2
The Nature and Value of Geohistorical Information

THE RATIONALE OF GEOHISTORICAL ANALYSIS

The study of ecological dynamics would benefit greatly from a tripartite research effort similar to that used so successfully in climate and ocean dynamics (see Figure 2.1), where information about past systems—their variability in rates, states, and composition; apparent boundary conditions (including those without modern analogs); spatial scaling effects; correlated relationships evident only over periods longer than modern monitoring programs—has proven to be an essential partner to modeling and studies of present day conditions. In fact, it is difficult to imagine understanding climate and ocean systems as well as we do today without the insights gained from geohistorical data. In the study of climate and ocean dynamics, geohistorical records—including ice cores, sediment cores, tree rings, and coral skeletons—have provided the first evidence for the existence of many important phenomena, including climate flickers and other abrupt climate change, the reorganization of deepwater circulation patterns, and Milankovitch oscillations in temperature and atmospheric composition. Geohistorical records also have provided the impetus and an essential empirical framework for modeling and experiments (see Box 2.1). The intellectual impact of individuals and projects that cross methodological interfaces—using some combination of geohistorical data, modeling, and present day observation and experimentation—has been at least as important as original research within subdisciplines in propelling our understanding of these complex environmental systems, including



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change 2 The Nature and Value of Geohistorical Information THE RATIONALE OF GEOHISTORICAL ANALYSIS The study of ecological dynamics would benefit greatly from a tripartite research effort similar to that used so successfully in climate and ocean dynamics (see Figure 2.1), where information about past systems—their variability in rates, states, and composition; apparent boundary conditions (including those without modern analogs); spatial scaling effects; correlated relationships evident only over periods longer than modern monitoring programs—has proven to be an essential partner to modeling and studies of present day conditions. In fact, it is difficult to imagine understanding climate and ocean systems as well as we do today without the insights gained from geohistorical data. In the study of climate and ocean dynamics, geohistorical records—including ice cores, sediment cores, tree rings, and coral skeletons—have provided the first evidence for the existence of many important phenomena, including climate flickers and other abrupt climate change, the reorganization of deepwater circulation patterns, and Milankovitch oscillations in temperature and atmospheric composition. Geohistorical records also have provided the impetus and an essential empirical framework for modeling and experiments (see Box 2.1). The intellectual impact of individuals and projects that cross methodological interfaces—using some combination of geohistorical data, modeling, and present day observation and experimentation—has been at least as important as original research within subdisciplines in propelling our understanding of these complex environmental systems, including

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change FIGURE 2.1 Schematic illustration showing how interactions among geohistorical data, observations and experiments, and modeling serve to increase understanding of the biotic response to environmental change. Until recently, geohistorical data have not been used extensively by biologists. BOX 2.1 CO2 Through Geologic Time: Observations, Models, and Geohistorical Data Royer et al. (2004) illustrated the strength of model-data comparisons in their analysis of the role of CO2 in driving climate during the past 600 Ma. How CO2 functions as a greenhouse gas is now well established through observations, experiments, and modeling of the present day atmosphere as well as the analysis of climatic proxies and CO2 in the Quaternary geohistorical record. However, substantiating CO2’s role in the regulation of pre-Quaternary climate is difficult because of the lack of direct information on the composition of the atmosphere through most of geologic time. Geochemical models, based primarily on sedimentary weathering rates, have been devised to predict the evolution of CO2 through geologic time (e.g., Berner, 2004; Berner and Kothavala, 2001). These predictions can be evaluated by comparing the model results to estimates of past CO2 based on geochemical and paleobotanical indicators in the rock record. Such indicators are said to be “proxy indicators,” in that they substitute for direct measurements. The development and calibration of these and other proxy indicators depends on observations and experiments with modern systems. Royer et al.’s (2004) review of proxy records indicates a consistent relationship between various CO2 indicators, geologic evidence of glaciations, and CO2 predictions from geochemical models. Accordingly, model-proxy comparisons support the hypothesis that CO2 has acted as the primary driver of climate during the past 600 Ma, identifies key time intervals in which proxy indicators and model predictions do not agree, and evaluates the extent to which proxy indicators differ.

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change Comparison of atmospheric CO2 composition predicted by GEOCARB III model calculations (range shown in pink) and proxy indicators. A. Individual proxy indicators (paleosols = fossil soil δ13C; phytoplankton = δ13C of phytoplankton remains; Stomata = density of stomatal pores in leaves of plants; Boron = δ11B in remains of planktonic foraminifera. B. Mean (black line) and ±1σ (gray bars) of proxy estimates in 10 Ma time steps. C. Frequency distribution of proxy data through time. SOURCE: Royer et al. (2004); used with permission.

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change the especially critical issue of the relative roles of anthropogenic and non-anthropogenic processes in causing biotic change. The biosphere is an extremely complex system—analogous to climate and ocean systems—with great diversity and variability in the behavior of its components, an enormous variety of interactions and feedbacks that are context specific and probabilistic, and a long memory and residence time for many of its components. Despite this complexity, improved techniques for extracting information on the environments, biotas, and ages of geohistorical records now combine to ensure that geohistorical analysis is well positioned to become a full and essential partner in the analysis of ecological dynamics. Here, following a brief discussion of strategies for working with incomplete records, we review some of the techniques used to acquire geohistorical data on ecological dynamics, organizing the discussion around questions that are commonly raised by workers unfamiliar with this topic: With what detail and confidence can environmental conditions be inferred for the past? What kinds of biological information are obtainable from the geologic record? How precisely and accurately can the ages of past records be determined? COMPLETENESS OF THE GEOLOGIC RECORD Taken as a whole, the geological record is incomplete—not every individual or species that ever lived is preserved, all the environmental conditions for every part of the globe are not recorded in the sediments, and deposition of rocks has not been continuous everywhere on the face of the earth and throughout geologic time. This heterogeneity in the quality of geohistorical data is a fundamental property of the geologic record, just as completeness and bias in written documents and instrumental records are properties that must be considered in the analysis of human cultural history and instrument-based environmental histories. Research over the last ~20 years (e.g., Kidwell and Flessa, 1995; Foote and Raup, 1996; Koch, 1998; Foote et al., 1999; Foote and Sepkoski, 1999; Behrensmeyer et al., 2000; Briggs et al., 2000; Alroy et al., 2001; Kidwell and Holland, 2002) has demonstrated that uniquely valuable data can be extracted from geohistorical records of diverse quality, thus increasing the effective availability of geohistorical information for biosphere analysis. Some of the key elements of this modern strategy, and the stereotypes that are being overturned, are summarized below.

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change Valuable insights can be gained from sparsely sampled time series and from isolated samples that are well positioned before and after significant biotic or environmental events or transitions. Although long, closely sampled time series may provide important geohistorical information, they are not always required. For example, ecological monitoring programs for modern environments may not have begun before human impact. Under these circumstances, geohistorical records of known age—but not directly connected to the present by a continuous series of comparable samples—are unique and valuable sources of information on local pre-monitoring conditions (e.g., the species composition or dominance structure of a benthic or forest community before an environmental change or an extinction bottleneck, or the typical body sizes of animals before commercial extraction or exploitation). This strategy of opportunism—using such records for their unique insights and aggressively seeking out geohistorical records from previously little-explored areas—has proved remarkably productive in bringing valuable geohistorical perspective on modern management issues in both terrestrial (e.g., Steadman, 1995; Swetnam et al., 1999) and marine settings (Kowalewski et al., 2000; J.B.C. Jackson et al., 2001; Pandolfi et al., 2003; Aronson et al., 2003). Similarly, biotic response to events in the deeper past can be analyzed in the absence of complete time series by using geohistorical records that are appropriately positioned in time or space, for example by comparing samples from before and after mass extinctions (e.g., Raup and Jablonski, 1993; Jablonski and Raup, 1995; Jablonski, 1995; Erwin, 1996; Foote, 2003). In fact, widely spaced samples over longer periods can be superior to closely spaced samples over shorter time series for the analysis of phenomena that occur over periods greater than a few years, and indeed many important biological processes unfold on timescales of millennia or even millions of years. Paleontologic samples of relatively coarse scale (i.e., millions of years) have revealed striking patterns in the loss and subsequent recovery of morphological richness, a measure of biodiversity that may capture functional variety far more effectively than simple counts of higher taxa (e.g., Foote, 1997, 1999; Lockwood, 2003). For many purposes, time-averaged (time-exposure) information—as is typically captured by sedimentary records—is adequate to the questions being addressed and may be better than instantaneous snapshot samples. Few geohistorical samples represent instants in time (snapshots) comparable to ecological samples but instead represent accumulations of individuals and species over some period (time exposures, or time averaged). Such time-averaged samples tend to filter out short-term (seasons to centuries) variation in composition, relative abundance, and diversity. If such samples are large enough to encompass the full range of variability (e.g., Olszewski [1999] indicates that samples of as few as 29

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change individuals may be sufficient), they can be employed to characterize long-term, average conditions (Kowalewski et al., 1998; Olszewski, 1999; Hadly, 1999). In addition, if the time-averaged specimens can be directly dated (e.g., by radiocarbon dating or amino acid racemization), then a continuous temporal record of specimens can be reconstructed (Kowalewski et al., 1998). When the goal is an understanding of larger-scale trends and general principles relevant to developing realistic conservation and management policies, time-averaged geohistorical data provide essential biological data. For example, the richness and dominance structure of most communities are inherited in part from periods predating scientific study. Dominant species in some communities may be relict populations whose dominance reflects incumbency and lifespan rather than present day conditions. In such situations, a one-time or very short-term sample could reflect ephemeral conditions, whereas a time-averaged sample is more likely to reflect average conditions of local diversity and structure. Because standing diversity may be the product of long-term as well as short-term processes, and because local populations are highly volatile, combining data from both living communities and geohistorical records allows tests of hypotheses, provides strength to the interpretations, and supplies new insights into ecological processes. The need for truly long-term geohistorical perspectives on species is especially critical in understanding extinction risks and the potential for speciation and other evolutionary innovation. Raw rates of speciation and extinction (numbers of taxa per unit time), as well as variability in raw rates and the frequency distributions of evolutionary durations, can all only be determined using the fossil record of evolutionary first and last appearances. Without such empirical information, rates of speciation, for example, would have to be assumed to be correlated with present day standing diversity, whereas high standing diversity could reflect either high speciation rates in that region, low extinction rates, high immigration of species that originated elsewhere, or some combination of these variables. Similarly, in the absence of geohistorical information on species ranges, extinction risk would have to be evaluated only in light of the population dynamics measurable today. These are examples in which the unique qualities of time-averaged paleobiological information permit questions to be answered that—if only modern-day biological data were available—could only be addressed indirectly. High-quality geohistorical records are present throughout the ~4 billion year record of Earth, although the total quantity of records decreases with increasing age. Geohistorical records from the relatively recent past contain a wider variety of extractable information (e.g., ancient DNA is more likely to be preserved, and isotopic signatures are less likely

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change to be overprinted), than more ancient records. The taxonomic and functional characteristics of biological remains are more readily inferred by direct comparison to living forms, and dating resolution is better. Because interpretations based on agreement of multiple lines of evidence (whether direct or proxy) are judged more robust than inferences based on single lines of evidence (Mann, 2002), younger records are more likely to yield high-quality information about both environmental change and its biotic response. Variation in the quantity and quality of data through time occurs over all timescales. For example, archived materials are subject to loss and deterioration even under the best of conditions, and—due simply to scientific and technological progress—workers during the last decades generally have used more precise and diverse means of characterizing and analyzing objects and systems than workers of 50 to 100 years ago. In Earth history—as in other areas—such variation does not invalidate historical analysis but only requires that analysis be done thoughtfully and with standard scientific rigor. Both biologically and geohistorically, some regions, habitats, and kinds of organisms are better known than others at any moment in time. Information quality is not randomly distributed with respect to geographic region, environment, or taxonomic group; for example, biotic systems and geohistorical records are better known in regions with long-standing academic programs or field stations (e.g., Western Europe and North America); some groups are better known because of their economic importance (e.g., insects as pests and vectors of disease, or benthic mollusks as food for commercial finfish); and some habitats and systems are better understood because of the greater ease of surveys and experiments (e.g., rocky intertidal settings, rodents and insect population dynamics). This unevenness can persist into more mature stages of data acquisition, because additional work on an already well-studied group or area benefits from past insights and investments. As a consequence, some systems are not as well explored as others despite the likely existence of materials (living communities, geohistorical records) that could yield valuable insights. In any empirical scientific analysis, the ideal is for sampling to be at regular intervals along a gradient, or through time over the course of an experiment, at the same time holding other conditions constant. The greater the density of sample points along the gradient or during time, the greater the confidence in recognizing the effects of processes operating at fine spatial or temporal scales, discovering correlated relationships among the variables and the greater the likelihood of understanding their underlying significance. Such regular sampling is not always possible in natural systems; instead, it is often necessary to interpolate among irregularly

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change spaced points or forced to compare the end points of a suspected spatial gradient or temporal trend. Geological analysis of ecological dynamics is as subject to these challenges as is field-based biology, where for example, the availability of rocky shores would determine the spacing of samples in a latitudinal analysis between Baja California and Oregon. Whether such limitations on continuous sampling constitute a significant impediment depends on the question at hand, but clearly continuous sampling is not required to answer all questions, nor is it always possible. What is most important in comparative analyses, whether biological or geohistorical, is to (1) have a good match between the question and the quality and quantity of samples that can be acquired; (2) identify the specific variables to be explored (e.g., gradients in environment, variation over geologic age or geographic region); (3) hold other variables constant (for example, sampling protocol—in geohistorical analysis this would also include the preservation quality of the local records sampled); and (4) have a reasonable number of replicate samples per treatment (i.e., per site, per region, per geologic period). With appropriate diligence, together with adjustment of questions as necessary, these challenges can be met in geohistorical analyses with a success rate that is similar to that achieved in field biology experiments. INFERRING ENVIRONMENTAL CONDITIONS—RESOLUTION AND CONFIDENCE Various elements of the geologic record capture environmental information in indirect, or “proxy,” form. For example, variations in air temperature and rainfall affect the width of tree rings and, accordingly, variations in the widths of tree rings are proxy indicators of variation in temperature and moisture. Similarly, variation in sea-surface temperature and ice cap volume are reflected in fluctuations of oxygen isotope ratios in skeletal remains preserved in marine sediments. In many cases, proxy indicators can be calibrated based on experiments or on their observed variation in known environments, allowing their use for the determination of precise quantitative measures of environmental conditions. The composition of biotic assemblages has long been used as a basis for quantitative inference of climate, salinity, redox, pH, water depth, productivity, and a variety of other environmental factors (e.g., Imbrie and Kipp, 1971; Bartlein et al., 1986; Fritz et al., 1991; Charles et al., 1994; Charman, 2001; Smol et al., 2002a,b). The discrimination of paleotemperatures with a standard error of only ± 2°C using shape and size characteristics of leaves is a good example of a proxy with relatively high precision based on modern calibration (Wilf, 1997), and the Stomatal Index (SI) of fossil leaves can provide historical records of past

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change levels of CO2 (Wagner et al., 1999). Similarly, stable isotopes incorporated into fossils and other geohistorical records are providing extraordinarily detailed insights into population size (e.g., Finney et al., 2000), home ranges (Hoppe et al., 1999), diet (e.g., Burton et al., 2001), photosynthetic pathways (e.g., Fox and Koch, 2003), river discharge (e.g., Dettman et al., 2004), salinity (e.g., Rodriguez et al., 2001), climate (e.g., Edwards et al., 1996; Huang et al., 2002), ocean pH (Lemarchand et al., 2000), and elevation (Morrill and Koch, 2002), and effects of CO2 and climate on vegetation (Huang et al., 2001; Street-Perrott et al., 2004). Some proxies provide only semi-quantitative or qualitative measures of environmental conditions. For example, sediment grain size, bedforms, and associated paleocurrent evidence for the frequency and magnitude of fairweather and storm reworking permit the confident—but qualitative and comparatively coarse—determination of shallow-, intermediate-, and deep-water paleodepth estimates for continental shelf deposits. Because water depth is rarely an environmental parameter per se (compared, for example, with sediment grain size, frequency of disturbance, or organic content), the recognition of semi-quantitative depth categories in the geologic record is sufficient in most instances. In the terrestrial fossil record, geologists can similarly discriminate among the fluvial sub-habitats of main channel, channel bar, levee, and flood plain by using sedimentary features (e.g., grain size distribution, sediment color, presence and type of root traces) characteristic of water content and soil development. States of preservation of fossil material provide additional information on environmental conditions at and immediately below the depositional interface, including the frequency and agent of burial events. Preservational condition also provides insights as to whether the fossil material is indigenous or exotic to the depositional site and thus whether the inferred ecological tolerances of fossil taxa can be used as proxy environmental information (see next section). Environmental proxies thus may be based on physical, chemical, or biological features of the geologic record (see Table 2.1). The fossilized remains of organisms—shells, bones, non-mineralized organic tissues—typically carry several kinds of proxy information. For example, the abundance of fossils can reflect original population size, predatory damage can indicate trophic role, and the isotopic composition of their hard parts can capture information on growth temperatures, and ambient salinities (see Box 2.2). Proxy indicators can, of course, be affected by post-depositional alteration and destruction. Some have proven utility through much of the geologic record (e.g., carbon isotopes), some are limited to the Mesozoic and Cenozoic (e.g., stomatal density), and others can be used only in very young deposits under special circumstances (e.g., fossil DNA).

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change TABLE 2.1 Examples of Proxy Indicators and Their Interpretation Source Proxy Type of Information Individual fossils Number and variation of growth rings/bands, size Seasonality, growth rates, growth conditions, generation times, biomass, age, and season of death Fire scars in tree rings Fire intensity and frequency Stomatal density in leaves Atmospheric CO2 Herbivore or predator damage Intensity of herbivory or predation, herbivore or predator preferences, evolution of herbivore feeding strategies and plant responses Leaf shape and size Mean annual temperature, elevation Overgrowth relationships in encrusters Competitive relationships Larval shell type Mode of dispersal, mode of development, evolutionary history Body size Body size, biomass, size frequency distribution Density Individuals per unit area, biomass Location Habitat, dispersal history, geographic distribution Tooth wear Diet in mammals Preservational condition Exposure, hydraulic energy, sedimentation rate Fossil DNA (generally limited to deposits <130 Ka) Presence of species not otherwise preserved, evolutionary relationships and rates Composition and structure of fossil assemblages Taxonomic composition Taxa present, guild structure, environmental tolerances, diverse climatological, hydrological, and other environmental variables

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change Source Proxy Type of Information Composition and structure of fossil assemblages Presence/absence and relative abundance of plant functional types Physical structure of vegetation—albedo, surface roughness, evapo-transpiration Species richness, diversity, and evenness Species richness, diversity, and evenness Relative abundance of taxa Relative abundance of taxa, dominance, shape of abundance frequency distribution Geographic distribution Geographic distribution, climate Presence/absence and relative abundance of animal guilds Sediment consistency, suspended food supply, presence and structure of vegetation Geochemical and isotopic composition of skeletal hard parts Strontium isotopes Source and amount of freshwater, migration histories, home range Oxygen isotopes Temperature range and variability, salinity, ice volume, source of moisture Carbon isotopes Source of food, vegetation type, productivity, atmospheric CO2, humidity, temperature Nitrogen isotopes Source of food, trophic level Boron isotopes pCO2 concentrations Hydrogen isotopes Temperature, source of moisture Multi-isotope systematics Climate, hydrologic conditions, source of organic compounds, biogeochemical cycling Ba/Ca ratios Productivity Mg/Ca ratios Temperature Sr/N ratios Trophic level in mammals Sr/Ca ratios Temperature range and variability

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change FIGURE 2.2 Effective temporal range (vertical axis shows millions of years into the past) and precision (width of horizontal bars represents 2-sigma confidence intervals) of some dating techniques commonly used to date geologic materials. The numbers below each technique indicate 2σ variation expressed as a percent of calculated age; horizontal scale bars below each expanding triangle express the 2σ variation in millions of years. SOURCE: Doug Erwin, National Museum of Natural History; used with permission.

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change of β-particle emission during decay of 14C isotopes. Such measurement requires large amounts (usually ≥1 g) of carbon, which often necessitates mixing of sediment constituents or organic materials of different ages. Some aquatic plants and algae use HCO3− and CO3= in photosynthesis, which may derive from dissolution of ancient carbonate-rich bedrock or surficial materials. Many organic-rich materials (bone, peat, shell) can be contaminated by intrusion and replacement by younger or older carbon. Thus, conventional β-decay dates are subject to imprecision and error beyond the standard counting error. Development of accelerator mass spectrometry (AMS) as an alternative means of assessing isotopic composition of carbon samples helps circumvent some of these problems. Much smaller quantities of carbon are required (as little as 10 µg), which minimizes time-averaging , and has other practical advantages: Dates can be obtained on compound-specific organic fractions from samples, using the most stable organic molecules and avoiding those most likely to exchange carbon with the surrounding environment. This has been especially useful in the dating of bone, shells, and concentrations of pollen and other microfossils. Dates can be obtained on one or a few specimens (plant fossils, bones, shells), separate portions of long-lived individuals, and even individual carbon compounds (e.g., Ohkouchi et al., 2003). Dating can be restricted to materials that are most likely to yield reliable dates (e.g., terrestrial plant macrofossils, charcoal). Dates can be obtained on sediment samples spanning as little as a few mm of deposition, rather than several cm. The advent of AMS dating has yielded more precise and accurate sediment chronologies and age estimates for fossils, and has made it possible to assess whether fossils occurring in the same stratum were truly contemporaneous (or, alternatively, represent contamination or time-averaging ). Because the production rate of 14C atoms in the atmosphere varies owing to variation in cosmic radiation, the 14C content of the atmosphere varies through time. Radiocarbon dates are thus not equivalent to calendar-year dates. However, variations in atmospheric 14C content for the past 20,000 years are well characterized, and so it is possible to estimate the calendar-year age of a sample given its radiocarbon age2 (Stuiver and Reimer, 1993). 2   See also http://radiocarbon.pa.qub.ac.uk/calib/.

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change Uranium/thorium techniques also can be used to date fossil material as old as ~300,000 years. Although molluscan aragonite does not appear to provide a closed system and often yields unreliable dates, coral skeletons have proven to be especially suitable for U-series dating. The technique also has been applied to tooth enamel and bone, but additional work is needed to refine the analyses of these materials. Amino acid racemization techniques (Wehmiller, 1993; Rutter and Blackwell, 1995; Goodfriend et al., 2000) also can be applied directly to skeletal materials (bone, shell, eggshell), and when calibrated (typically with radiocarbon or uranium/thorium techniques) can yield estimates of calendar age (e.g., Kaufman, 2003; Kowalewski et al., 1998). Within the range of radiocarbon dating, calibrated amino acid dates can be reliable and economical alternatives to radiocarbon dating for particular sites. Amino acid racemization dating is also valuable beyond the range of the radiocarbon technique, although precision is lower. Electron spin resonance can also be applied directly to fossil materials, such as shell and enamel, but the method requires further calibration and dates are rarely as precise as those produced by radiocarbon dating (Blackwell, 1996). Dating Associated Geologic Materials Other techniques of dating (luminescence, lichenometry, cosmogenic nuclides, paleomagnetism, astronomical cyclicity, radiometric other than radiocarbon and uranium/thorium) cannot be applied directly to fossil specimens but can date material associated with the fossil in some way, and this is the primary means of age determination for fossils that cannot be dated directly. For example, cave deposits (speleothems) associated with fossils can be dated using uranium/thorium. Age estimation of sediments spanning the past two to five centuries using radiocarbon-based models is rendered difficult by several factors. First, landscape disturbance frequently causes changes in sediment accumulation rates in lake and wetland basins, and most parts of the world have been cleared or cultivated in the past few centuries. Second, the precision of radiocarbon age estimates has a finite limit, typically ± 50-100 years except in a few unusual cases. However, other dating methods can be applied. For example, varve counts can provide precise chronologies for lakes with annually laminated sediments. Certain short-lived isotopes that accumulate in sediments can be used for dating. Primary among these is 210Pb, which can provide reliable age models for sediments of the last 100-150 years (Noller, 2000). Cesium-137, which is produced during nuclear explosions, is useful for very recent sediments. Microscopic and macroscopic carbonaceous/metallic particles produced by industrial combustion can sometimes be used as stratigraphic markers, as can

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change biostratigraphic markers (e.g., Ambrosia, Plantago, Salsola, Cannabis, and other indicators of land clearance or cultivation). Uranium-lead and argon-argon are two of the more widely used systems for absolute age determination in older, pre-Quaternary rocks. Both generally depend on samples of detrital volcanic materials, generally from ash falls preserved in the sedimentary rock record. Minerals studied include monzanite, zircon, apatite, and uraninite. Both systems are highly accurate, and the two techniques complement each other by providing important cross-checks on dates. U/Pb encompasses two decay systems (uranium-235 and uranium-238), thus providing an internal check on the reliability of results; argon lacks this ability. Rocks amenable to argon dating are somewhat more common, increasing the scope of the technique. There is also the prospect that continuing uncertainties about the precise potassium-40 decay constant will be resolved in the next few years, thereby opening up a further technique for developing accurate chronologies. Two distinct techniques are applied to the U/Pb system—Super High-Resolution Ion Microprobe (SHRIMP) and Isotope Dilution Mass Spectrometry (ITDMS). SHRIMP studies have the advantage of analyzing extremely small areas on single zircon grains, allowing identification of later overgrowths on the rim of grains, a complication that can reduce the precision of results. SHRIMP studies are particularly powerful for Archean and Proterozoic rocks, although they are also useful through much of the Phanerozoic. Where dates are not available for every level in a local sedimentary sequence, it is generally possible to develop an age model to estimate the ages of intervening levels, especially when the record is from depositional systems that have generally uniform sedimentation rates (most notably lake and deep marine sediments). Rates of sedimentation are calculated by interpolation between the few dates that are available. This rate then can be used to model the chronology for intervening intervals and, by extrapolation, through deeper parts of the local record if this extends beyond the limits of a single dating technique (e.g., radiocarbon). Milankovitch periodicities (periodic oscillations induced by changes in orbital parameters occurring at 23 Ka, 41 Ka, 100 Ka, and 400 Ka timescales; see Hays et al., 1976) have been recognized in sedimentation patterns, isotopic ratios, and other features back in time to at least the Miocene (and sporadically to the Triassic). Such variations have been very useful for calibrating many long-term records, and are now routinely employed in paleoclimatology and have been successfully employed in paleoceanography and stratigraphy (e.g., Zachos et al., 2001; House and Gale, 1995; Hinnov, 2000).

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change Correlation Techniques Even when absolute ages cannot be determined for fossil deposits, it is often possible to correlate intervals between localities and to establish contemporaneity. When absolute ages are known in one location, correlation allows projection of those dates to a different location. Techniques used in correlation between localities are varied. Biostratigraphy relies on broadly distributed, short-ranging species that allow correlation between rock units deposited during the same interval of time. Pelagic marine microfossils and to a lesser extent pollen are the most useful such fossils because of their broad geographic range and rapid evolutionary turnover, but a wide variety of fossils have been used. Biozonation schemes are well established for some fossils (e.g. pollen, foraminifera, radiolarians, nannofossils), in large part because of the immense efforts during the Deep Sea Drilling Project (DSDP, from 1968 to 1983) and Ocean Drilling Program (ODP, from 1985 to 2003). In continental deposits of the Cenozoic and latest Cretaceous, land-mammal ages—intervals of time characterized by distinctive assemblages of mammals (Lindsay, 2003; Woodburne, 2004)—have proven especially useful. Many biozonation schemes are increasingly well integrated with radiometric age determinations of associated rocks and with zonations of relative time based on past reversals in Earth’s magnetic field. More recently, other correlation techniques have been applied: (1) using distinctive isochronous events (isochrons) that occurred at a single, discrete point in time; (2) wiggle matching; and (3) quantitative biostratigraphy. These approaches are especially valuable tools because they permit correlation between marine and terrestrial sequences, resolve the ordering of events, and greatly improve the fidelity of correlation over standard zonations. Isochrons are a powerful tool because if correctly identified, they represent a single time horizon. Examples include volcanic ash beds and rapid shifts in carbon isotopes that permit correlation over large distances. The boundaries of magnetic reversal intervals can also serve as isochrons; although individual reversals are probably drawn out over hundreds to thousands of years, this fuzziness will in many cases be insignificant given the level of resolution that is otherwise possible. Epiboles—thin, widespread beds of a single species—can similarly serve as time horizons (Brett and Baird, 1997), but are typically restricted to intra-basinal rather than global correlation. In those rare instances where the horizon of interest contains material that can be dated to a high level of resolution (e.g., the volcanic ashes that coincide with the latest Permian mass extinction in south China [Bowring et al., 1998]), the isochrons also provide constraints on the rates of ecological or evolutionary response to environmental

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change change. Long-term records of changing isotopic values (typically of oxygen, carbon, sulfur, or strontium) in stratigraphic sequences have proven to be especially valuable for correlation because many such shifts are global. Variations in isotopic values in a new sequence are matched to corresponding isotopic variations in a well-dated (calibrated) sequence using a procedure informally known as wiggle matching (e.g., Hoek and Bohncke, 2001). In the Quaternary, variation in atmospheric 14C concentrations can be used to increase dating precision in some circumstances. Such wiggle matching, in which a tightly spaced series of AMS 14C ages is plotted against sediment depth, and then matched to the 14C-calendar-year curve, can yield subdecadal precision for Sphagnum peats (Kilian et al., 2000; Blaauw et al., 2004). However, this application requires large numbers of AMS dates (10 to 100 or more per core), at high expense. Strontium isotopes in the oceans have varied through time as a result of changes in the composition of continental rocks, the extent of hydrothermal activity, and rates of continental weathering. However, because the change in strontium isotope composition was not monotonic during the Phanerozoic, and isotopic values show some scatter at any particular time, this approach does not yet yield high-resolution dates, and it cannot be used for high-resolution correlation. Newly developed techniques and applications in geochronology and quantitative biostratigraphy (see Agterberg and Gradstein, 1999; Buck and Millard, 2003; Sadler, 2004) seek to both order and correlate by integrating, sequencing, and calibrating a large number of events. These approaches are based on algorithms that emphasize the temporal ordering of events occurring in different locations, rather than the establishment of simultaneity of events, as with the search for isochrons. Thus, the principal result is an integrated stratigraphy that can then be assigned a temporal framework using high-precision dating. Such ordination techniques have been used with a variety of taxonomic groups in a variety of stratigraphic settings (e.g., Guex, 1991; Alroy, 1994; Sadler and Cooper, 2003). These approaches have the potential to improve temporal resolution to 10,000 to 50,000 years over time spans from 50 million to 100 million years (Sadler, 2004). As biostratigraphic databases increase in number and size, a growing challenge will be the provision of computing facilities to handle these computationally intensive procedures. The most recent geologic timescale (see Figure 2.3; Gradstein and Ogg, 2004; Gradstein et al., 2004) integrates currently available geochronologic and stratigraphic information and represents perhaps a two-fold increase in refinement over the timescale available 20 years ago. For example, 40-Ka resolution is now available in the Neogene, more than 200 radiometric dates provide calibration points, and error bars are now provided to estimate uncertainties associated with boundaries. Continued refinement of

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change FIGURE 2.3 The geologic timescale of Gradstein and Ogg (2004), with numerical ages from Gradstein et al. (2004). The “golden spikes” represent stratigraphic unit boundaries that are located by a particular point in an actual stratigraphic section. SOURCE: International Commission on Stratigraphy; used with permission.

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change the geologic timescale is now being systematically pursued3 and is likely to provide even greater precision and accuracy for future studies of the geologic record of ecological dynamics. IMPACT OF DATABASE TECHNOLOGIES AND SYNTHETIC STUDIES The development of databases and database technologies during the past 20 years has facilitated analyses and syntheses at broad geographic and temporal scales, stimulated broad, community-based research projects, and enabled cross-disciplinary work. The addition of web-based technologies for rapid data transfer and communication has greatly improved scientific communication and dramatically improved the dissemination of research results. Databases and associated research projects abound in both the biological and geological sciences. Examples in the biological sciences most relevant to this report are databases on the distribution of plants (e.g., SALVIAS—Spatial Analysis of Local Vegetation Inventories Across Scales4), fish (FishBase5), phylogenetic analyses (e.g., Tree of Life6), and many others available through links from the Natural Science Collections Alliance.7 The importance of the analysis and synthesis of wide-ranging datasets and databases to the ecological and evolutionary sciences has been recognized by the establishment of the NSF-supported National Center for Ecological Analysis and Synthesis (NCEAS—see Chapter 4) and the recent NSF funding of a Center for Synthesis in Biological Evolution (CSBE). In the geological sciences most relevant to this report, examples include databases for paleoclimatic reconstructions (National Oceanic and Atmospheric Administration’s [NOAA] World Data Center for Paleoclimatology8), the JANUS9 database of results from the Ocean Drilling Program (see Chapter 4), resources for paleogeographic reconstructions (e.g., Paleogeographic Atlas Project10 and PALEOMAP project11), data on the diversity of fossils through geologic time (Paleobiology Database12— 3   See http://www.stratigraphy.org. 4   See http://eeb37.biosci.arizona.edu/~salvias/mission.html. 5   See http://www.fishbase.org/home.htm. 6   See http://www.tolweb.org/tree/phylogeny.html. 7   See http://www.nscalliance.org/bioinformatics/index.asp. 8   See http://www.ngdc.noaa.gov/paleo/paleo.html. 9   See http://www-odp.tamu.edu/database/. 10   See http://pgap.uchicago.edu/PGAPhome.html. 11   See http://www.scotese.com/. 12   See http://paleodb.org.

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change FIGURE 2.4 The “Sepkoski curve” showing numbers of genera of marine animal fossils through Phanerozoic time. The blue portions of the curve indicate time intervals chosen for more detailed study of diversity change by Alroy et al. (2001). SOURCE: Newman (2001); used with permission. see Chapter 4), information on the stratigraphic distribution of fossils (e.g., NEPTUNE13 and Paleostrat14), and many more specialized databases (e.g., Panama Paleontology Project15 and Neogene fossil mammals of the Old World16). There is probably no better example of how data compilations and syntheses can shape entire disciplines than the graphs illustrating the diversity of families and genera of marine fossils through geologic time (e.g., see Figure 2.4). Laboriously assembled from the print literature by Sepkoski (1982, 1993, 2002), these simple figures summarized the state of 13   See http://services.chronos.org/databases/neptune/index.html. 14   See http://www.paleostrat.org/. 15   See http://www.fiu.edu/~collinsl/pppimagemapnew.htm. 16   See http://www.helsinki.fi/science/now/.

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change knowledge of diversity change though geologic time, and in many regards they have proven robust to taxonomic revision and phylogenetic approaches (e.g., Sepkoski and Kendrick, 1993; Robeck et al., 2000; Adrain and Westrop, 2000). They have attracted broad attention in the biological sciences because of the near-plateau in diversity over several hundred million years in the Paleozoic, the abrupt extinctions and rapid rebounds that punctuate the time series, and the dramatic post-Paleozoic diversity increase to the present day. Interpretation of all these features remains controversial, and countless research projects have been engendered by new approaches to the biological and geological factors underlying the entire curve and its many taxonomic and ecologic components (e.g., Sepkoski, 1984, 1996; Raup and Boyajian, 1988; Valentine et al., 1991; Eble, 1999; Bambach et al., 2002; Alroy et al., 2001; Foote, 2003; Jablonski et al., 2003). Sepkoski’s efforts helped launch a new, integrative, and synthetic style of research in paleontology, and were instrumental in the formation of the community-based Paleobiology Database, which has initiated a next-generation compilation using spatially explicit occurrence data. SUMMARY AND OUTLOOK—STRATEGIES FOR GEOHISTORICAL ANALYSIS Geohistorical data are essential for answering many kinds of questions, especially when the aims are either to discriminate between anthropogenic and non-anthropogenic effects or to understand phenomena that cycle or emerge over periods greater than a few years. Such discrimination is extremely difficult without recourse to historic records in the broadest sense—pre-scientific documents, archaeological materials, natural sedimentary records. Therefore, a rigorous strategy for evaluating ecological dynamics using geohistorical records—and for integrating geological and biological methods and insights—is essential. Intensified research on these issues over the last 15-20 years now enables the exploitation of uniquely valuable data lodged in geohistorical records (summarized from above): Important environments (notably lakes, peatlands, estuaries, and deep ocean basins) typically yield high-resolution time series data ideal for tracking ecosystem change. Valuable and unique data also can be acquired even from isolated before-and-after sample sets, and the time-averaging that characterizes many geological samples of past ecosystems is advantageous for statistical analyses of broad-scale changes and patterns. Newly developed geological and, especially, geochemical methods provide an increasing number of proxy indicators of environmental change. Thus, changes in the composition, relative abundance, distribu-

OCR for page 30
The Geological Record of Ecological Dynamics: Understanding the Biotic Effects of Future Environmental Change tion, and diversity of fossil biotas can now be used as measures of the biotic response to environmental change, rather than indicators of both the change and the response. An increasing number of ecologically valuable metrics can now be evaluated confidently, either using the conventional fossil record of preserved individuals, or using various chemical proxies of former biological populations (biomarkers, isotopic signatures). Relative and absolute age dating of geologic materials and time series has improved greatly in the last 20 years, and has crossed a key threshold with respect to analyzing ecological dynamics. Resolution now matches or exceeds what would be possible using conventional biostratigraphic methods, liberating geohistorical analysis of ecological dynamics from pitfalls of circularity in age and rate determinations. New database and web technologies have stimulated community-based efforts to assemble and analyze large amounts of ecological, paleoecological, and evolutionary data relevant to understanding the geologic record of ecological dynamics. Such efforts at synthesis are essential partners in efforts to acquire new data. The geologic column provides a wealth of geohistorical records that are ready to yield important data on ecological dynamics. It is not the record that is inadequate but rather the availability of resources to extract and analyze the record. Using geohistorical records to their fullest potential will require an effort focused on the biological analysis of groups known to have reasonable fossil records, and on the development of proxy methods for groups of great biological interest but poor fossilization potential. Biologists and earth scientists, working together, will need to frame research questions so that the answers take maximum advantage of the great potential of geohistorical analysis to provide major insights.