The Work

Specialist Report 2005 - Preliminary Analysis of Pollen and Spores from High Pasture Cave, Skye

Posted by steven on 05/04/2006 at 06:26 PM


Sandra Pratt - University of Edinburgh


This report provides a summary of a preliminary assessment of pollen composition, concentration and preservation for 36 spot samples provided by Steven Birch from a range of contexts at High Pasture Cave. 30 samples are taken from the major stratified layers in Trench 2, 3 samples from the high level fossil cave passage, one sample from the cave stream passage deposits, one sample from Trench 3 and one sample from Trench 5.


Samples were prepared at the University of Hull palynology laboratory using standard methods for pollen concentration as described in Moore et al. (1991). Two Lycopodium tablets were added to a 1 cm3 sub-sample from each context as a marker grain to assess pollen concentration (Stockmarr, 1971) prior to treatment with hot hydrochloric acid (HCl) to dissolve carbonates. Samples were then digested with hot potassium hydroxide (KOH) to break down humic matter. The samples were then coarse and fine sieved (mesh sizes 150 μm and 10μm) to remove detritus.

Two treatments with hot concentrated hydrofluoric acid (HF) were carried out to dissolve silicates, followed by hot HCl to minimise precipitation of fluorosilicates. Samples were then acetolysed to remove cellulose and the remaining pellets were then stained and transferred to silicon oil prior to mounting for light microscopy.

Each sample was counted at a magnification of 400x (1000x for critical identifications) until a minimum of 50 Lycopodium spores had been recovered, with a further 50 spores recovered (i.e. 100 in total) for those samples containing >5 pollen grains at the first count. This ensured that a similar volume of sediment was scanned for each sample. A general estimate of charcoal abundance was made using an index (A=abundant, F=frequent, O=occasional, R= rare or absent), and spore types were also counted and identified by type. Identifiable pollen grains were recorded as normal, crumpled, corroded, broken or degraded. The same categories were used for indeterminate grains but with an additional two categories for concealed and unknown grains.

The criteria selected for accepting samples as suitable for further analysis were i) a concentration of >3000 grains cm-3 ; ii) no less than 45% indeterminable pollen grains; iii) a range of taxa besides decay resistant types.


All samples formed pellets as expected except for C222, which formed a grey, gelatinous pellet during preparation, probably due to precipitation of fluorosilicates after hydrofluoric acid treatment. Although Lycopodium was recovered, no detritus, charcoal or pollen/spore material was present so this sample was not included in the analysis.


28 of 35 samples were found to contain at least some charcoal when viewed under the microscope. No sample was scored as having abundant charcoal but 6 samples had frequent charcoal and the remainder occasional charcoal.  In contexts C205, C221 and C003 the fragments had a rounded, nodular appearance. Other fragments were of more typical angular appearance.

Pollen and spores

A total of 24 pollen and 4 spore taxa were identified for samples from the trenches. 20 unknown grains (ten taxa) were also observed and require cross referencing with type material. Pollen and spore counts for all samples (except A2, A4 and A6 which are discussed in the text) are summarised in Table 1. The 17 most frequently occurring pollen taxa are shown, with the remaining taxa classed as ‘other’. The percentage of indeterminate pollen grains and the total pollen concentration (grains cm-3) are shown for each context, and relate to all taxa observed. Pteridium aquilinum (bracken) spore concentrations are also discussed.

Where pollen was present, preservation was generally quite good, with most grains being recorded as normal and on occasion degraded, broken or crumpled.  Calluna (heather) and Poaceae (grass) pollen are by far the most frequent types, followed by Asteraceae Lactuceae (dandelion type), Corylus (hazel), Alnus (alder) and Betula (birch). Most grass pollen appeared to be of wild type but occasional grains of cereal type with larger pore annuli (diameter c.5-7 μm) were seen in contexts C221, C226, C231, C238, C240, C241 and C245.

A total of 19 samples had a concentration greater than 3000 grains cm-3 with fewer than 45% indeterminable grains, so can be considered suitable for further analysis. Of these, C245, C239, C006 and C240 had the highest concentrations, ranging from c.21 000 to c.55 000 grains cm-3. There were 6 contexts with pollen > 7000 grains cm-3: C003, C221, C223, C241, C231 and C244. Of two samples from C205, HPC2 contained pollen at a concentration of c.11 000 grains cm-3 in sample HPC 2, whilst. HPC 21 contained only c.1500 grains cm-3.

High Level Fossil Passage in Cave-Trench 1

C003- High level fossil passage midden
This context comprised mainly fine organic detritus with occasional nodular charcoal fragments of c. 30 – 50 μm diameter. A fairly good concentration of pollen was observed (c.8500 grains cm-3) with 7 Calluna, 5 Asteraceae Lactuceae (dandelion type), including one of Anthemis (mayweed) type, degraded cf.Ilex (holly) and 2 degraded Pinus air sacs.

C006 – High level fossil passage, orange cave earth with bone and fire cracked stone
Abundant organic matter with occasional charcoal; a high concentration of pollen (c.30 800 grains cm-3), with most pollen grains degraded or crumpled reflective of mechanical damage and oxidation but still identifiable. Pollen was predominantly Calluna, with Asteraceae Lactuceae, Poaceae (wild type) and Alnus (alder).

Trench 2- U-shaped structure on surface

C205-Dark, gritty organic rich sediment with fire cracked stone
In sample HPC2 from this context, a diverse assemblage was seen, with Alnus (alder), Corylus (hazel), Salix (willow), probable Quercus (oak), Calluna (heather), Poaceae (grass), degraded Plantago lanceolata, Rumex (sorrel) and Asteraceae Lactuceae (dandelion type). Pteridium (bracken) and bryophyte (moss) spores and frequent nodular charcoal fragments were also observed.

C221- Trench 2, buff/brown greasy context with thin ash lens and animal bone
Around 20% of identifiable pollen was broken, degraded or crumpled, and a further 35% indeterminate due to degradation, suggesting mechanical damage as well as oxidation. The dominant pollen type here was Poaceae. Other taxa included Corylus (hazel), degraded Alnus (alder) and Conopodiaceae (pignut). There was an abundance of fine organic detritus with occasional nodular charcoal fragments.

C223- Trench 2, buff/brown greasy context with animal bone and teeth, part of rectangular clay feature F210
This comprised abundant organic detritus with quite well preserved pollen (c.11 000 grains cm-3) and frequent charcoal.  Calluna was the dominant type (25%), with Corylus, Asteraceae Lactuceae (dandelion type) and an intact Pinus.

C231- Trench 2, orange/black silty lenses with peat ash and charcoal, NW of rectangular clay feature F210
Pollen concentration was c.7400 grains cm-3, with 4 Asteraceae Lactuceae, 5 Poaceae (including 2 cereal type) and 3 Caryophyllaceae (campion).  There was also Rumex acetosella type (sorrel), Corylus and a further 4 unidentified types (degraded or crumpled).

C239- Trench 2 dark brown to black silt possibly relating to weathering and silting of sinkhole
Pollen concentration quite high at c.23 000 grains cm-3, and although preservation was poor, implying mechanical damage and chemical and/or biological oxidation, identification was still possible. Occasional charcoal and frequent, somewhat degraded trilete spores, probably Pteridium (bracken). The dominant pollen type was Calluna.

C240 – pit fill in Trench 2
This context contained the highest pollen concentration of any of the samples
(c.55 000 grains cm-3). Most grains were identifiable but occasionally degraded, with 31% being too degraded to identify, implying oxidation by chemical or biological agents. There was abundant organic and limestone detritus and occasional charcoal. The most frequent pollen type was Poaceae (grass), 36 of wild type but 7 of undifferentiated cereal type. Corylus (hazel), Alnus (alder) and Salix (willow) and bracken spores were also quite frequent.

C241- Trench 2, brown loam; post hole fill
Pollen concentration was c.18 500 grains cm-3, with Poaceae by far the dominant pollen type at 21 grains. Other taxa included Corylus, Asteraceae Lactuceae and Calluna. Two grains were unidentified, pending access to reference material. Abundant bracken spores and occasional charcoal were observed.

C244- Dark brown sediment comprising entrance passage and burial fill
This was quite a diverse assemblage, dominated by Asteraceae Lactuceae type (24%), and Poaceae (wild type, 17%). Ericaceae (bell heather or bilberry type), Ilex (holly), Salix (willow), Corylus, Caryophyllaceae (campion) and Betula (birch) also occurred. Three unidentified partially degraded grains may be of Brassicaceae (cabbage family) type. There were 2 trilete spores (one of bracken type) and abundant other undifferentiated moss and fungal spores.

C245 Dark brown sediment underlying burial fill
Pollen concentration was c.22 000 grains cm-3, of which 44% was comparable to Salix. Salix was not observed in such abundance in any other sample (where present, it occurs at c.4-5%, but up to 11% in C216 and 12% in C234). There were also 6 Poaceae, including one of Hordeum (barley) type, 6 Asteraceae Lactuceae, 5 Calluna, and 1 Apiaceae (carrot family).

Context 226 contained 44 indeterminable unknowns, 38 of which were of similar appearance and probably algal spores. The sample from the stream section A-B contained little pollen (c. 2900 grains cm-3), comprising 2 Corylus (hazel), 1 Calluna (heather) and 1 Poaceae (grass). Spores (mainly bracken) were also observed, at a concentration of c.9000 cm-3).

Samples A1, A4 and A6 from within the cave contained almost no pollen. No pollen or spores were seen in A1, whilst in A4, no pollen but abundant, degraded trilete spores at a concentration of c.42 000 grains cm-3 were observed. A6 contained a lower concentration of similar trilete spores (c.1800 grains cm-3), one other spore and an indeterminable crumpled pollen grain. Depending on the geology of the contexts, it is possible that these are of pre-glacial origin.

Discussion and Conclusions

The results presented here represent rapid and preliminary counts only, but at least 19 of the 36 contexts contain sufficient well preserved pollen (<45% indeterminate excluding unknowns) at reasonable concentration (>3000 grains cm-3) to allow for further work, including analysis of higher counts. As shown by the results for context C205, pollen concentrations may vary greatly within any single context, so more detailed sub-sampling of contexts of particular interest should not be excluded, even if the pollen concentration from the spot samples analysed here is borderline or low.

The samples from contexts C003, C006, C205, C221, C223, C231, C239, C240, C241, C244 and C245 have particularly good concentrations of pollen and a sufficiently diverse range of taxa to suggest that further analysis would be worthwhile.

As expected, the assemblages are dominated by taxa typical of the upland landscape together with grasses and ‘weedy’ species associated with grazing and human activity. In contexts where there are >45% indeterminable degraded or crumpled grains, it is likely that certain taxa are over- or under- represented, making ecological interpretation difficult. C228, for example, is dominated by Asteraceae Lactuceae (30%). This type is resistant to decay in palaeo-soils, so would tend to be over-represented and its distinctive morphology means it is likely to be identified even when very degraded.  Smaller pollen grains may also be under-represented, e.g. Filipendula (meadowsweet), as these may be lost during fine sieving. It is therefore suggested that future analyses proceed without fine sieving to avoid potential loss of smaller pollen grains.

There are other difficulties associated with interpreting assemblages such as these in any ecologically meaningful way, as both natural and anthropogenic reworking and disturbance of material during and after deposition will contribute to complex origins for the pollen. However, when considered with other context characteristics, including charcoal and plant macro-fossils, useful archaeological interpretations should be possible, e.g. relating to bringing of material onto the site by humans. It is therefore suggested that a detailed assessment of pollen and spore preservation, as well as assemblage characteristics, is carried out as part of any further analyses.

More specific identification of Poaceae types may also be a useful supplement to macrofossil analysis, if based on the analysis of features of a larger number of grains. The level of concentration of these grains in samples C221, C226, C231, C238, C245, C240 and C241 supports the existing macrofossil evidence that processed or partly processed cereal was handled or deposited in and around the site; it is unlikely that extensive arable cultivation took place in the area surrounding the site so this could not explain the relatively high concentrations of cereal pollen seen.

Interestingly the high level fossil passage deposits all have at least some pollen, including 30 -50% Calluna and frequent Poaceae. If these passages were last active during pre-glacial times, such pollen could have originated from pre-glacial vegetation, but the high proportion of Calluna and Poaceae does support the idea that material containing pollen was brought in by humans, e.g. as bedding or fuel, rather than being deposited by natural processes. Further analysis of these contexts is recommended as they appear to have good potential for aiding the interpretation of the cave deposits, in terms of age and origin.

Further analysis of the C245 burial fill would be useful to establish whether the predominance of cf. Salix pollen is a localised phenomenon or apparent elsewhere in this context; as Salix is usually under-represented the high concentration observed suggests that male catkins were locally present. Analysis in conjunction with any plant macrofossil material found may be of benefit here.

The regional context

In view of the limitations associated with making ecological interpretations from the on-site contexts as discussed above, there is a strong case for analysing a suite of sediment cores from bogs and lakes around the site where pollen is generally less disturbed and also well preserved. This would provide important information to enable the site to be understood as part of the wider dynamic landscape, and such a study could be designed to focus on the period of interest, e.g. to cover a time span prior to and during the development, occupation and abandonment of the site.

Such analyses would also serve as a valuable reference for interpreting and dating the contexts at High Pasture Cave. With further regards to dating, there is a strong possibility that volcanic ash (tephra) deposits occur in the sediments here; the Hekla-4 (2310±20 BC) and Glen Garry (c.2100BP) tephras have been recovered from peat deposits at Ben Gorm Moss in northern Skye (Langdon and Barber, 2004). Identification of tephra is a useful method for reconstructing past landscapes, as it allows the same ‘marker’ horizon to be identified across multiple sites. Site selection and sampling strategies could be designed to address questions of interest concerning themes such as changes in woodland type, fire regimes, and degree of landscape openness. Similar questions are now being successfully addressed for reconstructing past cultural landscapes in Britain and NW Europe using recently developed techniques using pollen modelling in conjunction with Geographical Information System software, e.g. the work of members of the POLLANDCAL* (Pollen/Landscape Calibration network).


Langdon, P.G. and Barber , K. E. (2004) Snapshots in time: precise correlations of peat-based proxy climate records in Scotland using mid-Holocene tephras

Moore, P D, Webb, J A, and Collinson, M E, (1991) Pollen Analysis 2nd edition Oxford: Blackwell
Stockmarr, J. (1971) Tablets with spores used in absolute pollen analysis. Pollen et Spores 13(4): 615-621.

*POLLANDCAL homepage:

Next entry: Specialist Report 2005 - Small Finds Assessment

Previous entry: Specialist Report 2005 - Fish Bone and Marine Mollusc Report