A quick background to the Pliocene

Jonathan Adams, Environmental Sciences Division, Oak Ridge National Laboratory

The Pliocene (5.4 - 2.4 million years ago) is the uppermost subdivision of the long Tertiary period which began 64 million years ago; it represents the final stages of a global cooling trend that led up to the Quaternary ice ages.

Generally, the Pliocene world was rather warmer than at present. The ancient distribution of warm-climate ocean plankton, and of animal and plant fossils on the land, shows that mean annual temperatures in the mid-latitudes were often several degrees higher. The greatest warming seems to have been in the Arctic and cool temperate latitudes of the Northern Hemisphere, where temperatures were often warm enough to allow species of of animals and plants to exist hundreds of kilometres north of the ranges of their nearest present-day relatives. Because there was less ice volume near the poles, sea levels may have been as much as 30m higher than at present during the warmest intervals. The peak phases of warmth during the Pliocene were mostly during the interval between 3 and 4 million years ago (the mid-Pliocene), although almost all of the Pliocene was warmer than today's world.

The causes of the generally warmer climates of the Pliocene are something of a mystery. The warmth may have been related to changes in ocean circulation patterns, perhaps combined with higher concentrations of greenhouse gases in the atmosphere, as this site by M. Chandler explains.

Problems in fitting together data from the last 5 myr

There are many problems associated with reconstructing lower Quaternary (2.4-0.9 myr) and Pliocene (5-2.4 myr) vegetation and climates.

Long ocean sediment cores from the deep waters off the coasts of Africa and Arabia have yielded several relatively detailed records of environmental conditions on the continent during the last few million years (DeMenocal & Bloemendal 1996, Partridge et al. 1996, PRISM members 1996), allowing climate changes on the timescale of several thousand years to be detected. Deep lakes in Africa, South America, Europe and Siberia have also yielded long cores spanning the last several million years. However, it is difficult to correlate these cores with one another, and with the scattered information available from other sites in the interior of the landmasses.

What the long sediment cores do show is that even relatively short subdivisions of the Pliocene (as well as the Quaternary) were characterised by almost continuous flux and oscillation in aridity and temperature, occurring on the timescale of several thousand years (often with a detectable periodicity of 19,000-21,000 years, probably linked to 'Milankovitch Cycles'; variations of the Earth's position in its orbit). It is probable that many other, abrupt and shorter timescale, events remain 'hidden' from our view of the Pliocene because of the lack of resolution of the long palaeoenvironmental records. The best that one can hope for in terms of summarizing Pliocene climates is to talk of them in a statistical sense; in terms of averages, mean variance, period of oscillation etc. as they appear in the record. These statistical characteristics themselves shift throughout the period of the last few million years going into the Quaternary Period, with a decline in mean temperature and a trend towards increased aridity, and broader oscillations in both temperature and aridity.

The published terrestrial record (PRISM members 1996, Bonnefille 1996, Scott 1996) is for the most part limited to a relatively few scattered exposures. Given the limitations on accuracy of dating this far back in time, and the evidence for frequent oscillations in the long ocean and lake cores, it is difficult to fit these into any broader framework.

Thus at present, a general shortage of Pliocene palaeodata and the difficulties in dating and correlation against this variable background make it impossible to confidently map vegetation or climate for any one particular point in time before about 130,000 years ago (the last interglacial). As a hypothesis one might suggest that a coherant 'cold and dry' and 'warm and wet' pattern of African and south Asian climates has operated over about the last 24 kyr holds true for Plio-Pleistocene time in general. If so, one can suggest that vegetation and climate maps of the Last Glacial Maximum (about 23-18 kyr), the present-potential (0 kyr) and early Holocene (about 8-7 kyr) be used respectively as a hypothesis of how the numerous colder and warmer oscillations of the past few million years would have looked. However, even this principle might be unreliable; some data suggest that at particular pre-LGM stages during the Last Glacial period, climates in Africa were cooler than present, but about as moist or even moister than at present. Other data from the same time periods contradict this, and suggest that the coherant 'greater cold = greater aridity' pattern has indeed held true during the last 130,000 years (the contradictions may be due to errors in the dating, though which side is correct is presently unclear).

Whether or not there was a close relationship between temperature and aridity during Pliocene, the general band of variability seen in the long cores suggests that even the most arid oscillations of the Pliocene were probably nowhere near as arid as the Last Glacial Maximum, and during the generally warmer parts of the Pliocene even the dry minima of individual oscillations may actually have been wetter than at present. It is to be hoped that eventually, more sophisticated climate-vegetation models coupled to palaeoenvironmental data will be able to perform the task of broadscale vegetation reconstruction for the Pliocene. Maslin (1996) notes ocean core evidence that the sinking zone of water that pulls heat northwards along the Gulf Stream was operating less strongly during at least some of the cooler oscillations of the Pliocene, suggesting analogous mechanisms between the Pliocene climate oscillations and those which occurred during the later Quaternary. Other oceanographic evidence suggests a stronger-than-present Gulf Stream flow during the generally warmer phase of the middle Pliocene, offering another possible analogy with the fluctuations in the Late Quaternary world.

African Pliocene environments.

The most extensive record of the Pliocene relates to the African continent, although even here the evidence of environments on land is fragmentary, and the ocean cores and lake cores are difficult to interpret and to correlate with one another. The general picture for the early to mid Pliocene (From 5 myr until around 3 myr ago) is of moister-than-present conditions in Africa, with greater forest and tree cover and less desert.

Before 3 myr, North Africa generally seems to have been much moister than present, with semi-arid vegetation covering the present-day Sahara belt, and tropical forest and moist savanna extending at times as far as 21 deg.N (PRISM members 1996).

In SW Ethiopea/NW Kenya, the area of various classic hominid discoveries, pollen (Bonnefille 1996) and soil carbon isotope (Kingston et al. 1994) evidence suggests that tropical open woodland vegetation was present in some areas during the lower Pliocene that are now virtually semi-desert or at best a treeless dry grassland. In the now treeless landscape of the Turkana Basin in Ethiopea, forest would have been present in the landscape, but probably only as isolated patches fringing lakes and rivers (Bonnefille 1996).

It is uncertain whether there was ever a continuous east-west forest belt across equatorial Africa during the Pliocene. Red lateritic soils of early Pliocene age occur in equatorial east Africa, east of the Rift Valley, and leave open the possibility that in the lower Pliocene tropical forest extended in a continuous band from coast to coast, at least intermittantly (H. Faure pers. comm. May 1996), although the soils might instead be the product of conditions that supported only an open woodland cover (R. Morley pers. comm. May 1996). A.T. Grove (pers. comm., May 1996) notes that if the anomalous weather conditions of 1961-2 were to persist for a few centuries one might expect rainforest to be able to form a continuous east-west belt across Africa. Record rainfall and river flows were recorded that year all across from west Africa (Ogowe, Congo, Oubangui and Chari) to western and central peininsular India. This was believed to be related to unusually high Arabian Sea and Indian Ocean temperatures, and illustrates that a relatively small shift in the mean of the climate (within the present range of year-to-year variability) could give a very different range of habitats across Africa. At present, however, the palaeoenvironmental evidence from the region is too sketchy to say if or when tropical forest last existed across equatorial east Africa.

In southern Africa, where Australopithecus was also present at the time, open woodlands and warm temperate forest were present in at least some areas that are now arid scrublands (Scott 1996).

Around 3 myr ago, African environments seem to have shifted more toward a drier and somewhat cooler state. In this time interval, pollen and other evidence suggests more widespread grasslands in east Africa, scrub in southern Africa, and extension of desert in north Africa.

The pollen, dust and isotope indicators of arid and cooler conditions generally increase into the Quaternary (starting 2.4 myr ago), during most of which environments were somewhat cooler and more arid than at present.

A probabilistic view of Plio-Pleistocene environments.

As mentioned above, rather than talking in terms of 'the' climate of a region during any particular interval of the Plio-Pleistocene, we should perhaps consider the period in terms of a probability distribution with extremes and averages of temperature and aridity that have shifted over time. An indicator of the general range of variation in Plio-Pleistocene global temperature (and also the linked factor of ice volume) is the fluctuation in oxygen isotopes in seawater recorded by deep ocean cores. There has been a band of fluctuation, with occasional extreme events of colder or warmer conditions. Whilst the averages shift, there is a large amount of overlap between the colder events of one period and the warmer events of the next. The change in climates and environments may well have been a quantitative shift in terms frequency of occurrence, rather than any real qualitative transition.

On a local scale, this background of variability should perhaps be considered when describing the environmental context of hominid sites. Even if one finds that a fossil-bearing or artefact-bearing horizon shows strong evidence of a woodland environment, this does not necessarily mean that the woodland cover remained stable on a timescale of thousands of years. Woodland might have been 'generally' predominant in a particular region during a particular time interval of tens of thousands of years, but for some parts of this time the woodland may have been in regression and replaced by grassland or even semi-desert. Short-lived arid phases often leave relatively little trace of their occurrence, because deposition is slowed, erosion is more likely to occur, and preservation of environmental indicator fossils is less likely.

Other variables which must be considered as part of the changing background to hominid ecology and evolution include the relative frequency and relative abruptness of transitions between wet and dry climate states, and the existence and duration of periods of relative stability (Foley 1993). For example, the oscillations seen in the deep cores change in frequency and amplitude during the period of the Plio-Pleistocene (DeMenocal & Bloemendal 1996). Between about 5 myr and 1.8 myr ago (extending from Pliocene to lower Quaternary), the period between each major peak and the next major peak in temperature or rainfall averaged about 19-23 kyr with a downturn in climate following within a few thousand years after each peak had been reached.

The periodicity of greatest variation in climate them shifted to 41 kyr, between 1.8 and 0.9 kyr. After 0.9 myr ago both the length and amplitude of the cycle lengthened, to about 100 kyr, and relatively stable warm interglacials lasting around 10 kyr tended to occur, interspersed with much longer and more variable glacial periods.

Shorter timescale variability during the last 130 kyr as an example of 'hidden' variability during the Pliocene and Quaternary.

Working back in time from the relatively detailed perspective of the last 130 kyr may have generalised lessons for understanding climates on the longer timescale of the last several million years. There are numerous signs that in the Atlantic and Sahara region, and probably also elsewhere, the cold dry last glacial period was punctuated by sudden but shortlived warmer and wetter episodes that often lasted between a few hundred or a couple of thousand years (van Andel & Tzedakis, in press), before an almost equally rapid return to the cold, dry glacial state. An example of the sudden onset of moist climates in the Sahara is the interstadial that occurred around 13,500-11,500 calendar years ago. This ended rapidly with the cold, dry Younger Dryas phase, that itself ended rapidly (possibly in the space of only a few years?) giving way to moister-than-present early Holocene climates. The Holocene itself seems to have been punctuated with intensely wet and dry phases in North Africa, that often lasted only a few decades or centuries before ending abruptly (Gasse & van Campo 1994, QEN members 1991). Examination of meteorological data for the past century or so likewise suggests that in many parts of Africa the decadal and year-to-year variability in rainfall is very great, with short, intense droughts alternating with periods of greatly increased rainfall (e.g. the extremely wet years in east Africa 1961-2, mentioned above, or the 'El Nino'-influenced drought in central African rainforests in 1982-3). It is possible, although now generally thought unlikely (van Andel & Tzedakis, in press), that the last interglacial (125-110 kyr) was subject to even greater climate variability (GRIP project members 1993).

If they also occurred during the Pliocene, these smaller timescale fluctuations may have occurred and affected hominid distribution, ecology and extinction patterns, as well as many other aspects of biogeography.

It is necessary to bear in mind all of this 'finer scale' variability when considering the nature of African environments over the last few million years. Whilst long cores do not have the time resolution necessary to pick up any early Quaternary or Pliocene equivalents of interstadials or 'El Nino' influenced climate events in Africa, this does not mean that such fluctuations were absent. However, it is thought that the brief, warm interstadials that are observed from the last glacial were the product of a world with a greatly enlarged northern ice extent, (GRIP project members 1994) such as only occurred during the last 2.4 myr or so. Before then, comparable events may have been less common.

Current knowledge of the global Pliocene climate system by the scientific community remains very incomplete, because of the difficulties of gaining a well-dated coverage of environmental data for this relatively remote time period; at present all that one can reasonably say is that during the Pliocene there were quite large aridity oscillations in northern Africa that quite possibly extended across central and eastern Africa and the south-central Asian belt. By analogy with the general pattern prevailing during much of the late Quaternary, the arid phases might have been linked to the colder parts of global climate fluctuations, with falls of several degrees celsius in mean annual temperature between the peak and the trough of a single climate fluctuation. Evidence of changes in North Atlantic current intensity accompanying these fluctuations suggests that somewhat similar mechanisms were operating between Quaternary and Pliocene climate oscillations.

Visit the colourful NASA-GISS site on global Pliocene climate reconstruction. Note that they use a somewhat different age to define the end of the Pliocene; 1.8 myr ago instead of the 2.4 myr ago used here. This is due to different definitions used by geologists, not a disagreement over the dating of the sediments.

The general background of climates and vegetation during the Cenozoic (last 64 Myr).

Discussion of past and future climates by T.J. Crowley.


van Andel T.H. & Tzedakis P.C., in press. Palaeolithic landscapes of Europe and environs, 150,000-25,000 years ago: an overview. Quaternary Science Reviews.

Foley R.A. (1994). Speciation, extinction and climatic change in hominid evolution. Journal of Human Evolution v.26 p.275-289.

GRIP project members. Climate intability during the last interglacial period recorded in the GRIP ice core. Nature v.364 p.218-220. (1993).

Bonnefille R. (1996). A reassessment of the Plio-Pleistocene pollen record of east Africa. p.299-310. In; Vrba E., Denton G.H., Partridge T.C. & Burckle L.H. Palaeoclimate and Evolution, with emphasis on human origins. Yale University Press, New Haven.

Gasse F. & Van Campo E. (1994). Abrupt post-glacial climate events in west Asia and North Africa monsoon domains. Earth and Planetary Sciences Letters 126, p.435-456.

Hamilton A.C. & Taylor D. (1991). History of climate and forests in tropical Africa during the last 8 Million years. Climate Change v.19 p.65-78.

Hill A. (1996). Faunal & environmental change in the Neogene of east Africa: evidence from the Tugen Hills sequence, Baringo District, Kenya. p.178-193. In; Vrba E., Denton G.H., Partridge T.C. & Burckle L.H. Palaeoclimate and Evolution, with emphasis on human origins. Yale University Press, New Haven.

Kennett J.P. (1996). A review of polar climatic evolution during the Neogene, based on the marine sediment record. p.49-64. In; Vrba E., Denton G.H., Partridge T.C. & Burckle L.H. Palaeoclimate and Evolution, with emphasis on human origins. Yale University Press, New Haven.

Kingston J.D., Marino B. & Hill A. (1994). Isotopic evidence for Neogene hominid palaeoenvironments in the Kenya Rift Valley. Science v.264 p.955-959.

Maslin M.A., Haug G.H., Sarnthein M. & R. Tiedemann (1996). The progressive intensification of northern hemisphere glaciation as seen from the North Pacific. Geol. Rundush v.85 p.452-465.

deMenocal P.B. & Bloemendal J. (1996). Plio-Pleistocene climatic variability in subtropical Africa and the palaeoenvironment of hominid evolution: a combined data-model approach. p.262-287. In; Vrba E., Denton G.H., Partridge T.C. & Burckle L.H. Palaeoclimate and Evolution, with emphasis on human origins. Yale University Press, New Haven.

PRISM project members (1994). Joint investigations of middle Pliocene climate, 1 PRISM palaeoenvironmental reconstructions. Global & Planetary Change v.9 p.169-195.

PRISM project members (1996). Middle Pliocene palaeoenvironments of the Northern Hemisphere. p.198-212. In; Vrba E., Denton G.H., Partridge T.C. & Burckle L.H. Palaeoclimate and Evolution, with emphasis on human origins. Yale University Press, New Haven.

QEN members (1995). Atlas and review of palaeovegetation since the Last Glacial Maximum. Web address; http://www.soton.ac.uk/~tjms/adams1.html

Ruddiman W.F., Shackleton N.J. & McIntyre A. (1986). North Atlantic sea-surface temperatures for the last 1.1 million years. In; North Atlantic palaeoceanography v.21 p.155-173. (ed. Summerhayes C.P. & Shackleton N.J.). Geological Society of London, London.

Scott L. (1996). Pollen evidence for vegetational and climatic change in southern Africa during the Neogene and Quaternary. p.64-193. In; Vrba E., Denton G.H., Partridge T.C. & Burckle L.H. Palaeoclimate and Evolution, with emphasis on human origins. Yale University Press, New Haven.

Vrba E. (1996). On the connection between palaeoclimate and evolution. p.24-42. In; Vrba E., Denton G.H., Partridge T.C. & Burckle L.H. Palaeoclimate and Evolution, with emphasis on human origins. Yale University Press, New Haven.

"Who the heck is this Jonathan Adams, anyway?"

Anything missing? Anything incorrect? Don't suffer in silence - let me know and I'll change it. (jonathan@elvis.esd.ornl.gov)