Full article available in PDF version.

Understanding the role of thermohaline circulation in past climate depends on proxy-based reconstructions of deep-water mass composition. A growing body of data indicates that the neodymium (Nd) isotopic composition of fossil fish debris found in deep-sea sediments can be used as a gauge of deep-water mass composition. This paper reviews the basis for the use of fossil fish debris Nd isotopic values as a proxy for ancient deep-water mass composition and then presents an example of how ancient thermohaline circulation patterns are reconstructed using records of fish debris Nd from deep-sea sediment cores. Fish teeth and debris acquire enhanced Nd concentrations (∼100s to ∼1000 ppm) during an early diagenetic reaction at the sediment/water interface. Data published in 2004 by Martin and Scher and presented in this work confirm the assertion that fish teeth and debris record and retain a deep-water mass signal. The widespread stratigraphic and geographic occurrence of fish remains in deep-sea sediments enables construction of much higher-resolution records than previously afforded by Nd isotopic records derived from Fe-Mn crusts. The utility of fish debris Nd isotopic records is apparent when applied to the late Paleocene–early Eocene interval ca. 55 Ma, an interval of time poorly represented in Fe-Mn crust records. Fish debris records from a suite of deep-sea sedimentary sections indicate a mode of thermohaline circulation fundamentally different from the modern–deep waters formed primarily in the Southern Ocean, with no source of deep-water formation in the North Atlantic. This work reinforces the validity and effectiveness of fish debris Nd isotopic records as a tool for reconstructing ancient thermohaline circulation patterns.

Sequential measurements of molluscan radiocarbon are demonstrated to be an effective proxy of seasonal and El Niño–related upwelling variation in coastal Peru. A Trachycardium procerum valve from southern Peru was measured through ontogeny for radiocarbon via accelerator mass spectrometry (AMS) as well as δ 13 C and δ 18 O. A specimen collected in 1984 near Casma, Peru (∼9.30°S) grew before and during the 1982–1983 El Niño/southern oscillation warm event. Shell morphology recorded El Niño warming as a shallow growth break with subsequent realignment of aragonite crystal microstructure. The presence of this growth pattern indicates that shell material was precipitated during the warm event and that each AMS sample could be independently identified to represent a defined period in the El Niño/southern oscillation cycle. Samples taken from portions of the shell precipitated prior to the El Niño warm event (before the diagnostic growth break) had a mean value of 99.8 percent modern carbon (pMC), with a maximum seasonal range of 2.1 pMC. During warming, as indicated by a negative excursion in δ 18 O and the growth break, there was an abrupt increase to 107.9 pMC. Aragonite precipitated near the margin of the valve, after the El Niño/southern oscillation event concluded, had radiocarbon values approaching those present before the growth break. We attribute this radiocarbon distribution to variations in vertical mixing of surface and deeper upwelled water of greater ventilation age. As an El Niño event begins and the thermocline deepens, less deep water reaches the surface. Thus, radiocarbon values in shell precipitated during El Niño appear younger (more positive) relative to non–El Niño periods, which represent periods of more intense upwelling. The results from the modern specimen validate the use of molluscan radiocarbon as a proxy of upwelling conditions related to El Niño/southern oscillation and suggest the utility of similar analysis of more ancient valves in both oceanography and paleoclimatology.

This paper presents a graphical model for high-resolution qualitative reconstructions of environmental conditions (temperature, food availability, and fresh-water influx) from bivalve mollusk shells. The growth rate–temperature model (GRT model) is based on the observation that temperature is the most important control of shell growth as shown by shells of Phacosoma japonicum (Reeve). A highly significant correlation ( p <0.0001) exists between daily shell growth rate of this species and water temperature. Nonparametric bivariate density fields (NBDF) describe the area in which 95% of the growth rate–temperature data are plotted (observed temperatures or temperatures reconstructed from oxygen isotope ratios of the shell). Growth rate–temperature values deviating from the NBDF indicate significant variations in food availability and/or fresh-water influx (salinity changes). The model is applicable to Recent and fossil bivalve mollusk shells. Its strength lies in the combination of sclerochronological (growth rate analysis) and stable isotope analyses. The integration of other environmental proxies (trace and minor elements and elemental ratios) archived in bivalve shells could further improve the GRT model and allow for high-resolution quantitative environmental reconstructions.

Although water balance in terrestrial settings is an important climate parameter, relatively few proxies are available for reconstructing effective moisture. Here, we investigate the possibility of using sulfur speciation as a proxy for paleorainfall in sediments recovered from a large paleolake in the Bogota Basin, Colombia and spanning the past 660 k.y. Samples were digested with a mild acidic solution to extract acid-soluble sulfur minerals. Sulfur species extracted through this procedure included monosulfides, whereas sulfur in the remaining acid-insoluble fraction was found to be bound to organosulfur compounds. Monosulfide/total sulfur ratios range from 0.2 to 0.9 in the studied sediments and exhibit a cyclic distribution with depth. Low (<0.4) monosulfide/total sulfur ratios are characteristic of glacial intervals, whereas interglacial intervals exhibit both low and high ratios. Because partitioning of sulfur between monosulfides and organosulfur compounds depends on iron availability, we interpret that elevated iron delivery to the paleolake occurs at relatively high precipitation rates and results in the sequestration of sulfur in monosulfide minerals. Conversely, incorporation of sulfur into organic matter occurs at low precipitation rates when iron supply is low. Time-series analysis of sulfur ratios reveals the influence of orbital parameters (i.e., eccentricity, obliquity, and precession). Given that rainfall in the tropics is primarily associated with the passage of the Intertropical Convergence Zone, we conclude that orbital forcing exerts a significant control on the intensity or zonality of the Intertropical Convergence Zone that ultimately affects rainfall patterns in the Colombian Andes.

Fluid inclusions trapped in speleothems represent samples of the drip water (dw) from which the speleothems grew. It is uncertain whether the original 18 O/ 16 O ratios may have subsequently been altered due to exchange with the calcite (ct). We can use their deuterium/hydrogen (D/H) ratios to reconstruct δ 18 O of the drip water using the local meteoric water line (MWL) giving the relationship between δ 18 O and δD. We can thus calculate paleotemperatures from the calcite-water isotopic fractionation if the MWL relationship is known. A Holocene speleothem from Santa Cruz, California, displays a large variation in δD, possibly due to changing amount effect, and yields temperatures at 4–3.5 k.y. B.P. −7 °C lower than modern. Two coeval, mid-Wisconsin speleothems deposited tens of meters apart in Reed's Cave, South Dakota, exhibit different δ 18 O ct values but, when combined with δ 18 O dw , we obtain similar growth temperatures. Thus speleothems may display divergent δ 18 O ct records, but still record the same long-term climate variations.

A paleokarst fill deposit from the Wichita Mountains, south-central Oklahoma, United States, consists primarily of sparry calcite, Fe-sulfides, and goethite. Previous cement-stratigraphic studies and paleontological finds suggest that calcite mineralization was initiated during Permian time, whereas goethite and other oxides apparently formed from oxidation of preexisting Fe-sulfides during Pleistocene time. Therefore, these deposits have the potential to offer insight into surficial hydrology and paleoenvironment in an upland setting from two time periods at a single site. δ 13 C PDB and δ 18 O SMOW measurements of 17 samples from growth bands in a single karst-fill calcite crystal range from −10.7‰ to −6.6‰ (mean = −8.6‰) and 27.1‰ to 28.3‰ (mean = 28‰), respectively. Large oscillations in the δ 13 C values through the growth series may originate from seasonal changes in the magnitude of biological productivity during Permian time. These δ 13 C oscillations contrast with the relative stability of the δ 18 O values, which are more positive than would be expected for isotopic equilibrium with local modern waters. The δ 18 O values of the calcite may reflect the δ 18 O values of ambient meteoric groundwaters in the Permian that were isotopically similar to waters in modern, seasonally dry, low-latitude coastal regions. Goethites are not in equilibrium with modern waters or coexisting calcites in the fissure-fill deposit as determined from δ 18 O and δD values of the goethites. Furthermore, the combined δ 18 O and δD values of the goethites are indicative of formation from meteoric waters at a temperature of ∼9 °C ± 3 °C. This inferred temperature is 7 °C ± 3 °C cooler than local modern mean annual temperature and corresponds well with independent studies that propose temperatures ∼6 °C cooler in this region during Pleistocene time. The mole fraction and δ 13 C values of the Fe(CO 3 )OH component in solid solution in the goethite sample are 0.0103 and −10.1‰, respectively. In combination, these values suggest that goethite formed in an environment characterized by mixing of three isotopically distinct CO 2 components: (1) oxidized biological carbon, (2) atmospheric CO 2 , and (3) CO 2 from dissolution of carbonate in the karst system. Oxidized biological carbon may have originated either from flora characterized by C3 or mixed C3:C4 photosynthesizers. Mass balance calculations between these three CO 2 end members correspond to an inferred soil CO 2 concentration [CO 2 contributed from (1) and (2) above] ranging from ∼8,000 ppmV to ∼16,000 ppmV for a local ecosystem dominated by C 3 flora. This inferred range of soil CO 2 concentrations is typical of grasslands characterized by relatively high biological productivity. If C4 flora were a significant source of oxidizing carbon, the higher calculated ambient CO 2 concentration at the time of goethite crystallization in the cave (∼20,000 ppmV) might be interpreted to correspond to an unusually productive C4 soil present at a time of generally cooler and drier conditions across the southern Great Plains of North America.