Indicator Species and Fossil Assemblages

The species and assemblages of about a dozen major groups are used routinely in paleoclimatology, and many more minor fossil groups are used less commonly (figure 2-8, table 2-7). What makes a species a good proxy of climate parameters? Although the answer to this question varies widely from group to group, Coope (1977) provides five attributes of species of Coleoptera (beetles) that also apply to most other groups: evolutionary stability, morphological complexity (species-specific fossilizable characters), abundance in sediments (allowing quantitative assemblage analyses), physiological constancy (difficult to ascertain directly), and rapid response to climate shifts. Coope

Plankton And Paleoclimatology

FIGURE 2-8 Several micropaleontological groups used as proxies in paleoclimatology. From upper left: high marsh benthic foraminifer (Trocham-mina inflata), ostracode (Schizocythere kishinouyei), planktonic foraminifer (Globigerinoides ruber), marine diatom (Thalassiosira antarctica). Courtesy of D. Scott, A. Tsukagoshi, H. Dowsett, A. Leventer, and R. Scherer.

FIGURE 2-8 Several micropaleontological groups used as proxies in paleoclimatology. From upper left: high marsh benthic foraminifer (Trocham-mina inflata), ostracode (Schizocythere kishinouyei), planktonic foraminifer (Globigerinoides ruber), marine diatom (Thalassiosira antarctica). Courtesy of D. Scott, A. Tsukagoshi, H. Dowsett, A. Leventer, and R. Scherer.

adopted an actualistic approach to indicator species and communities in reconstructing past climate whereby the modern ecological requirements of species are used to directly interpret the significance of fossil assemblages and infer past climates. Coope did not mince words in expressing his confidence in his beetles as indicators of climate change: "fossil assemblages when treated in this way make ecological sense; [they are] the best possible justification for a hypothesis, namely that it works'' (p. 317). In late Quaternary continental deposits of England, Europe, Canada, and elsewhere, fossil Coleoptera have been used extensively to document rapid climate change.

An actualistic approach is used by many other paleoclimatologists who use fossil evidence to reconstruct climatic parameters. For example, any study in which the statistical relationship between the modern distribution of species and a climate-related parameter like temperature or salinity be established takes an actualistic approach. Imbrie and Kipp (1971), in their seminal study of planktonic foraminiferal assemblages and Atlantic Ocean surface water masses, relied heavily on foraminiferal species biology (e.g., Be and Hamlin 1967, Be 1977) to develop the transfer function paleoecological method of SST estimation. They stated clearly that this approach involves ecological assumptions regarding the use of core-top foramini-feral assemblages to estimate SSTs. One of the most common assumptions is that the maximum abundance of a species in a sample indicates that it was inhabiting its preferred ecological niche. This assumption also implies that the species is evolutionarily "best adapted'' to a particular set of environmental conditions.

Establishing the modern ecological preferences of foraminiferal species can be accomplished by mapping species distributions using dead specimens from core-top material (Imbrie and Kipp 1971), living plankton tow samples, living and dead bottom samples (Cronin and Dowsett 1990), or sediment trap samples. These sampling methods can result in useful databases that can serve as a baseline for quantitative analyses of past assemblages, especially over large spatial scales. Biogeographic databases are even more valuable when additional experimental or ecological data are available on species' habitat preferences (e.g., Spero et al. 1991).

Although the ecological limits of species can be estimated from their modern distributions, two complications arise in the application of statistical associations of species to reconstruct past changes in climate-related parameters. First, modification of the original faunal or floral assemblage occurs by taphonomic processes. Taphonomy, a large subdiscipline of the geosciences, involves the study of processes that alter a living community of organisms after death. The proper interpretation of taphonomic effects of paleocommunities is not a trivial matter. Even if the ecological tolerances of species are well known, the transition from life to death to fossil involves a series of complex processes, each of which can further complicate paleoclimatological interpretation. In the example of modern pollen rain, many biases are introduced between the time plants pollinate, the time pollen settles into lakes and bogs, and the time it is recovered from lake sediments (Birks and Birks 1980). These biases include biological factors such as differential pollen production, dispersal capabilities from pollen grain aerodynamics, and preservation within the sediments. In the case of marine microfossils, such as foraminifers, ostracodes, and diatoms, processes of transport, burial, and differential dissolution must similarly be taken into account before an assemblage can be properly analyzed. A firm grasp of taphonomic processes is essential for the successful reconstruction of climate history using fossil organisms over all spatial and temporal scales.

The second complication is whether ecological or climatic factors are responsible for observed temporal variability in species' proportions. Some researchers, especially those investigating vegetation changes reconstructed from pollen spectra, have attributed changes in pollen assemblages to plant-type life histories, dispersal, or competition, rather than to climatic factors. Community disequilibrium among species in forest communities has been cited as the cause for the differential rate and dispersal of plant taxa after deglaciation (e.g., Davis 1981). Population dynamics and interspecific interactions have also been cited as processes more important than climate change to explain patterns of pollen spectra since the last glacial period (see Prentice et al. 1991). Because biological processes can influence paleo-climatological interpretations so heavily, they are discussed at length in chapter 3.

During the past few decades, research on many fossil groups has reached a point where the geographic and/or bathymetic distributions of species are well known, the biology of many key species is understood, and quantitative methods are routinely applied to fossil assemblages. One database of eastern North American surface pollen samples, for example, includes pollen census data from about 1000 modern pollen samples (Prentice et al. 1991); the European pollen database of Guiot (1990) consisted of 655 samples. The planktonic foraminiferal core-top data have grown to thousands of ocean core-top samples. Other regional faunal and floral data sets often include tens to hundreds or more modern samples (e.g., ostracodes [Cronin and

Dowsett 1990; Ikeya and Cronin 1993], coccoliths, benthic foraminifers [Polyak et al. 1995], diatoms [Leventer et al. 1996]).

Many statistical methods have also been applied to translate faunal and floral census data into paleoclimatic reconstructions. Of the dozens of techniques available, the most commonly used are the transfer function (TF) (Imbrie and Kipp 1971), the modern analog technique (MAT) (Overpeck et al. 1985), and the response surface (Bartlein et al. 1986) methods. The TF method first uses factor analyses to establish a small number of "factors,'' or assemblages, based on the relative proportions of different species over an environmental gradient. For example, Imbrie and Kipp (1971) defined six major Atlantic Ocean assemblages: polar, subpolar, subtropical, equatorial, gyre, and ocean margin. Once the factors are defined, multiple regression analysis is used to formulate equations that relate the factors to the climate parameter, usually temperature or salinity. These equations are used to calculate paleotemperatures from downcore assemblages.

In MAT, species census data are also used to reconstruct past conditions, but instead of factor analysis, the statistical measures known as coefficients of dissimilarity are used to compare two assemblages on the basis of the relative proportions of key taxa. Overpeck et al. (1985) explored MAT using a number of dissimilarity coefficients in a study of North American deglacial pollen records. In Europe, Prentice et al. (1991) compared last interglacial (Eemian) pollen profiles to a modern database on pollen distribution for 28 plant taxa in surface samples from across Europe and argued that plant types exhibit a nonlinear response to several climate variables.

The response-surface method (Bartlein et al. 1986) measures the response of common tree taxa to climate parameters such as precipitation and temperature, identifying the ideal habitat conditions for key plant taxa. The many excellent reviews of these and other methods include the papers by Guiot (1990) and Prentice et al. (1991).

+1 0

Post a comment