Several factors other than climatic changes can produce variations in the magnetic properties of sediments. For example, concentrations of magnetite produced by magnetotactic bacteria in sediment or the water column may bear no obvious relation to climate change [ Snowball, 1994]. On the other hand, Hesse [1994] indicates that the cyclic pattern of variation observed in the magnetic properties of Tasman Sea sediments may result from changes in the assemblage of magnetotactic bacteria living at the site in response to environmental changes that are driven by glacial-interglacial cycles. Transmission electron microscopy (TEM) studies of magnetic separates are necessary to determine the importance of magnetotactic bacteria as a source of magnetic material in a sedimentary environment, and it is advisable that they be done in conjunction with all magnetic-property studies of sediments. Moskowitz et al. [1993] and Oldfield [1994] have described potentially useful magnetic-property measurements to identify bacterial magnetite that could be used in conjunction with TEM studies.
Another problem is the effect of post-depositional diagenesis on the primary magnetic-property signal. Many workers before 1991, and recently Bloemendal et al. [1992], showed that alteration (diagenesis) under chemically reducing conditions can significantly influence sediment magnetic properties in areas of high biologic productivity, in part by preferentially removing the fine-grained magnetite fraction. In some cases, the climatic record will be either severely degraded or destroyed [ Bloemendal et al., 1992], whereas in others, such as the Arabian Sea, reductive diagenesis can destroy the magnetic ``noise'' from in-situ bacterial magnetite and leave behind a climate signal from coarse, eolian-derived magnetite [ deMenocal et al., 1991; Bloemendal et al., 1993]. In addition to early diagenesis at the top of the sediment column, reductive diagenesis and/or dissolution of magnetite can occur in deeply buried sediments [ Lu and Banerjee, 1994]. Clearly, detailed magnetic-property and geochemical tests are necessary to identify the effects of post-depositional processes and to avoid misinterpreting post-depositional variations as a primary climate signal. One recent approach to this problem [ Dekkers et al., 1994] uses magnetic and chemical properties (including the amounts of calcium carbonate, barium, manganese, and sulfur) as variables in a multivariate analysis with fuzzy c-means (FCM) cluster analysis. Four different clusters of diagenetically altered samples were identified in a lithologically variable core from the central Mediterranean Sea. This method, although complex, appears to be useful not only for identifying intermittent intervals affected by diagenesis but also for determining the type and degree of diagenesis.