Aquatic ecology

A lake's life is not its own

Article metrics

Imagine a landscape suddenly released from the frozen grip of a kilometre-thick glacier. Bare rock crumbles to pockets of dust that harbour sprouting seeds, and soon patches of plant species adapted to this environment appear. As the environment gradually changes, the ensuing development of the flora — a process known as succession —proceeds roughly in step with predictions from theories of plant community dynamics. As ecology goes, we have a good idea of what happens in these circumstances.

But what of the numerous lakes left in depressions after the glacier has retreated? Surely science can also predict their development? Well. . . no. Unlike terrestrial ecology, which is founded in part on numerous studies of plant succession, aquatic ecologists have only the sketchiest of ideas about succession in lakes; there is no workable theory of how their chemistry and biology evolve. On page 161 of this issue1, Engstrom et al. present a powerful data set that provides insights into what happens in the first few years of a lake's life — and also the following thousands of years. More importantly, they can give reasons why the lakes they studied evolved as they did.

Why aren't there any general theories of lake development? One of the original ideas from the founder of this field, the late Ed Deevey at Yale University, was that lakes become more productive as they age; that is, the amount of biomass created in the lake increases. Although that is sometimes true, and the notion is often taken as accepted fact, it has long since been evident that it simply fails to hold as a generality2. There are also studies of how lakes develop towards the end of their existence, such as Walker's3 classic analysis from the British Isles of how lakes become land. But the variable patterns of succession from open water to swamps, and from bogs to forest, were predictable only in the coarsest statistical sense.

Perhaps the big problem is that the chance of observing lake evolution is so rare. Yes, there are hundreds of artificial reservoirs that we could track through history, but most of them flush so freely that they never truly develop as lakes, surviving instead as wide, slowly flowing rivers. And the approach of comparing lakes formed at different times is difficult, because most lakes were formed at the end of the last glaciation, at least 10,000 years ago. That makes the gap between recent reservoirs and natural lakes too great, and interpolation between those two points forms an uninteresting line.

Enter Engstrom et al.1, who have taken advantage of a natural experiment — the advance and final retreat of ice at Glacier Bay, Alaska, over the past 12,000 years. The end result is a superb sequence of natural laboratories in the form of lakes ranging from 10 years to 12,000 years in age (see the map on page 162 of the paper). The authors looked first across all the lakes for patterns in diatom species (lake algae) and water chemistry that were related to the age of the lake. They next studied the history of each lake, as recorded in the form of fossil diatoms and the chemistry of sediments buried over time. The patterns they found in the results of these comparative and historical approaches match, which is wonderful as it adds great confidence to the group's conclusions.

Engstrom and colleagues show that their lakes have become more dilute, more acid and more coloured with dissolved organic matter over time (and, by inference, less rather than more productive). The authors' explanations are intricate but don't hide the main picture that the terrestrial surroundings have controlled lake evolution. But this control is not as simple as the depletion of easily weathered minerals in the catchment area causing the runoff waters to become progressively more dilute. Rather, there appears to be a 200-year lag in the initial changes in lake chemistry and biology, a lag which coincides with the development of soils and vegetation in the Glacier Bay area. Such soil development eventually excludes water from running through the deeper, mineral-rich substrates and into the lakes. Even the details of variation between individual lakes in Engstrom and colleagues' study can be explained by the geomorphological control of local hydrology.

These results1 will help to reshape our understanding of lake evolution. The eventual goal is to predict how aquatic ecosystems respond to the interaction of landscape geomorphology, climate change and ecological factors acting through terrestrial vegetation or within the lake itself. This is the real mystery story, with the potential factors and mechanisms being so intertwined that it seems hard to isolate a problem let alone an answer.

Yet there is hope. We now have Engstrom and colleagues' evidence that geomorphological controls can dominate lake development in some cases. And Birks et al.4 have shown that sudden variations in aquatic ecosystems resulted from rapid climate change (during the last glacial-to-interglacial transition, about 14,000 years ago), while longer-term variation since then resulted from the interplay of climate and changes in the catchment area and in the lakes themselves.

In a further twist5, it seems that a lake's response to a strong shift in climate may be overridden by geomorphological controls. For example, arctic Alaska experienced a regional change in climate about 7,000 years ago6. This resulted in dramatic alterations in the carbon chemistry of sediments in a lake on a young, topographically varied landscape, but little or no stratigraphic variation in a neighbouring lake on a much older, smoother landscape (Fig. 1). The two lakes were exposed to identical climates and are similar in nature. Yet these indicators of carbon inputs or lake metabolism varied greatly across the climate transition in only one of the lakes.

Figure 1: Variation of organic-carbon content and δ 13C value in the sediments of two lakes in arctic Alaska5.

Both lakes experienced the same climatic transition about 7,000 years ago. But the carbon chemistry (an indicator of processes happening in the catchment area or the lake itself) responded only in the lake on the young surface. This example shows how landscape characteristics can override the effects on lake development from even rapid and strong climate change.

In revising our ideas about lake development, it is clear that climate, geomorphology and ecology are all major players that can smother each other now and then. We must discover patterns of when and where this has happened in the past, as these will guide our thinking about how aquatic ecosystems will change in the future. This great paper1 pushes us along, and it'll be fun to see how it all works out.


  1. 1

    Engstrom, D. R., Fritz, S. C., Almendinger, J. E. & Juggins, S. Nature 408, 161–166 ( 2000).

  2. 2

    Livingstone, D. A. Am. J. Sci. 255, 364–373 (1957).

  3. 3

    Walker, D. in Studies in the Vegetational History of the British Isles (eds Walker, D. & West, R. G.) 117–139 (Cambridge Univ. Press, 1970).

  4. 4

    Birks, H. H., Batterbee, R. W. & Birks, H. J. J. Paleolimnol. 23, 91– 114 (2000).

  5. 5

    Hu, F. S., Oswald, W. W., Brubaker, L. B. & Kling, G. W. in Geomorphic and Climatic Controls over Holocene Soil Development on the Arctic Foothills of Alaska 204 (Ecol. Soc. Am. 85th Annual Meeting, 6–10 August; Snowbird, Utah, 2000).

  6. 6

    Brubaker, L. B., Anderson, P. M. & Hu, F. S. in Arctic and Alpine Biodiversity (eds Chapin, F. S. & Korner, C.) 111–125 (Springer, Berlin, 1995).

Download references

Author information

Correspondence to George W. Kling.

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.