Chile refuges

Article metrics

Where did warmth-loving species shelter from the cold during ice ages? Genetic analyses of a coniferous tree in South America provide an unexpected answer.

The past two million years have been tough for species that thrive in temperate conditions. This period, the Pleistocene epoch, has seen erratic, often rapid climatic fluctuations that have tested many species to their limits, and some beyond. The persistence of some temperate species into the present day indicates that they must have survived the trying climatic conditions — but where? The identification of the precise locations of glacial survival could prove important in understanding current patterns of species distribution.

This work has traditionally relied on the efforts of geologists and palaeontologists, but genetic analyses of present-day populations are proving a valuable source of further information. This kind of approach has been used in the Northern Hemisphere to study species ranging from lodgepole pines1 to woodrats2. Writing in the Journal of Biogeography3, Premoli and colleagues use genetic analyses to study the coniferous tree Fitzroya cupressoides (Fig. 1), and manage to unravel the glacial history of this species in South America.

Figure 1: Fitzroya cupressoides, a long-lived South American conifer.


Genetic analyses of living Fitzroya populations3 provide clues to the location of refuges for temperate organisms in the Southern Hemisphere during the last ice age.

The climatic shifts of the past two million years have selected in favour of those species that can adjust their patterns of distribution, moving from one place to another and persisting — often in suboptimal conditions — in isolated refugia during cold periods. Reliably dated fossils have provided clues to the locations of these refugia and to subsequent patterns of spread during warm episodes, but fossils do not always supply conclusive evidence. In addition, the resolution of radiocarbon dating at the start of the most recent interglacial period, some 10,000 years ago, is poor. This makes it difficult to determine precise patterns of species movements. Pollen analyses have also been used extensively to identify refugia, but these studies suffer from problems of interpretation. For example, how much pollen needs to be found to assure us of a plant's presence in an area?

The conditions in the southern part of South America provide a good example of these types of problem. The Andes mountain chain separating Argentina and Chile was glaciated during the last glacial maximum, around 22,000 years ago, while the western lowland part of Chile was ice-free. Pollen studies have indicated that these lowlands were occupied by trees throughout the glaciation, so this region probably acted as a refugium for temperate forest organisms. But forest with a similar composition has also developed on the eastern side of the Andes during the present-day interglacial period, so refugia may have existed there, too.

To investigate this question, Premoli et al.3 have taken a genetic approach, based on the argument that patterns of genetic diversity among living populations of a species should provide insight into its history. They studied the alerce, or Patagonian cypress (Fitzroya), which occurs today in lowland Chile, on the western slopes of the Andes, and in scattered populations on the Argentinian (eastern) side of the mountains. Fitzroya is a long-lived conifer that has been recorded to survive for over 3,600 years, a lifespan second only to that of the bristlecone pine. Current regeneration of the tree is poor, suggesting that it would not have been able to spread rapidly after the end of the last glaciation.

Premoli et al.3 studied 24 populations of Fitzroya from both sides of the Andes, looking at 11 enzymatic systems involving 21 putative genetic loci. This provided an adequate means of estimating genetic variability within the species. If the western side of the Andes constituted the only refugial area, then modern-day populations would be expected to have overall genetic similarity. One would also expect lower diversity in populations found in the east, because these would have been established by a limited number of individuals that managed to disperse across the ice-laden divide of the mountain chain. But if there were separate eastern and western refugia, distinct genetic differences would be expected between today's eastern and western populations, because they will have been isolated from each other since at least the start of the last glaciation.

The authors found that the eastern and western populations are genetically distinct, so it seems that the single-refugium hypothesis must be abandoned. On the face of it, it is surprising that warmth-demanding species survived in the interior of the South American continent, which one might expect to be colder than the coasts. Perhaps areas east of the Andes had relatively little snowfall (because of the precipitation shadow effect), limiting ice development and leaving some sites ice-free.

The genetic patterns of the eastern populations are also diverse, suggesting that there were many refugia on the Argentinian side of the Andes. The more southerly of the eastern populations are the most varied, implying that those in the northeast were derived from those in the south as a result of postglacial spread. This is also surprising: one would expect refugia for temperate trees to be found at warmer latitudes further north. Premoli et al. question whether the extensive latitudinal movements seen widely in the Northern Hemisphere are to be expected in the Southern Hemisphere, where very different climatic patterns and less extreme variations were experienced in the Pleistocene. It seems that the biotic impact of the glacial cycles, in terms of both the latitudinal extent of vegetation shifts and the frequency of extinctions, was greater in the Northern Hemisphere. For Northern Hemisphere biogeographers, the south is clearly a foreign country and plants do things differently there.


  1. 1

    Cwynar, L. C. & MacDonald, G. M. Am. Nat. 129, 463–469 (1987).

  2. 2

    Smith, F. A., Matocq, M. D., Melendez, K. F., Ditto, A. M. & Kelly, P. A. J. Biogeogr. 27, 483–497 (2000).

  3. 3

    Premoli, A. C., Kitzberger, T. & Veblen, T. T. J. Biogeogr. 27, 251– 260 (2000).

Download references

Author information

Correspondence to Peter D. Moore.

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.