Isotopic and palynological evidence for a new Early Jurassic environmental perturbation Palaeogeography, Palaeoclimatology, Palaeoecology

The Early Jurassic Epoch was a predominantly greenhouse phase of Earth history, but a comprehensive understanding of its climate dynamics is hampered by a lack of high resolution multi-proxy environmental records. Here we report a geologically brief (approximately several hundred thousand years) negative carbon isotope excursion (CIE) of 2 – 3 ‰ in both marine and terrestrial materials, recognised for the ﬁ rst time for the Late Sinemurian Substage (Early Jurassic, ~194 Ma) of eastern England. The Late Sinemurian carbon isotope excursion, which is termed the S-CIE herein, is accompanied by peaks in the abundance of the pollen grain Classopollis classoides and the dino ﬂ agellate cyst Liasidium variabile . Classopollis classoides was thermophilic and is a reliable proxy for hot/warm climatic conditions. Liasidium variabile is interpreted as thermophilic and eutrophic using multivariate statistics, its ﬂ uorescence properties being similar to living heterotrophic dino ﬂ agellate cysts, and its association with C. classoides . Moreover, the morphological and ecological similarities of L. variabile to the Cenozoic genus Apectodinium are noteworthy. The co-occurrence of the acmes of C. classoides and L. variabile with a negative CIE is interpreted here as having wide geographical signi ﬁ cance due to the marine and terrestrial carbon isotope signals being precisely in phase within an open marine setting. This is consistent with an oceanic – atmospheric injection of isotopically-light carbon, coupled with global warming and increased marginal marine nutrient supply, possibly the result of increased precipitation due to an enhanced hydrological cycle or a seasonally-strati ﬁ ed water column. A probable sea level rise of at least regional extent has been identi ﬁ ed at the L. variabile event in other records, which supports this putative phase of global warming. All these features are common to the Paleocene/Eocene thermal maximum (PETM, ~56 Ma), and there are also similarities with the Early Toarcian oceanic anoxic event (T-OAE, ~182 Ma). © 2012 Natural Environment Research Council. Published by Elsevier B.V. All rights reserved.


Introduction
The Jurassic Period was a mainly greenhouse phase of Earth history, characterised by global palaeotemperatures that were 5-10°C higher than modern levels based on palaeo-modelling results (Chandler et al., 1992;Rees et al., 1999;Sellwood and Valdes, 2008). During the Early Jurassic the supercontinent Pangaea fragmented into Laurasia and Gondwana, creating major seaways including the proto Central Atlantic Ocean and marine connections between western Tethys and the north polar region (Fig. 1). Sea levels rose steadily throughout the Early Jurassic (Haq et al., 1987(Haq et al., , 1988 and much of Europe and surrounding areas in the northern hemisphere were covered by warm, relatively shallow and stable epicontinental seas (summary in Coward et al., 2003). Sedimentation was dominated by rhythmic cycles of fossiliferous marine shelfal muds with subordinate carbonates and sands within large fault-bounded basins that were created by crustal extension. Northwest Europe probably lay around 10°south of its present latitude (Bradshaw et al., 1992). British Lower Jurassic stratigraphy has been summarised by Cope et al. (1980) and Simms et al. (2004).
The Early Toarcian oceanic anoxic event (T-OAE,~182 Ma according to the timescale of Gradstein et al., 2012) was a short-lived major phase of global warming during the late Early Jurassic, and is marked by a significant negative carbon isotope (δ 13 C) excursion (CIE) (Jenkyns, 2010). The T-OAE and other large CIEs have been interpreted as being due to the release of isotopically-light, largely biogenic methane from sub-seafloor accumulations of methane clathrate Kemp et al., 2005;Cohen et al., 2007). Similar, but smaller scale, events have also been recognised in the Early Jurassic at the Pliensbachian-Toarcian boundary (~183 Ma; Hesselbo et al., 2007;Bodin et al., 2010;Littler et al., 2010), and at the Sinemurian-Pliensbachian boundary (~191 Ma; Korte and Hesselbo, 2011).

Material and methods
The Copper Hill Borehole was an exploratory stratigraphical borehole drilled by the British Geological Survey (BGS) during 1991 in Copper Hill Quarry, near Ancaster, Lincolnshire, eastern England; the national grid reference is SK 9787 4265 ( Fig. 4; Berridge et al., 1999, fig. 19). The Lower and Middle Jurassic (Sinemurian to Bajocian) palynology of the Copper Hill Borehole was described by Riding (1992). The palynology samples were prepared using the standard acid digestion method including mild oxidation with cold nitric acid (Wood et al., 1996), and the rock samples, prepared organic residues, and microscope slides are all housed  Smith et al., 1994) with the locations of records of Liasidium variabile indicated by the 12 filled circles. The locations were taken from Brittain et al. (2010). Continental areas are indicated in grey, and the black lines depict selected modern coastlines.  Riding et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 374 (2013) [16][17][18][19][20][21][22][23][24][25][26][27] in the collections of the British Geological Survey, Keyworth, Nottingham NG12 5GG, United Kingdom.
For the carbon isotope analysis on the bulk and palynomorph concentrates, the samples were homogenised and treated with 5% hydrochloric acid to remove any calcite. 13 C/ 12 C analyses were performed by combustion in a Costech Elemental Analyser (EA) on-line to a VG TripleTrap and Optima dual-inlet mass spectrometer, with δ 13 C values calculated to the VPDB scale using within-run laboratory standards calibrated against NBS18, NBS-19 and NBS-22. Replicate analysis of well-mixed samples indicated a precision of ±b0.1‰ (1 SD). For the belemnite carbon and oxygen isotope analysis, approximately 50 μg of carbonate was used for isotope analysis using a GV IsoPrime mass spectrometer plus Multiprep device. The isotope values (δ 13 C, δ 18 O) are reported as per mille (‰) deviations of the isotopic ratios ( 13 C/ 12 C, 18 O/ 16 O) calculated to the VPDB scale using a within-run laboratory standard calibrated against NBS standards. Analytical reproducibility of the standard calcite (KCM) is b0.1‰ for δ 13 C and δ 18 O.
The interval studied in the Copper Hill Borehole is relatively fossiliferous and the standard ammonite biozones and biosubzones have been determined (Tables 1-3). The ammonite biostratigraphy, however, is of limited utility in terms of identifying the duration of relatively short-lived palaeoclimatic events such as the S-CIE. The S-CIE is not currently well-constrained chronostratigraphically because there are no accurate estimates for the absolute age ranges of the three Late Sinemurian ammonite biozones as suitable cyclic sedimentary successions have not been studied. A cyclostratigraphical calibration of the long-term decline in seawater 87 Sr/ 86 Sr values throughout the Early Jurassic gives an estimate for ammonite biozone duration of~1 Ma each (Weedon and Jenkyns, 1999;Gradstein et al., 2012). This means that the S-CIE may have lasted a few hundred   Table 1 The quantitative stratigraphical distribution of palynomorphs between 200 m and 150 m in the Copper Hill Borehole. The numbers in the cells are percentages. An 'X' denotes that the respective palynomorph has been recorded outside the count. An ellipsis (…) denotes the absence of the respective palynomorph. In the ammonite biozone/biosubzone column, the index taxa have been abbreviated; A.s. = Arnioceras semicostatum, E.r. = Euagassiceras resupinatum, C. tur. = Caenisites turneri, As.o? = Asteroceras obtusum?, Ox.ox. = Oxynoticeras oxynotum, Ox.si. = Oxynoticeras simpsoni, E. rar. = Echioceras raricostatum, E. rari. = Echioceras raricostatoides and L.m. = Leptechioceras macdonnelli. thousand years, which is comparable to the Paleocene/Eocene thermal maximum (PETM,~56 Ma). Other potentially analogous CIEs are at the T-OAE, which has a poorly constrained duration of~200-900 ka (Suan et al., 2008;Kemp et al., 2011), and at the Selli Event in the Early Aptian (~120 Ma). The Selli Event, or OAE 1a, was reported to have lasted 1.11 ± 0.11 Ma by Malinverno et al. (2010).

Palynomorph and kerogen assemblages
The palynofloras in 26 samples between 200 m and 150 m from the Sinemurian of the Copper Hill Borehole are illustrated in Table 1. The succession studied represents an open marine palaeoenvironment within a geographically extensive seaway (Coward et al., 2003). All the samples yielded abundant and well-preserved palynofloras of relatively low taxonomic diversity. The floras are overwhelmingly dominated by gymnospermous pollen grains, particularly bisaccate forms including Vitreisporites pallidus, Classopollis classoides and Perinopollenites elatoides. Classopollis classoides is a long-ranging species and is present throughout the entire Sinemurian to Bajocian succession drilled by the Copper Hill Borehole (Riding, 1992). The acme of this species is at 177 m, at which level it comprises 24.3% of the entire palynoflora and 32.5% of the terrestrial palynomorphs ( Fig. 2; Tables 1, 2). This maximum is coincident with the lowest percentage (26.6%) of P. elatoides (Table 1). Other pollen present comprise Araucariacites australis, Cerebropollenites macroverrucosus and Chasmatosporites spp. Pteridophyte spores are significantly subordinate to gymnospermous pollen, and are dominated by simple, smooth forms attributable to the genus Cyathidites. Other spores encountered include Baculatisporites/Osmundacidites spp., Concavissimisporites verrucosus, Coronatispora valdensis, Dictyophyllidites spp., Duplexisporites sp., Gleicheniidites senonicus, Ischyosporites variegatus, Kraeuselisporites reissingeri, Neoraistrickia gristhorpensis, Retitriletes austroclavatidites, Sestrosporites pseudoalveolatus, Todisporites spp. and Uvaesporites argenteaeformis. Rare reworked Carboniferous spores were recognised; these include Densosporites spp., Lycospora pusilla and Tripartites vetustus. The freshwater/brackish alga Botryococcus is present in low proportions (Table 1).
Marine palynomorphs are present in consistently lower proportions than terrestrially-derived taxa. Liasidium variabile is the only dinoflagellate cyst recognised and was recorded between 182 m and 172 m, within the Upper Sinemurian part of the Brant Mudstone Formation, where it is largely confined to the O. oxynotum ammonite Biozone ( Fig. 2; Tables 1-3). The acme of L. variabile is at 177 m, at which level it comprises 21.3% of the entire palynoflora ( Fig. 2; Tables 1, 2). The marine acanthomorph acritarch genus Micrhystridium is present throughout in significant proportions. Other marine palynomorphs are present in low numbers; these are foraminiferal test linings, Tasmanites spp. and Veryhachium sp. (Table 1). The palynomorph species recognised in this study are listed, with author citations, in Appendix 1. The palynology of the entire Lower and Table 2 Raw data on carbon and oxygen isotopes and the percentages of Liasdium variabile and Classopollis classoides between 200 m and 150 m in the Copper Hill Borehole. The geochemical data comprise δ 13 C org measured on the bulk organic carbon for 37 samples, δ 13 C bel and δ 18 O bel measured from five belemnite specimens and δ 13 C palynomorph concentrate measured on the organic residues from eight samples. The figures for L. variabile and C. classoides are percentages of the overall palynoflora. In the ammonite stratigraphy column, the ammonite biozones and biosubzones have been abbreviated; A.s. = Arnioceras semicostatum, E.r. = Euagassiceras resupinatum, C. tur. = Caenisites turneri, ?As.o. = ?Asteroceras obtusum, Ox.ox. = Oxynoticeras oxynotum, Ox.si. = Oxynoticeras simpsoni, E. rar. = Echioceras raricostatum, E. rari. = Echioceras raricostatoides and L.m. = Leptechioceras macdonnelli.

Substage
Ammonite biozones Depth (m) Middle Jurassic (Sinemurian-Bajocian) succession cored by the Copper Hill Borehole was documented by Riding (1992). The percentages of the main kerogen macerals are given in Table 3. The organic residues are generally dominated by palynomorphs. Black wood and various plant tissues are subordinate to palynomorphs. Amorphous organic material (AOM) is largely relatively rare, although it is most abundant at 179 m where it attained 32% (Table 3).

Carbon and oxygen isotope geochemistry
Carbon isotope values were obtained from bulk organic matter, palynomorph concentrates and belemnite fragments between 201 m and 150 m in the Copper Hill Borehole; these geochemical data are given in Table 2. This interval includes the acme of Classopollis classoides and is within and adjacent to the range of L. variabile ( Table 2). The materials analysed are both terrestrial and marine and thus allow a test as to whether any carbon isotope perturbation affected the global carbon cycle. Specifically, the palynomorph concentrates are dominated by terrestrial materials and the belemnite fragments represent the marine realm. A consistent negative CIE of 2-3‰ occurs in the Copper Hill Borehole (Fig. 2). The CIE as defined by the bulk sample δ 13 C TOC is broadly symmetrical, and is recorded between 187 m and 176 m. There are two separate peaks; an older one at 186 m and a younger, larger one at 179 m (Fig. 2). The negative CIE is also recorded in the δ 13 C of palynomorph concentrates, which are overwhelmingly dominated by terrestrial material such as wood fragments, pollen and spores, and hence in part reflect isotopic changes in atmospheric CO 2, in addition to a potentially minor unknown contribution from changing vegetation types. The δ 13 C values of the only five belemnites (nekto-benthic marine molluscs) recovered in this interval also show a negative CIE. The peak excursion of all three materials is temporally coincident with the acmes of Classopollis classoides and L. variabile (Fig. 2).
In contrast to the carbon isotope record, the δ 18 O belemnite values are relatively uniform ( Fig. 2; Table 2) (mean δ 18 O =− 1.6 ± 0.3‰). Although the negative CIE and the coincident Classopollis classoides and L. variabile peaks are inferred here to represent a period of global warming, the lack of a clear warming signal from belemnites may represent the effect of sea level rise due to thermal expansion of seawater, and hence locally deeper cooler water, as has recently also been suggested for the Sinemurian-Pliensbachian boundary (Korte and Hesselbo, 2011). Alternatively, water column stratification and/ or local upwelling of cool intermediate waters may have increased. It is not possible to use TEX 86 or MBT/CBT as temperature proxies because the succession studied here is too thermally mature. Indeed, most Triassic and Jurassic successions have been too thermally altered for these methods, which have been used successfully on Early Cretaceous to modern material (Kuypers et al., 2001;Schouten et al., 2002).
There has been considerable debate as to the precise palaeoecological preference of Classopollis-producing plants. The genus has been reported in high proportions from near-coastal settings such as sandy bars or coastal islands with well-drained soils (Batten, 1975;Srivastava, 1976;Alvin, 1982;Abbink, 1998). This distribution led Batten and MacLennan (1984) to propose that the family Cheirolepidiaceae were possibly salt marsh shrubs or trees resembling modern mangroves. However Filatoff (1975, fig. 10) and Vakhrameev and Doludenko (1977) suggested that the Cheirolepidiaceae preferred upland slope habitats, and that the abundant Classopollis pollen preserved in nearshore palaeoenvironments was therefore allochthonous. Furthermore, several authors, such as Vakhrameev (1970), considered the Cheirolepidiaceae to be drought-resistant, thermophilous, xerophytic shrubs and trees which resembled modern juniper bushes.
The occurrence of abundant Classopollis during a global warming event is somewhat inconsistent with this genus being a proxy for dry (semiarid to arid) conditions as suggested by, for example, Pocock and Jansonius (1961), Pocock (1972 andVakhrameev 1991). This is because global warming events are normally associated with an enhanced hydrological cycle, and Classopollis has an extremely wide geographical distribution. Hence the conclusions of Batten and MacLennan (1984) that the parent plants of Classopollis were salt marsh or mangrove shrubs/trees are supported herein. However, it is possible that the Cheirolepidiaceae lived in a wide variety of habitats and that they could also survive in semiarid to arid conditions.

Other pollen and spores
The remaining pollen and spore floras are typical of Sinemurian terrestrial palynofloras from Europe (e.g. Srivastava, 1987). The non-Classopollis gymnosperm pollen is dominated by coniferous plants from the orders Pinales and Cycadales. For example Perinopollenites elatoides belongs to the Family Cupressaceae (cypresses). These groups are generally widely distributed, and hence are not especially palaeoecologically significant. However, Chasmatosporites is a possible representative of the Order Cycadales (see Pocock and Jansonius, 1969), and these forms are typical of subtropical and tropical regions today. The majority of the spores are from ferns or club mosses. These groups have extensive geographical extents today. However, Ischyosporites variegatus belongs to the chiefly tropical Family Schizaeaceae (see Couper, 1958). Therefore, the other pollen and spores support the warm climate signal provided by Classopollis classoides, and are entirely consistent with an equable, subtropical setting.

The dinoflagellate cyst L. variabile
The dinoflagellate cyst L. variabile was originally described from the Upper Sinemurian strata of southwest Germany by Drugg (1978). Despite relatively little being known concerning the tabulation style of this species, the accepted view is that L. variabile is the oldest known representative of the order Peridiniales (Bujak and Davies, 1983, p. 56;Evitt, 1985, p. 176, 177;Fensome et al., 1993, p. 121). Its morphology was thoroughly reviewed by Below (1987, p. 128), who noted that the tabulation of this species is not fully developed and that the size and shape of the archaeopyle differ from most other peridinialean taxa in being unusually large. Furthermore the anterior margin of the archaeopyle is geniculate and it lies near to the apex, encroaching on the apical horn (Below, 1987, fig. 68). Riding and Hubbard (1999, p. 27) interpreted L. variabile as a warm water species based on evidence derived from the study of a comprehensive database of Jurassic dinoflagellate cysts using principal component analysis.
Liasidium variabile is an excellent marker for the Upper Sinemurian of the northern hemisphere (Riding, 1987;Poulsen and Riding, 2003;Van de Schootbrugge et al., 2005;Brittain et al., 2010). This species is especially prominent in, and characteristic of, the O. oxynotum ammonite Biozone (~192-191 Ma) of northwest Europe, and is confined to northwest Europe and offshore eastern Canada (Williams et al., 1990) and has not been reliably reported elsewhere (Fig. 1). Dinoflagellates have a planktonic habit, some taxa also having a benthic resting cyst phase (Taylor, 1987). Thus the apparent geographical restriction of L. variabile to northwest Europe and offshore eastern Canada is deemed to reflect the lack of suitable studies of the Lower Jurassic from Africa, the Middle East, South America and southern Asia, coupled with the dominance of non-marine facies in the Early Jurassic of regions such as, for example, Russia and Australia .
Lower Jurassic dinoflagellate cyst assemblages have low species richness, especially in the Hettangian to Lower Pliensbachian (Bujak and Williams, 1979;Woollam and Riding, 1983;Fensome et al., 1996Fensome et al., , 1999MacRae et al., 1996). Upper Sinemurian successions yielding L. variabile are frequently underlain and overlain by strata that are entirely devoid of other dinoflagellate cysts (e.g. Riding, 1987, fig. 3). Van de Schootbrugge et al. (2005, p. 87-88) discussed the highly unusual sudden appearance, abundance and disappearance of L. variabile in Europe and eastern Canada, and explained this stratigraphically isolated acme as being due to major palaeoceanographical change, possibly linked to the opening and flooding of the Hispanic Corridor. These authors suggested that L. variabile migrated into Europe and eastern Canada from western Panthalassa through the Hispanic Corridor into the western Tethys in response to changes in ocean circulation during a sea-level highstand. This contention, however, is somewhat speculative because there are no records of L. variabile from Central America and the eastern Pacific area (Brittain et al., 2010, p. 72). It is equally possible, however, that L. variabile migrated northwestwards into Europe from western Tethys. This putative northerly migration of L. variabile may be an indication that this species tracked warmer water northwards. This contention is consistent with the interpretation of L. variabile as a thermophilic species by Riding and Hubbard (1999), and certain similarities between L. variabile and the peridinioid genus Apectodinium.
Feist-Burkhardt (2009) suggested that the acme of L. variabile in the Late Sinemurian is similar to the abundant occurrence of Apectodinium at the PETM, and also demonstrated that some specimens of L. variabile in unoxidised material exhibit dark brownpigmentation with low autofluorescence intensities (Fig. 5). No pigmented forms were found in the Copper Hill Borehole material, but these samples were all oxidised using nitric acid during the preparation procedure which may have had a bleaching effect. Extant pigmented dinoflagellate cysts, such as produced by Archaeperidinium, Congruentidium and Protoperidinium, with little or no autofluorescence are all heterotrophic (Head, 1994(Head, , 1996Brenner and Biebow, 2001). These characteristics led Feist-Burkhardt (2009) to conclude that L. variabile may have been produced by a heterotrophic and thermophilic dinoflagellate by analogy with extant floras, supposedly like the (consistently non-pigmented) genus Apectodinium. However, the peridinioid affinity of L. variabile is not prima facie evidence of this species being indicative of elevated sea surface temperatures; as with Apectodinium, palaeotemperature is one of several parameters which controlled its distribution.
There are, however, possible alternatives to L. variabile being a warm-water species. For example, a rise in sea level may have caused salinity changes that led to the influx of L. variabile during the Late Sinemurian. Alternatively, upwelling cold oceanic waters may have introduced an influx of nutrients which caused the acme of L. variabile at this time. The low-resolution but apparently invariant levels of δ 18 O throughout the succession studied ( Fig. 2; Korte and Hesselbo, 2011) would be consistent with this hypothesis. Despite these caveats, multivariate statistical evidence based on principal component analysis of the thermophilic nature of this species developed by Riding and Hubbard (1999), together with the co-occurrence of 22 J.B. Riding et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 374 (2013) 16-27 abundant Classopollis classoides and the negative CIE, makes a compelling case for L. variabile being thermophilic. Hence this independent evidence for thermopyly in L. variabile, and the fact that the evidence for heterotrophy in L. variabile is sporadic pigmentation, means that the analogy to Apectodinium is not as strong as suggested by Feist-Burkhardt (2009).

The dinoflagellate cyst genus Apectodinium
The comparison of L. variabile to the Paleogene genus Apectodinium by Feist-Burkhardt (2009) mentioned above in Section 5.3 is an important one in the context of this contribution. Both these peridinialean taxa have relatively transient acmes, and Apectodinium has been considered to be both heterotrophic and thermophilic (e.g. Crouch et al., 2003b). It should however be emphasised that, other than a shared peridinalean affinity and therefore a probable heterotrophic feeding strategy, similarities between these taxa are relatively few. This is especially the case in terms of morphology. Apectodinium has a subpentagonal outline, exhibits a quadra-style archaeopyle, typically bears five horns, is cornucavate and is consistently unpigmented Downie, 1976, 1979;Harland, 1979). By contrast, L. variabile is biconical, normally with two horns, acavate, has a large, heptagonal anterior intercalary (type I) archaeopyle and is occasionally pigmented (Drugg, 1978;Bucefalo Palliani and Riding, 2000, Figs. 5, 6;Feist-Burkhardt, 2009). The similarities between L. variabile and Apectodinium are mostly autoecological, i.e. are related to their similar geological settings. Both these taxa probably exploited rising sea levels and warming climates that would have caused the widespread development of water-column stratification on the continental shelves. These conditions may have caused algal blooms, thereby providing a food source for the heterotrophic motile dinoflagellate stages.
The palaeoecological preference of Apectodinium has recently been extensively debated. It is known as a neritic genus that thrived in estuarine settings (Crouch et al., 2003a;Sluijs et al., 2005). At low-mid latitudes, Apectodinium typically comprises >40% of the dinoflagellate cyst associations during the PETM (e.g. Crouch et al., 2003b;Sluijs et al., 2007;Kender et al., 2012). By contrast, Apectodinium is less common (~20%) in the Arctic, where low salinity dinoflagellate cysts increase Harding et al., 2011). Several studies have concluded that, due to the low latitude preference of Apectodinium, it must have tracked warm water northwards during the PETM (e.g. Crouch et al., 2003a). However, Canonical Correspondence Analysis (CCA) data from the New Jersey Shelf indicates that Apectodinium has only a weak correlation with palaeotemperature based on TEX 86 Brinkhuis, 2009, p. 1791). Furthermore,  found that Apectodinium is generally absent throughout another hyperthermal event, the Eocene thermal maximum 2 (ETM2). Recently, Sluijs et al. (2011, fig . 4) reported the main peak in abundance of Apectodinium occurred well before the PETM at ODP Site 1172 on the East Tasman Plateau. The principal acme of Apectodinium at this locality attained >60% of the assemblage at~0.8 m below the thermal peak of the PETM. Following a decrease to around 2% there was a later peak of Apectodinium, attaining~30% of the assemblage, within the PETM (Sluijs et al., 2011, fig . 3). The fact that the older, more significant, acme of Apectodinium on the East Tasman Plateau is associated with normal sea surface temperatures (~26°C) clearly indicates that this genus responded to variables other than simply temperature. For example, abundant Apectodinium has been interpreted as being indicative of elevated nutrient levels, water stratification and low salinities by Crouch et al. (2003b), Sluijs et al. (2008) and Kender et al. (2012). This is also consistent with a warming world and elevated fluvial nutrient runoff from an associated enhanced hydrological cycle with greater precipitation in mid-high latitudes . In conclusion, Apectodinium is normally abundant during the PETM, but temperature is only one of several factors which controlled its abundance.
Several authors have stated that the genus Apectodinium was probably heterotrophic and fed on other plankton and/or organic detritus (e.g. Brinkhuis et al., 1992Brinkhuis et al., , 1994Crouch et al., 2003b;Sluijs et al., 2005). This interpretation of trophic mode is based on the morphological similarity of Apectodinium with modern cysts of the unequivocally heterotrophic genus Protoperidinium (subfamily Protoperidinioideae). The vast majority of living protoperidiniacean dinoflagellates (including the diplopsalioideans) produces pigmented cysts, thus indicating that pigmentation is connected to heterotrophy in this group of dinoflagellates (Head, 1994(Head, , 1996Rochon et al., 1999;Brenner and Biebow, 2001). However, Apectodinium belongs to the subfamily Wetzelielloideae (see Fensome et al., 1993). Apectodinium and cysts of Protoperidinium are hence both members of the suborder Peridiniineae, but there are substantial familial and subfamilial differences which are based on significant differences in tabulation style. The subfamily Wetzelielloideae has, for example, a quadra-style 2a plate, a stable epicystal tabulation and a fully subdivided cingulum. In contrast, cysts of the subfamily Protoperidinioideae are characterised by a highly variable cinctioid or bipesioid epicystal tabulation and a largely undivided cingulum (Fensome et al., 1993). The subpentagonal outline of Apectodinium is not a primary characteristic of the subfamily Protoperidinioideae (see Fensome et al., 1993, fig. 142). Hence, the evidence for heterotrophy in Apectodinium, based on similarities to cysts of Protoperidinium at the suborder level, appears to be purely circumstantial.
Several studies from marginal marine environments suggest an elevation in levels of marine nutrients in shelf seas, possibly in response to an intensification of increased weathering, terrestrial runoff and/or water column stratification during the PETM (e.g. Crouch et al., 2003b;Sluijs et al., 2006Sluijs et al., , 2007Harding et al., 2011;Kender et al., 2012). It is highly likely therefore that the abundance of Apectodinium at or around the PETM is an indirect response to higher global palaeotemperatures and/or eutrophication.

Similarities of the S-CIE to other environmental change events
The temporal coincidence of the highly restricted thermophilic/ neritic L. variabile and the thermophilic Classopollis acmes in the   Riding et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 374 (2013) 16-27 Copper Hill Borehole, with a marked negative CIE (Fig. 2), suggests that the inferred environmental changes were associated with a perturbation to the carbon cycle. This may have been global in nature due to isotope excursions in both marine and terrestrial materials. Furthermore, the palynological records appear to be at least regional in extent. Van de Schootbrugge et al. (2005, table 3) also recorded coincident abundances of Classopollis spp. and L. variabile in the Late Sinemurian strata of the Mochras Borehole, northwest Wales.
Our new records are similar in some respects to other Mesozoic and Cenozoic hyperthermal events such as the T-OAE and the PETM (Cohen et al., 2007), which exhibit negative CIEs coeval with global warming and major changes in marine palynomorph assemblages. Furthermore, the S-CIE association with the peridiniacean dinoflagellate cyst L. variabile exhibits a striking similarity to the PETM. The latter hyperthermal event, occurring~56 Ma ago, is characterised by a negative CIE and abundant levels of the peridiniacean genus Apectodinium (e.g. Crouch et al., 2003b). During the PETM, the proportion of isotopically-light carbon in the atmosphere and oceans rapidly increased, hypothetically due to the catastrophic destabilisation of methane clathrate, magmatic heating of organic-rich sediments or direct volcanic outgassing (e.g. Dickens, 2011). The PETM is a particularly well-documented hyperthermal event~56 Ma ago, that lasted approximately 120-220 ka (Röhl et al., 2007;Murphy et al., 2010). The duration of the S-CIE may have been a few hundred thousand years, which is therefore comparable to the PETM and the T-OAE (see above, Section 2).
The negative CIEs at the T-OAE, OAE 1a and the PETM are similar in form to that of the S-CIE ( Fig. 6; Gröcke et al., 1999;Mort et al., 2007;Kuhnt et al., 2011). The 2-3‰ magnitude of the S-CIE is also similar to that of carbonate records of OAE 1a (e.g. Mort et al., 2007;Jenkyns, 2010). However, the magnitude of the S-CIE is smaller than that of the organic PETM (~5‰; summary in McInerney and Wing, 2011) and many T-OAE records (6‰; e.g. Hesselbo et al., 2007;Hermoso et al., 2009), suggesting either a smaller volume of injected carbon and/or a heavier isotopic value. The pulsed onset of the S-CIE is similar to those of the T-OAE, OAE 1a, and PETM (Fig. 6). However the S-CIE is of a much longer duration (~1 Ma) at the Early-Late Sinemurian transition in the Copper Hill Borehole, being perhaps therefore closer in character to the excursion at the Pliensbachian-Toarcian boundary followed by the T-OAE (Hesselbo et al., 2007). Specifically, the δ 13 C (bulk) record of the S-CIE shows an initial isolated spike at 186 m in the upper Caenisites turneri and Asteroceras obtusum ammonite biozones, followed by a partial δ 13 C recovery towards pre-S-CIE levels, before the main part of the S-CIE at 179 m (Fig. 2). It is possible that the initial onset of the S-CIE at 190 m represents an early phase of warming that was temporarily reversed, either by a reduction in carbon injection and/or by negative feedbacks. It is noteworthy that the range base of L. variabile and the onset of the Classopollis classoides acme are coincident with the largest (second) negative pulse at~180 m (Fig. 2), which we therefore infer to represent the largest environmental shift. In the Copper Hill Borehole, the onset of this main phase of the S-CIE coincides with the Glebe Farm Bed, an erosional horizon containing a concentration of bored carbonate nodules ( Fig. 2; Brandon et al., 1990) that also represents the base of the O. oxynotum ammonite Biozone. Thus the record reported here is interrupted by a hiatus and is unlikely to represent the full extent of this palaeoenvironmental perturbation.
The aforementioned Toarcian, Aptian and Paleogene CIE events have been hypothesised to be due to injections of isotopically-light carbon into the ocean-atmosphere system (e.g. Dickens et al., 1997;Hesselbo et al., 2000), and this is likely also to be the case for the S-CIE. Another possible cause of a negative carbon excursion could be a contracting biosphere due to cooling, but this would be evident as major changes in biomes and very extensive glaciation for which there is no evidence at this particular time (Sellwood and Valdes, 2008;Korte and Hesselbo, 2011). The coincident and transient nature of the range and acme of L. variabile, the acme of Classopollis classoides and the S-CIE is highly suggestive of a global warming event associated with carbon injection event.
There is as yet no known major magmatic/volcanic event, or previously reported indication of significant global warmth, during the Late Sinemurian. Furthermore, there are no known planktonic and/ or vegetational changes in the Late Sinemurian that could have significantly affected the bulk carbon isotope signal. However, there is some evidence for a sea-level rise of at least regional scale over this interval (Haq et al., 1988), which is consistent with a period of global warmth. The upper part of the A. obtusum and the lower part of the O. oxynotum ammonite biozones represent an interval of shale  Cohen et al., 2007). B-carbon isotope records from the PETM of the Arctic Ocean (black), and the Southern Ocean (grey) (summarised in Cohen et al., 2007). C-carbon isotope records from the S-CIE of Lincolnshire, United Kingdom (this study). The light grey shading indicates the extent of the three CIEs.

24
J.B. Riding et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 374 (2013) 16-27 deposition associated with relative sea-level rise in the Cleveland Basin, northern England Van Buchem and Knox, 1998). Similarly, the lithology of Late Sinemurian age strata on Skye, in the Hebrides Basin, northwest Scotland, closely approaches black (i.e. organic-rich) shale in the O. oxynotum ammonite Biozone  and Brittain et al. (2010) reported L. variabile from coeval strata on Raasay in the Inner Hebrides. As well as indicating relative sea-level rise in this example, organic-rich shale is typical of deposition in stratified marine waters during periods of global warming and high nutrient levels such as the T-OAE, OAE 1a and PETM. The Late Sinemurian is represented by hiatus concretions and an omission surface on the Dorset coast in the Wessex Basin, southern England (Coe and Hesselbo, 2000), interpreted by Haq et al. (1988) and Hesselbo and Palmer (1992) to have resulted from bioerosion during a period of regional sediment starvation caused by sea-level rise (contra Hallam, 1999). The Glebe Farm Bed in the Copper Hill Borehole (Fig. 2), overlain by mudstone, grading upwards to sandstone, probably represents the same condensation phenomenon brought about by sea-level rise and deepening (upper A. obtusum and lower O. oxynotum ammonite biozones), followed by progradation and shallowing (upper O. oxynotum ammonite Biozone) (Hesselbo, 2008).

Conclusions
This study demonstrates that a significant abundance of the thermophilic pollen Classopollis classoides and the range and acme of the thermophilic/eutrophic dinoflagellate cyst L. variabile are linked with a marked coincident negative CIE (herein termed the S-CIE) in the Late Sinemurian O. oxynotum ammonite Biozone (~194 Ma) of eastern England (Fig. 2). The S-CIE reflects a perturbation to the whole exchangeable carbon reservoir because terrestrially-derived organic material (the palynomorph concentrate) and marine carbonate exhibit the same excursion (Fig. 2). The relatively brief record of L. variabile together with the transient acme of Classopollis classoides and the negative S-CIE is highly suggestive of a major climate-warming-driven biogeochemical perturbation, such as a carbon injection event, at~194 Ma, and indicates that Jurassic climates were relatively susceptible to significant and transient change. The level of the S-CIE can be confidently delineated by the range and acme of L. variabile. The S-CIE shares many characteristics with the global negative CIE event at the PETM, a phase of abnormally high global temperatures coincident with a peak in the apparently thermophilic/ eutrophic dinoflagellate cyst genus Apectodinium. The S-CIE also shares similarities to negative carbon isotope excursions at the T-OAE and OAE 1a, which are other major phases of global warming.
The S-CIE has so far only been recorded from a single succession in eastern England. However, the magnitude and remarkable synchronicity of the geochemical and palynological data from the Copper Hill Borehole indicate that the S-CIE probably represents a significant environmental change event. Eastern England is known to have been within an extensive, open seaway during the Early Jurassic (Coward et al., 2003), and hence may be representative of a large part of the Earth system at this time. It is considered that the marine and terrestrial geochemical and palynological signals documented here are not attributable to local factors, nor are they a result of background noise or oscillations in the carbon cycle. However it is fully acknowledged that more sections need to be studied to test our hypothesis. These should include more complete successions and those from localities outside Europe in order to verify this potentially important, apparently global, palaeoenvironmental change event.