When magnetotactic bacteria die, their magnetosomes can be deposited and preserved in sediments, resulting in a (post) depositional remanent magnetization. The SD size of magnetosomes makes them excellent recorders of the paleomagnetic field and their unique hexagonal crystal forms provide a means for identifying magnetosomes via electron microscopy of magnetic extracts from sediments. Fossil magnetosomes, sometimes preserving the chain structure, have been identified in sediments spanning the Phanerozoic [ Vali and Kirschvink, 1990 and references therein].
Although magnetotactic bacteria are ubiquitous in many present day aquatic environments, the eventual fate of magnetosomes and their relative contributions to remanent magnetization and the mineral magnetic record in sediments is not so obvious. Oldfield [1992] summarizes what he calls the ``detrital'' and ``biomagnetic'' interpretations of the source(s) of fine grained magnetite in Quaternary sediments with emphasis on the mineral magnetic signature of paleoenvironmental change. He argues that, whereas TEM observation of magnetosomes provides proof that a biogenic component is present, new magnetic methods are needed to quantify the biogenic contribution in sediments and establish sediment-source linkages.
The most effective magnetic approach for biogenic identification and quantification should allow whole sediment samples to be measured using magnetic methods sensitive to the SD size specificity, and possibly chain assemblage, of BOB-type biogenic systems. Magnetic methods have the advantage of being rapid and non-destructive but may suffer from an ambiguity in distinguishing biogenic SD particles from detrital SD or small multidomain particles. In addition, BIM-type magnetic minerals lack the SD size specificity of magnetosomes and, at least for magnetite produced by GS-15, resemble inorganic magnetite particles produced during soil formation [ Lovley, 1990]. Hence, crystal morphology alone is not a useful criterion for identifying BIM-type magnetite. Furthermore, unless one can show that all SPM particles in sediments or soils are biogenic, then magnetic identification of BIM-type particles is also precluded [ Moskowitz et al., 1993].
Two approaches to the biogenic problem have recently been presented. In the first approach, Oldfield [1994] suggests using a combination of low-field, frequency dependent, and anhysteretic susceptibilities to isolate a magnetosome magnetic response. The problems posed by an in situ BIM-type magnetic fraction or volume reduction of magnetosomes producing SPM particles by dissolution was not addressed. The method is calibrated with synthetic magnetite and natural samples from several environments where evidence suggests either detrital or non-detrital (ie., biogenic?) magnetite. However, no electron microscopy was done to check whether magnetosomes were actually present in the sediments yielding the ``biogenic'' signature. The second approach is based on low temperature behavior (20-300 K) of saturation remanence observed in pure cultures of MTB on warming through the cubic-monoclinic phase transition in magnetite near 100 K [ Moskowitz et al., 1994]. Unlike room-temperature remanence/susceptibility parameters that are sensitive to a specific SD particle volume distribution, the low-temperature results appear to depend on the unique chain arrangement of magnetite magnetosomes in MTB and may be sensitive enough to quantify this fraction in bulk samples. Nevertheless, the method has yet to be ``field-tested'' on natural samples with known biogenic components.
Several studies reported biogenic minerals in lake, marine, and continental eolian deposits [ Snowball, 1994; Hess, 1994; Evans and Heller, 1994; Hawthorne and McKenzie, 1993; McNeill and Kirschvink, 1993; Maher and Thompson, 1992]. Two papers dealt with the formation and destruction of biogenic magnetite in lake sediments combining magnetic, TEM, and geochemical methods [ Hawthorne and McKenzie, 1993; Snowball, 1994]. In a study of Lake Greifen (Switzerland) sediments spanning the past 300 years, Hawthorne and McKenzie [1993] conclude that dissolution and sulfidization of detrital and biogenic magnetite in the upper 30 cm occurred in response to a change in the depositional environment of the lake from aerobic to anoxic due to eutrophication associated with agricultural/industrial development in the area since 1887. In contrast, Snowball [1994] documents a high concentration of biogenic magnetite in the upper sediment levels with progressive dissolution of biogenic magnetite at depth from lake sediments in Sweden. Biogenic magnetite was confirmed based on TEM observations and comparative magnetic studies on catchment and sediment samples.
Hess [1994] reports abundant fossil magnetosomes in oxic to suboxic hemipelagic sediments from the southwest Pacific Ocean. He concludes that biogenic magnetite is the dominant fraction in these sediments and that down-core magnetic variations represent paleoenvironmental changes affecting bacterial paleoecology. Finally, Evans and Heller [1994] suggest that the magnetic enhancement observed in paleosols from the loess plateau of China is due to in situ formation of both BOB-type (SD) and BIM-type (SPM) biogenic magnetites. Although this is an intriguing idea, neither TEM identification of soil magnetosomes nor observation of extant species of MTB or dissimilatory iron-reducing bacteria in modern soils in the area were documented. Instead, biogenic confirmation was based solely on the similarity of magnetic parameters with deep-sea sediments containing fossil magnetosomes. In contrast, Maher and Thompson [1992] found magnetite particles resembling magnetosomes in paleosol based on TEM observations of extracts, but they conclude that this biogenic(?) fraction is a minor magnetic component of the paleosol.