Soil Development in Recently Deglaciated Terrain by R. A. Cranwell

This is a digitization of the work ‘An Analysis of Factors and Processes of Soil Development Operating in Recently Deglaciated Terrain in Svinafell South East Iceland’ by R. A. Cranwell produced in 1980.  You can find more of the work of Robert Cranwell at his website Amateur Emigrant which holds recounted stories of his travels to far flung places all over the world sharing insights into the peoples, the cultures, the built and natural environments:

Bob Cranwell
Bob Cranwell

www.amateuremigrant.com

Bob says: “Pick up a twig on a walk down the road, anywhere. Take it carefully home. Make a cuppa. Look. There are many things at work on that twig. Some we (I) can barely imagine. They are critical to our world. Just take a look for 10 mins. More if you like… Have a think”.  You can find him on social media too:

@cranwellbob

@amateuremigrant

Table of Contents

Dedication

This work is dedicated to my parents, although they may not understand it, without their help, neither would I.

Acknowledgements

My main acknowledgements for help in this work go to Pete Wilkinson, B.Sc., without whose enthusiasm and genuine interest I would not have been prompted to stay up, sometimes until 4 am completing tests. For that matter, it is unlikely that I would have completed the fieldwork without Pete and his trusty spade.

Acknowledgements, too go to the other members of the ICEX 79 Expedition, some of whom exhibited enthusiasm, and gave help, others of whom irritated me sufficiently to prompt me to spend more time in the field than in camp. Very many thanks to my sister Theresa Baker, for typing the scrawled manuscript, and to my parents and young brother for allowing me to abuse their hospitality whilst writing this.

Great appreciation must go to Mr. Gordon Forrest, for the endurance he displayed in trying to get the stepwise program to work, and for taking the time and trouble to explain what was happening. Thanks go to Mr. Mike Meadows of the Geography department, for encouraging and constructive comment on the manuscript in the final stages. Acknowledgements to Dr. Frank Nicholson, for providing some of the direction, especially at the earlier stages of the study.

Acknowledgements to Paul Elek Scientific Books Ltd., for permission to reproduce Tables 8-11 from “Techniques in Pedology” (Smith and Atkinson, 1975). Finally, thanks to the few members of staff in the Geography department who took the trouble to be of, practical and otherwise, help in the preparation of this work.

List of Figures in Text

Fig 1: Location of study area in Iceland
Fig 2: Area around Oraefajokull, S.E. Iceland
Fig 3: Area of Svinafell and Skaftafell glaciers
Fig 4: Morphological map of Svinafell moraines
Fig 5: Slope classes on the Svinafell moraines
Fig 6: Bulk Density variation with distance from ice
Fig 7: Percentage moisture variation with distance from ice
Fig 8: Mean organic matter variation with distance from ice
Fig 9: Proportions of silt and clay in selected profiles
Fig 10: pH variation through profiles with distance from ice
Fig 11: pH changes within profiles in selected pits
Fig 12: Percentage vegetation variation with distance from ice ( log x normal )
Fig 13: Matrix of correlation coefficients
Fig 14: Cluster analysis of correlation coefficients
Fig 15: Dendogram of cluster correlation levels
Fig 16: Matrix of cophenetic values
Fig 17: Graph of cophenetic values against original correlations
Fig 18: Display of associations from stepwise multiple regression
Fig 19 – 48: Soil profiles, and descriptions for 30 pits
Fig 49: Diagram of pouring cylinder for bulk density measurements
Fige 50: Diagram of solar oven

List of Tables in Text

Table 1. Soil moisture results
Table 2. Soil organic matter results
Table 3. Results of sedimentation tests
Table 4. Results of soil pH tests
Table 5. Results of stepwise multiple regression
Table 6. Matrix for analysis of variance on directional data
Table 7.Results of analysis of variance on directional data
Table 8. Soils structure terminology
Table 9. Soil consistence terminology
Table 10. Stoniness description terminology
Table 11. Soil texture description terminology
Table 12. Rainfall data
Tables 8-11 Inclusive taken from Smith and Atkinson,(1975)
Table 12 reproduced courtesy of J. Wood Esq.

Chapter 1: Introduction

This study is a serious attempt to identify and quantify the interactive relationships between a number of factors which are influential in, or symptomatic of, soil development. The work is primarily concerned with relationships developing at a very early stage, and it is for this reason that an area of study was required where such conditions were likely to exist.

In South Eastern Iceland lies the largest icecap in Europe, Vatnajokull, (Figs. 1,2), and for a number of decades the outlet glaciers of Vatnajokull have been in retreat. One such glacier is Svinafellsjokull, (Figs. 3,4), which, until 1935, was joined with its neighbour, Skaftafellsjokull, at their snouts. However, considerable thinning and retreat of both glaciers has left, especially in the case of Svinafell, a large area of hummocky, recent moraines which was ideal for a study of this nature. In order to better describe the area it is probably necessary to set the scene of the study in more detail. The Svinafell glacier runs roughly North East – South West, and is bounded on both sides by sharp ridges up to 1000m high.

The glacier leaves the main icecap in a steep icefall of roughly the same height, and travels downslope to a height of less than 100m at the snout. The outwash plain of Svinafell is of low gradient running to the sea, and is cut by three main meltwater streams, fast running and turbulent, which drain the glacier. This area is covered with surface mosses, probably Rhacomitrium, about 10-15cm deep. In contrast, the moraines are very varied in vegetation cover, being mainly of extremely stony surfaces. The moraines upon which this study was based varied in height by some 30m, however, this was not a regular variation by an means, and a complex series of ridges and hollows existed. Variation too, was evident in the vegetation, which was generally zero on areas of raw till, whilst older areas tended to have more vegetation, however, the relief of the area seemed to play a major role in determining vegetation cover.

It is in this type of area that other studies, few in number, notably Crocker and Major (1955); Lawrence (1953); have been carried out, and it is from interest in this type of work that this study was conceived. However, for the most part, work of this type has been confined to investigations of variation in single factors and only occasionally with interactive relationships between two or more factors. While this type of analysis is felt to be both necessary and worthwhile in itself, it was thought that some energy could usefully be diverted into investigating the relationships between a number of factors in this environment. To this end the study has been divided into three main areas of analysis:

  • the recording and analysis of variations in individual factors,
  • the interrelationships between paired combination of factors,
  • the interrelationships between multiple combinations of factors.

The main hypothesis motivating this study has been the possibility of developing an expression, or model, encompassing the factors investigated in a tentative guide to the mechanisms involved in soil development, applicable to the type of environment in which this study is based. A secondary motivation has been to assess the benefits and problems of using multivariate analysis in this area of work.

The fieldwork upon which this study is based took place over a four week period in July, 1979, when time was spent working on the moraines of Svinafellsjokull, a temperate valley glacier. In practical terms, the work was carried out on a four hundred metre transect from the ice front, thirty metres wide, across very young glacial deposits. These ranged in age from a few days, to, at the most slightly over seventy five years old. A large number of measurements were taken of various facets within the study and methods of measurement are described in Chapter 2.

Because the amount of previous work on the area of recently deglaciated terrain is relatively small, reference to the findings or methods of previous workers will be found at relevant places in the text, and in addition will receive some attention and discussion in the summary. A separate discussion would necessarily limit the discussion referring specifically to the present authors work.

In some respects, this study represents what may appear an elementary expedition into a field about which little is known and even less published. In many ways this is generally so but the limitations in which this study is placed serve only to underline the lack of real available knowledge in viewing the soil, and especially its early development from a holistic view point.

Area around Oraefajokll, S.E. Inland
Fig. 2: Area around Oraefajokll, S.E. Inland

 

Fig 3: Area of Svinafell and Skatafell glaciers. S. E. Iceland

 

Fig. 4 Part of Svinafellsjokull moraines, S.E. Iceland, 1979

MISSING FIG 5

Chapter 2: Description of Methods Used in Data Collection

2:1 Sampling

The sampling procedure presented considerable delays in the study, especially at the beginning. After a long period of reconnaissance an area was selected which, subjectively, was as representative of the area as could reasonably be defined. The area, a transect of 400 metres by 30 metres was then subdivided into 100 metres x 30 metre plots and by using random numbers pit sites were allocated to each area in turn for the first 20 sites. After completion of these pits a strong bias was detected in the bearing of these sites, consequently allocation by random number selection for the remaining 10 sites was undertaken with an exclusive selective bias in favour of those quadrants which were under-represented, this was the only bias which was deliberately introduced in the sampling.

2:2 Bearing

This was quite simply determined by hand-held compass, identifying the direction which slopes faced on which pits were to be located. Correction for magnetic variation was, of course, taken into account.

2:3 Bulk Density

These were completed in as short a time as possible to minimise variability in the soil moisture regime, since the soil extracted in bulk density measurements was also used to determine soil moisture (see section 2:6). The procedure was a standard method described by the British Standard Handbook. The apparatus used is illustrated and described in the appendices, as too is its calibration and accuracy. Basically the procedure involves excavating a cylindrical hole about five inches deep and diameter four inches using a template.

The hole is then filled with sand of a known bulk density. Meanwhile, the excavated soil is bagged and weighed both in normal state, and after drying, which was done in a solar oven (see appendices). From the results obtained, and from information gained in the calibration of the unit, the wet and dry bulk densities can be determined.

2:4 Distance

This is a measure, in metres of the distance at which the pit in question is located along the transect in a direction away from the ice front. This was done by tape measure.

2:5 Percentage Downslope

This is an assessed measure of the position of a pit in relation to the slope upon which it is located.

2:6 Percentage Moisture

Determined by standard methods of weighing before and after drying. Although difficulties exist in the measure of absolute moisture content, this method is a common and useful measure of relative moisture content. The solar oven was used for drying (see appendices). To minimise variations in soil moisture over the transect, all samples were taken within 6 hours of starting. Rainfall figures (courtesy of J. Wood) for the period immediately preceding taking of samples are presented in the appendices.

2:7 Organic Matter

Carried out by standard methods, but a brief description precedes the results in section 3:4 along with some problems encountered.

2:8 Particle Size

In order to gain a rough indication of the particle size distribution of the soils that were being inspected, it was decided to try rough sedimentation tests. Using two small drops of washing up liquid as dispersant, in 1000ccs., of meltwater, approximately 50ccs., were poured into a series of test tubes, each already containing 10g of finely ground material from the soil samples taken. These tubes were then shaken, not vigorously, but sufficiently to mix the sediments to a similar degree in each tube. The depth of sediment which had settled out was noted after 5 minutes, 5 hours, and 24 hours. The 5 minute reading giving an indication of sand fraction, 5 hours indicating silt fraction and 24 hours indicating clay fraction. These are obviously only very approximate but they did serve as indicators.

2:9 pH

Again, standard methods used, a brief explanatory section precedes the results in section 3:6.

2:10 Slopes

The degree of slope was determined by compass clinometer, a task which is easily undertaken. Slope type had to be assessed by eye, and obviously in a continuum from highly concave to highly convex a wide variation can be expected. However, for ease of determining this factor a simple split into concave, straight and convex was decided upon since it was felt that further elaboration was unnecessary in a study of this type (see fig. 5).

2:11 Vegetation Cover

This factor presents the most problematic area of the study since it is notoriously difficult to assess vegetation cover, and a large amount of work has been devoted to refining assessment methods, e.g., Daubenmire; 1959. However, to a large extent this type of assessment depends upon practice, and, since the author had both limited time to refine methods and prior experience of one method, viz., percentage cover, it was decided that reasonably accurate assessment on a scale which had previously been used would be more useful than a possible confusion resulting from use of familiar methods. As a result, percentage cover was assessed in an area of one metre radius of the pit site.

2:12 Evaluation of Methods Used and Problems Encountered

For the most part, the methods used were relatively rapidly carried out and easy to use, whether they were standard techniques or not. However, some methods presented considerable difficulty. The sand used in the bulk density measurements had to be found, dried, sieved and stored. After trying beach sand from a site some 50 miles distant, which was not suitable, several alternative sources were investigated, resulting in our use of windblown sand that had accumulated along the embankments upon which the road near the site was built. It entailed sieving and bagging of the sand in situ, transporting to camp and drying, which posed problems because of the volume of sand and unsettled weather. When the sand was finally air dry, it was sieved again, by hand, to get sand in the range 250 to 500 microns, a small range than the technique required, but ensuring greater accuracy. A further difficulty was the transporting of 60 kg of sand across the moraines to the site but this was overcome. The apparatus for measuring the Bulk Density was not available, so a pouring cylinder was constructed by the author from a blueprint (British Standard 1377) using plywood and 4 inch plastic drainpipe. The equipment aroused much amusement but proved to be an accurate and useful tool.

The drying of samples presented some difficulties, because of the large number of samples which many of the expedition required. The use of a solar oven, constructed by P. Wilkinson (see appendices) solved a great many problems but it is doubtful whether it achieved the same efficacy as oven drying at a constant temperature until constant weight is achieved.

However, under the circumstances it proved an admirable substitute and it is hoped that any inaccuracies that might have ensued from its use are at least systematic errors and apply equally to all the samples thus dried.

The premise upon which the study is based that is, that the sampling was correctly done may be open to some question on the basis that the area was subjectively chosen. This was done to try and ensure that a representative area was used, since random sampling of the whole area could have resulted in wholly unrepresentative data, and, in addition would have entailed far more work than could reasonably have been achieved in the fieldwork period. It is the present author’s opinion that a more representative area could not reasonably have been selected in the area in which the study was based.

The major area of criticism which would be substantiated lies in the vegetation sampling. It has been mentioned earlier that numerous methods of assessment exist, many of which are far more accurate than the method chosen. However, it must be admitted that it was felt more safe to use a method which had been used before than risk unfamiliar methods. The basic problem arises from the unwillingness of the author to try a different method because it would be a departure from familiar territory. However, the vegetation cover data is as accurate as the method would allow and, as a final word, the adoption of newer, more time consuming methods may have resulted in more unreliable or incomplete data in the hands of an inexperienced operator.

As a note to the particle size assessment method used, it must be borne in mind that accurate measures were not required since they were only undertaken as an indicator of the type of material with which we were working. It was not felt necessary to do elaborate analyses of particle size partly because of the nature of till fabric, and partly because of the scarcity of equipment to do extensive tests, since other people on the expedition had far more justification for the use of sieves, hydrometers, etc.
As a final note on problems in the fieldwork a good number of days and half days produced no results whatsoever, and so curtailed the extent of the study because of rain which was a frequent visitor to the site.

Chapter 3: Principal Data Used in Analysis; Analysis of Individual Elements.

3:1 Soil Pit Profiles

The profile are presented in the Appendices, along with profile data and slope data. The striking facet of the thirty soil profiles which were recorded is their similarity in a great many areas. The facets recorded for individual profiles are discussed below as group information.

Texture varied little over the whole transect, with the sole exception of the degree of sandy texture being much more pronounced in the earlier pits. Bands of entirely sandy material exist in some profiles but this is due to inwashing of sandy material from surface water and later covered by other deposits.

The variation of texture through profile took a rather more complex turn, since on elevated pits, i.e., on ridges, the highest proportion of silty material was found in the upper most horizon whereas in other pits there appeared to be a translocation of the silt fraction albeit slight, through the profile. A major cause of this has been suggested as frost action on exposed areas when snow lies on the ground. The ridge tops have snow cover blown off them and are more susceptible to frost action. This moves stones upward through the profile, to emerge at the surface, then they are quickly moved downslope by gravity and dislodged by the movements beneath them. This eventually results in a concentration of silt size particles in this region. Furthermore, the ability of small particles to hold water better than large particles renders silt concentrations more susceptible to ice lensing which may move them upward through profiles. As a general description the texture of these soils is soft and gritty with slight to strong cohesion when moist.

Structure: predominantly of fine to medium sized crumbs but with only slight to moderate development.

Consistency: Probably because of the unconsolidated nature of the material, consistency was predominantly soft with weak cementation, but some upper horizons when moist were slightly sticky and slightly plastic.

Stoniness: As can be expected in glacial material tremendous amounts of stones littered the surface, ranging from gravel to boulders. A similar situation was reflected in the soils, but a much higher proportion of smaller stones was found, all of which were angular to subangular. In no pit were stones not a problem to excavation and it is only in the narrow sandy horizons that stones were not found.

Colour: Dominant colours were in the range 7.5 YR to 10YR, with low hue and chroma values. Pew departures from this norm occurred except where in situ breakdown of a common green coloured agglomerate had occurred. In many cases the differences in colour between horizons is the only major difference, and this is only a slight alteration.

Boundaries of Horizons: Very poorly developed with few sharp delineations, occurring in conjunction with sandy horizons. In some cases the boundaries are so vague as to be counted as separate horizons themselves.

Organic Matter: Very little visible detritus, and roots small in terms of mass. However, some woody roots from Salix occurred, but the major form of root presence was a preponderance of extensive thin hairy roots from the herbaceous plants above, these extended both laterally and downward, as a response to the following factor, and to the shortage of available nutrients.

Drainage: Soils appeared to be well drained, and their coarse nature speeded surface drying out after rainfall. However, the pit profiles always showed moistness about a foot, sometimes less, beneath the surface. It is for this reason that in periods of heavy rainfall, surface gullying, especially near the ice front, occurred, creating gullies up to three feet in depth on the unvegetated slopes near the ice

3:2 Bulk Density

The Bulk Density values shown in Fig. 6 seem to confirm the expectations with which the investigation began. A definite trend toward lower densities of soil exists, from around 2000 kg/m3 near the ice front, to values of around l6-1700kg/m3 toward the end of the transect.

One probable main cause of this is the influence of vegetation, however, the influence is very variable, as will be seen in later analysis. Nevertheless, lateral root action still has a tendency to disturb the upper layers in the soil. As well as this, the action of soil animals probably plays an important part, although there was a dearth of mammals in the area generally, there were large numbers of beetles, and other soil fauna with hard exoskeletons which may assume a quite important role in this area. Because of the nature of the material, being unconsolidated, few soft bodied animals could operate in it without being quickly crushed. In fact, in a preliminary survey of earthworms, only five were found in a selection of ten pit sites.

As can be seen from Fig 6 a fairly distinct change in soil density occurs over the distance of the transect. Assuming that distance from the ice front may be taken as a surrogate for age, then the change in values correlates well with trends found by Crocker (1960), in the Mt. Shasta, and Glacier Bay sequences. In his investigation with Major, (Crocker& Major;1955), upon which part of the later work is based, he found that considerable changes in density occurred in a short space of time, the original density halving in about 500 years. Moreover, the change was confined to a great extent to the upper layers of the soil. The results obtained here reasonably accurately reflect both components of this change.

However, the origins of the changes are more problematic to locate, in Crocker’s (1955;1960), work, most emphasis was put upon vegetation, and it must be admitted that this was investigated in this study with the idea that the results were a fait accompli. As will be seen from later analyses in this study, the plant factor in changing soil densities is by no means the only component, nor does it act independently itself.

3:3 Moisture content

Moisture content is generally low 6 to 10 percent by weight and little variation can be seen with distance from the ice front.
(See fig 7) This, at the time of determination was thought to be unusual as it was thought that moisture retention would increase with organic matter content. However from subsequent analyses in Chapters 4 & 5 it will be seen that a much more complex explanation is required. However, the generally low contents can to some extent be explained by the granular free draining nature of the soil.

If reference too, is made to the rainfall data supplied by John Wood, (another expedition member) in the appendices it will be seen that little rainfall had occurred in the preceding few days. What was expected was a change in the relative amounts of water in the soil along the transect but as can be seen this did not occur. See also Table 1.

Table 1 Soil Moisture Results

Table 1: Soil Moisture Results

3:4 Organic Matter Content

(See Table 2, & Figure 8)
Organic matter, where it occurred tended to be concentrated in the upper horizon with appreciable differences in the second and lower horizons as with soil moisture the expected result of increase with distance from the ice front (as a surrogate for age) was not realised. Instead, the mean organic matter tends, if anything, to decline with distance along the transect and the main reason for this is the variability in vegetation cover and the variability of sampling.

In Chapters 4 & 5 it will again be shown that the distribution of organic matter is by no means as clear cut as might be imagined, the primary influence it appears, is not the age of the surface but its topography.

It is on the basis of this, that doubt could conceivably be thrown on other studies where organic matter is only correlated against one other variable, usually surface age, since unless sites are used which are in every other respect comparable, then results of organic matter content become very unconvincing indeed.

Soil Organic Matter: Table 2

Method of Determination
First, 10g of fine soil material, less than 2mm, was produced by crushing with a mortar and pestle, taking care to remove rock fragments. This material was first weighed on the most accurate available scale, only accurate to .1g, however, it was then weighed with the crucible in which it would be burned, and this weight noted. The crucible then had the lid placed on it, and was placed on a triangular stand over a bunsen burner, using a hot blue flame.

Three pilot samples were burned for first, 5 minutes, then weighed after cooling for 10 minutes, then burned for 10 minutes, cooled and weighed, then for 15 minutes, cooled and weighed.

From the results achieved from these tests, it was decided that the most efficient burning time would be 10 minutes for the remaining samples, partly because of the extra time involved, and partly because losses, if any, were minimal. In some cases samples gained weight slightly, and this was thought to be due to the hygroscopic effect of heating samples for longer periods. In order to minimise the latter, during the 10 minute cooling period, all crucible had their lids on.

Weights overleaf are for sample plus crucible

 

3:5 Particle Size Distribution

From the rough and ready method employed, (see section 2:8 ), not a great deal may be usefully inferred. However, the method suited the purpose of the moment, and confirmed much of what was expected. From the histograms in Figure 9, it can be seen that wide variations occur in the distributions of clay and silt fractions, and, thereby, in the coarser units. The only possible point which may be usefully made from these distributions is a comparison, albeit tentative, with Retzer, (1965). It is just about permissible to see a corroboration of his statement:

‘silts and clays occur in the greatest quantities in the upper two horizons, and decrease with depth almost without exception.’

However, because of the nature of the method employed in this case, it would be unreasonable to infer more than a vague similarity. What the tests were expected to show, and did show, was a large variation in the type of material with which we were working. This, it may be said with some confidence, is a typical characteristic of glacial deposits and so it came as no surprise when the tests confirmed it. (See also Table 3 ).

3:6 Variations in pH

The range of pH values along the transect varied only half a full point, from 6.7 to 7.2, for the majority of individual pits. There were several departures from this main grouping, however, taking the full range to one full point, from 6.3 to 7.4 .

In Fig.10, the variations within the upper two horizons of the profiles are graphed against distance from the ice front. The pH of the third horizons is patchy, since only a minority of pits exhibited anything which could be described as three horizons. However, something can be inferred from the plotting of the first two horizons. Firstly, in concord with what Retzer, (1965), describes in alpine soils, there is an increase in pH with depth, although the values recorded are generally higher than those he quotes, but this may be accounted for by differences in lithology, rather than differences in soil forming processes.

A further occurrence is that the differences between upper horizons appears to be less than differences between lower horizons. Although pH values for third horizons are not displayed, they are tabulated in Table 4. A main point to be noted is that the pH values appear to be stabilising, after 200m along the transect. The relationship between horizon pH values is obviously erratic within 200m of the ice front, partly due to the large amounts of water present within 20m of the ice, but clear evidence of attaining a temporary equilibrium between horizon pH values seems to exhibit itself as the effects of vegetation, translocation of minerals, and climate begin to show.

A further illustration of the “regularisation” of pH values is shown in Fig. 11, where the upper figure shows the change in horizon pH through several selected pits at the start of the transect, and the lower figure shows the same, but for pits at the end of the transect. It can again be clearly seen that the erratic nature of pH changes has been mollified, from sharp variations, and even reversals of trends in the early pits, to a more regularised variation in the later profile pH values.

Soil Ph Results

TABLE 4
Method
First, 10 grammes of soil material from each sample was taken, removing all rock fragments over 2mm., this was crushed with a mortar and pestle. The (weighed) 10g was then placed in a 50ml flask, with 25ml of distilled water, (see note below). This mixture was then shaken thoroughly and the W.P.A. Ph probe was then inserted, (after first calibrating with Ph7 Buffer solution). This was then read off, the probe was washed twice in distilled water and re-calibrated. This procedure was followed at least twice for each sample (three times where a notable difference occurred between 1st and 2nd readings), and the mean reading noted.

N.B, Distilled Water: very little distilled water was available, therefore it was only used to make up buffer solutions. Meltwater from supra-glacial streams was used instead. To account for the buffering effect this may have had, since its Ph was approximately 6.5, some ordinary vinegar was diluted to .25m with the meltwater, and its Ph value noted. A sample of the vinegar was taken back to England, where it too was diluted to .25m with distilled water, the mean Ph was then noted, The original pH was 3.5, and the sample tested in England proved to be 3.75, when diluted to .25M. This means a buffering effect existed which reduced the pH values by .25 of a pH unit.

3:7 Vegetation

Vegetation cover was very variable, and no real sign of increase with distance can be defected, although there is a slight statistical one. Rather, as shall be further demonstrated in Chapters 4 and 5, the cover varies with upslope and downslope variation, i.e., where hollows occur, vegetation occurs, and vice versa.

The areas nearest the ice front tended to attract the colonisers, such as Dryas, Galium, and some of the Graminae. (Common names are given in the appendices). The flowering plants tended to be alpine varieties, producing large numbers of seed, from which few survive in the harsh environment. Further from the ice, woody species, mainly Salix,spp. become established, but the pioneer species still remain in evidence.

A common feature of tie hollows was the colonisation by mosses, probably Rhacomitrium spp., which preceded other vegetation in most cases. The influence of vegetation could have been fairly quickly detected, especially the influence of the Nitrogen fixing species, such as Dryas, which carry nitrogen fixing bacterial nodules upon their roots.

A further influence of vegetation is the loosening effect upon the soil surface, and, in the case of the woody species, Salix, and Vaccinium spp., to lower levels. However, this feature is subject to considerable variation, and vegetation is by no means the sole influencing factor. Species comparisons with other work reveals remarkable correlation in coloniser species. At Mt.Robson, for instance, Tisdale et al (1966), found very many of the same species occupying the same ecological niche.

Chapter 4: Cluster Analysis of Correlation Coefficients

Paired Combinations of Factors

 

4:1 Method

Cluster analysis is a method which has a number of variations of technique, all of which hinge on the ability to produce numerical values to represent the closeness of relationships between two factors, which may be of differing nature. In a sense, it allows statistical comparison of apples and oranges.

Either a correlation coefficient may be used, or a standardised M-space Euclidean distance coefficient, as described by J.C.Davis (1973), The first step in cluster analysis is to produce a matrix of correlations, in this instance done by using the Pearson Product Moment correlation “r”. (See figures 13,14).

From this matrix, the highest positive mutual correlation is selected, the two elements then being combined to form the first group of variables. The matrix is then recomputed, using simple arithmetic averaging. Thus, the new correlation between a hypothetical group AB, and a third variable C, is provided by the following;

Once the matrix has been recalculated, the process is repeated until all the individual elements in the analysis have been subsumed into one of the final two groupings. The final matrix, of course, will be a 2 x 2 matrix between the last remaining clusters of elements.

A useful method of displaying the results of this method is by producing a dendogram of correlation levels. ( See diagram 15) It must be remembered, however, that the dendogram will only illustrate the general relationships between elements, since these relationships will become increasingly distorted by the averaging process. The degree of distortion can be seen by producing a matrix of Cophenetic values, which shows the apparent correlations contained within the dendogram, (see figure 16).

A useful means of displaying this distortion is to plot the cophenetic values against the original values, as in figure 17, for, if both matrices were the same, then a straight line would result from the plotting of these values, however, deviations will inevitably result, and the distance of plotted points from the ideal straight line shows the degree of distortion incurred. Distance above of below the straight line shows the positive or negative distortion involved. The actual mechanics this method are described at some length, and in detail, in Davis, (1973).

4:3 Analysis of Results

As can be seen from the dendogram in Figure 15, the factors investigated tend to group themselves in two main groups, those associated with distance downslope, and those associated with distance from the ice. Further, that these two main trends seem to be in opposition to one another, since their resultant correlation level is a negative one.

In the first group, the group associated with downslope distance, we can see fairly strong associations between mean pH values and bulk density values, however, it is probable that this arises from the influence of distance downslope upon them, in the main, although since vegetation will not become established in soils of very high density, over 2000 kg/m3, (Donahue et al,l971), then the influences of organic acids will be absent. It is in many ways strange to find bulk density negatively associated with distance from the ice front, since this conflicts with results obtained by Crocker and Major,(1955), where vegetation played a strong role in reducing densities. This phenomenon is more clearly explained in the following chapter. The main cause of this distortion of relationships is probably the relatively minor role, according to this analysis, of vegetation, becoming correlated to this group only at a low level.

The angle of slope is of only minor significance too, although it would have been thought to have played a significant part in determining vegetation establishment.

The change in pH values through the profiles,pH2, shows an association with distance from the ice, although only at a lower level. Nevertheless, the development of discrete, albeit minor alterations in horizon pH values is a concrete indicator of soil development.

Although the mean organic matter has not been included in this part of the analysis, it seems likely that it would play a part in the development of these differences in horizon pH, despite the fact that vegetation appears to decline with distance from the ice.

Some doubt, however, can be thrown on this- part of the analysis, by referring to the above Figure. It can be clearly seen that the extent of deviation from the straight line is large, and includes a large number of variables. This represents a great degree of distortion being introduced by the analysis of the coefficients. The majority of them being apparently associated at much higher levels than they actually are.

The probable main cause of this distortion, and therefore of the element of distrust of this part of the analysis, is due to not having converted the values of the different variables into z-scores, that is, statistical values to describe the amount of deviation of an individual value from the mean of the values of that group in numbers of standard deviations. With this doubt thrown on the results, it seems of little value to continue with their analysis, but to proceed to a more reliable form of analysis in the next chapter.

Chapter 5: Stepwise Multiple Regression

5:1 Method

Since the data collected for this study is of far too extensive a nature to enable successful interpretation of interactions between components in any simpler way, a stepwise multiple regression program for computers was used.
The program used was a derivation of one used in the PET microcomputer, but the capacity of this facility is limited to 240 “units”, that is, in this instance, that computation capacity was limited to combinations not exceeding 30 samples, and 8 variables, (8 x 30 = 240). The program was therefore adapted to the Systime 1000 system, of much larger capacity.
The program consists of the operator selecting a dependant variable, which the computer then matches up with the remaining variables in successive combinations, in increasing order of correlation. Beyond stages where the increase in the correlation coefficient value is less than .002, Davis, (1973), suggests that the significance is minimal. However, the results are recorded in toto in Table 5.

5:2 Analysis

The actual analysis of the data proved to be difficult to get under way. This was partly due to the author’s less than comprehensive knowledge of computational techniques, but also partly due to the presence of “bugs” in the program, which arose in the process of translation from one system to another. Once these “bugs” had been, overcome, the analysis was rapid, and reliable, in the context of the data used.
Multiple regression analysis is computed after the formulation of a “z” matrix, which allows the computer to handle the information more rapidly. The regressions are calculated according to the general formula;

5:3 Results and Analysis of Results

It is important to be aware that the cumulative “r” in the results does not take account of the sign of the correlation, therefore, for guidance, the sign of the original Pearson Product moment correlation has been inserted.

Table 5

It must be borne in mind, that due to the nature of the results received, analysis must of necessity be fragmented and complex. It is far beyond the capacity of the human brain to evaluate and conceptualise the complexity of the interactions between all components, therefore analysis has to be of the nature of describing trends observed from data, with incidental comment on the derivation of these trends, where relevant. Because of the massively complex task of interpretation of this information, apologies are given in advance for any inadvertent short comings in the analysis.

It appears that two major trends are involved which are associated, firstly those which are associated with distance downslope and those associated with distance from the icefront. (See figure 18).

With regard to factors strongly associated with distance downslope, the development of vegetation is most strongly evident. This may be for a number of reasons. In this environment winter frost heaving is most strongly associated with areas of little or no snow cover, conversely, the areas beneath snow cover, that is, the hollows are little affected by this disturbance, thus giving vegetation a more significant chance of establishment, and of protection from the harsh wintered climate, particularly freezing temperatures. In addition, the regular disturbances occurring upslope with frost heaving may increase the rate of downwashing of available nutrients into hollows.

A second feature that appears is the tendency for moisture content to decrease downslope. This is more difficult to analyse, but is probably likely to be due to the variations in texture which exist. The samples which were used to determine moisture content were taken in the upper ten to fifteen cm of the soil and because of frost action there is a tendency for silt and clay sized particles to predominate here especially on the upper parts of slopes. It is probable that the variation in moisture content arises from this alone, the smaller sized particles having a tendency to hold water more strongly by adsorption than the coarser material at the surface in the hollows. The vegetation cover in the hollows however tends to be so small in absolute terms, that is has little effect on water retention in the soils.

Because bulk density appears to increase downslope it may indicate that the effects of vegetation on decreasing bulk density are less than had been thought although the effect is still detectable, it may be more true to say that frost action has more significant effect upon soil density than vegetation does.The strong association of mean pH with distance downslope presents an unusual problem. According to the data, pH increases with distance downslope, however, this also means that pH increases with vegetation cover. This can be viewed in two lights, firstly, in complete contrast to what is expected, since pH tends to decrease with vegetation (Thompson and Troeh; 1978; pl69) since the organic acids produced, although increasing cation exchange capacity,tend to decrease per cent base saturation and pH. However, an alternative perspective may be adopted. If the rate of decrease of pH on bare soils exceeds the rate of decrease under vegetation, as could possibly happen, the pH values will statistically appear to increase downslope, not because they actually are increasing, but because the rate of decrease is slower. The effect leaching on bare soils of this type is illustrated in Crocker; (1960). In this respect bases released from organic matter and from accelerated weathering of the soil may be enough to prevent the soil from becoming acid (Thompson and Troeh; 1978; pl69). Therefore on balance it seems the latter explanation is more valid.

The trend most noticeable in terms of distance from the ice front is the decrease of the mean pH. Despite the variation with respect to distance downslope, it is probable that vegetation is the single most significant factor in this context, although the relationship is by no means clear cut since vegetation appears to increase only slightly with distance from the ice. However, the distribution of vegetation is highly affected by distance downslope and the proportion that hollows occupy in the landscape may not be accurately reflected in the sampling. In addition, leaching of the bare surfaces, is still occurring although probably at slower rates as time progresses. Bulk density also varies strongly with distance, a combination of vegetational and frost action effects.

One element of which little notice has hitherto been taken is the action of soil fauna. Earthworms, from an initial survey are few in number, therefore loosening and mixing of soil is likely to be affected slightly more by insects and other hard shelled fauna, which appeared to be more widely distributed in the later stages of the transect.

Moisture content tends to increase along the transect partly as a result of vegetational colonisation in hollows, (which although appears of relatively lesser importance in individual instances, will have effect in a longtitudinal context along the transect). The moisture content, as noted before, will also be affected by the concentrations of silt and clay sized particles on ridge tops.

Organic matter content decreases with distance, although only very slightly. This is only a slight departure from the marginal increase of vegetation with distance so the association is not completely without foundation in this instance. However, analyses involving organic matter results are becoming increasingly difficult to explain in this study as variation become more complex. Where organic matter does show a logical and expected association is in the change in pH through profiles. The increase in carbonic acids (which occurs as does organic matter, in upper horizons), produces a corresponding increase in variation of pH between horizons. A further, unexpected association is a negative one between organic matter and moisture content, that is in the main, explained by the anomalies occurring in moisture content with fine particle accumulation in upslope areas, which then produces the apparent association of decrease of moisture content with increase of vegetation, whereas this in fact is not the case, but is obscured by stronger relationships between particle size and moisture, as explained before.

Some factors appear to have stronger relationships with factors other than distance or percentage downslope. For instance, the change in pH between horizons, although exhibiting a wider differentiation between horizons with distance from the ice, and smaller differentiation with increased distance downslope the relationship with bulk density is much more pronounced. The changes between horizons decreases with a decrease in the density of material it is possible that this may be a distortion introduced by the reversals of pH trend which occur in the higher density soils near the ice front, however since this phenomenon occurs in only a few profiles the effect will be limited in a sample size of 30.

The angle of slope, which is generally higher near the ice front shows quite strong effects on some factors. Vegetation and moisture content both decrease with slope angle both of which are understandable in the context of the area. Slope instability and free draining characteristics combine to reduce vegetation and moisture on high angle slopes. Average pH also increases, with slope angle possibly due to the absence of vegetation, but mainly as a relative phenomenon, that is, it decreases less quickly than areas which are vegetated.

This section of the study shows quite clearly that much more complex mechanisms are operating than would at first appear in single factor analysis. However at times the explanation and interpretation of these mechanisms requires much more information than is available since data alone are insufficient for a worthwhile analysis of soil forming processes and almost a sense of intuition regarding the interrelationships of these mechanisms becomes necessary. Never-theless the use of the stepwise regression program has proved both illuminating and instructive in revealing major trends and the minor complexities which operate within the soil forming processes, even at such early stages.

Chapter 6: Analysis of Variance on Directional Data

6:1 Method

The compass bearings of each pit site, or to be more precise the bearing of the slope upon which the pit was located, was divided into eight quadrants, as can be seen from Table 6, and the vegetation cover data also inserted in the matrix.
It had become increasingly obvious that a statistical treatment of this type would be necessary to analyse the effects of aspect upon, primarily, vegetational development, since directional data were unusable in the multiple regression program.
The problem mainly arises from the circular nature of the data. The major drawback can be illustrated by an extreme example, if for instance, the observed values in a sample size of two are 359 degrees and 1 degree, then the arithmetic mean and variance will give absurd results. “Therefore,” as Mardia, (1972; pl8) remarks, “the usual linear measures are inappropriate to circular distributions “. In short, there was no easy and reliable method for transforming data of a circular nature for use in a linear regression program.
The method is both well known and widely used, the analysis depending upon the probability of the sample populations arising statistically from the same population. If the probability is high, then the chances are, in this case that bearing would have no significant effect upon vegetation.
The analysis of variance can be both tedious and time consuming, therefore the “ANALV 1” cassette program for the PET microcomputer was used for speed of computation. Later, as will be seen, the Student’s T-test program was used in the same facility.

Therefore, the probability of the sample populations arising from the same statistical population is only 43.387%. However, far from confirming that bearing has a strong influence upon vegetation, it leads us to question that assumption, since the variation between the groups is not as great as the variation within the groups themselves, since the V-ratio, i.e.

and moreover, does not exceed tabulated F. There fore, it appears that variation is greater within blocks than between them.
To test whether this variance arose in any particular sector, the blocks were further split into North and South facing sectors, and later into East and West facing sectors, and the Students T-test applied to trace the source of the variance, the results were as follows

Meaning that the chances of Northern data differing from Southern data only by chance are 81.343 %

Meaning that the chances of Eastern data differing from Western data only by chance are 17.973 %

Therefore, we can see that the main source of variation in the data arises mainly along an East-West axis, with little variation along the North-South axis.

This type of variation may seem difficult to interpret, since what is expected is a significant variation on a North South axis, corresponding to variations in the amount of insolation which a particular slope receives. However the apparent disparities can be explained in two main ways.

Firstly, there is a constant downflow of cold air from the glacier which must lower surface temperatures considerably, although no data are available, it seems a reasonable assumption to make. The axis of the glacier lies mainly in a North-East, South-West direction which must surely further confuse the effects of the cold air upon vegetation.

However, the major effect which confuses the effect of insolation must be the topography of the area. It has already been noted in the Introduction and especially in Figure 4, that the topography is extremely hummocky, and the main source of variation is likely to arise from this. The effects of shadowing from the ridge and hollow topography upon vegetation are even more important when one considers that the main sites for colonisation are in the bottom of the hollows. Therefore, not only do they suffer from shadowing from the surrounding ridges, but this shadowing effect is likely to be an extremely complex one. The degree of complexity leads us on then, to a final, major problem.

This, quite simply is the question of whether 30 randomly chosen sites can accurately reflect the true variations in differential insolation. On the basis of the results given by the data collected, it seems that this objection is upheld, notwithstanding the acknowledged complexity of the situation. In this instance the wide range of factors affecting the vegetational cover would seem to merit a much wider range of sampling cover, and, if nothing else, this part of the study has established that there is a wide and complex series of factors in operation, and the concomitant difficulty of establishing a clear relationship.

Chapter 7: Summary and Conclusions

7:1 Summary

This study has established that vegetational colonisation in recently deglaciated terrain follows a much more complex pattern than may have been inferred from previous authors.

This is reflected in the higher colonisation in the hollows of the hummocky moraine topography. Indeed, topographical influences are remarkably strong in influencing soil textures, promoting, through frost action, a concentration of smaller sized particles along ridges, whilst material in the hollows is of a much more coarse nature. This effect incidentally is reflected too in the moisture in the soil, being differentially affected by the capacity of silt and clay sized particles to hold water by adsorption to their surfaces.

Changes in pH over the area appear to be confused by the topography too, since leaching is occurring on bare surfaces which are normally the higher and steeper slopes in this terrain. Differences in pH values between horizons is more strongly affected on areas of vegetational colonisation, understandably, through the influences of organic acids produced.

However, some overlap of influence appears to exist in the decline of pH values along the transect, since the expected decline in pH suggested by Crocker (1960), and Crocker and Major (1955), with surface age, and with vegetational cover is confused by the decline in pH caused by leaching. Nevertheless, the influences can be separated. Vegetational colonisation with age of sunface, as suggested by the above authors also appears to be a more complex influence, as suggested above.

One influence that is quite easily detected is the change in bulk densities along the transect. This is not however, solely due to the influence of vegetation, which as has been shown, is quite variable, it is also partly due to the action of winter frost heave, and probably to the action of soil fauna. As well as there being a distinct lontitudinal effect on soil densities, the topographical influence is strong too, with higher densities along ridge tops.

The manner in which these influences have been detected has been to investigate each factor or element individually, and establish the apparent trends of each element in relative isolation. The next step was to combine each variable with all other variables, in a method of cluster analysis to try and identify some of the main associations implied in the data. However this part of the analysis was not entirely successful, in main due to the manner of handling of the data, but also due to the nature of the data itself. It did however, establish the main trends in the data collected, which were amplified by the multiple regressions.

The stepwise multiple regressions on the data showed a degree of complexity in the interactions between elements which in cases was difficult to explain, and on the basis of the small amount of information upon which the analysis was based, only tentative conclusions could be drawn.

However, a great amount of insight into the interactions of most of the elements could be discerned from this type of analysis, and undoubtedly helped in interpreting the complexity of the associations of factors involved in the formation of soils at an early stage.

Finally, it was seen that in the case of Svinafell, at any rate, there existed such a plethora of influences upon the establishment of vegetation, that the influence of aspect, thought to be a major factor, was impossible to isolate, although a larger amount of information could, it is thought, provide a much more clear picture of the degree of influence exerted by this factor.

7:2 Conclusions

Having completed the fieldwork, and the analysis of the data collected, it is now necessary to assess the limitations and values of this study.

Despite the fact that some problems and criticisms have been mentioned and discussed in each section, one or two major criticisms may be levelled against this work.

In the first place, it can be said that, given the complexity of the area in which the work was done, and, moreover, the complexity of the associations which were detected, can a sample of 30 pits reliably and accurately reflect these sometimes confusing interactions? In many ways, this must be warranted criticism, especially when the sites were randomly chosen in an area of wide ranges of variability in all factors.

To attempt to justify this point, it must be said that time was limiting, and therefore only a certain amount of work could be completed in the time available. It is also true to say that random sampling was chosen because the author did not have the necessary experience and expertise to reliably choose sites which would reflect the patterns of the area, and the processes involved. Undoubtedly, in any future work, sites would be both randomly sampled and chosen subjectively, to enable the work to encompass as many variations as possible, in an attempt to achieve a more reliable data base.

Secondly, following on from this, given the limited sample for the data base, it could be said that the analysis with which the data has been treated is not warranted by the nature of the data. Again, this is a warranted criticism. However, very little could have been inferred from the data without extensive analysis, since, limited though the data is, relatively speaking, it nevertheless represents a considerable body of information which would be exceedingly difficult if not impossible to unravel without relatively sophisticated analytical techniques.

Where, in some cases the analysis has been of additional use has been where it actually underlines the need for more information, as in Chapter 6. Without analysis the data could only have been useful in a descriptive frame-work, therefore it is only reasonable to say that the analysis has been useful to identify the range and scale of its variation, and that results should be viewed in the limitations which the data places upon them, but not to decry the use of extensive analysis in a possibly marginal situation.

On a more practical than conceptual level, two main problems have arisen. Firstly, a fault was made in accepting organic matter data as accurate in a relative sense. That is to say, despite the difficulty of achieving absolute measures for organic matter, it was thought that relative measurements could be attained. As analysis continued, the pattern of organic matter has become increasingly difficult, and confusing, to explain. This, it is now considered, has been primarily due to inaccurate data.

Given the difficulties inherent in dependence on this data, it would probably have been better to have brought samples back to England and to carry out proper ignition tests. The necessity of planning and testing useful techniques before fieldwork cannot be pressed too strongly and, in this respect, a resounding failure must be recorded.

Secondly, the work in Chapter 4, in retrospect must have the limitations mentioned in that chapter underlined.

This technique, proven in psychology and expanded in geology by Davis (1973), is sound and reliable. However, this only applies where data is handled correctly, and, with such wide variations in data, it would obviously have been better to have converted the data to z-scores, despite the limitations of the data itself.

In terms of where this study has achieved something of value, we must look to Crocker (1952). A major point which this study has underlined, it is hoped, is the necessity for a linkage of factors and processes in soil genesis. As far back as 1952, Crocker stated that

“work in soil genesis is for the most part concerned with soil dependence upon environmental factors, or with the actual processes involved in the formation of soil material, but rarely with both at the same time.”

It seems painfully obvious from the literature that this boundary is rarely crossed, even now. If nothing else, this study has been an attempt to do that, by not only identifying the processes at work in the soil at an early stage, but also the parameters, determined by environmental factors, within which these processes operate.

To turn to the hypotheses which motivated this study, mentioned in the Introduction, it must be admitted that the first hypothesis, that of the possibility of developing an expression as a guide to the mechanisms involved in soil formation, has been left open ended. Clearly, it has been an over ambitious objective for this type of study, but, it is hoped, if more work in the area becomes possible, then this objective could still be realistically achieved.

The secondary motivation, of assessing the benefits and problems of using multivariate analysis in this area of work, has, it is hoped, been adequately covered.

The analyses used, particularly the multiple regression have been of great value in tracing and unravelling the interactions involved, interactions which would have been far more difficult to identify if analysis had been limited to individual elements.

In conclusions, it is customary to allude to any further areas of work which may be of interest, and connected to the work in hand, and one glaring omission from the published literature is a time-based study of a whole range of variables. Both Cooper (1916;1923;1939), and Lawrenc (1953;1958), have spent time in investigating vegetational successions and effects, but little published work is available on the wider influences, either due to vegetation, or from other sources (e.g.,microclimate, topography ),with the notablet;but dated exception of Crocker and Major, (1955).

Finally, to return to the concept outlined by Crocker (1952), more recently Birkeland,(1974), has stressed the need to distinguish more clearly between factors and processes, a view it seems which would lead even more to a situation where research is devoted to effects of environmental factors, cr to mechanisms and processes of soil formation. However, lately, more and more, as clearly pointed out by Paton, (1978, p 107), to solve the problem of soil formation it is seen as

“necessary to acknowledge the full complexity of the problem, and attempt to solve it in a holistic, rather than particularate manner.”

It is hoped that this work has been a step in that direction.

Bibliography

Birkeland, P.S.,1974, Pedology, weathering, and geomorphological research; Oxford Univ Press
Bishop, O.N., 1966, Statistics for biology; Longman
Cooper, W.S., 1916, Plant successions in the Mt. Robson area of British Columbia; The Plant World; 19;211-238
Cooper, W.S., 1923, The recent ecological history of Glacier Bay, Alaska; II, The present vegetation cycle; Ecology;4;233-246
Cooper, W.S., 1939, A fourth expedition to Glacier Bay, Alaska; Ecology;20;130-155
Crocker, B.L.,1952, Soil genesis and pedogenic factors; Quart. Rev. Biol.;27;139-168
Crocker, R.L.,1960, The plant factor in soil formation; in Proceedings, 9th Pacific Sc. Cong., Vol 18;84-90; Pub.,Secretariat,9th pac. Sc. Cong.,Dept, of Science, Bangkok.
Crocker, R.L., & Major, J. 1955, Soil development in relation to vegetation and surface age at Glacier Bay, Alaska;J.Ecol.;43;428-448
Daubenmire, R.1959, A canopy coverage method of vegetational analysis;Northwest Scientific; 33;43-66
Davies, J.C., 1973, Statistics and data analysis in geology; G. Wiley & son.
Dickson, B.A. & Crocker, R.L. 1953, A chronosequence of soils and vegetation near Mt. Shasta, California;Journ.Soil Sc. 4;2;142-154
Donahue, R.L., Shickluna,J.C. Robertson, L.S. 1971, Soils, an introduction to soils and plant growth; 3rd Edition; Prentice Hall
Fitzpatrick, E.A.1971, Pedology, a systematic approach to soil science; Oliver & Boyd, Edinburgh.
Harris, J.C., 1973, Some factors affecting rates and processes of periglacial mass movements; Goegrafiska Annaler;55a;1;24-28
Hill, d.e.,& Tedrow, J.C.F. 1961, Weathering and soil formation in the Arctic environment; Amer. J. Soil Sci; 259;84-101
Howarth,P.J. Some brief notes on the excursion to the outlet glaciers of Oraefajokull; Mimeographed report, Univ. of Glasgow
Jenny, H. 1941, Factors in soil formation; New York,McGraw-Hill
Lawrence, D.B.,1953, Development of vegetation and soil on deglaciated terrain of S.E. Alaska;Mimeo.report, Univ. of Minnesota;39PP
Lawrence, D.B.,1958, Glaciers and vegetation in S.E. Alaska;Amer. Sci.;46;89-122
Mardia, K.V. Norris,J.M.& Dale, M.B. 1972, Statistics of directional data ;Academic Press 1971, Factor analysis and numerical taxonomy of soils;Soil Sci.Soc.of Amer.;35;3;487—491
Paton, T.R., 1978, The formation of soil material; George Allen & Unwin
Retzer, J.L., 1965, Present soil formation factors in arctic & alpine regions; Soil Science;99,38-44
Russell,E.W., 1961, Soil conditions and plant growth; London, Lowe and Brydone
Smith, R.T.,& Atkinson, K. 1975, Techniques in pedology; Paul Elek Scientific Books
Taylor, J.A., 1960, Methods of soil study;Geography;45;52-67
Thompson, L.M.& Troeh, F.R. 1978, Soils and soil fertility; 4th Ed;McGraw-Hill
Tisdale, E.W., Fosberg,M.A.,& Poulton,C.E. 1966, Vegetation and soil development on recently deglaciated terrain near Mt. Robson,British Columbia;Ecology;47;4;517-523
Webster, R. 1977, Quantative and numerical methods in soil classification and survey;Oxford Univ Press


Appendix 1: Soil Profiles and Descriptions Pits 1 – 30

N.B. Exaggeration exists on vertical scale
horizontal scale is 1 : 10
vertical scale is 1 : 5

Pit No 1

Grid Co-ordinates: 04,25
Slope Angle: 11
Slope Bearing: 323
Slope Type: Concave
Distance Downslope: 100%

PROFILE

I Texture: Very gritty, almost all sand.
Structure: Crumby, medium, moderate. Consistency: (moist) non sticky, non plastic.
Stoniness: Very stony, small stones, subangular.
Colour: 10YR 2/2 Narrow Boundary

II Texture: Very gritty, almost all sand.
Structure: Crumby, fine, moderate. Consistency: Slightly sticky, non-plastic.
Stoniness: Very stony, small stones, subangular.
Colour: 10YR 2/1

ORGANIC MATTER: None. No visible roots.

Free, but lower 10cm still waterlogged after heavy rain raised water level in adjacent stream.


 

Pit No 2

Grid Co-ordinates: 20,07
Slope Angle: 9.5
Slope Bearing: 311
Slope Type: Concave
Distance Downslope: 100%

PROFILE

I Texture: Very gritty, almost entirely sand, can maintain no shape when moist.
Structure: Crumby, fine, weak.
Consistency: Slightly sticky, non-plastic when moist.
Stoniness: Stony, medium stones, subangular.
Colour: 5Y 3/1

Merging Boundary
II Texture: Very gritty, almost all sand.
Structure: Structureless mudflow, (waterlogged).
Consistency: Sticky, non-plastic.
Stoniness: Slightly stony, gravel, subangular.
Colour: 10YR 4/1

ORGANIC MATTER: None, no roots visible.
DRAINAGE: Free to 30cm, waterlogged below.
N.B. The banding in this profile did not constitute a separate horizon in itself, but was more accurately, a part of the upper horizon, where layers of coarser sand had been intermittently laid down by streamflow. This banding was entirely sand, loose and granular with no stones; colour 7.5Y 2/1


 

Pit No 3

Grid Co-ordinates: 49,25
Slope Angle: 20
Slope Bearing: 313
Slope Type: Straight
Distance Downslope: 10%

PROFILE
I Texture: Soft and gritty, with slight cohesion when moist.
Structure: Granular, fine, moderate.
Consistency: Friable, with weak cementation.
Stoniness: Very stony, large stones, subangular.
Colour: 7.5YR 2/1

Merging Boundary
II Texture: Friable, slightly plastic when moist.
Structure: Crumby, fine, moderate.
Consistency: Slightly plastic slightly sticky.
Stoniness: Stony, large stones, subangular.
Colour: 10YR 2/2

ORGANIC MATTER: No visible roots, or organic matter in profile.
DRAINAGE: Free.


 

Pit No 4

Grid Co-ordinates: 74,10
Slope Angle: 7
Slope Bearing: 323
Slope Type: Concave
Distance Downslope: 70%

PROFILE

I Texture: (dry) soft and gritty, slight cohesion when Miost.
Structure: crumby, fine, strong
Consistency: soft, weak, cementation (dry)
Stoniness: stony, medium stones, subangular
Colour: 10YR 2/2

Merging Boundary
II Texture: Plastic (moist) but only moderately gritty
Structure: Crumby, moderate, strong
Consistency: Very plastic, sticky (moist)
Stoniness: Slightly stony, large stones, rounded
Colour: 10YR 3/2

Sharp Boundary
III Texture: Very gritty, almost all sand
Structure: Granular, fine, strong
Consistency: Loose with weak cementation
Stoniness: Stoneless
Colour: 5Y 2/1

ORGANIC MATTER: Roots common in upper horizon, but few below, fine roots stretching to 40cm.
DRAINAGE: Free


 

Pit No 5

Grid Co-ordinates: 82,04
Slope Angle: 22
Slope Bearing: 281
Slope Type: Concave
Distance Downslope: 20%
Pit inspected 12h after heavy precipitation (.4 Inches in 24h)

PROFILE
I Texture: Friable, slightly plastic when moist.
Structure: Crumby, medium, moderate.
Consistency: Friable with weak cementation.
Stoniness: Stony, large stones, subangular.
Colour: 10YR 2/2

Sharp Boundary
II Texture: Soft and gritty, slight cohesion when moist.
Structure: Granular, fine, strong.
Consistency: Soft with weak cementation.
Stoniness: Stoneless.
Colour: 7.5YR 3/2

Merging Boundary
III Texture: (Moist) slight cohesion.
Structure: Crumby, medium, moderate.
Consistency: Slightly sticky, slightly plastic.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/2

ORGANIC MATTER: Roots common in upper horizon to 15cm., few below, and absent in lower horizon.
DRAINAGE: Free.


 

Pit No 6

Grid Co-ordinates: 98,11
Slope Angle: 0
Slope Bearing: 157
Slope Type: Convex
Distance Downslope: 0%
Colour: 10YR 3/1
Pit inspected 12h after heavy precipitation (.4 inches in 24h)

PROFILE
I Texture: (moist), sticky and slightly gritty.
Structure: Crumby, fine, strong.
Consistency: Slightly sticky, plastic.
Stoniness: Stony, nediun stones, subangular.
Colour: 10YR 1.7/1

Merging Boundary
II Texture: (moist), slightly plastic.
Structure: Crumby, medium, Moderate.
Consistency: Slightly sticky, plastic.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/1

ORGANIC MATTER: Few roots to 10cm, some fine roots extending to 20cm (rarely). Absent below 20cm.
DRAINAGE: Free.


 

Pit No 7

Grid Co-ordinates: 103,04
Slope Angle: 17
Slope Bearings: 247
Slope Type: Convex
Distance Downslope: 30%
Profile inspected 12h after heavy rainfall (.4 inches in 24h).

PROFILE
I Texture: (moist) slightly plastic, probably high silt fraction.
Structure: Crumby, fine, strong
Consistency: Slightly sticky, slightly plastic, (moist)
Stoniness: Stony, medium stones subangular.
Colour: 10YR 2/2

Merging Boundary
II Texture: Slightly plastic, (moist) as I
Structure: Crumby, medium, moderate.
Consistency: Slightly sticky, slightly plastic, (moist).
Stoniness: Stony, large stones, subangular
Colour: 10YR 2/3
ORGANIC MATTER: Roots, few in top horizon to 10cm, absent below.
DRAINAGE:


 

Pit No 8

Grid Co-ordinates: 120,20
Slope Angle: 10
Slope Bearing: 71
Slope Type: Concave
Distance Downslope: 25%

PROFILE
I Texture: Soft. but slightly gritty, sandy laom type material.
Structure: Crumby, fine, moderate.
Consistency: Soft, with weak cementation.
Stoniness: Slightly stony, small stones, subangular.
Colour: 10YR 2/1

Merging Boundary
II Texture: (moist) slightly plastic, similar to I
Structure: Crumby, medium, moderate.
Consistency: Friable, weak cementation.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/1

ORGANIC MATTER: Roots common in upper horizon to 8cm, few below, absent in II.
DRAINAGE: Free.


 

Pit No.9

PROFILE
I Texture: Soft and gritty, silty sand slight cohesion when moist.
Structure: Crumby, medium, stoney.
Consistency: Friable, weak.
Stoniness: Stony, small stones, subangular.
Colour: 10YR 2/1

Boundary Merging
II Texture: Sandy silt, slightly plastic
Structure: Crumby, fine, medium.
Consistency: Non sticky, slightly plastic.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/2

Drainage: Free
Organic Material: Roots, few to 10 cm, rare below.


 

Pit No.10

 
Grid. Co-ordinates: 162,06
Slope Angle: 8
Slope Bearing: 349
Slope Type: Convex
Distance Downslope: 40%

PROFILE
I Texture: Soft and gritty silty sand type material, slight cohesion when moist.
Structure: Crumby, medium, strong.
Consistency: Friable, with weak cementation.
Stoniness: Stony, small stones, subangular.
Colour: 10YR 2/3

Narrow Boundary
II Texture: Sandy silt type material, slightly plastic (moist).
Structure: Crumby, fine, moderate.
Consistency: Non sticky, slightly plastic.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 2/2

ORGANIC MATTER: Roots, few to 8cm, in I, absent in II.
DRAINAGE: Free.


 

Pit No.10

Grid Co-ordinates: 172,03
Slope Angle: 11
Slope Bearing: 363
Slope Type: Convex
Distance Downslope: 20%

PROFILE
I Texture: Soft and gritty, slight cohesion when moist
Structure: Crumby, fine, strong.
Consistency: Friable with weak cementation.
Stoniness: Slightly stony, medium stones, subangular.
Colour: 10YR 3/2

Merging Boundary
II Texture: Friable, slightly plastic when moist.
Structure: Crumby, medium, moderate.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 4/2

ORGANIC MATTER: Roots common to absent below
DRAINAGE: Free.


 

Pit No.12

Grid Co-ordinates: 198,22
Slope Angle: 4
Slope Bearing: 271
Slope Type: Concave
Distance Downslope: 25%

PROFILE
I Texture: Silty material, friable but slightly plastic when moist.
Structure: Crumby, fine, strong.
Consistency: Loose, with weak cementation.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 2/2

Narrow Boundary
II Texture: Friable, slightly plastic when moist.
Structure: Crumby, fine, strong.
Consistency: Soft, weak cementation.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/2

Narrow Boundary
III Texture: Soft and gritty.
Structure: Crumby, fine, moderate.
Consistency: Slightly sticky, slightly plastic.
Stoniness: Stony, small stones, subangular.
Colour: 10YR 4/3

Narrow Boundary
IV Texture: Friable, slightly plastic when moist.
Structure: Crumby, fine, moderate.
Consistency: Slightly sticky, slightly plastic.
Stoniness: Stony, small stones, subangular,
Colour: 10YR 3/3

ORGANIC MATTER: Roots few in I to 6cm., rare in II, absent below. Roots woody in upper horizon.
DRAINAGE: Free.


 

Pit No.13

Grid Co-ordinates: 210,14
Slope Angle: 10
Slope Bearing: 275
Slope Type: Straight
Distance Downslope: 20%

PROFILE
Texture: Silty material, soft and gritty.
Structure: Crumby, fine, strong Consistency: Friable, weak cementation.
Stoniness: Stony, large stones, subangular.
Colour: 10YR 2/2

Narrow Boundary
I Texture: Very gritty, almost all sand.
Structure: Granular, fine, strong.
Consistency: Loose, weak cementation.
Stoniness: Stony, large stones, subangular.
Colour: 10YR 1.7/1

Narrow Boundary
II Texture: Very gritty, almost all sand.
Structure: Granular, fine, strong.
Consistency: Soft, weak cementation.
Stoniness: Stony, large stones, subangular.
Colour: 2.5 YR 4/3

Narrow Boundary
III Texture: Soft and gritty Structure: Crumby, medium, moderate.
Consistency: Soft, weak cementation.
Stoniness: Stoneless.
Colour: 10YR 3/2

Narrow Boundary
IV Texture: Soft and gritty
Structure: Crumby, medium, moderate
Consistency: Soft, weak cementation
Stoniness: Stoneless
Colour: 10YR 3/2

ORGANIC MATTER: Roots few to 10cm., rare to 17cm, absent below.
DRAINAGE: Free.


 

Pit No.14

Grid Co-ordinates: 243,20
Slope Angle: 12
Slope Bearing: 271
Slope Type: Straight
Distance Downslope: 10%

PROFILE
Texture: Friable and gritty.
Structure: Crumby, medium, moderate.
Consistency: Soft, weak cementation.
Stoniness: Stony, small stones, angular.
Colour: 10YR 2/3

Narrow Boundary
I Texture: Very gritty, almost entirely sand.
Structure: Granular* fine* strong.
Consistency: Soft, weak, cementation.
Stoniness: Very stony, medium stones, subangular.
Colour: 2.5Y 4/4

Mottles: Texture, structure, consistency, all the same as horizon II.
Colour: 7.5 YR 5/8. Colour almost certainly due to in situ breakdown of iron-bearing materials present in greenish agglomerate.

ORGANIC MATTER: Roots common to 10cm in upper horizon* rare below.
DRAINAGE: Free.


 

Pit No.15

Grid Co-ordinates: 288,19
Slope Angle: 12
Slope Bearing: 39
Slope Type: Concave
Distance Downslope: 100%

PROFILE
Texture: Sticky and gritty (wet).
Structure: Granular, fine, moderate.
Consistency: Sticky, non-plastic.
Stoniness: Very stony, gravel, subangular.
Colour: 10YR 2/3

Merging Boundary
I Texture: Forms flowing mass, water-logged, moderate amount of coarse sand and gravel.
Structure: Structureless, fine mud with sand and gravel.
Consistency: Non-sticky, non-plastic.
Stoniness: Stony, gravel, subangular.
Colour: 10YR 3/2

ORGANIC MATTER: Roots common in top 10cm., few below.
DRAINAGE: Very poor.

N.B. Due to a large amount of water in this profile, especially in lower horizon, and to gravelly nature of material, this pit collapsed almost completely after less than five minutes.


 

Pit No.16

Grid Co-ordinates: 295,19
Slope Angle: 12
Slope Bearing: 22
Slope Type: Straight
Distance Downslope: 95%

PROFILE
Texture: Sticky and gritty, (wet).
Structure: Granular, fine, moderate.
Consistency: Sticky, non-plastic.
Stoniness: Very stony, gravel, subangular.
Colour: 10YR 2/3

Merging Boundary
I Texture: Forms flowing mass, water- clogged, moderate amount of coarse sand and gravel.
Structure: Structureless, fine mud with sand and gravel.
Consistency: Non-sticky, non-plastic,
Stoniness: Stony, gravel, subangular,
Colour: 10YR 3/2

ORGANIC MATTER: Roots common in top 10cm., few below.
DRAINAGE: Very poor.


 

Pit No.17

Grid Co-ordinates: 302,04
Slope Angle: 20
Slope Bearing: 39
Slope Type: Concave
Distance Downslope: 15%

PROFILE
I Texture: Soft and gritty, (moist).
Structure: Crumby, medium, moderate.
Consistency: Slightly sticky, slightly plastic.
Stoniness: Stony, small stones, angular.
Colour: 10YR 3/2

Merging Boundary
II Texture: Soft and gritty (moist).
Structure: Crumby, medium, moderate.
Consistency: Slightly sticky, slightly plastic.
Stoniness: Very stony, small stones, subangular.
Colour: 10YR 4/2
Mottles: Sandy texture, colour 7.5 Y 5/3, due to in situ breakdown of material, rather than drainage.

ORGANIC MATTER: Roots few in top 15cm., rare below.
DRAINAGE: Free.


 

Pit No.18

Grid Co-ordinates: 344,18
Slope Angle: 26
Slope Bearing: 285
Slope Type: Straight
Distance Downslope: 50%

PROFILE
I Texture: Soft and gritty (moist).
Structure: Crumby, fine, moderate.
Consistency: Soft, weak cementation.
Stoniness: Stony, small stones, subangular.
Colour: 10 YR 3/2

Harrow Boundary
II Texture: Soft and gritty, (moist).
Structure: Crumby, medium, moderate.
Consistency: Friable, weak cementation.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 2/2
Mottles: Some very slight mottles, very small, less than 2mm., colour 7.5 6/6. Due to in situ breakdown of material.

ORGANIC MATTER: Roots common ,to 10cm., few below.
DRAINAGE: Free.


 

Pit No.19

Grid Co-ordinatess: 366,05
Slope Angle: 12
Slope Bearing: 37
Slope Type: Convex
Distance Downslope: 30%

PROFILE
Texture: Soft and gritty (moist).
Structure: Crumby, medium, strong.
Consistency: Slightly sticky, slightly plastic.
Stoniness: Stony, small stones, subangular.
Colour: 10YR 4/3

Narrow Boundary
I Texture: Friable, slightly plastic.
Structure: Crumby, medium strong.
Consistency: Slightly sticky, slightly plastic.
Stoniness: Very stony, large boulders, subangular.
Colour: 10YR 3/3

ORGANIC MATTER: Roots common to 15cm, rare below.
DRAINAGE: Free


 

Pit No.20

Grid Co-ordinates: 311,19
Slope Angle: 16
Slope Bearing: 321
Slope Type: Convex
Distance Downslope: 40%

PROFILE
Texture: Friable, slightly plastic when moist.
Structure: Crumby, medium, moderate.
Consistency: Soft, weak cementation.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 2/3

Sharp Boundary
I Texture: Very gritty, almost entirely sand.
Structure: Granular, medium, strong.
Consistency: Soft, weak cementation.
Stoniness: Stoneless.
Colour: 7.5YR 3/2

Sharp Boundary
II Texture: Soft and gritty (moist).
Structure: Crumby, medium, moderate.
Consistency: Slightly sticky, slightly plastic.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/2

Sharp Boundary
III Texture: Soft and gritty (moist)
Structure: Crumby, medium, moderate.
Consistency: Slightly sticky, slightly plastic.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/2

ORGANIC MATTER: Roots common to 15cm., few below.
DRAINAGE: Free.


 

Pit No.21

Grid Co-ordinates: 26,14
Slope Angle: 9
Slope Bearing: 141
Slope Type: Concave
Distance Downslope: 50%

PROFILE
Textures: Friable, slightly plastic when moist.
Structure: Crumby, fine, moderate.
Consistency: Soft, weak cementation.
Stoniness: Stony, small stones, subangular.
Colours: 10YR 2/1

Sharp Boundary
I Textures: Very gritty, almost entirely sand.
Structure: Granular, medium, strong.
Consistency: Loose, weak cementation.
Stoniness: Stony, gravel, subangular.
Colour: 10YR 2/2

Sharp Boundary:
II Texture: Soft and gritty, (moist)
Structure: Crumby, medium, moderate.
Consistency: Sticky and plastic.
Stoniness: Stony, medium stones. subangular.
Colour: 10YR 3/1

Sharp Boundary:
III Texture: Soft and gritty, (moist)
Structure: Crumby, medium, moderate.
Consistency: Sticky and plastic.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/1

ORGANIC MATTER: Roots common to 15cm., few below to 35cm., rare below 35cm.
DRAINAGE: Free.


 

Pit No.22

Grid Co-ordinates: 47,05
Slope Angle: 12
Slope Bearing: 147
Slope Type: Concave
Distance Downslope: 85%

PROFILE
I Texture: Soft and gritty (moist).
Structure: Crumby, fine, moderate.
Consistency: Soft, weak cementation.
Stoniness: Stony, small stones, subangular.
Colour: 7.5YR 2/2

Merging Boundary
II Texture: Soft and gritty (moist).
Structure: Crumby, fine, moderate.
Consistency: Friable, weak cementation.
Stoniness: Very stony, medium stones, subangular.
Colour: 10YR 2/3

ORGANIC MATTER: Roots common to 15cm., few below to 30cm., ra below to 30cm., rare thereafter.


 

Pit No.23

Grid Co-ordinates: 76,00
Slope Angle: 23
Slope Bearing: 237
Slope Type: Concave
Distance Downslope: 20%

PROFILE
Texture: Soft and gritty, slightly plastic when moist.
Structure: Crumby, fine, moderate.
Consistency: Soft. weak cementation.
Stoniness: Stony, small stones, subangular.
Colour: 10YR 3/2

Narrow Boundary
I Texture: Soft and gritty, slightly plastic when moist.
Structure: Crumby, fine, moderate.
Consistency: Soft, weak cementation.
Stoniness: Stony, small stones, subangular.
Colour: 10YR 2/2

Narrow Boundary
II Texture: Friable and gritty, slightly plastic when moist.
Structure: Crumby, medium, moderate.
Consistency: Slightly sticky, non-plastic.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 2/3

Narrow Boundary
III Texture: Friable and gritty, slightly plastic when moist.
Structure: Crumby, medium, moderate.
Consistency: Slightly sticky, non-plastic.
Stoniness: Stony, medium stones, subangular
Colour: 10YR 2/3

ORGANIC MATTER: Roots common to 15cm., few below
DRAINAGE: Free


 

Pit No.24

Grid Co-ordinates: 90,01
Slope Angle: 28
Slope Bearing: 226
Slope Type: Concave
Distance Downslope: 40%

PROFILE
I Texture: Soft and gritty.
Structure: Crumby, fine, moderate.
Consistency: Soft, weak cementation.
Stoniness: Stony, small stones, subangular.
Colour: 10YR 3/2

Narrow Boundary
II Texture: Friable and gritty.
Structure: Crumby, medium, moderate.
Consistency: Soft, weak cementation.
Stoniness: Stony, small stones, subangular.
Colour: 10YR 2/2

Merging Boundary
III Texture: Friable, clayey.
Structure: Blocky, (angular), fine, strong.
Consistency: (moist), sticky and plastic.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 2/1

ORGANIC MATTER: Roots common to 12cm., few below to 30cm., and rare thereafter.
DRAINAGE: Free.


 

Pit No.25

Grid Co-ordinates: 128,03
Slope Angle: 10
Slope Bearing: 199
Slope Type: Concave
Distance Downslope: 10%

PROFILE
Texture: Soft and gritty, slightly plastic when moist.
Structure: Crumby, fine, moderate.
Consistency: Soft, weak cementation.
Stoniness: Stony, small stones, subangular.
Colour: 7.5YR 2/2

Sharp Boundary
I Texture: Very gritty, almost entirely sand.
Structure: Granular, fine, strong.
Consistency: Loose, weak cementation.
Stoniness: Stoneless.
Colour: 10YR 2/2

Sharp Boundary
II Texture: Friable, (dry) slightly plastic when moist.
Structure: Crumby, medium, moderate.
Consistency: Friable, weak cementation.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/2

Sharp Boundary
III Texture: Friable, (dry) slightly plastic when moist.
Structure: Crumby, medium, moderate.
Consistency: Friable, weak cementation
Stoniness: Stony, medium stones, Subangular.
Colour: 10YR 3/2

ORGANIC MATTER: Roots few to 10cm., rare below.
DRAINAGE: Free.


 

Pit No.26

Grid Co-ordinates: 162,05
Slope Angle: 5
Slope Bearing: 101
Slope Type: Concave
Distance Downslope: 100%

PROFILE
I Texture: Friable, slightly plastic when moist.
Structure: Crumby, fine, moderate.
Consistency: Friable, strong cementation
Stoniness: Stony, small stones, subangular.
Colour: 10YR 3/2

Merging Boundary
II Texture: Soft and gritty, slightly plastic (moist).
Structure: Crumby, fine, moderate.
Consistency: Slightly sticky, slightly plastic.
Stoniness: Very stony, medium stones, subangular.
Colour: 10YR 3/3

ORGANIC MATTER: Roots abundant to 5cm., then common to 20cm., few below,
DRAINAGE: Free


 

Pit No.27

Grid Co-ordinates: 233,26
Slope Angle: 25
Slope Bearing: 225
Slope Type: Concave
Distance Downslope: 5%

PROFILE
I Texture: Friable, slightly gritty, slightly plastic when moist.
Structure: Crumby, fine, moderate.
Consistency: Loose, weak cementation.
Stoniness: Stony, small stones, subangular.
Colour: 10YR 2/3 (top 3cm., which were very dry was 10YR 5/2).

Narrow Boundary
II Texture: Silty, slightly plastic (moist).
Structure: Crumby, fine, moderate.
Consistency: Friable, weak cementation.
Stoniness: Very stony, medium stones, subangular.
Colour: 10YR 2/2

ORGANIC MATTER: Roots common to 20cm., few below
DRAINAGE: Free.


 

Pit No.28

Grid Co-ordinates: 244,24
Slope Angle: 15
Slope Bearing: 221
Slope Type: Concave
Distance Downslope: 30%

PROFILE
I Texture: Friable, slightly gritty.
Structure: Crumby, fine, moderate.
Consistency: Soft, with weak cementation.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 2/3

Merging Boundary
II Texture: Soft and gritty, slightly plastic (moist).
Structure: Crumby, fine, moderate.
Consistency: Slightly plastic, non- -sticky.
Stoniness: Stony, large stones, subangular.
Colour: 10YR 2/2

ORGANIC MATTER: Roots common to 20cm., few from 20 – 35cm., absent below.
DRAINAGE: Good.


 

Pit No.29

Grid Co-ordinates: 348,06
Slope Angle: 16
Slope Bearing: 125
Slope Type: Straight
Distance Downslope: 100%

PROFILE
I Texture: Soft and gritty, slight cohesion when moist.
Structure: Crumby, medium, stoney.
Consistency: Soft, weak cementation
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/3

Narrow Boundary
II Texture: Friable (dry) slightly plastic when moist.
Stnucture: Crumby, medium, moderate.
Consistency: Friable, weak cementation.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/2

Merging Boundary
III Texture: Soft and gritty, slight cohesion when moist.
Structure: Crumby, medium, moderate.
Consistency: Slightly plastic, slightly sticky.
Stoniness: Stony, small stones, subangular
Colour: 10YR 3/3

ORGANIC MATTER: Roots common to 20cm, few from 20 – 35cm., absent below.
Drainage: Free


 

Pit No.30

Grid Co-ordinates: 349,03
Slope Angle: 18
Slope Bearing: 70
Slope Type: Straight
Distance Downslope: 40%

PROFILE
Texture: Soft and gritty, slight cohesion when moist.
Structure: Crumby, medium, stony.
Consistency: Soft, weak cementation.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/3

Narrow Boundary
I Texture: Friable (dry) slightly plastic when moist.
Structure: Crumby, medium, moderate.
Consistency: Friable, weak cementation.
Stoniness: Stony, medium stones, subangular.
Colour: 10YR 3/2

Merging Boundary
II Texture: Soft and gritty, slight cohesion when moist.
Structure: Crumby, medium moderate.
Consistency: Slightly plastic, slightly sticky.
Stoniness: Stony, small stones, subangular.
Colour: 10YE 3/3

Merging Boundary
III Texture: Soft and gritty, slight cohesion when moist.
Structure: Crumby, medium moderate.
Consistency: Slightly plastic, slightly sticky.
Stoniness: Stony, small stones, subangular.
Colour: 10YR 3/3

DRAINAGE: Free


Appendix 2: Bulk Density Pouring Cylinder

A bulk density pouring cylinder was constructed from plywood, plastic drainpipe and a plastic funnel, to a similar design to the British Standard model (BS 1377).

This apparatus was calibrated as follows.
The weighed cylinder was filled with 2kg. of fine grained sand, seived by hand from windblown sand on the sandur, some distance from the camp. This was close-grained material, 250 – 500 microns, and air dried. The cylinder was then opened over a plate of glass, to determine the weight of sand in the cone, (see fig 49). This was repeated a number of times, and the mean wight noted, (W1).

A tin can was filled brimfull of water several times, and a mean weight obtained, from which the volume of the can was determined, assuming 1g of water occupied 1cc. The puring cylinder was then placed over the empty can, and sand allowed to flow into the can until the flow stopped. This was repeated a number of times, and the weight of the sand left in the cylinder was noted, (W 2).

From the wight of the sand left in the cylinder after pouring, and the weight of the sand in the cone, a figure could be reached for the wight of sand required to fill the can. Using the volume of the can, determined by water, a density for the sand could then be calculated. Details of the actual calibration are recorded below. It should be noted that the actual calibration of this apparatus was by no means as simple to complete as it is here described, but nevertheless, an accuracy of around 1.3% was achieved.

From the above figures, it is possible to give a figure for the bulk density of the sand, since, by calculation, given an initial wt of 2000g of sand, then the sand required to fill the can is

2000 – W1 – W2

Thus, the weight of sand is

2000 -173.1 -722.3 =1104.6g

When the weight of the can is added to this figure, it becomes 1220.0g.
The actual weight of the sand plus can, after pouring and carefully scraping the surface of the sand level with the rim of the can, was

1236.2g

which was a mean figure of three measurements.
Accuracy of the above calibration can be estimated by the difference between actual and expected weights, which was

This gives a margin of error of

However, given that the scales used were only accurate to .1g, then a cumulative error of several grammes either way could result. Nevertheless, the accuracy of the apparatus was accepted as reliable for the purposes of the study.


Appendix 3: Profile Description Terminology

(Reproduced from Smith and Atkinson, 1975)


 

Appendix 4: Diagram of Solar Oven Used in the Field


Appendix 5: Rain Fall Data

Figures supplied courtesy of J.Wood, another expedition member.
Bulk Density and soil moisture tests carried out within 8 hours of last figure, during which no rain fell whatsoever.
Rainfall recorded at camp raingauge, approximately 1km from area of transect.


Appendix 6: Vegetation Species


 

Note for the reader

Any typo or omission from this document is the fault of the digital transcriber, Alex Dunedin, and what you see here is a work in progress. If you notice any detail which needs corrected please leave a comment or get in touch via the contact page, it will be much appreciated.  This project is working towards a fully typeset digital copy of this work in PDF format as a valuable contribution to the field. Thank you for your understanding and help.

 

Leave a comment

name*

email* (not published)

website