear to be of interest in considering other forms of erosion. Conditions on the scree simulate conditions of aridity. There is no continuous plant cover or soil. There is a large amount of coarse, loose waste on the surface and, although the amount of water reaching the scree may
be considerable, it flows only beneath the surface. Also it is suggested that, because the supply of waste is in excess, the flowing water is always laden to capacity and acts solely as a transporting agent, moving waste material in more or less broad sheets down slope. The water can have no cutting action on the terrain.
In arid and semi-arid climates there are many wide plains with little or no soil and scanty vegetation. They are covered with loose waste material that is highly absorptive like the scree. On these porous plains there will be little or no runoff from such a surface and water, therefore, will not collect in streams or rivers to any extent. It is suggested that what water there is after rain will flow largely within the mantle of waste, as it does on the scree, spreading the stones in layers, and so tending to produce a flat surface.
On the bare residual mountains of the arid landscape, however, the runoff of all precipitation will be immediate and complete. Where the total precipitation though irregular is yet considerable—e.g., in the savanna type of climate, the runoff from the mountains will be very active. It will sweep all the waste from the continuing decay of the residuals away from their bases and may form watercourses or even rivers around the foot. If, where the more arid climate prevails, there is not sufficient rainfall to spread all the waste from the residuals, there will be some accumulation of scree around the base of the mountains. Thus the action of flowing water in these deserts will be to transport the waste resulting from the chemical and physical weathering of exposed rocks in such a way that nearly level plains are formed from which the steep sided residuals arise abruptly.
In the early development of an arid plain when a change from normal to arid climate occurs, great accumulations of waste may result from the accelerated erosion on steeper slopes following the loss of vegetative cover. In any case under arid conditions large amounts of waste are considered characteristic of the early stages of the cycle. The higher and steeper parts of the landscape which become bare are freely exposed to the atmosphere and continue to weather. Periodic sheet floods spread the waste material so that extensive plains of aggradation result (Cotton, 1942, p. 251). Parts of the terrain which are deeply buried will not weather so fast as parts freely exposed to the atmosphere. Weathering occurs down to the level of the ground water (Cotton, 1942, p. 21) apparently because the oxygen of the atmosphere circulates in the cavities not filled by water. This level may be at a considerable depth below the surface on an arid plain.
Following the period of aggradation, the action of the periodic floods of water is to spread the waste material out in sheets at lower and lower levels. The more elevated parts of the terrain beneath the plain of aggradation will become more exposed to the action of the atmosphere and also to the physical action of heat and cold and will decay faster than parts more deeply buried. As the level of the plain is gradually lowered the continued weathering of the buried rock nearest to the surface will tend to produce a plane rock surface beneath a veneer of waste material.
The statement that rock decay proceeds faster under a thin layer of waste than under a thick layer may be quoted also in an explanation of the development of convex land forms. The soil cover which constitutes the bulk of the mantle of waste under these conditions when it is saturated with water and flows in a mass (soil “creep”) may be assumed to have considerable coherence and to flow as an extremely viscous liquid. On a convex surface the effect of such flow would be the thinning of the layer of waste on the flat top of the convex hill, in the same way that a layer of pitch laid over a dome will flow away first from the highest point. On a hill with a convex surface the top has the thinnest cover of soil, and is presumed to be the point where rock decay will be most rapid. In these terms an explanation may be made of the development of hills of continually gentler slope by the production and removal of rock waste more rapidly from the almost flat top of the convex surface than from its more steeply sloping sides.
The conclusion may be drawn that while valleys generally are the result of the work done by the rivers which have occupied them, the interfluves are mainly reduced by the processes of weathering. The removal of the products of weathering by flowing water in the form of sheet-floods and rill wash and by mass movements aided by water allow the processes of rock decay to continue. The material from the convex hills is delivered into the valleys where graded rivers which have finished their work as agents actively shaping the landscape, continue the work of transportation of the waste. While the hills are slowly reduced, the valley floor will be maintained at about the same level, so that river and valley bottom may show but little change in a period during which the hills on either hand are considerably lowered.
Many of the observations on which this note is based were made while the writer was employed by the Wellington City Engineer's Department. The City Engineer has kindly given permission to publish this work.
Cotton, C. A., 1941. Landscape. Cambridge University Press.
—— 1942. Geomorphology. Whitcombe & Tombs, Ltd., Christchurch.
Sharpe, C. F. S., 1938. Landslides and Related Phenomena. Columbia University Press, New York.
Zotov, V. D., 1940. Certain Types of Soil Erosion and Resultant Relief Features on the Higher Mountains of New Zealand. N.Z. Journ. Sci. and Tech., 21: 256B–262B.