Search

The timing of glacial-interglacial cycles shows a clear dependence on the periodic variations in the Earth’s orbital parameters. However, the Earth’s climate is an extremely complex, non-linear system, with many internal feedback mechanisms and there are still features of the climate record for which a definitive explanation remains elusive. Understanding reef history is important due to significant predicted feedbacks between changes in global climate and carbonate production via the carbon cycle; phases of rapid reef growth in shallow water areas being associated with increased release of carbon dioxide to the atmosphere. Previous work on Pleistocene reef history, investigated via reef boreholes, shows a large global expansion of reefs between 800-400ka; approximately concurrent with one of the major unexplained alterations in the climate system, the Mid-Pleistocene Transition (MPT). Quaternary reef history is usually investigated via reef boreholes, which provide limited spatial information and are subject to dating uncertainties of the order of ±100kyrs. This means that any inferences made about the relationship between reef expansion and specific changes in the climate system are not well constrained. This thesis instead, presents a novel approach to reconstructing reef growth history, using a trial site near the island of New Caledonia, in the south west Pacific. The initiation of carbonate production on shallow shelves is known to produce a signal in the surrounding deeper basins, via sediment shedding. Therefore, this research set out to independently verify the proposed expansion of the New Caledonia Barrier Reef at ~ 400ka (Marine Isotope Stage [MIS] 11) by examining the composition of turbidites deposited in the New Caledonia Trough. Deep sea sediment core MD06-3019, was collected south west of the New Caledonia Barrier Reef (22oS, 165oE; 3,500m water depth). It is predominantly composed of pelagic carbonate ooze, into which 79 sandy turbidite layers have been deposited. These layers interrupt, but do not disturb, the background sedimentation and source material from the shallow shelf, which is carried to the deep sea via submarine canyons. A core age model based principally on orbital tuning, yields a core bottom age of 1,260ka, ~MIS38. This chronology has allowed the timing of deposition of the turbidite layers within the core to be assigned to within ±10kyr. Turbidite layers vary in width (1-35cm), grain size (φ=4 to -2) and composition, containing among other shelf derived material, well preserved coral fragments from 1.26Ma through to the present day. Patterns in turbidite timing and frequency, grain size and composition (investigated via point counting, carbonate coulometry and aragonite content) have been analysed, to assess whether there are any temporal changes which may reflect variation in shallow shelf reef extent. This included the development of XRF scanning measurements for [Sr], as a new proxy for the aragonite content of samples. A calibration line with the equation: Aragonite %=0.0011* Sr count +2.64 (R2 =0.6105, p-value < 0.001) was obtained for turbidite samples from sediment core MD06-3019. The method shows significant promise as a new proxy for quickly establishing the aragonite content of sediment samples. Corroborating the work of previous investigators, turbidites deposited since MIS11 show an increase in average bulk carbonate and aragonite content, a greater dominance of shallow water bioclasts and a higher occurrence of coral fragments. Additionally, both coarse and fine grained turbidites are present, whereas directly before this period only fine grained turbidites occur. However, there is another significant shift in depositional style further back in the record. Prior to MIS23 both coarse and fine grained turbidites are present, the average carbonate content of turbidite layers is higher and there is a greater dominance of shallow water biota. Coral abundance for turbidites at the base of the core can equal values for turbidites at the top of the core. These results challenge the assumption that the only significant evolution on the western New Caledonia margin over the last 1.2Myrs was the expansion of the barrier reef at MIS11. This suggests that the history of the western New Caledonia margin may be more complicated than initially anticipated. These temporal variations in turbidite deposition are interpreted as reflecting changes in the level of carbonate production on the shallow shelf over the course of the 1.26Myr record. Shallow water carbonate production having decreased substantially during the period MIS23-MIS11. There are many possible controls on the shallow water carbonate production; such as: sea surface temperature and salinity, sea-level and nutrient availability. However, it is hypothesised that the principal control is glacial-interglacial sea-level change. It is proposed that prior to MIS23 sea-level was high enough during certain interglacial periods for significant carbonate production to occur on the shelf. However, from MIS23 onwards the climate proceeded into a period of ‘lukewarm’ interglacials which were both cooler and had lower sea-levels. It is hypothesised that during this period sea-level did not rise enough during highstands to flood the shelf sufficiently to allow for significant shallow water carbonate production. The high sea-levels of the long, warm MIS11then allowed for the expansion of the barrier reef (perhaps on substrates provided by former siliciclastic coast lines, deposited between MIS23-11) and its continuation during subsequent interglacial periods until the current day. This pattern of shallow shelf carbonate production is similar to those proposed for the Belize margin and the Gulf of Papua over the last 1.2Myrs. This thesis provides one of the first detailed investigations of gravity deposits in the New Caledonia Trough, providing information on their composition and timing over an unprecedented 1.26Myr time period. This study demonstrates that deep sea turbidites, sourced from shallow shelf areas, can be used to help reconstruct tropical reef growth histories. The results corroborate the work of previous researchers in the area and provide new insights into the history of reefs along the western New Caledonian margin. The main advantage of this method, compared to traditional borehole techniques, is the 10-fold reduction in the age uncertainty of events, to ±10 kyr. In addition, because turbidite material is sourced from a wide area along the coast, the method is able to provide information on reef history over a larger spatial area than single reef boreholes. This method can now be extended globally to help improve knowledge of the timing and history of tropical reef growth during the Quaternary. This will enable a better understanding of how reefs have impacted on, and been affected by, changes in climate, linked by feedbacks mechanisms via the global carbon cycle.

In this study, specimens of Tridacna sp., which are reef dwelling bivalve molluscs and have been shown to live up to 60 years, collected from the Huon Peninsula, Papua New Guinea, were sampled for geochemical profiles. The Huon Peninsula is in the heart of the Western Pacific Warm Pool (WPWP), which plays a key role in ENSO dynamics. The uplifted reef terraces of the Huon Peninsula have been extensively studied, and are well dated, which gives the opportunity to reconstruct the local climate of this region at key intervals during the past. Previous work on Tridacna sp. has shown that they precipitate their aragonite shell in equilibrium with the surrounding seawater, and the δ18O profile of a modern T. gigas from the Huon Peninsula has been shown to correlate with precipitation and temperature anomalies, and the Niño 3.4 temperature anomaly record. Fossil samples from this region are therefore assumed to have the ability to capture changes in δ18O attributable to ENSO. Seasonally resolved δ18O measurements from Tridacna sp. from early Holocene and Marine Isotope Stage 3 (MIS3) reefs were used to reconstruct changes in mean climate, seasonality and inter-annual variability (e.g. ENSO). Reconstructions of the mean state tend to agree with previously published studies of Holocene and MIS3 climate, showing similar temperatures to today during the early Holocene, and an average cooling of 2- 3°C during MIS3. The early Holocene Tridacna sp. samples show a reduction in seasonality, consistent with the reduction in seasonal insolation at this time, while those from MIS3 show variable seasonality between 30-60ka. ENSO appears to have been supressed during the early Holocene by up to 50% compared with the late 20th century, which is consistent with coral data and modelling studies. During MIS3, ENSO appears to have been more variable with some records showing anomalous warm and cool events as strong as those seen in the modern T. gigas, used here as a benchmark. Trace element profiles derived from the Tridacna sp. used in this study show a tentative link with temperature and local productivity, but these relationships are subject to species specific and intra-shell effects.

The end-Permian extinction at 252 Ma is widely regarded as the most severe of the Phanerozoic mass-extinctions and enabled the evolution of the modern carbon cycle and ecosystem structure. The cause of the extinction is still debated but the synergistic pressures of global climate change, such as anoxia and ocean acidification, were clearly important. The extinction occurred in two phases and is marked by a uniquely protracted recovery period of ~ 5 Myrs where diversity fails to reach pre-extinction levels until the Middle Triassic. This period is characterized by an unstable global carbon cycle, secondary extinctions, reef, chert and coal gaps, and changes in the carbonate factory from reef to microbial and abiotic dominated deposition. This thesis focuses on using geochemical data from the Arabian Margin to investigate the carbon cycle record and the links between kill mechanisms and carbon cycle dynamics. A new record of carbon cycling is presented for the Tethys in the form of a carbon isotope record for the entire Early Triassic from the Musandam Peninsula, United Arab Emirates (UAE). The Musandam carbon isotope record can be broadly correlated with global isotopic events but also resolves additional secondary excursions. These new short-lived events are probably related to the occurrence of the more widely recognized Early Triassic excursions, and may represent fluctuations in the driving mechanisms superimposed on the continued instability of the global carbon cycle in the aftermath of the end-Permian extinction. To unravel palaeo-depositional redox conditions this work utilizes geochemical proxies based on Fe systematics (Fe-speciation). To date, however, these proxies have only been calibrated in relation to modern and ancient siliciclastic marine sediments. This clearly limits the use of the Fe-speciation proxy, particularly in relation to carbonate-rich sediments and rocks. This thesis explores the use of Fe-speciation in carbonates using compiled literature and new data from modern oxic and anoxic settings. This new assessment expands the utility of Fe-based redox proxies to also incorporate carbonate-rich rocks that contain significant total Fe (>0.5 wt%), providing care is taken to assess possible impacts of diagenetic processes such as dolomitization. Based on this calibration work Fe-speciation is used to reconstruct the redox structure for the Arabian Margin mixed carbonate and clastic sediments, from the late Permian to the Middle Triassic. Fe-S-C systematics are utilized to identify the spatial and temporal dynamics of anoxia for a Neo-Tethyan shelf-to-basin transect. The unique spatial resolution afforded by this transect allows a direct link to be drawn between biodiversity, carbon cycling and anoxic events. For the first time we can directly observe a switch from deep-ocean dominated anoxia to a dynamic anoxic wedge at the end-Permian extinction. Additionally the data suggest that ferruginous conditions (anoxic non-sulphidic) were dominant in the Tethyan Ocean throughout the Early Triassic, proposing that euxinia was restricted regionally with potential implications for nutrient recycling, carbon cycle models and driving mechanisms. Redox dynamics may have had important implications for the wider carbonate cycle. These two themes are particularly inter-related with regards to oceanic alkalinity and pH. This thesis presents the first shallow water boron isotope record for the Permian Triasssic Boundary, used as a proxy for pH. The record demonstrates some unexpected results; firstly a sudden increase in pH is observed, prior to the first phase of the extinction and interpreted to reflect alkalinity supply from the development of slope anoxia. Secondly there is no evidence for an acidification event at the first phase of the extinction where pH remains stable. A rapid acidification event is, however, seen in the earliest Triassic, contemporary to the second phase of the mass extinction, but delayed compared to the main negative carbon isotope excursion that indicates the main phase of Siberian Trap volcanism. These events may be explained by dramatic changes in ocean the ocean’s buffering capacity linked to changes in alkalinity supply and the carbonate factory.

This thesis investigates changes in temperature extremes between 1950-2005, analysing gridded data sets of observations and climate model simulations. It focuses on changes in the frequency of extreme temperatures occurring in single days or over periods of six or more consecutive days. The study aims to quantify the significance of changes in extreme temperature events and answer the following questions. Are external or human-induced forcings together with natural forcings responsible for the observed change in temperature extremes or can these changes be explained due to natural climate variability alone? Are the observed changes consistent with those from climate model simulations? And are the changes in extremes linked only to changes in the mean climate, or only to those in climate variability or both? The analysis concentrates on changes from global to regional scale and from annual mean to seasonal scale. A detection method is applied to assess if changes are significantly different with respect to the internal climate variability. Results show that there has been a significant increase in warm daily extremes and a decrease in cold ones, both on large and small spatial scales. The increase in warm extremes has been found to be highly correlated with the increase in mean temperature. The changes in daily extremes are well represented in climate model simulations. Changes in the persistent extremes show a detectable increase in the frequency of warm and a decrease in cold events and are reproducible by models.

A multi-disciplinary study of sediment core records from the Mingulay Reef Complex, a cold-water coral reefs system off western Scotland, highlights the potential of cold-water corals from which detailed centennial-scale palaeo-environmental reconstructions can be derived. This study provides a new insight on the mechanisms controlling Lophelia pertusa reef build-up, shifts in biodiversity, the physical/chemical/biological processes and the sedimentary regime. A detailed record of Mingulay Complex growth history shows unprecedented high average accumulation rates of 3 – 4 mm a-1. Marine radiocarbon reconstruction derived from paired 14C and U-series dated fossil corals revealed substantial abrupt oceanic shifts during the Holocene that have repetitively affected cold-water coral growth, eventually causing local disappearance. These periods of reduced accumulation rates are synchronous with other coral structures from the NE Atlantic illustrating basin wide events. Finally, trace/minor element ratios reproducibility within coral skeleton was investigated to test if palaeo-environmental reconstructions could be made from cold-water corals.

A comprehensive theory explaining the relationship between periodic variations in the Earths orbital parameters and the response of the climate system remains elusive. One of the key challenges is that of the Mid-Pleistocene Transition (MPT), during which the dominant mode of glacial/interglacial climatic variability shifted without any corresponding change in the mode of orbital forcing. Subtropical climate on orbital time-scales is sensitive to variability in both the low-latitude ocean/atmosphere circulation regime and the global carbon-cycle (through its effect on atmospheric greenhouse gas levels), both of which may have played a role in the shift in mode of global climate response to orbital forcing during the MPT. This thesis presents a series of multi-proxy (foraminiferal stable isotope and trace-metal) paleoceanographic reconstructions from the subtropical southwest Pacific, as seen in marine sediment core MD06-3018, from 2470m water depth and 23ºS in the New Caledonia Trough, southern Coral Sea. The core age-model, based upon magnetic stratigraphy and orbital tuning, yields a mean sedimentation rate at the site of 20mm/ka and a core-bottom age of 1600ka. The MD06-3018 reconstruction of New Caledonia Trough deep water chemistry, based on benthic 13C measurements, shows that the spatial chemistry gradient within the Southern Ocean between deep waters entering the Tasman Sea and the open Pacific was greater during glacial (relative to interglacial) stages over at least the past 1100ka. This gradient was, however, generally reduced on the >100kyr time-scale across the MPT, consistent with it being a period of reduced deep water ventilation in both hemispheres. The MD06-3018 Mg/Ca-derived reconstruction of subtropical southwest Pacific Sea Surface Temperature (SST) shows glacial/interglacial variability of 2-3ºC but no significant trends on the >100kyr time-scale over the duration of the record. An estimate of the uncertainty associated with the SST reconstruction demonstrates that no significant changes in reconstructed southern Coral Sea mean-annual SST can be identified between interglacial stages across the MPT. It is, therefore, unlikely that regional climatic change constituted the main cause for the observed middle Pleistocene expansion of coral reef systems. The >100kyr time-scale stability of southern Coral Sea SST means that the position of the southern boundary of the Pacific warm pool has also been stable over at least the past 1500ka. Comparison with other low-latitude Pacific reconstructions shows that the early Pleistocene warm pool was consequently more hemispherically asymmetric than its present configuration, with the latter being established by ~1000ka and implying significant changes in meridional atmospheric heat and moisture fluxes prior to the MPT. On orbital time-scales, the SST reconstruction shows a clear shift from dominant 40kyr to 100kyr modes of variability over the MPT, although significant 40kyr structure is also retained into the middle/late Pleistocene. In contrast, reconstructed hydrological cycle variability (based on coupled 18O-Mg/Ca measurements) shows only limited coherence with the obliquity cycle and a stronger relationship with the precession cycle. The decoupling of the reconstructed subtropical SST and hydrological cycle responses places constraints on the extent of orbitally paced fluctuations in the low-latitude ocean/atmosphere system. Instead, comparison of the MD06-3018 SST reconstruction with others from across the lowlatitude Pacific supports a dominant role for greenhouse gas forcing in low-latitude western Pacific glacial/interglacial SST variability across the Pleistocene. The subtropical multi-proxy climate reconstructions presented here show that the timing and sense of long-term (>100kyr time-scale) changes in the low-latitude ocean/atmosphere circulation regime are consistent with that system having been important in the expansion of northern hemisphere ice-volume during the early part of the MPT. However, the subtropical reconstructions also suggest that neither the low-latitude ocean/atmosphere circulation system nor the global carbon-cycle underwent a fundamental change in mode of response to orbital forcing during the transition. Instead, the origin of the 100kyr glacial/interglacial mode was most likely related to thresholds in the dynamics of the expanding northern hemisphere icesheets, leading in turn to the existence of significant inter-hemispheric asymmetry in the orbital time-scale climate response over the middle/late Pleistocene. Summary for Non-Specialists. Over the past five million years of its history, the Earths climate has undergone a series of regular, or nearly regular, fluctuations between warmer and colder states. These fluctuations take tens to hundreds of thousands of years to occur and are known as the ‘glacial/interglacial cycles’ on account of the associated changes in ice-sheet extent in the high-latitudes. The origin of these cycles is widely held to be the regular variations in form of the Earths orbit around the sun. In spite of decades of research, however, no complete ‘orbital theory of climate’ exists, mainly because the patterns of past climate variability, as reconstructed using ‘proxies’ for variables such as surface temperature, is much more complex than that of the orbital variations themselves. It follows that processes within the Earth system, especially those associated with large ice-sheets, the carbon-cycle and the ocean circulation system, act to substantially modify the climate response to the orbital variations. Over the past ten years, new observations from both ice-cores and low-latitude marine sediment cores have suggested that the dominant system(s) involved in setting the Earths response to the orbital variations may potentially be the carboncycle and/or the low-latitude ocean/atmosphere circulation regime rather than highlatitude ice-sheet dynamics, as was generally supposed previously. If this new view is correct, it has profound implications for the general sensitivity of the climate to the carbon-cycle on a range of time-scales - making its evaluation a scientific objective of considerable current importance. This thesis presents a series of reconstructions of aspects of climate and carbon-cycle variability for the subtropical southwest Pacific, as based on proxy measurements in a marine sediment core than spans the past 1,600,000 years at around 5000 year resolution. The key focus is on an interval called the ‘Mid- Pleistocene Transition’, during which time the mode of glacial/interglacial variability changed, indicating a fundamental change in one or more aspects of the response to the orbital forcing. The study site is well placed to investigate variability in both the carbon-cycle and low-latitude ocean circulation over the climatic transition as it lies between the Southern Ocean, a key source of carbon-cycle variability and the equatorial Pacific, where the modern El-Niño system arises. By characterizing variability in these systems, the potential role played by both systems in causing the change in mode of glacial/interglacial variability can be evaluated. The key findings of the thesis are that; firstly, changes in the long-term state of the low-latitude ocean circulation system may well have been important for the expansion of northern hemisphere ice-sheets during the early part of the Mid- Pleistocene Transition. Secondly, it provides further support for a close connection between variability in the carbon-cycle and low-latitude climate on orbital timescales but suggests that there is no clear evidence for either system undergoing a fundamental change in sensitivity to the orbital forcing during the transition.