THE ISLE OF WIGHT
On the southwest coast of the Isle of Wight, between Brook Chine and Atherfield Point, a continuous sequence of late Wealden (Barremian and early Aptian) mudstones with subordinate sandstones, making up the upper part of the Wessex Formation and the Vectis Formation, can be seen gently dipping and younging in the direction of Atherfield Point (Fig. 8.1). This continuous sequence presents the opportunity to investigate whether there was a change in the general taphonomic signature with time through the latter part of the Wealden in the Wessex sub-basin.
For this purpose, whilst at the same time ensuring that each sample analysed was of a suitable size, the Wessex Formation was subdivided into three units, with the Vectis Formation being considered separately. It should be clearly understood that it was only the exposed Wessex Formation that was subdivided and therefore although the three resulting units were called the lower, middle and upper Wessex Formation they are not literally such. It should also be noted that the three units, having no context outside this research, were not based on any previously accepted subdivisions. The precise placement of lower/middle and middle/upper Wessex Formation boundaries, respectively just below the Chilton Chine Sandstone and along the base of the Grange Chine Black Band, was quite arbitrary but was intended to divide the exposed Wessex Formation into three roughly equal thicknesses. In most cases it was possible to associate each Wessex Formation vertebrate find with one of the three subdivisions, since the associated records for specimens recovered from the Isle of Wight are generally good. However a handful of specimens could not be associated with one particular unit and had instead to be included in two analyses. For example, a bone for which the records said no more than that it was recovered from somewhere between Sudmoor Point and Chilton Chine would have been included in both the lower Wessex Formation analysis and the one for the middle Wessex Formation.
The lower Wessex Formation analysis is based on a total of 291 specimens, 170 of which were examined taphonomically (Appendix F). It should be noted that some of these bones came from northwest of Brook Chine but none came from beyond Hanover Point. Beyond Hanover Point faulting makes correlation with the sequence to the southeast of Brook Chine difficult and so it is not possible to say with confidence which of the three subdivisions of the Wessex Formation is exposed at any particular point. The middle Wessex Formation analysis is based on a total of 150 specimens, 57 of which were examined taphonomically (Appendix G). The upper Wessex Formation analysis is not examined in this chapter because the sample was dominated by bones from the Hypsilophodon Bed (31 of the 61 bones in the sample). Instead the specimens from the Hypsilophodon Bed are analysed separately (Appendix H). Finally, the Vectis Formation sample numbered 198 specimens in total, of which 126 were examined taphonomically (Appendix I).
Top of page Homepage
8.2 The lower Wessex Formation
Earlier in this dissertation the Wessex Formation was described as having been deposited by a meandering river system. The bulk of the formation consists of floodplain deposits, with subordinate sandstone units representing meandering rivers. The taphonomic signature for the lower Wessex Formation can quite easily be interpreted in terms of such a sedimentary context.
Figure 8.2. Grain sizes of attached matrix in the lower Wessex Formation sample (n = 59).
An examination of attached matrix in the lower Wessex Formation sample confirms that the predominant sedimentary context of the bones was that they ended up being buried in floodplain deposits (Fig. 8.2). The majority of the bones in the sample (89.8%) have fine silty matrix attached to them. These bones will almost certainly have been recovered from the floodplain deposits making up the bulk of the lower Wessex Formation.
Figure 8.3. Faunal composition of the lower Wessex Formation sample (n = 253).
The bar chart above illustrates the faunal composition of the lower Wessex Formation sample (Fig. 8.3). Clearly terrestrial groups (including Iguanodon) dominate the sample (75.9%), with aquatics and semi-aquatics being very poorly represented (22.9%). The representation of terrestrial groups far exceeds anything seen previously, even in the Cuckfield analysis. However this is perhaps not unexpected given the sedimentary context of the vertebrate material recovered from the lower Wessex Formation. The fluvial channels at Cuckfield would have sampled terrestrial animals living on or close to their banks and the crocodiles and turtles living in them. The Cuckfield sample therefore included a good proportion of both terrestrial and semi-aquatic groups. The lower Wessex Formation meandering rivers, represented by its subordinate sandstone units, would have sampled the fauna present in the same way as the Cuckfield channels but the adjacent floodplains would have only sampled the terrestrial animals that wandered across them. Since floodplain deposits make up the bulk of the lower Wessex Formation, the sample of vertebrate material recovered from it therefore includes a much higher proportion of terrestrial groups than was seen in the Cuckfield sample.
Figure 8.4. Trampling fragmentation in the lower Wessex Formation sample (n = 161).
The trampling fragmentation displayed by the lower Wessex Formation sample resembles quite closely that seen in the Hollington analysis (Fig. 8.4). Just as was seen for the Hollington sample, there is a fairly even spread across all stages of trampling fragmentation. Having said that, the balance is tipped slightly in favour of no trampling fragmentation, with the largest proportion of the sample (31.7%) belonging to this stage, whereas the balance in the Hollington sample was tipped slightly the other way, with the largest proportion of that sample (31.1%) belonging to stage 3 trampling fragmentation.
However before taking these observations at face value some consideration should be given to bone sizes in the lower Wessex Formation sample. The bar chart below (Fig. 8.5) illustrates the range of original bone volumes in the sample. Unlike the Hollington sample, which included few bones with original volumes greater than 999cc, there is quite an even spread of original volumes in the lower Wessex Formation sample. Given that larger bones are more robust and therefore less likely to be broken by trampling, it might be expected that if some account of this size bias was made then it would become apparent that the lower Wessex Formation sample actually shows greater trampling fragmentation than the Hollington sample.
Figure 8.5. Original volumes in the lower Wessex Formation sample (n = 166).
However the bar charts below (Fig. 8.6), illustrating the trampling fragmentation displayed by bones from the lower Wessex Formation sample in each of the successive ranges of original volume, suggest that the size bias is not important. In each volume range it is generally true to say that similar but slightly less trampling fragmentation is seen than was seen in the same volume range in the Hollington analysis, which is merely a confirmation of the observations made above after considering the lower Wessex Formation sample as a whole. It should be noted here that the trend of decreasing fragmentation with increasing volume seen in previous analyses, particularly the Cuckfield one, is not at all clear in this case. Indeed there is really no discernible trend at all.
Figure 8.6. Trampling fragmentation in each of the ranges of original volume in the lower Wessex Formation sample. A, <10cc (n = 36). B, <100cc (n = 23). C, <1000cc (n = 44). D, <10000cc (n = 36). E, >9999cc (n = 20).
Turning now to the pattern of weathering displayed by the lower Wessex Formation sample as a whole, it is again more or less identical to that seen in the Hollington analysis (Fig. 8.7). In the Hollington analysis, even though the majority of bones showed no weathering at all, only 59% of the sample belonged to this stage and the weathering seen ranged up to stage 3. The same is the case for the lower Wessex Formation sample, although in this case only 57.8% of the sample shows no weathering at all.
Figure 8.7. Weathering in the lower Wessex Formation sample (n = 147).
Again the question that arises is whether the size bias in the lower Wessex Formation sample is a factor that must be taken into consideration. The bar charts below (Fig. 8.8), illustrating the weathering displayed by bones from the lower Wessex Formation sample in successive ranges of original volume, help to answer that question.
Figure 8.8. Weathering in each of the ranges of original volume in the lower Wessex Formation sample. A, <10cc (n = 39). B, <100cc (n = 22). C, <1000cc (n = 38). D, <10000cc (n = 26). E, >9999cc (n = 18).
In a similar fashion to what was seen when the trampling fragmentation in the sample was broken down by original volume, all the bar charts show roughly the same pattern as was seen for bones in the same volume range in the Hollington analysis. The greater proportion of larger bones in the lower Wessex Formation sample compared with the Hollington sample, which might have been expected to skew the pattern of weathering displayed by the former to give the impression of less weathering, therefore seems to be of little importance. Additionally, in other analyses a trend of decreasing weathering with increasing volume was seen. This is not at all discernible in the above bar charts.
Figure 8.9. Abrasion in the lower Wessex Formation sample (n = 160).
Finally, the bar chart above (Fig. 8.9) illustrates the abrasion displayed by the lower Wessex Formation sample as a whole. It indicates that although there are bones in the sample which may be classified as stage 3 it is only a very small proportion of the sample that shows such heavy abrasion or anything approaching it. In fact the largest proportion of the sample (43.8%) shows only slight abrasion (stage 0-1). In this respect the pattern of abrasion displayed is more closely similar to what was seen for the Cuckfield sample, of which the majority of specimens (53%) also showed slight abrasion. By contrast the majority of the Hollington sample (51.2%) displayed stage 1 abrasion. To summarise, the lower Wessex Formation sample can be described as showing slightly greater abrasion than the Cuckfield sample and nothing like the abrasion displayed by the Hollington sample.
Figure 8.11. Abrasion in each of the ranges of pre-transport volume in the lower Wessex Formation sample. A, <10cc (n = 39). B, <100cc (n = 24). C, <1000cc (n = 45). D, <10000cc (n = 26). E, >9999cc (n = 15).
Since in this case it is the Cuckfield sample that compares most closely with the lower Wessex Formation sample the slight difference in the patterns of abrasion displayed by the two samples may be taken pretty much at face value. There is little need to worry about a size bias since the two samples have a more or less similar spread of volumes. However it is still quite instructive to break down the abrasion displayed by the lower Wessex Formation sample by pre-transport volume (Fig. 8.11). The reason for this is that no discernible trend in the pattern of abrasion displayed with increasing volume is revealed, just as was case in the trampling fragmentation and weathering analyses. Instead more or less the same pattern of abrasion is seen from one volume range to the next and therefore in all cases the pattern of abrasion approximately resembles the pattern for the sample as a whole. It should be noted that this somewhat spoils the comparison made with the Cuckfield sample. Since the Cuckfield sample showed a strong trend of decreasing abrasion with increasing volume, the resemblance in the patterns of abrasion displayed by the two samples becomes less close when larger bones are compared. It is therefore the larger bones in the lower Wessex Formation sample that account for the slightly greater abrasion seen for the sample as a whole when compared with the Cuckfield sample.
Having outlined the taphonomic signature for the lower Wessex Formation, interpretations can be made in the context of the physical environment represented, in this case a floodplain environment. As stated earlier, the analyses of attached matrix and faunal representation in the lower Wessex Formation sample are easily interpreted in terms of such a sedimentary context but what of the taphonomic modification seen, specifically the high degree of trampling fragmentation and weathering combined with low abrasion. Floodplains are characterised by very low burial rates and therefore any bones lying on them will be subject to prolonged periods of subaerial exposure. This will result in heavy weathering and also means that there is a good chance of breakage by animals trampling across the floodplain. Occasionally floods will occur and transport any unburied bones, although perhaps not on distal parts of the floodplain where flood strengths are likely to be much dissipated. Given that burial rates are typically low on a floodplain, such floods are likely to affect most bones at least once. However before significant abrasion can occur any flood transport must be repeated and/or prolonged and so normally minimal abrasion will be seen.
There remains only one loose end and that concerns the explanation for the absence of any trends with increasing volume in the trampling fragmentation, weathering and abrasion displayed by the lower Wessex Formation sample. The smaller bones in the sample would have been more susceptible to modification by each of these taphonomic processes, as seen in the analyses of previous chapters. However the larger bones, although less susceptible to such taphonomic modification, would have remained unburied and subject to that modification for much longer than the smaller bones due to the low burial rates on the lower Wessex Formation floodplain. It is this difference in the length of time that bones of different sizes were exposed to the various taphonomic processes that accounts for the absence of any trends with increasing volume in the sample.
The Wealden lagoon, like the lower Wessex Formation floodplain, was also characterised by low burial rates. Therefore in the Hollington analysis it might also have been expected to have seen an absence of any trends of taphonomic modification with increasing volume. This was not however the case. At Hollington, although burial rates were normally low, the successive returns of the Wealden lagoon following periods of emergence were accompanied by raised sedimentation rates. Most bones left unburied prior to a return of the lagoon would therefore have been buried by its return and few bones would thus have been subject to taphonomic modification for much longer than any other bone. With the time factor therefore effectively eliminated the size dependent trends were able to show through more clearly. However these trends were not as clear as in the Cuckfield analysis, probably suggesting that the above discussion is a bit of an oversimplification and that the time factor can not be totally ignored in the case of Hollington.
Top of page Homepage
8.4 The Hypsilophodon Bed
A taphonomic analysis of the upper Wessex Formation sample revealed a very different signature to that seen for the lower two subdivisions of the formation. This is not what might have been expected given that no change in the general sedimentary context is represented. The reason for the different taphonomic signature lies in the fact that half the bones on which the analysis was based were from the Hypsilophodon Bed. Bones recovered from the Hypsilophodon Bed found themselves there in the first place as a result of a single taphonomic event. They have therefore been taphonomically modified in a way that specifically reflects that event and this was what largely determined the taphonomic signature revealed by the analysis of the upper Wessex Formation sample, bones from the Hypsilophodon bed accounting for such a large proportion of those included. There is therefore little point discussing any further the upper Wessex Formation analysis. Instead the focus here is an analysis of only those bones specifically recovered from the Hypsilophodon Bed.
Figure 8.13. The Hypsilophodon bed as it is exposed in the cliffs just northwest Cowleaze Chine on the southwest coast of the Isle of Wight.
The Hypsilophodon Bed marks the top of the Wessex Formation and is exposed in the cliffs between Barnes High and Cowleaze Chine (Fig. 8.13). According to Insole and Hutt (1994) it is composed of interbedded floodplain and crevasse splay deposits and telltale sedimentary structures suggest that these would have been heavily waterlogged when first laid down (Stewart, 1978a and 1978b). Turning now to the palaeontological finds from the Hypsilophodon Bed, it is of note that they are more or less exclusively various highly articulated remains, representing several individuals, some adult, some subadult, all of the same species (Hypsilophodon foxii) (Galton, 1974). Catastrophic burial of a herd of animals is suggested. Given the telltale sedimentary structures suggesting heavy waterlogging, Insole and Hutt (1994) went with the suggestion that the herd died as a result of being trapped in quicksand. Alternatively the herd may have been caught up in one of the flood events responsible for the crevasse splays of which the Hypsilophodon Bed is composed. However these suggestions are based on a somewhat superficial taphonomic analysis compared with the sort of analysis utilised previously in this dissertation. That is not to say that the purpose of an analysis of the latter sort would be to find evidence to contradict the idea of catastrophic burial. The purpose would be to establish a more detailed taphonomic history for the bones recovered from the Hypsilophodon Bed. Such an analysis is presented below.
Figure 8.14. Trampling fragmentation in the Hypsilophodon Bed sample (n = 29).
Turning first to the trampling fragmentation displayed by the sample as a whole, all stages of fragmentation are seen, with the largest proportion of the sample (34.5%) belonging to stage 3 (Fig. 8.14). This is not dissimilar from what was seen in the Hollington analysis. Indeed this comparison can be taken at face value because there is no size bias to take account of, there being a similar distribution of bone volumes in both samples, specifically with larger bones being very poorly represented (Fig. 8.15). The poor representation of larger bones in the Hollington sample was a result of aquatic and semi-aquatic reptiles making up the majority of the sample. In the case of the Hypsilophodon Bed sample the absence of larger bones is because, as described earlier, practically every find from the bed may be attributed to Hypsilophodon foxii, a creature that measured only 2m in length.
Figure 8.15. Original volumes in the Hypsilophodon Bed sample (n = 31).
The bar charts below (Fig. 8.16) break down the trampling fragmentation displayed by the Hypsilophodon Bed sample by original volume. Although some caution is required because the number of specimens represented by each bar chart is decidedly small, what is seen again closely resembles what was seen in the Hollington analysis. In each successive volume range more or less the same pattern of fragmentation is displayed as was seen in the same volume range in the Hollington analysis. The only notable difference is in the less than 10cc volume range, where less fragmentation is suggested than was seen in the Hollington analysis (Fig. 8.16A). In the Hollington analysis there was some discussion of how it was that bones in this volume range, given their relative fragility, could show less evidence of trampling fragmentation than bones in the succeeding volume ranges. Clearly in the case of the Hypsilophodon Bed sample, what with there being even less evidence of fragmentation in the less than 10cc volume range, the case is more extreme. This point will be returned to later.
Figure 8.16. Trampling fragmentation in each of the ranges of original volume in the Hypsilophodon Bed sample. A, <10cc (n = 10). B, <100cc (n = 12). C, >99cc (n = 7).
Moving on to the weathering displayed, not one of the bones examined showed any evidence of weathering, even with them all being that much more susceptible as a result of their relatively small sizes. It is more or less the same story as far as abrasion is concerned, with 83.9% of the sample showing no abrasion at all and no bone showing anything greater than stage 1 abrasion (Fig. 8.17). This is again despite all the bones in the sample being relatively small in size and therefore more easily transported. Breaking the sample down by pre-transport volume, it is perhaps a surprise to find that those few bones that do show signs of abrasion are not some of the smallest bones in the sample but in fact some of the largest (Fig. 8.18). The explanation for this is related to what was seen in the above analysis of trampling fragmentation in the sample and will be returned to later.
Figure 8.17. Abrasion in the Hypsilophodon Bed sample (n = 31).
Figure 8.18. Abrasion in each of the ranges of pre-transport volume in the Hypsilophodon Bed sample. A, <10cc (n = 18). B, >9cc (n = 11).
Having described the taphonomic modification typical of bones recovered from the Hypsilophodon Bed, it is necessary to interpret what has been seen and assess what new light is thrown on the taphonomic history of the bones. There is no doubt, given that the majority of the bones recovered belong to the same species, with several individuals, both adults and subadults, being represented, that this is a case of catastrophic burial of a herd of animals. The more likely scenario is that the herd got caught up in one of the floods that laid down the crevasse splay deposits making up the Hypsilophodon Bed. The alternative suggestion that the herd stumbled into quicksand can be rejected because of the evidence of trampling fragmentation. It would have been rather difficult for other animals to have later trampled through the quicksand without being caught up in it themselves. The quicksand theory can also be rejected on the basis that proper quicksand would surely have obliterated all sedimentary structures and yet this is not seen.
After the flood the crevasse splay deposits, although only thin, would have more or less buried the Hypsilophodon carcasses caught up in them, Hypsilophodon being a small creature. Burial does not allow for much in the way of disarticulation, weathering or abrasion and this agrees well with the taphonomic characteristics of the Hypsilophodon Bed material. However an explanation is still required for the trampling fragmentation seen. Since the burial can only have been very shallow and must have been in heavily waterlogged, soft sediments, as indicated by the telltale sedimentary structures seen, it was still entirely possible for there to be some trampling fragmentation. In fact trampling not only caused bone breakage, it was probably also responsible for what little disarticulation that occurred. In addition it provides an explanation for the abrasion seen in a few cases. This abrasion was not due to transport but resulted from the action of the feet of the trampling animals. Finally, it only remains to explain why the smallest bones in the sample showed less trampling fragmentation and abrasion than larger bones. The soft sediments in which these bones were buried is again part of the explanation for this. The smaller bones, when stepped on, would simply have been pushed deeper into the sediments in which they were buried. Only those bones with dimensions similar to or larger than those of the feet of the trampling animals would have ended up displaying any signs of trampling related breakage or abrasion.
Top of page Homepage
In this chapter the two end member Wealden taphonomic model of previous chapters has become a three end member model. Having defined, through analyses of samples from Cuckfield and Hollington respectively, the taphonomic signatures typical of Wealden fluvial channels and the Wealden lagoon, the taphonomic signature typical of a Wealden floodplain has been added through analyses of samples from the Isle of Wight, namely the lower and middle Wessex Formation samples. More specifically the typical Wealden floodplain taphonomic signature has been revealed to be characterised by high levels of trampling fragmentation and weathering but low levels of abrasion. This is easily explained given that on a floodplain burial rates will be low and there will be an absence of any transporting currents, except for during the occasional flood. A bone that finds itself on a floodplain will therefore normally lie subaerially exposed without being moved very far from the place of death of the animal to which it once belonged for some considerable period of time. This will result in precisely the sort of taphonomic modification seen in the lower and middle Wessex Formation analyses. Independent confirmation that these analyses have indeed revealed the taphonomic signature typical of a floodplain sedimentary context is provided by the work of Fiorillo (1988) on the taphonomy of Miocene material from Hazard Homestead Quarry, Nebraska. The material, having been recovered from floodplain deposits, was found, just like the material in the lower and middle Wessex Formation samples, to be heavily modified by trampling and weathering but not particularly abraded.
Having established the typical floodplain taphonomic signature, the Hypsilophodon Bed presented the opportunity to focus in on the taphonomic signature associated with a particular floodplain process, specifically minor breaks in channel banks resulting in the creation of crevasse splays. It would seem that such channel bank failures presented a real danger to herds of animals inhabiting the Wealden floodplains. The Hypsilophodon bed analysis revealed that subsequent burial in heavily waterlogged crevasse splay deposits ensured isolation from weathering and also largely abrasion, but not trampling fragmentation.
Finally, the Vectis Formation analysis merely confirmed a return of the Wealden lagoon prior to the Lower Greensand marine transgression.
Top of page Homepage