INTRODUCTION Unlike the familiar Phanerozoic history of megascopic life, evolution during earlier and much longer Precambrian time was character- ized by development of metabolic capabilities in microscopic prokaryotes (Schopf, 1994). Be- cause modern prokaryotes are highly diverse physiologically and those having similar ap- pearances can differ markedly in metabolism, the sorting out of such capabilities among Pre- cambrian fossil microorganisms requires more than traditional morphology-based paleontol- ogy. To date, this need has been addressed by isotopic analyses of whole rock acid-resistant carbonaceous residues (kerogens), studies that have traced isotopic signatures of biological carbon fixation to ~3500 Ma (Hayes et al., 1992; Schidlowski et al., 1983) and possibly earlier (Schidlowski, 1988), and evidently of both methanogenic and methanotrophic microbes to ~2700 Ma (Hayes, 1994). Although useful, isotopic analyses of bulk samples cannot provide information specific to the apparent physiology of, or the phylogenetic relations among, individual microscopic fossils. However, because the metabolic pathways that typify different groups of autotrophic prokaryotes can fractionate carbon isotopes by differing amounts—for example, the acetyl-CoA pathway can yield substantially greater fractionation than the Calvin cycle (House, 1999; Preuβ et al., 1989), and the Calvin cycle more than the reduc- tive tricarboxylic acid (TCA) cycle (House, 1999; Preuβ et al., 1989)—isotopic evidence of metabo- lism and phylogenetic affinities might be pre- served in the cell walls and other kerogenous con- stituents of fossil microorganisms. If so, accurate, 1‰–2 ‰ precision measurements of the carbon isotopic composition of individual Precambrian microfossils could provide significant new means to more clearly define the composition and evolu- tionary history of the early biota. The ion microprobe can be used for in situ analysis of the isotopic composition of very small amounts of carbonaceous material (Ireland, 1995; McKeegan et al., 1985). This instrument was used by Mojzsis et al. (1996) for carbon isotopic analy- sis of individual ~5 μm carbonaceous inclusions hosted within apatite crystals from ~3800 Ma banded iron formation (BIF) sediments. That pio- neering study identified isotopically light carbon and inferred fractionation of sufficient magnitude to suggest the presence of biological activity at the time of sedimentation. However, the analytical uncertainties of individual analyses were typically too large (~5‰) to confidently reveal diversity in the isotopic record and, moreover, the BIF studied was severely metamorphosed, eliminating the potential for structural preservation of micro- fossils. Here we report results of analyses of microorganisms cellularly petrified in two Pre- cambrian microbial communities, those of the Bitter Springs (~850 Ma) and Gunflint (~2100 Ma) microbiotas. Techniques used are such that iso- topic measurements have been made with suffi- cient accuracy and precision to reveal potential isotopic variability in the fossil record on the level of a single microscopic organism. METHODS The University of California, Los Angeles, CAMECA ims 1270 ion microprobe was used to analyze the carbon isotopic composition of indi- vidual microfossils exposed at the upper surfaces of polished petrographic thin sections (~100 μm thick) of subgreenschist facies stromatolitic car- bonaceous cherts from the ~850 Ma Bitter Springs Formation of central Australia (Schopf, 1968) and the ~2100 Ma Gunflint Formation of southern Canada (Barghoorn and Tyler, 1965). Optical microscopy was used to select microfossils of diverse morphology and free of detectable impuri- ties such as carbonate or hematite. Secondary C – ions were sputtered by a ~10- to 15-μm-diameter Cs + beam from the microfossils (Figs. 1 and 2) and analyzed at a mass resolving power between 3000 and 4000 in order to eliminate molecular ion (hydride) interferences. Instrumental mass frac- tionation was calibrated by interposing analyses of individual microfossils with analyses of a car- bonaceous chert specimen from the Precambrian Swaziland Supergroup (Precambrian Paleobiol- ogy Research Group [PPRG] sample 215-1), a “standard” that is mineralogically similar to the fossiliferous samples and contains kerogen of Geology; August 2000; v. 28; no. 8; p. 707–710; 3 figures; 3 tables. 707 Carbon isotopic composition of individual Precambrian microfossils Christopher H. House * J. William Schopf Department of Earth and Space Sciences, and IGPP Center for Astrobiology, University of California, Kevin D. McKeegan Los Angeles, California 90095-1567, USA Christopher D. Coath T. Mark Harrison Karl O. Stetter Lehrstuhl für Mikrobiologie und Archaeenzentrum, Universität Regensburg, 93053 Regensburg, Germany * Present address: Department of Geosciences, Penn State University, 208 Deike Building, University Park, PA 16801. ABSTRACT Ion microprobe measurements of carbon isotope ratios were made in 30 specimens repre- senting six fossil genera of microorganisms petrified in stromatolitic chert from the ~850 Ma Bitter Springs Formation, Australia, and the ~2100 Ma Gunflint Formation, Canada. The δ 13 C PDB values from individual microfossils of the Bitter Springs Formation ranged from –21.3 ± 1.7‰ to –31.9 ± 1.2‰, and the δ 13 C PDB values from microfossils of the Gunflint Formation ranged from –32.4 ± 0.7‰ to – 45.4 ± 1.2‰. With the exception of two highly 13 C-depleted Gun- flint microfossils, the results generally yield values consistent with carbon fixation via either the Calvin cycle or the acetyl-CoA pathway. However, the isotopic results are not consistent with the degree of fractionation expected from either the 3-hydroxypropionate cycle or the reductive tri- carboxylic acid cycle, suggesting that the microfossils studied did not use either of these pathways for carbon fixation. The morphologies of the microfossils suggest an affinity to the cyanobacteria, and our carbon isotopic data are consistent with this assignment. Keywords: Precambrian microfossils, Calvin cycle, carbon isotopes, kerogen. Figure 1. Optical photomicrographs (A and C) and 12 C – scanning ion images (B and D) of Cephalophytarion (A and B), a filamentous (oscillatoriacean) cyanobacterium, and Myx- ococcoides (C and D), cyanobacterial (Chro- ococcacean) unicells clustered in a sheath, exposed at upper surface of polished petro- graphic thin section of stromatolitic Bitter Springs chert. Ion images show that carbon analyzed is derived predominately (>90%) from microfossils rather than from particulate kerogen diffused throughout mineral matrix.