Portal de Periódicos da CAPES

Exploring Globorotalia truncatulinoides coiling ratios as a proxy for subtropical gyre dynamics in the northwestern Atlantic Ocean during late Pleistocene Ice Ages

2016; American Geophysical Union; Volume: 31; Issue: 5 Linguagem: Inglês

10.1002/2016pa002927

1944-9186

Katharina Billups, Charlotte Hudson, Hans Kunz, I. Rew,

Isotope Analysis in Ecology

PaleoceanographyVolume 31, Issue 5 p. 553-563 Research ArticleFree Access Exploring Globorotalia truncatulinoides coiling ratios as a proxy for subtropical gyre dynamics in the northwestern Atlantic Ocean during late Pleistocene Ice Ages K. Billups, Corresponding Author K. Billups School of Marine Science and Policy, University of Delaware, Lewes, Delaware, USA Correspondence to: K. Billups, kbillups@udel.eduSearch for more papers by this authorC. Hudson, C. Hudson Cape Henlopen High School, Lewes, Delaware, USASearch for more papers by this authorH. Kunz, H. Kunz School of Marine Science and Policy, University of Delaware, Lewes, Delaware, USASearch for more papers by this authorI. Rew, I. Rew Oberlin College, Oberlin, Ohio, USASearch for more papers by this author K. Billups, Corresponding Author K. Billups School of Marine Science and Policy, University of Delaware, Lewes, Delaware, USA Correspondence to: K. Billups, kbillups@udel.eduSearch for more papers by this authorC. Hudson, C. Hudson Cape Henlopen High School, Lewes, Delaware, USASearch for more papers by this authorH. Kunz, H. Kunz School of Marine Science and Policy, University of Delaware, Lewes, Delaware, USASearch for more papers by this authorI. Rew, I. Rew Oberlin College, Oberlin, Ohio, USASearch for more papers by this author First published: 27 April 2016 https://doi.org/10.1002/2016PA002927Citations: 8AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract We explore the use of the coiling direction of planktic foraminifer Globorotalia truncatulinoides in sediment cores from the northwestern subtropical Atlantic Ocean as a proxy for variations in the intensity of the western boundary of the subtropical gyre over the past 280 kyr. Core-top sediments from the study region are dominated by the left coiling variety consistent with the deep permanent thermocline at the study sites (KNR140-37PC and Ocean Drilling Program Site 1059). Downcore G. truncatulinoides (sinistral) maxima occur in conjunction with 14 out of the 25 (Northern and Southern Hemisphere) precession maxima contained in the study interval. The agreement between the dominance of left coiling tests and the precession index of the Southern Hemisphere, in particular, supports a link between a deep thermocline in the northwestern subtropical Atlantic and northward flow of equatorially sourced warm surface currents, a situation analogous to the Late Holocene. Interglacial marine oxygen isotope stage (MIS) 5 lacks G. truncatulinoides (s) minima attesting to the relative stability of the western boundary during an interval of prolonged global warmth. G. truncatulinoides (s) disappear during the glacial extremes of MIS 2, 6, and 8 implying a weaker western boundary current at these times. Our results support that the coiling direction of this species is sensitive to variations in hydrography of the western boundary of the subtropical gyre. Because of the association between G. truncatulinoides (s) and precession maxima in both hemispheres, results support the importance of oceanic heat transport in half-precession climate variability in the North Atlantic. Key Points A 280 kyr record of Globorotalia truncatulinoides coiling ratios in the subtropical Atlantic Left coiling tests dominate during Northern and Southern Hemisphere precession maxima Coiling ratios track changes in the intensity of the western boundary current 1 Introduction The western boundary of the North Atlantic subtropical gyre, the Gulf Stream, is a major component of the climate system [Schmitz and McCarthy, 1993]. This current transports as much as 30 sverdrups of warm water to high northern latitudes leading to warm anomalies of up to ~10°C [Rahmstorf, 2002]. The intensity of the flow correlates with North Atlantic warming and freshening with implications on the strength of thermohaline overturning circulation [Sallenger et al., 2012]. On the more regional scale, a recent weakening of the Gulf Stream system may be contributing to the 3 to 4 times higher than global rise in sea level along the mid-Atlantic coast [Sallenger et al., 2012; Ezer et al., 2013]. Clearly, understanding the long-term stability of the subtropical gyre and its boundary, the Gulf Stream system, has important societal implications on the global as well as regional scale. Proxy-based reconstructions suggest that while the position of the Gulf Stream appears to have remained relatively constant, its intensity is quite variable through time. Planktic foraminiferal (Globorotalia truncatulinoides) δ18O values indicate that the position is similar to today even during the climatic cold extreme of the Last Glacial Maximum [Matsumoto and Lynch-Stieglitz, 2003; LeGrande and Lynch-Stieglitz, 2007]. At the same time, the intensity of Gulf Stream flow as reconstructed from benthic foraminiferal δ18O gradients may have been reduced by 30–50% [Lynch-Stieglitz et al., 1999]. Significant variations in the intensity of the Gulf Stream also occur during decadal-scale cold events of the Holocene, such as the Little Ice Age [Lund et al., 2006]. Here we explore the potential of a faunal proxy, namely, downcore changes in the coiling ratio of Globorotalia truncatulinoides, as a potential proxy for the stability of upper ocean hydrography in the northwestern subtropical Atlantic Ocean (Figure 1). We focus our study interval on the past 280 kyr capturing three full interglacial (marine oxygen isotope stages, MIS, 1, 5, and 7) and three full glacial cycles (MIS 2, 6, and 8). Our study sites (KNR140-37PC and Ocean Drilling Program Site 1059, both at ~75°W, 32°N) are located underneath the northwestern limb of the subtropical gyre, to the southeast of the northern edge of the Gulf Stream (Figure 1a). The region is characterized by large seasonal changes in the wind-driven upper layers of the water column (Figure 1b). Westerly winds are strongest at the end of the winter, and the boundary of the subtropical gyre is then relatively diffuse [Hogg and Johns, 1995]. At the end of the summer when winds are weaker, Gulf Stream flow is strongest, and a seasonal thermocline develops above the study site. The permanent thermocline remains deep (500–900 m) throughout the year (Figure 1b) characteristic of the presence of relatively warm, subtropical gyre water [Pickard and Emery, 1991]. Figure 1Open in figure viewerPowerPoint Seasonal ((a, top) January–March, (a, bottom) July–September) sea surface temperatures in the North Atlantic [Levitus and Boyer, 1994]. Black star indicates the location of the study sites southeast of the Gulf Stream (approximated by the black arrow). (b) Corresponding seasonal water column structure (January–March, grey and July–September, black) at the study site, which is characterized by a deep permanent thermocline (500–1000 m). Panels were generated using the interactive website of the Lamont Doherty Earth Observatory. 1.1 Globorotalia Truncatulinoides Coiling Direction in the Subtropical Atlantic G. truncatulinoides are among the deepest dwelling of the planktic foraminifera [e.g., Fairbanks et al., 1980; Hemleben et al., 1989; Lohmann and Schweitzer, 1990]. Their geochemistry has been used quite extensively to assess changes in the thermal structure of the upper ocean [Mulitza et al., 1997; Jian et al., 2000; Matsumoto and Lynch-Stieglitz, 2003; LeGrande and Lynch-Stieglitz, 2007; Cleroux et al., 2007; Bahr et al., 2013]. Molecular analyses indicate that there are four genetically distinct species occupying different hydrographic regimes [de Vargas et al., 2001]. “Species 2” defined by de Vargas et al. exhibits both, the left and right coiling test varieties, and it occupies the warm water of the subtropical oceans. Phylogenetic analysis of G. truncatulinoides from plankton towns in the North Atlantic Ocean confirms that it is this particular type that dominates in the Sargasso Sea near our study sites [Ujiié et al., 2010; Quillévéré et al., 2013]. The left coiling variety of species 2 prefers warm, nutrient depleted waters such as the center of the subtropical gyre, while the right coiling variety favors the still relatively warm, but more nutrient-rich waters associated with the gyre margins [Renaud and Schmidt, 2003; Ujiié et al., 2010]. This general pattern was first recognized in an expansive study of the distribution of left versus right coiling tests in core tops from the North Atlantic [Ericson et al., 1954] (Figure 2). Ujiié et al. [2010] present an updated, spatially more resolved data set that confirms the dominance of the left coiling variety in the subtropical gyre, while the right coiling becomes more abundant toward the gyre margin and dominates the equatorial regions, the Gulf of Mexico, and the northeast Atlantic Ocean. Figure 2Open in figure viewerPowerPoint Distribution of Globorotalia truncatulinoides (sinistral) and Globorotalia truncatulinoides (dextral) in surface sediments from the North Atlantic (Reprinted from Deep-Sea Research, Volume 2, David B. Ericson, Goesta Wollin, and Janet Wollin, Coiling direction of Globorotalia truncatulinoides in deep-sea cores, pages 152–158 (1954), with permission from Elsevier). The left coiling variety dominates the subtropical of the gyre, while the right coiling are found primarily in equatorial regions, the Gulf of Mexico, and the northeast Atlantic Ocean. The spatial distribution is consistent with the vertical habitat. Plankton tows across the western region of the subtropical gyre illustrate that left coiling tests dominate at 200–400 m water depth in the Sargasso Sea, where the permanent thermocline is deep [Ujiié et al., 2010]. The study also shows that right coiling tests become increasingly more abundant, but at 200–100 m water depths, toward the southern gyre margin as the permanent thermocline shoals. Although not supported by such direct evidence, we point out that the region where left coiling tests give way to right coiling tests in the northern gyre margin (for example, at about 70°W, 37°N, Figure 2) coincides with an abrupt shoaling of the permanent thermocline (Figure 3). Figure 3Open in figure viewerPowerPoint Temperature profile along a north-south transect across the Gulf Stream region in the northwestern Atlantic ((top) looking southwest). Graph was generated using the Ocean Data Viewer [Schlitzer, 2014]. Slope water is relatively cold with a shallow permanent thermocline, while the subtropical gyre is associated with a deep permanent thermocline (~500–1000 m) [Pickard and Emery, 1991]. White boxes illustrate the inferred depth habitat of Globorotalia truncatulinoides (sinistral) and (dextral) based on the surface distribution shown in Figure 2, the distribution of this species in tropical to subtropical Atlantic plankton tows [Ujiié et al., 2010], and the conceptual model of Feldmeijer et al. [2014]. (bottom) The white star indicates the location of the study region with respect to the location of the cross section. Studies have proposed that downcore variations in the coiling direction of G. truncatulinoides reflect paleoenvironmental changes. Thiede [1971] recognizes a shift in the coiling ratio in the eastern Atlantic ascribing it to water mass changes. Renaud and Schmidt [2003] observe glacial to interglacial changes in the coiling ratio in the subtropical South Atlantic invoking changes in the species' habitat. Feldmeijer et al. [2014], in particular, provide a conceptual model demonstrating that the left coiling test variety, G. truncatulinoides (sinistral), dominates North Atlantic sediments when the permanent thermocline is deep, while G. truncatulinoides (dextral) are more abundant when the permanent thermocline is shallow. As a deep permanent thermocline defines the subtropical gyre [e.g., Pickard and Emery, 1991], we infer that downcore changes in the dominant coiling direction of G. truncatulinoides can be used to monitor the relatively stability of the subtropical gyre. Given the location of our study sites, and the apparent stability of the Gulf Stream on glacial to interglacial time scales [e.g., Matsumoto and Lynch-Stieglitz, 2003], we infer that G. truncatulinoides coiling direction reflects the intensity of northward flow associated with the western boundary current. We hypothesize that interglacial intervals are dominated by the left coiling variety reflective of the dominance of subtropical gyre water above the core site, while glacial intervals are characterized by the right coiling variety reflecting a generally weaker, more diffuse western boundary current. 2 Methods We reconstruct downcore changes in the G. truncatulinoides coiling direction spanning the past 280 kyr. We sampled the latest Pleistocene portion (MIS 5 through 1) from piston core KNR140-37PC (31°41 N, 75°25 W, ~3000 m water depth), and we extend the record through the late stages of MIS 8 using Ocean Drilling Program Site 1059 (31°40 N, 75°24 W, 2985 m water depth). Both cores have been studied extensively, and we picked G. truncatulinoides from previously processed samples [Oppo et al., 2001; Hagen and Keigwin, 2002; Billups and Scheinwald, 2014]. From each core interval, all G. truncatulinoides tests were selected from the >355 µm size fraction (adult samples according to Ujiié et al. [2010]). Feldmeijer et al. [2014] show that changes in the coiling direction of this species are consistent across size fractions, and it is thus not necessary to further size fraction the counts. At Site 1059, all G. truncatulinoides were counted in each sample interval (ranging from n = 2 to n = ~600 individuals, see data archive). In KNR 140-37PC, intervals with a high volume of planktic foraminifera were split prior to size fractioning. The number of tests counted ranges from ~5 to ~400. From counts of left versus right coiling tests we report downcore variations in the percent G. truncatulinoides (s). To check the reproducibility of the counts, and the sensitivity of the signal to low test abundance, we duplicated percent G. truncatulinoides (s) data across an interval of low total test abundance but spanning the entire range of ratios (between 60 and 120 ka, n = 28). We aimed at a millennial-scale resolution of between 500 and 1000 years counting tests at a rate of 8 to 16 cm from KNR140-37 PC and every 5 cm from Site 1059. We compare the percent G. truncatulinoides (s) data and total test counts to the percent sand size fraction of the sediments, which is a qualitative estimator of general foraminiferal versus nannofossil productivity, carbonate dissolution, and dilution with terrigenous sediments. For Site 1059, percent sand size fraction had been recorded during sediment processing by weighing the dried, bulk sediment and, after wet sieving and drying, the >63 µm size fraction. For KNR140-37PC, we have the bulk sample weight corresponding to the 2 cm core intervals for about half of the record (~0–60 ka, L. Keigwin, personal communication, 2016). To complete the record, we use the average sample weight (15.3 g) to derive the percent sand fraction by weighing the remaining >63 µm size fraction. We estimate that the uncertainty related to estimating the dry, bulk sediment weight from the average weight corresponding to a 2 cm slice of the core is on average 0.5%. Having removed foraminifera before weighing the >63 µm fraction introduces an uncertainty on the order of 0.1% for every 100 G. truncatulinoides tests removed (ignoring G. ruber and benthic tests removed in prior studies). Thus, we estimate an overall uncertainty in the percent sand record on average ~ 0.5–2%. 3 Age Model Robust chronology already exists for both of the cores [Hagen and Keigwin, 2002; Billups and Scheinwald, 2014]. For KNR140-37PC, high-resolution age control is based on oxygen isotope correlation between planktic foraminifer Globigerinoides ruber and the Greenland Ice Sheet Project 2 ice core record [Hagen and Keigwin, 2002]. In order to extend this type of chronology through the bottom of KNR140 (MIS 6), we readjust the G. ruber δ18O minima to warming events in the synthetic Greenland ice core δ18O record of Barker et al. [2011] (Figure 4). In keeping with the original age model, the approach implies that millennial-scale δ18O minima are synchronous with warming events on Greenland [Hagen and Keigwin, 2002], which is supported by the nature of temperature variations in this region [Vautravers et al., 2004]. The Site 1059 G. ruber δ18O record was tuned to Northern Hemisphere summer insolation yielding excellent agreement with the synthetic Greenland ice core record [Billups and Scheinwald, 2014]. No further adjustments are necessary. The entire 280 kyr long record thus reflects a consistent chronology. Between ~130 ka and 140 ka, there is a short interval of overlap between the two records. Based on the presence of a small yet distinct δ18O minimum at the beginning of the deglaciation, we splice them together at 134 ka. Figure 4Open in figure viewerPowerPoint Revised age model for KNR140-37PC by aligning minima in the Globigerinoides ruber δ18O record [Hagen and Keigwin, 2002] with interstadial events expressed in the synthetic Greenland ice core record (Gsyn) [Barker et al., 2011]. Vertical dashed lines indicate the age control points. We linearly interpolate between the age control points to calculate sedimentation rates. Inherent uncertainties in the age models include interpolation between age control points, which can be separated by as much as 20 kyr. Furthermore, the age model of the synthetic Greenland ice core target record carries with it an uncertainty of about 2–3 kyr [Barker et al., 2011]. 4 Results Large variations in the relative abundance of left coiling G. truncatulinoides tests occur throughout the study interval (Figure 5a). G. truncatulinoides (s) peaks are outlined by multiple data points attesting to the robustness of the overall signal. Duplicate G. truncatulinoides (s) data are on average within 4% regardless of the test count or coiling ratio (Figure 5a). The error is small compared to the overall range of the signal (0–100% Figure 5a). We are therefore confident that the G. truncatulinoides (s) signal is reliable despite the low number of tests per sample in some of the intervals. Figure 5Open in figure viewerPowerPoint (a) Percent Globorotalia truncatulinoides (sinistral) over the past 280 kyr, (b) the number of G. truncatulinoides tests counted per sample, (c) the percent sand fraction of the sediments and the (d) linear sedimentation rates (LSR) obtained by interpolating between age control points. In black are the data from KNR140-37PC with grey crosses in Figure 5a denoting duplicate counts. In grey are the data from Ocean Drilling Program Site 1059. Vertical grey boxes highlight the interglacial intervals (marine oxygen isotope stages, MIS, are labeled across the top). In general, intervals with maxima in sinistral test abundance tend to be intervals in which the total G. truncatulinoides abundance (counts/sample) and the percent sand fraction of the sediments is also relatively high (MIS 1, MIS 5, MIS 7, Figures 5a–5c, respectively). Conversely, when the percentage of G. truncatulinoides (s) is low, the total test abundance is also low (MIS 2, MIS 6, and MIS 8). Exceptions occur during MIS 3, 4, late MIS 6, and late MIS 8 when distinct G. truncatulinoides (s) maxima are defined by fewer than 100 counts. Variations in sedimentation rate accompany the general pattern being highest during glacial MIS 2, MIS 6, and MIS 8 and lowest during interglacial MIS 5 and 7 (Figure 5d). To a first order, variations in the percent G. truncatulinoides (s) are associated with glacial versus interglacial climate change as exemplified by the G. ruber δ18O record (Figure 6). Percent G. truncatulinoides (s) clearly dominate the Holocene (MIS 1, ~11–0 ka) and interglacial MIS 5 (~130–78 ka). G. truncatulinoides (s) also reach maximum levels during the relatively cold interglacial MIS 3. Glacial MISs, on the other hand, are generally characterized by low abundance, or lack, of this morphotype. G. truncatulinoides (s) disappear numerous times during MIS 2, 6, and 8, which as noted above, are times when the total G. truncatulinoides samples count is also very low (Figure 5b). Glacial MIS 4 is characterized by high variability in abundance of the left versus right coiling tests. That the signal is robust is indicated by the duplicate counts across this interval of time. Figure 6Open in figure viewerPowerPoint (a) Percent Globorotalia truncatulinoides (sinistral) over the past 280 kyr, the (b) previously published Globigerinoides ruber δ18O record (0–134 ka: KNR140-37PC) [Hagen and Keigwin, 2002; Oppo et al., 2001; 134–280 ka; Ocean Drilling Program Site 1059 Billups and Scheinwald, 2014], and the (c) synthetic Greenland ice core record (Gsyn) [Barker et al., 2011]. In black are the data from KNR140-37PC with grey crosses in Figure 6a denoting duplicate counts. In grey are the data from Site 1059. Vertical grey boxes highlight the interglacial intervals (marine oxygen isotope stages, MIS, are labeled across the bottom), and vertical hatched bars reflect the timing of Heinrich events, which are numbered across the top. The timing of Heinrich events 14 and 15 are estimated from Lototskaya et al. [1998] and Feldmeijer et al. [2014]. While the glacial to interglacial pattern appears to be the primary mode, there are a number of instances when percent G. truncatulinoides (s) maxima prevail during glacial intervals but disappear during interglacial intervals (Figure 6a). Percent G. truncatulinoides (s) increase toward Holocene levels during glacial MIS 8 (~250–260 ka) and during the later phases of glacial MIS 6 (at ~140 ka). Conversely, G. truncatulinoides (s) disappear during interglacial MIS 7 (at 235 ka and 215–220 ka). And, an about 40% reduction in their numbers occurs during the Holocene at ~10 ka. All told, variations in percent G. truncatulinoides (s) occur more frequently than the glacial to interglacial background. In fact, G. truncatulinoides (s) peaks at and above 50% tend to coincide with maxima in the Northern and Southern Hemisphere precession indices (Figures 7a and 7b, respectively). To within a quarter cycle (5 kyr) G. truncatulinoides (s) maxima coincide with Northern Hemisphere precession peaks at 267 ka, 198 ka, 127 ka, 83 ka, 58 ka, 33 ka, and 11 ka (Figure 7a) and with Southern Hemisphere precession peaks at 252 ka, 230 ka, 208 ka, 137 ka, 72 ka, 46 ka, and 1 ka (Figure 7b). Figure 7Open in figure viewerPowerPoint Comparison of the percent Globorotalia truncatulinoides (sinistral) record with the (a) Northern and (b) Southern Hemisphere precession indices. Vertical grey bars highlight those respective precession maxima that are accompanied by G. truncatulinoides (s) counts > 50 ± 4%. The width of the bars reflects one quarter cycle (5 kyr). Solid symbols reflect the data from KNR140-37PC [Hagen and Keigwin, 2002; Oppo et al., 2001], and open symbols are the data from Ocean Drilling Program Site 1059 [Billups and Scheinwald, 2014]. Evidently, not all precession maxima are accompanied by percent G. truncatulinoides (s) maxima. Missing from the pattern are precession maxima (from either hemisphere) during glacial MIS 8 (242 ka), MIS 6 (at 187 ka, 175 ka, 164 ka, and 150 ka) and glacial MIS 2 (20 ka). MIS 5, on the other hand, is characterized by a general lack of pronounced G. truncatulinoides (s) minima. Thus, while it appears that the coiling ratio is sensitive to precession forcing, during climatic extremes, either cold or warm, the proxy response can be muted. Variations in the percent G. truncatulinoides (s) do not appear to be sensitive to higher-frequency millennial-scale variations. Sinistral test maxima do not seem to be associated consistently with minima in the G. ruber δ18O record or Greenland ice core interstadials nor do they readily respond to the classical Heinrich events (Figure 6). Heinrich events 14 and 15 may provide the exception as both contain maxima in sinistral test abundance. These latter results echo observations at a site downstream in the North Atlantic [Feldmeijer et al., 2014], perhaps reflective of the subtropical gyre/Gulf Stream/North Atlantic Drift connection. 5 Discussion Results show that, in general, changes in the coiling ratio are associated with variations in the total G. truncatulinoides count, the percent sand size fraction of the sediments, and the sedimentation rates (Figure 5). Largely, these variations correspond to variations in the glacial to interglacial climate background. At the water depth of the two sites (~3000 m), sedimentation rates increase primarily due to the focusing of sediments by the Deep Western Boundary Current [Keigwin et al., 1998]. Thus, the decrease in total G. truncatulinoides counts (and the percent sand) during the glacial intervals, which is when sedimentation rates tend to be highest, reflects dilution with terrigenous material [Keigwin et al., 1998]. This means that the abundance of left and right coiling G. truncatulinoides (and other foraminifera) is equally affected by changes in the depositional environment. Carbonate dissolution, which affects the amount of whole planktic foraminifera in the sediments and hence the percent sand size fraction in a sample, should not discriminate between coiling direction of planktic foraminifera. We are therefore confident that the percent G. truncatulinoides (s) record reflects changes in the upper ocean thermal structure although the counts are, many times, based on fewer than 100 tests. It then follows that there are distinct variations in the depth of the permanent thermocline above the core sites and thus the northwestern reaches of the subtropical gyre. According to our working model, when the sinistral variety dominates, as during the late Holocene, the permanent thermocline is deep. As the late Holocene G. truncatulinoides (s) maximum corresponds to a Southern Hemisphere precession maximum, we infer that during past times of Southern Hemisphere precession maxima the region is characterized by a northwestern subtropical gyre margin with dynamics similar to today. Conversely, when the sinistral variety decreases, which implies that the right coiling tests dominate, the permanent thermocline must be shoaling. Given that the position of the gyre boundary, the Gulf Stream, remains stationary [Matsumoto and Lynch-Stieglitz, 2003], a shallower permanent thermocline might indicate weaker more diffuse flow in the western boundary current allowing a relatively high proportion of slope waters to reach the study sites. Using the modern analog, we think that the observed association of G. truncatulinoides (s) maxima with Southern Hemisphere precession is consistent with a response related to changes in the intensity of the Gulf Stream, because the Gulf Stream is fed, in part, with warm surface waters originating in the Southern Hemisphere [e.g., Pickard and Emery, 1991]. As the Gulf Stream is the western boundary current of the subtropical gyre, variations in its strength should communicate to the study sites. Interestingly, this holds during glacial MIS 8 (255 ka) and the latest part of glacial MIS 6 (140 ka). If our interpretation of the coiling direction changes is correct, it illustrates that the upper limb of the thermohaline overturning cell can be relatively strong despite glacial boundary conditions. The lack of correspondence with the precession index of either hemisphere is also of paleoenvironmental significance. The lack of a percent G. truncatulinoides (s) maximum during the Southern Hemisphere precession maximum of the Last Glacial Maximum (MIS 2, ~20 ka) is consistent with reduced Gulf Stream flow and reduced meridional overturning circulation at that time [Lynch-Stieglitz et al., 1999]. Continued dominance of G. truncatulinoides (s) during interglacial MIS 5, on the other hand, should then attest to the relative stability of the subtropical gyre and Gulf Stream during intervals of prolonged global warmth. This latter observation contrasts the relatively large and rapid δ18O (presumably temperature) oscillations recorded by the Greenland ice core (e.g., Figure 5a versus Figure 5c). And it contrasts terrestrial climate instabilities in the southeastern United States as observed by downcore pollen assemblage changes at Site 1059 [Heusser and Oppo, 2003]. The discrepancies highlight the different response of the sea surface and mixed layer versus deeper portions of the upper water column justifying the closer look at the subsurface environment for a more complete understanding of subtropical gyre dynamics and flow in the western boundary current. Association of the percent G. truncatulinoides with precession maxima of both hemispheres is reminiscent of half precession cycles, which are usually interpreted as being caused by low-latitude insolation forcing. The conceptual model is based on the seasonal analog. The sun crosses the equator twice a year with additional temperature maxima associated with the solstices [Short et al., 1991; Berger et al., 2006]. The Southern Hemisphere temperature maxima are then transferred into the Northern Hemisphere via cross-equatorial surface currents feeding the western boundary current leading to an observed double beat [Short et al., 1991; Hagelberg et al., 1994; McIntyre and Molfino, 1996] or its multi