Changing the surface of our planet — results from studies of the global ecosystem



During the last few years, it has become almost commonplace that the surface of the Earth is changing rapidly due to human action, and that further changes are likely in coming decades. These changes, collectively referred to as ‘Global Change’, include both large-scale modifications of land use, and unintended effects of developments. They include socio-economic changes affecting land use decision-making, the rapidly increasing CO2 concentration of the atmosphere, large-scale deposition of nitrogen from atmospheric pollution, and the possible alteration of temperature and rainfall patterns. Land ecosystems are known to be sensitive to all of these changes, but the question has been, for more than a decade: are the expected environmental changes large enough to raise concern about the future of the biosphere as we know it today?

A range of research tools are available, or under development, that can be used to shed light upon this important issue. They range from experimental studies of ecosystem processes under modified environmental conditions, through simulation models of varying levels of complexity and spatial comprehensiveness, to systematic studies of the changing behaviour of land use managers and political decision-makers. Early on in this endeavour, it was recognized that international co-operation is crucial to addressing these complex issues. The International Geosphere-Biosphere Programme (IGBP) and the International Human Dimensions Programme (IHDP) each exemplify this co-operative approach, building upon the voluntary contributions of hundreds of scientists to a common goal. The Joint Open Science Conference of the IGBP Core Project ‘Global Change and Terrestrial Ecosystems’ (GCTE) and the IGBP/IHDP Core Project ‘Land Use and Cover Change’ (LUCC), held in Barcelona, Spain, on 14–18 March 1998, was the first major event arising from these two programmes to bring the ecological and the social science research communities together. Over 800 scientists met and the participants demonstrated an impressive range of both approaches and results.

Among the key findings of the conference was the general recognition that current CO2 and climate change scenarios for the next century indeed represent major changes in environmental boundary conditions for the Earth's vegetation, but that these cannot presently be identified locally because the regional pattern of a modified climate remains unpredictable. Evidence was also presented that broad-scale direct modifications of the land surface affect ecosystems at all scales: from the problem of maintaining biological diversity at the local and landscape scale up to the possible feed-backs that altered vegetation structure may have on atmospheric circulation. Many examples also demonstrated that powerful new tools exist to observe‘Earth's Changing Land’ (the title of the conference) directly: either from space-borne sensors that now provide more than a decade of valuable observations and much additional potential for the future, or from intelligent aggregations of available statistical data, such as derived from agricultural and forest inventories. Yet the conference also showed that we are still far from having the ability of effectively managing the planet towards the goal of sustainable use of its biological resources. We are still scratching the surface of a multifactorial problem, and large additional efforts are needed to provide answers to the questions that are now being asked by policy makers, such as those negotiating the details of the Kyoto Protocol.

The importance of dealing satisfactorily with ecological feed-backs at the continental or global scale, both in constructing climate scenarios and in understanding the implications for terrestrial ecosystems, have previously featured in contributions carried by this journal (e.g. Claussen & Gayler, 1997; Solomon & Kirilenko, 1997; Woodward & Beerling, 1997; Lavorel et al., 1998; Cowling, 1999). Thus, while the results from the Barcelona conference will appear in several different journals, those selected for Global Ecology and Biogeography all originate from the need to address issues relating to land-use change, climate change, and interactions of these forcing factors with terrestrial ecosystem, directly at the macro-scale. All of these GCTE contributions have been subject to the journal's normal peer-review process, and in total, 16 peer-reviewed papers will appear in this and forthcoming issues of Volumes 8 and 9 of the journal.

Clearly, ‘ecological’ research at this scale precludes experimental manipulation in the classical sense of laboratory or field studies, and the development of computer models is therefore as imperative as is the analysis of large data sets from different sources. Satellite-borne sensors have long been seen as critical for this work, but the papers by De Fries & Townshend, Frolking et al. (this issue, pp. 367–379 and 407–416, respectively) and Roderick et al. (issue 6, pp 501–508) leave no longer any doubts that we have finally passed the stage of feasibility studies, and that ecological issues can be studied across entire continents or even the globe. Additionally, data are being gathered from archives that allow spatially comprehensive assessments of land cover backwards in time, as is illustrated by Ramankutty & Foley (this issue, pp. 381–396) and in both papers by Houghton and co-authors (to be published in Volume 9, 2000). Indeed, just this basic capacity to measure the spatial extent and intensity of land cover change over large areas and time frames of a century or more, has been one of the most critical limiting factors for understanding the biosphere's role in the carbon cycle. The Amazon basin is one area of particularly interesting global change processes, both due to rapid change in land use, and due to the still poorly known carbon budget of the area. Two papers, by Kleidon & Heimann, (this issue, pp. 397–405) and by Tian et al. (to be published in Volume 9, 2000) look at these processes from quite different perspectives: the former shows that the role of this particular ecosystem in the Earth System as a whole only can be modelled if we take account of rooting depth in a better way than before. The latter makes the convincing case that interannual variability of carbon fluxes through this ecosystem is larger than previously assumed and that it can be modelled effectively using a validated ecosystem model.

An ecosystem with high responsiveness to climate change is the treeline ecotone—investigated at the global scale by Jobbagy and Jackson (to be published in Volume 9, 2000). Their analysis of temperature effects provides a basis both for interpreting past fluctuations and for predicting future responses of forest limits to climate change. For some time, we have known that regional changes in ecosystem characteristics may have global climate implications and that these can feed back on the ecosystem. Such a case is made by Levis et al. (issue 6, pp. 489–500) who show the sensitivity of global climate to arctic land cover conditions. Inevitably, a question arises about the uncertainty of these broad-scale assessments, whether these come from data assimilations or models, or both. Knorr (to be published in Volume 9, 2000) convinces us that this uncertainty, despite being large, can now be assessed quantitatively for ecosystem models. Phillips et al. (to be published in Volume 9, 2000) make a similar case for the carbon budgets from forests, which are becoming increasingly relevant for policy making.

Several groups are currently working towards simulating vegetation dynamics at the global scale, and some of these aim at a high level of integration of the related feedback processes with atmosphere and the ocean. These studies are made at widely differing levels of complexity, e.g. using conceptual (Füssler & Gassmann, to be published in Volume 9, 2000) or complex (Potter & Klooster, issue 6, pp. 473–488) biosphere models. Betts et al. (to be published in Volume 9, 2000) use a vegetation dynamics model as a component of an Earth System model, showing that physiological responses in vegetation actually may enhance CO2-related warming, but that this might well be offset by structural changes in vegetation. A model of the new class of ‘Intermediate Complexity Models’ has been used by Brovkin et al. (issue 6, pp. 509–517) to make a qualitative assessment of the role of historical land cover change for climate in the 20th century.

Efforts to provide a unified synthesis of global change-related research on the dynamics of the land surface are underway (e.g. Watson et al., 1996; Walker et al., 1999), but as the papers in this GCTE series make clear, multiple approaches and further refinements of the models developed so far, will continue to be necessary in order to provide the sort of scientific inputs needed by policy-makers. These GCTE papers show just how much these techniques have advanced over the past few years. It is now recognized that numerical models not only provide valuable insights, but also that they may address different questions using widely differing levels of model complexity. Uncertainty analysis is also increasingly being used to test systematically whether model results can be used for extrapolation into the next century or not. Land cover change is also increasingly being observed with some confidence at the global scale, as is shown by several of the papers in this series. This adds plausibility to those modelling efforts that aim to identify the sensitivity of the climate system to such changes.


As editors of the GCTE papers we wish to thank all those who have acted as referees for this series of contributions.