The Leg 188 cores from all sites in Prydz Bay show some evidence of cyclic patterns on a wide range of timescales, from millions to tens of thousands of years (Shipboard Scientific Party, 2001a). Our usage of "cyclic patterns" ranges from noting patterns that are regular and systematic (e.g., periodic variations at Milankovitch frequencies) to patterns that have widely variable data parameters and that may or may not have good temporal control (e.g., stacked lithostratigraphic units and unconformities and variable physical properties of sediments—with variable quality age control). Our usage means similarity but not exact repeatability of data parameters in subsequent cycles. Such cyclicity is common in drill core from around Antarctica and elsewhere and, where preserved, provides a useful guide to inferring depositional paleoenvironments and processes that are common to dynamic glacial and oceanographic systems.
On the continental rise, recovery was very good, as were downhole logging records, thereby yielding good records of cyclic patterns. Age dating in this setting was hampered by extensive silica diagenesis, which resulted in loss of silicious microfossils at great depth and dissolution of magnetite in the upper part of the hole. Both losses meant degradation of age control in these parts of the section, which in turn affected the ability to decipher accurate periodicities for cyclic patterns. The rarity of calcareous microfossils on the rise due to large water depths and high polar latitudes prevented the definition of the oxygen isotope stratigraphy, useful for calibrating ages and deciphering paleoceanographic Milankovitch periodicities. On the continental slope, logging-while-drilling data provided surprising and useful information on cyclic patterns, but the absence of diatoms in the fan deposits precluded making reliable estimates of cyclic periodicities. On the continental shelf, the poor core recovery and the large natural variability and incompleteness of the sediment section permitted only sporadic observation of cyclic patterns and no resolution of periodicities.
Cyclic patterns are best documented in the continental rise drift deposits from Site 1165. Below ~60 mbsf, the patterns can be visually observed as systematic fluctuations between darker- and lighter-colored sediments that have different physical properties and biogenic contents (Figs. F10, F11). The fluctuations occur throughout the hole, but are masked above 70 mbsf by the uniform brown color of the sediment and masked below ~400 mbsf by low color contrast between dark-colored terrigenous and biogenic units. In these sections of the core, the cyclic patterns are best seen in the laboratory and downhole logging measurements, respectively.
Warnke et al. (this volume) illustrate the cyclic patterns in high-resolution measurements from the upper 50 m of the section, and these are recorded in clay-mineral assemblages, grain-size distribution, spectrophotometer reflectance curves, and 
18O measurements (in the upper 17 mbsf, where calcareous microfossils are found). The authors do not attempt to establish cyclic periodicities but do relate the fluctuations in parameters to a dynamic behavior of the ice margin in Prydz Bay. A tentative identification of marine isotope stages (MIS) 19 and 71 is proposed, consistent with interglacial–glacial variations.
In the upper part of the core in the late Miocene–age section, the cyclic pattern is visually evident and was described initially by the Shipboard Scientific Party (2001b) as alternations between a lighter-color greenish gray facies and a darker-color gray to dark gray facies. Color variations may indicate changes in Fe2+ (green) and organic carbon (gray) (Potter et al., 1980). These cyclic patterns are also seen in the lightness factor values from the spectrophotometer (Damuth and Balsam, this volume). The lighter facies have lower density and magnetic susceptibility values and greater biogenic (principally diatoms) content than the darker facies. The patterns in the upper part of the Site 1165 core are more thoroughly described by Rebesco (this volume).
Cyclic patterns in the lower part of the core appear as systematic and coincident spikes in resistivity (positive) and gamma ray (negative) values (Fig. F12). The spikes occur at cemented beds with lower clay content (Facies III-2)—beds that are inferred to have had higher biogenic content, now diagenetically altered, than encasing beds (Shipboard Scientific Party, 2001b). Initial onboard studies suggested the cemented beds were calcareous, but postcruise studies (Williams et al., 2002) indicate that they are predominately silicious. The cyclic patterns throughout the hole demonstrate that sedimentation since the early Miocene (bottom of hole) has systematically varied between principally terrigenous and biogenic sources.
The cyclic patterns appear to fall in the range of Milankovitch periodicities, within the limits of errors in the age vs. depth curve for the hole (Fig. F11). For the interval 83–100 mbsf (late Miocene) the shipboard-determined periodicities in the lightness and bulk density values are ~94, 41, 31, 21, and 18 k.y. (Shipboard Scientific Party, 2001b). Postcruise studies of the same interval (6.76–7.21 Ma) by Grützner et al. (2003) incorporate additional high-resolution measurements of color spectra, X-ray fluorescence (XRF), Fe, grain size of IRD, and magnetic susceptibility. They use cross-spectral analyses to refine the periodicities to 84.1, 41.9, 20.0, and 16.8 k.y. and note that the correlation of spectral peaks with astronomic cycles of obliquity (41 k.y.) and precession (19–23 k.y.) suggests an orbital origin for the observed variations in color and iron concentration.
For the lower section of the hole (~675–999 mbsf; ~18 to ~22 Ma), the postcruise studies (Williams et al., 2002; T. Williams, pers. comm., 2004) include analysis of the cyclicity in the downhole resistivity and gamma radiation values (Fig. F12). They suggest that the regular spacing of Facies III-2 rocks with resistivity peaks in the downhole logs could also be astronomically paced. Between 18 and 20 Ma the average thickness of cycles is 1.55 m, which gives a periodicity of ~15–23 k.y., depending on sedimentation rates. Between 20 and 22 Ma, the average cycle thickness of 13 m gives a periodicity of ~135 k.y., based on the age model in Fig. F6, but may be shorter with their alternate age model. Studies of lithologic cyclicity for nearly the same time period in Antarctica are reported for the Cape Roberts area of the Ross Sea adjacent to the Transantarctic Mountains (Naish et al., 2001). Here the site is proximal to the front of an East Antarctic outlet glacier, and they tie the cycles to orbital influences at periods of 40 and 125 k.y.
Cyclic patterns of terrigenous and biogenic units in continental rise drift deposits at Site 1165 adjacent to the East Antarctic Ice Sheet can be compared with similar cyclic sediments in rise drift deposits at ODP Site 1095 adjacent to the Antarctic Peninsula (West Antarctic Ice sheet). At Site 1095 initial shipboard analyses of alternating terrigenous (gray) and biogenic (brownish) units pointed to the variations having Milankovitch periodicities (Shipboard Scientific Party, 1999). However, detailed postcruise studies by Lauer-Leredde et al. (2002) and Pudsey (2002) established that the observed lithologic cyclic patterns are far from regular and included Milankovitch as well as other periodicities ranging from ~20 to 110 k.y. The irregularity is due in large part to the inherent variability of different age models for the drill site (e.g., magnetostratigraphy and diatom/radiolarian biostratigraphy). Pudsey (2002) notes that in the late Miocene section (6.2–6.4 Ma), which is nearly equivalent to the period that Grützner et al. (2003) studied at Site 1165, the irregularity may be related to drainage basin geology and tectonic uplift masking a well-defined orbital cyclicity. The Prydz Bay region is an older passive margin setting (e.g., Cooper et al., 1991a) than the Antarctic Peninsula, and depositional patterns near Site 1165 have not been greatly affected by basin tectonics since the early Miocene; therefore, such processes would not mask a climatic cyclic pattern.
Cyclic patterns in cores from Site 1166 on the continental shelf are poorly defined due to generally low core recovery and the highly varied lithologies encountered (Fig. F5). Lithologic cyclicity is observed in the late Eocene to early Oligocene as alternating nonmarine massive sands and silty sands of logging Unit 4A (163–195 mbsf). Here, the cyclic pattern may result from variations in fluvial/tidal environments resulting from changing sea levels or autocyclic processes in the depositional environment. In the early Pliocene and younger age glacial section (0–135 mbsf), three possible cyclic intervals in grain density are noted and may represent three advances and retreats of the continental ice sheet (Shipboard Scientific Party, 2001c). Isotopic measurements (18O) on foraminifers from this section by Theissen et al. (this volume) show large variations that are consistent with glacial–interglacial changes in ice volume/temperature, but the values are too sparse to establish a clear cyclic pattern.
Site 1167 on the continental slope is marked by irregular cyclic patterns in lithologies, magnetic susceptibilities, and 18O values at widely differing scales. The age control is, however, inadequate to establish periodicities.
Cyclic lithologies at Site 1167 are denoted by thick debris flow units (as thick as ~50 m) alternating with thin mud horizons (as thick as 3 m). The irregular cyclic pattern can be seen in cores but is more evident in the continuous downhole logs that were recorded in the upper 260 m of the section (Fig. F9). In the logs, the thin and poorly recovered mud layers appear as coincident spikes in resistivity and gamma ray values and delineate ~15 debris–mud cycles of variable thickness (Shipboard Scientific Party, 2001d). In this part of the hole, as noted above, a cyclic sawtooth pattern was identified in shipboard magnetic susceptibility values within four broad intervals (50–80 m thick) that are characterized by linear upward-decreasing susceptibility values (Shipboard Scientific Party, 2001d) (Fig. F9). These intervals are bounded by mud layers. Cyclic patterns are also evident in 18O measurements on planktonic and benthic foraminifers and are of variable periodicity superimposed on a nearly linear uphole increase in 18O values (Theissen et al., this volume). Even where measurements are sparse, below 50 mbsf, cyclic variations are systematic, sometimes large, and sometimes (but not always) bounded by mud horizons. Above 50 mbsf, Theissen et al. (this volume) tentatively identify MIS 16–21, in which each stage has a cyclic period of about 4–8 m, and is interpreted as a glacial-interglacial cycle.
The cyclic patterns at Site 1167 can generally be explained by fluctuations in oceanographic and depositional paleoenvironments in response to changes in the Antarctic cryosphere and depositional source areas over the past 1–2 m.y. Periods of increased ice volumes, when glaciers extended to the continental shelf edge, are times of enhanced sedimentation and debris flows, and contrast with periods of diminished ice and low sedimentation and resulting mud layers. Isotopic measurements in slope debris flow units also support this general concept. Yet many detailed questions for Site 1167 remain unanswered, including
18O measurements, consistently match apparent paleoclimate markers (e.g., mud layers as possible markers of interglacials), as commonly postulated? The drilling data from Site 1167 provide the first detailed deep measurements from an Antarctic trough-mouth fan to help answer these questions. The new data illustrate, however, that prior conceptual models of ice sheet depositional processes may not apply to the world's largest glacier outlet system in Prydz Bay. The cyclic patterns are real, but the regional data are still inadequate to explain them.
Leg 188 drilling provides further evidence that paleoenvironments and paleoclimates for the Prydz Bay continental margin have strong cyclic components throughout the Cenozoic since earliest glacial times. There is a long-term cyclicity over millions of years in the climatic and ocean cooling, as seen in the changes in palynology (e.g., Site 1166) (Macphail and Truswell, this volume a), the systematic increases in 18O isotope values (Site 1167) (Theissen et al., this volume, 2003), and ice volume and sediment delivery as reflected by systematic increases in IRD and decreases in sedimentation rates since the middle Miocene (Site 1165) (Fig. F6) (Shipboard Scientific Party, 2001a). There is also a shorter-term cyclicity over tens of thousands of years in glacial and interglacial periods reflected in changing lithologic properties across the continental margin. These include the fluctuations in physical properties and diatom contents of stacked strata on the shelf (Site 1166) (Shipboard Scientific Party, 2001c), the alternating debris flows and mud layers on the slope (Site 1167) (Shipboard Scientific Party, 2001d), and systematic fluctuations between terrigenous and biogenic facies on the rise (Site 1165) (Shipboard Scientific Party, 2001b). The glacial and interglacial cyclicity is seen also in variations in 18O isotope measurements at all three Leg 188 drill sites (Theissen et al., this volume, 2003; Warnke et al., this volume), but is best documented for Leg 188 at Site 1167.
The explanations for the cyclic patterns are equivocal based solely on Leg 188 data because there are many contributing and interrelated factors that have been widely discussed regarding cyclicity—factors that include changing ocean currents and temperatures, sea levels, ice volumes, terrigenous erosion rates, biogenic productivity, CO2 levels, and others. The relative importance of these factors changes with location and time. On the continental shelf, the late Eocene cyclic sand–shale sequences point to controls of changing sea levels, but in the early Oligocene and again in the late Neogene the greatest control on cyclic patterns was glacial extent. In the early Miocene, the cyclic patterns on the continental rise may have been largely controlled by distribution of abundant fine-grained terrigenous debris by downslope and bottom-contour ocean currents, whereas in the middle Miocene and younger times there was an increasing relative effect of biogenic productivity and climatic cooling (with resulting cold-ice cover and lower terrigeneous sediment supply) that dominates cyclic patterns. On the flanks of the Kerguelen Plateau (~1000 km northeast of Prydz Bay), cyclic terrigeneous–biogenic lithologic patterns in the late Miocene and younger sections at ODP Sites 745 and 746 result from variable biogenic productivity and terrigenous sediment supply (Ehrmann et al., 1991; Ehrmann and Grobe, 1991). Generally, for the late Neogene, the greatest controls on cyclic patterns throughout the Prydz Bay region are from volume changes of glacier ice affecting erosion and redistribution of sediments and from paleoceanographic changes modulating ocean-biogenic productivity.
