DISCUSSION

This section is not meant to solve all the questions that arose during study of this material, rather it presents the species observed in these cores and their present-day ecological restrictions and some inferences about the transported vs. in situ nature of the diatoms (see "Taxonomy and Ecological Preferences", for complete listing).

As was previously noted, present-day in situ sedimentation and preservation of diatoms at Site 1063 are quite rare (see Table T1 for species observed in surface sediments). This trend extends downcore with a few periods of high diatom numbers to 9 mbsf (see Fig. F1; Tables T2). One species, Hemiaulus hauckii, was confined to occurrences within a restricted biostratigraphic range (Samples 172-1063D-2H-2, 23-24 cm, to 83-84 cm). This is an extant species that presently occurs in warm to temperate water regions. Apart from this species, all others occur throughout the entire study interval. Species not recorded for each sample will clearly have noncontinuous ranges.

Maynard (1976) discussed the correlation between phosphate values in the waters and the diatoms per gram of sediment in the underlying surface sediments. The correlation between areas with abundant phosphate and high primary productivity and subsequent deposition of diatoms in the sediment is great. Very little lateral transport occurs, with the biocoenoces differing very little from the underlying thanatocoenoces. As nutrient supply is the necessary ingredient for abundant diatom production, a nutrient source should be explained before the intervals of abundant diatoms at Site 1063 is inferred to be in situ. But the idea can not be ruled out solely because this situation does not currently exist in the area. The diversity and preservation of the assemblage supports the interpretation that previous conditions were favorable for periodic abundant diatom growth.

Several discernible patterns are recognized from the biostratigraphic data of this study, presented below in no particular order of importance.

1. As is illustrated in Figure F2 and listed in Table T5, the species observed in this study exist today in a variety of biological habitats. Some habitats, such as coastal and littoral settings, are more restrictive than neritic or planktonic habitats and offer more precise information than the latter. The known present ecological range of individual species can aid in the determination of the history of the material studied. Species that have been observed in either benthic habitats or have strong associations to continental coasts include C. disculoides, C. lorenziana, D. surirella, Diploneis spp., Podosira stelliger, Suriella spp., Fragilariopsis doliolus, H. hauckii, R. amphiceros, and Thalassionema nitzschioides.

   When the abundance data for these and other species are plotted against depth (Fig. F3), it is apparent that the above-mentioned species are more abundant in the lower samples of the study with fewer occurrences in samples from the upper part of the hole, whereas species with oceanic/planktonic habitat preferences are observed throughout the study interval. These data support the idea of a change in currents or sediment source between the periods of deposition for the lower and upper sections of the study interval. Chaetoceros lascinosus and Thalassiosira nordenskoldii, two species that are more commonly found today in colder waters, have a higher abundance in the lower samples (below 21 mbsf). D. surirella, common along coasts in colder climes, does not decrease towards the top of the hole but assumes a more sporadic pattern of appearance.

2. The diatom assemblages in the low-density intervals are not uniform in character; all species do not experience the coincident increases and decreases in abundance. This would indicate that these intervals are not merely the result of a decrease in other sedimentary material (such as calcareous nannofossils or terrigenous material). Conversely, the periods with higher density can not have been created solely through dilution by terrigenous material. This is apparent when the sieved material for the high density intervals is examined. Some samples have abundant and rich diatom assemblages, whereas others are barren. Changes do occur in the accumulation patterns of other sedimentary material, but this can not entirely explain the diatom abundance changes observed.

   An example of the noncontinuous quality of the diatom assemblage can be found between 22.63 and 21.63 mbsf. Here, Ethmodiscus rex is observed in low numbers, whereas T. nitzschioides has a large abundance peak. Also, Ditylum brightwellii disappears almost completely from the sediment record over the interval 26.03-21.23 mbsf (Stage 4). Curiously, this last species is not presently observed in polar regions. This might seem to indicate a change in water temperature over this site or the source site for the sediment at this time, except that other species restricted from cold waters do not also decrease in numbers during this interval.

   Dilution is indeed occurring, because the species diversity of the assemblage does not change between the intervals with higher and lower abundances of diatoms. One does not find merely the very robust, corrosion-resistant species in the intervals with low total abundance. In addition, the valves observed in the sieved fraction do not show any more breakage or dissolution during these high-density intervals than those from the raw material with more abundant diatoms. In fact, there is very little dissolution of the fine features of the diatom valves at all. Abrantes (1991) uses several species as indicators of dissolution, two of them being Bacteriastrium hyalinum and Thalassiosira spp., because of the fragile nature of their valves. These species continue to be part of the assemblage, though decreased in number, throughout the study interval. Two of the species that she places in the resistant category are P. sulcata and Actinocyclus nodulifer. These species do experience gaps in their range during periods with low total diatom abundance, further supporting the idea of dilution with increased amounts of other sedimentary particles.

3. Five species (B. hyalinum, C. lascinosus, Nitzschia marina, Paralia sulcata, and T. nitzschioides) exhibit a pattern of abundance that increases throughout Stage 4. This pattern is not completely smooth, but the general trend is from lower abundances at the beginning of Stage 4 (~25.63 to ~23.63 mbsf) to higher near the end (~23.63 to ~21.33 mbsf). The species in which this pattern is the most pronounced is T. nitzschioides. B. hyalinum, T. nitzschioides, and C. lascinosus all occur today in temperate waters, P. sulcata is a cosmopolitan species, and N. marina is more common in warm waters. B. hyalinum and N. marina live in the pelagic realm, whereas the other three are associated with neritic environments. The variety of present ecological and regional affinities does not assist in determining a cause for this observed increase. All five species also show a slight to pronounced decrease in abundance just prior to the onset of Stage 4.

   Several other species (D. surirella, F. doliolus, T. eccentrica, T. lineata, and T. oestrupii) have a brief interval of high abundance in the upper part of Stage 4 but do not exhibit a general steplike increase throughout this period. One species that is obviously missing during most of Stage 4 is D. brightwellii. The only ecological affinity found in the literature for this species observes that it is a cosmopolitan species that is not recorded from the polar regions (Hasle and Syvertsen, 1996). For the rest of the species with abundance vs. depth data illustrated in Figure F3, there is no discernible pattern or noticeable peaks or lows during Stage 4.