PP41B-01 INVITED 08:00h
Climatic Calibration of Paleoecological Data and Data/Model Comparisons
When accurately and precisely dated, paleoecological data along with geochemical data and geomorphological features indicate past changes in climate. Calibrating these data in terms of standard climatic (e.g. July mean temperature) or bioclimatic (e.g. winter extreme temperatures) variables has allowed construction of paleoclimatic maps and time series that not only show the magnitude and extent of the difference in past climates but also can be compared to the output from global climate and earth-system models. Key to judging the reliability of the derived paleoclimatic estimates is the qualitative and quantitative comparison of estimates from different paleoclimatic indicators to show whether consistent or discrepant patterns exist. Here "multi-proxy" studies of independent paleoclimatic indicators from single cores, single basins, and multiple sites are providing many new opportunities to test calibration methods and results. The multi-proxy studies are also allowing researchers to develop a multivariate view of climate because different types of data, whether paleoecological or geochemical, often are sensitive to different aspects of climate change. These can be temperature changes and/or moisture changes as well as changes in seasonality or climatic extremes. Recent research mapping late Quaternary pollen, geochemical, and lake-level data across eastern North America in 1000-yr intervals illustrates some the advances gained from this multi-proxy approach to the reconstruction of past climates since the last glacial maximum 21,000 years ago.
PP41B-02 08:15h
Evaluating Climatic Controls of Proxy Indicators: A Calibration Study Using Annually Laminated Sediments From a Minnesota Lake
The reliability of paleoclimatic inferences from sedimentary proxy data depends on a solid understanding of what environmental factors control the proxy indicator. Despite the rapid accumulation of paleoclimatic data, few rigorous calibration studies exist that compare proxy records with instrumental weather data. We analyzed annually laminated sediments of the past 100 years from Steel Lake (46o58' N, 94o41' W), Minnesota, for a suite of commonly used climatic proxies, including varve thickness (VT), grey-scale intensity (GSI), oxygen and carbon isotopic compositions (d18O and d13C), biogenic silica (BSi), inorganic carbon (IC), and organic carbon (OC). These proxy data were compared with time series of weather variables, including precipitation (P), temperature (T), P minus evapotranspiration (P-ET), and the Palmer drought severity index (PDSI), as well as with indices of the climate modes thought to influence the regional climate, especially the Pacific Decadal Oscillation (PDO) and the Atlantic Multidecadal Oscillation (AMO). We confirmed the annual nature of laminations using radiocesium (137Cs) dating and petrographic thin sections. The time series of d18O and the mean annual T share little similarity, but d18O covaries with both effective moisture (based on P-ET or PDSI) and PDO, with high d18O values generally corresponding to low P-ET and warm phase PDO. During the warm PDO phase, the climate in the midwestern United States is generally warm and dry (low P-ET) especially during the winter months, which would elevate d18O. Varves are typically thick during the period of low d18O and high P-ET, suggesting that thick varves tend to form under wet weather conditions that promote sediment deposition through increased production and/or formation of amorphous iron minerals. The accumulation rates of various sediment components (e.g., BSi, IC) show similar trends to VT. The time series of GSI and weather parameters share little similarity. d13C does not covary with d18O, but it exhibits trends similar to those of %BSi, the inverse of %IC, and to a less extent, %OC. These proxies are likely influenced by lake productivity, sediment diagenesis, and land-use history, which may compromise their comparisons with weather time series. Our calibration results demonstrate the utility of some of the proxies (especially d18O and VT) for reconstructing drought and large-scale climatic modes, but they also illustrate that not all the proxies commonly used for paleoclimatic reconstructions are climatically sensitive. In conjunction with the findings of several other recent studies, our data suggest that the controlling factors of a proxy indicator are region and potentially site specific, which further underscores the importance of proxy calibration.
PP41B-03 08:30h
Using Multiproxy Influx Data to Clarify the Origin of Carbon-Isotope Signals in Organic Matter from Late Quaternary Lake Sediments: Application to an Altitudinal Transect of Lakes on Mt. Kenya, East Africa
Considerable ambiguities arise in the interpretation of both bulk and molecular stable-isotope analyses from tropical lake sediments, due to past variations in the relative importance of different local organic-matter sources, including C3 and C4 terrestrial plants and emergent macrophytes, submerged aquatic macrophytes, algae, and bacteria, as well as atmospheric particulates. Although comparisons are often made between isotope curves and other sedimentological, geochemical, or palaeoecological indicators, these usually rely on concentration or percentage data, which may be biased by large temporal fluctuations in sedimentation rate, bulk density, and proportions of major sediment components or common fossil taxa. It is also difficult to make interlake comparisons by this means. We present new insights into the 13C records from four small lakes at 2350 - 4595m a.s.l. on Mt. Kenya, spanning the period 38,000 - 0 cal. yr BP, based on the accumulation rates (influxes) of an array of sediment components, including total sediment, magnetic minerals, TOC, TN, n-alkyl lipids, pollen, charred grass fragments, diatoms, diatom C and N, and green microalgae. The results suggest that Sacred Lake (2350m a.s.l.) has always been oligotrophic and dominated by organic-matter inputs from surrounding rainforest and aquatic macrophytes, whereas the higher algal productivity of lakes above treeline ($>$3000m a.s.l.) was stimulated by nutrients transported by periglacial activity from surrounding slopes, and aeolian inputs of dust and organic detritus. Comparisons between lakes help to distinguish between isotopic coherence resulting from shared atmospheric forcing, and purely local signals reflecting differences in lake depth, hydrology, vegetation, and geomorphological processes.
PP41B-04 08:45h
Terrestrial Plant Biomarkers Preserved in Cariaco Basin Sediments: Records of Abrupt Tropical Vegetation Response to Rapid Climate Changes
Organic-rich sediments from the anoxic Cariaco Basin, Venezuela, preserve high concentrations of biomarkers for reconstruction of terrestrial environmental conditions. Molecular-level investigations of organic compounds provide a valuable tool for extracting terrestrial signals from these annually laminated marine sediments. Differences in hydrogen isotopic fractionation between C$_{16-18}$ and C$_{24-30}$ {\it n}-alkanoic acids suggest a marine source for the shorter chain lengths and a terrestrial source for the longer chains. Records of carbon and hydrogen isotopes, as well as average carbon chain length (ACL), from long-chain {\it n}-alkanoic acids parallel millennial-scale changes in vegetation and climate between the late Glacial and Preboreal periods, 15,000 to 10,000 years ago. Data from all terrestrial chain lengths were combined to produce single $\delta$D and $\delta ^{13}$C indices through deglaciation, exhibiting enrichment during the late Glacial and Younger Dryas and depletion during the B$\o$lling-Aller$\o$d and Preboreal periods. $\delta$D reflects the hydrogen isotopic composition of environmental water used for plant growth, combined with evaporative enrichment within leaf spaces, and as such may act as a proxy for local aridity. Leaf wax $\delta ^{13}$C, which is a proxy for C$_{3}$ versus C$_{4}$ metabolic pathways, indicates that C$_{3}$ plants predominated in the Cariaco watershed during warm/wet B$\o$lling-Aller$\o$d and Holocene periods, and C$_{4}$ plant biomass proliferated during cool/dry Glacial and Younger Dryas intervals. Coupled carbon and hydrogen isotopic measurements together clearly distinguish deglacial climatic periods as wetter with C$_{3}$ vegetation versus drier with C$_{4}$ vegetation. High resolution biomarker records reveal the rapidity of vegetation changes in northern South America during the last deglaciation. The leaf wax data reveal that local vegetation biomass, although not necessarily entire assemblages, shifted between arid grassland and wetter forest taxa on timescales of decades. Comparison of ACL versus $\delta ^{13}$C for Cariaco Basin and NW African leaf waxes indicate that biomarkers reflect real changes in local South American vegetation and not contamination from long-distance transport during cold windy climates. The precise temporal relationship between tropical vegetation shifts and climate changes is measured by direct comparison of terrestrial vegetation and climate proxies from the same core. Abrupt deglacial climate shifts in tropical and high-latitude North Atlantic regions were synchronous, whereas changes in tropical vegetation consistently lagged climate shifts by several decades.
PP41B-05 09:00h
Seasonally Resolved Oxygen Isotope Paleoclimate Proxy in Tree-Ring Cellulose from the Southeastern U.S.
Stable isotopes in precipitation reflect changes in climate, moisture source, and extreme events such as tropical cyclones, and an oxygen isotope proxy record of these changes through time and space is preserved in tree-ring cellulose. Extreme climate events such as droughts and hurricanes are formidable natural disasters in the southeastern United States, and considerable efforts have been made to understand factors controlling their frequency, whether natural or anthropogenic. Tree rings offer an unusually well-resolved, dateable record of climate events extending beyond modern or historical (documentary) records. Oxygen isotopes in alpha-cellulose of shallowly-rooted conifers predominately reflect the composition of precipitation. Tropical storm convection results in marked $^{18}$O depletion in storm precipitation, to -15$\permil$ relative to source seawater (~0$\permil$). The depletion increases towards the eyewall of the cyclone, however, isotopically depleted precipitation may extend outward many 100's of km. Storm water $^{18}$O depletion translates to soil water $^{18}$O depletion that may persist for many weeks until ameliorated by soil water evaporation. Tree growth during that time will take up the anomalous isotopic compositions. Distinctive earlywood (EW ~March-June) versus latewood (LW ~July-October) growth allows the rings to be resolved at an intra-annual (seasonal) scale. By comparison to average soil water, droughts result in $^{18}$O-enriched soil water compositions. Seasonal drought or years of continued drought will be similarly captured in the isotope compositions of tree-ring cellulose. A 227-year (1770-1997) seasonally-resolved record of tropical cyclone and drought activity was obtained from cross-sections of felled slash pines ({\it Pinus elliottii} Engelm.) and remnant longleaf pines ({\it Pinus palustris} Mill.) from southern Georgia. Interpretations of drought or hurricane events were tested by comparison with recent, detailed meteorological records. The 227-year record reveals most previously established hurricane events, including Florence (1953) and the Great Hurricane of 1780. Newly recognized tropical storms such as 1857 are also evident. Significant seasonal droughts such as 1955, 1927, 1904 and 1896, are observed for southeastern Georgia. Larger-scale climate oscillations appear to overprint the EW and LW isotope series, displaying periods of relatively large or small differences in EW and LW $\delta$$^{18}$O values. The oscillations are interpreted to reflect dominant climate modes that influence moisture source or seasonal temperature variation. The tree-ring record potentially extends many centuries. A preliminary record through a portion of the North American "Little Ice Age" (1580-1650) indicates a significant reduction in tropical cyclone activity.
PP41B-06 09:15h
Lignin Phenol and $\delta^{13}$C Evidence of Quaternary Vegetation Change in South America
The very low CO$_{2}$/O$_{2}$ ratio at the last glacial maximum would have been physiologically challenging for C$_{3}$ land plants by enhancing photorespiration and water stress (Cerling {\it et al}., 1997). Strong palaeoclimatic gradients, along a transect from Venezuela to NE Chile, make it possible to separate the effects of lower pCO$_{2}$ (global) and variations in precipitation (regional). All lake sites lie below the upper limit of C$_{4}$ graminoids in the Andes ($\sim$ 4200m a.s.l) and should have been sensitive to past increases in C$_{4}$ biomass resulting from lower atmospheric CO$_{2}$ and/or decreased summer-monsoon rainfall. Lakes situated at the southern end of the transect exhibit lower $\delta^{13}$C values during the last glacial, when periods of moister climate and higher lake levels occurred, and higher values during the Holocene. In contrast, lakes at the northern end exhibit an opposite temporal trend, with higher isotope values during the Late Pleistocene and lower values during the Holocene. During the last glacial, lignin phenol distributions suggest a dominance of woody plants at the southern end of the transect and a dominance of grasses at the northern end. The opposite occurred during the Holocene, with grasses dominant in the south and woody plants in the north. We conclude that glacial / interglacial variations in precipitation amount and seasonality, driven mainly by orbital forcing, had a greater impact on the distribution of C$_{3}$ and C$_{4}$ biomass in the tropical Americas than low glacial-age CO$_{2}$, which on its own would have favoured the spread of C$_{4}$ plants. Cerling T.E. {\it et al}. (1997) Nature 153-158.
PP41B-07 09:30h
A Significant Thirteenth-century CO$_{2}$ Increase in Stomatal Frequency an ice core Records
The climate of the first half of the last millennium is characterized by a transition from a relative warm period during medieval times (the Medieval Climatic Optimum) towards a more cooler period during the Little Ice Age (~AD 1300-1850). In several global and Northern-Hemispheric air-temperature reconstructions for the last millennium, this natural climate variability is represented by an air-temperature anomaly in the range of 0.2 to 1 $^{O}$C. In contrast to the strongly increased atmospheric CO$_{2}$ levels of the last century, ice-core CO$_{2}$ measurements constrain the pre-industrial CO$_{2}$ variability to a maximum of 12 ppmv, precluding a significant role for CO$_{2}$ as the primary forcing factor of air-temperature changes during the last millennium. As an alternative to ice-core measurements, atmospheric CO2 reconstructions are currently available for the last millennium from stomatal frequency analysis performed on fossil leaves.A period where both methods consistently provide evidence for natural CO$_{2}$ changes is the 13th century.The results of the two independent methods differ significantly in the amplitude of the estimated CO$_{2}$ changes (10 ppmv ice, versus 34 ppmv stomatal frequency). Here, we compare stomatal frequency and ice core results by using a firn-diffusion model in order to assess the potential influence of smoothing during enclosure on the temporal resolution as well as the CO$_{2}$ mixing ratios. The seemingly large discrepancies between the CO$_{2}$ levels estimated by the contrasting methods, diminish when effects of natural smoothing of the ice-core record is simulated for the raw data of the stomatal frequency record. Results indicate that the differences derived by the two methods may be less significant than previously thought. Climate model calculations show that the 34 ppmv shift as detected in the stomata record could generate Northern-Hemisphere and global air-temperature responses that stay well within the constrained range of reconstructed air-temperature variability of the last millennium. To assess the exact influence of a dynamic CO$_{2}$ regime on the air-temperature history of the last millennium, model studies simulating the air-temperature response to fluctuating CO$_{2}$ levels in interaction with the other forcing factors are needed. However, this study already indicates that CO2 should be reconsidered as a significant pre-industrial climate forcing factor during the last millennium.
PP41B-08 09:45h
Stable Isotopic Variations in Columnar Cacti: are Responses to Climate Recorded in Spines?
The behavior of the North American monsoon (NAM), particularly with respect to times of continental drought and its relationship to the Pacific-North American (PNA) teleconnection pattern and the El Nino/Southern Oscillation (ENSO) is of great interest to paleoclimatologists and water managers. Long-term instrumental precipitation and tree ring records in the southwestern United States and northwestern Mexico at low elevations are sparse and this has hindered research on NAM variability at interannual timescales. Saguaro cacti (Carnegiea gigantea) and other columnar cacti in North and South America are long-lived and have the potential to record climate variability on land with high temporal and spatial resolution. The vertical sequence of spines on the saguaro's exterior represents a high resolution (4 to 6 per year), and long (over 150 years) record of environmental change. We present results from an experiment where we tracked the oxygen isotopic values in the source waters, stem tissue waters and spine tissue for three treatments over the course of three months. These data are then compared to a previously developed mechanistic model of isotopic variation that reflects the physiological responses of Saguaro to climate variation over seasonal to century long time-scales. We also present the rationale for a new method to determine the growth rate of columnar cacti using the radiocarbon bomb spike. Our measurements reveal that oxygen and hydrogen isotopic variation among the sequentially produced and persistent spines covering the saguaro body record fluctuations in saguaro water balance. The model successfully predicts isotopic variation in spines and constrains controlling variables, yielding a powerful and high-resolution stable isotope index of water stress in the low desert. The development and refinement of an isotopic model for saguaro will serve as the basis for models applied to other species of columnar cacti in North and South America. The role of the tropics in global climate change is poorly understood and precise chronologies of tropical climate change are needed to place empirical constraints on competing theories and models. In particular, the use of continental records from columnar cacti in South America could identify ENSO periods in the last century and provide empirical constraints on the inputs of Atlantic (monsoonal) versus Pacific (winter) moisture to the Altiplano during ENSO and other important climatological phenomena.