Coral reefs record glacial cycles
Terraces offshore of several of the Hawaiian Islands are sequences of ancient coral reefs. Reefs can grow fast enough to keep up with the continuous subsidence of the islands, but they “drown” and die when sea level rises suddenly, such as when glaciers melt. Software called “ReefGrow” has been developed as part of our project for modeling this sequence.
The timing of the interglacial periods in the Pleistocene is not well constrained by other paleoclimate records such as ice cores, but the ages of the tops of the reef terraces should coincide with major meltwater pulses as glaciers receded. The reefs off the Big Island date from 15,000 (the shallowest reef) to 400,000 years, and the reefs off Lanai, Oahu, and Niihau record the glacial cycles up to several million years ago. Intriguing, isn’t it, that one goes to the tropics to study glacial cycles?
The paleoenvironments experienced by the fossil reefs provide other information about the history of the islands, such as evidence that giant tsunami hit Lanai in the past, and the tilting history of the flexural arch.
Our research on drowned coral reefs around Hawaii
Impact of mid-Pleistocene climate change on coral reef growth
MAUI – The submarine reef terraces (L1–L12) of the Maui Nui Complex (MNC—the islands of Lanai, Molokai, Maui and Kahoolawe) in Hawaii provide a unique opportunity to investigate the impact of climate and sea level change on coral reef growth by examining changes in reef development through the Mid-Pleistocene Transition (900–800 ka). We present an analysis of the biological and sedimentary composition of the reefs that builds directly on recently published chronological and morphological data. We define nine distinct limestone facies and place them in a spatial and stratigraphic context within 12 reef terraces using ROV and submersible observations. These include oolitic, two coral reef, two coralline algal nodule, algal crust, hemi-pelagic mud, bioclastic and peloidal mud facies. These facies characterise environments from high energy shallow water coral reef crests to low energy non-reefal deep-water settings. Combining the bottom observations and sedimentary facies data, we report a shift in the observed sedimentary facies across the submerged reefs of the MNC from dominant shallow coral reef facies on the deep reefs to coralline algae dominated exposed outcrop morphology on the shallower reefs. We argue that this shift is a reflection of the change in period and amplitude of glacioeustatic sea level cycles (41 kyr and 60–70 m to 100 kyr and 120 m) during the Mid- Pleistocene Transition (MPT, ~800 ka), coupled with a slowing in the subsidence rate of the complex. The growth of stratigraphically thick coral reef units on the deep Pre-MPT reefs was due to the rapid subsidence of the substrate and the shorter, smaller amplitude sea level cycles allowing re-occupation and coral growth on successive cycle low-stands. Longer, larger amplitude sea level cycles after the MPT combined with greater vertical stability at this time produced conditions conducive to deep-water coralline algae growth which veneered the shallower terraces. Additionally, we compare reef development both within the MNC, and between the MNC and Hawaii. Finally we suggest that climatic forcings such as sea-surface temperature and oceanographic currents may also have influenced the distribution of coral species within the sample suite, e.g., the disappearance of the Acropora genus from the Maui Nui Complex in the Middle Pleistocene.
Reference: Faichney, I.D.E., J.M. Webster, D.A. Clague, J.C. Braga, W. Renema, D.C. Potts (2011) The impact of the Mid-Pleistocene Transition on the composition of submerged reefs of the Maui Nui Complex, Hawaii. Palaeogeogr., Palaeoclim., Palaeoecol. 299: 493-506, doi:10.1016/j.palaeo.2010.11.027.
Growth of coral reefs offshore Lanai during the Pleistocene
LANAI – A sequence of submerged terraces (L1–L12) offshore Lanai was previously interpreted as reefal, and correlated with a similar series of reef terraces offshore Hawaii island, whose ages are known to be < 500 ka. We present bathymetric, observational, lithologic and 51 87Sr/86Sr isotopic measurements for the submerged Lanai terraces ranging from − 300 to − 1000 m (L3–L12) that indicate that these terraces are drowned reef systems that grew in shallow coral reef to intermediate and deeper fore-reef slope settings since the early Pleistocene. Age estimates based on 87Sr/86Sr isotopic measurements on corals, coralline algae, echinoids, and bulk sediments, although lacking the precision (~ ± 0.23 Ma) to distinguish the age–depth relationship and drowning times of individual reefs, indicate that the L12–L3 reefs range in age from ~ 1.3–0.5 Ma and are therefore about 0.5–0.8 Ma older than the corresponding reefs around the flanks of Hawaii. These new age data, despite their lack of precision and the influence of later-stage submarine diagenesis on some analyzed corals, clearly revise the previous correlations between the reefs off Lanai and Hawaii. Soon after the end of major shield building (~ 1.3–1.2 Ma), the Lanai reefs initiated growth and went through a period of rapid subsidence and reef drowning associated with glacial/interglacial cycles similar to that experienced by the Hawaii reefs. However, their early Pleistocene initiation means they experienced a longer, more complex growth history than their Hawaii counterparts.
Reference: Webster, J.M., Clague, D.A., Faichney, I.D.E., Fullagar, P.D., Hein, J.R., Moore, J.G., Paull, C.K. (2010) Early Pleistocene origin of reefs around Lanai, Hawaii, Earth Planet. Sci. Lett. (2010), 290(3-4): 331 – 339, doi:10.1016/j.epsl.2009.12.029.
Oolite growth during Younger Dryas in Maui-Nui complex
MAUI – We describe the first recorded occurrence of oolite in the main Hawaiian Islands. Well-cemented oolite was recovered at several locations at −140 to −150 m depth south of Lana’i. The ooids contain foraminiferal, skeletal, and peloidal nuclei, coated by thin to moderately thick (20–50% of ooid radius) tangential cortices, and are cemented by fibrous aragonite needles. The extent of amino acid racemization (AAR) analysed on 127 ooid grains and 55 pristine foraminifera in the same deposits confirm that the particles formed during a brief interval estimated at 1900 yr duration. A single, large benthic foraminiferan (Amphisorus sp.) intermixed with the ooid grains yielded a calibrated 14C age of 12.2 ± 0.3 ka. In contrast, a series of progressive leaches on two splits of ooid grains yielded a sequence of 14C ages ranging from ~ 27 ka for the nuclei to < 15 ka from the outermost cortices, similar to recently reported results of deeply submerged ooids from the Great Barrier Reef, Australia. In contrast, a similar leach experiment using AAR showed no evidence for this 12,000-year range of apparent ages within the ooids. Based on modern analogues, individual ooids are unlikely to have formed continuously during 12 ka of deep submergence without leaving evidence of corrosion or bioerosion. At this time, we have no definitive explanation for the range 14C leach ages, and further research is needed. However, assuming that the age of the ooids lies between the 14C ages of the youngest cortex (14.9 ka) and the Amphisorus (12.2 ka), then eustatic sea-level history records suggest two intervals of slowing sea-level history rise that could accommodate ooid formation. These occurred immediately preceding melt-water pulse (MWP) 1A (~ 15 ka) and MWP1B (~ 12.5 ka). In the area of the oolite deposit, there is no terrace or hardground morphology that coincides with a −100 m sea stand preceding MWP1A. Thus, we suggest that conditions required for ooid formation converged when a flat carbonate hardground in the broad Au’au Channel was initially flooded with energetic shallow water as sea level rose to and slowed at −60 to −55 m during the onset of the Younger Dryas interval (return of glacial conditions) around 12.7 ka. Ooid formation ensued for a short period (~ 1900 yr) and then abruptly ceased; most reasonably as a result of a 7.5 m deepening of the sea during MWP1B around 11.5 ka. The oolitic sediments were subsequently transported longshore and offshore to southern Lana’i.
Reference: Hearty, P.J., Webster, J.M., Clague, D.A., Kaufman, D.S., Bright, J., Southon, J., Renema, W. (2010) A pulse of ooid formation in Maui Nui (Hawaiian Islands) during Termination I, Marine Geology, 268: 152-162, doi:10.1016/j.margeo.2009.11.007.
Evolution of drowned coral reefs in Maui-Nui complex
MAUI – Reef drowning and backstepping have long been recognised as reef responses to sea-level rise on subsiding margins. During the Late Pleistocene (~ 500–14 ka) Hawaiian reefs grew in response to rapid subsidence and 120 m 100 kyr sea-level cycles, with recent work on the submerged drowned reefs around the big island of Hawaii, and in other locations from the last deglacial, providing insight into reef development under these conditions. In contrast, reefs of the Early Pleistocene (~ 1.8–0.8 Ma) remain largely unexplored despite developing in response to significantly different 60–70 m 41 kyr sea-level cycles. The Maui-Nui Complex (MNC — forming the islands of Maui, Molokai, Lanai and Kahoolawe), provides a natural laboratory to study reef evolution throughout this time period as recent data indicate the reefs grew from 1.1 to 0.5 Ma. We use new high resolution bathymetric and backscatter data as well as sub-bottom profiling seismic data and field observations from ROV and submersible dives to make a detailed analysis of reef morphology and structure around the MNC. We focus specifically on the south-central region of the complex that provides the best reef exposure and find that the morphology of the reefs varies both regionally and temporally within this region. Barrier and pinnacle features dominate the steeper margins in the north of the study area whilst broad backstepping of the reefs is observed in the south. Within the Au’au channel in the central region between the islands, closely spaced reef and karst morphology indicates repeated subaerial exposure. We propose that this variation in the morphology and structure of the reefs within the MNC has been controlled by three main factors; the subsidence rate of the complex, the amplitude and period of eustatic sea-level cycles, and the slope and continuity of the basement substrate. We provide a model of reef development within the MNC over the last 1.2 Ma highlighting the effect that the interaction of these factors had on reef morphology.
Reference: Faichney, I.D.E, J.M. Webster, D.A. Clague, C. Kelley, B. Appelgate, J.G. Moore (2009) The morphology and distribution of submerged reefs in the Maui-Nui Complex, Hawaii: New insights into their evolution since the Early Pleistocene, Marine Geology, 265: 130-145. [Abstract]
Coral reefs on subsiding margins are an archive of sea-level and climate changes
PAPUA NEW GUINEA AND HAWAII – A series of well-developed submerged coral reefs are preserved in the Huon Gulf (Papua New Guinea) and around Hawaii. Despite different tectonic settings, both regions have experienced rapid subsidence (2-6 m/kyr) over the last 500 kyr. Rapid subsidence, combined with eustatic sea-level changes, is responsible for repeated drowning and backstepping of coral reefs over this period. Because we can place quantitative constraints on these systems (i.e., reef drowning age, eustatic sea-level changes, subsidence rates, accretion rates, basement substrates, and paleobathymetry), these areas represent unique natural laboratories for exploring the roles of tectonics, reef accretion, and eustatic sea-level changes in controlling the evolution of individual reefs, as well as backstepping of the entire system.
A review of new and existing bathymetric, radiometric, sedimentary facies and numerical modeling data indicate that these reefs have had long, complex growth histories and that they are highly sensitive, recording drowning not only during major deglaciations, but also during high-frequency, small-amplitude interstadial and deglacial meltwater pulse events. Analysis of five generalized sedimentary facies shows that reef drowning is characterized by a distinct biological and sedimentary sequence. Observational and numerical modeling data indicate that on precessional (20 kyr) and sub-orbital timescales, the rate and amplitude of eustatic sea-level changes are critical in controlling initiation, growth, drowning or sub-aerial exposure, subsequent re-initiation, and final drowning. However, over longer timescales (>100-500 kyr) continued tectonic subsidence and basement substrate morphology influence broad scale reef morphology and backstepping geometries. Drilling of these reefs will yield greatly expanded stratigraphic sections compared with similar reefs on slowly subsiding, stable and uplifting margins, and thus they represent a unique archive of sea-level and climate changes, as well as a record of the response of coral reefs to these changes over the last six glacial cycles.
Reference: Webster, J.M., J.C. Braga, D.A. Clague, K. Riker-Coleman, C. Gallup, J.R. Hein, D. Potts, R. Riding, W. Renema, E. Silver, L.M. Wallace (2009) Coral reef evolution on rapidly subsiding margins, Global and Planetary Change, 66: 129-148, doi:10.1016/j.gloplacha.2008.07.010.
Reef development during the last two glacial cycles
HAWAII – Drowned coral reefs on rapidly subsiding margins possess a unique archive of sea level and climate changes, generally unavailable from stable or uplifting margins. Using available field observations and sedimentary, radiometric age, and numerical modeling data, we propose a new model of submerged reef development around Hawaii during the last two glacial cycles (250 kyr). This model provides a quantitative predictive stratigraphy for the reefs that we argue, if drilled, will yield new information on sea level and climate changes, as well as coral reef response over the last 250 kyr. Comparing the observational and numerical modeling data, combined with sensitivity testing, we present our ‘‘best case’’ scenario for the evolution of the drowned lowstand reefs now at -400 (H2) and -150 m (H1). We find that growth rates of 2.5–2.85 m/kyr for the main shallow reef building facies, a subsidence rate of 2.5 m/kyr, and a variable basement substrate configuration best explain the observational data. Modeling of the internal stratigraphic succession of the reefs shows that the number and thickness of shallow reef units, as well as the frequency and duration of subaerial exposure and reef-drowning events, are sensitive to the frequency and amplitude of eustatic sea level variations but not the rate of subaerial erosion. H2 and H1 initiated growth during stable eustatic sea level conditions during highstands circa 222 ka (MIS7) and circa 126 ka (MIS5e), respectively. Both H2 and H1 have a long and complex growth history, growing episodically for ~90 kyr. Precessional (~20 kyr) and higher-frequency, suborbital eustatic sea level fluctuations dominate, with each reef experiencing repeated but brief (<5–10 kyr) drowning and subaerial exposure, producing a complex layer cake stratigraphy of shallow (0–30 m) coral reef units separated by either subaerial exposure horizons or thin, intermediate (30–60 m) coralgal units. Final drowning of H2 and H1 occurs during the penultimate (133–134 ka) and last deglaciation (12–14 ka). These findings are consistent with available age data and qualitative predictions of previous studies around Hawaii.
Reference: Webster, J.M., L.M. Wallace, D.A. Clague, J.C. Braga (2007) Numerical modeling of the growth and drowning of Hawaiian coral reefs during the last two glacial cycles (0–250 kyr), Geochem. Geophys. Geosyst., 8, Q03011, doi:10.1029/2006GC001415.
Diagenesis affects geochemistry of coral skeletons and requires careful analytical techniques
HAWAII – The geochemistry of coral skeletons may reflect seawater conditions at the time of deposition and the analysis of fossil skeletons offers a method to reconstruct past climate. However the precipitation of cements in the primary coral skeleton during diagenesis may significantly affect bulk skeletal geochemistry. We used secondary ion mass spectrometry (SIMS) to measure Sr, Mg, B, U and Ba concentrations in primary coral aragonite and aragonite and calcite cements in fossil Porites corals from submerged reefs around the Hawaiian Islands. Cement and primary coral geochemistry were significantly different in all corals. We estimate the effects of cement inclusion on climate estimates from drilled coral samples, which combine cements and primary coral aragonite. Secondary 1% calcite or ~2% aragonite cement contamination significantly affects Sr/Ca SST estimates by +1 °C and −0.4 to −0.9 °C, respectively. Cement inclusion also significantly affects Mg/Ca, B/Ca and U/Ca SST estimates in some corals. X-ray diffraction (XRD) will not detect secondary aragonite cements and significant calcite contamination may be below the limit of detection (~1%) of the technique. Thorough petrographic examination of fossils is therefore essential to confirm that they are pristine before bulk drilled samples are analysed. To confirm that the geochemistry of the original coral structures is not affected by the precipitation of cements in adjacent pore spaces we analysed the primary coral aragonite in cemented and uncemented areas of the skeleton. Sr/Ca, B/Ca and U/Ca of primary coral aragonite is not affected by the presence of cements in adjacent interskeletal pore spaces, i.e. the coral structures maintain their original composition and selective SIMS analysis of these structures offers a route to the reconstruction of accurate SSTs from altered coral skeletons. However, Mg/Ca and Ba/Ca of primary coral aragonite are significantly higher in parts of skeletons infilled with high Mg calcite cement. We hypothesise this reflects cement infilling of intraskeletal pore spaces in the primary coral structure.
Reference: Allison, N., A.A. Finch, J.M. Webster, D.A. Clague (2007) Palaeoenvironmental records from fossil corals: The effects of submarine diagenesis on temperature and climate estimates, Geochimica et Cosmochimica Acta, 71, 4693-4703, doi:10.1016/j.gca.2007.07.026.
Evidence of large tsunami on Lanai
LANAI – The origin of subaerial coral conglomerate deposits on the Hawaiian islands of Lanai and Molokai is controversial, primarily because these deposits are difficult to interpret and the vertical motion of these islands is poorly constrained. Based on bathymetry, dive observations, sedimentary and radiocarbon data from coralline algal dominated deposits from two submerged terraces at –150 and –230 m off Lanai, Lanai has experienced relatively little vertical movement over the last 30 ka. Using internally consistent age versus depth relationships, paleowater depths, and published sea level data, we estimate that Lanai has experienced maximum rates of uplift of 0.1 m/kyr or subsidence of 0.4 m/kyr over this period. Our analysis of possible uplift mechanisms, published geophysical, numerical modelling, and recent tide data suggests that this is also the maximum uplift rate for the last several hundred thousand years. Taken together these data support the interpretation that coral conglomerates at elevations higher than +35 m on Lanai are tsunami deposits with a minimum wave run up > 170 m, rather than shoreline deposits formed during the last two interglacials, then uplifted to their present elevations.
Reference: Webster, J.M., D.A. Clague, J.C. Braga (2007) Support for the Giant Wave Hypothesis: evidence from submerged terraces off Lanai, Hawaii, Int J Earth Sci, 96(3): 517-524, doi:10.1007/s00531-006-0107-5.
Coralline algal deposits off Lanai indicate little vertical movement
LANAI – We present detailed bathymetry, remotely operated vehicle (ROV) and submersible observations, and sedimentary and radiocarbon age data from carbonate deposits recovered from two submerged terraces at -150m (T1) and -230m (T2) off Lanai, Hawaii. The tops of the terraces are veneered by relatively thin (<5 m) in situ accumulations of coralline algal nodule, coralgal nodule, Halimeda and a derived oolitic facies deposited in intermediate (30-60 m) to deep fore-reef slope settings (60-120 m). The data are used to develop a sedimentary facies model that is consistent with eustatic sea-level variations over the last 30 thousand years (ka). Both nodule facies on T1 and T2 initiated growth 30-29 ka following a fall in sea level of ~50 m and increase in bottom currents during the transition from Marine Isotope Stage 3 to 2. The nodules accreted slowly throughout the Last Glacial Maximum when sea-level was relatively stable. Drowning occurred during the early deglaciation (17-16 ka) and was marked by the complete drowning of coralline algal nodules facies on T2 and incipient drowning of coralgal facies on T1. Abrupt sea-level rise during the middle deglaciation, perhaps associated with global meltwater pulse 1A (14-15 ka), finally drowned the coralgal facies on T1, which in turn was overlain by a deep-water Halimeda facies or an oolitic facies derived from upslope. Our data indicates that Lanai has experienced relatively little vertical tectonic movement over the last 30 ka. Using paleobathymetric data derived from the sedimentary facies, age vs. depth relationships, and published sea-level curves, we estimate that Lanai could be either slowly uplifting or subsiding, but at rates <0.1 m/kyr (uplift) or <0.4 m/kyr (subsidence) over this 30 kyr period.
Reference: Webster, J.M., D.A. Clague, J.C. Braga, H. Spalding, W. Renema, C. Kelley, B. Applegate, J.R. Smith, C.K. Paull, J.G. Moore, D. Potts (2006) Drowned coralline algal dominated deposits off Lanai, Hawaii: carbonate accretion and vertical tectonics over the last 30 ka. Marine Geology 225: 223-246, doi: 10.1016/j.margeo.2005.08.002.
Drowning of coral reef associated with melting of glaciers
HAWAII – Sea-level rise was not smooth and continuous when the glaciers melted, and we propose that it was during the meltwater pulses, periods of dramatic, unusually fast rise of sea-level, that the reef-building corals could not grow vertically fast enough and died. We present evidence that the drowning of the -150 m coral reef around Hawaii was caused by rapid sea-level rise associated with meltwater pulse 1A (MWP-1A) during the last deglaciation, when sea-level rose about 35 meters in less than 500 years.
New U/Th and 14C accelerator mass spectrometry dates, combined with reinterpretation of existing radiometric dates, constrain the age of the coral reef to 15.2-14.7 ka (U/Th age), indicating that reef growth persisted for 4.3 thousand years following the end of the Last Glacial Maximum at 19 ka. The drowning age of the reef is roughly synchronous with the onset of MWP-1A between 14.7 and 14.2 thousand years ago. Dates from coralline algal material range from 14 to 10 cal ka (calibrated radiocarbon age), 1-4 thousand years younger than the coral ages. A paleoenvironmental reconstruction incorporating all available radiometric dates, high-resolution bathymetry, dive observations, and coralgal paleobathymetry data indicates a dramatic rise in sea level around Hawaii ca. 14.7 thousand years ago. Paleowater depths over the reef crest increased rapidly above a critical depth (30-40 m), drowning the shallow reef-building Porites corals and causing a shift to deep-water coralline algal growth, preserved as a crust on the drowned reef crest.
Reference: Webster, J.M., D.A. Clague, K. Riker-Coleman, C. Gallup, J.C. Braga, D. Potts, J.G. Moore, E.L. Winterer, and C.K. Paull (2004) Drowning of the -150m reef off Hawaii: a casualty of global meltwater pulse 1A? Geology, 32(3): 249-252. [Abstract]