Carbon dioxide and terrestrial ecosystems

BOOK

REVIEWS

Continuous change is the one constant factor The Ecology and Biogeography of Nothofagus Forests edited by Thomas 7: Veblen, Robert S. Hill and Jennifer Read Yale University Press, 1996. f50.00 hbk (viii + 403 pages) ISBN 0 300 06423 3 iewing our globe from the south, we V see the vast area of Antarctic ice, surrounded by cold southern seas, and the southern ends of South America, New Zealand, Australia and Africa, all widely sep arated by ocean. This book is about the tree genus Nothofagus,one of the key genera to understand the evolution and migration of the southern biota of this part of the Southern Hemisphere. The book resulted from a series of meetings of scientists interested in the ecology and biogeography of the South Temperate Zone. Its goal is to overcome the isolating effects of distance, language, and disciplinary speciality, by synthesizing the widely scattered current knowledge. (A related book1 resulted from the 1993 Hobart Conference - Southern Temperate Ecosystems: Origin and Diversification.) Useful concepts are discussed around the unifying theme of change in Nothofagus forests, from highland New Guinea at the equator to the southeastern tip of South America, at different temporal and spatial scales. Twenty authors and co-authors of the 12 chapters succeeded in integrating perspectives from historical biogeography and ecological dynamics to develop a basis for predicting future vegetation change in relation to anthropogenic climate changes. Infrageneric classification based on rationalization of the taxonomic division with revised pollen groupings, tested by cladistic analysis, provided a firmer basis to consider the history of Nothofagus and its 35 current species. The extensive fossil record of Nothofaguspollen is used to discuss compositional and physiognomic changes in the disparate southern landmasses, at temporal scales of millennia to millions of years, and spatial scales of tens of square kilometres. Absence of this abundant and distinctive fossil pollen from Africa and India is a case of ‘negative evidence’. Land-based dispersal was long considered the only migration option for the extremely poor dispersal of the dry Nothofagusfruit. Recent pollen evidence suggests that rare long-distance gene flow (by fruit dispersal, or floating live trees or parts of trees) has occurred over the Tasman Sea from Australia to New Zealand during the Cenozoic, when the introduction 526

0 1996,Elsevier Science Ltd

of new species required transoceanic dispersal. Changes in climate, photoperiod, carbon dioxlde levels, and position of the southern landmasses, further affected species distributions and interactions at the community level. Interesting accounts are given of attempts to reconstruct forest structure during the late Cretaceous from a combination of fossil data and the ecophysiological response of living species. An intriguing aspect of the extensive tracts of species-poor, often mono-specific, Nothofagusforests is the occurrence of conspicuous stand-level dieback. A variety of complex interactions among predisposing factors (climatic conditions, site conditions and stand-age structure) and triggering and accelerating exogenous factors (e.g. insects, pathogens, wind and landslides) cause repeated localized to widespread forest destruction, on timescales that are shorter than the maximum longevity of the Nothofagus trees. These stand-level diebacks lead to synchronized regeneration over wide areas. This in turn implies that most forest stands are cohort rather than all-aged in structure, and has implications for stand diversity and for sustainable resource harvesting. Interesting observations of replacement of one species assemblage by another (succession) were made at various temporal and spatial scales. Tree size-class distributions in ecotonal areas between beech forest and other forest types suggest that beech is slowly invading across the boundary, which could be interpreted as unstable ecotones due to climatic change. Along gradients (altitudinal, rainfall, soils, and gradient overlaps), species responded individualistically according to their tolerances and competitive abilities, resulting in a sequence of overlapping ranges, that is, supporting the individualistic model of community structure. Also, the response pattern of Nothofagusassociations to climatic variation resulted in new assemblages due to the individualistic range changes of different taxa. Nothofagusgaps, that is, apparently suitable sites where the species is absent between extensive areas of contagious distributions, each have a unique history of a complex of causal factors. Surprisingly, most Nothofagusspecies are not fire-tolerant although fire has been a significant environmental factor for several million years. The constant switch from present-day ecology to the history and paleoecology of the genus in each of its main areas of current distribution, develops a better perception of the importance of history to understand the present, but also for the need to adapt to continuous environmental change. This is what fascinated me when I had the opportunity to visit many different Nothofagusforests from tropical and temperate Australia, New Zealand, southern Chile and Argentina. This book takes the reader along such a fas-

cinating and stimulating journey, and does so with much clarity. Coert J. Geldenhuys Division of Water, Environment and Forestry Technology, CSIR, PO Box 395, 0001 Pretoria, South Africa

References 1

Enright, N.J.and.Hill,R.S.,eds (1995)Ecology of the Southern Conifers, Melbourne University Press

Ecosystem responses to CO, increase Carbon Dioxide and Terrestrial Ecosystems edited by G. W Koch and H.A. Mooney Academic Press, 1996. $79.95 hbk (xviii + 443 pages)

ISBN 0 12 505295 2

T

he now notorious increasing concentration of atmospheric carbon dioxide has been both a major indicator and a principle driver of global environmental change. The International Panel on Climate Change recently concluded that if CO, emissions are maintained at 1994 levels, then the atmospheric concentration would reach 500 ppm (approaching twice the pre-industrial concentration) by the end of the next century’. The Panel agreed that ‘there is a discernible human influence on global climate”: their projections of amounts and rates of climate change and their possible impacts on ecosystems2 make disturbing reading. However, such large-scale summaries inevitably hide much important detail, and necessarily underplay scientific uncertainties. Six years ago, reviewing the information on ecological effects of increased CO, in this journal, I discussed3 results from the only two studies measuring the response of intact ecosystems to CO,: in Alaskan tundra and in a Chesapeake Bay saltmarsh. Now, as this book details, several further studies have been carried out, and some have been pursued for several years. The studies now include both broadleaved and conifer tree species, grown as seedlings, saplings or even as mature trees (with bags to enrich CO, around individual branches), annual grassland communities, prairie grass stands, alpine vegetation, and several annual crops (cotton, wheat, field bean and rice). Not only have the number of different experiments increased substantially in the past decade, but the approach has changed both technically (gradually moving from open-topped chambers to free-air CO, enrichment, or FACE) and scientifically. Several chapters describe the linked experimental TREE

vol.

II,

no. 12 December

2996

POSTSCRIPT and modelling approach being pursued by the different groups, and others provide stimulating reviews of both the necessity for and the difficulties of mechanistic models in this ecosystem scale work, particular for forests where experimental approaches are so limited. Several authors highlight the gaps and problems in available forest response information (measurements only on seedlings or juvenile material, with ‘step changes’ in CO, concentration, and recording shortterm responses only, without interactions through nutrient cycling). The FACEforest systems being set up now4 in the USAwill help answer some but not all of these questions. As emphasized by Amthor and Loomis (Chapter 18) we have yet to produce robust models for the considerably simpler crop systems, thus designing reliable models of forest-ecosystem response to change remains a major challenge, requiring components on species composition changes, demography, and productivity and nutrient cycling. It is the linkage of these components that determines the ecosystem level response. Furthermore, this in turn determines the net ecosystem productivity, which rep resents the net impact of ecosystems on the carbon cycle as Amthor and Koch point out (Chapter 21) information that is critical in global projections of the carbon cycle and climate change (e.g. Refs 5,6). Some common themes emerge from the

wide range of case studies described in this

book. One is that while increased CO, often increases net photosynthesis (unless there is down-regulation), it may not necessarily stimulate aboveground productivity, resulting instead in greater carbon flow to the belowground part of the system. This has many far-reaching implications, in particular to decomposition and nutrient cycling. The second is that there are substantial changes in plant material composition, at least in these ‘step-change’experiments. However, predicting the influence of these on the whole system is clearly fraught with difficulty because, as Lindroth (Chapter 7) indicates, it is not just the plant C:Nratio that matters to herbivores, and because plant detritus may not exhibit the changes in nutrient composition that green leaves do, as O’Neill and Norby (Chapter 6) point out. There are several surprises, too: for example, in the study by Owensby et al. in the C4 dominated tall grass prairie, the C3 species declined over five years of CO,-enrichment (contrary to the usual expectation of increasing CO, benefitting C3 species more than C4 species) because the increased CO,

improved the water status of the C4 species. The editors make a brave summary of the breadth of information contained in this book, information that spans the scales from leaf level to global responses. They suggest that there may be a possible correlation between absolute ecosystem productivity and responsiveness to increased

Two truths about discountingand their environmentalconsequences iscounting is the process by which we D translate the value of expected future costs and benefits into present-day values. It is a fact of life that, taken as a whole, society discounts future values. Economists note that as borrowers people are willing to pay an exponential interest penalty to acquire resources now; conversely, savers expect to be rewarded for their forbearance with an exponential interest premium on bank deposits and investments. Using this exponential logic, it can appear to make economic sense to harvest rain forests rapidly or to fish the great whales to extinction if the discount rate is higher than the natural asset’s growth rate’. Similarly, if the high costs of decommissioning a nuclear power station are incurred far into the future, they have negligible impact on our present-day decisions. Ethologists, however, take a hyperbolic approach to discounting. The difference is of considerable importance over long TREE

uol.

II,

no.

12 December

I996

timeframes, with important implications for sustainability. Under the standard economic perspective*J, the present value, VP,of a future value, V,,received after a delay of time, t, is: VP= v,/(l t r)’

where ris the (exponential) discount rate. The rationale for this exponential relationship is the cost of borrowing capital; bank deposits and liabilities grow exponentially. However, ethologists, although they agree with economists on the principle of discounting, disagree with them as to the nature of discounting and maintain that: VP= V,/(l + r-J)

(2)

where r, is the (hyperbolic) discount rate. This conclusion is based on studie&s of human choice, where different sized rewards are linked to differing time delays, and on studies of other animals, such as 0 1996, Elsevier

Science

Ltd

CO,, though there is no clear pattern yet, and no data from forest ecosystems. They also stress the variability in productivity re sponse observed from year to year in the various longer-term studies, variability due to the interaction with the climate, and particularly with water availability. The book is a comprehensive statement of progress and prospects in assessing ecosystem response to rising CO,: there are still many gaps in our knowledge and the major tropical ecosystems (both grassland and forest) have still to be examined. James I.L. Morison Dept of Biological and Chemical Sciences, University of Essex, Colchester, UK CO4 3SQ

References Houghton, J.T. et al., eds (1996) Climate Change 1995: the Science of Climate Change,

Cambridge University Press Watson, R.T., Zinyowera, M.C. and Moss, R.H., eds (1996) Climate Change 199.5: lmpacfs, Adaptations

and Mitigation

Scientific-Technical

of Climate Change:

Analyses,

University Press Morison, J.I.L. (1990) Trends

Cambridge Ecol. Evol. 5,

69-70

Ellsworth, D.S. et al. (1995) Oecologia 104, 139-146 Lloyd, J. and Farquhar, G.D. (1996) Funcf. Ecol. 10,4-32 Wang, Y.P. and Polglase, P.J. (1995) Plant Cell En&on. l&1226-1244

rats and pigeons, which have also been shown to discount hyperbolically4J. When integrated over time from zero to infinity the exponential function produces a finite value, whereas the hyperbolic function produces infinity. In practice, policy makers are rarely concerned with the extreme long-term consequences of resource-use decisions, but do sometimes take the more immediate long term (50 to 200 years) into account. Under the exponential view, at normal rates of discount (r= 0.04-0.10) costs and benefits occurring more than 50 to 100 years into the future’carry negligible weight when converted to present values. Under hyperbolic discounting, the distant future has more import. The distinction is critical when trying to value projects such as forestry replantings, dam development or species and habitat preservation - all projects whose costs and benefits extend far into the future. To illustrate this we show data collected by Cropper et al.6, who asked members of the American public to choose between alternative government spending programmes. The first programme alternative would save 100 (anonymous) lives today, while the second would save PII: SO169-5347(96)20083-7

52 7