Odekon M. Encyclopedia of paleoclimatology and ancient environments

Подождите немного. Документ загружается.

RADIOCARBON DATING

Radiocarbon (

C) dating, now in its fifth decade of routine

use, remains the most widely employed method of inferring

chronometric age for organic materials from the late Pleisto-

cene and Holocene. It provides the principal time scale for

reconstruction of the history of late Quaternary environments,

including the temporal scale for climate proxy records, and

documents chronometric relationships for prehistoric human

cultures on a world-wide basis.

Radiocarbon dating model

The natural production of

C is a secondary effect of cosmic-

ray interactions with atmospheric gas molecules, with the resul-

tant production of neutrons (Figure R1). Most

C is formed

by the reaction of neutrons with

N. It is then rapidly oxidized

to form

. In this form,

C is distributed throughout the

Earth’s atmosphere by stratospheric winds, becoming generally

well-mixed by the time

C-tagged CO

molecules reach

the Earth’s surface. Most

C is absorbed in the oceans, while

1–2% becomes part of the terrestrial biosphere, primarily by

means of photosynthesis. Plant materials and animals that are,

directly or indirectly, dependent on plants, are thus all labeled

with

Metabolic processes in living organisms maintain the

content in approximate equilibrium with atmospheric

C con-

centrations. While

C decays in living tissue, it is continually

replaced through the biochemical processing of plant or animal

tissue. Once metabolic processes cease, as at the death of

an animal or plant, the amount of

C begins to decrease by

radioactive decay – in the case of

C, by beta decay – at a rate

measured by the

C half-life (Taylor, 1987, 1997a).

The radiocarbon age of a sample is based on a measurement

of its residual

C content. For the

C age of a sample to be

equivalent to its true or solar/calendar age at a reasonable level

of precision, a set of assumptions needs to be satisfied. These

assumptions are that: (a) the

C ratio in materials of zero

age in each carbon reservoir has remained essentially constant

over the entire

C time scale, (b) complete and rapid mixing

C occurs throughout the various carbon reservoirs on a

worldwide basis, (c) carbon isotope ratios (e.g.,

C as well

C) in samples have not been altered except by

decay since the death of an organism, (d) the half-life of

is accurately known with reasonable precision, and (e) natural

levels of

C can be measured to appropriate levels of accuracy

and precision (Taylor, 1996).

A significant portion of contemporary

C studies involves

two major types of efforts. First, there are investigations

designed to examine and compensate for the effects of viola-

tions of the primary assumptions as applied to specific sample

types or portions of the carbon reservoirs. Second, there are

efforts to document explicitly the physical association between

the organic sample on which a

C age estimate is obtained

and the object, feature, or geological context for which an

age determination is desired.

Radiocarbon age estimates are generally expressed in terms of

a set of characteristic parameters that define a conventional radio-

carbon age. These parameters, introduced in the mid-1970s and

explicitly defined by Stuiver and Polach (1977), include (a) the

use of 5,568 (30) years as the

C half-life (8,033 year mean life)

used to calculate age, although the actual half-life value is prob-

ably closer to 5,730 (40) years; (b) the use of one of the United

States National Institute of Standards and Technology (formerly

U.S. National Bureau of Standards, NBS)-distributed oxalic acid

preparations – or a secondary standard with a known relationship

to the primary standard – as a contemporary or modern reference

standard to define a “zero”

C age; (c) the use of AD 1950 as the

zero point from which to measure

C time; (d) to account for frac-

tionation effects, a normalization of the measured

C content in

all samples to a common

C(d

C) value of –25% (wrt

PDB standard); and, finally, (e), an assumption that

C ratios

in all reservoirs have remained constant over the

C time scale. In

addition, each

C determination is expected to be accompanied

by an expression that provides an accurate estimate of the overall

experimental or analytical uncertainty. Since counting statistics

associated with the measurement of

C concentrations in samples

are usually the dominant component of the analytical uncertainty,

this value is informally referred to as the “statistical error .”

This “” term is suffixed to all appropriately documented

Cage

estimates and is typically expressed as  one standard deviation

(1 s).

For samples from some carbon reservoirs, conventional

contemporary standards may not define a zero

C age. A reser-

voir corrected radiocarbon age can sometimes be calculated by

documenting the apparent age exhibited in control samples and

correcting for the observed deviation. Reservoir effects are

most often observed in samples from fresh water lakes and

marine environments. A calibrated radiocarbon age takes into

consideration the fact that

C activity in living organisms,

which in most cases directly reflects atmospheric

C levels,

has not remained constant over the

C time scale because of

changes in

C production rates or reservoir mixing parameters

of the carbon cycle. The study of the various factors responsi-

ble for the variability in

C production rates and changes

in the world-wide distribution of

C in various carbon reser-

voirs over the late Pleistocene and Holocene has occupied the

attention of many in the

C research community for over four

decades (Taylor, 2001). The various issues raised by the nature

of the variability over time in the

C time scale – with the

documentation of this variability now extended into the late

Pleistocene – will be discussed in greater detail in the next

section.

In the current nomenclature of the journal Radiocarbon

“

C years BP” is expressed only as “BP” with the “

C years”

implied (e.g., 2,510  50

BP). Calibrated ages, i.e., calendar

or solar years, are expressed with only the “cal” (calibrated)

designation attached (e.g., cal

AD 1520, 3,720 cal BC), with

the “years” implied. The journal American Antiquity adopted

the nomenclature of Radiocarbon except for punctuation

(

B.P. rather than BP). In the early 1970s, the journal Antiquity

adopted a terminology which distinguished conventional (unca-

librated) and calibrated

C values whereby “bp” and “ad/bc”

are employed to designate conventional

C values while “BP”

and “

AD /BC” are used to designate calibrated

C values.

The

C time scale now extends from about 300 years to

between 40,000 and 60,000 years. The limitations of using

Figure R1 Radiocarbon dating model: production, distribution, and decay of

C. Figure adapted from Taylor (1997a: Fig. 3.1).

864 RADIOCARBON DATING

the

C time scale over the last few hundred years are a conse-

quence of the complex interplay of various recent natural

deviations in

C content (informally referred to variously as

“warps” or “wiggles” or more formally as “de Vries effects,”

see below) and recent anthropogenic (e.g., “fossil fuel” and

“bomb”

C) effects that have also caused recent deviations in

C content. The maximum

C ages that can be inferred using

decay counting technology depend on characteristics of differ-

ent laboratory instrumentation and experimental configurations –

e.g., background and sample blank values, counter size, length

of counting – and, to some degree, the amount of sample avail-

able for analysis. Currently, for laboratories with counters

designed to work with older samples, routine upper limits using

moderate sample sizes (3–5 grams of carbon) range upward

to about 55,000 years. Employing relatively large sample sizes

(>15 grams of carbon), a few laboratories have developed

the capability to obtain age estimates up to about 70,000 years.

With isotopic enrichment – again using relatively large sample

amounts – ages up to 75,000 years have been reported on a

small number of samples (Grootes et al., 1975; Erlenkeuser,

1979). There are efforts now underway to exploit accelerator

mass spectrometry (AMS) technology (see below) to extend

the

C time scale to as much as 70,000–90,000 years using

sample weights of less than a gram of carbon.

Development of radiocarbon dating studies

Stages in the history of the development of

C dating can

be distinguished partly by the type of detection technology

employed and, partly, on understandings concerning the rela-

tionship between radiocarbon and “real” or solar/calender time.

The first generation of

C studies began with the appearance of

the first

C date list in 1950. In the early 1970s, second genera-

tion studies began with the recognition of the existence of sys-

tematic temporal offsets in the

C time scale (Suess, 1970).

In the late 1970s, third generation

C studies were ushered in

by a proposal (Muller, 1977) to develop a new approach to

C measurement by AMS, although routine operation of

AMS technology was not achieved until the mid-1980s (Linick

et al., 1989).

Several recent technical and methodological advances have

provided contexts for a series of significant applications of

the

C method in a number of different disciplines. In addition

to the application of AMS technology, these include the exten-

sion of the calibrated

C time scale into the late Pleistocene

and a more detailed characterization of Holocene de Vries

effects where relatively rapid changes in atmospheric

C con-

centrations complicate the inherent precision with which

content can be converted into inferred age. The extension

of the calibration of the

C time scale beyond that provided

with tree ring data has been accomplished primarily by the

use of AMS-based

C and TIMS (thermal ionization mass

spectrometry)-based uranium-thorium (U-series) measurements

on marine coral samples.

Extending the ca libration of the

C time scale

Comparisons of

C with dendrochronological data based on

Irish oak and German oak and pine for the earliest portion

and Douglas fir, sequoia and bristlecone pine for later periods

now document about 12,000 years of solar time (cal

BP)

(Stuiver et al., 1998a; Reimer et al., 2002). Most of the

measurements comprising these data sets are characterized as

“high-precision,” referring to carefully measured counting

uncertainties typically at the  1s level of <20 years for the

C values used to provide the dendrochronologically-based

calibration data. The validity of the stated uncertainties has

been extensively examined by detailed interlaboratory compar-

isons among a number of the laboratories producing the cali-

bration data (e.g., Scott et al., 1990). For the period before

dendrochronological controls were available, paired uranium/

thorium (

234

230

Th) and

C samples from cores drilled into

coral formations (Bard, 1998; Edwards et al., 1993) provide

the principal data on which a late Pleistocene

C calibration

curve has been provisionally extended to 24,000 cal

BP or, as

expressed in

C time, to 20,300 BP (Bard et al., 1998).

With the extension of the

C calibration framework using

the uranium-series data on corals, it has now become apparent

that what was first thought to be a “sine-wave”-like characteri-

zation of the Holocene

C calibration curve was an artifact of

the limited time frame documented by the original middle and

late Holocene tree-ring/

C data set. Figure R2 represents a

plot of the off-set between

C and assumed “true” ages out

to about 30,000 years based primarily on the most current com-

bined dendrochronological/

C and coral uranium-series/

data, as presented in the INTCAL98 data set (Stuiver et al.,

1998b). Based on the expanded calibration data, it now appears

that the long-term secular variation

C anomaly over this per-

iod can be characterized as representing a slow-decay function

on which has been superimposed middle- and short-term de

Vries effects perturbations.

For the period before about 30,000 cal

BP, age estimates

obtained from

C and other Quaternary dating methods applied

to samples obtained from a variety of contexts are, to varying

degrees, inconsistent with regard to indications of the magni-

tude of the

C age offsets. For example, comparisons between

C values (adjusted by 3% to take into consideration the

most likely

C half-life) and thermoluminescence age esti-

mates on materials from hearths at Lake Mungo, Australia sug-

gest deviations in the range of 3,500–5,000 years between

27,000 and 31,000

BP (Bell, 1991). Studies of variations in

the intensity of the Earth’s dipole magnetic field – a major

cause of the long-term

C secular variation trend – as well as

comparisons of precisely determined

C and associated

Ar/

Ar values from volcanic deposits, suggest somewhat

lower age offsets for this period. However, the geomagnetic

data suggest an increasing

C/solar time offset in the range

of 1,500–2,700 years between 25,000 and 40,000 years but

predict that there will be good agreement between

C and solar

time at about 45,000–50,000 years (Mazaud et al., 1992;

Southon et al., 1995). In addition, there also may have been a

major “spike” in atmospheric

C concentrations between

30,000 and 35,000

BP (Voelker et al., 1998: Fig. 6). Further

detailed studies will be necessary to clarify the magnitude of

the offsets in

C activity for this period (Geyh and Schlüchter,

1998; Reimer et al., 2002).

Holocene de Vries effects

A fuller rendering of the entire Holocene

C time scale permits

researchers to review more precisely the timing and character-

istics of the de Vries effect “wiggles” over the last ten millen-

nia. Figure R3 plots the series of defined Holocene “time

warps” reflecting the time ranges of the de Vries effect pertur-

bations in the calibration curve (Taylor et al., 1996: Figs. 3A

and 3B). The top panel of Figure R3 represents 0–5,000

and the bottom panel the 5,000–10,000 BP period in

C years.

RADIOCARBON DATING 865

The inherent uncertainties and fluctuations in the calibration

curve cause discrete calibrated ages to broaden into ranges,

even when, as in this figure, the

C ages are plotted with a

hypothetical zero variance. Thus,

C age determinations deriv-

ing from time periods with larger calibration range values will

be inherently less precise. For example, on average, calibrated

age values for 8,000–10,000

BP are inherently much less

precise than those in the 0–2500

BP period because of the

large

C de Vries effect “time warps” occurring during in those

periods.

In Figure R3, 12 major and 5 intermediate de Vries effect

perturbations can be identified. Major de Vries effects have

been defined as “time warps” exceeding 250 cal years; inter-

mediate de Vries effects are those that exhibit ranges in excess

of 140 cal years, and minor de Vries effects have ranges of less

than 140 cal year. The 17 major and intermediate de Vries per-

turbations have been assigned Roman numeral and letter com-

binations. Roman numerals identify the

C millennium, i.e.,

I=0–1,000

BP, II = 1,000–2,000 BP, while lower case letters

identify the perturbation in chronological order within each

C millennial period. It should be noted that the INTCAL98

calibration data set shifts the time scale for early Holocene

C time warps slightly but does not affect their magnitude.

Various strategies to increase the precision of

C-based age

inferences in light of the variable effects of de Vries-type secu-

lar variations include the use of Bayesian statistical approaches

to calibration procedures (Buck and Christen, 1998) and

“wiggle-matching,” in which the known pattern of the de Vries

perturbations are matched on an unknown sample of wood

containing a sequence of tree rings by measuring

C activity

in a set of adjacent tree rings in that sample and matching that

pattern with the known de Vries pattern. An example of the lat-

ter approach includes the determination of a cutting date for a

timber from a Japanese tomb to

AD 320 with a precision

of 5 years (Kojo et al., 1994). When precision at the level

of <30 years can be statistically justified, concerns about

latitude-dependent differences in

C atmospheric concentra-

tions become relevant. Suggestions of the magnitude of the

offset between Northern and Southern Hemispheric samples

currently range between 24  3 (Stuiver et al., 1998b)to

40  13

C years (Hogg et al., 2002 ).

Accelerator mass spectrometry

From the initiation of

C studies until the late 1970s, the

basis of inferring

C concentrations, and thus the

C age of

samples, exclusively employed decay counting technology.

In decay counting, isotopic concentrations are measured by

counting decay events in an ionization or scintillation detector

and comparing the count rate observed in an unknown-age

sample to that exhibited by appropriate standards under a com-

mon set of experimental conditions. For

C, this involves

counting beta particles, i.e., negatively charged electrons

emitted from the

C nucleus. In decay counting, a relatively

small fraction of the

C atoms present in a carbon sample are

actually detected over the course of measurement. While there

are approximately 5.9  l0

atoms of

C in 1 gram of modern

“pre-bomb” carbon, on average, over a one-minute period,

less than 14 of these atoms will decay and be available for

Figure R2 Late Pleistocene and Holocene deviation of

C from dendrochronologial-, uranium-series-, and varve counted marine sediment-based

age data (Data from Stuiver et al., 1998a; figure adapted from Taylor 2001: Fig. 2.2).

866 RADIOCARBON DATING

detection. In large part, it was this inherent inefficiency in

decay counting that gave impetus to efforts to develop direct

or ion counting technology using a form of mass spectrometry.

Within a year of the initial publication of the concept under-

lying AMS technology, the first published AMS-based

determination was obtained using the 88-inch (224-cm) cyclo-

tron at the University of California Lawrence Berkeley Labora-

tory (Muller, 1977). Because of technical problems with using

cyclotron-based AMS systems for natural level measurements,

most routine AMS measurements are currently undertaken

on another type of particle accelerator, an electrostatic tandem

accelerator. Although initially HEMS (high energy mass

spectrometry) was used to refer to this technology, typically,

TAMS (tandem accelerator mass spectrometry) is now used.

“Tandem” here refers to the fact that particle acceleration in

electrostatic systems is accomplished in a two-step “pull-push”

process. The TAMS approach proved to be a practical

AMS technology for a wide range of rare, cosmogenic isotopes

including

C and, to date, most AMS

C age determinations

have been obtained on a TAMS instrument. By way of a histor-

ical footnote, the question of who first conceived the idea

that led directly to the development of AMS technology

is the subject of continuing discussion (Gove, 1996; 1999;

Taylor, 1997b).

Three advantages of AMS technology in the measurement

C were anticipated as a result of the greatly enhanced

detection efficiency. First, major reductions in sample sizes

would be possible – from gram amounts of carbon to milligram

amounts and, with additional efforts, to the level of less than

100 micrograms. Second, major reductions in counting times

would be possible. Reductions from several days for conven-

tional systems and even weeks and months with micro- and

mini-decay counting system could be achieved with several

minutes of counting for AMS systems to achieve 1% count-

ing statistics. Finally, it was anticipated that significant

increases in the applicable dating time frame would be

possible – from the currently routine 40,000–50,000 years to

as many as 100,000 years (Muller, 1977; Gove, 1992).

The first two of the three originally-anticipated benefits of

AMS technology – major reductions in sample size and count-

ing times – have been fully realized over the last decade

(Taylor et al., 1984; Taylor, 1991; Hedges, 1995). For both

sample sizes and counting times, order-of-magnitude reductions

have been made possible on a routine basis. However, the pro-

jected third advance has not occurred due to the inability, at

present, to exclude microcontamination of samples during sam-

ple preparation. The source of a significant portion of this con-

tamination results from the current requirement in most

laboratories that samples must be converted to graphitic carbon

for use in the ion source of an AMS system. Parts per million

of modern carbon contamination translates into background

levels which generally limit the maximum ages that can be

resolved to between 40,000 and 50,000 years. One laboratory

has developed a CO

gas source but reports similar background

values (Bronk and Hedges, 1987). To date, the oldest age

reported on a graphitized background blank (Pliocene wood)

is 60,500

BP (Kirner et al., 1997).

The development of AMS technology has provided the tech-

nical means by which very low organic carbon content sample

types, such as organic extracts from bone and ceramics, along

with microsample materials such as single seeds and small

amounts of hair, can now be routinely dated by

C. It should

be emphasized that AMS-based

C age determinations are

not necessarily more or less accurate or precise than decay

counting. Currently, the principal advantage of AMS technol-

ogy for

C applications is the ability to obtain meaningful

dates with samples containing microgram amounts of carbon

(e.g., Kirner et al., 1995).

Applications

Over the last decade, the expanding use of AMS-based

analysis has continued to create new and expanded areas of

research where

C data have yielded important new under-

standings that would not have been possible or practical with

decay counting. There have been several issues and topics that

have been significantly impacted by the new capability to

obtain

C measurements on milligram amounts of sample.

Figure R3 Holocene de Vries effects: Ranges (“time warps”) in

calibrated years (cal yr) obtained from the calibration of conventional

C ages. Figure taken from Taylor et al. (1996) based on data

from Stuiver and Reiner (1993: Figs. 3A–3D). Upper panel represents

0–5,000

BP and the lower panel represents 5,000–10,000 BP. Ranges

produced for ideal hypothetical samples with zero

C sample

standard deviation that had been formed during a 20-year or

shorter interval and with the youngest calibrated age obtained for

each

C age set at zero.

RADIOCARBON DATING 867

Hedges (1995) has noted that the advantages of microsam-

ple

C analysis include the ability to increase the reliability

C-inferred age estimates as well as the generation of new

chronological information. With regard to the former, the cap-

ability of remeasuring problematical dating results, the feasibil-

ity of undertaking highly selective chemical pretreatment

strategies, and the ability to compare different chemical frac-

tions of the same sample and select the most relevant sample

material have been made routinely feasible as the result of the

development of AMS-based

C analyses.

Some AMS-based

C values have been obtained in situa-

tions where larger amounts of sample were available. However,

those having responsibility for the unique historical object

would consent to the removal of only a small portion of the lar-

ger sample. Such was the case with the AMS

C dating of the

Shroud of Turin, now perhaps the most widely known sample

dated by the

C dating method. This 4.3 by 1 m rectangular-

shaped linen cloth housed in the Cathedral of St. John the

Baptist in Turin, Italy has been alleged to have been the burial

sheet of Jesus of Nazarath since 1353 when its existence was

first documented. The calibrated

C age of this sample mea-

sured by three AMS laboratories indicates that the flax from

which the linen was fabricated was most probably growing

sometime during the later part of the thirteenth or the four-

teenth century

AD, exactly the period during which the shroud

was first historically documented (Damon et al., 1989).

Arguments that

C activity in portions of the shroud could

have been altered by exchange of atmospheric CO

with the

linen as the result of high temperature effects (“scorching”)or

other purported chemical reactions have been refuted by

experiments specifically designed to test the validity of the

hypothetical mechanisms (Jull et al., 1996; Long, 1998).

A good illustration of the effect of being able to target the

most relevant sample is illustrated by a study of maize speci-

mens excavated from two rock shelters in the Tehuacan Valley,

Mexico. Samples of Zea mays from these sites had been

regarded as the earliest example of domesticated maize – the

most important New World domesticated plant. In the early

1970s, their age had been estimated on the basis of conven-

tional (decay counting)

C determinations obtained on char-

coal assumed to be stratigraphically associated with the maize

samples in the Tehuacan Valley sites. In contrast to the

5,350–7,000

BP values on the associated charcoal, the range

C values directly obtained on milligram amounts of maize

samples using AMS-based

C analysis was 1,560–4,700 BP for

the samples from San Marcos Cave and 450–4,090

BP for the

specimens from Coxcatlan Cave (Long et al., 1989). The sig-

nificantly later occurrence of maize at Tehuacan raised ques-

tions concerning assumptions about where the center(s) of

maize domestication in the New World may have been and

thus the timing of the spread of domestication of this plant in

the Western Hemisphere.

Conclusion

The impact of

C dating on the conduct of research in a

number of disciplines has been, in some aspects, clear and

explicit and, in others, subtle. In addition to providing a com-

mon chronometric time scale for the entire late Quaternary, an

important contribution of the

C method is that the technique

provides a means of providing a chronometric time scale using

means independent of any assumptions about environmental

physical parameters other than radioactive decay, carbon reser-

voir source, and geochemical conditions.

Radiocarbon data continues to provide the foundation

on which most of the most secure chronometric time scales in

most areas of the world for the last 40,000–50,000 years are

constructed, directly or indirectly, particularly in contexts

and areas lacking historical or textual documentation. While

currently not often stressed, the influence of

C data on the

pursuit of paleoenvionmental and archaeological studies

continues to be profound and pervasive.

R. Ervin Taylor

Bibliography

Bard, E., 1998. Geochemical and geophysical implications of the radiocar-

bon calibration. Geochimica et Cosmochimica Acta, 62, 2025–2038.

Bard, E., Arnold, M., Hamelin, B., Tisnerat-Laborde, N., and Cabioch, G.,

1998. Radiocarbon calibration by means of mass spectrometric

230

Th-

234

U and

C ages of on corals. Radiocarbon, 40, 1085–1092.

Bell, W.T., 1991. Thermoluminescence dates for the Lake Mungo aborigi-

nal fireplaces and the implication for the radiocarbon time scale.

Archaeometry, 33,43–50.

Bronk, C.R., and Hedges, R.E.M., 1987. A gas ion source for radiocarbon

dating. Nucl. Instrum. Methods, B29,45–49.

Buck, C.E., and Christen, J.A., 1998. A novel approach to selecting sam-

ples for radiocarbon dating. J. Archaeol. Sci., 24, 303–310.

Damon, P.E., Donahue, D.J., Gord, B.H., Hatheway, A.L., Jull, A.J.T.,

Linick, T.W., Sercelo, P.J., Toolin, L.J., Bronk, C.R., Hall, E.T.,

Hedges, R.E.M., Housley, R., Law, I.A., Perry, C., Bonani, G.,

Trumbore, S., Wolfli, W., Ambers, J.C., Bowman, S.G.E., Leese, M.N.,

and Tite, M.S., 1989. Radiocarbon dating the shroud of Turin. Nature,

337,611–615.

Edwards, R.L., Beck, J.W., Burr, G.S., Donahue, D.J., Chappell, J.M.A.,

Bloom, A.L., Druffel, E.R.M., and Taylor, F.W., 1993. A large drop

in atmospheric

C and reduced melting in the Younger Dryas, docu-

mented with

230

Th ages of corals. Science, 260,962–968.

Erlenkeuser, H., 1979. A thermal diffusion plan for radiocarbon isotope

enrichment from natural samples. In Berger, R., and Suess, H.E.

(eds.), Radiocarbon Dating. Berkeley, CA: University of California

Press, pp. 216–237.

Geyh, M.A., and Schlüchter, C., 1998. Calibration of the

C time scale

beyond 22,000

BP. Radiocarbon, 40, 475–482.

Gove, H.E., 1992. The history of AMS, its advantages over decay

counting: Applications and Prospects. In Taylor, R.E., Long, A., and

Kra, R.S. (eds.), Radiocarbon after Four Decades an Interdisciplinary

Perspective. New York: Springer, pp. 214–229.

Gove, H., 1996. Relic, Icon, or Hoax? Carbon Dating the Turin Shroud.

Bristol: Institute of Physics.

Gove, H., 1999. From Hiroshima to the Iceman: The Development and Appli-

cations of Accelerator Mass Spectrometry. Bristol: Institute of Physics.

Grootes, P.M., Mook, W.G., Vogel, J.C., de Vries, A.E., Haring, A., and

Kismaker, J., 1975. Enrichment of radiocarbon for dating samples up

to 75,000 years. Zeitschrift füer Naturforschung, 30A,1–14.

Hedges, R.E.M., 1995. Radiocarbon dating by accelerator mass spectrome-

try. Am. J. Archaeol., 99 , 105–108.

Hogg, A.G., McCornac, F.G., Higham, T.F.G., Reimer, P.J., Baillie, M.G.I.,

and Palmer, J.G., 2002. High-precision radiocarbon measurements

of contemporaneous tree-ring dated wood from the British Isles and

New Zealand:

AD 1850–950. Radiocarbon, 44, 633–640.

Jull, A.J.T., Danahue, D.J., and Damon, P.E., 1996. Factors affecting

the apparent radiocarbon age of textiles: A comment on “Effects of

fires and biogractionation of carbon isotopes on results of radiocarbon dat-

ing of old textiles: The Shroud of Turin.” J. Archaeol. Sci., 23,109–121.

Kirner, D., Taylor, R.E., and Southon, J.R., 1995. Reduction in backgrounds

of microsamples for AMS

C dating. Radiocarbon, 37, 697–704.

Kirner, D., Burky, R., Taylor, R.E., and Southon, J.R., 1997. Radiocarbon

dating organic residues at the microgram level. Nucl. Instrum. Methods

Phys. Res., B123, 214–217.

Kojo, Y., Kalin, R.M., and Long, A., 1994. High-precision ‘wiggle-

matching’ in radiocarbon dating. J. Archaeol. Sci., 21, 475–479.

Linick, T.W., Damon, P.E., Donahue, D.J., and Jull, A.J.T., 1989. Accelera-

tor mass spectrometry: The new revolution in radiocarbon dating.

Quaternary Int., 1,1–6.

868 RADIOCARBON DATING

Long, A., 1998. Attempt to affect the apprent

C age of cotton by

scorching in a CO

environment. Radiocarbon, 40,57–58.

Long, A., Benz, B.F., Donahue, D.J., Jull, A.J.T., and Toolin, L.J., 1989.

First direct AMS dates on early maize from Tehuacan, Mexico.

Radiocarbon, 31, 1035–1040.

Mazaud, A., Laj, C., Bard, E., Arnold, M., and Tric, E., 1992. A geomag-

netic calibration of the radiocarbon time-scale. In Bard, E., and

Broecker, W.S. (eds.), The Last Deglaciation: Absolute and Radio-

carbon Chronologies. Berlin: Springer, pp. 163–169.

Muller, R.A., 1977. Radioisotope dating with a cyclotron. Science, 196,

489–494.

Reimer, P.J., Hughen, K.A., Guilderson, T.P., McCormac, G., Baille, M.G.L.,

Bard, Ed. Barratt, P., Beck, J.W., Buck, C.E., Damon, P.E., Friedrich, M.,

Kromer, B., Ramsey, C.B., Reimer, R.W., Remmele, S., Southon, J.R.,

Stuiver, M., van der Plicht, J., 2002. Preliminary report of the first work-

shop of the INTCAL04 radiocarcbon calibration/ comparison working

group. Radiocarbon, 44, 653–661.

Scott, E.M., Long, A., and Kra, R., (eds.) 1990. Proceedings of the Inter-

national Workshop on Intercomparison of Radiocarbon Laboratories.

Radiocarbon, 32(3), 253–397.

Southon, J.R., Deino, A.L., Orsi, G., Terrasi, R., and Campajola, L., 1995.

Calibration of the radiocarbon time scale at 37KA

BP. Abstract of

Papers, 209th American Chemical Society National Meeting,

Part 2, p. 10.

Stuiver, M., and Polach, H.A., 1977. Discussion: Reporting of

C data.

Radiocarbon, 19, 355–363.

Stuiver, M., and Reimer, P.J., 1993. Extended 14C data base and revised

CALIB 3.0 14C age calibration program. Radiocarbon, 35, 215–230.

Stuiver, M., Long, A., and Kra, R.S., (eds.) 1998a. INTCAL 98:

Calibration Issue. Radiocarbon, 40, 1041–1159.

Stuiver, M., Reimer, P.J., Bard, E., Beck, J.W., Burr, G.S., Hughen, K.A.,

Kromer, B., McCormac, G., van der Plicht, J., and Spurk, M., 1998b.

INTCAL98 Radiocarbon age calibration 24,000–0 cal

BP. Radiocarbon,

40, 1041–1083.

Suess, H.E., 1970. Bristlecone-pine calibration of radiocarbon time 5200

to present. In Olsson, I.U. (ed.), Radiocarbon Variations and Absolute

Chronology. Stockholm: Almqvist & Wiksell, pp. 303–312.

Taylor, R.E., 1987. Radiocarbon Dating An Archaeological Perspective.

San Diego, CA: Academic Press.

Taylor, R.E., 1991. Radioisotope dating by accelerator mass spectrometry:

Archaeological and paleoanthropological perspectives. In Göksu, H.Y.,

Oberhofer, M., and Regulloi, D. (eds.), Scientific Dating Methods.

Dordrecht (Netherlands): Kluwer, pp. 37–54.

Taylor, R.E., 1996. Radiocarbon dating: The continuing revolution.

Evol. Anthropol., 4, 169–181.

Taylor, R.E., 1997a. Radiocarbon dating. In Taylor , R.E., and Aitken, M.J.

(eds.), Chronometric Dating in Archaeology. New York: Plenum,

pp. 65–96.

Taylor, R.E., 1997b. Review of H.E. Gove, Relic, Icon or Hoax?

Carbon Dating the Turin Shroud. Radiocarbon 39,115–117.

Taylor, R.E., 2001. Radiocarbon dating. In Brothwell, D.R., and

Pollard, A.M. (eds.), Handbook of Archaeological Sciences. New York:

Wiley, pp. 23–34.

Taylor, R.E., Donahue, D.J., Zabel, T.H., Damon, P.E., and Jull, A.T.J.,

1984. Radiocarbon dating by particle accelerators: An archaeological

Perspective. In Lambert, J.B. (ed.), Archaeological Chemistry III.

Washington, DC: American Chemical Society, pp. 333–356.

Taylor, R.E., Stuiver, M., and Reimer, P.J., 1996. Development and exten-

sion of the calibration of the radiocarbon time scale: Archaeological

applications. Quaternary Sci. Rev., 15, 655–668.

Voelker, A.H.L., Sarnthein, M., Grootes, P.M., Erlenkeuser, H., Laj, C.,

Mazaud, A., Nadeau, M.-J., and Schleicher, M., 1998. Correlation of

marine

C ages from the Nordic seas with the GISP2 isotope record:

implications for

C calibration beyond 25 ka BP. Radiocarbon, 40,

517–534.

Cross-references

Carbon Isotopes, Stable

Dating, Dendrochronology

Dating, Radiometric Methods

Isotope Fractionation

Uranium-Series Dating

RADIOLARIA

Radiolarians are marine protozoa with a geological record

ranging from Precambrian to Present. Radiolarians are exclu-

sively marine and are found in all the oceans, in all temperature

zones and at all depths. They are also neritic in their distribu-

tion, being found in Norwegian fjords, in the plankton (9,000

specimens m

–3

), more abundantly in the tropics (18,000 speci-

mens m

–3

), and rarely in the polar oceans. The highest numbers

of Radiolaria are found in the upper 50–200 m of the water

column. Those radiolarians possessing a skeleton may contri-

bute significantly to the deep sea eupelagic sediments, the

so-called radiolarian ooze, which are found deposited in the

central parts of the Pacific and Indian Oceans.

Evolution of this group has resulted in a high diversity of

both extinct and extant species. Their opal skeleton is charac-

teristic and in many cases is radial, from which the name

Radiolaria is derived. Some, however, possess a bilateral skele-

ton, whilst others have no skeleton at all. The inner part of the

protoplasm (endoplasm) houses the nucleus (reproduction),

organelles such as mitochondria (respiration) and Golgi bodies

(secretion), as well as lipid droplets and vacuoles, all enveloped

by an organic central capsule. The outer part of the protoplasm

(ectoplasm), the calymma, communicates through openings,

or pores, in the central capsular wall (Figure R4). The calymma

is a frothy or bubble-like body, which is interpreted as a

floating device, besides housing algal symbionts (zooxanthel-

lae), vacuoles, and mitochondria. The outer margin of the

protoplasm is characterized by a network of anastomosing

Figure R4 Principal structure of a polycystine radiolarian, showing soft

and hard body features (Modified after Kling, 1978).

RADIOLARIA 869

pseudopodia as well as different irregular rhizopodia, long and

stiff radial axopodia (Figure R4), which are clearly seen as a

bundle of long, straight spines in the living specimen of Spiro-

cyrtis scalaris (Figure R5), and delicate filpodia (Figure R4).

The skeletal spines are enveloped by the protoplasm, and there-

fore protected from being dissolved by seawater. Prey is cap-

tured by the pseudopodia and digested in the ectoplasm.

Radiolarian taxonomy was essentially established by Ernst

Haeckel when he published an extensive monograph in 1887,

based on material collected on the Challenger Expedition

(1873–1876) that circumnavigated the world’s oceans. This is

not a natural taxonomic system, but a useful system to group

similar morphotypes. Radiolarian taxonomic studies were

initiated in Germany, but subsequent schools developed in

France, Russia, and the U.S.A., where taxonomy was based

predominantly on intricate skeletal morphologies, as shown in

Callimitra carolotae (Figure R6), with the exception of the

French group who in addition based their taxonomy on micro-

scopic cellular structures. There is no consensus on the higher

rank taxonomy within the radiolarians, but the following is

often used:

Kingdom PROTISTA Haeckel, 1866

Phylum SARCODINA Hertwig, 1876

Class ACTINOPODA Calkins, 1910

Subclass RADIOLAIRA Müller, 1858

Superorder PHAEODARIA Haeckel, 1879

Superorder POLYCYSTINA Ehrenberg, 1838, emend.

Riedel, 1967

Order SPUMELLARIA Ehrenberg, 1875

Order NASSELLARIA Ehreneberg, 1875

Recent work on radiolarian DNA suggests another relation-

ship of the main Radiolaria groups, in addition to that proposed

by Haeckel (1887), above. Yuasa et al. (2003) showed that

the Phaeodaria made a separate clade, while the Acantharia,

Spumellaria and Nassellaria made another clade. The latter

made two subclades where solitary Spumellaria grouped with

the Acantharia and the colonial Spumellaria grouped with the

Nassellaria. This is supported by other DNA studies (Polet

et al., 2004).

The diversity and relatively rapid rate of evolution of Radi-

olaria through time makes this group very important in Tertiary

and Quaternary biostratigraphy, especially in deep-sea sedi-

ments where carbonate fossils are dissolved. Levine (1963)

calculated the total number of living and fossil radiolarians to

be 4,791 and 2,389, respectively. Goll and Merinfeld (1979)

concluded that these numbers are inflated by unwarranted taxo-

nomic splitting and suggested numbers that are more conser-

vative: about 200 Acantharia, more than 600 Phaeodarians,

and about 1,500 extant and extinct polycystine radiolarians,

the latter number is very uncertain.

Polycystine radiolarians are important in deep sea stratigra-

phy. The Cenozoic is divided into 32 biozones (Recent through

Eocene, Paleocene is still unzoned). Radiolarians are also

important for paleotemperature proxy data, and estimates for

the last ca 13,400

C years BP have been successfully applied

in the Norwegian Sea (Dolven et al., 2002). Here, the calcula-

tions show an average temperature of 12  1



C during the

Holocene (the last 10,000

C years). During the Younger

Dryas (10,000–11,000

C years BP), the temperature decreased

to about 7



C in only a few centuries. In the Holocene, the

“8,200-year event,” as observed in the GISP2 ice core, is also

Figure R6 Callimitra carolotae Haeckel. Recent, from surface sediments

in the Orca Basin, Gulf of Mexico. Photograph by K. R. Bjørklund and

R. M. Goll.

Figure R5 Spirocyrtis scalaris Haeckel, a living nassellarian radiolaria

taken from the surface water near Okinawa Island, Japan. Photograph

provided by Dr. Atsushi Matsuoka.

870 RADIOLARIA

observed in the radiolarian paleotemperature estimates, with a

temperature drop of ca. 3



C. The associated radiolarian assem-

blages changed during this time interval with a relative

abundance minimum of the warm water species Lithomelissa

setosa and Pseudodictyophimus gracilipes, while Stylodictya

validispina displayed an absolute maximum (26%). This evi-

dence proves that radiolarians are a sensitive and powerful

paleoclimatic tool.

Adolfo Molina-Cruz and Kjell R. Bjørklund

Bibliography

Calkins, G.N., 1910. Protozoölogy. London: Bailliere, Tidall and Cox,

349pp.

Dolven, J.K., Cortese, G., and Bjørklund, K.R., 2002. A high-resolution

radiolarian-derived paleotemperature record for the Late Pleistocene-

Holocene in the Norwegian Sea. Paleoceanography, 17(4), 1072,

doi:10.1029/ 2002PA000780,2002. (Electronic journal).

Ehrenberg, C.G., 1838. Uber die Bildung der Kreidefelsen und des Kreide-

mergels durch unsichtbare Organismen. Konigliche Akademie der

Wissenschaften zu Berlin, Abhandlungen, Jahre 1838,59–147.

Ehrenberg, C.G., 1875. Fortsetzung der mikrogeologischen Studien

als Gesammt-Uebersicht der mikroskopischen Palaontologie gleichartig

analysirter Gebirgsarten der Erde, mit specieller Rucksicht auf den

Polycystinen-Mergel von Barbados. Konigliche Akademie der Wis-

senschaften zu Berlin, Abhandlungen, Jahre 1875,1–225.

Goll, R.M., and Merinfeld, G., 1979. Radiolaria. In Fairbridge, R.W., and

Jablonski, D. (eds.), The Encyclopedia of Paleontology. Encyclopedia

of Earth Sciences, vol. VII. Stroudsburg, PA: Dowden, Hutchinson &

Ross, pp. 673–684.

Haeckel, E., 1866. Generelle Morphologie der Organismen. Teil II. Allge-

meine Entwicklungsgeschichte der Organismen. Berlin: Reimer,

pp. 1–462.

Haeckel, E., 1879. Ueber die Phaeodarien, eine neue Gruppe kieselschali-

ger mariner Rhizopoden. Jenaische Zeitschrift für Naturwissenschaft,

14, 151–157.

Haeckel, E., 1887. Report on the Radiolaria collected by the H.M.S.

Challenger during the Years 1873–1876. Report on the Scientific Results

of the Voyage of the H.M.S. Challenger, Zoology, vol. 18, 1803pp.

Hertwig, R., 1876. Zur Histologie der Radiolarien. Untersuchungen uber

den Bau und die Entwicklung der Sphaerozoiden und Thalassicolliden.

W. Engelmann, Leipzig, Germany, 91+ errata.

Kling, S.A., 1978. Radiolaria. In Haq, B.U., and Boersma, A. (eds.),

Introduction to Marine Micropaleontology. New York: Elsevier,

pp. 203–244.

Levine, N.D., 1963. Protozoology today. In Progress in Protozoology. New

York: Academic Press, pp. 39–40.

Müller, J., 1858. Über die Thalassicollen, Polycystinen und Acanthometren

des Mittelmeeres. Konigliche Preussische Akademie der Wissenschaften

zu Berlin, Abhandlungen, Jahre 1858,1–62.

Polet, S., Berney, C., Fahrni, J., and Pawlovski, J., 2004. Small-subunit

ribosomal RNA gene sequences of Phaeodarea challenge the mono-

phyly of Haeckel’s Radiolaria. Protist, 155,53–63 (http: / / www.

elsevier-deutschland.de / protist).

Riedel, W.R., 1967. Subclass Radiolaria. In Harland, W.B., et al., (eds),

(eds.), The Fossil Record. London, UK: Geological Society of London,

pp. 291–

298.

Yuasa, T., Takahashi, O., and Mayama, S., 2003. Molecular Phylogeny of

the solitary shell-bearing Polycystinea. In Diserens, M.-O., and Jackett,

S.-J. (eds.), Tenth Meeting of the International Association of Radiolar-

ian Palaeontologists, University of Lausanne, September 8–12, 2003,

Abstract and Program: p. 118.

Cross-references

Marine Biogenic Sediments

Radiocarbon Dating

The 8,200-Year bp Event

Younger Dryas

RED BEDS

Red beds are red-colored clastic sedimentary deposits

(Figure R7). The term is widespread in the literature and has

been used to describe both marine and non-marine rocks, so

it is non-genetic. The presence of oxidized (ferric) iron (usually

as fine-grained hematite) disseminated throughout the matrix of

the deposits is the unifying feature of red beds. Reddening typi-

cally forms as a result of the post-depositional dehydration of

oxyhydroxides such as goethite to hematite and indicates +u

(positive) Eh conditions. Rocks previously called red beds

include paleosols (including laterites), banded iron formations,

sandstones, siltstones, conglomerates, and deep sea shales.

Marine red beds are rare and form only in areas with little

organic matter and slow sedimentation rates that allow all of

the organic matter to be broken down prior to burial. Only

red bed sequences containing or composed of paleosols or eva-

porites are useful for paleoclimatic reconstructions.

Paleosols in red bed sequences

Red beds have been interpreted as desert deposits by analogy

with modern deserts, which are often red (Walker, 1967), but

subsequently it was recognized that the modern red deserts

are associated with ancient red beds (e.g., the Simpson Desert

of Australia where Folk (1976) showed that the red color of

the modern dunes was due to prior laterite formation). More

recent work has emphasized a role for strongly seasonal cli-

mates (Dubiel and Smoot, 1994), such as monsoons with short

wet seasons between long dry seasons during which the soils

could dry out substantially, promoting formation of iron oxy-

hydroxides and a high degree of aeration. However, it should

be noted that any well-drained soil has the same aerated condi-

tions, and well-drained soils can form in a variety of climatic

settings other than in monsoonal belts (Sheldon, 2005). Diag-

nostic features of monsoonal settings include concentrically

layered carbonate or iron oxide concretions (Retallack, 1991;

Table R1), dispersed calcium carbonate throughout the deposits

(Retallack, 1991; Table R1), and soil types including Vertisols

Figure R7 Eocene red beds from the Wasatch Formation of

Wyoming, US.

RED BEDS 871

with their characteristic deep cracking due to seasonal shrink-

ing and swelling of clays, or Mollisols with crumb peds. All

of these features are also preserved in paleosols. Model results

indicate monsoon-like climatic circulation patterns for the Pan-

gean supercontinent (Kutzbach and Gallimore, 1989), but so

far insufficient field work has been performed to confirm mon-

soonal features in either the sediments or paleosols preserved in

the Pangean red bed sequences.

Paleosols are fossil soils, typically buried by subsequent

deposition. Many sequences of terrestrial fluvial red beds pre-

serve paleosols that developed on flood deposits, but other

preservational settings include sequences of lava flows, and

lagoonal, lake, and ocean margins. Each paleosol represents a

significant hiatus in deposition ranging in duration from per-

haps one hundred years up to millions of years in the case of

ultisols, oxisols, and laterites. Because paleosols form in direct

contact with the atmosphere, they record both paleoenviron-

mental and paleoclimatic conditions. Identification of paleosols

within a sequence of red beds is made on the basis of a number

of morphological features (Table R1). Some of the features

of individual paleosols also give information about the depth

of the paleo-water table beneath the surface of the paleosol,

and by extension, information about how well-drained the

paleosol was. In addition to gaining this paleohydrological

information about the depositional setting of the red bed

sequence, a number of other features of paleosols are quantita-

tively useful as paleoclimatic indicators. Mean annual precipi-

tation may be estimated from the degree of chemical

weathering (Sheldon et al., 2002) or from the depth to a Bk

(carbonate-bearing) horizon in a paleosol (Retallack, 1994),

and atmospheric pCO

can be estimated by measuring the iso-

topic composition of carbonate nodules (Cerling et al., 1989)or

pedogenic goethites (Yapp and Poths, 1992).

Geologic record of red beds

Because the red color in red bed forms is due to oxidized iron, the

geologic record of red beds reflects the Precambrian rise of atmo-

spheric oxygen levels. The oldest universally agreed upon red

bedsare red sandstones, siltstones, and mudstones of South Africa,

immediately overlying the so-called Hekpoort Paleosol, which has

been dated to about 2.2 Ga before present. Red beds have been

common in the geologic record ever since, with their greatest

extent in the Phanerozoic, especially during the Paleozoic-

Mesozoic transition during which red beds are one of the charac-

teristic depositional facies of the Pangean supercontinent.

Nathan D. Sheldon

Bibliography

Birkeland, P.W., 1999. Soils and Geomorphology, 3rd edn. New York:

Oxford University Press, 430pp.

Cerling, T.E., Quade, J., Wang, W., and Bowman, J.R., 1989. Carbon iso-

topes in soils and palaeosols as ecology and palaeoecology indicators.

Nature, 341, 138–139.

Dubiel, R.F., and Smoot, J.P., 1994. Criteria for interpreting paleoclimate

from red beds – a tool for Pangean reconstructions. In Embry, A.F.,

Beauchamp, B., and Glass, D.J. (eds.), Pangea: Global Environments

and Resources. Calgary, AB: Canadian Society of Petroleum Geolo-

gists, Memoir 17, pp. 295–311.

Folk, R.L., 1976. Reddening of desert sands: Simpson Desert, N.T., Australia.

J. Sediment. Petrol., 46,604–615.

Kutzbach, J.E., and Gallimore, R.G., 1989. Pangean climates: Mega-

monsoons of the megacontinent. J. Geophys. Res., 94, 3341–3357.

Retallack, G.J., 1991. Miocene Paleosols and Ape Habitats of Pakistan and

Kenya. Oxford, UK: Oxford University Press. Oxford Monographs on

Geology and Geophysics, no. 19, 346pp.

Retallack, G.J., 1994. The environmental factor approach to the interpreta-

tion of paleosols. Soil Sci. Soc. Am. Special Publication Number, 33,

31–64.

Table R1 Some features of red bed sequences useful for paleoenvironmental reconstructions

Feature type Diagnosis

Soil characteristics:

1. Gilgai (small-scale anticlinal relief) 1. Pronounced wet-dry seasonality

2. Slickensides, ped morphology, relict bedding 2. Relative weathering intensity

3. Horizon thickness and color 3. Duration of weathering

4. Mottling associated with organic matter (roots) 4. Micro-reducing conditions within the soil

Diagnostic soil types:

5. Caliche /calcrete soils 5. Sub-humid to arid climatic conditions

6. Duripan /silcrete soils 6. Semi-arid to arid climatic conditions

7. Vertisols 7. Pronounced wet-dry seasonality

8. Laterites 8. Tropical conditions

Soil nodules:

9. Size 9. Large imply longer formation times

10. Distribution 10. Disbursed implies strong seasonality

11. Concentric growth and alternating carbonate and oxide layers indicate 11. Strong seasonality and/ or monsoonal climatic conditions

Evaporites

12. Gypsum and halite 12. Arid conditions

Eolian deposits

13. Proportion of silt in thin section point counts conditions 13. Semi-arid to arid climatic

Frost wedges

14. Clastic dikes 14. Cold (taiga or tundra) climatic conditions

Water table characteristics:

15. Rooting depth 15. Deeply penetrating implies well-drained

16. Burrowing depth 16. Deeply penetrating implies well-drained

17. Density and tiering of ichnofabrics 17. High density implies low flood frequency

Invertebrate fossils with modern analogs Depends on the modern analog

Vertebrate fossils with modern analogs Depends on the modern analog

Information after Birkeland (1999), Dubiel and Smoot (1994), and Retallack (2001).

872 RED BEDS