Odekon M. Encyclopedia of paleoclimatology and ancient environments

Подождите немного. Документ загружается.

Late Devonian glaciation (Glacial I) and paleoclimate

The onset of Southern Hemisphere glaciation in the Late Devo-

nian is a logical place to begin our consideration of late

Paleozoic paleoclimates. Although this glaciation was rela-

tively short, perhaps less than 0.5 million years (McGhee,

1996; Streel et al., 2000), it signals the beginning of the transition

between the early Paleozoic “hot-house” and the late Paleozoic

“ice-house” or glacial climates. Isbell et al. (2003)referredtothis

glaciation as Glacial I (Figure L6). Late Famennian (Glacial I)

sediments appear in a number of South American basins (Bolivia,

Madre de Dios, Solimoes, Amazonas, Paranaiba) and possibly in

Libya (Isaacson et al., 1999;Streeletal.,2000). This glaciation

coincided with the deposition of the Hangenberg Shale, a black,

organic-rich shale deposited on top of ramp and platform carbo-

nates in low and mid-latitudes (Caplan and Bustin, 1999). The

Hangenberg is one of a series of organic-rich shales and limestones

deposited on Devonian and early Mississippian continental

shelves, termed “Kellwasser event beds” after the extensive

Kellwasser Horizon that marks the Frasnian-Famennian boundary

(House, 1985). During Kellwasser events, anoxic or dysoxic water

flooded continental shelves and the burial rate of organic carbon

increased (Buggisch, 1991).

The terrestrial floras and marine faunas of the Late Devo-

nian and Mississippian were extremely cosmopolitan (Ziegler

et al., 1981; McGhee, 1996; Streel et al., 2000; Raymond and

Metz, 2004), suggesting a shallow pole-to-equator temperature

gradient. In the Mesozoic and early Cenozoic, shallow latitudi-

nal temperature gradients reflect extremely warm poles (Valdes,

2000). In the Late Devonian, however, cosmopolitan floral and

faunal assemblages may reflect equatorial cooling. Evidence

for cool equatorial temperatures associated with Late Devonian

glaciation includes contraction of the ranges of calcareous

foraminifera toward the equator (Kalvoda, 1986) and the

increasing diversity of deep-water siliceous sponges in the lat-

est Devonian (McGhee, 1996).

Late Devonian glaciation was preceded by mass extinction

at the Frasnian/Famennian boundary: 13–38% of marine

families, 29–60% of marine genera and 82% of marine species

became extinct (Sepkoski, 1986; McGhee, 1996). This mass

extinction appears to have been related to equatorial cooling

and flooding of continental shelves with anoxic or dysoxic

water, although other mechanisms have been proposed (see

Raymond and Metz, 2004 and references therein). In the Late

Devonian mass extinction, equatorial taxa suffered higher per-

centages of extinction than southern cold-temperate and polar

taxa; and, shallow water, presumably warm-adapted corals suf-

fered greater percentages of extinction than deep water, pre-

sumably cold-adapted, corals (Scrutton, 1988; McGhee, 1996).

Late Devonian glaciation coincided with a second interval

of increased extinction at the Devonian-Carboniferous bound-

ary: approximately 16% of marine families and 21% of marine

genera became extinct (Sepkoski, 1986). However, these

extinctions show no clear latitudinal pattern.

Witzke (1990) reviewed the use of sedimentological indica-

tors (coals, bauxite, kaolinite, evaporites and calcretes) to

define climate zones. Based on sedimentological indicators,

we can identify the following Late Devonian climate zones:

(a) a wet, equatorial zone, coincident or slightly to the north

of the paleoequator; (b) two arid, subtropical zones to the north

and south of the equatorial zone and a southern wet, warm tem-

perate zone; and (c) a southern cool temperate zone associated

with South Polar glaciation (Figure L7).

Mississippian paleoclimate

Late Devonian glaciation may have persisted into the earliest

Mississippian (Isaacson et al., 1999; Isbell et al., 2003),

although earliest Mississippian glacial sediments are poorly

dated (Streel et al., 2000). Aside from these putative earliest

Mississippian glacial deposits, the first two stages of the Mis-

sissippian, the Tournaisian and Visean, were apparently free

of significant glacial ice.

Based on sedimentary indicators, the following climate

zones can be identified: (a) a zone of wet “equatorial” climate,

which lay to the north of the paleoequator on the northwest

side of Pangaea and coincided with the paleoequator on the

east side of the continent; (b) a zone of arid climate to the

south; (c) a southern warm temperate zone to the south

of the arid zone; (d) a southern cool temperate zone to the

Figure L6 The timing and duration of late Paleozoic glaciation.

d: Glacial I, II, and III of Isbell et al. (2003); i: mid-to-late Westphalian

glaciation of Veevers and Powell (1987), Frakes et al. (1992), and Crowell

(1999). While glaciation may have persisted until the end of the

Permian (Frakes et al., 1992; Crowell, 1999; Isbell et al., 2003); others

have suggested that glaciation ended in the mid-Permian (Veevers and

Powell, 1987; Gibbs et al., 2002; Rees et al., 2002).

LATE PALEOZOIC PALEOCLIMATES 499

south of the arid zone; and (e) a northern warm temperate zone

to the north of the equatorial wet zone (Figure L8).

Rowley et al. (1985) reviewed the phytogeography of Tour-

naisian and Visean floras. In both intervals, northern warm

temperate floras from Siberia are distinct. In both, wet tropical

assemblages from Europe and South China appear distinct from

arid tropical assemblages found in the Canadian Maritimes. Few

Tournaisian Gondwanan assemblages were included in their

study; however, Visean Gondwanan assemblages from the warm

temperate zone have a distinct floral assemblage (Iannuzzi and

Pfefferkorn, 2002).

The migration patterns of marine invertebrates allow us to

trace the relative climate of the Mississippian. In the mid-to-

late Visean, a statistically significant percentage of articulate

brachiopod genera expanded their range toward the poles,

suggesting that both poles became warmer in the late Visean.

Although the biostratigraphic resolution of land plant migra-

tions is not as precise as that of brachiopods, Northern Hemi-

sphere land plants show a similar range expansion toward the

pole in the Visean (Raymond, 1985; Raymond et al., 1990;

Kelley and Raymond, 1991).

Late Mississippian-Permian glaciation (Glacial II and III)

and paleoclimate

Marine invertebrate migration patterns suggest that the North

Pole remained warm into the Sepukhovian (Raymond et al.,

1990). In the Southern Hemisphere, however, a significant

percentage of genera migrated away from the South Pole in the

late Visean to Sepukohovian, suggesting South Polar cooling

during the Serpukovian (Kelley and Raymond, 1991). A growing

Figure L8 Lower Carboniferous (Mississippian) paleogeographic reconstruction showing climate zones and the distribution of sedimentary

indicators of early–middle Mississippian (Tournaisian and Visean) climate (unpublished data, A. J. Boucot, C. Xu, and C. R. Scotese).

Figure L7 Late Devonian paleogeographic reconstruction showing climate zones and the distribution of sedimentary indicators of Late Devonian

climate (based on unpublished data of A. J. Boucot, C. Xu, and C.R. Scotese).

500 LATE PALEOZOIC PALEOCLIMATES

body of evidence suggests that this migration away from the

South Pole was related to the onset of significant glaciation in

the Southern Hemisphere (Isbell et al., 2003). Global regression

in the late Serpukhovian has been linked to the onset of Southern

Hemisphere glaciation (Bouroz et al., 1978; Ross and Ross,

1988). Glacial sedimentary deposits of Namurian through ear-

liest Westphalian age occur in South America and Australia

(Isbell et al., 2003). In addition, shifts in the isotopic composition

O and d

C) of marine carbonates from North America, Eur-

ope, and the Russian Platform in the Serpukhovian may signal

the onset of significant glaciation (Bruckschen et al., 1999; Mii

et al., 2001).

Although many lines of evidence suggest that glacia-

tion began in the Serpukhovian, the nature of this glaciation

remains uncertain. Isbell et al. (2003) argued that Glacial II

sediments resulted from alpine rather than continental glacia-

tion. Others (Veevers and Powell, 1987; Crowell, 1999) argued

for continental glaciation during this interval. Resolution of

this controversy will require additional data; nonetheless, the

biogeographic patterns discussed below indicate that the onset

of glaciation in the Serpukhovian had a profound effect on

Carboniferous biotas. The biotic response to Serpukhovian

glaciation was similar to the biotic response to continental gla-

ciation in the late Tertiary-Recent and very different from the

biotic response to Late Devonian glaciation.

The biogeographic differentiation of land plants and marine

invertebrates increased dramatically in the early Pennsylvanian

(Ziegler et al., 1981). This increase in biogeographic differen-

tiation reflected the biotic response to the establishment of a

steep pole-to-equator temperature gradient as a result of glacia-

tion. Crame and Rosen (2002) noted a similar biogeographic

response to late Cenozoic glaciation. Late Devonian glaciation

occurred during an interval of low biogeographic differentia-

tion, which persisted into the early Mississippian (Raymond

and Metz, 2004).

As the poles became frigid, the equatorial climates of the

Pennsylvanian became wetter and perhaps warmer as well. Late

Visean “equatorial” coal deposits have a broad latitudinal distribu-

tion, ranging from 20



S paleolatitude to 30



Npaleolatitudeon

a variety of paleogeographic reconstructions (Raymond, 1996).

A pronounced paleoequatorial coal belt, centered on the paleoe-

quator, became established in the Serpukhovian and continued

throughout the Pennsylvanian and into the Permian (Figures

L8–L10). As the equatorial coal belt became established, the

diversity of equatorial land plants rose relative to the diversity of

mid-latitude and high-latitude land plants (Raymond, 1996). The

presence of a pronounced paleoequatorial coal belt and the

increased diversity of paleoequatorial land plants during this inter-

val suggest that equatorial climates became wetter during glacial

onset. Bande and Prakash (1986) noted a similar increase in

the diversity of equatorial land plants between the early and late

Cenozoic, which may be tied to late Cenozoic glaciation.

Coal distribution data suggest that the paleoequator became

wetter during glacial onset. Marine invertebrate migration patte-

rns suggest that it may have become warmer as well. A small,

but statistically significant percentage of Northern Hemisphere

brachiopod genera vacated the equatorial zone in the Serpukho-

vian by moving their southern boundary away from the equator

(Raymond et al., 1989; Raymond et al., 1990). Valentine (1984)

noted similar migration patterns for marine invertebrates asso-

ciated with late Cenozoic glaciation, which he interpreted as evi-

dence for equatorial warming. During Late Devonian glaciation,

calcareous foraminifera contracted their ranges toward the pole

(Kalvoda, 1986). This migrational pattern is exactly opposite from

that of Carboniferous brachiopods and late Tertiary marine inver-

tebrates, and suggests that the low-latitudes became cooler during

Late Devonian glaciation.

Ziegler et al. (1987) linked the presence of equatorial coal

to cold or glaciated poles. Otto-Bliesner (1993) attributed

Pennsylvanian paleoequatorial coal to the orientation and lati-

tudinal placement of the Ouachita and Alleghany Mountains,

along an east to west trend, close to the paleoequator. Paleocli-

mate models suggested that east-west trending equatorial

mountain belts might constrain the position of the inter-tropical

convergence zone (the zone of ever-wet, ever-warm climate) to

a narrow latitudinal belt, centered on the equator. In the latest

Mississippian and early Pennsylvanian, glacial onset and conti-

nental collision were so closely spaced in time that it is impos-

sible to distinguish the contribution of mountain building and

glaciation from the formation of the paleoequatorial coal belt.

Figure L9 Upper Carboniferous (Pennsylvanian) paleogeographic reconstruction showing climate zones and the distribution of sedimentary

indicators of Pennsylvanian climate (unpublished data, A. J. Boucot, C. Xu and C. R. Scotese).

LATE PALEOZOIC PALEOCLIMATES 501

The continuity of glaciation during the rest of the Pennsylva-

nian and into the Permian remains controversial. Isbell et al.

(2003) argued for substantially ice-free poles during most of the

Westphalian, followed by renewed glaciation in the late Pennsyl-

vanian (Stephanian). Other workers suggested that Gondwana

remained glaciated from the Namurian into the Permian

and argued for peak glaciation in the Westphalian (Veevers and

Powell, 1987;Frakesetal.,1992; Crowell, 1999). Isotopic studies,

which suggest warming rather than cooling at the Westphalian/

Stephanian boundary (Bruckschen et al., 1999;Miietal.,2001),

appear to contradict Isbell et al. (2003).

The abundance of 4th and 5th order transgressive-regressive

cycles, known as cyclothems, in Carboniferous and Permian

strata from the Paleo-Northern Hemisphere has been inter-

preted as evidence for continental glaciation (Ross and Ross,

1988; Heckel, 1994). According to Heckel and others (see

Heckel, 1994 and references therein), Milankovitch orbital

cycles caused late Paleozoic continental glaciers to advance

and retreat, which resulted in the deposition of repeated

transgressive-regressive sedimentary cycles in equatorial and

low latitudes. Cyclothems are usually linked to continental gla-

ciation because only continental glaciers contain enough water

to affect global sea level (Isbell et al., 2003). If cyclothems

reflect glacial-eustasy, the presence of Westphalian cyclothems

in North America, Europe, the Russian Platform, the Moscow

Basin, and the Urals argues for significant Westphalian glacia-

tion. Conversely, Isbell et al. (2003), who felt that the middle

and late Westphalian were ice-free, suggested that Carbonifer-

ous and Permian cyclothems could have other causes aside

from glaciation.

A full understanding of the significance of 4th and 5th order

transgressive-regressive cycles known as cyclothems requires

more research. The earliest reported cyclothems occur in the

early Visean of Great Britain (W right and Vanstone, 2001),

before other signs of significant glaciation, such as increased

biogeographic differentiation, the development of the paleoequa-

torial coal belt, and isotopic shifts indicating global cooling and

the growth of glaciers (Bruckschen et al., 1999; Mii et al., 2001;

Raymond and Metz, 2004). Some Carboniferous 4th and 5th

order trangressive-regressive cycles attributed to glacial-eustasy

may have a different cause. However, glacial-eustasy appears

to be a reasonable explanation for repeated 4th and 5th order

transgressive-regressive cycles that can be correlated across

North America (Heckel, 1994).

Widespread glaciation may have ended in the Middle Permian

(Veevers and Powell, 1987; Gibbs et al., 2002; Rees et al., 2002)

(Figures L10 and L11). Others (Frakes et al., 1992; Crowell,

1999;Isbelletal.,2003) suggest that Gondwanan glaciers may

have persisted to the end of the Permian. Pennsylvanian and

Permian marine and terrestrial biotas retained a high amount of

biogeographic differentiation, suggesting a relatively steep pole

to equator temperature gradient throughout the Pennsylvanian

and Permian (Ziegler et al., 1981; Rees et al., 2002). Nonetheless,

the presence of fossils forests, reconstructed as cool-temperate

deciduous vegetation, in Late Permian sediments from Antarctica,

suggests that the Late Permian South Pole was warmer than

it was during the Pennsylvanian and Early Permian (Rees et al.,

2002). Compared to Pennsylvanian and Early Permian climate,

the climate of the Middle and Late Permian was more influenced

by monsoonal circulation patterns (Parrish, 1993). As a result,

the equatorial zone of the supercontinent Pangea was relatively

arid in the Middle and Late Permian and the paleotropical

coal belt was confined to island continents, such as South

China, which straddled the paleoequator to the east of Pangea at

that time.

Summary

In summary, the late Paleozoic was a time of dramatic climatic

shifts, punctuated by glaciation events. Clues to these cli-

mate changes can be found in sediments, the stable isotopic

composition of marine carbonates, and the biogeographic

distribution of land plants and marine animals. Late Devonian

glaciation (Glacial I of Isbell et al., 2003), which started the

interval, appears more similar to Late Ordovician “hot-house”

glaciation than subsequent ice-house glaciations in the mid-

Carboniferous-to-Permian and late Cenozoic (Raymond and

Metz, 2004). The glaciation event that began in the late Serpu-

kohovian or at the Mississippian-Pennsylvanian boundary

(Glacial II of Isbell et al., 2003) appears very similar to late

Cenozoic glaciation. During both intervals, the poles were cold,

Figure L10 Lower Permian paleogeographic reconstruction showing climate zones and the distribution of sedimentary indicators of early Permian

climate (unpublished data, A. J. Boucot, C. Xu, and C.R. Scotese).

502 LATE PALEOZOIC PALEOCLIMATES

the equator was warm, and there was abundant biogeographic

differentiation. Unlike the hot-house glaciations of the Late

Ordovician and Late Devonian, Pennsylvanian-Permian glacia-

tion was not marked by mass extinction. Sources differ on the

continuity of glaciation in the mid-to-late Pennsylvanian (mid

Westphalian through Stephanian) and after the Middle

Permian (Veevers and Powell, 1987; Frakes et al., 1992;

Bruckschen et al., 1999; Crowell, 1999; Isbell et al., 2003).

However, the persistence of biogeographic differentiation

throughout the Pennsylvanian to the end of the Permian argues

for steep latitudinal temperature gradients, cold poles, and

warm, wet equatorial climates until the end of the Paleozoic

(Ziegler et al., 1981; Rees et al., 2002).

Anne Raymond and Christopher R. Scotese

Bibliography

Bande, M.B., and Prakash, U., 1986. The Tertiary flora of Southeast Asia

with remarks on its palaeoenvironment and phytogeography of the

Indo-Malayan region. Rev. Palaeobot. Palynol., 49, 203–234.

Bouroz, A., Einor, O.L., Gordon, M., Meyen, S.V., and Wagner, R.H.,

1978. Proposals for an international chronostrtigrahic classification of

the Carboniferous. Huitieme Congres International de Stratigraphie et

de Geologie Carbonifere, Compte Rendu, 1, 36–52.

Bruckschen, P., Oesmann, S., and Veizer, J., 1999. Isotope stratigraphy of

the European Carboniferous: Proxy signals for ocean chemistry, climate

and tectonics. Chem. Geol., 161, 127–163.

Buggisch, W., 1991. The global Frasnian-Famennian ‘Kellwasser’ Event.

Geologische Rundschau, 80,49–72.

Caplan, M.L., and Bustin, R.M., 1999. Devonian-Carboniferous Hangenberg

mass extinction event, widespread organic-rich mudrock and anoxia:

Causes and consequences. Palaeogeogr. Palaeoclimatol. Palaeoecol.,

148,187–207.

Crame, J.A., and Rosen, B.R., 2002. Cenozoic palaeogeography and the

rise of modern diversity patterns. In Crame, J.A., and Owen, A.W.

(eds.), Palaeobiogeography and Biodiversity Change: The Ordovician

and Mesozoic-Cenozoic Radiations. London: Geological Society of

London, pp. 153–168.

Crowell, J.C., 1999. Pre-Mesozoic Ice Ages: Their Bearing on Understand-

ing the Climate System. Boulder, CO: Geological Society of America,

pp. 1–106.

Frakes, L.A., Francis, J.E., and Syktus, J.I., 1992. Climate Modes of

the Phanerozoic. Cambridge, UK: Cambridge University Press, 274pp.

Gibbs, M.T., Rees, P.M., Kutzbach, J.E., Ziegler, A.M., Behling, P.J., and

Rowley, D.B., 2002. Simulations of Permian climate and comparisons

with climate-sensitive sediments. J. Geol., 110,33–55.

Hambrey, M.J., and Harland, W.W., 1981. Earth’s Pre-Pleistocene Glacial

Record. Cambridge, UK: Cambridge University Press, 1004pp.

Heckel, P.H., 1994. Evaluation of evidence for glacio-eustatic control over

marine Pennsylvanian cyclothems in North America and consideration

of possible tectonic events. In Dennison, J.M., and Ettensohn, F.R.

(eds.), Tectonic and Eustatic Contrls on Sedimentary Cycles. Tulsa,

OK: SEPM, pp. 65–88.

House, M.R., 1985. Correlation of mid-Paleozoic ammonoid evolutionary

events with global sedimentary perturbations. Nature, 313,17–22.

Iannuzzi, R., and Pfefferkorn, H.W., 2002. A pre-glacial, warm-temperate

floral belt in Gondwana (Late Visean, Early Carboniferous). Palaios,

17, 571–590.

Isaacson, P.E., Hladil, J., Shen, J.-W., Kalvoda, J., and Grader, G., 1999.

Glaciation in South America and marine offlap on other continents. In

Feist, R., Talent, J.A., and Daurer, A. (eds.), Noth Gondwana: Mid-

Paleozoic Terraines, Stratigraphy, and Biota. pp. 239–257.

Isbell, J.L., Miller, M.E., Wolfe, K.L., and Lenaker, P.A., 2003. Timing of

late Paleozoic glaciation in Gondwana: Was glaciation responsible for

the development of northern hemisphere cyclothems? In Chan, M.A.,

and Archer, A.W. (eds.), Extreme Depositional Environments: Mega

end Members in Geologic Time. Boulder, CO: Geological Society of

America, pp. 5–24.

Kalvoda, J., 1986. Upper Frasnian and lower Tournaisian events and evolu-

tion of calcareous foraminifera - close links to climatic changes. In

Walliser, O. (ed.), Global Bio-Events. Berlin: Springer, pp. 225–236.

Kelley, P.H., and Raymond, A., 1991. Migration, origination, and extinc-

tion of Southern Hemisphere brachiopods during the Middle Carboni-

ferous. Palaeogeogr. Palaeoclimatol., Palaeoecol., 86,23–39.

McGhee, G.R. Jr., 1996. The Late Devonian Mass Extinction. New York:

Columbia University Press, 303pp.

Mii, H.-S., Grossman, E.L., Yancey, T.E., Chuvashov, B., and Egorov, A.,

2001. Isotopic records of brachiopod shells from the Russian Platform -

evidence for the onset of mid-Carboniferous glaciation. Chem. Geol.,

175, 133–147.

Otto-Bliesner, B., 1993. Tropical mountains and coal formation: A climate

model study of the Westphalian (306 MA). Geophys. Res. Lett., 20,

1947–1950.

Parrish, J.T., 1993. Climate of the Supercontinent Pangea. J. Geol., 101,

215–233.

Raymond, A., 1985. Floral diversity, phytogeography, and climatic ameli-

oration during the Early Carboniferous (Dinantian). Paleobiology, 11,

293–309.

Raymond, A., 1996. Latitudinal patterns in the diversification of

mid-Carboniferous land plants: Climate and the floral break. In Leary,

Figure L11 Middle and Upper Permian paleogeographic reconstruction showing climate zones and the distribution of sedimentary indicators of

middle and late Permian climate (unpublished data, A. J. Boucot, C. Xu, and C.R. Scotese).

LATE PALEOZOIC PALEOCLIMATES 503

R.L. (eds.), Patterns in Paleobotany: Proceedings of a Czech-U.S. Car-

boniferous Paleobotany Workshop. Springfield, IL: Illinois State

Museum, pp. 1–18.

Raymond, A., and Metz, C., 2004. Ice and its consequences: Late Ordovi-

cian, Late Devonian, Pennsylvanian-Permian, and Cenozoic glaciation

compared. J. Geol., 112, 655–670.

Raymond, A., Kelley, P.H., and Lutken, C.B., 1989. Polar glaciers and life

at the equator: The history of Dinantian and Namurian (Carboniferous)

climate. Geology, 17, 408–411.

Raymond, A., Kelley, P.H., and Lutken, C.B., 1990. Dead by degrees:

Articulate brachiopods, paleoclimate, and the mid-Carboniferous

extinction event. Palaios, 5, 111–123.

Rees, P.M., Ziegler, A.M., Gibbs, M.T., Kutzbach, J.E., Behling, P.J., and

Rowley, D.B., 2002. Permian phytogeographic patterns and climates:

New data and model comparisons. J. Geol., 110,1–31.

Ross, C.A., and Ross, J.A., 1988. Late Paleozoic transgressive and regres-

sive deposition. In Wilgus, C.K., Hastings, B.S., Kendall, C.G.S.C.,

Posamentier, H.W., Ross, C.A., and Wagoner, Van J.C. (eds.), Sea-Level

Changes: An Integrated Approach. Tulsa, OK: Society of Economic

Paleontologists and Mineralogists, pp. 227–247.

Rowley, D.B., Raymond, A., Parrish, J.T., Lottes, A.L., Scotese, C.R., and

Ziegler, A.M., 1985. Carboniferous paleogeographic, phytogeographic,

and paleoclimatic reconstructions. Int. J. Coal Geol., 5,7–42.

Scrutton, C.T., 1988. Patterns of extinction and survival in Paleozoic corals.

In Larwood, G.P. (eds.), Extinction and Survival in the Fossil Record.

Systematics Association, pp. 65–88.

Sepkoski, J.J. Jr., 1986. Phanerozoic overview of mass extinctions. In

Raup, D.M., and Jablonski, D. (eds.), Patterns and Processes in the

History of Life. Berlin: Springer, pp. 277–295.

Streel, M., Caputo, M.V., Loboziak, S., and Melo, J.H.G., 2000. Late-

Frasnian-Famennian climates based on palynomorph analysis

and the question of Late Devonian glaciations. Earth Sci. Rev., 52,

121–173.

Valdes, P.J., 2000. Warm climate forcing mechanisms. In Huber, B.T.,

Macleod, K.G., and Wing, S.L. (eds.), Warm Climates in Earth History.

Cambridge, UK: Cambridge University Press, pp. 3–20.

Valentine, J.W., 1984. Neogene marine climate trends: Implications for

biogeography and evolution of shallow-sea biota. Geology, 12,

647–650.

Veevers, J.J., and Powell, C.M., 1987. Late Paleozoic glacial episodes

in Gondwanaland reflected in transgressive-regressive depositional

sequences in Euramerica.

GSA Bull., 98, 475–487.

Witzke, B.J., 1990. Paleoclimates in Laurentia and Euramerica. In

McKerrow, W.S., and Scotese, C.R. (eds.), Paleozoic Palaeogeography

and Biogeography. Bath, UK: Geological Society of London,

pp. 57–73.

Wright, V.P., and Vanstone, S.D., 2001. Onset of Late Paleozoic glacio-

eustasy and the evolving climates of low latitude areas: A synthesis of

current understanding. J. Geol. Soc. (Lond.), 158, 579–582.

Zachos, J., Pagani, M., Sloan, L., Thomas, E., and Billups, K., 2001.

Trends, rhythyms, and aberrations in global climate 65 Ma to present.

Science, 292, 686–693.

Ziegler, A.M., Bambach, R.K., Parrish, J.T., Barrett, S.F., Gierlowski, E.H.,

Parker, W.C., Raymond, A., and Sepkoski, J.J.J., 1981. Paleozoic bio-

geography and climatology. In Niklas, K.J. (eds.), Paleobotany, Paleoe-

cology, and Evolution. New York: Praeger, pp. 231–266.

Ziegler, A.M., Raymond, A.L., Gierlowski, T.C., Horrell, M.A.,

Rowley, D.B., and Lottes, A.L., 1987. Coal, climate and terrestrial

productivity: The present and early Cretaceous compared. In Scott, A.C.

(eds.), Coal and Coal-Bearing Strata: Recent Advances. London: Geologi-

cal Society, pp. 25–49.

Cross-references

Animal Proxies, Invertebrates

Coal Beds, Origin and Climate

Cyclic Sedimentation (Cyclothems)

Early Paleozoic Climates (Cambrian-Devonian)

Glacial Eustasy

Glaciations, Pre-Quaternary

“Greenhouse” (warm) Climates

“Icehouse” (cold) Climates

Mass Extinctions: Role of Climate

Paleobotany

Plate Tectonics and Climate Change

Sedimentary Indicators of Climate Change

Stable Isotope Analysis

LATE QUATERNARY MEGAFLOODS

The last major deglaciation of planet Earth (the last portion of

Marine Isotope Stage 2, from about 20,000 to 11,000 calendar

years ago) involved huge fluxes of water from the wasting con-

tinental ice sheets. It has become increasingly apparent that

much of this water was delivered as floods of immense magni-

tude and relatively short duration. These late Quaternary mega-

floods had short-term peak flows, comparable in discharge to

the more prolonged fluxes of ocean currents (The unit of dis-

charge for both ocean currents and megafloods is the Sverdrup,

equivalent to 1 million cubic meters per second). Some out-

burst floods likely induced very rapid, short-term effects on

Quaternary climates. The late Quaternary superfloods also

greatly altered drainage evolution and the planetary patterns

of water and sediment movement to the oceans.

Historical context

During the eighteenth and early nineteenth centuries there were

many studies of landscape phenomena that posited the action

of phenomenally large floods. Some of these studies invoked

a kind of biblical catastrophism, but other studies merely

employed hypotheses of immense floods because these phenom-

ena seemed to provide the best explanations for such features as

scoured bedrock and accumulations of huge, water-transported

boulders. Unfortunately, a logically flawed doctrine of uniformi-

tarianism was promoted in the later eighteenth century (Baker,

1998), effectively retarding progress on understanding the role

of cataclysmic flooding in Quaternary processes. The renais-

sance in catastrophic flood studies of the later twentieth century

was made possible because of a long-lasting controversy over

the origin of the Channeled Scabland in the northwestern United

States. Extending from the 1920s to the 1960s, the great “scab-

lands debate” eventually led to general acceptance of the cata-

clysmic flood origin that was championed by J Harlen Bretz, a

professor at the University of Chicago.

Based on Bretz’s studies of the Channeled Scabland, the

distinctive erosional and depositional features of cataclysmic

flooding are well known. These include scabland bedrock ero-

sion (Figure L12), streamlining of residual uplands, large-scale

scour around obstacles, depositional bars, giant current ripples

(fluvial gravel dunes), and huge sediment fans. These features

can be identified by orbital remote sensing imagery and by

field reconnaissance. These features could be used to discover

and document cataclysmic flood effects on more river basins

globally.

Megaflood landscapes

Cordilleran

The Channeled Scabland landscape developed south of the

Cordilleran Ice Sheet, in the northwestern United States. The

Purcell Lobe of the ice sheet extended south from British

Columbia to the basin of modern Pend Oreille Lake in northern

Idaho. It thereby impounded the Clark Fork River drainage to

504 LATE QUATERNARY MEGAFLOODS

the east, forming glacial Lake Missoula in western Montana. At

maximum extent, this ice-dammed lake held a water volume of

about 2,500 km

with a depth of 600 m at the dam. Repeated

failures through subglacial tunnels are inferred to have

occurred between about 18,000 and 15,000 years ago (Waitt,

1985). However, these multiple outburst events were of greatly

differing magnitudes. About 15 events exceeded 3 Sverdrups,

and at least one of these reached about 10 Sverdrups in dis-

charge (Benito and O’Connor, 2003). The largest failure or fail-

ures probably involved a different source mechanism than that

envisioned by Waitt (1985).

Upon reaching the Pacific Ocean, the Missoula floodwaters

continued flowing down the continental slope as hyperpycnally-

generated turbidity currents (Normark and Reid, 2003). The

sediment-charged floodwaters followed the Cascadia submarine

channel into and through the Blanco Fracture Zone and out onto

the abyssal plain of the Pacific. As much as 5,000 km

of sediment

may have been carried and distributed as turbidites over a distance

of 2,000 km west of the Columbia River mouth.

Laurentide

As the Laurentide Ice Sheet of central and eastern Canada

retreated from its late Quaternary maximum extent, it was

bounded to the south and west by immense meltwater lakes,

which developed in the troughs that surrounded the ice. As

the lake levels rose, water was released as great megafloods,

which carved numerous spillways into the drainages of the

Mississippi, St. Lawrence, and Mackenzie Rivers (Kehew and

Teller, 1994). The greatest megafloods developed from the last

of the ice-marginal lakes, a union between Lake Agassiz in

south-central Canada and Lake Ojibway in northern Ontario.

The resulting megalake held about 160,000 km

of water,

which was released subglacially about 8,400 years ago, under

the ice sheet and into the Labrador Sea via the Hudson Strait

(Clarke et al., 2003).

Eurasian spillways

In the 1970s, Mikhail G. Grosswald recognized that the late

Quaternary ice-sheet margins of northern Eurasia, like those

of northern North America, held huge proglacial lakes, and

great spillways that developed for the diversion of drainage.

Not only was meltwater diverted to the south-flowing Dnieper

and Volga Rivers, but the great north-flowing Siberian rivers,

the Irtysh, Ob, and Yenesei were impounded by ice sheets that

covered the modern Barents and Kara Seas. Though Grosswald’s

reconstructions created considerable controversy, recent work

shows that these impoundments did indeed occur. However, the

largest diversions are now dated to about 90,000 years ago, when

ice-sheet growth was enhanced by the climatic influenceof the ice-

dammed lakes (Krinner et al., 2004). The largest megalake, Lake

Mansi, drained southward through the Turgai divide of north-

central Kazakhstan to the basin of the modern Aral Sea. The Aral,

in turn, drained though a spillway at its southwestern end into the

Caspian, which expanded to a late Quaternary size over twice its

modern extent, known as the Khvalyn paleolake. The Khvalyn

paleolake spilled westward through the Manych spillway into

the Don valley, then to the Sea of Azov, and through the Kerch

Strait to the Euxine Abyssal Plain (the floor of the modern Black

Sea). The huge size of the Manych spillway, up to 35 km wide,

implies that flows may have reached a peak of 10 Sverdrups

(Baker, 1996). During the last glaciation, the Black Sea was

isolated from the Mediterranean Ocean, and it filled with fresh

water derived from the great system of glacially augmented rivers

entering it from the north. The freshwater faunal facies of the

Black Sea show that the Caspian was spilling through the Manych

spillway around 18,000 years ago and the Black Sea was spilling

into the Sea of Marmara, which, in turn, spilled via the Hellespont

into the Agean Sea.

Central Asian uplands

A number of cataclysmic flood landscapes are now recognized

in the mountain areas of central Asia. The best studied of these

regions is the Altai Mountains of south-central Siberia (Carling

et al., 2002). The Altai flooding derived from the Chuja-Kuray

ice-dammed lake, which may have held as much as 1,000 km

of water at a maximum depth of up to 900 m (Baker et al.,

1993). Downstream of this ice dam, the Chuja and Katun River

valleys are characterized by immense gravel bars, emplaced

into various valley-side embayments by the flooding. The

bars are up to 5 km long and 120 m in height. The bar surfaces

and associated run-up layers of flood-transported gravel indi-

cate maximum flow depths of about 320 m. Flow modeling

of the associated paleoflood discharges retrodicts a peak flow

of about 11 Sverdrups and mean flow velocities of about

30 m s

1

(Herget and Agatz, 2003).

Another region with extensive evidence of late Quaternary

megafloods is Tuva. The Tuvan paleofloods derived from an

ice-dammed lake in the Darkhat depression of north-central

Mongolia (Grosswald and Rudoy, 1996). The paleolake held

about 250 km

of water with a depth of 200 m at its ice dam.

The paleoflooding entered the upper Yenesei River, and it

was remarkable for the emplacement of spectacular trains of

giant current ripples (gravel dunes) near the Tuva capital city

of Kyzyl (Figure L13).

In Kirgizstan, Lake Issyk-Kul, the world’s second largest

modern mountain lake, was transformed into an even larger

ice-dammed lake during the last glaciation. Iceberg-emplaced

landforms rise to 300 m above the present lake level (Gross-

wald et al., 1994). Spectacular erosion at the full-glacial

lake outlet, Boam Canyon, indicates that the paleolake failed

by cataclysmic outburst flooding. This flood emplaced an enor-

mous outwash fan, extending 85 km from the mouth of Boam

Figure L12 Aerial view of giant potholes and butte-and-basin scabland

eroded into the basalt bedrock of eastern Washington by the

cataclysmic Missoula megafloods, approximately 15,000 years ago.

These potholes measure about 15 m in depth and 40 in width. They are

in Lenore Canyon, a part of the lower Grand Coulee. Lake Lenore is

visible at the upper right.

LATE QUATERNARY MEGAFLOODS 505

Canyon. A similar but smaller fan occurs on the upper Yenesei

River, where the Tuvan floods emerged from canyons of the

West Sayan Mountains and entered the Abakan Basin. Another

area of probable glacial cataclysmic flooding is Lake Baikal,

which may have expanded during full-glacial periods by ice

blockage of its outlet, thereby raising its level of spill points.

Current controversies

Black Sea floodi ng

Although late Quaternary freshwater inundation of the Black

Sea is well recognized, it is usually correlated with enhanced

proglacial meltwater flow via the many rivers draining south-

ward from the northern Eurasian ice sheets, as noted above.

A recent controversy has arisen, however, over the marine

influx to the Black Sea that occurred during rising Holocene

sea level. This influx caused Mediterranean water to spill

through the Hellespont and Bosporus, reaching the Black Sea

about 9,500 years ago. One view (Ryan et al, 2003) holds that

the global ocean rose to the level of a relatively shallow sill of

the Bosporus outlet and catastrophically spilled into the Black

Sea basin, which then was partly filled with freshwater to a

level about 85 m below that of today. The resulting cataclysmic

inundation presumably displaced a large human population in a

calamity that is equated to the Noachian flood myth. An alter-

native view is that much of the Bosporus is underlain by fresh-

water facies of late Pleistocene age, derived from the Black

Sea. The minimal erosion of these sediments is not consistent

with the idea that cataclysmic flooding led to the overlying

Holocene marine facies of Mediterranean origin.

Subglacial megafloods

A variety of enigmatic landforms, involving water erosion and

deposition, develop beneath very large continental ice sheets.

These landforms include drumlins, Rogen moraines, large-scale

bedrock erosional marks, and tunnel channels (valleys). Though

commonly explained by subglacial ice deformational processes,

the genesis of these features cannot be observed in modern glaciers

that are much smaller than their late Quaternary counterparts. The

assemblage of landforms as part of an erosional/depositional

sequence beneath continental ice sheets that precedes regional

ice stagnation and esker formation was explained by Shaw

(1996) as resulting from a phase of immense subglacial sheet

floods, which, in turn, follow ice-sheet advances that terminate

with surging, stagnation and melt-out. Shaw (1996)proposedthat

peak discharges of tens of Sverdrups are implied by the late

Quaternary subglacial landscapes of the southern Laurentide Ice

Sheet. Shoemaker (1995) provided some theoretical support for

Shaw’s model, though at smaller flow magnitude levels.

Global consequences

Ocean currents are integral to Earth’s climate engine, distri-

buting heat between the equator and the poles. For example,

the Gulf Stream flows at discharges of up to 100 Sverdrups,

involving relatively slow moving water 1 km in depth and

50–75 km wide. As this seawater evaporates and becomes

more saline, it increases in density, eventually sinking in the

northern Atlantic. This process forms a portion of the great

thermohaline circulation, which acts as a conveyor belt for heat

in the oceans. Cataclysmic megafloods can introduce relatively

low-density, freshwater lids over large areas of ocean surface.

Freshwater can enter the oceans very rapidly, at flow rates of

up to 10 Sverdrups and volumes of thousands of km

. The

resulting disruption of sea-surface temperatures and density

structure can drastically alter climates, at least on time scales

of decades to centuries.

The connections between Pleistocene ice sheets and the

oceans are still very poorly understood. An important link is

inferred from relatively brief (100- and 500-year) intervals in

which thick marine layers of ice-rafted material were widely

distributed across the North Atlantic. Called “Heinrich events,”

these layers are thought to record episodes of massive iceberg

discharge from unstable ice sheets. The youngest Heinrich

events date to 17,000 (H1), 24,000 (H2) and 31,000 (H3) years

ago. Closely related is the Younger Dryas (YD) event, a global

cooling at 12,000 years ago, which is also associated with

North Atlantic ice-rafted rock fragments. There is evidence

from corals in Barbados that sea level rose spectacularly, about

20 m during H1 and about 15 m just after YD. Such rises

would require short-term freshwater fluxes to the oceans of

about 14,000 and 9,000 km

1

, respectively. (Fairbanks,

1989). These events are thought to relate to ice-sheet collapse,

reorganization of ocean-atmosphere circulation, and release of

subglacial and proglacial meltwater, most likely during epi-

sodes of cataclysmic megaflooding.

Victor R. Baker

Bibliography

Baker, V.R., 1996. Megafloods and glaciation. In Martini, I.P. (eds.), Late

Glacial and Postglacial Environmental Changes. New York: Oxford

University Press, pp. 98–108.

Baker, V.R., 1998. Catastrophism and uniformitarianism: Logical roots and

current relevance. In Blundell, D.J., and Scott, A.C. (eds.), No. 143,

Lyell: The Past is the Key to the Present. London: Geological Society

Special Publications, pp. 171–182.

Baker, V.R., Benito, G., and Rudoy, A.N., 1993. Paleohydrology of late

Pleistocene superflooding, Altai Mountains, Siberia. Science, 259,

348–350.

Benito, B., and O’Connor, J.E., 2003, Number and size of last-glacial

Missoula floods in the Columbia River valley between Pasco

Basin, Washington and Portland, Oregon. Geol. Soc. Am. Bull., 115,

624–638.

Figure L13 Aerial photograph of giant current ripples (gravel dunes)

located near Kyzyl, Tuva. The ripples occur on a gravel bar just north of

the Yenesei River (bottom of photograph) about 25 km southwest

of Kyzyl. There are nearly 100 ripples on a bar 6 km in length.

Ripple heights vary from 2 to 3 m. The photograph shows an

area 6 by 4 kilometers. The paleoflood flow was moving from

right (east) to left (west).

506 LATE QUATERNARY MEGAFLOODS

Carling, P.A., Kirkbride, A.D., Parnachov, S., Borodavko, P.S., and

Berger, G.W., 2002. Late Quaternary catastrophic flooding in the Altai

Mountains of south-centra l Siberia: A synoptic over view and introduc -

tion to flood deposit se dimentology. I n Martini, I .P., Baker, V.R., a nd

Garzon, G. (eds.), Floo d a nd Megaflood Processes an d Deposits :

Recent and Ancie nt Examples. Inter national Association of Sedimen-

tologists, Special Publication No. 32, pp. 17–35.

Clarke, G., Leverington, D., Teller, J., and Dyke, A., 2003. Superlakes,

megafloods, and abrupt climate change. Science, 301, 922–923.

Fairbanks, R.G., 1989. A 17,000-year glacio-eustatic sea level record:

Influence of glacial melting rates on the Younger Dryas event and deep

ocean circulation. Nature, 342, 631–642.

Grosswald, M.G., and Rudoy, A.N., 1996. Quarternary glacier-dammed

lakes in the mountains of Siberia. Polar Geogr., 20, 180–198.

Grosswald, M.G., Kuhle, M., and Fastook, J.L., 1994. Wurm glaciation of

Lake Issyk-Kul area, Tian Shan Mts: A case study in glacial history of

central Asia. Geojournal, 33, 273–310.

Herget, J., and Agatz, H., 2003. Modelling ice-dammed lake outburst floods

in the Altai Mountains (Siberia) with HEC-RAS. In Thorndycraft, V.R.,

Benito, G., Barriendos, M. and Llasat, M.S. (eds.), Palaeofloods, Histor-

ical Data and Climatic Variability: Applications in Flood Risk Assess-

ment. Madrid, Proceedings of the PHEFRA Workshop, Centro de

Ciencias Medioamientales, pp. 177–181.

Kehew, A.E., and Teller, J.T., 1994. History of the late glacial runoff along

the southwestern margin of the Laurentide Ice Sheet. Quaternary Sci.

Rev., 13, 859–877.

Krinner, G., Mangerud, J., Jakobsson, M., Crucifix, M, Ritz, C., and

Svendsen, J.I., 2004. Enhanced ice sheet growth in Eurasia owing to

adjacent ice-dammed lakes. Nature, 427, 429–432.

Normark, W.R., and Reid, J.A., 2003. Extensive deposits on the Pacific

Plate from Late Pleistocene North-American glacial lake bursts.

J. Geol., 111, 617–637.

Ryan, W.B.F., Major, C.O., Lericolais, G., and Goldstein, S.L., 2003. Cat-

astrophic flooding of the Black Sea. Annu.Rev. Earth Planet. Sci., 31,

525–554.

Shaw, J., 1996. A meltwater model for Laurentide subglacial landscapes. In

McCann, S.B., and Ford, D.C. (eds.), Geomorphology sans Frontieres.

New York: Wiley, pp. 182–226.

Shoemaker, E.M., 1995. On the meltwater genesis of drumlins. Boreas,

24,3–10.

Waitt, R.B. Jr., 1985. Case for periodic, colossal jokulhlaups from Pleisto-

cene glacial Lake Missoula. Geol. Soc. Am. Bull., 96, 1271–1286.

Cross-references

Cordilleran Ice Sheet

Drumlins

Glacial Geomorphology

Glacial Megalakes

Heinrich Events

Last Glacial Termination

Laurentide Ice Sheet

Proglacial Lacustrine Sediments

Thermohaline Circulation

Younger Dryas

LATE QUATERNARY-HOLOCENE VEGETATION

MODELING

Introduction

Vegetation modeling is central to understanding the role that

plants have played in shaping the Earth’ s atmospheric compo-

sition and climate in the past, and the potential responses and

feedbacks of vegetation to future climate changes. Over the

past decade, a series of global vegetation models have been

applied to the subject of late Quaternary and Holocene climate

change. Because a vegetation model provides a complete

picture of the Earth’s land surface, and because it produces a

diagnostic that can be directly compared to paleo-observations

(e.g., plant assemblages), vegetation models have been particu-

larly useful for evaluating the output of general circulation

models (GCMs). Progress in vegetation modeling has enabled

the researcher to simulate vegetation patterns with more detail,

and to evaluate other variables with paleoproxies, such as dust

fluxes, methane emissions, and carbon storage. Current direc-

tions in late Quaternary and Holocene vegetation modeling

include the incorporation of a vegetation model directly into

the land surface scheme of GCMs so that vegetation may inter-

act in real time with the climate system through the atmosphere

and oceans.

Over the past decade, a variety of process-based, global vegeta-

tion models have been developed to study a variety of problems

in biogeography, biogeochemistry, and atmospheric composition.

These models have been primarily applied to the present-day, but

are increasingly being used to study the recent geological past. For

paleo-applications, global vegetation models play an important

role, because other methods of inferring past vegetation patterns

and processes leave the investigator with information gaps, either

spatially, temporally, or both. Paleoproxy data from a variety of

sources (e.g., pollen, packrat middens) can be used to infer vegeta-

tion history and biogeography at a point, but extrapolation is diffi-

cult given the sensitivity of vegetation to climatic and edaphic

factors beyond the catchment area of the pollen source. The com-

position of trace gases measured in gas bubbles from ice cores

contains some signal of the biogeochemical processes active in

the terrestrial biosphere, such as vegetation productivity, methane

emissions from wetlands, and compounds produced during bio-

mass burning. However, these ice core records integrate the cli-

matic or environmental signals over entire hemispheres, and thus

provide little information about their spatial distribution, which

is important in understanding the influence of the biosphere on

the climate system and interactions between vegetation and the

animal world. Furthermore, the land surface signal in ice cores

must often be separated from oceanic activity. For some radia-

tively important trace gases, including CO

, this is a difficult task.

Thus, to study late Quaternary and Holocene vegetation history, a

global vegetation model is an essential tool: as an integrator over

both space and time and as a method to estimate the contribution

of the land surface to the budgets of greenhouse gases, an impor-

tant quantity not only at present but also in the past.

Vegetation model structure and processes

The past decade has seen great progress in the development

and application of global vegetation models. The latest vegeta-

tion models incorporate both biogeographical and biogeochem-

ical processes to simulate vegetation distribution and the fluxes

of carbon, water, nutrients, and other trace gases through the

terrestrial biosphere. This capability makes current vegetation

models a powerful tool for investigating the Earth’s surface

during the late Quaternary and Holocene, as a variety of pro-

blems may be addressed.

Methods

The Plant Functional Type (PFT) is the basic unit of calculation

in most global-scale vegetation models. Each PFT is defined to

be a generic representation of all plant taxa that share specific

characteristics such as growth form (e.g., tree or grass), pheno-

logical habit (e.g., deciduous or evergreen; summergreen or

raingreen), leaf morphology (e.g., broadleaf or needleleaf),

LATE QUATERNARY-HOLOCENE VEGETATION MODELING 507

and other parameters such as root depth distribution and maxi-

mum photosynthetic rate. At the computational core of a vege-

tation model is a scheme that determines the combination of

PFTs that make up the vegetation at a certain location. PFT

combinations are often combined into assemblages, termed

“biomes”; these biomes represent the major vegetation types

found on earth (Table L2).

The simplest form of vegetation model (e.g., BIOME1;

Prentice et al., 1992) uses a set of rules based on absolute biocli-

matic limits to determine the presence or absence of PFTs and then

assigns a biome based on a simple hierarchy of these PFTs. More

advanced vegetation models, e.g., BIOME4 (Kaplan, 2001), use a

coupled carbon and water flux scheme to determine the leaf area

index (LAI) that maximizes net primary productivity (NPP) for

any given PFT. The PFT hierarchy, and thus the resulting biome,

is determined by the list of PFTs ranked by LAI and NPP. Addi-

tional rules or climate-based indices may also be used to further

refine the biome classification. An advantage of this type of

coupled biogeography-biogeochemistry model is that both vegeta-

tion structure and the impact of vegetation on global biogeochem-

istry may be simultaneously investigated. Finally, dynamic

vegetation models, e.g., LPJ or IBIS, explicitly simulate the com-

petition among PFTs as they evolve through time.

In the most recent global vegetation models (e.g., BIOME4,

LPJ), model operation is based on 12 or 13 PFTs representing

physiologically distinct classes from arctic/alpine cushion

forbs to tropical evergreen trees. Each PFT is assigned biocli-

matic limits that determine whether its net primary productivity

(NPP) is calculated for a given grid cell.

Equilibrium versus dynamic global vegetation models

Despite the recent focus on developing dynamic global vegeta-

tion models (DGVMs) for carbon cycle research (Cox et al.,

2000; Sitch et al., 2003), equilibrium vegetation models that

provide a “snapshot” view of the Earth’s surface at a given time

remain important for late Quaternary and Holocene vegetation

modeling. Equilibrium vegetation models are currently far

more detailed and precise than DGVMs in their ability to simu-

late biogeographic patterns. Additionally, operation of a DGVM

requires 10 to 100 times more overhead than an equilibrium

vegetation model in terms of the size of input and output datasets

and computing power. As such, a DGVM is neither necessary

nor appropriate for addressing every research problem. Equili-

brium vegetation models provide an efficient method to investi-

gate global ecosystem theory and a required test-bed for

developing methods for simulating the behavior of vegetation

that may later be applied in a DGVM context. A DGVM is

required to study periods of rapidly changing climate where eco-

system “memory” is important, e.g., biogeochemical feedbacks

in the anthropogenically-modified environment of the present

and the near future. Equilibrium models are more suited to “time

slice” analyses where an equilibrium approximation is adequate

and interest focuses on regional phenomena that are not as well

simulated with current DGVMs.

On time scales greater than ca. 10

years, vegetation

dynamics are controlled by two main factors: shifts in absolute

limits of tolerance for certain plant types and competition

among coexisting plant taxa due to shifts in disturbance regime

or optimum climate. Thus, most modeling to date for late

Quaternary and Holocene studies focuses on simulating the

potential, natural equilibrium vegetation at an instant in time

(i.e., a time slice). A major reason for choosing time slices

for vegetation reconstructions is that paleodata have been

assembled for particular time slices, which may be used to

evaluate the performance of the model. Furthermore, the com-

putational requirements of producing a time slice simulation

of vegetation are much simpler than for a dynamic scenario.

Paleovegetation simulation

For simulation of vegetation of the past, a vegetation model

requires paleoclimatology data, which are usually generated

from one or an ensemble of paleoclimate simulations from a

GCM. Vegetation models can be coupled to the GCM climatol-

ogy in a variety of ways (Figure L14). In the simplest case, the

vegetation model is run “offline” through one-way forward

coupling to the GCM climatology. This method has the advan-

tage of being computationally efficient and enables a great deal

of flexibility in the choice of GCM output and particular vege-

tation model used.

To investigate certain phenomena where particular feed-

backs between the terrestrial biosphere and the climate system

are thought to be important, forms of coupling that are more

complicated have been applied. Asynchronous coupling uses

a GCM simulation to drive an offline vegetation model, which

then provides the land surface conditions (primarily the distribu-

tion of the major biomes) to make a subsequent run of the GCM.

After three or four iterations, the GCM climatology is usually in

equilibrium with the vegetation model. Asynchronous coupling

is particularly useful for simulating long-timescale climate

changes, such as those that occur on millennial scales, and has

been successfully applied to the problem of glacial inception

and late Pleistocene deglaciation. Finally, a vegetation model

Table L2 Biomes simulated by BIOME4

Tropical forests

Tropical evergreen broadleaf forest

Tropical semi-evergreen forest

Tropical deciduous broadleaf forest and woodland

Temperate forests

Temperate deciduous broadleaf forest

Temperate evergreen needleleaf forest

Warm-temperate evergreen broadleaf and mixed forest

Cool mixed forest

Cool evergreen needleleaf forest

Cool-temperate evergreen needleleaf and mixed forest

Cold forests

Cold evergreen forest

Cold deciduous forest

Forest, Shrub and Grassland combinations

Tropical savanna

Tropical xerophytic shrubland

Temperate xerophytic shrubland

Temperate sclerophyll woodland and shrubland

Temperate deciduous broadleaf savanna

Temperate evergreen needleleaf open woodland

Cold parkland

Grassland

Tropical grassland

Temperate grassland

Tundra

Graminoid and forb tundra

Low- and high-shrub tundra

Erect dwarf-shrub tundra

Prostrate dwarf-shrub tundra

Cushion-forb, lichen, and moss tundra

Sparse or no vegetation

Desert

508 LATE QUATERNARY-HOLOCENE VEGETATION MODELING