Paleozoic Data of Climatological Significance and Their Use for Interpreting Silurian-Devonian Climate
Oregon State University
University of Oregon
In principle, enough geologic data of climatological import have been available for many years to enable climatologists to attempt global climatic reconstructions for the Paleozoic. However, only in the past few decades, particularly with the advent of plate tectonic concepts, has this information begun to be synthesized in a manner that makes it usable to the climatologist. Lithofacies data with climatic implications, for example, are now available for the Paleozoic on a global scale, and biogeographic data are being synthesized for the major time intervals. But most classes of information are not adequately compiled, and others for which syntheses exist are for such relatively broad time intervals that they will confuse attempts to gain climatological understanding. Our purpose in this paper is to review categories of information useful in reconstructing Paleozoic climates. We begin with those most readily available and most important and conclude with some that, while little exploited, may provide important supplementary data. We present some examples of how these data have been used climatologically in the Paleozoic, and finally, using several categories of information, we provide an exam ple of how these data might be used to interpret Silurian-Devonian climates.
PALEOZOIC CLIMATIC INFORMATION
The data of Paleozoic biogeography are readily available in several volumes, edited by Middlemiss et al. (1971), Hallam (1973), Hughes (1973), Ross (1974), and Gray and Boucot (1979). These volumes include papers covering many Paleozoic organisms but chiefly shallow-water, marine invertebrates. The coverage is by no means encyclopedic, but it is extensive enough to provide an approximate biogeography for the Paleozoic Periods.
What data does biogeography provide for the climatologist? In view of what we understand of modern biogeography we can conclude that the animals and plants of any biogeographic unit were in reproductive contact and that the environment within the unit had a certain level of uniformity. Thus, a shallow-water biogeographic unit of continental-shelf
depth based on sea shells (as are most of the Paleozoic units) indicates a water mass having relatively uniform properties and a shallow-water circulation system permitting reproductive communication. Although the larval stages of some marine benthos are capable of long-range transport within the plankton (teleplanic species), it appears that most species do not belong to this category. The taxa characterized by long-range transport and consequently more cosmopolitan distribution do, however, yield information about some aspects of long-range, shallow-current circulation.
In the nonmarine environment, Paleozoic data are almost exclusively for post-Devonian woody plants. Animals, vertebrate and invertebrate, of the nonmarine environment are too rare as fossils to be of practical value, although future study may alter this condition. With plants we deal chiefly with megascopic structures, but global studies of pollen and spores are beginning to provide supporting data.
Additionally, Paleozoic biogeography provides direct information about climates and climatic gradients for the Periods. Taxonomic diversity at all levels from the species through the superfamily tends to be significantly lower in cold or cool (temperate to glacial) climates than in warm (warm temperate to tropical) climates (Stehli et al., 1967; Stehli, 1968). Boucot (1975) summarized diversity data of this type for the Silurian and Devonian. Warm and cold or cool biogeographic units may also be indicated by the diversity present in benthic marine animal communities. Cold- to cool-water communities tend to have a much smaller number of species per community than do warm-water communities. There is also a tendency for the shells of organisms of cold- or cool-water units to be smaller and thinner than those of warm water units (Nicol, 1967).
The climatologist should recognize that levels of provincialism increase and decrease during the Paleozoic (see Boucot and Gray, 1979, for a brief summary of provincialism from the Cambrian through the Permian). These changes are probably the results of many interacting forces such as changes in global climatic gradients, paleogeography, and shallow-water circulation systems.
Boucot and Gray (1979, 1980) emphasized in general terms the climatic importance of carbonate and noncarbonate sedimentary rock sequences. Heckel and Witzke (1979, pp. 99–103) provided a detailed discussion of how carbonate rocks may be used for purposes of climatic analysis in the Paleozoic. Evidence indicates that carbonate rocks (limestone and dolomite of sedimentary origin) usually denote tropical to warm temperate conditions and that regions lacking such rock types in the marine facies probably were temperate to cooler. We have summarized above some biologic data that parallel the lithofacies evidence. Cool-climate, noncarbonate sedimentary rocks commonly are richer in unweathered mica (which imparts a glitter to the bedding planes) than warm climate noncarbonate rock sequences. When noncarbonate sequences are found in warm climates they are commonly associated with at least some redbeds (marine and nonmarine) and other evidence favoring an interpretation of warm climate.
However, climatic interpretation of a noncarbonate rock sequence requires a large sample; it cannot be based on single rock samples, rocks of a single roadcut, or rocks of a single limited geographic area. When an entire region has been studied, however, the climatic conclusions have a high degree of reliability.
The climatic significance of coal has been discussed briefly by Boucot and Gray (1980), who cited previous reviews by Krausel (1964) and Schopf (1972) dealing with the distribution and paleoclimatic significance of coal. Coal deposits indicate high humidity in association with either high or low temperature, as well as conditions adequate for the preservation of organic material. Similar observations were made by Davis (1913).
In the Paleozoic, coals occur in both types of climatic regimes—in the cool, high-humidity climate of the Gondwana Realm of the Late Carboniferous-Early Permian and in the major warm, high-humidity climate of the extra-Gondwanic Realms (Gray and Boucot, 1979). Heckel (1977) showed that the North American distribution of major warmclimate Pennsylvanian coals is well removed from contemporary evaporites. The areal distribution of coal belts relative to evaporite belts as well as scattered paleosols can thus provide climatic information. Heckel’s approach assumed that local orographic, rain-shadow effects will probably not affect the geologic record enough to prevent interpretation of broad climatic belts. However, some Paleozoic coal deposits will probably have to be explained in terms of local orographic perturbations. Heckel (The University of Iowa, personal communication, 1981) suggested, for example, that thin, nonminable coals and evaporites of Pennsylvanian age in Colorado were deposited in the rain shadow of the ancestral Rockies, although other interpretations are possible. Knowledge of local phenomena should make an orographic appeal rational in some cases.
Meyerhoff (1970) provided an excellent global summary of the distribution of Paleozoic coal deposits. Coals are virtually absent prior to the Late Devonian, when, presumably for the first time, enough land plant debris became available in suitable preservational sites to provide material for extensive coal deposits.
Evaporites, derived chiefly from marine water, form today in arid regions with ready access to the sea. In addition to an arid climate in which evaporation exceeds freshwater input, it is necessary that there be limited circulation between the evaporating basin and the oceanic reservoir lest the waters recombine before those of the evaporating basin approach salinities compatible with precipitation of evaporite minerals. Thus, evaporite deposits may be absent in regions believed from other evidence to have been arid during the Paleozoic. Moreover, the presence of evaporites during every Paleozoic Period does not imply that all parts of the Period were arid. For example, during the Silurian, an interval of about 30 million years (m.y.), evaporites are widespread in middle latitudes
during the Late Silurian but are rare during the Early Silurian. In addition, the width of past evaporite belts is partly a function of the global climatic gradients present during each time interval. For a further discussion of the climatic significance of evaporites see Chapter 10.
Evaporite deposition is incompatible with the climatic conditions characteristic of coal formation. Therefore, the distribution of coeval coal and evaporite deposits can provide an index to humid and arid climatic belts (Heckel, 1977; Heckel and Witzke, 1979), although coals are not sufficiently abundant prior to the Late Devonian for this purpose, Meyerhoff (1970) provided a useful Period-by-Period summary of Paleozoic evaporite distribution.
In the geologic record, a high correlation exists between warm climate indicators and the presence of both marine and nonmarine redbeds. Redbeds are defined here as detrital sedimentary sequences containing many red interbeds, as well as yellow and orange rocks. Geologic-mineralogic studies indicate that much of the red, yellow, and orange pigment consists of iron minerals oxidized under near-surface terrestrial conditions before being incorporated in the sedimentary record (Walker, 1967, 1975). However, older redbeds may be eroded and redeposited under climatic conditions entirely different from those under which they formed. Modern intertidal-shallow subtidal red sediments found near the head of the Bay of Fundy, Nova Scotia, and derived from the weathering and redeposition of Triassic beds are a good example of this phenomenon. In addition, there are diagenetic processes capable of producing red, yellow, and orange minerals that have little to do with the ordinary surface climatic condition responsible for the formation of these pigments.
Gray and Boucot (1979) reviewed some of the previously synthesized data relating to the distribution of Paleozoic redbeds. Global summaries for many of the Periods do not exist, although many raw data are available in the geologic literature. Nor are syntheses available for the distribution of redbeds within the Paleozoic Periods, even from secondary sources, such as regional geologies. What data have been synthesized, however, are consistent with the conclusion that redbeds are indicators of warm climates both past and present. For additional discussion of climate and redbeds, see Chapter 11.
The presence of redbeds in one region for a part of any Paleozoic Period does not necessarily indicate that conditions suitable for redbed formation characterized that region during the entire Period. The Paleozoic Periods are lengthy, and conditions changed from place to place within as well as between Periods.
Plant Morphological Features
A variety of gross morphological and anatomical characters recognizable in Paleozoic plant megafossil remains represent adaptations to environmental parameters that can be climatically interpreted (see Potonie, 1911; White, 1913, 1925, 1931; Noe, 1931; Krassilov, 1975, for Paleozoic examples). The basis for such interpretations are the autecological adaptations found among living plants, many of which are summarized by Daubenmire (1974), Richards (1979), and others. To use characters found in fossil plants as a basis for paleoenvironmental interpretations, it is necessary to assume that plants growing under similar circumstances and stresses have always adapted in the same morphological and anatomical manner. It is also necessary to assume that analogous features in fossil plants carry the same environmental connotations that they do in living plants. It must be borne in mind, however, that the adaptive significance of morphological characters of largely extinct plants, unrelated or only distantly related to living plants, may be different, particularly if there is conflicting biological and/or physical information. Moreover, the precise climatic variable—for example, temperature or moisture—to which the adaptation is a physiological response is not always clear, even when interpreting varied morphological features in modern plants (see Dolph and Dilcher, 1979). Finally, the physiological response to different climatic variables, perhaps even the opposite climatic variable, may lead to the same morphological adaptation (White, 1931, p. 272). However, when the varied morphological and anatomical data reinforce one another, and when they reinforce and bolster other data from the physical environment, they can prove an important source of paleoclimatic information.
Leaves are often regarded as among the most sensitive of plant structures to climatic conditions as they are the most exposed; roots the least as they are seldom exposed. There are specific types of root adaptations, however, that provide definite environmental information. Among the varied morphological and anatomical sources of environmental data in the Paleozoic are diverse features related to leaves (size, texture, gross morphology, and anatomy), tree rings, and other anatomical features connected with woody stems, varied adaptations connected with reproductive structures and roots, and growth habits.
Specific attributes of leaves that are regarded as environmental adaptations are texture (soft, delicate, thick, leathery, or coriaceous), size, surface hairs, scales and glands, the arrangement, number and position of stomates and varied anatomical adaptations related to cell size and cell wall thickness, intercellular spaces, and the development of certain tissue types. Two examples illustrate how this information may be interpreted environmentally. Many of the leaves of the Hermit Shale (Permian, Arizona) are very thick and leathery and have a scaly covering. White (1929, p. 21) interpreted such information, which reinforces other data bearing on the climate of that area, as indicating “a semi-arid climate with a long dry season, …” The “reduced, coriaceous and generally densely villous leaves” of Mississippian plants from Illinois suggested to White (1931, p. 272) similar unfavorable conditions of growth with a climate “characterized by severe droughts,” or with soils possibly “overdrained during dry seasons,” providing in either case evidence of seasonality,
Tree rings or growth rings in woody plants have been used extensively as a basis for making climatic deductions throughout geologic time. Compilations of tree-ring records for various intervals of geologic time are provided by Goldring (1921), Antevs (1925), and Chaloner and Creber (1973). Such rings are known to be common in living woody plants in envi-
ronments with marked seasonality, whether wet-dry seasonality or a seasonality that reflects temperature changes. In these circumstances, tree rings represent annual increments of wood laid down following a period of growth dormancy. Conversely, the absence of growth rings is believed to imply a more equable climate, in which growth occurred throughout the year.
In deducing information about ancient climates from tree rings, it is important to recognize that rings of a nonannual kind may be correlated with a variety of factors including fire, drought, disease, frost, floods, and defoliation among others. Antevs (1925), Tomlinson and Craighead (1972), and Chaloner and Creber (1973) discussed possible limitations to the use of tree rings in environmental interpretation. Tomlinson and Craighead (1972) and Chaloner and Creber (1973) noted a variety of nonclimatic factors that influence growth ring formation. In addition, it is essential to recognize that the presence or absence of tree rings in woody stems may be an unreliable guide in some circumstances to present climatic conditions, and by implication to climatic conditions of the past. For example, Tomlinson and Craighead (1972) noted that in the distinctive woody flora of subtropical south Florida, which mixes tropical and temperate plants, there is so much variety and variability in growth ring formation that it is difficult to find any climatic correlation. As a result of their studies in this region they write “…trees within a single climatic zone and vegetation type may or may not exhibit growth rings…. One is forced to the conclusion that the ability to develop growth rings is primarily determined by the genetic make-up of the individual species and only in a limited number of species is there a correlation with climate such that one distinct ring per year is produced” (Tomlinson and Craighead, 1972, p. 49).
The Paleozoic examples that follow presume that the presence of growth rings implies seasonality and the absence of growth rings lack of seasonality. The possible complications in this interpretation, noted above, should be borne in mind, especially where there may be conflicting biological and/or physical information. Arnold (1947, p. 391) provided a striking illustration of a stem of Callixylon from the Upper Devonian of New York State that shows what appear to be well-developed growth rings. However, Chaloner and Creber (1973) discussed other examples of Callixylon from Upper Devonian strata, including materials from the eastern and central United States and Europe, that show no growth rings or only very obscure growth rings. Chaloner and Creber (1973) concluded from their limited survey of Devonian woody plants that some show growth rings, but they suggest that the rings are less pronounced than would be expected. from woody plants at comparable latitudes at this time. White (1931) used the slight development or even “obscurity of annual rings” in Carbondale age floras of Indiana and Illinois together with other features of the vegetation to suggest general equability of temperature. Plumstead (1963) reported an example of silicified wood from Lashly Mt. (Antarctica) of Middle to Late Devonian age that shows “closely set annual rings indicating slow growth and marked seasons.” Examples of Permian gymnospermous woods from Antarctica that show broad and well-marked growth rings are provided by Schopf (1972; see also Plumstead, 1965).
Other anatomical features of fossil wood with possible climatic significance are discussed by Noe (1931).
The habit of cauliflory (bearing of reproductive organs directly on the stem) is typical today of many tropical rainforest trees. Although some examples of cauliflory are known for the Paleozoic (see White, 1913, 1931; Noe, 1931), Potonie (1953, as cited in Krassilov, 1975, pp. 118–119) suggested that the habit in Paleozoic plants may have been an adaptation specifically for protection of the reproductive organs against heavy rain, rather than having any temperature significance.
Roots are regarded as among the least sensitive of the major plant organs to climatic variables. There are special cases, however, where they have environmental significance as, for example, in the occurrence of pneumatophores or breathing roots found in swamp plants. The presence of these structures or of subaerial roots of the type found growing flat near the ground surface suggest poorly aerated soil and plants growing in bogs or swamps with a shallow but permanent water cover. The anatomical structures of such roots (i.e., the presence of air chambers) may confirm their function of regulating air supply in a swamp environment. Aerial or stilt roots of the type found among modern mangroves have also been reported for some Paleozoic plants. The complex of anatomical and morphological features found in the roots of Amyelon, a Pennsylvanian cordaitalean, for example, can only be matched among modern mangroves. This conclusion led Cridland (1964, p. 201) to suggest that this gymnosperm must have occupied a similar habitat in “tropical or subtropical saline swamps of sheltered marine shores and estuaries….” Additional examples of the use of roots in interpreting the environment of Carboniferous plants are provided by Potonie (1911), White (1913, 1925, 1931), and Noe (1931).
Luxuriance of growth as predicted from size and abundance of plants has also been used as a basis for climatic conclusions. The small size of plants from the Hermit Shale (Permian, Arizona) reinforces sedimentary data relative to the unfavorable climate under which these plants lived (White, 1929). The stunted appearance of Chester age plants from Illinois and Indiana, together with the general poverty of the flora and the presence of xerophytic characters, led White (1931) to suggest unfavorable conditions of growth. By contrast, the “lush” Carbondale floras of Indiana and Illinois together with the large size of the leaves and trunks were “proof of ample rainfall.” Dilation of tree bases, a condition found among modern swamp plants, also occurs in certain Carboniferous plants (Calamites and Sigillarians) and has been used to reinforce interpretation of their growth in permanent swamps.
An increase in plant size during the Paleozoic (as a general index to luxurious growth) might be interpreted to indicate more favorable conditions for growth relative to the availability of soils, oxygen content of the atmosphere, and insulation from ultraviolet radiation. Chaloner and Sheerin (1979) summarized data related to the maximum observed axis diameter of Late Silurian and Devonian plants. They noted that the potential for tree size was achieved at least by the end of the Devonian. Chaloner and Sheerin also pointed out that more complex plant communities would have been possible as a result of an increase in plant size, because stratification possible with plants of different sizes would have led to “different
micro-environments at different levels of light intensity and humidity.”
During the Early Paleozoic we have the first evidence for the presence of higher, green land plants and, following them, land animals. Although the time interval for the initial advent of land plants remains in question, the earliest known vascular plant megafossils have been found in the Wenlock (early Late Silurian) of Great Britain (Edwards and Feehan, 1980) and possibly in the late Llandovery or Wenlock of North Africa (Boureau et al., 1978); the earliest specimens attributed to nonvascular land plants in the Llandovery (Early Silurian; Pratt et al., 1978) and records of varied land plant microfossils (trilete spores, spore tetrads, and cuticle remains) have been found from Pridoli (latest Silurian) to the Caradocian portion of the Ordovician (Middle Ordovician; Gray and Boucot, 1971, 1978; Gray et al., 1974, 1980). From this information it can be concluded that the land estate was possibly gained as early as the Middle Ordovician. This benchmark has varied meaning to the climatologist with regard to the oxygen budget, CO2 content of the atmosphere, incidence of ultraviolet radiation, soil formation, and other factors.
A great deal is known about the climatic significance of various soil types, although Paleozoic paleosols have largely escaped attention except for studies related to the significance of the mineralogy of Permo-Carboniferous underclays in coal deposits.
Boucot and Gray (1980) synthesized data on paleosols and soil products currently recognized in the Paleozoic, including kaolins and bauxites, as well as gibbsite, boehmite, and emery deposits, the last being the regionally metamorphosed equivalent of alumina-rich soils. Additional data on Paleozoic paleosols is summarized by Retallack (in press).
At present, Paleozoic paleosol data are too limited for independent climatic conclusions, although the data may provide useful constraints. The distribution of Paleozoic calcretes, for example, provides insight into regions of very moderate seasonal rainfall. Such information may be combined with similar data for evaporite and coal distribution to make climatic conclusions more realistic. Pre-Carboniferous Paleozoic calcretes are largely restricted to the Devonian. Dineley (1963) and McKerrow et al. (1974) described Late Ludlovian or Early Pridolian age calcretes from the upper Red Member of the Moydart Formation of Nova Scotia. Allen (1974) described Downtonian (Pridolian) calcretes from England and has documented a number of occurrences of Lower and Upper Devonian calcretes in nonmarine sequences in England, Wales, and Scotland. Dineley and Hickox (1974) described calcareous nodules and conglomerates from the Lower Devonian Knoydart Formation of Nova Scotia that are similar in all respects to the same age calcretes from Britain described by Allen (1974). Woodrow et al. (1973) found similar calcretes in the Middle and Upper Devonian of New York. McPherson (1979) notes Upper Devonian calcrete from Antarctica. D.L. Woodrow (Hobart and William Smith Colleges, personal communication, 1979) mentioned that Brian G. Jones, University of Wollongong, finds Upper Devonian calcretes in Australia. Loope and Schmitt (1980) reported calcrete from the Pennsylvanian of Wyoming. We have recently observed Lower Devonian calcrete in eastern Yunnan Province, south-western China.
Yaalon (The Hebrew University, personal communication, 1979) wrote that “calcrete develops best under semiarid conditions, with an optimum [rainfall] of about 300–350 mm, i.e., when the soil absorbs all the moisture and there is practically no recharge to groundwater. Above 600 mm leaching is too strong and too frequent to form secondary CaCO3. In fully arid regions, calcrete forms slowly and at shallow depths and gyperete is more common, provided sulphate is available… laterite plus bauxite…require much more humid conditions, but…seasonal aridity (about 2 months) is frequently a prerequisite, to enable groundwater level changes and/or the drying of seepages that lead to the irreversible hardening of the accumulated sesquioxides.”
Widespread Lower and Upper Devonian calcretes in the Caledonian Belt of Britain, and from Nova Scotia to New York in the Appalachian Belt, as well as in the Upper Devonian of Antarctica and the Lower Devonian of southwestern China help to extend the arid zone into those regions, although the absence of evaporites in all of them previously provided somewhat ambiguous climatic information. The occurrence of Upper Devonian coal in North America (Heckel and Witzke, 1979) and in Spitsbergen and Bear Island far to the north of the Devonian calcretes indicates a consistent climatic difference.
Possible evidence regarding ancient oceanic circulation patterns of climatic importance can be obtained from the distribution of marine phosphorite deposits. Boucot and Gray (1980) summarized the Paleozoic occurrence of these deposits and discussed some of the principal possibilities for their genesis. There is reasonable evidence that at least a few of the Paleozoic phosphorites, such as those of the Permian in western North America, are best considered to indicate upwelling from deep water on the western margins of mid- to low- but not lowest-latitude major land areas. However, not all phosphorite deposits of the Paleozoic can be regarded as having had such an origin.
Boucot and Gray (1979, 1980) and Crowell (Chapter 6) synthesized the major sources of data pertaining to Paleozoic glacial deposits. The latest Ordovician provides evidence for extensive southern hemisphere Gondwana-region glaciation; so does the Late Carboniferous-Early Permian. The regional limits of the areas affected differ. Evidence for the Late Paleozoic suggests that a number of glacial centers were present but that they were not simultaneously active. No evidence exists for extensive glacial activity at sea level during the Paleozoic at times other than those specified above. It is reasonable to assume, however, that many Paleozoic mountain belts were
glaciated at elevations greater than 1000 in above sea level, although evidence is lacking because of the destruction of high-elevation regions by erosion.
Mountain belts are a common feature for many intervals of the Paleozoic; these belts are now preserved only as their eroded roots. The distribution in time and space of the widespread Paleozoic orogenic belts would provide an additional set of climatic clues. The time interval of orogeny must be accurately dated to know when such orogenic belts presumably had climatologically significant elevations.
Both mountain building and erosion may be relatively rapid geologic processes, each occupying no more than a few million years. Although the average or maximum elevation of Paleozoic mountain belts may be estimated from the volume of debris eroded from them, from their area, and from the interval of time during which the mountain belt was being uplifted and eroded, such estimates are too imprecise to be of value to the climatologist. Nevertheless, orographic effects may be useful in helping to explain some climatological anomalies, and this possibility should be kept in mind.
Summaries of mountain-belt locations are readily available in the literature for the Paleozoic, period by period. The reliability of the data decreases with time. For the Cambrian particularly, there are some large uncertainties in the currently available data.
For many Paleozoic time intervals, relatively detailed information on changing shoreline positions is available. This is the principal data of paleogeographic maps. When synthesized on a global scale, the maps provide a measure of the percentage of the continental areas covered with shallow seawater. These data should provide insight into changing albedo values.
The climatologist should understand that the relative amount of seawater cover on the continents varies from time interval to time interval and from continent to continent. North America, for example, had a large part of its area covered by shallow seas during the Paleozoic. Africa in contrast had a large part of its area above sea level during the Paleozoic. The Early Cambrian had far more overall continental area above sea level than did the mid-Silurian. A time sequence of paleogeographic maps (i.e., Boucot and Gray, 1979, 1980) is necessary, therefore, to evaluate this regression and transgression of the Paleozoic seas from the continents.
There have been several attempts to use sedimentary structures formed by wind as a measure of wind direction of the past. For the later Paleozoic, Poole (1964), Runcorn (1964), Glennie (1972), and Van Veen (1975) provided good summaries of the kinds of data employed. Although aeolian deposits are neither common nor widespread during the Paleozoic, advantage should be taken of the data to provide information on average wind direction, bearing in mind that a large sample is necessary before conclusions have any real significance.
Krinsley and Wellendorf (1980) suggested that microsculpturing on quartz grains provides evidence about the last wind velocity to which the grain was subject. If these data prove reliable, they should provide wind velocities for some Paleozoic aeolian deposits, as well as velocities for “floating” quartz grains found here and there in marine sediments.
SUMMARY OF SILURIAN AND DEVONIAN CLIMATES
We have discussed categories of data that have proved useful, or are potentially useful, in reconstructing Paleozoic climates. We now provide an example of how the data can be used to determine something about the climates of the Silurian and Devonian.
The reconstruction of past climates is intimately involved with the reconstruction of past geographies. We begin with a brief consideration of one possible interpretation of Early Paleozoic paleogeography—the Pangaeic. The climatologist may wish to test this and the many alternative reconstructions for the Paleozoic generated by interest in plate tectonic concepts.
Figures 21.1 and 21.2 present a possible interpretation of Late Silurian and Early Devonian paleogeography within the Pangaeic framework. The canons on which this interpretation is based and the data used to support it have been discussed in detail elsewhere and need not be repeated here (Boucot and Gray, 1979, 1980). We would like to comment, however, about one facet of physical geology that may influence acceptance or consideration of the Pangaea—we refer to the assumption of some geologists that all mountain belts, tectonic zones, and ophiolite occurrences represent suture zones for areas previously separated by thousands of kilometers. Such an assumption often sets up situations that are biogeographically and oceanographically unworkable in terms of oceanic circulation patterns necessary to explain the available data. For example, Scotese et al. (1979) provided a set of period-by-period paleogeographic maps for the Paleozoic that assume that most orogenic and ophiolite belts indicate the location of ancient plate boundaries to either side of which there was a large amount of movement. They also assumed that their selected paleomagnetic data are reliable. Their maps suggested a fragmentation of Asia during the Paleozoic for which good geologic and paleontological evidence to the contrary exists. Chang (1981, p. 184) comments “…Ziegler et al. (1977) placed Chinese Tibet on the equator contiguous with the rest of western China and directly opposite Australia across a seaway. Two years later the same group of authors [Scotese et al. (1979)] placed Chinese Tibet adjacent to India but very distant from China.” No justification for this major change was provided. Their interpretation was also inconsistent with what we now conclude about surface ocean-current circulation based on the biogeography of the Cambrian through the Devonian.
Hall (1980) indicated why one such tectonic belt, located in the Mediterranean region, had nothing to do with significant
seafloor spreading. Similarly, the orogenic belts of later and mid-Paleozoic age situated along the eastern margins of North America (the Appalachian system), and corresponding units of the western margins of Europe (the Caledonide systems in particular, plus some parts of the Hercynian), provide us with little solid data with regard to their Paleozoic geographic relations to each other. Although a water body separated much of eastern North America from western Europe and northern Africa during the Late Silurian and Early Devonian (the “Iapetus” of some writers), the physical data are unconvincing whether it was of the magnitude of the Mediterranean, the Pacific, or something in between. Thus, each belt or zone must be considered on its own merits.
Although we do not regard Figures 21.1 and 21.2 as maps in the strict sense, we believe that they provide a more rational synthesis of the varied pre-Carboniferous lithofacies and biofacies data on which they are principally based than do other types of pre-Carboniferous paleogeographic reconstructions. Engel and Kelm (1972) employed a similar Pangaeic reconstruction for the Precambrian, although basing it on data different than ours because of the earlier time interval. In addition to biogeography and lithofacies information, our re-
constructions outline surface oceanic circulation patterns that will reproductively connect the same biogeographic unit present in two or more areas, i.e., surface current patterns that are consistent with the present known biogeographic data.
For the Silurian, the chief source of climatic information, in addition to biogeography, is the distribution of evaporites and carbonate (including reefs) and noncarbonate rock sequences. The carbonate and reef facies, denoting tropical to warm temperate conditions, are confined to the North Silurian Realm. Silurian coals have not been recognized. Paleosol data of climatic significance are limited: calcretes, indicative of semiarid or seasonally arid regions, are confined to the latest Silurian where their distribution complements known areas of major Late Silurian evaporites. The significance of phosphorites is too poorly understood within the Silurian to employ them as clues or even as definite indicators of oceanic circulation.
Paleoclimatic data for the Devonian are more varied. Figure 21.2 shows biogeographic subdivisions for the Early Devonian as well as some lithofacies data. As in the Silurian, carbonate rocks continue to be absent from the Malvinokaffric Realm. For the Late Devonian the distribution of coals (humid indicators) is shown latitudinally separated from that of evaporites (arid indicators) in North America. Similar data are available for the Devonian of Eurasia and Australia for both the Middle and Late Devonian (Oswald, 1968). As in the Silurian, the distribution of Devonian calcretes complements the distribution of evaporites. Lacustrine beds in northern Scotland yield authigenic aegirine compatible with a semiarid, possibly seasonal climate (Fortey and Michie, 1978). These data are consistent with low-latitude humid climates and middle- to high-latitude dry or seasonally dry climates. Paleosols of seasonally humid-arid implications (bauxites, kaolin, and other lateritic products) are prominent in the Uralian region and are also found in Siberia and Iran. Their distribution implies a regional climatic anomaly inconsistent with a latitudinal, completely parallel distribution of humid climate in low and middle latitudes and relatively arid and seasonal climate in higher latitudes.
In sum, lithofacies and biofacies evidence for most of the Silurian and Devonian indicate the presence of a major southern hemisphere high-latitude region (the Malvinokaffric Realm) lacking carbonate rocks, reefs, redbeds, evaporites, coal deposits, and other evidences of warm climate. During the Silurian and Devonian there are widespread evaporite belts deduced to have been present in low (but not lowest) to middle latitudes. These are best preserved during the Late Silurian, the Middle Devonian, and the Frasnian (lower half of the Late Devonian) but are poorly represented in both the Early Silurian and Early Devonian. Varied paleosols geographically complement the distribution of evaporites for the most part. Silurian-Devonian redbeds are widespread in regions ringing the Malvinokaffric Realm in consort with other evidences of warm climate. Coal deposits are present in the Late Devonian in regions deduced to represent low latitudes.
During most of the Silurian and Devonian the climate may be characterized as having had a high latitudinal gradient, less than during intervals of extensive glaciation such as the Permian and Quaternary but higher than that of the Lower Carboniferous and much of the Mesozoic. Continental glaciation in the southern hemisphere also marks the end of the Ordovician (the Ashgillian) as a time of highest climatic gradient. The absence of positive evidence for continental glaciation in the Silurian and Devonian leads us to conclude that the overall climatic gradient was lower than in the Late Ordovician. Carozzi (1979) mentioned Upper Devonian tillite on the north side of the Amazon Basin, but no paleontologic evidence is provided to document that this tillite could not be of PermoCarboniferous age. Sometime during the Late Devonian, however, the climate of the high-latitude Malvinokaffric Realm appears to have ameliorated, and the Realm no longer existed as a biogeographic unit. These changes are consistent with the presence of a globally low climatic gradient similar to that which characterized the Early Carboniferous on a global scale.
SUMMARY AND CAVEATS
To understand Paleozoic climates, detailed time-sequence maps are necessary on which shoreline positions for the continents are indicated. On these should be plotted the boundaries of biogeographic units, lithologic data (distribution of redbeds, carbonate-noncarbonate rock sequences, evaporite and coal deposits, paleosol and glacial deposits, and phosphorites of the upwelling type), mountain belts, and biological information of environmental significance. Only then can climatic deductions and constraints be suggested for the Paleozoic. Maps such as Figures 21.1 and 21.2 and those by Boucot and Gray (1979, 1980) are a beginning but include only a fraction of the usable data. Maps for the Devonian by Heckel and Witzke (1979) are well-documented attempts to synthesize more varied climatic indicators, with emphasis on sedimentary data.
Probably the chief warning to the climatologist attempting to reconstruct Paleozoic climate and climatic events is to employ as short a time interval as possible but one lengthy enough to provide sufficient global data to permit informed speculation. For the Paleozoic, time intervals of 10–15 m.y. are usable, and it is improbable that units much shorter than that can be synthesized in the near future. In geologic terms this means that a geologic Period, measured in tens of millions of years, is commonly too large a unit for which to collect meaningful data because there is good evidence that globally significant shifts in the position of climatic belts, as well as changes in the global climatic gradient, have taken place during these lengthy intervals. For example, White (1931) contrasted the plant remains of the Late Mississippian (Chester) with those of the earlier Pennsylvanian of the Eastern Interior, noting that the former have a “generally stunted” appearance and “more or less distinctly xerophytic characters,” whereas the general luxuriance of the Pennsylvanian vegetation together with its large size and other features clearly indicates not only equable climates but ample rainfall. In the Carboniferous of the Northern Appalachian region (New England through Newfoundland) the Early Carboniferous (Mississippian) is characterized by the deposition of evaporites such as
those of the Windsor Group, whereas the Late Carboniferous (Pennsylvanian) was sufficiently humid to permit the accumulation and preservation of enough plant material to form economically important coal beds. In a recent publication Habicht (1979) confused this paleoclimatic evidence by plotting evaporites and coals together on the Carboniferous map of the Northern Appalachian region without regard for the fact that they accumulated at different times. Habicht (1979) presented a similar anomaly for the Cambrian of North Africa, where the Early Cambrian includes many marine carbonate beds, indicators of warm climatic regime, whereas the Middle and Late Cambrian lacks carbonates and is characterized by a cool climate marine fauna. We have pointed out earlier (Boucot and Gray, 1979, 1980) that in the Antarctic Devonian, the Early Devonian is of cold climate, Malvinokaffric Realm type, whereas the Late Devonian is of warm type and even includes a recently discovered calcrete (McPherson, 1979).
Recognition that the time correlation of the features being discussed here have different values is crucial. Marine carbonate rocks rich in fossils can probably be correlated with a precision within 2–3 m.y. The dating of nonmarine bauxites and varied paleosols may be much more approximate and in some instances no more precise than 10–15 m.y. Therefore, the climatologist should continually consult with the biostratigrapher-paleontologist to make certain about the reliability of the correlation and the precision of the data being employed.
Allen, J.R.L. (1974). Sedimentology of the Old Red Sandstone (Siluro-Devonian) in the Clee Hills Area, Shropshire, England, Sed. Geol. 12, 73–167.
Antevs, E. (1925). The climatologic significance of annual rings in fossil woods, Am. J. Sci. 9 (5th series), 296–302.
Arnold, C.A. (1947). An Introduction to Paleobotany, McGraw-Hill, New York, 433 pp.
Boucot, A.J. (1975). Evolution and Extinction Rate Controls, Elsevier, New York, 427 pp.
Boucot, A.J., and J.Gray (1979). Epilogue: A Paleozoic Pangaea? in Historical Biogeography, Plate Tectonics, and The Changing Environment, J.Gray and A.J.Boucot, eds., Oregon State U. Press, Corvallis, pp. 465–482.
Boucot, A.J., and J.Gray (1980). A Cambro-Permian pangaeic model consistent with lithofacies and biogeographic data, in The Continental Crust and its Mineral Deposits, D.W.Strangway, ed., Geol. Assoc. Canada Spec. Pap. No. 20, pp. 389–419.
Boureau, E., A.Lejal-Nicol, and D.Massa (1978). A props du Silurien et du Devonien en Libye. Il faut reporter au Silurien la date d’apparition des plantes vasculaires, C.R. Acad. Sci. Paris 286, ser. D, pp. 1567–1571.
Carozzi, A.V. (1979). Petroleum geology in the Paleozoic clastics of the Middle Amazon Basin, Brazil, J. Petrol. Geol. 2, 55–74.
Chaloner, W.G., and G.T.Creber (1973). Growth rings in fossil woods as evidence of past climates, in Implications of Continental Drift to the Earth Sciences, D.H.Tarling and S.K.Runcorn, eds., Academic, New York, pp. 425–437.
Chaloner, W.G., and A.Sheerin (1979). Devonian macrofloras, in The Devonian System, M.R.House, C.T.Scrutton, and M.G. Bassett, eds., Palaeontol. Assoc. Spec. Pap. in Palaeontol. No. 23, pp. 145–161.
Chang, W.T. (1981). On the northward drift of the Afro-Arabian and Indian Plates, R. Soc. Victoria Proc. 92, 181–185.
Cridland, A.A. (1964). Amyelon in American coal-balls, Palaeontology 7, 186–209.
Daubenmire, R.F. (1974). Plants and Environment: A Textbook of Plant Autecology (3rd ed.), Wiley, New York, 422 pp.
Davis, C.A. (1913). Origin and formation of peat, in The Origin of Coal, D.White and R.Thiessen, Bureau of Mines Bull. 38, pp. 165–186.
Dineley, D.L. (1963). The “Red Stratum” of the Silurian Arisaig Series, Nova Scotia, Canada, J. Geol. 71, 523–524.
Dineley, D.L., and C.F.Hickox (1974). Arisaig Group: Knoydart Formation, in Geology of the Arisaig Area, Antigonish County, Nova Scotia, Geol. Soc. Am. Spec. Pap. 139, pp. 68–76.
Dolph, G.E., and D.L.Dilcher (1979). Foliar physiognomy as an aid in determining paleoclimate, Palaeontographica 170, 151–172.
Edwards, D., and J.Feehan (1980). Records of Cooksonia-type sporangia from late Wenlock strata in Ireland, Nature 287, 41–42.
Engel, A.E.J., and D.L.Kelm (1972). Pre-Permian global tectonics: A tectonic test, Geol. Soc. Am. Bull. 83, 2325–2340.
Fortey, N.J., and U.McL.Michie (1978). Aegirine of possible authigenic origin in Middle Devonian sediments in Caithness, Scotland, Mineral. Mag. 42, 439–442.
Friend, P.F., and M.Moody-Stuart (1970). Carbonate sedimentation on the river floodplains of the Wood Bay Formation (Devonian) of Spitsbergen, Geol. Mag. 107, 181–195.
Glennie, K.W. (1972). Permian Rotliegendes of northwest Europe interpreted in light of modern desert sedimentation studies, Am. Assoc. Petrol. Geol. Bull. 56, 1048–1071.
Goldring, W. (1921). Annual rings of growth in Carboniferous woods, Bot. Gaz. 72, 326–330.
Gray, J., and A.J.Boucot (1971). Early Silurian spore tetrads from New York: Earliest New World evidence for vascular plants? Science 173, 918–921.
Gray, J., and A.J.Boucot (1978). The advent of land plant life, Geology 6, 489–492.
Gray, J., and A.J.Boucot, eds. (1979). Historical Biogeography, Plate Tectonics, and The Changing Environment, Oregon State U. Press, Corvallis, 500 pp.
Gray, J., S.Laufeld, and A.J.Boucot (1974). Silurian trilete spores and spore tetrads from Gotland: Their implications for land plant evolution, Science 185, 260–263.
Gray, J., D.Massa, and A.J.Boucot (1980). Caradocian land plant microfossils from Libya: The oldest known land plant spores, Geol. Soc. Am. Abstr. Programs 12, p. 436.
Habicht, J.K.A. (1979). Paleoclimate, Paleomagnetism, and Continental Drift, Am. Assoc. Petrol. Geol. Studies in Geology No. 9, Tulsa, Okla., 31 pp.
Hall, R. (1980). Contact metamorphism by an ophiolite peridotite from Neyriz, Iran, Science 208, 1259–1262.
Hallam, A., ed. (1973). Atlas of Palaeobiogeography, Elsevier, New York, 531 pp.
Heckel, P.H. (1977). Origin of phosphatic black shale facies in Pennsylvanian cyclothems of mid-continent North America, Am. Assoc. Petrol Geol. Bull. 61, 1045–1068.
Heckel, P.H., and P.J.Witzke (1979). Devonian world palaeogeography determined from distribution of carbonates and related lithic palaeoclimatic indicators, in The Devonian System, M.R.House,
C.T.Scrutton, and M.G.Bassett, eds., Palaeontol. Assoc. Spec. Pap. in Paleontol. No. 23, pp. 99–123.
Hughes, N.F., ed. (1973). Organisms and Continents Through Time, Palaeontol. Assoc. Spec. Pap. in Palaeontol. No. 12, 334 pp.
Krassilov, V.A. (1975). Paleoecology of Terrestrial Plants: Basic Principles and Techniques, Wiley, New York, 283 pp. (translated from Russian by H. Hardin).
Krausel, R. (1964). Introduction to the palaeoclimatic significance of coal, in Problems of Palaeoclimatology, A.E.M.Nairn, ed., Wiley-Interscience, New York, pp. 53–56, 73–74.
Krinsley, D., and W.Wellendorf (1980). Wind velocities determined from the surface textures of sand grains, Nature 283, 372–373.
Loope, D.B., and J.G.Schmitt (1980). Caliche in the Late Paleozoic Fountain Formation: Rediscovery and implications, Geol. Soc. Am. Abstr. Programs 12, p. 473.
McKerrow, W.S., A.J.Boucot, and C.F.Hickox (1974). Arisaig Group: Beechhill Cove Formation, Ross Brook Formation and others, in Geology of the Arisaig Area, Antigonish County, Nova Scotia, Geol. Soc. Am. Spec. Pap. 139, pp. 34–68.
McPherson, J.G. (1979). Calcrete (caliche) palaeosols in fluvial red-beds of the Aztec Siltstone (Upper Devonian), southern Victoria Land, Antarctica, Sed. Geol. 22, 267–285.
Meyerhoff. A.A. (1970). Continental drift-. Implications of paleomagnetic studies, meteorology, physical oceanography, and clima-tology, J. Geol. 78, 1–51.
Middlemiss, F.A., P.F.Rawson, and G.Newall, eds. (1971). Faunal Provinces in Space and Time, Seel House Press, Liverpool, 236 pp.
Nicol, D. (1967). Some characteristics of cold-water marine pelecypods, J. Paleontol. 41, 1330–1340.
Noe, A.C. (1931). Evidences of climate in the morphology of Pennsylvanian plants, State of Ill., State Geol. Surv. Bull. 60, 282–289.
Oswald, D.H., ed. (1968). International Symposium on the Devonian System, Calgary 1967, Calgary, Alberta Soc. Petrol. Geol., Vol. 1, 1055 pp.; Vol. 2, 1377 pp.
Plumstead, E.P. (1963). Palaeobotany of Antarctica, in Antarctic Geology, SCAR Proc., XI Palaeontology, pp. 637–654.
Plumstead, E.P. (1965). Glimpses into the history and prehistory of Antarctica, Antarktiese Bull. 9, 1–5.
Poole, F.G. (1964). Palaeowinds in the western United States, in Problems in Paleoclimatology, A.E.M.Nairn, ed., Wiley-Interscience, New York, pp. 394–405, 423–424.
Potonie, H. (1911). Die Tropen-Sumpfflachmoor-Natur der Moore des Productiven Karbons, Kgl. Preuss. geol. Landesanst. Jahrb. (1909) 30, 389–443.
Potonie, R. (1953). Zür Palaobiologie der karbonischen Pflanzenwelt, Naturwissenschaften 40, 119–128.
Pratt, L.M., T.L.Phillips, and J.M.Dennison (1978). Evidence of non-vascular land plants from the Early Silurian (Llandoverian) of Virginia, U.S.A., Rev. Palaeobot. Palynol. 25, 121–149.
Retallack, G. (in press), Fossil soils—indicators of ancient terrestrial environments, in Evolution, Paleoecology and the Fossil Record, K. Niklas, ed., Praeger, New York.
Richards, P.W. (1979). The Tropical Rain Forest: An Ecological Study, Cambridge U. Press, Cambridge, 450 pp.
Ross, C.A., ed. (1974). Paleogeographic Provinces and Provinciality, Soc. Econ. Paleontol. Mineral. Spec. Publ. No. 21, 233 pp.
Runcorn, S.K. (1964). Paleowind directions and palaeomagnetic latitudes, in Problems in Palaeoclimatology, A.E.M.Nairn, ed., Wiley-Interscience, New York, pp. 409–421, 424.
Schopf, J.M. (1972). Coal, climate and global tectonics, in Implications of Continental Drift to the Earth Sciences, Vol. 1, D.H. Tarling and S.K.Runcorn, eds., Academic, New York, pp. 609–622.
Scotese, C.R., R.K.Bambach, C.Barton, R.Van der Voss, and A. M.Ziegler (1979). Paleozoic base maps, J. Geol. 87, 217–277.
Stehli, F.G. (1968). Taxonomic diversity gradients in pole location; The Recent model, in Evolution and Environment, E.T.Drake, ed., Yale U. Press, New Haven, Conn., pp. 163–227.
Stehli, F.G., A.L.McAlester, and C.E.Helsley (1967). Taxonomic diversity of Recent bivalves and some implications for geology, Geol. Soc. Am. Bull. 78, 455–466.
Tomlinson, P.B., and F.C.Craighead (1972). Growth-ring studies on the native trees of sub-tropical Florida, in Research Trends in Plant Anatomy (Chowdhury Commemorative Volume), Ghouse and Yunus, eds., McGraw-Hill, New York, pp. 39–51.
Van Veen, F.R. (1975). Geology of the Lemen gas-field, in Petroleum and the Continental Shelf of Northwest Europe, Vol. 1, A.W. Woodland, ed., Halsted, New York, pp. 223–231.
Walker, T.R. (1967). Formation of red beds in modern and ancient deserts, Geol. Soc. Am. Bull. 78, 353–368.
Walker, T.R. (1975). Red beds in the western interior of the United States, in Paleotectonic Investigations of the Pennsylvanian System in the United States, Pt. II, L.D. McKee, E.J.Crosby, and others, U.S. Geol. Surv. Prof. Pap. 853, pp. 593–606.
Wang Yu, A.J.Boucot, Rong Jia-Yu, and Yang Xue-Chang (in press). Silurian and Devonian biogeography of China, Geol. Soc. Am. Bull.
White, D. (1913). Physiographic conditions attending the formation of coal, in The Origin of Coal, D.White and R.Thiessen, Bur. of Mines Bull. 38, pp. 52–84.
White, D. (1925). Environmental conditions of deposition of coal, Trans. Am. Inst. Min. Metal. Eng. 71, 3–34.
White, D. (1929), Flora of the Hermit Shale, Grand Canyon, Arizona, Carnegie Institution of Washington Publ. No. 405, 119 pp.
White, D. (1931). Climatic implications of Pennsylvanian flora, State of Ill., State Geol. Surv. Bull. 60, 271–281.
Woodrow, D.L., F.W.Fletcher, and W.F.Ahrnsbrak (1973). Paleogeography and paleoclimate at the depositional sites of the Devonian Catskill and Old Red facies, Geol. Soc. Am. Bull. 84, 3051–3064.
Ziegler, A.M., K.S.Johnson, M.E.Kelly, C.R.Scotese, and R.Van der Voo (1977). Silurian continental distributions, paleogeography, climatology, and biogeography, Tectonophysics 40, 13–51.