Pergamon

Geochimica et Cosmochimica Acta, Vol. 65, No. 6, pp. 885900, 2001 Copyright 2001 Elsevier Science Ltd Printed in the USA. All rights reserved 0016-7037/01 $20.00 .00

PII S0016-7037(00)00582-2

The chemical structure of Gloeocapsomorpha prisca microfossils: Implications for their origin
PETER BLOKKER,1,* PIM
1

VAN

BERGEN,2 RICH PANCOST,1, MARGARET E. COLLINSON,3 JAN W. 1,2 and JAAP S. SINNINGHE DAMSTE

DE

LEEUW,1,2

Department of Marine Biogeochemistry & Toxicology, Netherlands Institute for Sea Research, PO Box 59, 1790 AB Den Burg, Texel, The Netherlands 2 Department of Earth Sciences, Utrecht University, PO Box 80021, 3508 TA, Utrecht, The Netherlands 3 Department of Geology, Royal Holloway University of London, Egham, Surrey, TW20 0EX, UK (Received May 1, 2000; accepted in revised form July 23, 2000)

AbstractTwo Estonian Kukersites (Ordovician) and two samples from the Guttenberg Member (Ordovician) of the Decorah formation (North America) containing botryoidal aggregates of Gloeocapsomorpha prisca were investigated by RuO4 chemical degradation, FTIR, and ash pyrolysis-GC/MS to obtain information about the polymeric structure of these microfossils. The products formed upon oxidation by RuO4 were analysed by GC/MS and revealed the presence of a wide range of carboxyl and/or carbonyl moiety containing compounds with carbon skeletons ranging from C5 to C20. The Estonian Kukersites reveal the presence of a characteristic set of mono-, di-, and tricarboxylic acids. These compounds suggest that the Estonian Kukersites are composed of a polymer consisting of mainly C21 and C23 n -alkenyl resorcinol building blocks. Similarly, although the tricarboxylic acids are not present, the RuO4 degradation product mixtures of the Guttenberg Member samples, suggest a poly( n -alkyl resorcinol) structure. The higher thermal maturity is most likely responsible for the different chemistry and morphology of the G. prisca microfossils in these samples. Because compounds like n -alkenyl resorcinols are known to polymerise under oxygenated conditions even in an aqueous environment, it is not per se necessary that these microfossils are composed of a selectively preserved biopolymeric cell wall. It is also possible that G. prisca microfossils are composed of a cell wall or sheath component that polymerised during senescence or diagenesis of the organism. Copyright 2001 Elsevier Science Ltd
1. INTRODUCTION

The origin and chemical composition of Gloeocapsomorpha prisca has long been a topic of debate in geochemistry and palynology and many researchers have focussed on this subject since the recognition of these microfossils in Ordovician sediments (Zalessky, 1917). Zalessky (1917) made the rst link between the morphology of these microfossils and the extant cyanobacteria Gloeocapsa and thus named this microfossil G. prisca. Despite numerous investigations the biological afnity of G. prisca is still uncertain (Fowler, 1992 for review). For example, Reed et al. (1986) suggested a non-photosynthetic, prokaryotic, benthonic, mat-forming organism while Hoffmann et al. (1987) concluded that the organism was phototrophic planktonic, and possibly eukaryotic. Foster et al. (1989, 1990) proposed based on investigations of G. prisca in Estonian samples, that it must have been a mat-forming cyanobacteria similar to Entophysalis major from an intertidal environment. They described three morphotypes designated as Type I, II, and III. Stasiuk and Osadetz (1990) and Stasiuk (1991) suggested that the microfossils probably originate from a photosynthetic cyanophyte with a complex three-stage life cycle. Derenne et al. (1990, 1992) suggested that the G. prisca microfossils are the selectively preserved algaenan cell walls of a green microalga like Botryococcus braunii.
* Author to whom correspondence should be addressed (blokker@chem.vu.nl). Present address: Organic Geochemistry Unit, University of Bristol, Cantocks Close, Bristol BS8 1TS, UK. 885

To date it is generally accepted that ancient microfossils are often composed of selectively preserved nonhydrolysable and insoluble biopolymers. In recent years, it has become evident that resistant biopolymeric materials are fairly common in nature, and especially green micro-algae from the order of the Chlorococcales comprise several members biosynthesizing such a diagenetically resistant material (algaenan) (Blokker et al., 1998; Largeau and de Leeuw, 1995). To date, the monomeric components of a number of resistant biopolymers in green microalgae have been elucidated. For example, B. braunii algaenan is composed of mainly C32 ,-dialdehydes (Berthe as et al., 1999; Gelin et al., 1994; Metzger et al., 1991), algaenans from Pediastrum, Scenedesmus, Sorastrum, Coelastrum, and Tetraedron are composed of even, unsaturated C30C34 -hydroxy fatty acids (Blokker et al., 1998; Blokker et al., 2000), and Nannochloropsis algaenan is composed of C28-C36 diols and C30 and C32 alkenols (Gelin et al., 1997). Because it was shown that the isolation procedure for resistant materials can give rise to the formation of artefacts (Allard et al., 1997), the discovery of resistant biopolymers in cyanobacteria (Chalansonnet et al., 1988) has become questionable. Consequently, the cell walls of green microalgae seem to be one of the most important sources of selectively preserved biopolymers especially in the lacustrine environment. Due to the presence of a resistant algaenan in the green microalga B. braunii and some morphological similarities between this alga and G. prisca, B. braunii, or a related alga was suggested as a possible source organism for the G. prisca microfossils (e.g., Traverse, 1955; Glikson et al., 1989;

886

P. Blokker et al.

Derenne et al., 1992). However, Burns et al. (1982) suggested that such resemblance is only supercial and probably erroneous. Derenne et al. (1992) observed that upon increasing salinity of the B. braunii culture medium, a signicant phenolic component was biosynthesized. Because the presence of phenolic compounds like n -alkyl-phenols and n -alkyl-resorcinols in the pyrolysates of G. prisca is one of the most striking chemical features of these microfossils (Derenne et al., 1990; 1992) these ndings are in favour of the B. braunii origin theory. Although B. braunii was shown to grow under more saline conditions, this microalgae is principally a freshwater organism. In contrast, the deposition of G. prisca occurred in a marine environment as indicated by intercalated limestones containing the remains of typical marine fauna, indicating a marine origin for this organism (Foster et al., 1989; Bekker, 1921). The bituminous fraction of G. prisca-rich rocks typically reveals an exceptional distribution of odd-over-even n -alkanes ranging between C9 and C19 in carbon number and n -alkylcyclohexanes exposing a similar distribution (Fowler, 1992 and references cited therein). To date, it is still unknown if this fraction in some way reects the chemical structure of the kerogen. In general details about the polymeric structure of G. prisca microfossils are still scarce. Recently, improved techniques have rened the knowledge of polymeric materials in organisms and kerogen (e.g. Eglinton, 1994; Largeau and de Leeuw, 1995; Blokker et al., 2000), which may assist in obtaining new information about G. prisca microfossils. In this study RuO4 degradation is used as a tool to obtain additional chemical information about the polymeric substance making-up the G. prisca microfossils. This technique allows a detailed look at the complex chemistry of these microfossils and sheds new light on their biological afnity.
2. GEOLOGICAL SETTING

40%. The organic-rich layers are associated with the presence of G. prisca microfossils (Jacobson et al., 1988). In the Guttenberg member, the organic matter is represented by stromatolitic as well as disseminated A maceral based on the classication of Stasiuk et al. (1993). No closed (phenol-rich) morphotype (Derenne et al., 1992) and disseminated B maceral (Stasiuk et al., 1993) have been reported in these samples (Pancost et al., 1998).
3. EXPERIMENTAL 3.1. Determination of Hopanoid Epimerisation The G. prisca-rich rocks were ground with a mortar and subsequently extracted with DCM/hexane (4, 1 : 1) and hexane (2). After evaporation of the solvent, the apolar hydrocarbon fractions were obtained from the extracts by column chromatography on activated AlO2 using hexane as an eluent (elution volume 4 column volume). These fractions were analyzed by GC/MS. The 22S/(22S 22R) C31 and C32 17,21(H)-homohopane ratios were determined from the integrated peak areas using the total ion cur (TIC) trace. 3.2. Rock-Eval Rock Eval pyrolysis was performed following standard procedures (Peters, 1986). 3.3. G. prisca Microfossil Preparation After grinding the sediments the samples were ultrasonically suspended in dichloromethane (DCM) and centrifuged for 1 min at 3000 rpm. The DCM oat consisted of pure G. prisca microfossils as shown by light microscopy and was used for further analysis. To simulate the effect of increased temperature on G. prisca, the Kukersite I sample was heated under an N2 atmosphere at 100C for 10 h. 3.4. Microscopy Light microscopy examination was performed on an Olympus BH microscope (New Hyde Park, NY, USA). Specimens for SEM were mounted onto a negative lm from a water droplet following the methods described by Moore et al. (1991) for modern pollen. Specimens for TEM were incorporated into an agar pellet by centrifugation and the pellet was embedded in Spurr resin. Sections 60-nm thick were cut by using a ReichertJung ultracut microtome, no xatives and no staining were used in order to examine the material as close to the chemically analyzed conditions as possible. 3.5. Fourier Transform/Infrared Spectroscopy Fourier transform infrared spectroscopy was performed on a Bruker IFS28 scanning (Billerica, MA, USA) over a frequency range of 400/cm to 4000/cm using KBr pellets containing 2 mg of dry sample. 3.6. Ruthenium Tetroxide Treatment This treatment was modied from a procedure reported earlier (Blokker et al., 1998). Briey: 5 mg of microfossils was ultrasonically suspended in a mixture of 1 mL of chloroform, 1 mL of acetonitrile, and 2 mL of an aqueous solution of NaIO4 (0.2 M, pH 3 4). After addition of 6 mg Ru(III)Cl3, the two phase system was allowed to react in an ultrasonic bath. After 4 h, 3 mL of water, and 2 mL of hexane were added. The organic layer was removed and quenched with methanol (MeOH) (0.5 mL). The aqueous layer was extracted with 2 mL of hexane (1) and 2 mL of dichloromethane (2) and these organic layers were transferred to the rst organic extract. The black Ru-salts were precipitated from the organic layer by centrifugation. The remaining supernatant was washed with 0.5 mL of Na2S2O3 solution (5% in H2O). The extract was dried over Na2SO4, evaporated to dryness under a nitrogen ow and derivatized with diazomethane prior to analysis.

Both Estonian samples are from the Kohtla quarry, northeast Estonia. The section in this quarry represents the lowermost part (the Kivioli Member, Viivikonna Fm.) of the Kukruse Stage. This age assessment is based on the palynology (undertaken by Dr. Zwier Smeenk) and other known markers. This section is part of Graptolite zone N. gracilus (Middle LandeiloEarliest Caradoc Epoch; Middle Ordovician). Kukersite II comes from a real kukersite layer (bed D in this quarry) denoted as an argillaceous kukersite with sparse and few skeletal debris (TOC 2550%). This sample was collected 20 cm above Kukersite I, which is from a kukersineous limestone [Total Organic Carbon (TOC) 525%; bed C/D in this quarry] characterized as a kukersineous ne-grained argillaceous skeletal wackestone. The North American samples are taken from the Guttenberg member of the Decorah Formation obtained from the Cominco SS-9 core (205.4 m and 206.6 m), Millbrook Farms, (Jackson Co., IA, USA). The Guttenberg Member is composed predominantly of carbonate mudstones with occasional skeletal wackestone and packstone lenses interpreted as tempesite beds (Ludvigson et al., 1996). In general CaCO3 content ranges from ca. 40% to 80% and thin (ca. 1 m) brown-coloured G. prisca-rich partings are common and associated with a high TOC content. The TOC values in the Guttenberg member range from 0.5% to

The chemical structure of G. prisca microfossils 3.7. GC/MS, GC/FID and Curie-point py-GC/MS GC/MS, GC/FID, and py-GC/MS equipment and conditions were as previously described (Blokker et al., 1998). For the analysis of the apolar fractions, the GC column temperature was programmed from 70C to 130C at 20C/min followed by 4C/min to 320C with a nal isothermal period of 10 min. For the analysis of the RuO4 product mixture the following temperature program was used: 1 min at 40C and subsequently heated to 320C at 4C/min. The nal temperature was held for 10 min. Of all the compound series identied here, with the exception of compounds 8 and 12, a reference mass spectrum was available in the NIST database. The relative retention time and mass spectrometric characteristics were used to identify other members of these series. Compound series 7, 8, 9, 11, and 12 were quantied by integration of the peak areas in the mass-chromatograms of m / z 58 and 88. The obtained peak areas were multiplied by a response factor compensating for differences in mass spectral responses. This response factor was determined from the relative intensity of the ion used for mass chromatography in the mass spectrum of a well-resolved peak of the homologue series. 4. RESULTS

887

2851/cm (strong), 1710/cm (medium), 1610/cm (medium, broad), 1457/cm (medium, broad), 1375/cm (medium, broad), 1283/cm (medium, broad), 1200/cm (medium, broad), 1094/cm (medium, broad), 974/cm (medium, broad shoulder), 837/cm (small), 720/cm (small). The samples from the Guttenberg Member contain an additional strong absorption at 1025/ cm. 4.4. Flash Pyrolysis The TIC traces of Estonian Kukersite I and Guttenberg II pyrolysates are shown in Figure 2. The Kukersite pyrolysates are dominated by a large unresolved complex mixture (UCM) (Fig. 2a). Important contributors to the UCM are phenolic compounds (van Bergen, unpublished results) such as n -alkyl and n -alkenyl resorcinols (5-n -alkyl-benzene-1,3-diol). n -Alkyl and n -alkenyl resorcinols can be revealed by the m / z 124 mass chromatogram (Figs. 2b) but depending on their concentration they also appear as resolved peaks. For Kukersite I this mass chromatogram reveals some mono-, diunsaturated, and saturated C21 and C23 n -alkyl resorcinols, which were not found in any of the other samples but were reported before in Kukersite pyrolysates by Derenne et al. (1990, 1992). All samples also display a series of n -alkenes and n -alkanes ranging from C7 to C20. These compounds are more dominant in the Guttenberg pyrolysates (Fig. 2c). In the TIC traces of the Guttenberg pyrolysates the UCM is much less pronounced, although the mass chromatogram of m / z 124 (Fig. 2d) still reveals the contribution of n -alkyl resorcinols, 4.5. RuO4 Degradation Degradation of polymeric substances with RuO4 is based upon oxidative cleavage of cross-links such as ether-bonds, which are known to give aliphatic biopolymers like algaenans their structural integrity (Blokker et al., 1998, 1999; Schouten et al., 1998; Gelin et al., 1997). Upon RuO4 degradation, aromatic moieties like n -alkyl benzenes are oxidised resulting in the generation of fatty acids containing one more carbon atom than their aliphatic chains (e.g. Trilieff et al., 1992). G. prisca microfossils are degraded with RuO4 within 1 h and no nal residue was left after the complete reaction time of 4 h. Figure 3 shows the TIC traces of the RuO4 product mixtures of the four G. prisca microfossil samples. Close examination of these traces shows the presence of a number of homologous series of sometimes coeluting compounds. Mass chromatograms of m / z 74, 88, 58, characteristic for compounds containing a carboxyl, a 2-carboxyl and a carbonyl function, respectively, reveal the presence of these compounds in the Estonian Kukersite I (Fig. 4). The most complex product mixtures are obtained from the Estonian Kukersites. In general the compounds in these product mixtures can be subdivided into three major groups on the basis of the number of carboxyl moieties (Fig. 5). Group I contains the compounds with one carboxyl moiety, Group II has two carboxyl moieties, and Group III has three carboxyl moieties. The relative abundances of these components in the RuO4 degradation product mixtures of the four samples are plotted as bar diagrams in Figures 6 and 7. The compounds displayed in Figures 6a and 7a show a strong odd-over-even carbon number predominance, especially in the

4.1. Thermal Maturity The T max for both Kukersite samples is 430C, while the Guttenberg samples reveal a T max of 444C for sample I and 442C for Sample II (Jacobson et al., 1988), indicating that the Kukersites are thermally more immature than the Guttenberg samples. The thermal maturity of the samples as determined from the biomarkers, also indicate that the organic matter of the Kukersites was exposed to less thermal stress than the Guttenberg samples. This is expressed by the presence of a C31 17,21(H)-homohopane in the Kukersites and 22S/(22S 22R) C31 and C32 17,21(H)-homohopane ratios of 0.3. The Guttenberg samples lack the presence of a C31 17,21(H)homohopane and display homohopane ratios of 0.6, which suggests a thermal maturity equal to or greater than the onset of oil generation (Seifert and Moldowan, 1986). 4.2. Microscopy Light microscopic examination of the microfossils that were isolated from the sediment by preparing a DCM oat showed that this material consisted of virtually 100% G. prisca remains in all samples. The G. prisca microfossils in the Estonian Kukersites are yellowish to brown. Although the size of the obtained microfossils is quite variable, they generally range from 10 to 80 m. SEM (Fig. 1a) shows that the particles consist of budding colonies with a smooth surface. TEM (Fig. 1b) shows that the colonies are comprised of concentrical layers of organic matter around what appear to be cell voids. The Guttenberg Member G. prisca microfossils are comprised of some larger brown aggregates and are more homogenous in colour. SEM (Fig. 1c) illustrates that only some remnant morphology has been retained. TEM (Fig. 1d) shows that the spherical bodies composing the colonies are amorphous solid masses with only minor changes in image density, which are possibly the remains of the layers observed in the Kukersites. 4.3. FTIR The FTIR spectra of all four samples are virtually identical and all display absorption bands at: 3650/cm (small, broad, shoulder), 3340/cm (strong, very broad), 2927/cm (strong),

888

P. Blokker et al.

Fig. 1. SEM and TEM of G. prisca specimens from the Estonian Sample I (a,b) and Guttenberg Sample II (c,d). Scale bars 10 m.

The chemical structure of G. prisca microfossils

889

Fig. 2. The py-GC/MS chromatograms of the Estonian Kukersite I and Guttenberg Member II. Mass-chromatograms ( m / z , 124) (b) and (d) reveal the n -alkyl resorcinols. Numbers denote carbon number. UCM unresolved complex mixture.

higher molecular weight region. All compounds in Figures 6b and 7b except the dicarboxylic acids (6) display an even-overodd carbon number predominance, which is particularly evident for the higher carbon numbers. 4.5.1. Estonian Kukersites Of the series of compounds displayed in Figure 5 only 1 and 5 are not present in the RuO4 product mixtures of the Estonian

Kukersites. From group I only the fatty acids (2) display a broad range of carbon numbers, ranging from C7 to C18 with an even-over-odd carbon number predominance in the C15C18 area. Although these normal fatty acids are not abundant in the product mixtures, their counterparts containing a carbonyl group are present in higher relative concentrations. Both 1oxo- and n-oxo-fatty acids (3 and 4, respectively) reveal a strong even-over-odd predominance in the C15C18 area. Al-

890

P. Blokker et al.

Fig. 3. The GC/MS chromatograms of the RuO4 product mixtures of the G. prisca microfossils.

though the carbonyl moiety of series 4 is present at multiple positions as illustrated by the broad peaks in the m / z 58 mass chromatogram (Fig. 4), the oxo-fatty acids are dominated by the 1 and 9-oxo counterparts. Compounds from Group II (Fig. 5) can all be found in the RuO4 degradation product mixtures of the Kukersites. The dicarboxylic acids (6) represents the most dominant series in these product mixtures and are found in the range of C5 to C14 with a maximum at C9. In contrast to the other compound

series, the dicarboxylic acids do not show a specic odd or even carbon number predominance and display a preference for the lower carbon numbers. The 10-carboxy 1-oxo-fatty acids (7) and 11-carboxy 9-oxo-fatty acid (8), both contain an additional carboxyl moiety in comparison to compounds 3 and 4 (Group I). Series 7 displays a restricted carbon range C17 C19, dominated by the C19 member. Series 8 is only represented by the C19 member. The position of the mid-chain carboxyl moiety was determined to be at 10 for all members of series

The chemical structure of G. prisca microfossils

891

Fig. 4. Total ion current and mass chromatograms of m / z 74, 88, and 58, showing fatty acids, 2-methyl carboxyl moieties, and carbonyl containing compounds in the RuO4 product mixture of the Estonian Sample I. Numbers denote longest linear chain length including the carbon of eventual terminal carboxyl moieties.

7, while 8 contains an 11 carboxyl moiety, Since no reference mass spectrum was available for compound 8, this identication remains tentative and is only based on its mass spectrum (Fig. 8a) and relative retention time. The last series within the dicarboxylic acid group are the 2-methyl dicarboxylic acids (9), which range from C5 to C19 in Kukersite II and from C5 to C14 in Kukersite I. Like the former two compound scries the carbon

number distribution of the 2-methyl dicarboxylic acids in Kukersite II maximizes at C17 and C19. The last group in the Estonian Kukersite RuO4 degradation product mixtures contains compounds with three carboxyl moieties (Group III in Fig. 5). Series 10, like the previous three groups, exhibit a strong odd-over-even carbon number predominance and in both Kukersites product mixtures is dominated

892

P. Blokker et al.

Fig. 5. Compounds generated upon RuO4 degradation of the G. prisca microfossils. The compounds are grouped on the basis of the number of carboxyl moieties. x and y are variable and the total chain length of the molecules are mentioned in the text and in Figs. 6 and 7.

by the C17 member. The mid-chain carboxyl moiety of this series is predominantly at the 10 position. The same holds for the series of 2-methyl tricarboxylic acids (11), containing the mid-chain carboxyl at the 10 position. In contrast to all other compound series, series 11 shows a carbon number preference

of C18 and C20 in both Kukersites product mixtures, The last compound eluting in the chromatograms of the Kukersite product mixtures is the 8-carboxy-9-oxo-nonadecanedioic acid (12). Since no reference mass spectrum is available for this compound, this structure is only identied on the basis of its relative

Fig. 6. Bar diagrams representing the relative concentration of compounds in the RuO4 product mixtures of the Estonian Kukersites. Diagram (a) contains the compounds showing an odd-over-even predominance and (b) contains the compounds showing even-over-odd predominance. The abundance of the most abundant compound is set at 100%. The height of the bars representing the dicarboxylic acids has been multiplied by a factor of 0.5. The inset shows the relative amount of the compound groups from Fig. 5 in the product mixtures.

The chemical structure of G. prisca microfossils

893

Fig. 7. Bar diagrams representing the relative concentration of compounds in the RuO4 product mixtures of the Guttenberg Members. Diagram (a) contains the compounds showing an odd-over-even predominance and (b) contains the compounds showing even-over-odd predominance. The abundance of the most abundant compound is set at 100%. The height of the bars representing the 2-methyl fatty acids has been multiplied by a factor of 10. The inset shows the relative amount of the compound groups from Fig. 5 in the product mixtures.

retention time and interpretation of the mass spectrum (Fig, 8b). The shorter C17 and C16 homologues of this compound are present in the Kukersite product mixtures, but only in very low concentrations. 4.5.2. Guttenberg Members Compared with Kukersites, the RuO4 product mixtures of the isolated G. prisca microfossils from the Guttenberg Members do not show such a strong dominance of certain compounds nor a strong preference for odd or even carbon numbers. Most of the compounds observed in the product mixtures display a broad distribution resulting in a series of coeluting compounds as displayed in Figure 3. Most striking is the fact that no compounds could be found in the product mixture containing a mid-chain carboxyl moiety as observed in compound series 7, 8, and 10 to 12 (Group III) in the Kukersite product mixtures. However, in addition to the series present in the Kukersites, a series of methyl ketones (1) and 2-methyl fatty acids (5) are observed. Since the concentration of methyl ketones is low and because they coelute with some other compounds, they can only be revealed by the m / z 88 mass chromatogram. They are found in the product mixtures ranging from C12 to C19 exhibiting a odd-over-even predominance. From group I (Fig. 5) the fatty acids are dominant in the

product mixtures, especially in the Gutterberg I sample. In both Guttenberg samples they are observed in the carbon number range of C7 up to C18 displaying a slight even-over-odd predominance. In contrast to the Kukersites, the 1-oxo- and n-oxo-fatty acids (3, 4) occur in a very broad range of carbon numbers from C9 to C18 with a much less pronounced, but still present, even-over-odd predominance. Consistent with the Kukersite RuO4 product mixtures, the carbonyl moiety of the n-oxo-fatty acids is present at multiple positions, but is most prominently present at 9. 2-Methyl fatty acids (5) are like series 1 unique for the Guttenberg product mixtures. They can be found with carbon numbers ranging from C8 to C20, revealing an even-over-odd predominance above C12. The dicarboxylic acids (6) are as prominent as the fatty acids in the Guttenberg G. prisca RuO4 degradation product mixtures. They range from C6 to C18, maximising at C10C11. As in the Estonian Kukersite product mixtures, the dicarboxylic acids do not display a preference for odd or even carbon numbers. The only other member from compound group II present in the RuO4 degradation product mixtures are the 2-methyl dicarboxylic acids (9). These also display a wide range of carbon numbers, generally from C6 to C19. Especially for the longer chain-length members there is a preference for odd carbon numbers.

894

P. Blokker et al.

Fig. 8. Mass spectra of the compounds tentatively assigned as (a) 2-(2-oxo-decyl)-nonanedioid acid dimethyl ester and (b) 8-methoxycarbonyl-9-oxononanedecanoic acid dimethyl ester.

5. DISCUSSION

5.1. Chemical Structure of G. prisca Microfossils in Estonian Kukersites Studies using ash pyrolysis have revealed that n -alkyl resorcinols (Derenne et al., 1990, 1992), but also alkylated heterocyclic sulphur compounds, alkylbenzenes, and alkylcyclohexanes (Douglas et al., 1991), can be important constituents of G. prisca microfossils. In the pyrolysates of the Estonian Kukersites (Fig. 2a,b) n -alkyl resorcinols show a broad distribution ranging from C7 to C20 maximizing at C14. The broad distribution indicates that these pyrolysis products are probably derived from larger polymer building blocks that are thermally degraded. The Kukersite I pyrolysate contains C21, and C23 n -alkyl resorcinols (Fig. 2), thus containing an n -alkyl side chain comprised of 15 and 17 carbon atoms, respectively. This specic distribution, only displaying the odd homologues, could indicate that these resorcinols represent fragments that did undergo only minor structural changes upon thermal cracking during pyrolysis. In addition, the presence of the mono- and diunsaturated homologues (see also Derenne et al., 1990, 1992) suggests the presence of intermolecular linkages on the n -alkyl side chain in the polymer that were thermally cleaved resulting in the formation of a double bond at the cleavage position. Consequently, if we consider the C21 and C23 n -alkyl resorci-

nols as important building blocks of the polymer, the shorter saturated and unsaturated n -alkyl resorcinols are most likely fragments from a thermal cleavage at a mid-chain position of the alkyl side chain, possibly due to the presence of a linkage to another monomer at this position. Since these shorter homologues are dominated by the C14 saturated and unsaturated-n alkyl resorcinols, this suggests that such a cross-link is positioned around the 8 position on the alkyl side-chain. Upon RuO4 degradation of C21 and C23 n -alkyl resorcinols, the resorcinol moieties of these compounds will be oxidised to a carboxyl moiety generating products that contain the C15 and C17 n -alkyl chain and the additional carbon from the carboxyl moiety, thus C16 and C18 fatty acids, respectively. However, although fatty acids are present in the RuO4 product mixtures, they are only minor components (Fig. 6). The most dominant compounds in the RuO4 product mixture of the Estonian Kukersites are series 7, 8, and 10 (Fig. 5), which all contain a mid-chain carboxyl moiety and have a chain-length preference of C17 and C19 instead of C16 and C18. This shift in distribution by one carbon number seems the consequence of the additional mid-chain carboxyl moiety, which is probably a remnant of a C-C bond between an n -alkyl resorcinol and another moiety. Such a moiety linked at the mid-chain position of the n -alkyl chain of the original n -alkyl resorcinol is most probably a resorcinol unit from another monomer. More specically, the

The chemical structure of G. prisca microfossils

895

Fig. 9. Rationale for the formation of compounds series 7 (a), 10, (b) and 8 (c) upon RuO4 degradation of the Estonian Kukersites.

compounds of series 7 were probably linked to two other n -alkyl resorcinols via a C-C bond and a C-O-C bond generating a carboxyl moiety and a carbonyl moiety, respectively (Fig. 9a). The specic distribution of compound series 10 suggests that these also originate from the oxidative linkage of an n -alkyl resorcinol linked via a C-O-C and a C-C bond (Fig. 9b). In case of compound 8 the C-O-C bond between the resorcinol unit and the alkyl chain was positioned at the 10 position and the C-C bond at the 8 position (Fig. 9c). Although the carboxyl and carbonyl moieties of compound 8 may originate from two different monomers linked to the alkyl chain, it is also possible that a chroman-like structure as in Fig. 9c is responsible for this particular degradation product. On the basis of the former information and considerations, a structure for an n -alkyl resorcinol-based polymer can be proposed (Fig. 10a). The 10-carboxy 1-oxo-fatty acids (7) may be generated upon RuO4 degradation of unit e and the 11carboxy 9-oxo-fatty acid (8) from n -alkyl resorcinol (a). Tricarboxylic acids (10) are most likely formed upon degradation of a monomer (n) linked at a mid-chain position to a carbon of another n -alkyl resorcinol and at the terminal carbon of the n -alkyl via an ether linkage. Not only are these former compounds are explained by this model, the formation of all other compounds in the RuO4 product mixtures can also be explained by this polymer model (Fig. 10a). Fatty acids (2) can be formed from RuO4 degradation of terminal n -alkyl resorcinols (b, m), providing the long-chain fatty acids with a

predominance at C16 and C18. The shorter homologues can, for example, be formed upon cleavage of a vicinal set of etherlinkages ( n -alkyl resorcinol unit (g). 1 (3) and n-oxo fatty acids (4) may originate from cleavage of n -alkyl resorcinols containing a single ether cross-link at the alkyl chain (c, l, o, p), although they also may originate from n -alkyl resorcinols units containing a carbonyl moiety (h). Long-chain dicarboxylic acids (6) can be formed upon degradation of n -alkyl resorcinol (f), whereas the shorter homologues may originate from cleavage of a vicinal set of ether-linkages (g). 2-Methyl dicarboxylic acids (9) are formed upon oxidation of, for example, n -alkyl resorcinol (i) linked via a C-C bond at the 1 position to another resorcinol unit. RuO4 degradation can generate 2-methyl tricarboxylic acids (11) from an n -alkyl resorcinol (k) linked to two resorcinol moieties via C-C linkages at a midchain and 1 position. The 8-carboxy-9-oxononadecanedioic acid (12) is possibly formed from a unit such as d, which is linked to three other monomeric units by two ether and one C-C bond. Considering that the Kukersites were completely degraded by RuO4 degradation, we can regard all the degradation products to be representative of the entire polymer, which suggests that the G. prisca microfossils in the Kukersites are completely composed of n -alkyl resorcinols like in Fig. 10a. FTIR analysis of Estonian Kukersite microfossils also supports the polymer structure presented above. The FTIR data obtained in this study are comparable to those of G. prisca microfossils reported by Derenne et al. (1990, 1994) and Sta-

896

P. Blokker et al.

Fig. 10. Generalised and simplied reconstruction of the proposed n -alkyl resorcinol based polymer composing the G. prisca microfossils from the Kukersites (a) and from the Guttenberg (b). The numbers refer to compounds generated from these polymeric moieties upon RuO4 degradation shown in Fig. 5.

siuk et al., (1993). In all FTIR spectra the oxygen functionalities are expressed as absorptions at 3340/cm (hydroxyl groups), 1710/cm (non-conjugated carbonyl groups), 1200/cm (probably phenolic or alkyl aryl ether moiety), and the area between 1094/cm and 974/cm consisting of a combination of peaks and shoulders (possibly alkyl aryl ethers or dialkyl ethers). The bands at 2927/cm, 2851/cm, 1457/cm and 720/cm are due to the presence of alkyl CH2. The peak broadening at 2927/cm, 2851/cm be ascribed to the presence of CH3 stretching absorptions together with the absorption at 1375/cm. The nal group of absorptions originates from the C-H of benzene moieties. Because these absorptions are generally rather weak and found in areas around 3000/cm, 1650 1400/cm, and between 1200 400/cm they overlap with some other broad signals in these spectra, which makes it difcult to assign absorption bands to these moieties. Although some of these are consequently double assigned, the 974/cm, 837/cm, and 720/cm may reect 1,3,5-substituted benzene together with those at 1610/cm. However the latter can also result from some carboxylate moieties, which could provide and alternative explanation for the 1375/cm absorption.

The 13C-nuclear magnetic resonance data presented by Derenne et al. (1990) are also in agreement with the polymer model in Figure 10. Their spectrum clearly shows the dominant aliphatic signal at 29 ppm and some smaller and broader peaks at chemical shifts 155, 142, 108, 103, 60, and a shoulder at 15 ppm. The rst four of these peaks can be ascribed to an aromatic resorcinol moiety and the latter is the chemical shift of the CH3. The chemical shift at 60 ppm most likely corresponds to the aliphatic carbon of the ether-linkages that form the cross-links between the monomers. 5.2. Relation Between G. prisca Microfossils in Kukersites and the Guttenberg Member The FTIR spectra of the Guttenberg G. prisca microfossils are virtually identical to those of the Kukersites, indicating that these are probably also composed of a poly( n -alkyl resorcinol). Although the Guttenberg samples display an additional absorption at 1025/cm this can most likely be ascribed to some inorganic residues in these samples (see also Derenne et al., 1994). Upon ash pyrolysis of the Guttenberg G. prisca mi-

The chemical structure of G. prisca microfossils

897

crofossils a clear n -alkyl resorcinol series is observed next to the series of n -alkenes and n -alkanes (Fig. 2c and d). Although the abundance of n -alkyl resorcinols generated is lower and some relative differences between the n -alkenyl and n -alkyl resorcinols exist when compared with the Estonian samples, the distribution is very similar to that observed in the Kukersite pyrolysates (Fig. 2a,b). Not only do the FTIR spectra and pyrolysates of the two different sets of samples display similar features; also upon RuO4 degradation identical compound series are observed (Fig. 7). Compound series 2, 6, and 9 in the Guttenberg RuO4 product mixture display virtually the same broad distribution as those in the Kukersite product mixtures. In addition series 3 and 4 are also present, displaying a preference for carbon numbers C16 and C18 similar to their counterparts in the Kukersite product mixtures. Although compound series 2, 6, and 4 are not specic G. prisca RuO4 products and can be found amongst the degradation products of for example algaenans (though having another chain-length distribution) (Blokker et al., 1998, 1999), series 3 and 9 have never been observed in RuO4 degradation products before and might be highly specic degradation products of G. prisca microfossils. The presence of these compounds in the Guttenberg G. prisca RuO4 product mixture, together with the other data obtained for these samples suggests that the Guttenberg microfossils are composed of a similar polymer as found in the Kukersites. However, there are some important chemical differences between the RuO4 product mixtures of the Guttenberg Member samples and Estonian Kukersites. For example, series 3 and 4 display a broader carbon chain-length distribution in the RuO4 product mixtures of the Guttenberg Member samples. The most striking difference is the lack of compounds bearing a mid-chain carboxyl moiety (pie diagrams, Figs. 6 and 7) and the presence of two additional compound series 1 and 5 in the Guttenberg product mixture. Considering that the polymer as presented in Figure 10a forms the basis of also the Guttenberg microfossils, these differences must be the consequence of cleaved C-C bonds at a mid-chain position. Such cleavages could be the consequence of the higher thermal maturity of the Guttenberg samples as implied by the hopanoid epimerization ratios, which indicate that these samples have reached the early stage of oil generation. Simulation experiments with the more immature Kukersite I sample revealed that upon heating of this sample for 10 h at 100C under an inert atmosphere, the pyrolysate changed, resulting in the loss of C21 and C23 n -alkyl resorcinols and a decreased abundance of the UCM. These changes in pyrolysate composition illustrate that upon minimal thermal stress indeed chemical changes occur that could ultimately result in a material with the same chemical properties as the Guttenberg G. prisca microfossils. Although this is a simplied representation, Figure 10b illustrates that early catagenetic cleavage of the thermally labile C-C bonds to the resorcinol moieties linked at a mid-chain position, will result in the loss of series 7, 8, 10 12 in the RuO4 degradation mixture. Furthermore, the moieties generated by this process will be responsible for the formation of series 1 and 5 upon RuO4degradation, although the long-chain homologues of these series are more difcult to explain by this process. Another effect of such C-C bond cleavages is that all compounds generated upon RuO4degradation will show a more broad carbon chain length distribution for compound series 3 and 4.

Because thermal maturity apparently has had an important impact on the chemistry of the Guttenberg Member samples, this process also should have affected the morphology of G. prisca microfossils in these samples. Although the Guttenberg Member G. prisca morphotypes supercially resemble those of the Estonian Kukersite G. prisca microfossils (Fig. 1a and c), some morphological differences are indeed observed. Although some of the differences could be the effect of variations in the salinity levels in the paleoenvironment or other differences in growth conditions (Derenne et al., 1992), the complete lack of discernible cell voids in the Guttenberg G. prisca microfossils, as illustrated by TEM analysis (Fig. 1d), strongly points towards a post-depositional thermal process. This effect has been simulated under laboratory conditions by Stasiuk and Osadetz (1993), who showed that G. prisca microfossils display a sort of melting phenomenon upon heating up to 290C. Although the G. prisca microfossils in the Kukersites and Guttenberg Members are both composed of a poly( n -alkyl resorcinol), the results presented here indicate that both the chemical and some of the morphological differences can be explained by post-depositional thermal processes. 5.3. Biological Occurrence of Resorcinols n -Alkyl resorcinols (5-n -alkyl-benzene-1,3-diol) are quite common in nature and especially the C21 and C23 members can be abundant in some higher plants such as the Ginkgoaceae and Anacardiaceae (Kozubek and Tyman, 1999, for a review). Recently, the occurrence of resorcinol lipids has been demonstrated in an increasing number of tissues and organisms including fruits, seeds, bacterial cysts, bacterial vegetative cells and senescent organs or cells (Kozubek and Tyman, 1999). A series of unique resorcinolic lipids were found in the green microalga B. braunii involving C31, C33, and C35 n -alkenyl resorcinols linked via a resorcinolic C-O linkage to long-chain n -alkenes, n -alkadienes and n -alkatrienes forming high molecular weight molecules (Metzger and Largeau, 1994). However, in most cases the reported alkyl resorcinols occur as C21 and C23 homologues, generally having different degrees of unsaturation (Kozubek and Tyman, 1999 and references therein). The function of alkyl resorcinols in nature is only partially known. However, it is clear that this family of compounds plays an important role in organism-organism interactions (Kozubek and Tyman, 1999 and references therein), Furthermore, it was shown that longchain 5-n -alk(en)ylresorcinols can prevent membranes from oxidation processes and even polymerise under oxic conditions (Tymann, 1991; Kozubek and Tyman, 1999). Some plants like the oriental lacquer trees (e.g. Toxicodendron verniciuum, Rhus vernicifera) contain types of compounds very similar to n -alkyl resorcinols namely n -alkyl catechols (3-n -alkyl-benzene-1,2-diol) with carbon numbers of predominantly C21 and C23. These n -alkyl catechols are the main constituents of the sap that has been used as natural varnish and adhesive for more than 5000 yr (Du et al., 1984; Kumanotani, 1998). It was shown that two main mechanisms are involved in the polymerisation of these compounds: a coupling of the catecholic units producing diphenyls and dibenzofurans and a reaction of the catechol unit with the unsaturation on the alkyl chain. The observed dimeric structures as

898

P. Blokker et al.

Fig. 11. Proposed catalytic radical polymerisation mechanism of alkyl resorcinols and formed dimeric structures and their RuO4 oxidation products.

proposed by Kumanotani (1995, 1998) clearly illustrate that the type of linkages found in the G. prisca microfossils (Figs. 9 and 10) are easily formed upon oxygen-induced polymerization of such compounds. Although to date we can only guess what kind of polymerisation mechanism was exactly involved in case of G. prisca, it seems possible that some kind of radical mechanism initiated the polymerisation process in parallel to the peroxidase induced radical polymerisation of the phenolic units in lignins (Freudenberg and Neish, 1968) and aerobic oxidative or laccase-catalysed polymerisation of urushiol (Kumanotani, 1995). The polymerisation of both lignin and urushiol starts with the abstraction of an H radical from a phenolic hydroxyl group, upon which the activated species can react with other moieties. In the case of G. prisca probably the resorcinol moiety of the n -alkyl resorcinols did undergo such H radical abstraction and reacted with the other monomers, which probably contained one or more unsaturations on the alkyl chain. Consequently most of the intermolecular linkages will involve the resorcinol as illustrated in Figure 11, showing that both carbon-carbon bonds and ether-linkages can be created this way. Although this mechanism is tentative, the structures expected from such a polymerisation are in perfect agreement with the proposed polymer structure (Fig. 10a). 5.4. Implications for the Origin of G. prisca Although the inference that the G. prisca microfossils are composed of a poly( n -alkyl resorcinol) is an important contribution to the knowledge of these microfossils, it still remains unclear if the poly( n -alkyl resorcinol) is a biopolymer or a structure that is formed from free n -alkyl resorcinols after the organism died, The inner morphology of the Kukersite G. prisca microfossils clearly shows the voids of the original cells (Fig. 1b) embedded in concentrically layers of organic matter. Derenne et al. (1992) interpreted these layers as the remnants of the cellular matrix of a B. braunii-like organism, thus composed of the biopolymer algaenan. Because B. braunii also

makes resorcinolic lipids, though having much longer alkylchains of 2529 carbon atoms instead of 1517 as observed in the G. prisca polymer, a relation between these organisms is possible (Metzger and Largeau, 1994). Algaenan-containing algae like Tetraedron minimum, Pediastrum boryanum normally release their daughter cells from the algaenan containing mother cell wall via a ssure at a specic position in the wall (van Hoek et al., 1995) and B. braunii daughter cells grow out of the mother cell cup (Burns, 1982). In this way these algaenan-containing algae avoid being trapped inside their polymeric algaenan cell wall. Although Derenne et al. (1992) clearly illustrated that the morphology of B. braunii becomes more G. prisca-like under increased salinity, the inner morphology observed for the Kukersite microfossils, in which the daughter cells stay within the mother cell, is unlikely for organisms having an algaenan-containing cell wall, since this rigid polymeric structure would obviously restrict cell division and growth. Furthermore, in case of B. braunii the resorcinolic lipids are not the main building-blocks of the algaenan cell walls, which is composed of linear polyaldehyde (Metzger et al., 1993; Gelin et al., 1997, Berthe as et al., 1999). In G. prisca microfossils, a poly( n -alkyl resorcinol) forms the basis of the polymeric structure and no evidence is found for the presence of a linear polyaldehyde or other aliphatic polymer structure from the RuO4 degradation products. Although compounds like unsaturated n -alkyl resorcinols can be polymerised by enzyme catalysis as illustrated by urushiol (Kumanotani, 1995) and most likely also by the resorcinolic lipids in B. braunii, the sensitivity of these compounds towards polymerisation suggests the possibility that G. prisca microfossils consist of a polymer formed after the organism died. It is possibly that G. prisca lived under conditions which required protection of a sheath containing anti-microbial agents, anti-oxidants and/or UV-lters. If G. prisca excreted n -alkyl resorcinols for this purpose the specic chemical properties of these compounds could have resulted in polymerisation during senescence or after death. Polymerisation of the

The chemical structure of G. prisca microfossils

899

n -alkyl resorcinols could have been catalysed by oxygen itself, trace metals and even lytic enzymes, released upon senescence of the cells or by certain bacteria thriving on its remains. Based on the sheath morphology of some cyanobacteria G. prisca was suggested to be closely related to these microorganisms (Foster et al. 1989, 1990; Stasiuk and Osadetz, 1990; Stasiuk et al., 1991). Cyanobacterial sheaths often contain large amounts of secondary protective compounds like the natural UV-lter scytonemin, sometimes in concentrations up to 25 mg/g protein (Dillon and Castenholz, 1999). To date not much is known about the chemical composition of cyanobacterial sheaths, but it is known that this mucus can be readily preserved (Boon and de Leeuw, 1987; de Leeuw and Largeau, 1993) and is even reported to partially polymerise under certain conditions (Humm and Wicks, 1987). Cyanobacterial sheaths probably do not have intrinsic resistant properties like algaenan containing cell walls, but fossilisation obviously does occur due to the specic preservation conditions (de Leeuw and Largeau, 1993). Because the water-containing sap of oriental lacquer trees resembles to a certain degree the sheath of a cyanobacteria (Humm and Wicks, 1987), maybe such sheaths serve a suitable microenvironment for the polymerisation of excreted n -alkyl resorcinols under certain conditions. Although this theory seems to favour the cyanobacterial origin of G. prisca, it remains pure speculative because a poly( n -alkyl resorcinol) has never been observed in such organisms before.
6. CONCLUSIONS

Ludvigson and B. Witzke of the Iowa Geological Survey for supplying samples from the SS-9 core and T. Brain of Kings College London for assistance with the TEM. This is NIOZ contribution 3489. Associate editor: M. C. Kennicutt II REFERENCES Allard B., Templier J., and Largeau C. (1997) Artifactual origin of mycobacterial bacteran. Formation of melanoidin-like artifact macromolecular material during the usual isolation process. Org. Geochem. 26, 691704. Bekker H. (1921) The Kucker stage of the Ordovician rocks of NE Estonia. Acta et Commentationes Univsitalis Dorpatensis A II 1, 191. Berthe as O., Metzger P., and Largeau C. (1999) A high molecular weight complex lipid, aliphatic polyaldehyde tetraterpenediol polyacetal from Botryococcus braunii (L race). Phytochemistry 50, 85 96. Blokker P., Schouten S., Van den Ende H., de Leeuw J. W., Hatcher P. G., and Sinninghe Damste J. S. (1998) Chemical structure of algaenans from the fresh water algae Tetraedron minimum, Senedesmus communis and Pediastrum boryanum. Org. Geochem. 29, 1453 1468. Blokker P., Schouten S., de Leeuw J. W., Sinninghe Damste J. S., van den Ende H. (1999) Molecular structure of the resistant biopolymer in the zygospore cell walls of Chlamydomonas monoica. Planta 207, 539 543. Blokker P., Schouten S., de Leeuw J. W., Sinninghe Damste J. S., van den Ende H. (2000) A comparative study of fossil and extant algaenans using ruthenium tetroxide degradation. Geochim. Cosmochim. Acta 64, 20552065. Boon J. J. and de Leeuw J. W. (1987) Organic geochemical aspects of cyanobacterial mats. In Cyanobacteria: Current research (eds. P. Fay and C. van Baalen), pp. 471 492. Elsevier. Burns D. A. (1982) A transmission electron microscope comparison of modern Botryococcus braunii with some microfossils previously referred to that species. Rev. Esp. Micropaleontol. 14, 165185. Challansonnet S., Largeau C.. Casadevall E., Berkaloff C., Peniguel G., and Couderc R. (1988) Cyanobacterial resistant biopolymers. Geochemical implications of the properties of Schizothrix sp. resistant material. In Advances in Organic Geochemistry 1987 (eds. L. Mattavelli tnd L. Novelli), pp. 10031010. Pergamon Press. Derenne S., Largeau C., Casadevall E., Sinninghe Damste J. S., Tegelaar E. W., and de Leeuw J. W. (1990) Characterization of Estonian Kukersite by spectroscopy and pyrolysis: Evidence for abundant alkyl phenolic moieties in an Ordovician, marine, type II/I kerogen. Org. Geochem. 16, 873 888. Derenne S., Metzger P., Largeau C., van Bergen P. F., Gatellier J. P. Sinninghe Damste J. S., de Leeuw J. W., and Berkaloff C. (1992) Similar morphological and chemical variations of Gloeocapsomorpha prisca in Ordovician sediments and cultured Botryococcus braunii as a response to changes in salinity. Org. Geochem. 19, 299 313. Derenne S., Largeau C., Landais P., and Rochdi A. (1994) Spectroscopic features of Gloeocapsomorpha prisca colonies and of interstitial matrix in kukersite as revealed by transmission micro-FT-i.r.: Location of phenolic moieties. Fuel 73, 626 628. Dillon J. G. and Castenholz R. W. (1999) Scytonemin, a cyanobacterial sheath pigment, protects against UVC radiation: implications for early photosynthetic life. J. Phycol. 35, 673 681. Douglas A. G., Sinninghe Damse J. S., Fowler M. G., Eglinton T. I., and de Leeuw, J. W. (1991) Unique distributions of hydrocarbons and sulphur compounds released by ash pyrolysis from the fossilised alga Gloeocapsomorpha prisca, a major constituent in one of four Ordovician kerogens. Geochim. Cosmochim. Acta 55, 275291. Du Y., Oshima R., and Iwatsuki H. (1984) High-resolution gas-liquid chromatographic analysis of urushiol of the lac tree, Rhus vernicifera, without derivatization. J. Chrom. 295, 179 186. Eglinton T. I. (1994) Carbon isotopic evidence for the origin of macromolecular aliphatic structures in kerogen. Org. Geochem. 21, 721735. Fowler M.G. (1992) The inuence of Gloecapsomorpha prisca on the

The results obtained from FTIR, ash pyrolysis and most importantly RuO4 degradation presented in this study illustrate that the G. prisca microfossils from the Kukersites and the Guttenberg Member are composed of a polymeric structure based on unsaturated C21 and C23 n -alkyl resorcinols. These monomeric units are intermolecularly linked via C-C and C-O bonds between resorcinol moieties and the n -alkyl chains. Both the chemical and some morphological differences between the microfossils from these different settings are assigned to the higher thermal maturity of the G. prisca obtained from the Guttenberg Member, which altered the chemical properties of the poly( n -alkyl resorcinol) structure by cleavage of specic linkages within the polymer. This process disrupted the integrity of the polymer allowing some of the morphological differences between the Guttenberg and the Kukersite microfossils. Because poly(n -alkyl resorcinols) are unknown from nature it is not possible to directly link the chemistry to a certain organism and answer the question if these microfossits represent a biopolymer or a polymer formed during senescence or after the cells died, However, the specic chemical, physical, and biological properties of n -alkyl resorcinols could indicate that they served as a sheath component forming a part of the matrix in which the original cells were embedded. In this case polymerisation must have resulted in a resistant polymeric substance, which could have been selectively preserved in the sediment in such a fashion that the cell morphology was retained in the shape of readily discernible cell voids in the immature G. prisca microfossils in the Kukersites.
AcknowledgmentsThis study was supported by the Research Council for Earth and Life Sciences (ALW) with nancial aid from the Netherlands Organisation for Scientic Research (NWO). We thank G.

900

P. Blokker et al. Metzger P. and Largeau C. (1994) A new type of ether lipid comprising phenolic moieties in Botryococcus braunii. Chemical structure and abundance, and geochemical implications. Org. Geochem. 22, 801 814. Metzger P., Largeau C., and Casadevall E. (1991) Lipids and macromolecular lipids of the hydrocarbon-rich microalgae Botryococcus braunii. Chemical structure and biosynthesis. Geochemical and biotechnological importance. Prog. Org. Nat. Prod. 57, 170. Moore P. D., Webb J. A., and Collinson M. E. (1991) In Pollen Analysis. Blackwell Scientic. Pancost R. D., Freemann K. H., Patzkowsky M. E., Wavrek D. A., and Collister J.W.(1998) Molecular indicators of redox and marine photoautotroph composition in the late Middle Ordocician of Iowa, U.S.A. Org. Geochem. 29, 1649 1662. Peters K. E. (1986) Guidelines for evaluating petroleum source rock using programmed pyrolysis. Am. Assoc. Pet. Geol. Bull. 70, 318 329. Reed J, D., Illich H. A., and Horseld B. (1986) Biochemical evolutionary signicance of Ordovician oils and their sources. In Advances in Organic Geochemistry 1985 (eds. D. Leythaeuser and J. Rullko tter), pp. 347358. Pergamon Press. Schouten S., Moerkerten P., Gelin F., Baas M., de Leeuw J. W., and Sinninghe Damste J. S. (1998) Structural characterization of aliphatic, non-hydrolyzable biopolymers in freshwater algae and leaf cuticle by ruthenium tetroxide degradation. Phytochemistry 49, 987 993. Seifert W. K. and Moldowan J. M. (1986) Use of biological markers in petroleum exploration. In Methods in Geochemistry and Geophysics 24 (ed. R. B. Johns), pp. 261290. Elsevier Scientic. Stasiuk L. D., Kybett B. D., and Bend S. L. (1993) Reected light microscopy and micro-FTIR of Upper Ordovician Gloeocapsomorpha prisca alginite in relation to paleoenvironment and petroleum generation, Saskatchewan, Canada. Org. Geochem. 20, 707719. Stasiuk L. D. (1991) Organic petrology and petroleum formation in Palcozoic rock of nothern Williston basin, Canada. Ph.D. thesis, University of Regina, Saskatcheman, Canada. Stasiuk L. D. and Osadetz K. G. (1990) The life cycle and phyletic afnity of Gloeocapsomorpha prisca Zalessky 1917 from Ordovician rocks in the Canadian Williston Basin. Geol. Surv. Canada Paper 89-ID, 123137. Stasiuk L. D. and Osadetz K. G. (1993) Thermal maturity, alginitebitumen transformation, and hydrcarbon generation in upper Ordovician source rocks, Saskatchewan, Canada. Energ. Sources 15, 205237. Stasiuk L. D., Osadetz K. G., Goodarzi F., and Gentzis T. (1991) Organic microfacies and basinal tectonic control on source rock accumulation: A microscopic approach with examples from intracratonic and extensional basin. Int. J. Coal. Geol. 19, 457 481. Traverse A. (1955) Occurrence of the oil-forming alga Botryococcus braunii in lignites and other tertiary sediments. Micropaleont. 1, 343350. Trilieff S., Sieskind O., and Albrecht P. (1992) Biological markers in petroleum asphalthenes: Possible mode of incorporation. In Biological Markers in Sediments and Petroleum (eds. J. M. Moldowan, P. Albrecht, and R. P. Philip), pp. 350 369. Prentice-Hall. Tyman J. H. P.(1991).The chemistry of non-isoprenoid phenolic lipids. In Studies in Natural Products Chemistry 9 (ed. Atta-ur-Rahman), pp. 313379. Elsevier. Zalessky J. E. (1917) On marine sapropelite of Silurian age formed by a blue-green alga. Izvtiya Academiya Nauk IV 1, 318.

organic geochemistry of oils and organic-rich rocks of Late Ordovician age from Canada. In Early Organic Evolution: Implications for Mineral and Energy Resources (eds. M. Schidlowski et al.), pp. 336 356. Springer. Foster C. B., Wicander R., and Reed J. D. (1989) Gloeocapsomorpha prisca Zalessky, 1917: A new study. Part I: taxonomy, geochemistry and paleoecology. Geobios 2, 735759. Foster C. B., Wicander R., and Reed J. D. (1990) Gloeocapsomorpha prisca Zalessky, 1917: A new study. Part II: origin of Kukersite, a new interpretation. Geobios 23, 133140. Freudenberg K. and Neish A. C. (1968) Constitution and Biosynthesis of Lignin. Springer-Verlag. Gelin F., Boogers I., Noordeloos A. A. M., Sinninghe Damste J. S., Riegman R., and de Leeuw J. W. (1997) Resistant biomacromolecules in marine microalgae of the classes Eustigmatophyceae and Chlorophyceae: Geochemical implications. Org. Geochem. 26, 659 675. Gelin F., de Leeuw J. W., Sinninghe Damste J. S., Derenne S., Largeau C., and Metzger P. (1994) Scope and limitations of ash pyrolysisgas chromatography-mass spectrometry as revealed by the thermal behaviour of high-molecular-weight lipids derived from the green microalga Botryococcus braunii. J. Anal. Appl. Pyrol. 28, 183204. Glikson M., Lindsay K, and Saxby J. (1989) Botryococcus-A planctonic green alga, the source of petroleum through the ages: transmission electron microscopical studies of oil shales and petroleum source, rocks. Org. Geochem. 14, 595 608. van Hoek C., Mann D. G., and Jahns H. M. (1995) Algae: An Introduction to Phycology. Cambridge University Press. Hoffmann C. F., Foster C. B., Powell T. G., and Summons R. E. (1987) Hydrocarbon biomarkers from Ordovician sediments and fossil alga G. prisca Zalessky 1917. Geochim. Cosmochim. Acta 51, 2681 2697. Humm H.J., and Wicks S. R. (1987) Introduction and Guide to the Marine Bluegreen Algae. Wiley-Interscience. Jacobson S. R., Hatch J. R., Teerman S. C., and Askin R. A. (1988) Middle Ordovician organic matter assemblages and their effect on Ordovician-derived oils. Bull. AAPG 72, 1090 1100. Kumanotani J. (1995) Usushi (oriental lacquer)-A natural aesthetic durable and future-promising coating. Prog. Org. Coat. 26, 163195. Kumanotani J. (1998) Enzyme catalyzed durable and authentic oriental lacquer: a natural microgel-printable coating by polysaccharideglycoprotein-phenolic lipid complexes. Prog. Org. Coat. 34, 135 146. Kozubek A., Tyman J. H. P. (1999) Resorcinolic lipids, the natural non-isoprenoic amphiphiles and their biological activity. Chem. Rev. 99, 126. Largeau C. and de Leeuw J. W. (1995) Insoluble and highly aliphatic macromolecular constituents of microbial cell walls. In Advances in Microbial Ecology 14 (ed. J, G. Jones), pp. 77117. Plenum Press. de Leeuw J. W. and Largeau C. (1993) A review of macromolecular organic compounds that comprise living organisms and their role in kerogen, coal and petroleum formation. In Oganic Geochemistry (eds. M. H. Engel and S. A. Macko), pp. 2372. Plenum Press. Ludvigson G. A., Jacobsen S. R., Wizke B. J., and Gonzalez L. A. (1996) Carbonate component chemostratigraphy and depositional history of the Ordovician Decorah Formation, Upper Mississippi Valley. In Paleozoic Sequence Stratigraphy: Views from the North American Craton (eds. B. Witzke, G. Ludvigson, and J. Day), pp. 67 86. Geological Society of America Special Paper 306. Metzger P., Pouet Y., Bischoff R., and Casadevall E. (1993) An aliphatic polyaldehyde from Botryococcus braunii (A race). Phytochemistry 32, 875 883.