September 19, 2008

Discussion on Palaeoecology of the Late Triassic Extinction Event in the SW UK

By Radley, Jonathan Twitchett, Richard J; Mander, Luke; Cope, John

Journal, Vol. 165, 2008, pp. 319-332 Jonathan Radley writes: The Penarth Group (of Late Triassic and possibly ranging to Early Jurassic age) of the southern UK marks a marine transgression and the establishment of a shallow epicontinental seaway (Hallam & El Shaarawy 1982; Warrington & Ivimey-Cook 1992), influenced by regressive-transgressive pulses and characterized by rapid facies changes (Hallam & Wignall 2004; Hesselbo et al. 2004). A well- documented invertebrate macrofauna includes corals, brachiopods, molluscs and echinoderms (Swift & Martill 1999). Facies developments range from storm-influenced, shallow-marine mudrocks (Westbury Formation) upwards into calcareous mudstones, siltstones and essentially fine-grained carbonates (Lilstock Formation) demonstrating extremely shallow-water conditions: wave ripples, storm beds, desiccation and omission surfaces (Swift 1999a; Wignall 2001; Hallam & Wignall 2004). The low-diversity macrofauna of the Lilstock Formation has frequently been attributed to salinities deviating from that of normal seawater within shallow-water, partly landlocked settings, isolated from the open sea (Hallam & El Shaarawy 1982; Swift 1995; Allison & Wright 2005). However, widespread occurrences of stenohaline marine macrofossils such as corals and echinoderms (Swift & Martill 1999) suggest that salinity fluctuations were at most slight (Wignall 2001).

Mander et al. (2008) have presented high-resolution benthic macrofaunal data, principally from molluscan macrofossils, for significant palaeoecological change within the Penarth Group of two key sections in the SW UK. Their detailed results demonstrate loss of taxonomic richness within the uppermost Westbury Formation-lower Lilstock Formation (Gotham Member) interval (latest Rhaetian), followed by a poorly fossiliferous interval represented by the latest Rhaetian or early Hettangian upper Lilstock Formation (upper Gotham Member-lower Langport Member). They attributed this pattern to an extinction event followed by a post-extinction 'Dead Zone'; the latter coinciding with the onset of a negative organic carbon- isotope excursion, and an extinction event recorded in the terrestrial palynoflora. The relatively diverse shelly macrofauna of the upper Langport Member and overlying Blue Lias Formation, ranging to the Early Jurassic, is taken to mark recovery. Invertebrate body sizes remained low until the later part of the Hettangian Stage. Mander and co-authors discuss these palaeoecological changes with reference to the widely recognized Late Triassic extinction episode.

I would like to draw the attention of the authors to a widespread stromatolitic horizon within the Lilstock Formation, which has potential significance in the case for an extinctionrelated benthic crisis. Well-preserved stromatolites occur at the top of the Gotham Member at a number of sites in the SW UK (Hamilton 1961; Warrington & Ivimey-Cook 1992), marking the peak of the carbon-isotope excursion mentioned above (Hesselbo et al. 2004). The best- documented examples (comprising the 'Gotham' or 'Landscape' Marble of the Bristol district, western England) are typically fine- grained, well-laminated micritic stromatolites associated with a sparse shelly epifauna of marine aspect (Mayall & Wright 1981; Ivimey-Cook et al. 1999), scours, shallow channels and mud-flake breccias (Hamilton 1961). Further afield, reworked fragments of stromatolitic limestone occur at the base of the Langport Member in central England (Radley 2002).

Phanerozoic stromatolites typify environmentally stressed marginal-marine or lacustrine facies, where rates of biogenic disruption to microbial mat ecosystems are significantly reduced (Walter 2001). Some workers (e.g. Schubert & Bottjer 1992; Whalen et al. 2002; Sheehan & Harris 2004) have documented stromatolite developments within Phanerozoic carbonate units interpreted as subtidal in origin, immediately post-dating mass extinction events. Such occurrences have been taken to indicate relaxation of ecological constraints during the early stages of faunal recovery, notably the loss of inhibiting bioturbators and grazers.

The chronostratigraphy of the Cotham-Langport Member boundary across the SW and central UK is poorly resolved. Acceptance of synchroneity or slight diachroneity (Warrington et al. 1980) would imply widespread stromatolite development in what has been interpreted, in the SW UK at least, as a transgressive marine, shallow-water succession (Hesselbo et al. 2004). At face value this could be taken as evidence for establishment of disaster forms in a shallow-marine setting as recovery commenced (Schubert & Bottjer 1992). Equally, association with epifaunal molluscs, mud-flake breccias and channels suggestive of intertidal conditions raises the possibility that the inhibiting fauna was excluded by other factors such as unsuitable substrates, desiccation, and/or a tendency towards hypersalinity in evaporating pools.

Recognition of an extinction event within the Penarth Group is potentially complicated by the 'noise' of rapid facies changes, erosion surfaces (Swift 1999a), a strong substrate control on the distribution of shelly macrobenthos (Ivimey-Cook et al. 1999), and the likelihood of transient environmental stress including salinity fluctuations. With reference to the case for stromatolites as disaster forms, I would welcome the authors' views on the palaeoecological and palaeoenvironmental significance of this regionally extensive stromatolitic horizon within the Lilstock Formation.

13 March 2008

Richard J. Twitchett & Luke Mander reply: We would like to thank Jon Radley for bringing to our attention some implications of our recent palaeoecology-based interpretation of the fossil records of the Westbury and Lilstock formations of the SW UK that, we confess, had escaped our notice. Radley is correct that the changing facies and palaeoenvironments of the Triassic-Jurassic rock record of the SW UK do complicate matters, but if a key control on fossil diversity in this interval is the biotic consequences of a major extinction event then certainly the lowdiversity benthic assemblages of the Lilstock Formation do not necessarily have to reflect changes in salinity as previously suggested. A low-diversity benthic macrofossil assemblage containing stenohaline taxa would be expected in normal-salinity marine settings in the immediate aftermath of a major extinction event (e.g. Twitchett 2006).

The question of whether the Gotham Member stromatolites represent post-extinction 'disaster forms' (sensu Schubert & Bottjer 1992) is intriguing and something we had not considered. Bottjer et al. (1996) expanded the Schubert & Bottjer (1992) dataset and confirmed that the greatest peak in occurrences of 'normal marine' stromatolites in the Phanerozoic is recorded in the Early Triassic, which they interpreted as reflecting the postextinction loss of metazoans. Since then, more examples of Early Triassic microbialites have been documented (e.g. Pruss et al. 2004) and post-extinction stromatolites have also been identified in the aftermath of the Late Devonian Frasnian-Famennian extinction (Whalen et al. 2002).

These examples are not universally accepted as representing post- extinction disaster forms (Shen & Webb 2004; Twitchett 2006, p. 204) and aspects of this debate are perhaps crucial to the interpretation of the Gotham Member examples; in particular, stratigraphie position and depositional setting. As Radley notes, stratigraphie position is correct; that is, in the immediate aftermath of the event, when metazoan diversity and abundance, infaunal tiering, bioturbation depth and intensity are all much reduced (Barras & Twitchett 2007; Mander et al. 2008). Perhaps the critical question is whether the depositional setting is one typically associated with stromatolites at non-extinction times (e.g. hypersaline lagoons and tidal flats), in which case there would be no reason to label them as post- extinction disaster forms.

Like many Induan (Early Triassic) examples, the upper Gotham Member stromatolites formed in an early transgressive systems tract (Hesselbo et al. 2004). Algae within the stromatolitic layers indicate deposition within the photic zone (Mayall & Wright 1981) and flat-topped wave-ripples from the upper Gotham Member imply water depths of just a few centimetres at some horizons (Mayall 1983; Hesselbo et al. 2004). Ostracodes indicative of hypersaline conditions have not been recorded from the Gotham Member, although Boomer et al. (1999) suggested that monospecific 'floods' of the marine taxon Ogmoconchella in the overlying Langport Member may reflect hypersaline lagoonal conditions. The ostracodes of the Gotham Member are either marine (e.g. Bairdia, Eucytherura) or freshwater (e.g. Darwinula), with mixed assemblages interpreted as being brackish (Anderson 1964; Boomer et al. 1999). Although we cannot find published records of ostracodes recovered from the Gotham Member stromatolite horizon itself, it appears that a hypersaline depositional setting can be rejected.

The Gotham Member stromatolites apparently grew in a schizohaline (i.e. generally brackish, fluctuating salinity) coastal setting (Mayall & Wright 1981). Is this a setting in which subtidal stromatolites may normally be expected to occur? In the schizohaline, calcite-oversaturated, tropical, coastal waters of Chetumal Bay, Belize, modern stromatolite reefs are flourishing and have been described by Rasmussen et al. (1993). The Chetumal stromatolites may be a good analogue for the Gotham Member stromatolites, and perhaps Chetumal Bay is a good depositional model for the Gotham Member as a whole. Unless further work demonstrates that the Gotham Member stromatolites grew under normal marine conditions, their existence is probably better explained by depositional setting rather than by a postextinction disaster taxon interpretation. 20 March 2008

John Cope writes: Mander et al. (2008) present a detailed analysis of molluscan extinctions across the Triassic-Jurassic boundary at two localities in southwestern Britain. However, much of the study concerns the Penarth Group, a stratigraphical unit that was deposited in very shallow water, and in which faunal variation is more likely to reflect changes in salinity, water depth, clastic input and contiguity with the open sea than any extrinsic forces. These factors affect any potential contribution to the debate on the magnitude and precise stratigraphical position of any western peri- Tethyan or even global extinction event.

At the end of the Triassic Norian Epoch the British area was one of desert plains surrounded by low-lying hills. The initial Rhaetian transgression flooded the plains with a very shallow sea, leaving the hills of the Norian still above water (Warrington & Ivimey-Cook 1992); recent confirmation of this is the Early Rhaetian age now assigned to the karstic fissure fills yielding land vertebrates in the neighbouring Carboniferous Limestone (Whiteside & Marshall 2007).

The fauna of the lower (Westbury) Formation of the Penarth Group is dominated by shallow-water bivalves. Notable absences (or exceptional rarities) for rocks of this age are all the stenohaline invertebrate groups, including ammonoids, conodonts, bryozoans, calcareous brachiopods, sponges and echinoderms (bar one species of ophiuroid; perhaps the one group of echinoderms that could tolerate a slightly reduced salinity). In such shallow waters the salinity must have varied considerably and connection with the open sea to the SW was probably restricted. Both selected sections were very close to land, so that any runoff would have affected the salinity locally. The black shale facies exhibits conspicuous cyclicity and it would be expected that the fauna would vary in frequency and size in response to changing local conditions.

The 'abrupt faunal extirpation' at the top of the Westbury Formation with '52% of species disappearing' (Mander et al. pp. 328, 323) is at odds with the record of Waters & Lawrence (1987, based on detailed collecting by H. C. Ivimey-Cook), who noted that the lower part of the Gotham Member has a diverse marine fauna that includes most of the bivalves found in the Westbury Formation. The fact that the upper part of the Gotham Member is 'virtually devoid of fossil specimens' surely does not reflect a 'Dead Zone', but merely a drop in salinity at this level, coinciding with shallowing and isolation from the sea that lay to the SW (Swift & Martill 1999). A non- marine influence is confirmed by the fauna of freshwater ostracodes and the abundant occurrence in some localities in the SW of the bryophyte Naiadita lanceolata, which has been interpreted as living in lagoonal fresh to brackish water (Hemsley 1989). However, dinoflagellate cysts, indicating a marine influence, are also present (Hounslow et al. 2004). Desiccation phenomena are common in the Gotham Member and occasional hypersalinity is indicated by gypsum and baryto-celestite. These changes were admirably summarized by Warrington & Ivimey-Cook (1992, p. 102): 'The marine fauna of the earliest [Gotham] beds rapidly dies out upwards. ... Beds with marine phytoplankton alternate with ones containing fish scales (Pholidophorus), Euestheria, darwinulid ostracods and plants such as Naiadita. The facies is widespread, but its sparse fauna and argillaceous lithology ... suggest ... a very extensive lagoonal environment ...[with]... periodic incursions of fresh water alternating with more saline and even hypersaline conditions. The redistribution of periodic ashfalls could account for some of its unusual lithological and biogeographical features.' Thus the 'Dead Zone' invoked by Mander et al. is a purely an effect of salinity fluctuations over a large plain that lay close to sea level.

Depositional conditions for the Langport Member were clearly more normally marine, so that corals, echinoids and conodonts occur (Swift 1999b-d). However, desiccation features are common and the water was clearly still very shallow and subject to salinity variation. It is this last feature that appears most likely to have governed the size and diversity of the fauna. The two sections described have abnormally small thicknesses of this Member (2.5 m at Lavernock Point and c. 1.5 m at St. Audrie's Bay). Thicker successions of the Langport Member (to c. 8 m) in Devon or the Midlands show that significant condensation has occurred in the two sections considered, and this will have seriously affected the value of the authors' data. This condensation results in significant non- sequences at the many hardgrounds in these attenuated successions, which would have biased the record in favour of epifaunal forms. Loss of infauna merely reflects these changing conditions, but local disappearances need not mean extinctions. It is difficult to see how global extinction levels can be identified when there are yet so few records of faunas from open marine environments recorded across this interval, and there is no means of providing accurate correlation from them to the Penarth Group.

In the Blue Lias, an open marine environment was rapidly established a little above the base, with ammonites re-colonizing the NW European seaways. However, faunal evidence from the two sections studied may well be taphonomically biased, because in the Blue Lias aragonitic fossils are commonly absent (Wright et al. 2003.). The extent of this loss can be demonstrated where early post- depositional silicification has preserved the originally aragonitic shells; there, the recorded faunas are significantly richer. Thus little can be concluded about the original Blue Lias fauna from the lists published by Mander et al. (2008).

The Triassic-Jurassic Boundary Extinction Event has generated a plethora of papers in the past decade. Unfortunately, some of the published information on the faunas lacks the critical rigour necessary to allow a full evaluation of its relevance. Mander et al. have not recognized that the geological setting of the Penarth Group renders their results of limited value outside of the immediate vicinity, and their presentation of their results is misleading. For example, their figures 3-6 are actually records of discontinuous spot-sampling that should be presented as histograms. They are not continuous records, as the lines connecting the samples might imply; in fact, these connecting lines have no meaning.

Until full documentation of the faunas from open marine successions across the Triassic-Jurassic boundary is available, little value can be attached to data from quasi-marine successions, such as those, for example, from NW Europe. The number of such open marine successions known so far hardly exceeds a handful, reflecting a historic eustatic lowstand of sea level. It is likely that horizons of extinctions in such successions will be found to differ significantly from those recorded from successions in which salinity fluctuations and lithofacies changes provide a dominant faunal control.

16 May 2008

Richard J. Twitchett & Luke Mander reply: We would like to thank John Cope for his comments and will endeavour to address all of his points in the order he presents them. We are pleased that Cope agrees with us that the depositional setting of the lower Westbury Formation was shallow water, comprising, at least in part, 'restricted lagoons subject to fluctuating salinity' (Mander et al. 2008, p. 320). Certainly, we would expect the biota to respond to local environmental conditions in such settings and one possible interpretation of the body size fluctuations within the lower Westbury Formation is indeed that they reflect salinity; we never suggested otherwise. However, to suggest that the highly diverse and abundant assemblage recorded in the upper Westbury Formation indicates restricted salinity is untenable: brackish settings are species-poor and typically inhabited by organisms near the limits of their tolerances (e.g. Remane 1934; Barnes 1989).

Salinity, biotic crisis and the 'Dead Zone '. We are pleased that Cope agrees with us that the lower Gotham Member fossil fauna is diverse, marine and contains many taxa also present in the underlying Westbury Formation (Mander et al. 2008, pp. 227-228). It is this fauna that is strongly affected by the biotic crisis. Cope's suggestion that the drop in diversity and subsequent 'Dead Zone' is merely due to local salinity changes is questionable. First, significant numbers of taxa have their global last appearances at this level (Mander et al. 2008, figs 3 and 4). Second, as clearly stated by Cope and in the references he cites (e.g. Hemsley 1989; Warrington & Ivimey-Cook 1992; Hounslow et al. 2004), the upper Gotham Member strata were not all deposited under restricted salinity conditions, but fully marine beds are also present and alternate with ones containing freshwater taxa. In addition, small freshwater taxa may be transported into nearshore marine settings, and it is perhaps pertinent that 'non-marine bivalves are unknown' from the Penarth Group (Ivimey-Cook et al. 1999, p. 83). Until and unless it is demonstrated unequivocally that the beds we sampled were deposited under nonmarine conditions we reject salinity as the key control on the palaeoecological changes we recorded. Third, as Cope himself again notes, the depositional setting of the overlying Langport Member is 'clearly more normally marine', as is the Blue Lias Formation, yet none of the marine taxa that disappeared in the lower Gotham Member reappear in these marine beds. There was apparently no significant difference in overall sea level or salinity between the lower Gotham Member and the Langport Member (Hesselbo et al. 2004, fig. 4, p. 368), so if these were the dominant controls on diversity, ecology, and presence or absence, as Cope implies, the assemblages of these two units should be similar. The marine assemblages of the upper Gotham Member, Langport Member and lower few metres of the Blue Lias Formation are, however, characterized by low diversity, low abundance, small size, low evenness and high dominance, in stark contrast to the marine assemblages of the uppermost Westbury Formation, lower Gotham Member, and younger horizons within the Blue Lias (Mander et al. 2008, figs 5 and 6). Choice of sections. Cope suggests that our choice of sections has 'seriously affected the value of [our] data'. Much as it would have been desirable to have sampled every bed in every known section, practically this is not possible. The St. Audrie's Bay section was chosen as it is well known, well exposed and was recently proposed as a Global Stratotype section and Point candidate section (Warrington et al. 1994); the Lavernock section was chosen to provide a comparison. Cope's suggestions that the Langport Member data are biased 'in favour of epifaunal forms' at 'many hardgrounds' and the 'loss of infauna merely reflects [this]' are not supported by the facies we sampled or the taxa we actually recorded. Of the bivalves we recovered from the Langport Member only Liostrea hisingeri and Oxytoma fallax have epifaunal life habits, and these taxa were found in association with semi-infaunal and/or infaunal bivalves (Mander et al. 2008, figs 3 and 4).

Correlation of the Penarth Group. We feel that Cope's view that there is 'no means of providing accurate correlation from [open marine sections] to the Penarth Group' is over-despondent and does not give due credit to the recent progress made in this area (e.g. Hesselbo et al. 2007, fig. 1, p. 2). Of central importance to the improving Triassic-Jurassic stratigraphie framework is magnetostratigraphy (e.g. Hounslow et al. 2004) and the carbon isotope record (e.g. Hesselbo et al. 2004), which have proved valuable methods of global correlation. Boundary sections from the Queen Charlotte Islands, British Columbia, are of particular relevance to this discussion because they represent an outer shelf to upper slope depositional setting (Longridge et al. 2007), and preserve a dramatic extinction event among the Radiolaria that is coincident with a negative anomaly in organic carbon isotope values (Williford et al. 2007). The biotic crisis we have recorded in the SW UK 'apparently began just prior to the onset of the negative carbon isotope excursion' (Mander et al. 2008, p. 328) and can, therefore, be plausibly correlated with the radiolarian crisis reported from open marine sections in British Columbia (e.g. Longridge et al. 2007) and from deep-sea cherts in Kurusu, Japan (Hori et al. 2007). Cope notes that 'there are few records of faunas from open marine environments recorded across [the Triassic- Jurassic] interval'. There are certainly few published records, compared with those of other extinction intervals (Twitchett 2006), which is why Longridge et al. (2007, p. 143) commented that, 'any stratigraphie section that spans the Triassic-Jurassic boundary is extremely important because of its potential contributions to understanding the [Late Triassic biotic crisis]'. We concur, and note that there is no global shortage of marine sections; hitherto they simply have not received the attention they deserve. Fortunately this is changing, and rapidly so: watch this space!

Dissolution and the preservation of aragonitic taxa. Cope highlights the fact that skeletal material composed of aragonite is typically less well represented in the fossil record than that composed of calcite. This is indeed a potentially important bias for all studies of past biodiversity and is one of the taphonomic filters that means that fossil assemblages never precisely census the original, living fauna. Post-depositional dissolution certainly occurred in the Westbury Formation, where known calcitic macrofossils occur as empty moulds (Hesselbo et al. 2004), and at some localities, fossil assemblages from the Blue Lias Formation are biased in favour of calcitic taxa (Wright et al. 2003). Evidence for carbonate dissolution in the Lilstock Formation is, however, more limited and radial calcite ooids, for example, are well preserved in the upper Gotham Member (Hesselbo et al. 2004, fig. 7, p. 370). Thus, samples with the lowest recorded diversities derive from the upper Gotham Member, where conditions apparently favoured carbonate preservation, whereas the highest diversity samples are found in the Westbury Formation, where conditions apparently favoured carbonate dissolution. Additionally, there appears to be little change in the proportion of calcitic and originally aragonitic taxa in our upper Westbury Formation and Lilstock Formation samples from St. Audrie's Bay (i.e. spanning the biotic crisis), and it therefore appears unlikely that post-mortem shell dissolution or preferential dissolution of aragonitic taxa was the primary control on measured diversity at this time. We acknowledge, however, that post-mortem dissolution of aragonitic taxa may have affected our perceptions of the nature and timing of the post-crisis recovery interval recorded by the Blue Lias Formation (see also Mander & Twitchett 2008).

Presentation of our results. Cope takes issue with the presentation of our results, stating that the figures are 'misleading' because we connected our discrete sampling points with lines. Notwithstanding the facts that this is standard practice when displaying discrete numerical data, such as isotope values (e.g. Hesselbo et al. 2004), against stratigraphic sections, and that neither reviewers nor the Editor found this in any way misleading, the reason we did this was simply to reinforce the fact that the siliciclastic and carbonate samples were treated separately so as to avoid the pitfall of comparing mudstone with limestone. Histograms are used to present frequency data from a set of data binned into artificial classes (Dytham 2003, p. 47). None of our data were binned, so histograms are inappropriate in this case.

The value of our data. Our study represents the first such high- resolution, quantitative, palaeoecological analysis of Triassic- Jurassic benthic assemblages from any section worldwide. We interpreted the palaeoecological changes recorded in these shallow- water environments of the SW UK as reflecting changes associated with the apparent mass extinction event that is evident from global- level datasets. If this hypothesis is correct then we would expect similar results from other shallow marine Triassic-Jurassic sections in the region and worldwide, recognizing that there may be some regional, or environmental, variation in timing and magnitude. The scientific value of our study is demonstrably clear: novel, quantitative data collected and analysed in a rigorous manner, building on a previous scientific study (Kiessling et al. 2007) and resulting in a hypothesis with predictions that are testable from the fossil record. If other sections outside the NW European region also show evidence of a biotic crisis at a similar stratigraphie level (e.g. Queen Charlotte Islands, British Columbia) then, far from our results having 'little value outside the immediate vicinity' as Cope states, our interpretation would be fully justified. Even if the interpretation needs to be modified or rejected in light of future work, the value of our data is in no way diminished; this is, after all, how science progresses.

3 June 2008


ALLISON, P.A. & WRIGHT, V.P. 2005. Switching off the carbonate factory: A-tidality, stratification and brackish wedges in epeiric seas. Sedimentary Geology, 179, 175-184.

ANDERSON, F.W. 1964. Rhaetic Ostracoda. Bulletin of the Geological Survey of Great Britain, 21, 133-174.

BARNES, R.S.K. 1989. What, if anything, is a brackish water fauna? Transactions of the Royal Society of Edinburgh, Earth Sciences, 80, 235-240.

BARRAS, C.O. & TWITCHETT, R.J. 2007. Response of the marine infauna to Triassic/Jurassic environmental change: ichnological data from southern England. Palaeogeography, Palaeoclimatology, Palaeoecology, 244, 223-241.

BOOMER, I.D., DUFFIN, C.J. & SWIFT, A. 1999. Arthropods 1: Crustaceans. In: SWIFT, A. & MARTILL, D.M. (eds) Fossils of the Rhaetian Penarth Group. Palaeontological Association, Field Guides to Fossils, 9, 129-148.

BOTTJER, D.J., SCHUBERT, J.K. & DROSER, M.L. 1996. Comparative evolutionary palaeoecology: assessing the changing ecology of the past. In: HART, M.B. (ed.) Biotic Recovery from Mass Extinction Events. Geological Society, London, Special Publications, 102, 1- 13.

DYTHAM, C. 2003. Choosing and Using Statistics: a Biologist's Guide. Blackwell Science, Oxford.

HALLAM, A. & EL SHAARAWY, Z. 1982. Salinity reduction of the end- Triassic sea from the Alpine region into northwestern Europe. Lethaia, 15, 169-178.

HALLAM, A. & WIGNALL, P. 2004. Discussion on sea-level change and facies development across potential Triassic&Jurassic boundary horizons, SW Britain. Journal of the Geological Society, London, 161, 1053-1056.

HAMILTON, D. 1961. Algal growths in the Rhaetic Gotham Marble of southern England. Palaeontology, 4, 324-333.

HEMSLEY, A.R. 1989. The ultrastructure of the spore wall of the Triassic bryophyte Naiadita lanceolata. Review of Palaeobotany and Palynology, 61, 89-99. HESSELBO, S.P., ROBINSON, S.A. & SURLYK, F. 2004. Sea-level change and facies development across potential Triassic/Jurassic boundary horizons, SW Britain. Journal of the Geological Society, London, 161, 365-379.

HESSELBO, S.P., MCROBERTS, C.A. & PALFY, J. 2007. Triassic- Jurassic boundary events: problems, progress, possibilities. Palaeogeography, Palaeoclimatology, Palaeoecology, 244, 1-10.

HORI, R.S., FUJIKI, T., INOUE, E. & KIMURA, J. 2007. Platinum group element anomalies and bioevents in the Triassic-Jurassic deep- sea sediments of Panthalassa. Palaeogeography, Palaeoclimatology, Palaeoecology, 244, 391-406.

HOUNSLOW, M.W., POSEN, P.E. & WARRINGTON, G. 2004. Magnetostratigraphy and biostratigraphy of the Upper Triassic and lowermost Jurassic succession, St. Audrie's Bay, UK. Palaeogeography, Palaeoclimatography, Palaeoecology, 213,331-358.

IVIMEY-CooK, H.C., HODGES, P., SWIFT, A. & RADLEY, J.D. 1999. Bivalves. In: SWIFT, A. & MARTILL, D.M. (eds) Fossils of the Rhaetian Penarth Group. Palaeontological Association, Field Guides to Fossils, 9, 83-127.

KIESSLING, W., ABERHAN, M., BRENNEIS, B. & WAGNER, P.J. 2007. Extinction trajectories of benthic organisms across the Triassic/ Jurassic boundary. Palaeogeography, Palaeoclimatology, Palaeoecology, 244, 201-222.

LONGRIDGE, L.M., CARTER, E.S., SMITH, P.L. & TIPPER, H.W. 2007. Early Hettangian ammonites and radiolarians from the Queen Charlotte Islands, British Columbia and their bearing on the definition of the Triassic-Jurassic boundary. Palaeogeography, Palaeoclimatology, Palaeoecology, 244, 142-169.

MANDER, L. & TWITCHETT, R.J. 2008. Quality of the Triassic- Jurassic bivalve fossil record in northwest Europe. Palaeontology, in press.

MANDER, L., TWITCHETT, R.J. & BENTON, MJ. 2008. Palaeoecology of the Late Triassic extinction event in the SW UK. Journal of the Geological Society, London, 165, 319-332.

MAYALL, MJ. 1983. An earthquake origin for synsedimentary deformation in a late Triassic lagoonal sequence, southwest Britain. Geological Magazine, 120, 613-622.

MAYALL, MJ. & WRIGHT, V.P. 1981. Algal tuft structures in stromatolites from the Upper Triassic of south-west England. Palaeontology, 24, 655-660.

PRUSS, S., FRAISER, M. & BOTTJER, D.J. 2004. Proliferation of Early Triassic wrinkle structures: implications for environmental stress following the end-Permian mass extinction. Geology, 32, 461- 464.

RADLEY, J.D. 2002. The Late Triassic and Early Jurassic succession at Southam Cement Works, Warwickshire. Mercian Geologist, 15, 171-174.

RASMUSSEN, K.A., MCINTYRE, IJ. & PRUFERT, L. 1993. Modern stromatolite reefs fringing a brackish coastline, Chetumal Bay, Belize. Geology, 21, 199-202.

REMANE, A. 1934. Die Brackwasserfauna. Verhandlungen der Deutschen Zoologischen Gesellschaft, 36, 34-74.

SCHUBERT, J.K. & BOTTJER, D.J. 1992. Early Triassic stromatolites as post-mass extinction disaster forms. Geology, 20, 83-886.

SHEEHAN, P.M. & HARRIS, M.T. 2004. Microbialite resurgence after the Late Ordovician extinction. Nature, 430, 75-78.

SHEN, J.W. & WEBB, O.E. 2004. Famennian (Upper Devonian) stromatolite reefs at Shatang, Guilin, Guangxi, South China. Sedimentary Geology, 170, 63-84.

SWIFT, A. 1995. A review of the nature and outcrop of the 'White Lias' facies of the Langport Member (Penarth Group: Upper Triassic) in Britain. Proceedings of the Geologists' Association, 106, 247- 258.

SWIFT, A. 1999a. Stratigraphy (including biostratigraphy). In: SWIFT, A. & MARTILL, D.M. (eds) Fossils of the Rhaetian Penarth Group. Palaeontological Association, Field Guides to Fossils, 9, 15- 30.

SWIFT, A. 1999b. Echinoderms. In: SWIFT, A. & MARTILL, D.M. (eds) Fossils of the Rhaetian Penarth Group. Palaeontological Association, Field Guides to Fossils, 9, 161-167.

SWIFT, A. 1999c. Corals. In: SWIFT, A. & MARTILL, D.M. (eds) Fossils of the Rhaetian Penarth Group. Palaeontological Association, Field Guides to Fossils, 9, 169-174.

SWIFT, A. 1999d. Conodonts. In: SWIFT, A. & MARTILL, D.M. (eds) Fossils of the Rhaetian Penarth Group. Palaeontological Association, Field Guides to Fossils, 9, 183-190.

SWIFT, A. & MARTILL, D.M. (EDS) 1999. Fossils of the Rhaetian Penarth Group. Palaeontological Association, Field Guides to Fossils, 9.

TWITCHETT, R.J. 2006. The palaeoclimatology, palaeoecology and palaeoenvironmental analysis of mass extinction events. Palaeogeography, Palaeoclimatology, Palaeoecology, 232, 190-213.

WALTER, M.R. 2001. Stromatolites. In: BRIGGS, D.E.G. & CROWTHER, P.R. (eds) Palaeobiology II. Blackwell Science, Oxford, 376-379.

WARRINGTON, G. & IVIMEY-COOK, H.C. 1992. Triassic. In: COPE, J.C.W., INGHAM, J.K. & RAWSON, P.P. (eds) Atlas of Palaeogeography and Lithofacies. Geological Society, London, Memoirs, 13, 97-106.

WARRINGTON, G., AUDLEY-CHARLES, M.G. & ELLIOTT, R.E. ET AL. 1980. A Correlation of D-iassic Rocks in the British Isles. Geological Society, London, Special Reports, 13.

WARRINGTON, G., COPE, J.C.W. & IVIMEY-COOK, H.C. 1994. St. Audrie's bay, Somerset, England: a candidate Global Stratotype section and Point for the base of the Jurassic System. Geological Magazine, 131, 191-200.

WATERS, R.A. & LAWRENCE, D.J.D. 1987. Geology of the South Wales Coalfield, Part III, the Country around Cardiff. Memoir of the British Geological Survey.

WHALEN, M.T., DAY, J., EBERLI, G.P. & HOMEWOOD, P.W. 2002. Microbial carbonates as indicators of environmental change and biotic crises in carbonate systems: examples from the Late Devonian, Alberta basin, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology, 181, 127-151.

WHITESIDE, D.I. & MARSHALL, J.E.A. 2007. The age, fauna and palaeoenvironment of the Late Triassic fissure deposits of Tytherington, South Gloucestershire, UK. Geological Magazine, 145, 105-147.

WIGNALL, P.B. 2001. Sedimentology of the Triassic-Jurassic boundary beds in Pinhay Bay (Devon, SW England). Proceedings of the Geologists ' Association, 112, 349-360.

WILLIFORD, K.H., WARD, P.D., GARRISON, G.H. & BUICK, R. 2007. An extended organic carbon-isotope record across the Triassic-Jurassic boundary in the Queen Charlotte Islands, British Columbia, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology, 244, 290-296.

WRIGHT, V.P., CHERNS, L. & HODGES, P. 2003. Missing molluscs: field testing taphonomic loss in the Mesozoic through early large- scale aragonite dissolution. Geology, 31, 211-214.

JONATHAN D. RADLEY, Warwickshire Museum, Market Place, Warwick CV34 4SA, UK (e-mail: [email protected])

RICHARD J. TWITCHETT, School of Earth, Ocean and Environmental Sciences, University of Plymouth, Drake Circus, Plymouth PM 8AA, UK (e-mail: [email protected]

LUKE MANDER, School of Biology and Environmental Science, University College Dublin, Belfield, Dublin 4, Ireland

J. C. W COPE, Department of Geology, National Museum of Wales, Cathays Park, Cardiff CF10 3NP, UK

Scientific editing by John Marshall.

Copyright Geological Society Publishing House Sep 2008

(c) 2008 Journal of the Geological Society. Provided by ProQuest LLC. All rights Reserved.