The Cave

Appendix 1

Posted by steven on 04/03/2005 at 04:19 PM

The geomorphology of Uamh an Ard Achadh (High Pasture Cave), Isle of Skye: a prliminary report

Tim Lawson & Ivan Young
Grampian Speleological Group

Uamh an Ard Achadh (NG 59425 19706) is a 320 m long active cave in the Coille Gaireallach area in Strath Suardal.  The cave was visited in May 2004 by the authors and a brief investigation of its morphology and sedimentary deposits was made.  This forms the basis of this preliminary report.

1.  Past work on the cave’s geomorphology

Discovered in 1972, a Grade 5c survey of the cave was made by the Moldywarps Speleological Group in 1973; this survey is the most accurate one published to date, and forms the basis of the illustrations in this report (Ryder 1995).  Ryder included a discussion of the possible development of the cave in a more general article on the caves of this area (Ryder 1974).  He suggested three phases of development within the cave: a possibly ‘Pre-Glacial’ phase represented by all the dry side passages and the high-level oxbows; a subsequent phase (possibly ‘Pre-Glacial’ or ‘Inter-Glacial’) during which the active streamway was created; some potentially ‘Post-Glacial’ development of the previous passages.  This will be discussed below.

2.  Cave morphology

The cave underlies a shallow, north-east trending valley, partly dry, on the north-west side of Beinn an Dubhaich.  A small stream sinks about 25 m south-west of the present entrance, and re-emerges approximately 300 m away at NG 596198.  The undulating ground between these two points contains the cave under investigation. . At the downstream end of the navigable cave is Terminal Chamber (NG 59546 19767). This lies directly beneath loosely packed boulders at the head of a shallow valley. This at one time was the main resurgence for the cave and probably still acts as such in extreme flood conditions. Between Terminal Chamber and the present day resurgence the passage is completely flooded, much smaller in cross section, and has not been penetrated by cavers for more than a few metres.  The entrance to this cave is via a shaft excavated through a bouldery fill, giving access to a roomy stream passage (‘Tin Can Alley’) exhibiting a phreatic origin.  The presence of a rounded, cruciform cross-section (e.g. cross-section c on the cave survey) indicates formation of this passage beneath a water table at the intersection of a vertical joint or fault plane and an approximately horizontal bedding plane.  Downstream of the duck, cave cross-sections have a more angular appearance as the roof of the active streamway often follows an horizontal or inclined surface, representing a bedding or thrust plane of some sort.  Downstream of the waterfall, the cave exhibits stronger vadose development as the active stream has incised into the rock.  Breakdown processes have modified the original cave cross-section where the passage has been widened by lateral stream erosion.

The presence of at least ten igneous intrusions (dykes) was noted during the survey (see Fig. 1).  Sometimes the cave passage runs along the side of a dyke, which forms one of the passage walls (e.g. cross-section n in the Lower Streamway).  More intriguingly, the dykes are often seen to transect the passageway, which begs the question how the cave was able to form along its present orientation.  Examination of surface outcrops of several of the dykes shows that they are often relatively thin sheets of igneous rocks.  Although they would have acted as impenetrable barriers in the initial stages of cave formation beneath a local water table, they may have contained localised zones of weakness that might have initially been exploited chemically, and then by hydraulic action and abrasion as the local water table was drawn down by glacial erosion of Strath Suardal.  As an hypothesis, this is difficult to test, but the chemical composition of the intrusions may be worthy of further study.  It was noted that most of the intrusions in the cave appear to break down very readily into small angular fragments and lumps can easily be removed by hand. Peculiarly, they seem to have offered a path of less resistance to erosion (other than by carbonation/solution) than the limestone.  It is also clearly necessary to map the outcrops of intrusions on the surface and attempt to correlate with those seen in the cave, to determine how continuous they are across the area where the cave has formed.  In certain locations, the presence of a dyke has had a significant influence on the cave passage morphology.  The 3 m-high waterfall that separates the Upper Streamway from the Lower Streamway cascades over one such intrusion, which seems to have been particularly resistant to erosion and has acted as a real barrier to stream incision upstream.

Fig. 1:  Occurrence of igneous intrusions in Uamh an Ard Achadh

The published cross section illustrates two major planes of development within this cave system (Fig 2).  Although it is tempting to interpret such information in terms of the initial development of an upper level followed by the subsequent development of successive lower levels as the local water table falls due to external influences, it is necessary to discount the possibility of contemporaneous development of cave passages at more than one level by solution beneath a single water table – perhaps along structural lines of weakness at more than one level over large parts of the cave.  Indeed, this must have been the case between cross-sections n and o on the cave survey where there are two distinct levels. The lower level must have been forming when the main stream was still flowing along the upper level.  The almost continuous presence of an upper series of passages in the form of a high-level canyon or abandoned oxbows high on the passage walls along the Lower Streamway is strong evidence for a single, formerly active streamway flowing along this path in vadose conditions (i.e. freely flowing water, with an airspace above).  Under such conditions both lateral erosion (to form the meanders and oxbows) and incision (to form the canyons) can occur.  It could be that for most of its life the cave resurged at the head of the ‘dry’ valley close to the position of Terminal Chamber, and that is what set the level of the local water table. 

That a subsequent period of downcutting occurred to create the present active passages of the Lower Streamway is in no doubt, but the reason for its initiation may not be due solely to glacial erosion in the valley outside the cave.  The longitudinal section on the survey shows that the top of the waterfall is approximately at the height of the upper level of passages downstream of it.  The presence of the intrusion that forms the waterfall has had a bearing on the evolution of the Upper Streamway series in that it has prevented the development of a lower passage level, preventing the necessary downcutting from occurring.  Another intriguing feature of the cave plan is the proximity of the two calcited chokes in the middle part of the cave.  Both comprise truncated sections of an upper level of passages, and appear to end within 5 m of each other at about the same altitude.  They must also be very near to the surface at this point, the choke most probably representing some form of collapse of the surface.  Preliminary investigations suggest that no obvious depression occurs on the surface here, nor evidence of former stream channels, although these may have been smoothed over by glacial activity.  However, if it can be demonstrated that a former water course sank at this point on the surface, this would have some bearing on the relationship of the ages of the Upper and Lower streamways.  However, it is more likely that the choke is blocking what was once a continuous passage linking the higher level remnants found upstream of the waterfall.

Fig. 2 Distribution of high-level passage fragments in Uamh an Ard Achadh

The age of the upper and lower cave passage levels, and their relationship to the advance of the last ice sheet across this area (circa 26,000-14,000 years ago), will only be firmly established by a combination of radiocarbon dating of organic remains and uranium-thorium dating of speleothems.

3.  Sediments

The preliminary survey of the clastic sediments in the cave suggests that a number of separate facies are present.  Facies 1 is a pale yellow or yellow-grey silt which occupies isolated, sheltered locations such as crevices and pockets in the cave wall; this is truncated by Facies 2 deposits, a mixture of rounded and sub-rounded cobbles and gravel – many of which are darkly stained and partly indurated with calcite.  This may represent a lag gravel, the coarser fraction of a more extensive sand and gravel deposit whose finer particles have been winnowed away by streams.  The dark staining is possibly an oxide of magnesium or manganese, derived from humic-rich, ‘peaty’ surface water.  Facies 3 is a red-brown gritty sand and fine gravel, whose particle size range varies throughout the cave, depending on environment of deposition.  Facies 4 is a brown silt which is invariably found at the top of sediment sequences in sheltered parts of the cave.  It seems to have been washed into the cave by groundwater via fissures in the local rock.  Facies 5 is a breakdown deposit, derived from spalling or collapse of the roof or wall of the cave.

A schematic diagram of the postulated sedimentary sequence is given in Fig. 3.  This is based on a sedimentary stratigraphy discovered at Location A (Fig. 4) in the Upper Streamway.  The oldest sediments appear to be the yellow or pale yellow-grey silt, indicating deposition in a low-energy environment in ponded water or flooded cave passages.  They may represent the by-products of glacial abrasion of the land surface, washed into the caves by meltwater, but at this stage this is entirely speculative.  Subsequent erosion of these deposits by flowing water has resulted in their patchy survival throughout the cave.  The occurrence of Facies 5 (breakdown deposits) is dependent on local conditions of structural instability within the rock, but such deposits are found at the base of the stratigraphy at location A; it was difficult to ascertain whether the breakdown cut the pale yellow-grey silt, or whether the latter occupied a pocket within the breakdown, although the former was considered the most likely scenario.  A very thin layer of flowstone overlay the breakdown, indicating a period of increased roof drip and low water level in the cave.  Above this was a mixed sand and gravel deposit, comprising a greater proportion of water-worn, dark stained igneous clasts in the lower level.  Analysis of samples taken from the upper and lower levels of this deposit has shown a degree of lithological similarity throughout, although limestone and chert clasts are more prevalent in the upper level and more concretions of reprecipitated calcite are found nearer the base.  The upper levels have a surprisingly high proportion (31%) of dark-stained igneous clasts although only 5% of these showed any degree of rounding; in contrast, the upper level of this stratum possessed 35% of such material, 30% of which showed the effect of attrition carried out in flowing water.  This part of the sedimentary sequence at Location A was topped by flowstone, being actively contributed to in some places, but also partly covered by a deposit of muddy silty-sand.  This stratum contains fragments of charcoal which are most noticeable in its lower level. 

Fig. 3 Schematic lithostratigraphy

Two samples of lag gravel deposits have been examined from the upper passage levels.  At Location C (Fig. 4), angular gravel clasts contain some rounded cobbles have been partly indurated by calcite deposited from dripwater.  This sample differed from the gravel fraction studied at Location A in that a significant proportion of the clasts (23%) appeared to be of a coarse-grained igneous rock, thought to be the local granite, and 4% of the clasts were red sandstone.  This indicates that a sizeable proportion of the gravel has been derived from glacial deposits on the surface.  Some of the dark-stained igneous clasts exhibited sign of rounding of edges, but the majority were angular in shape and clearly derived from the nearby dyke that crosses the passage a few metres to the south.  At Location D (Fig. 4), only 7% of the clasts were limestone, calcited concretions or chert and the great majority were fragments of various igneous rock types.  66% of the material was stained dark and, of these clasts, 33% of them showed signs of rounding of edges.  Gravel samples from Location F have yet to be examined fully, but again show a surprisingly low content of calcareous material.  Analysis of fine sediment samples, which requires specialist equipment, has yet to be carried out.

Fig. 4 Diagramatic representation lag gravel deposits

4.  Future work

(a) Cave morphology.  Future work should attempt to identify all occurrences of igneous intrusions cropping out in the cave passages, and a survey of their thickness and petrology should be undertaken to determine any variety and thereby their relative resistance to erosion, as they appear to have had only limited influence on the evolution of the cave.  This should be matched with a survey of dyke outcrops and joint patterns above ground.  A survey of the land surface immediately above the cave would show how far below the ground lie each of the passages.  This would help to rule out the possibility of a former water sink into the cave at the point above the choked high-level passages that appear to end so close to one another.

(b) Sediment analysis.  Fine-grained sediments need to be characterised by particle size analysis, and an attempt should be made to ascertain their chemical composition to see if they have been derived purely from within the cave by weathering of the bedrock, whether they represent ‘rock flour’ created by subglacial abrasion on the surface and washed into the caves by meltwater, or whether they are derived from organic-rich, modern surficial deposits washed into the cave by dripwater.  Further lithological characterisation of the various gravel deposits and the mapping of facies units in the cave should take place so that the relationship of various strata to the development of the cave, and the relationship of ‘natural’ sediments to those of archaeological significance in the excavated side passage, can be ascertained.

(c) Dating.  Once the above has been progressed it will be necessary to try to pin down the chronostratigraphy of the cave’s morphological development and sedimentary history by uranium-series dating of speleothems and radiocarbon assay of organic remains.

5.  References

Ryder, P.F. (1974) The Caves of the Beinn an Dubhaich Area, Isle of Skye.  BCRA Trans. 1 (2) 101-124.

Ryder, P.F. (1995) Caves of Skye.  Grampian Speleological Group, occ. Publ. No. 7, Edinburgh, 73 pp.

Figure 28.  Alex Kozikowska emerging from the excavated entrance to High Pasture Cave, Skye.

Previous entry: References