Shipboard classification of primary volcanigenic sediments encompassed three principal groups: pumices, pods, and layers (Sacks, Suyehiro, Acton, et al., 2000). In this initial study we have excluded pumice analysis, principally because, although pumices may be deposited shortly after their eruption, they may equally drift or reside in the environment to be deposited/transported many thousands of years later (e.g., see Binns, 1972; Newton and Dugmore, 1995; Bryan, 1968). However, this does present some problems in attempting to establish a tephrostratigraphic correlation with the terrestrial record, as some eruptions may be represented in more distal marine deposits by pumice alone.
The process of pod formation is not clear. Cambray et al. (1993) suggest that pods in the Pacific Ocean are formed by the disruption of thin layers of tephra by the burrowing action of benthic dwellers such as Zoophycos and Phonolites. In this mode of pod formation, geochemical integrity of the reworked tephra is likely to be maintained, albeit the proportion of nonvolcanigenic grains may be reduced by dilution. An additional pod formation process whereby strong currents and steep bottom topography play a role in controlling the tephra form has been proposed by Fujioka (1986). If such formation occurs while the tephra remains exposed on the seafloor, its stratigraphical context is not lost and its geochemical characteristics are preserved, albeit with potential dilution of the volcanigenic content. If, however, tephra redistribution by bottom-current/turbidite activity occurs after burial, the potential stratigraphic significance may be lost, as such large-scale reworking processes are likely to concentrate the products of more than one eruption. This reworking process, which can affect tephra layers as well as pods, will result in decreased homogeneity of the unit compared to a primary deposit. In light of these uncertainties, there is some debate regarding the usage of pods, especially their inclusion in the recording of the frequency of volcanic activity, with some workers discounting pods entirely (see discussion in Cadet and Fujioka, 1980). If the pods are to be used as reliable stratigraphical markers, they must have the same likelihood as layers of representing the products of only one eruption. It follows from these assertions that for the layers and pods to function as tephra marker horizons, their geochemistry should be either (1) well-trended (i.e., displaying a clear fractionation/evolution signal) or (2) relatively homogeneous, with no significant difference in homogeneity between pods and layers.
Leg 186 shipboard smear slide data (Sacks, Suyehiro, Acton, et al., 2000) show that none of the pods contain >85% vitric shards in their total framework composition, whereas 50% of layers do (Fig. F2; Table T3). This contrasts with smear slide data from tephras from nearby Deep Sea Drilling Project (DSDP) Leg 57, where 25% of both layers and pods contain >85% vitric shards (Fujioka et al., 1980); homogeneity indices may be a more robust indicator of the suitability of pods for tephrochronological purposes. Homogeneity indices have been proposed (Boyd et al., 1967; Potts et al., 1983) for investigating the "quality" of potential analytical standards. Both of these methods, however, require reference to instrumental counting statistics and are not readily applicable to published data. We have therefore taken the more straightforward, but nonetheless valid, approach of comparing standard deviations of key major elements (Si, Al, K, and Na) for all analyses of each layer/pod. We observed no significant difference between the standard deviations (i.e., homogeneity) of layers and pods (Fig. F3). We conclude, therefore, that layers and pods can be regarded equally in terms of their tephrostratigraphic utility and that pod formation occurs without introducing substantial allochthonous volcanic and nonvolcanic components. From these data however, we cannot comment on the relative merits of the biogenic and current-driven redistribution models that have been used to explain pod formation. In general, the tephra units analyzed in this study are well clustered, indicative of good homogeneity and encouraging for their potential tephrochronological utility, as demonstrated in the later section on geochemistry and correlations.