78 0169-5347/99/$ – see front matter © 1999 Elsevier Science. All rights reserved. TREE vol. 14, no. 2 February 1999 Pines and eucalypts: stress-tolerators or strategic switchers? Eucalypt Ecology: Individuals to Ecosystems edited by Jann Williams and John Woinarski Cambridge University Press, 1997. £95.00 hbk (xii + 430 pages) ISBN 0 521 49740 X Ecology and Biogeography of Pinus edited by David M. Richardson Cambridge University Press, 1998. £95.00 hbk (xvii + 527 pages) ISBN 0 521 55176 5 C omparative ecology is a powerful tool for investigating adaptation when se- ries of closely related taxa have radiated into diverse habitats, and more than one such series is available for comparison. These compendia of the vast and scattered literature on the ecology of Pinus and Euca- lyptus can be used to illustrate this ap- proach. Both genera are speciose, Euca- lyptus (c. 700 species) much more so than Pinus (111 species), comprised of mostly (Eucalyptus) or entirely (Pinus) evergreen woody plants, and largely characterized by their dominance in habitats where short- age of water and/or nutrients (along with a high frequency of fire) are major con- straints to plant productivity. Conveni- ently, the native ranges of the two genera do not overlap, although exotic plantations of Pinus are now widespread within the native range of Eucalyptus and vice versa. What characteristics of Pinus and Eucalyp- tus have allowed each to cope with the resource-poor environments that they oc- cupy in their respective ranges? These books allow us to evaluate two competing hypotheses. In the search for generalizations about plant life-history strategies, many ecolo- gists have attempted to construct robust classifications of functional types based on constellations of covarying morpho- logical and physiological characteristics. The best-known of these is undoubtedly Grime’s triangular ordination of plant func- tional types 1 , in which plants of resource- poor environments – stress-tolerators – are identified as a coherent group with a common suite of characteristics. These traits include a low maximum potential rela- tive growth rate (RGR), long-lived leaves, inflexibility in form and physiology in response to changing external resource supply, and a low proportion of annual car- bon assimilation devoted to reproduction. The stress-tolerator concept has been influ- ential in the development of life history research and is very widely cited to the present day. Recently, Grubb 2 has drawn attention to numerous plants that grow in resource- poor environments but do not possess the stress-tolerant traits in whole or part. Rather than lumping all such plants to- gether, Grubb has proposed dividing them into three groups, with strategies iden- tified, respectively, by their low flexibility (corresponding essentially to Grime’s stress- tolerators), their capacity to switch their RGR and other characteristics during de- velopment (they correspond to Grime’s stress-tolerators as seedlings but not as adults) and by their capacity to ‘gear- down’ their respiration rate in response to shortage of resources, whilst remaining flexible in other aspects of their form and physiology. The comparison of Pinus and Eucalyptus represents a useful model system for test- ing these ideas, because many (but not all) of their species might be regarded as stress-tolerators par excellence, and as- pects of their ecology are very well known. The most obvious paradox for the unitary stress-tolerator concept is that the adult trees of many pine and eucalypt species are highly productive in forest plantations in temperate climates when removed from the constraints of low water and nutrient supply 3 . Under these conditions, the stress- tolerator evergreen species have higher productivity than the native deciduous species over the lifetime of a plantation, which is the opposite of the prediction of Grime’s model 1 . The basis for the fast growth of the evergreen trees in plan- tations is their accumulation of a high bio- mass of long-lived foliage, which enables them to develop a high leaf-area index and light-interception capacity relative to de- ciduous species, which have to replace their leaves each spring. Why, then, are broadleaved deciduous trees dominant in cool temperate climates where nutrients and water are conducive to rapid plant growth? The answer, originating from Bond 4 , lies in the slow RGRs of the seedlings of Pinus and those other evergreen conifers that attain rapid growth as adults. Seed- lings of Eucalyptus species of dry en- vironments are also known to grow slowly, but adult growth rates are poorly known. Therefore, for some Pinus species there is clear evidence of a transition from a rela- tively slow-growing seedling stage to much faster growth as adults, whereas for Euca- lyptus species of drier sites the evidence is equivocal. The conventional explanation for the slow seedling RGRs in switchers is that they possess low assimilation rates per unit leaf-area and a low leaf-area per mass. These characteristics reflect the costs of defending long-lived leaves in resource- poor environments. The reviews in these books illustrate mechanisms for the ‘switching’ strategy not considered pre- viously. For Pinus, the greatest differential between seedling and adult growth rates probably occurs among species that pos- sess a ‘grass stage’ of seedling devel- opment, so-named because the shoot of young seedlings resembles a tuft of grass as a result of restriction of internode extension. This characteristic occurs in a small number of pine species of well drained, nutrient-poor and fire-prone environments, such as Pinus palustris of coastal sandy soils of southeastern USA, and P. merkusii, which grows in similarly unproductive sites in southeast Asia. After five to 20 years in the grass stage, internode ex- pansion rates suddenly increase and the seedling gains in height very rapidly, and P. merkusii, in particular, is renowned for its fast growth as an adult tree. Despite the restriction in above-ground growth, assimilates are allocated to root growth and to storage of carbohydrates in the stem during the grass stage, which buffers the seedling against the droughts and fires common to these unproductive sites 5 . The grass stage of some Pinus species has evolved in response to strong selection for pre-emption of below-ground resources in the short-term to facilitate rapid growth at a later stage, and provides an extreme example of the capacity of some plants of resource-poor environments to switch allocation of resources between different functions during ontogeny. Height growth of the seedlings of some Eucalyptus species of dry or fire-prone envi- ronments can also be constrained in the short-term whilst assimilates are allocated to root growth and development of a ligno- tuber (a basal swelling of the stem). The lignotuber acts as a storage organ and provides a reservoir of dormant buds that are released after destruction of the crown during fires, leading to rapid recovery of canopy structure, as discussed by Bell and Williams in Eucalypt Ecology. Most euca- lypts develop a lignotuber at the seedling stage, and it is highly likely that lignotuber development carries a cost in terms of above-ground seedling growth. The grass stage is present in only a small number of Pinus species, but all pines and many eucalypts are obligately ectomycorrhizal, and it is well known that a major function of this mutualism is to facilitate improved acquisition of limiting nutrients. In the Pinus volume, Read em- phasizes that carbon allocation by the plant to ectomycorrhizas is more appropriately BOOK REVIEWS