The disturbance of forest ecosystems: the ecological basis ... · PDF fileThe disturbance of forest ecosystems: ... ecological basis for conservative management ... The extensive literature

Forest Ecology and Management, 63 (1994) 247-300 247 Elsevier Science B.V.

The disturbance of forest ecosystems: the ecological basis for conservative management

Peter M. Attiwill School of Botany, The University of Melbourne, Parkville, Vic. 3052, Australia

(Accepted 30 August 1993 )

Abstract

The extensive literature on natural disturbance in forests is reviewed in terms of the hypotheses: ( 1 ) that disturbance is a major force moulding the development, structure and function of forests; and (b) that management of forests for all their benefits can be controlled so that the effects can be contained within those which result from natural disturbance. The causal factors of natural disturbance are both endogenous and exogenous; there are major difficulties in the formal characterization of disturbance and of recovery after disturbance. As to the latter, the acceptance of classical generali- zations of the nature of succession has led to particular difficulties in the assessment and interpretation of recovery.

Tree fall, which creates gaps, is fundamental to the development of many forests, and has been most intensively studied in tropical forests of Central America and the Amazon and in temperate forests of North America. Tree fall is part of autogenic change; mechanisms of gap-filling and subsequent growth and species composition vary widely with forest type and geography.

Disturbance by wind is particularly difficult to characterize. Wind varies along a continuum; the blow-down of an individual tree may be mostly due to autogenic processes of ageing and decay, whereas catastrophic hurricanes and cyclones may be defined as wholly exogenous. Nevertheless, the resilience in terms of species diversity of tropical forests following catastrophic disturbance by hurricane is remarkable. A number of studies support the view that the tropical forest in hurricane-prone areas is not a stable steady-state ecosystem but rather that heterogeneity is maintained by catastrophe. The ability to regenerate by suckers and the coincidence of regenerative space and gregarious flowering are important components of the response of rainforest following disturbance.

For much of the world, 'fire is the dominant fact of forest history'. As examples, fire and its effects are reviewed for the northern boreal forests, oak-pine forests and north-western sub-alpine forests of North America. The effect of fire on species composition varies with intensity and frequency. That, together with the popular view of fire as unnatural and therefore unacceptable, places great demands on management of forests for all of their benefits, including national parks and reserves. These difficulties also affect management of other ecosystems, such as Mediterranean-type shrublands and heathlands where species diversity, productivity and cycles of regeneration and degradation are gov- erned by fire as a natural disturbance.

Shifting agriculture is a traditional form of agriculture used by at least 240 million people in the humid tropics. Shifting agriculture, together with wind, lightning and fire, is an exogenous disturbance which has little effect on soil fertility and on structure and composition of the rainforest which re-establishes after abandonment. As the intensity of disturbance of rainforest increases, resilience of the forest decreases and the current problems of extensive clearing for improved pasture and of uncontrolled logging are resulting in degraded ecosystems.

Regeneration follows the often extensive death of trees caused by outbreaks of insects in many coniferous forests of northern America. This disturbance by herbivory halts increasing stagnation (as

© 1994 Elsevier Science B.V. All rights reserved 0378-1127 /94 /$07 .00 SSDI 0 3 7 8 - 1 1 27 ( 9 3 ) 0 3 2 8 9 - U

248 P.M. Attiwill / Forest Ecology and Management 63 (1994) 247-300

measured by decreasing rates of ecosystem production and nutrient cycling) and reinitiates succession. Other disturbances to forests occur through damage from ice-storms, snow avalanches, erosional and earthquake landslides, and volcanic activity; the development of N o t h o f a g u s forests in Chile and New Zealand is determined by such catastrophic mass movements.

An extensive literature supports the hypothesis that natural disturbance is fundamental to the development of structure and function of forest ecosystems. It follows that our management of natural forest should be based on an ecological understanding of the processes of natural disturbance. Whether or not we want to do this, and the extent to which we want to derive all of the benefits from the forest, including timber, depends on social attitudes. Whereas humanism may treat conservation as the wise husbanding of forests in the interests of social traditions and harmony, animism may give nature unalienable rights. The conclusion from this review is that the ecological framework of natural disturbance and the knowledge of its component processes and effects provides the basis on which we can manage our forests as a renewable resource which can be utilized so that the forests 'retain their diversity and richness for mankind's continuing benefit'. Nowhere is this management more desperately needed than for the protection of the world's tropical forests, its peoples and their cultures.

Introduction

The literature on disturbance and on the associated concepts of stability, susceptibility and the steady state of ecological systems is very large. Much of this literature was synthesized in the multi-authored book The Ecology of Nat- uralDisturbance andPatch Dynamics (Pickett and White, 1985a). Since 1985, there has been an increase in the number of studies of disturbance and there has been a particular concentration of work on forested ecosystems, much of which has been summarized in detail by Oliver and Larson ( 1990 ).

Pickett and White (1985b) state in their conclusion that: 'There can be no doubt that disturbance is an important and widespread phenomenon in nature... Disturbance is common to many different systems. It functions or has functioned at all temporal and spatial scales and levels of organization of ecological and evolutionary interest'.

The hypothesis proposed that harvesting a forest for all of its benefits, including its t imber products, can be controlled so that it creates a disturbance, the effects of which do not differ from those of natural disturbance. In this first paper, disturbance of forests is reviewed with an emphasis on the recent (post- 1985 ) literature. The second paper (Attiwill, 1994) reviews natural disturbances in eucalypt forests in Australia with the aim of determining whether or not the effects of forest harvesting can be contained within the framework provided by the effects of natural disturbance.

Our judgement about the effects of human disturbance (or management) will always have a level of uncertainty and the degree of uncertainty which can be tolerated increases with the social acceptability of management and its objectives (or politics; Walters and Holling, 1990). Thus, there is room for great disparity between the degree of ecological uncertainty and social objec-

P.M. Attiwill /Forest Ecology and Management 63 (1994) 24 7-300 249

tives so that decreasing the degree of scientific uncertainty may not of itself influence social acceptability. The rapidly changing social acceptability of forest harvesting in Australia is based on a number of considerations. Most of them are political in nature (e.g. what is the correct balance between conservation values and timber values?; do we want to export wood-chips?; what are the employment opportunities in forestry?; is there inherent value in old- growth forest?; are industries running on subsidies from the state forest own- ers?; should we aim to get all of our timber only from plantations?) and science to a large degree is secondary. Nevertheless, science is brought to the forefront where the various proponents see that they might use it to validate their primary social objectives.

This paper does not embrace the proposition that all human disturbance produces effects which differ from those of natural disturbance. Rather, I ar- gue that there is a scale of human disturbance as there is a scale of natural disturbance. At points on these scales, we could judge that the effects of human disturbance on ecosystem structure and function are not significantly different from those of natural disturbance. The hypothesis is illustrated by the effects of human disturbance on rainforest in eastern Amazonia (Fig. 1, from Buschbacher et al., 1988; and see also Jordan, 1985; Uhl et al., 1988a). Traditional slash-and-burn agriculture leads to secondary succession and the recovery of diversity and productivity of the mature forest. The greater the

Mature rain forest " ~ , , ~

., I \

Pasture I{.eco~ er~ :

uloclerate'use /

Seeon(lar} l'ol'e~t ~

Accidental fire \ J

Herbicide and Mechanized hea~v use fertilizer ~x j / pasture, ' ~ ~ : \ De! aded '

(2C(IS~ stein Fig. 1. Degradation and recovery following various intensities of deforestation and modification of forest ecosystems in eastern Amazonia (from Buschbacher et al., 1988, with the permission of Blackwell's, Oxford).

2 5 0 P.M. Attiwill / Forest Ecology and Management 63 (1994) 247-300

disturbance (bulldozing, ploughing, fertilizers, herbicides) the less is the probability that succession after abandonment will effect the recovery of diversity and productivity.

Disturbance, recovery and stability of forested ecosystems

The nature of disturbance

Terrestrial plant communities throughout the world are affected by a variety of major, natural disturbances. Major, natural disturbances listed by White (1979 ) and White and Pickett (1985 ) include fire; hurricanes, windstorms and gap dynamics; ice storms, ice push, cryogenesis and freeze damage; landslides, avalanches and other earth movements, including coastal erosion and dune movement; coastal flooding; lava flows; karst processes; droughts, flash floods, rare rainstorms, fluctuating water levels, alluvial processes and salin- ity changes; biotic disturbances including insect attack, fungal disease, brows- ing and burrowing animals, invasion by plants (weeds); and disturbance caused by man.

Natural disturbance is difficult to characterize because the causal factors of disturbance are both endogenous and exogenous and because they operate over wide ranges of size, frequency, predictability, timing (season of the year) and magnitude (or intensity) of impact (Webb et al., 1972; Grime, 1979; White, 1979; Bormann and Likens, 1979a; Pickett and White, 1985b; Waring and Schlesinger, 1985; Hopkins, 1990 ). Rykiel ( 1985 ) proposed a set of formal definitions of the various terms disturbance, stress and perturbation in an ecological context; most authors, however, have used the terms in a general rather than formally defined sense, an approach which I adopt in this paper. A few examples show the difficulties of formal characterization.

Wind Bormann and Likens (1979a) point out that 'most agents of disturbance

occur as a continuum'; wind ranges from zero to hurricane force. Where wind blows over or breaks trees that have been weakened by growth stresses and decay (autogenic development ), the disturbance is undoubtedly endogenous, 'an integral part of the developmental process of an ecosystem' (Bormann and Likens, 1979a). At some intensity, where wind destroys healthy as well as weakened trees, the disturbance might be considered to be exogenous and catastrophic.

F i r e

The literature on the ecological effects of fire on plant communities throughout the world is enormous and it has been regularly reviewed (e.g. Ahlgren and Ahlgren, 1960; Kozlowski and Ahlgren, 1974; Spurr and Barnes,

P.M. Attiwill / Forest Ecology and Management 63 (1994) 247-300 2 51

1980; Rundel, 1981; Gill et al., 1981; Mooney and Godron, 1983; Attiwill, 1985; Oliver and Larson, 1990). Fires range widely both in intensity and frequency. They are mostly caused by lightning or by humans, and occasionally by spontaneous combustion and volcanic activity. There is a well-established history of fires in pre-European settlement times in both North America and Australia. In Australia, for example, there is a marked and significant increase in charcoal in lake deposits dating from 120 000 years ago (Singh et al., 1981 ). The spread and intensity of fire depends not only on prevailing climate (moisture, temperature and wind) but on endogenous factors, such as the quantity of fuel and its combustibility. 'Fire is not an entirely exogenous factor and may be as much a result of communi ty structure and composition as of environment ' (White, 1979).

Insect attack Plants produce a wide variety of compounds, some of which are more or

less nutritious, and some more or less repellant, to insects (Harborne, 1988; and for general discussion see Waring and Schlesinger, 1985; Aber and Mel- illo, 1991 ). Recent studies indicate that the intensity of insect attack is related both to the availability of nitrogen in the foliage (Ohmart et al., 1985 ) and to the concentration of free amino acids (White, 1984; Cockfield, 1988 ). A number of changes in biochemistry, which affect both nutritional quality and palatability, can result from changes in environment. There is evidence that nitrogenous solutes which act as osmotica accumulate during periods of water stress, and that this accumulation is associated with intensive insect attack (Poljakoff-Mayber et al., 1987 ). Dr. M.A. Adams and I observed total defoliation of Eucalyptus sideroxylon forest in Victoria, Australia, in the middle of the worst drought on record (1982-1983); we observed the trees re- covering by epicormic shoots in the crown later in the drought, in midsum- mer. We measured dawn water potentials of - 6 MPa at that time.

Insect attack varies from normal and generally low levels of herbivory (e.g. in most eucalypt forests) (Attiwill, 1979; Ohmart et al., 1983; Lamb, 1985 ); to moderate to severe herbivory, particularly where trees are placed under stress (e.g. as remnant patches or individuals after forest clearing) (Dreistadt et al., 1990; Landsberg et al., 1990; Adams and Atkinson, 1991 ); to severe, episodic outbreaks epitomized by the spruce budworm Choristoneura fumi- ferana in North America (e.g. MacLean, 1980; Ostaff and MacLean, 1989). To what extent is insect attack a part of the ecosystem (part of autogenesis) and to what extent is insect attack an exogenous catastrophe?

These simple examples illustrate the difficulties of rigorous classification of natural disturbance. White (1979) summarized the position: 'There would seem to be a cont inuum from factors that are relatively endogenous to those that are relatively exogenous in vegetation composition, and, in addition, a cont inuum from chronic, normal environmental factors to acute, cata-

252 P.M. A ttiwill/Forest Ecology and Management 63 (1994) 247-300

strophic ones. Classifications posed for phenomena along these continua... lend an artificiality to vegetation concepts that depend on them'.

Timber harvesting and acid rain are examples of disturbances which are clearly exogenous. Fire started by lightning is presumably endogenous and fire started with malicious intent or which has unintentionally escaped is presumably exogenous. It is less easy to assign a place on the endogenous-exogenous cont inuum of fire, often of catastrophic scale and intensity, as an integral part of human development over 10 000 years in the Americas (Schule, 1990), over tens of thousands of years in Australia (Singh et al., 1981; Singh and Geissler, 1985 ) and over 1.5 million years in Africa (Schule, 1990).

Recovery after disturbance

The recovery of an ecosystem after disturbance can be described in terms of resilience 'the persistence of relationships within a system' and 'a measure of the ability (of the system) to absorb changes of state variables, driving variables, and parameters, and still persist' and stability 'the ability of a system to return to an equilibrium state after a temporary disturbance' (Holling, 1973; Denslow, 1985 ). Webster et al. ( 1975 ) defined resistance as the extent to which the ecosystem is displaced (equivalent to resilience sensu Holling, 1973 ) and resilience as the rate of recovery of the ecosystem (equivalent to stability sensu Holling, 1973 ). These terms and their definitions (Fig. 2 and Table 1, from Swank and Waide, 1980) are essentially conceptual and their quantification in forest ecosystems is difficult for at least two reasons. First,

_.E

-q

igh Resilience

~ a n c e . Low Resilience

IDislurbance ~ / - Time

Fig. 2. Concepts of resistance and resilience following a disturbance (from Swank and Waide, 1980, with the permission of Oregon State University Press). Several responses are shown, each differing in magnitude of deviation from the pre-disturbance level of function (resistance) and in the rate of recovery, or return to pre-disturbance condition (resilience).

P.M. Attiwill / Forest Ecology and Management 63 (1994) 247-300 2 5 3

Table 1 Definitions of resistance and resilience, two components of ecosystem stability (from Swank and Waide, 1980, with the permission of Oregon State University Press)

Component Analogous terms Meaning

Resistance Inertia Rigidity

Resilience Elasticity Restoration time Classical stability

1 Inversely proportional to the amount by which some functional property or measure of the state of the ecosystem deviates from a nominal level in response to a specific disturbance

2 Directly proportional to the extent of disturbance required to produce a certain magnitude of deviation in some ecosystem functional property

1 Directly proportional to the rate at which some functional measure of the state of the ecosystem recovers or returns to a predisturbance or nominal level

2 Inversely proportional to the time required for ecosystem recovery following disturbance

they require definition of a reference, or pre-disturbance condition. Secondly, since forest ecosystems are diverse, complex and long lived, their quantification is without limit.

Westman ( 1986 ) defined inertia as 'ecosystem resistance to change under stress' (equivalent to resistance as defined by Webster et al., 1975 ) and resilience as the 'degree, manner and pace of restoration of initial structure and function in an ecosystem after disturbance' essentially the same concept as that of Webster et al. ( 1975 ). Westman (1986) proposed that resilience has components of elasticity (the rapidity of return to pre-disturbance), ampli- tude (the extent of displacement ), hysteresis (recovery under disturbance vs recovery when not under stress ), and malleability (how similar the recovered state is to the pre-disturbance state). Other workers have used some or all of these terms in various ways (see Table 1 in Westman, 1986), and this again is probably due to the difficulties of rigorous measurement and quantification. In my view, the simple, conceptual framework of resistance and resilience (Webster et al., 1975) provides a sufficiently complex yet comprehen- sible framework to aim for in studies on the recovery of forest ecosystems.

The conceptual model of Webster et al. ( 1975 ) embraced the range of ecosystems from streams and oceans to grasslands and forests. They concluded that resistance 'results from the accumulated structure of the ecosystem' and is associated with large biomass storage, large amounts of recycling and long turnover times while resilience 'reflects dissipative forces inherent in the ecosystem' and is associated with low storage, fast recycling and short turnover times. Neither term, it seems, is definitive. Harwell et al. (1977) reanalysed the Webster et al. ( 1975 ) model and reached quite different conclusions, es- pecially in the ranking of resistance as it is related to biomass storage.

There is, however, surprisingly little empirical work on these themes in plant


communi t ies . Recovery after disturbance is most often assessed by indices of c o m m u n i t y structure and species diversity (e.g. Ewel, 1983; Denslow, 1985; Halpern, 1988; Olsen and Lamb, 1988; Unwin et al. 1988 ) and diversity profiles (Lewis et al., 1988), and there is some, l imited evidence that resistance (sensu Webster et al., 1975) increases with increasing diversity (Mc- Naughton, 1977; Walker, 1982; Frank and McNaughton , 1991 ). At the process level of ni trogen cycling in forests and the potential for losses after disturbance, resistance (sensu Webster et al., 1975 ) is characteristic of less fertile sites and resilience is characteristic of more fertile sites (Vitousek et al., 1982; Polglase et al., 1986; Weston and Attiwill, 1990).

In summary, both disturbance and recovery after disturbance are difficult to classify, conceptually and quanti tat ively, with precision. Fragility is often used descriptively, but it rarely appears in the ecological literature and appears not to have a strict ecological definit ion. It is implicit in all of these terms that there is a succession leading toward a defined reference.

Disturbance, succession and climax

Ecosystem development - the northern hardwoods model The model for ecosystem deve lopment following disturbance proposed by

Bormann and Likens (1979a,b) is based on the rate of accumulat ion of biomass (living and dead) . The model has four phases (Fig. 3 ). After distur-

T c~ ca.)

e-.

>

,,,.d

E ©

er~

[ - -

I I I

Clear Cut

S I

Second- I growth I forest

I I

~ Reorganizanon Phase I

Aggradation Transilion Phase Phase

I

I

I

D

Steady State Phase

Time ,~ Fig. 3. Phases of ecosystem development after clear-felling a second-growth hardwood forest, north eastern United States (from Bormann and Likens, 1979a,b, with the permission of Sprin- ger-Verlag, New York). There is no exogenous disturbance after clear-felling.

P.M. A ttiwill / Forest Ecology and Management 63 (1994) 247-300 2 5 5

bance, there is an immediate decrease in the total amount of biomass, the reorganization phase, where increased rates of respiration and decomposition follow increases in water availability and temperature. The rate of biomass accumulation by the regenerating forest is then maximum during the aggradation phase, declines during the transition phase and eventually reaches an irregula~ shifting mosaic steady-state phase. The model was developed to de- scribe changes following clear-felling of the mixed hardwood forest at the Hubbard Brook Experimental Forest, north-eastern USA. The forest therefore starts as even-aged and remains so to the end of the aggradation phase, at which time it is probably described as old-age (or virgin, climax, pristine or steady-state, depending on one's point of view; Bormann and Likens, 1979a). This even-aged forest is dominated by a few old trees per unit area; as these trees die, gaps are created within which new vegetation develops to become dominant, thereby creating a shifting mosaic of ages and resulting in a decrease in total biomass averaged over the whole. Thus 'an individual plot will never exhibit a prolonged steady-state in which biomass accumulation will approximate zero' (Bormann and Likens, 1979a).

Ecosystem development in mountain ash (Eucalyptus regnans)

The 'northern hardwood model ' (Fig. 3 ) is driven essentially by autogenic processes; gaps are created by trees falling over or being blown over, regeneration in gaps eventually producing an all-aged forest. In many forests, however, regeneration depends wholly on disturbance. Australian eucalypts of wet sclerophyll forest (Beadle and Costin, 1952) or tall open-forest (Specht, 1970), such as E. regnans (mountain ash), are outstanding examples of such a dependency (Ashton, 1981 a,b). These forests regenerate naturally only after major fire, the result being extensive tracts of even-aged forest. The forest at 250 years is dominated by perhaps 20 trees ha-1, and gaps created by death or windblow of these trees are filled only by the relatively low understorey. The even-aged structure of the dominant trees is maintained; multi-aged forest is restricted to no more than two or three age classes, generally in small edge-patches where subsequent, non-lethal ground-fires have intruded. We therefore expect that the relative decrease in biomass, averaged over the whole, as the eucalypt tall open-forest progresses from aggrading to steady-state, will be greater than for the northern hardwood forest (Fig. 3 ).

The pattern of water yield or streamflow from mountain ash forest with age (Fig. 4, from Kuczera, 1987; see also Attiwill, 1991; O'Shaughnessy and Jay- asuriya, 1991 ) implies a high rate of evapotranspiration in the aggrading forest and a decreasing rate in the maturing to old-age forest. A likely explana- tion of this pattern with age is that water use depends on sapwood area. In the young, aggrading forest, all of the wood is sapwood. As the forest ages, the area of sapwood decreases (Dunn and Connor, 1991 ). The decrease in water


b r-

:~ 12

o d w 10

N 8 o _.1 kL

< 6

__1

z o 2

u.l

0

-j

"L X \ // / / ,,% \ / / / , /

a I a I n I n I n I

40 80 120 160 200

FOREST AGE (years) Fig. 4. Relationships between water yield and stand age for Eucalyptus regnans forested catch- ments (from Kuczera, 1987 ). The idealized response curve is for E. regnans in the high rainfall region of the Central Highlands of Victoria, Australia, annual rainfall in the range 1800-2000 ram. (See Attiwill, 1991. )

use of the old-aged forest (at about 160 years, Fig. 4 ) relative to the aggrading forest is very large.

We could accept that the phases of reorganization, aggradation and transition of the northern hardwood model (Fig. 3 ) are universally applicable, but that the steady state phase is entirely forest specific. The northern hardwoods of New Hampshire, USA, are relatively fire-resistant, to the extent that they have been called 'asbestos forests' (Bormann and Likens, 1979a); histori- cally, large-scale disturbances by fire (Russell, 1983) and windthrow (Seis- chab and Orwig, 1991 ) were rare. There is a mixture of dominant species, shade tolerance is high, and regeneration or release of suppressed trees follows the creation of gaps caused by endogenous disturbance. The shifting mosaic steady state therefore involves shifts in both diversity and net productivity in time and space. The frequency of catastrophic exogenous disturbance such as large-scale winds or fire which would initiate a new reorganization phase over extensive areas is relatively low. It follows, that if an exogenous disturbance such as clear-felling (followed by regeneration ) is imposed on such a system at a frequency which differs markedly from the frequency of catastrophic disturbance, the 'asynchronous' age structure of the shifting mosaic steady state will be regulated.

P.M. Attiwill / Forest Ecology and Management 63 (1994) 24 7-300 257

In contrast, Eucalyptus regnans (mountain ash ) in south-eastern Australia is a truly fire climax forest (Ashton, 1981 a,b ); its natural presence is due to a major fire in the past. At its best development, away from ecotones, it grows as pure, even-aged stands. It is shade intolerant, so that mortality in the aggrading phase is high. Trees never remain suppressed, to be liberated by the creation of a gap. Seedlings do not develop in gaps, nor do trees re-grow from coppice shoots or root suckers. The 'steady state' phase (perhaps from about 160 years, Fig. 4) is therefore neither a steady state nor a shifting mosaic; it is a progression from domination of the site by trees to domination by understorey species of shrubs (understorey trees such as Acacia themselves being dependent on fire for their regeneration). Given time, species characteristic of cool temperate rainforest may become established. To a major degree, however, this is speculation, since the frequency of catastrophic fire which regenerates the forest is high enough to preclude the gathering of empirical evidence. It follows that the effect of an exogenous disturbance such as clear- felling (followed by regeneration) will affect the age structure of the mountain ash forest to the extent that the frequency is related to the stochastic frequency of catastrophic fire.

Succession and climax

Successional change in community structure and composition is often ob- vious, but the end-product of succession cannot be predicted with the cer- tainty implicit in the classical view of the stable and self-perpetuating climax that prevailed earlier this century (see discussions and reviews by White, 1979; Bormann and Likens, 1979a; Denslow, 1980; Noble and Slatyer, 1980; Spurr and Barnes, 1980; Oliver, 1981; Oliver and Larson, 1990; Hopkins, 1990). Rather, there is very great variation, both in time and space, in the forces which influence and direct the development of the structure, composition and functioning of plant communities (Whittaker, 1953, 1956; Egler, 1954; Drury and Nisbet, 1973; Horn, 1974; Connell and Slatyer, 1977; White, 1979; Shu- gart, 1984). Disturbance is one of these forces, to the extent that 'disturbance gradients may parallel physical environmental gradients (and so ) the two may be difficult to distinguish' (Harmon et al., 1983). Some disturbances (espe- cially fire) have had such profound evolutionary influence on the survival and regeneration of plant communities that periodic disturbance is essential if diversity is to be maintained, and a delay in the recurrence of disturbance leads to a reduction in diversity (e.g. Loucks, 1970; Whelan and Main, 1979; Walker, 1982 ).

These various themes are encompassed by Shugart and West ( 1981 ): 'For- ests often represent a tranquillity associated with the apparent changelessness of large trees. Ironically, this tranquillity is largely a product of the human perception of time... Forest ecosystems are dynamic entities that may not be


static in either t ime or space. Many of our concepts of ecosystem management are based on the hope that if a forest is left alone it will gradually return to its natural state. 'Succession', 'wilderness', 'virgin forests' and 'climax forests' are all concepts that appeal to the basic notion that forest systems should approach some equilibrium state with time.'

Loucks (1970) proposed the hypothesis that maximum diversity and productivity of forest ecosystems are maintained by random periodic disturbance, a hypothesis which is directly applicable to the mountain ash forest, as well as to other forests which depend on fire for their renewal (Bormann and Likens, 1979a). There is no autogenically derived steady state, and the system is entirely dependent on exogenous and catastrophic disturbance for its recycling. This disturbance, random in time, is therefore fundamental to the stability of the ecosystem (Loucks, 1970). Loucks ended his paper in the strongest way, concluding that the elimination of disturbance by modern humans 'will be the greatest upset of the ecosystem of all time... It is an upset which is moving us unalterably toward decreased diversity and decreased productivity at a t ime when we can least afford it, and least expect it'.

This system, in which stability is achieved through random disturbance, is found in many plant communit ies including heathlands of Mediterranean regions (Specht, 1979, 1981c). Coastal heaths in south-eastern Australia include a large number of species (e.g. Allocasuarina paradoxa, Banksia mar- ginata, Hakea sericea and Leptospermum myrsinoides) which regenerate either (or both) vegetatively from shoots and suckers or (and) from seed held in bradysporous fruit (Specht, 198 la,b). Like the chaparral in Califor- nia (Force, 1981 ), diversity is highest immediately after disturbance by fire. If fire is excluded from heathland, diversity of both plants and animals quickly decreases (Force, 1981; Walker, 1982 ) and dominance is exerted by one or a few original or invading (Burrell, 1981 ) species. The capacity of the heath to respond to a disturbance when it eventually occurs is then greatly diminished (Walker, 1982). Not only would it be practically difficult to regenerate the original set of species but there would be a decrease in diversity, maximum diversity being maintained at the historic frequency of disturbance (Den- slow, 1980).

The autogenically derived steady state for the northern hardwood forest (Fig. 3 ) is predictable for the whole, but less predictable for a space (or plot ) within the whole. Whether or not the mountain ash forest and the heathland reach a steady state, and what this steady state will be for a given space, are for the most part matters for conjecture. Furthermore, I suggest that steady state is an inappropriate term for an ecological system, since it implies both predictability and stability at a level we can neither expect nor assess.

It is apparent then that the process of succession depends both on the behaviour and responses of species (e.g. Drury and Nisbet, 1973; Westman and O'Leary, 1986) and on the regime of disturbance. Noble and Slatyer (1980)


identified three sets of attributes of plants which are vital to their role in a developing plant community: ( 1 ) the method of arrival or persistence of the species at the site during and after a disturbance; (2) the ability to establish and grow to maturity in the developing community; and (3) the time taken for the species to reach critical life stages. However, at the level of the plot, the unit on which we must assess resistance and resilience, Halpern ( 1988 ) concluded that 'variation in the long-term response of communities reflects complex interactions between species life history, disturbance intensity, and chance, suggesting that both deterministic and stochastic factors must be considered in evaluating community stability and response to disturbance'. While the approach through vital attributes and life histories is wholly rational, au- tecological data even for common understorey species is mostly limited, as Halpern (1989) noted for Pseudotsuga forests. The problem in tropical forests is much greater. G6mez-Pompa and V~zquez-Yanes (1981 ) acknowl- edged the 'enormous importance' of understanding life cycles, but concluded that: 'There exists an enormous diversity of tropical ecosystems in the world... Of the milli.ons of species of plants and animals in the tropics, we know with some detail (although superficially) the life cycles of only a few. Until our knowledge increases, it will not be possible to arrive at a general concept of succession in the tropics'.

Natural disturbances in forests

Tree fall and gap dynamics

The topic of tree fall and gap dynamics has been developed principally as the study of autogenic changes in species composition and community structure and function at a given site (the dynamics of tree crowns of existing dominants and their occupation of the space created by the fall ofa neighbour has been given less attention). Gap dynamics is therefore the essence of the shifting mosaic steady state (Fig. 3 ); it is concerned with the development of species which require high levels of resources and grow rapidly (the terms 'large gap specialists' (Pickett, 1983 ) and 'pioneer' species (Whitmore, 1989 ) appear to be synonyms) and of species which require low levels of resources and grow slowly ('small gap specialists' or 'non-pioneer' species). These latter, shade tolerant species vary widely in their tolerance of shade and hence in their responses to gaps (Canham, 1989; Whitmore, 1989 ).

The foundation for tree fall and gap dynamics is 'pattern and process' (Watt, 1947 ), and much of the current work is based on Watt's defined phases, 'gap', 'building' and 'mature'. The literature to 1985 has been intensively reviewed: by Runkle ( 1985 ) for temperate forests; by Brokaw ( 1985 ) for tropical forests; and by Veblen (1985a) for Nothofagus forests in Chile. Since that time, a special feature on treefall gaps has been published in Ecology (Platt and

260 P.M. Attiwill /Forest Ecology and Management 63 (1994) 247-300

Strong, 1989 ) and a collection of papers from a symposium was published in Canadian Journal of Forest Research (Denslow and Spies, 1990 ).

It seems that the diversity of results support Grubb's (1977) conclusions: tree fall creates a wide diversity of gaps, and different species which might fill the gaps have different requirements. For example, Denslow and Spies ( 1990 ) concluded that the importance of gaps depends on the peculiarities of the system and its component species, whether or not there are established understorey shrubs which might compete with regenerating trees, whether or not there are established seedlings and if not, what is the pattern and composition of seedfall?

Most of the work on tree fall and gap dynamics has been in tropical forests and in temperate forests of the USA. The mean proportion of mature temperate forest opened to gaps each year is within the range 0.005-0.02 (Runkle, 1985 ) so that the turnover time for the whole forest is 50-200 years. Data for tropical forest are within this range (e.g. 80-138 years (Hartshorn, 1978 ); 60-135 years (Brokaw, 1985, with the exclusion of one much longer t ime); 137 years (Lang and Knight, 1983 ); 47 years (Martinez-Ramos et al., 1988 ) ). Both frequency and intensity of treefall in tropical rainforest vary widely with age, the proportion opened to gaps in individual years varying from < 0.01 to about 0.09 (Martinez-Ramos et al., 1988). This proportion is significantly and positively correlated with annual rainfall and Martinez-Ramos et al. ( 1988 ) suggest that it is this temporal and spatial variance of treefall disturbance which gives greater diversity in tropical forest than in temperate forest. Strong (1977) commented on the increase in the number of vines and epiphytes in wet lowland forest toward the equator; he suggested that the weight of vines and epiphytes in the canopy would lead to greater disturbance, a sort o f ' domino ' effect. As far as I am aware, however, there has been no detailed analysis of the relationship between species diversity in tropical forest and frequency and intensity of treefall along a latitudinal gradient.

The size of gaps ranges typically from 50 to 200 m 2 for single tree falls (Hartshorn, 1978; Brokaw, 1985; Arriaga, 1988; Howe, 1990; Runkle, 1990) up to 300-500 m 2 for multi-tree falls (Canham et al., 1990). The light regime within the gap depends on tree height (Canham et al., 1990) and on the nature of canopy damage caused by the fall of trees and limbs (Lawton, 1990 ). Light penetration into the gap decreases with decreasing latitude but even in gaps as large as 1000 m 2, direct sun reaches the forest floor for less than 4 h day-1 (Canham et al., 1990). The light regime in gaps is therefore more heterogeneous and changes more significantly with size in temperate forest than in tropical forest (Poulson and Platt, 1989). Poulson and Platt (1989) proposed that, because of this heterogeneity and variability, variation in growth rates of individual species will be greater in temperate forest than in tropical forest. This relationship has been developed in a general model (Fig. 5, from Howe, 1990) in which size of the gap has increasing influence on the availa-

P.M. Attiwill / Forest Ecology and Management 63 (1994) 247-300 261

1 O0

. 0~0~0 ~ 0 d ~ ~ O ~ O ~ O ~

-75 • ~ o d = 6

/ o~O ~ O ~-~ • ~ o d=10

"'~ / / • em / • ~ • 50 • o ~ ~o d:14

= / 0 • • • •

/ ./" • /

I I I I I

l0 20 30 40 50

Gap Radius (m) Fig. 5. Habitat available to colonizing tree species as a function of gap size, represented by the radius of a circular gap. The variable d is the distance (m) into the gap required by different species to avoid suppression. Shade-tolerant species have smaller values of d than light-demanding species (from Howe, 1990, with the permission of Oikos).

bility of space as shade tolerance decreases. Thus shade-tolerant species establish more densely in small gaps than in large gaps in tropical cloud forest in Costa Rica, while shade-intolerant species establish more densely in large gaps (Lawton and Putz, 1988 ), and in tropical lowland forest on B arro Col- orado Island, gap size has a lesser and longer lasting effect on species composition of shade-tolerant than shade-intolerant species (Brokaw and Scheiner, 1989).

Most studies of gap dynamics in tropical forest have been based in Middle America (Uhl et al., 1988b) and have been mainly concerned with post-gap regeneration in the tropical environment where seedfall is plentiful and diverse and the soil seedbank is rich (Young et al., 1987; Brandani et al., 1988; Alvarez-Buylla and Martinez-Ramos, 1990; Denslow and G6mez Diaz, 1990). In contrast, however, Uhl et al. (1988b) found that advanced growth of non- pioneer species rather than post-gap regeneration from seed was wholly dominant (advanced growth accounted for 97% of trees in single-treefall gaps) in gaps in Amazonian rainforest at San Carlos de Rio Negro. While gap size markedly affected rates of growth, it had little effect on plant density and mortality. Thus for the Amazonian forest, 'the abundance of advance regeneration and (its) high survivorship.., minimizes size and microhabitat effects' (Uhl et al., 1988b).


It is clear that there are many strategies involved in gap dynamics. In tropical cloud forest at Costa Rica, seedlings germinating from buried seed and developing from epiphytic seedlings in the crowns of fallen trees are more important in gap dynamics than are seedlings germinating from post-gap seedfall (Lawton and Putz, 1988). Sprouters appear to be as important as seeders in gap dynamics in beech (Fagus crenata) forest in Japan (Hara, 1987) and Putz and Brokaw (1989) found a strong survival of sprouts from broken trees in tropical forest at Barro Colorado Island. While sprouts were produced by a number of species after treefall in beech-hemlock forest in eastern USA, their survival was poor (Peterson and Pickett, 1991 ).

A feature of treefall in temperate forest is soil disturbance in the form of pits and mounds which may last for decades or centuries (e.g. Hutnik, 1952; Stephens, 1956; Raup, 1981; review by Schaetzl et al., 1989) and which may be of significance in soil development (Armson and Fessenden, 1973 ). This microsite variation associated with treefall is of considerable importance in maintaining species diversity in temperate forest (Hutnik, 1952; Nakashi- zuka, 1989; Peterson et al., 1990) and boreal forest (Jonsson and Esseen, 1990). In contrast, mounds and pits formed by treefall in tropical forests level out rapidly and are therefore not prominent (Putz, 1983 ). In the Amazonian tropical forest at San Carlos de Rio Negro, microhabitat within gaps had no measurable effect either on fine root biomass (Sanford, 1990) or on nutrient loss (Uhl et al., 1988b). Understorey shrubs planted across gaps in tropical forest at Costa Rica responded to the changing light environment, but not to the addition of fertilizers (Denslow et al., 1990). It seems, therefore, that the effect of microsite in tropical forest is less important than in temperate forest, in terms both of soil upheaval and of light penetration (Canham et al., 1990 ). I have not seen detailed studies on the effects of disturbance due to soil upheaval on species diversity in tropical forest apart from Barro Colorado Is- land where Putz (1983) reported that the uprooting of buried seed and the baring of mineral soil favoured the establishment of pioneer species.

There are many other examples of gap creation having an important influence on species composition, the co-existence of Abies lasiocarpa and Picea engelmannii in subalpine forest in the Rocky Mountains, Colorado (Veblen, 1986); the persistence of Picea sitchensis in coastal forests of northwestern USA where Tsuga heterophylla is the putative climax (Taylor, 1990 ) and the regeneration of the putative climax Tsuga canadensis in forests of Pseudot- suga menziesii (Spies et al., 1990) in western Oregon and Washington (to choose but a few recent examples). Pseudotsuga menziesii regenerates only on mineral soil and so it requires large-scale disturbance, a response which is typical of many species which grow in even-aged stands (e.g. Eucalyptus reg- hans in Australia (Ashton, 1981 a); Nothofagus spp. in South America (Ve-


blen, 1989) ). For these forests, the relevance of gap dynamics depends on 'the kind and heterogeneity of seedbeds, structures and resources' (Spies et al., 1990).

In summary, gaps are a fundamental part of the life of forests. The mechanisms of gap filling and the responses of growth and species composition vary widely with forest type and geography. Lieberman et al. ( 1989 ) warn of the dangers of the gap paradigm, of treating gaps as holes in an otherwise homo- geneous forest: 'considering forests as a Swiss cheese of gaps and non-gaps does not even begin to do justice to the daunting complexity of real forests. In fact, non-gaps are as heterogeneous as gaps'.

Lieberman et al. ( 1989 ) place their emphasis on the forest canopy, so that the response of individual species can be tested against the continuum of canopy closure and thereby be extrapolated to predict behaviour over the range of canopy openings. Brokaw and Scheiner ( 1989 ) developed a similar theme: 'More work is needed on dynamic processes determining the composition and structure of the understorey beneath a closed canopy. This 'subcommunity' covers much more area than do gaps (and) contains more individuals... Cre- ation of a gap in the canopy acts on patterns already established in the understorey of the closed phase'.

Large-scale disturbance by wind

Disturbance by wind occurs over a continuum, and the nature of disturbance may therefore be difficult to characterize. The blow-down or snapping of an individual tree is at one extreme, which might be defined as wholly endogenous (or the result of autogenic processes). At the other extreme, catastrophic hurricanes may be defined as wholly exogenous (an allogenic force on the community); on the other hand, there is a current view (see Boucher, 1990 ) that the composition of some rainforests is maintained by catastrophic disturbance, that periodic disturbances of rainforest preclude an equilibrium condition ever being reached.

The regeneration of balsam fir (Abies balsarnea ) forest which grows at high altitudes in the north-eastern USA is a nice example of the difficulty of characterization. The trees have a life-span of 60-80 years and at senescence, local disturbance of the weakened trees expands windward as a wave (Sprugel and Bormann, 1981 ). Regeneration of the disturbed areas follows, resulting in wind-induced waves of death and regeneration which move through the forest at a rate of 0.75-3.3 m year-~ in a windward direction. The maintenance of this forest therefore depends on disturbance by wind (Sprugel and Bor- mann, 1981 ).

In a sub-alpine forest of the Rocky Mountains, Colorado, disturbance by hurricane-force winds has actually accelerated succession (Veblen et al., 1989). The forest was dominated by lodgepole pine (Pinus contorta var. la- tifolia ) which had regenerated after fire with a sub-canopy ofAbies lasiocarpa

264 P.M. Attiwill / Forest Ecology and Management 63 (l 994) 247-300

and Picea engelmannii. Blow-down of the pine has released the sub-canopy species, thereby accelerating successional development.

There has been a long and sustained interest in the role of disturbance in north-central and north-eastern USA (see, for example, early accounts of disturbance in the pre-settlement forest in Stearns (1949) and Mclntosh (1961) ). Return times for catastrophic windthrows ( > 1 ha) in pre-settlement forests are in the range 1000-2000 years (as judged from original sur- vey records; Canham and Loucks ( 1984 ); Seischab and Orwig ( 1991 ) ); most windthrows were in the size range 1-50 ha with a lesser number in the size range 50-1000 ha (from Canham and Loucks, 1984; Seischab and Orwig, 1991 ). Hurricanes strike central New England every 20-40 years, and major, catastrophic storms occur every 100-150 years (Foster 1988a,b).

The area devastated by severe winds and hurricanes can be very large. A severe windstorm (gusts in excess of 150 km h - ~ ) in north-western Ontario, Canada, uprooted or snapped up to 50% of trees in valley bottoms and 100% of trees on hilltops over an area of 350 km 2 (Schindler et al., 1980) and the 1938 hurricane in New England, USA (gusts in excess of 200 km h - l ) , blew down more than half the trees on 43% of the area of the Harvard forest (Spurr, 1956). Damage and subsequent recovery from the 1938 hurricane to the Hubbard Brook Experimental Forest in New Hampshire, USA, was assessed from 1942 and 1978 aerial photographs (Peart et al., 1992). Damage was patchy, and the most heavily damaged area (72% of the area open) had substantially recovered 40 years later (9% open); '40 years of canopy development effectively masked the patchy effects of hurricane damage' (Peart et al., 1992).

The susceptibility of a forest to damage from severe windstorms depends on species, on stand age and structure (older and more open forests are more susceptible than younger, denser forests), on the characteristics of the site including slope and soil depth, on the physiography of the region, on the surrounding forest structure, and on the nature of the prevailing storm (Canham and Loucks, 1984; Foster, 1988a,b; Webb, 1989).

Canham and Loucks (1984) painted the broad picture of pre-settlement disturbance in forests of the northern USA, low levels of disturbance giving more or less steady state communit ies in the north-east (the northern hardwoods model of Bormann and Likens, 1979a,b, Fig. 3), intermediate levels of disturbance by severe winds, thunderstorms and fire in the less mesic forests (Brewer and Merritt, 1978; Frelich and Lorimer, 1991 ) toward the north- central region, and catastrophic disturbance in the fire-dependent coniferous forests in the north-west. For the central part of this scenario, the mixed forests of northern Wisconsin, Stearns (1949) wrote: 'There are many instances of natural catastrophe noted in the surveyors' field notes and they are still occurring in the residual stands. Windfall alone or with the other agents of mortality, fire, drought, glaze storm, insect or fungus infestation and senes-


cence keeps the forest in a constant state of flux which results in continuous variation in composition, both in space and time, within the limits of the species present'.

In tropical latitudes, catastrophic disturbance by windstorm is frequent, with trees being extensively damaged either by uprooting or snapping (Putz et al., 1983). Tropical cyclone Winifred (1 February 1986) passed through the coastal area around Innisfail, Queensland, with winds peaking at 185 km h - l ; some 1500-2000 km 2 of rainforest was damaged, an area of 300-700 km 2 being moderately to severely (boles or crowns of most trees broken, smashed or windthrown) damaged. Within 40 days, structural resilience was demonstrated by 'an overwhelming mode of short term recovery'; virtually all tree species were resprouting from leaf shoots or coppice shoots or both (Unwin et al., 1988). Webb (1958) concluded that the rainforests of north Queensland lowlands and foothills suffer severe or general damage every 3- 40 years: 'Present observations in north Queensland suggest that cyclones are a potent ecological factor which regularly upsets forest equilibrium, with far- reaching consequences for the regeneration, suppression and reproduction of species. The heterogeneity of the tropical biological environment, preserved by cyclones, has important implications for speciation' (Webb, 1958 ).

This view is re-stated for the tropical rainforests of Puerto Rico in a study of the effects of Hurricane Hugo, 1989: 'Hurricanes strike Puerto Rico every 10-30 years... Such a high frequency of hurricanes overrides other ecological factors and creates a cyclic steady state in the forest ecosystem... A stable, steady state ecosystem in which there is no net change in total biomass over time is not possible' (Basnet et al., 1992; and compare with the model for northern hardwoods, Fig. 3 ).

Hurricane Hugo was the largest hurricane to strike the Luquillo Experi- mental Forest in Puerto Rico since 1932. Basnet et al. (1992) report that every tree was damaged, and some 85% were severely damaged, the largest trees being the most susceptible to damage. It is interesting that there was greater damage in the valleys than on ridges and slopes; the possible expla- nations are that the valley soils are deep and poorly drained and hence un- stable, that rocks on slopes provide anchorage for trees, and that valley trees suffer damage from tree and branchfall on the slopes (Basnet et al., 1992 ).

Hurricane David (August 1979) passed through the Carribean island of Dominica at a speed of 22.5 km h-1, with wind velocities at its centre reaching 241 km h -~ (Lugo et al., 1983). Some quarter of a million hectares of tropical forest, one-third of the island, was heavily damaged. Regrowth of seedlings and saplings was rapid after the hurricane, with densities up to 20 plants m -2 within 1 month, and the number of species in the regrowth was greater than in forest that had not been damaged. Recovery time, to predisturbance condition, was estimated at 50 + years (Lugo et al., 1983).

In 1988, Hurricane Joan struck the east coast of Nicaragua, and winds up


to 250 km h-1 toppled some half a million hectares of rainforest. The broadleaved forest before the hurricane was very rich, with a total of 79 species of dicotyledonous plants and the ten most abundant species making up only 43% of the total (Vandermeer et al., 1990; Yih et al., 1991 ). After the hurricane, the forest regenerated profusely and predominantly by re-sprouting from the damaged trees, together with seedlings which had probably established before the hurricane. It is of great significance that there was no reduction in the diversity of plant species; that is, regeneration both of sprouts and seedlings was characteristic of primary forest, rather than of secondary pioneers. Such a result would not be anticipated from the conventional sequence coming from gap dynamics (Fig. 5 ).

Yih et al. ( 1991 ) conclude that: ' If disturbances like Hurricane Joan have occurred repeatedly in the history of Central American rainforests (on a scale of centuries ), we can no longer view them as stable equilibrium communities. As with many other ecosystems, their structure depends on patterns of natural damage and recovery. The appropriate metaphor for natural ecosystems is not eternal constancy, but rather cycles of death and resurrection'.

Lugo et al. ( 1983 ) developed a similar theme on the problems of interpret- ing succession, climax and the steady state: 'With such high frequencies, intense hurricanes could easily override the usual ecological factors that im- pinge on an ecosystem. Thus stable steady-state ecosystems probably cannot develop in hurricane-prone regions'.

It is worth recalling the views of some earlier writers. In 1958, Webb strongly questioned the concept of a 'stable tropical forest climax'. Webb quoted Rich- ards (1952) who outlined the mosaic theory of regeneration developed in 1938 by A. Aubrrville. The theory holds that species composition at a given site of mixed tropical forest is variable both in space and time. No set of species is more stable than is any other, and none is in permanent equilibrium with the environment; the forest as a whole is a constantly changing mosaic. Richards gives a translation from Aubrrville; it is a delightful passage to include here: 'To borrow a comparison from algebra, one could say that the conditions of equilibrium between the determining factors of the habitat and the numerous independent variables constituted by the characteristic species of the communi ty form a number of equations much fewer than the number of independent variables. All kinds of solutions are thus possible satisfying the equations of equilibrium' (Richards, 1952, pp. 49-50 ).

An interesting feature of rainforest and its response to major disturbance is gregarious flowering (Richards, 1952 ). Hopkins and Graham ( 1987 ) counted 54 species of shrubs and trees flowering a month after cyclone damage to tropical lowland rainforest in northern Queensland. Twelve of these 54 species had already flowered and fruited before the cyclone. Levey (1988) noted a higher intensity of fruit production in gaps in lowland rainforest in Costa Rica, the intensity increasing with gap size; it is possible that this too was due to


gregarious flowering. Lugo et al. (1983) noted 'conspicuous' flowering of Chimarrhis cymosa 1 month after very heavy hurricane damage to tropical forest on the island of Dominica. The coincidence of regenerative space and an enhanced seed supply due to gregarious flowering may be an important component of response of rainforest following disturbance (Hopkins and Graham, 1987 ).

Large-scale disturbance by fire

Disturbance by fire in forests of North America Spurr and Barnes (1980) wrote that 'fire is the dominant fact of forest his-

tory': 'The great majority of the forests of the world - excepting only the per- petually wet rain forest, such as that of southeastern Alaska, the coast of northwestern Europe and the wettest belts of the tropics - have been burned over at more or less frequent intervals for many thousands of years. Even under present-day conditions, marked by a great awareness of forest fires, the separation of forest tracts by intervening tracts of farmland and settlement, and the crossing of the forest by many roads and trails, fire continues to be a major disturbing factor in much of the North American forest' (pp. 421-422 ).

Over the past 50 years or so, the recognition of the importance of fire as an integral component of forest ecosystems has been increasing, rapidly so over the past 20 years. The role of fire in coniferous forests of North America is well documented: for the redwood (Sequoia sempervirens) and the giant sequoia (Sequoiadendron giganteum ) in California (e.g. Kilgore, 1973; Weaver, 1974; Kilgore and Taylor, 1979 ); for ponderosa pine (Pinus ponderosa ) and for Douglas-fir (Pseudotsuga menziesii) in western and north-western USA (e.g. Weaver, 1951, 1974; Cooper, 1960, 1961; Agee, 1991 ); for longleafpine (Pinus palustris) in south-eastern USA (e.g. Heyward, 1939; Garren, 1943; Komarek, 1974); for pitch pine (Pinus rigida ) communit ies in northeastern USA (Forman and Boerner, 1981; Seischab and Bernard, 1991 ) to choose but a few examples.

In the following sections, the recent literature on disturbance by fire of selected forest communit ies in North America and of Mediterranean-type communities is reviewed. Disturbance by fire of eucalypt forests in Australia will be reviewed in a second paper (Attiwill, 1994).

Northern boreal forest. Boreal coniferous forest of f ir-pine-spruce in northern America forms a band as wide as 1000 km (Larsen, 1980; Van Cleve and Dyrness, 1983 ) from Newfoundland to Alaska, a vast area greater than half the forest area of North America (Heinselman, 1981 ). There is a high frequency of lightning strikes, with fire return times of 100-300 years (Rowe and Scotter, 1973 ). In the main boreal forest, fires are high-intensity crown fires or severe surface fires of large size, often greater than 10 000 ha and

268 P.M. At#will / Forest Ecology and Management 63 (1994) 247-300

sometimes greater than 40 000 ha (Heinselman, 1981 ) or 200 000 ha (Dyr- ness et al., 1986 ). In extreme fire years, the total area burned in Alaska ranges from 1 million to 2 million hectares (Dyrness et al., 1986 ). The main boreal zone is a typical fire-dominated communi ty (Zasada, 1986 ). The vegetation is a mosaic of even aged patches of forest; there are no communit ies at steady- state, no unidirectional succession of communi ty development, diversity decreases with age, and the random perturbation by fire restarts succession, re- juvenates the growing stock and reprimes the cycling of carbon and nitrogen (Dix and Swan, 1971; Shaft and Yarranton, 1973; Heinselman, 1981, Bonan, 1990).

The dominance of fire in boreal forest is well-illustrated by a study of forest in northern Sweden dominated by Picea abies, Pinus sylvestris, Populus tre- mula and Betula spp. with a ground layer of Vaccinium spp. and Calluna vul- garis. Using tree-ring chronology up to a max imum age of 600 years, Zackris- son (1977) showed that the mean return time of fire has doubled to 155 years since fire suppression started in the 19th century. Zackrisson (1977 ) argues that: 'The vegetational diversity maintained by forest fires in the past was probably a pre-requisite for the long-term stability of boreal forest ecosystems as implied by paleoecological records... In the past the forest was in a state of continuous change, and climax stages in the classical sense can hardly have existed anywhere... Due to efficient fire suppression during the past two centuries, fire is no longer a rejuvenating factor in the forests'.

Black spruce (Picea mariana) forms extensive, almost monospecific stands in north-eastern Canada. The age of the trees increases along a northward gradient, this increase being associated with a decrease in fire return time (Sirois and Payette, 1989). Black spruce has semi-serotinous cones and is dependent on fire for regeneration (Dyrness et al., 1986 ). Seed germination and seedling survival is almost exclusive to severely burned sites where the mineral soil has been bared (Zasada et al., 1983 ).

Sirois and Payette ( 1989 ) cite evidence that treeless areas in the sub-arctic boreal-tundra zone originated from fire-induced deforestation over the last 3000 years. A number of destructive fires can therefore lead to exclusion of trees in the subarctic zone, and present evidence suggests that tundra is ex- panding into the upper boreal zone (Sirois and Payette, 1991 ).

In the northern areas of the boreal zone of North America, fires have not been (and are not capable of being) controlled to the extent that they have been in the south. At the southern transition of the boreal zone, 'near boreal forest' (e.g. in the Boundary Waters Canoe Area in northern Minnesota, Heinselman, 1973), the mean return time of fire over the past 1000 years is 60-70 years with a range of20-100 years (Swain, 1973 ). Heinselman ( 1973 ) writes of this forest: 'In much ecological literature as well as the popular press, fire has been viewed as unnatural, man-caused, and a 'disturbance' somehow not part of the ecosystem... It must now be seen as just a perturbation within


the system... (as) an essential factor in maintaining.., long-term stability and diversity'.

The stultifying effects of earlier ecological concepts of succession and climax are evident in the literature on boreal forest. Larsen (1980) concluded that each successional pathway may have its own development and progression according to the environmental conditions it creates; 'the fortunes of succession.., make most hypotheses very general indeed'. Larsen's (1980) comments on the role of fire are most relevant: 'Fire is without question the most important cause of the initiation of pioneer stages of succession. It is so ubiquitous that Dix and Swan (1971) could make no conjecture as to the nature of the forest that might develop were the area free of fire for a pro- tracted period of time. They point out: 'For Candle Lake forest as a whole, succession does not seem to be important. It seems more important that the area has probably undergone an infinite number of vegetational readjust- ments. ' They cite Rowe ( 1961 ) and confirm his opinion that the western boreal forest is a disturbance forest, one that does not fit classic concepts: 'In view of these observations, any attempt to fit the vegetation into the mold of a climax concept would be unreal and, in our opinion, unjustified.' (Larsen, 1980, p. 317).

Heinselman (1973) and others (e.g. Viereck, 1973) throw out the chal- lenge for the introduction of a programme of prescribed fires and monitored lightning fires aimed at restoring and maintaining communi ty structure and diversity. Rowe and Scotter (1973) placed the problem in context for the land manager: 'The management of national parks, nature reserves, and wilderness areas poses many questions about the use of fire. The near exclusion of wildfires in such places has had profound effects. If the major goal of such areas is to perpetuate samples of as many landscapes as possible - with the recognition that fire is an inseparable part and natural agent in the ecology of many ecosystems - then land managers must 'unsell' the false impression that all fires are bad and be prepared to use both prescribed fires and natural lightning fires in landscape management ' .

Oak-pine forests of eastern USA. The history and development of forests in eastern USA dominated by oak (Quercus alba and Quercus velutina) has been documented by Abrams ( 1992 ) and Abrams and Nowacki ( 1992 ). Oak forests occupied much of the country (except for the north-east), and a change from pine (Pinus strobus) dominat ion to oak dominat ion some 9000-10 000 years ago is associated with an increase in charcoal abundance. Abrams ( 1992 ) suggests that fire return times of 50-100 years promoted dominance and stability of oak, and that 'regions of eastern North America that lacked oak domination before European settlement probably burned too frequently (tallgrass prairie) or too infrequently (hemlock-northern hardwoods) for oak to pros- per'. Abrams concludes: 'Fire played a vital role in maintaining oak domi-


nance before European settlement. If in the current oak forest factors antag- onistic to oak regeneration, such as a lack of fire, persist into the twenty-first century, a retrenchment of oak dominance would seem inevitable'.

Subalpineforests of north-western USA. The vegetation sequence of mountain ranges of the inland north-western USA (e.g. the Grand Teton National Park, Loope and Gruell ( 1973 ) and see detailed example, Fig. 6, from Romme and Knight ( 1981 ) ) is Douglas-fir (Pseudotsuga menziesii) at lower elevations (to about 2000 m or more on south-facing slopes) followed by lodgepole pine (Pinus contorta) at intermediate elevations ( 1800-2400 m), both occurring as essentially even-aged stands from fire origin. Subalpine fir (Abies lasiocarpa) grows from 1850 m to the timberline at about 3050 m, and En- gelmann spruce (Picea engelmannii) mixes with fir in the upper part of this range. The evidence for the role of fire in maintaining the diversity of these communities over thousands of years has been clearly recognized. For example, Taylor ( 1973 ) showed that numbers of species of plants, birds and small mammals, and number of individuals of birds and small mammals, was max-

3000

PINE-SPRUCE-FIR

2900 SPRUCE-FIR FOREST OR F O ~

28O0 MEADOW COMMUN,T E E , R J

J FO : S T J LODGEPOLE 2700 f j PINE FOREST

~ N Z 2600 FOREST

DOUGI.AS / FI R

"~ 2500 WILLO~ FOREST ~" -ALDER MESIC [.I.] COMM- ASPEN "-'] FOREST k~ UNITY SAGEBRUSH

2400 DOUGLAS COMMUNITY FIR

FOREST

2300

2250 NORTll E W SOUTH SLOPE SLOPE SLOPE

I RIDGETOPI

MOST MESIC 9 ~ MOST XERIC

TOPOGRAPHIC POSITION

Fig. 6. Vegetation of the Savage Run Watershed (Medicine Bow Mountains, Wyoming) in relation to elevation and topography (from Romme and Knight, 1981, with the permission of the Ecological Society of America).


imum in the first 25 years and decreased thereafter. In alpine krummholz and heath communit ies of the North Cascades, Washington, fire results in a dis- tinctive and persistent flora which 'makes a substantial contribution to the diversity of alpine areas' (Douglas and Ballard, 1971 ). Taylor ( 1973 ) concluded that: 'It is necessary that fire be accepted as one of the natural and important environmental factors of the Park (Yellowstone National Park). Older lodgepole pine forests must be periodically burned to perpetuate natural plant and animal communi ty life cycles and to promote greater biotic diversity. This could be done by allowing certain lightning-caused fires to burn and by controlling only those that endanger Park facilities and human lives. A system of zoning that would allow quick decisions on which fires to control and which to allow to undergo natural development ' .

A policy of halting fire suppression and allowing natural fires to run their course in selected areas of Yellowstone National Park was made in 1972 (Houston, 1973; Habeck and Mutch, 1973; Loope and Gruell, 1973) and there is now intensive work on formulating strategies to manage fire as a natural ecosystem process (Keane et al., 1990 ). This is a difficult task since both fuel loadings to support fire and regeneration strategies after fire vary greatly in the steep mountainous region and are therefore to a large extent unpredict- able (e.g. Lunan and Habeck, 1973; Agee and Smith, 1984). An example of variability comes from a study of western larch (Larix occidentalis) and lodgepole pine forests in the Glacier National Park, Montana (Barrett et al., 1991 ). There are two distinct fire regimes: short intervals of 25-75 years between fires of variable intensity on dry sites and longer intervals of 120-350 years between stand-replacing fires on moist sites. Fire suppression over the last 50 years has resulted in a major reduction in fire frequency on dry sites but there has been no change on moist sites. However, 50 years is a short t ime within the primeval fire interval for stand-replacing fires, and Barrett et al. ( 1991 ) cite evidence to suggest that, given sufficient t ime for the accumulation of fire-supporting fuels, climate is the ultimate controlling factor in the occurrence of large, stand-replacing fires. Further to the north in the Glacier National Park, British Columbia, weather patterns produce the combination of intense drying of fuel, high winds and lightning and the potential for fire and its rapid spread is therefore very great. From an assessment of historical fire frequencies, Johnson et al. (1990) concluded that neither fire suppression policies nor the accidental starting of fires following European settlement have significantly altered the fire frequency. Barrett et al. ( 1991 ) conclude: '(Glacier National Park) contains one of North America's last expanses of unlogged larch-lodgepole pine forest, and large fires will continue to be important perturbations. Still, managers have tried to extinguish most fires to protect park visitors and facilities, evidently with varying degrees of suc- cess. Fire history reveals the inherent irony and futility of this approach, and consequently the need for new management strategies'.


The complexities of management strategies are shown by Romme's ( 1982 ) reconstructions of vegetation mosaics in a 73 km 2 watershed of Yellowstone National Park, Wyoming (Fig. 7 ). The diversity of landscape was maximum in the early 1800s following major fires; total exclusion of fire is predicted to increase evenness and to reduce diversity (Fig. 7 ). A carefully planned programme of prescribed burning may increase diversity but it may also have negative effects and Romme ( 1982 ) suggests 'that such an experiment would be wholly inappropriate in a natural area such as Yellowstone National Park'. Romme ( 1982 ) concluded that: 'The current fire management plan probably will be effective in maintaining natural landscape patterns in the subalpine zone if most lightning-caused fires are allowed to burn naturally, including the very large fires... Managers should expect, and allow, an occasional fire

L ( R )

91)

81/"

711

(a) Richness

~'~,~, %- / SELECTIVE

EXCLUSION • L _ . . . . . . . . . . . • J . •

>-

811.

> 70 E '..:d 611

< 50, ...] ted e¢ b-- Z "J 17 Z O ~- 16 ©

15-

(b) Evenness EXCLUSION o - o - • ° . . . . . . . .

. 4 - - . . . . • "" . ~ o~- - - ~ . :q. o- . . . . ~ -~:'- ~'* ~ SELECTI VE

14

4 ( / -

2 0 -

(c) Patchiness

NATURAL

% Watershed Area Burned Under Natural Fire Regime

i 9 17i5 1834 19115 1949

1778 1798 18'18 1838 1858 18~]8 1898 19'18 1938 1958 1978

YEAR

Fig. 7. Reconstructions of diversity in a watershed of subalpine forest, Yellowstone National Park in relation to natural and imposed fire regimes (from Romme, 1982, with the permission of the Ecological Society of America). Relative richness, relative evenness and relative patchiness are reconstructed since 1778 under three fire regimes: (a) natural, all fires allowed to burn; (b) selective, large fires excluded, small fires allowed to burn; and (c) exclusion, total fire exclusion.


covering many square kilometres; these are the fires that will exert a predom- inant influence on landscape composition and diversity for many decades to follow. Such large fires should not be viewed as unusual events occurring because of unusually high fuel accumulations'.

These views were put fiercely to the test in 1988 when 562 000 ha of the greater Yellowstone area, including 396 000 ha within Yellowstone National Park, were burned (SchuUery, 1989 ). However, Romme and Despain ( 1989 ) consider that, since the reconstructed fire history shows fires of similar magnitude in the early 1700s, the 1988 fires should not be viewed as an abnormal event. But the views of the American people in general and ecologists and foresters in particular vary. Elfring (1989) quotes Romme as saying that Americans think of parks as static curiosities rather than as dynamic ecosystems, and concludes that people see the role of national parks as 'preserving pretty places' rather than 'protecting remnants of functioning natural systems'.

On this theme, the use of prescribed fires to reduce the hazards of wildfire, for regeneration, for enhancement of wildlife habitat, for insect and disease control and for conservation of diversity of forested ecosystems in Canada has been recently reviewed (Weber and Taylor, 1992). They concluded that prescribed burning is both ecologically compatible and cost effective; it requires a 'vigorous public awareness campaign', particularly where it is to be used in parks and reserves, so that people are more aware of the dynamic nature of ecology and of the ecological goals of management.

Clark ( 1990a,b ) investigated the highly complex problems of determining fire regimes and the effects of fire suppression in the face of climatic change. The evidence is that fire frequency would have increased by 20-40% this century because the climate has become warmer and drier. Indeed, 'direct effects of climate change on forest production and decomposition may well be swamped by the more drastic effects of climate change on fire regime' (Clark, 1990a). However, fire suppression in northwestern Minnesota has resulted in significant increases in organic matter storage in litter and coarse woody debris, as well as in significant compositional changes.

Disturbance by fire in Mediterranean-type ecosystems Mediterranean-type ecosystems (sensu DiCastri, 1981; Margaris, 1981;

Trabaud, 1981; Specht and Moll, 1983) are found in the Mediterranean (monte bajo and tomillares in Spain, maquis and garrigue in France, macchia and gariga in Italy, choresh and batha in Israel, xerovuni and phrygana in Greece, gatha nabati in Syria and Lebanon), in California (chaparral and coastal sage ), in Chile (espinal and matorral ), in South Africa (renosterveld, strandveld and fynbos) and in Australia (mallee or mallee open-scrub and heathland). The fire dependence of Mediterranean-type vegetation has long been recognized (e.g. Biswell, 1974 ). For the Mediterranean region itself, Na-


veh ( 1975 ) wrote less than 20 years ago: 'Due to the destructive combination of overgrazing and land abuse with burning, wildland fires have been re- garded in the Mediterranean Region as a wholly condemnable, man-made element. This might be the reason for the few, systematic studies of fire effects and the neglect of their evolutionary role in Mediterranean vegetation. In pre- vious studies, fire has been mentioned only briefly as a destructive factor or in connection with man-induced regression stages, leading away from the 'maquiclimax'. Only (recently has fire been recognized) as one of the major ecological factors which shaped the Mediterranean landscape and affected its mosaic-like pattern of regeneration and degradation stages' (see also Naveh, 1974).

This recognition is now extensively documented; fires at frequencies in the general ranges 10-25 years or 20-50 years are fundamental to the maintenance of composition and structure of Mediterranean-type shrublands (e.g. Keeley, 1986) and heathlands (e.g. see papers in Specht 198 l c). Mediterra- nean-type ecosystems regenerate rapidly after fire (Margaris, 1981; Hanes, 1971 ). In Australian heathlands, for example (Specht, 1981a,b) plants are characterized by sclerophyllous and often resiniferous leaves, heat-resistant bark, lignotubers, protected buds which can produce epicormic shoots, and woody (bradysporous) fruits which retain viable seed for many years. In wetter communities, plants regenerate mainly from sprouts whereas in more open drier communities, plants established from seed can withstand the competi- tion from the sprouters (the strategies of maquis (or wetter) ecosystems and phryganic (or drier) ecosystems, Margaris ( 1981 ) ). The oils and resins, together with an accumulation of litter, makes the vegetation extremely fre- prone, but regenerative adaptations to fire result 'in a wealth of regenerating plants and seedlings ... thus repairing the catastrophic effect of the fire' (Specht 198 l a). In modelling these responses for the Californian chaparral, Hilbert (1987 ) suggests that resprouting might be the dominant strategy where the fire interval is short ( < 5 years) or long ( > 60 years), and seed production and seed survival would be dominant at intermediate intervals, and both would be equal in the frequency range 40-60 years.

Change in diversity of plant species in Mediterranean-type ecosystems following fire (Fig. 8, from Kruger, 1983 ) is not dramatic. Diversity is generally greatest in the first few years after fire (see also Keeley et al., 1981 ) and then declines slowly (Fig. 8). Diversity of small mammals in Australian heathland, however, appears to be entirely dependent on a mosaic of disturbance: 'The members of the small mammal community have evolved by adapting to changes in the vegetation, which provides their food and shelter, in the same way that the heath vegetation has evolved in response to specific fire regimes... The heathlands with the highest small-mammal diversity seem to be in areas where this regime has not been greatly altered and other man-made

P.M. Attiwill / Forest Ecology and Management 63 (1994) 247-300 27 5

1 O0

"' "~x Protea fynbos

.~ 8 0

- - ~ ~ ~ Heathland, Wilson's Promontory I

o 40 o - . . . . . . . . . . . . . . . .

u . . . . . Manzanita Chaparral

0

2 0 " . . Healhland, Dark Island Z ' ' -

" "b Chamise Chaparral

t ! i i

10 2( ) 30 40

T i m e s ince fire, yr.

Fig. 8. Trends in plant diversity in Mediterranean-type communities with time after fire (from Kruger, 198 3, with the permission of Springer-Verlag, Berlin).

disturbances have been minimal. Where man's impact has altered the frequency of burning the range of different seral stages is reduced' (Fox, 1983 ).

The Californian chaparral is dominated by Adenostoma fasciculatum (chamise) which regenerates after fire both by seeding and resprouting (Rundel et al., 1987 ), it is interesting that the seed-bank of chamise increases with t ime since fire, an unusual strategy for older, resprouting dominants in Mediterranean-type ecosystems (Zammit and Zedler, 1988). Short-lived or ' temporary' cover appears to be important to ecosystem stability immediately after fire (Horton and Kraebel, 1955 ) and species diversity is greatest in the first few years and decreases rapidly thereafter (Parsons, 1976; and see limited data for chamise chaparral, Fig. 8 ). Growth is most rapid in the earliest years (Hanes, 1971; Rundel and Parsons, 1979). Chaparral stands older than 60 years are 'decadent ' (Hanes, 1971 ), characterized by little growth and an accumulation of dead material, both on the shrubs and in the litter layer. There may be an accumulation of phytotoxic materials - materials which have an allelopathic effect and thereby inhibit regeneration (Hanes, 1971; Christen- sen and Muller, 1975; Keeley et al., 1985). The foliage of younger stands of chaparral is relatively rich in nitrogen and phosphorus; Rundel and Parsons (1980) suggest that fire is of critical importance in recycling nutrients from nutrient-poor 'senescent' stands to nutrient-rich regenerating stands, and that

276 P.M. A ttiwill / Forest Ecology and Management 63 (1994) 247-300

this may well be an important factor in determining the quality and quantity of foliage for vertebrate herbivores in the chaparral.

The role of allelopathy in the Californian chaparral (as in a number of other fire-dependent communities) remains uncertain. The hypothesis is that phy- totoxins are produced by plants which inhibit germination and which are denatured by fire. However, as Keeley et al. ( 1985 ) have argued, much of the evidence for allelopathy is circumstantial and lacks rigour; their data for the chaparral suggest that species have a diversity of life histories and germination cues. Fire appears not to have determined the germination response of some species, so that species diversity is enhanced by maintaining a variable frequency of fire within a given area. (Keeley, 1987 ).

The maintenance of diversity in Mediterranean-type ecosystems presents a problem for the land-manager. The accumulation of large amounts of slowly decomposing dead material (e.g. in chaparral (Hanes, 1971 ); in fynbos (Van Wilgen, 1982)) together with hot, dry summers provides the conditions for high inflammability and for fires of high intensity. Mediterranean-type shrublands and heathlands form a mosaic with other vegetation types which may differ in their fire dependency (Wells, 1962 ). They are therefore at some risk from surrounding land and Zedler et al. ( 1983 ) documented the case of major modification of chaparral after two fires in less than one year, leading to near elimination of some species. On the other hand, fire suppression in national parks causes a different set of problems; that theme will be developed later in this paper with reference to heathlands in national parks in south- eastern Australia.

Shifting agriculture in the tropics

The distribution of natural woody vegetation in 1980 in tropical latitudes (rounded areas in millions of hectares, from Lanly, 1982) is given in Table 2.

The present rate of deforestation of tropical moist forest (as defined by Myers ( 1991 ), roughly equivalent to closed broadleaved forest in Table 2 ) is 13.86X 106 ha year -1, almost double the rate in 1979. If the same rate of deforestation applies to open forest, the total rate is about 22 X 106 ha year- (Houghton, 1990 ). The rate of deforestation in tropical America is 7.68 X 106 ha year- 1, more than half the total rate (Houghton, 1990; Myers, 1991 ). The original extent of tropical moist forest was 1360 X 106 ha and the present area is 778X 106 ha of which 532X 106 ha is primary forest (Myers, 1991 ). The latter figure is 20% less than the area of 668 X 106 ha of undisturbed and pro- ductive closed broadleaved forest in 1980 (Lanly, 1982). The tropical moist forests of Brazil represent 28% of the world total, and the annual amount deforested in Brazil is more than one-third of the annual deforestation of the world's tropical moist forests (Myers, 1991 ). On a global basis, it has been

P.M. AttiwiU / Forest Ecology and Management 63 (1994) 247-300 277

Table 2 Distribution of natural woody vegetation in tropical latitudes (based on data in Lanly, 1982 )

Formation Vegetated area Fallow area

Closed broadleaved forest 1160 228 Coniferous forest 34 10 Bamboo forest 6 1 Open broadleaved forest 734 170

Total forest 1934 409

Shrub 624

Total woody vegetation 2558

Closed broadleaved forest includes 15 X 10 6 ha of mangrove forest. Fallow areas include the complex of shifting cultivation and secondary forest growth within closed broadleaved forest (almost 20% of the area) and coniferous forest which has been degraded by fire, over-exploitation and overgrazing.

estimated that 1 X 10 6 ha of land developed after forest clearing for agriculture is abandoned each year (Salati and Vose, 1983 ).

The mean rate of destruction of the world's tropical moist forest is 1.8% per year (Myers, 1991 ). Rates in Ecuador, Madagascar, Mexico, Philippines, Thailand and Vietnam are in the range 4-10% per year, and on the Ivory Coast and in Nigeria, 14-16% per year. The relatively small area of tropical forest on Haiti was almost completely destroyed by the end of the last century (Lewis and Coffey, 1985) and continuing rates of forest clearing on other islands is giving cause for concern (Lugo et al., 1981 ) (3.3% per year rate of deforestation from 1980 to 1986 in Jamaica (Eyre, 1987) ).

For Latin America, 370 × 106 ha (range 313-412 × 106 ha) were replaced by some other ecosystem over the period 1850-1985 (Houghton et al., 1991 ). This area is 28% of the forest area in 1850; 44% of the 370× 10 6 ha has been converted to pastures, 25% to croplands, 20% is degraded land, and shifting cultivation accounts for 10%. The rate of conversion of forests to managed lands in Latin America was 5.6 × 106 ha in 1980, increasing to 6.6 × 106 ha in 1985 (Houghton et al., 1991 ).

Shifting cultivation (or swidden, bush fallowing or slash-and-burn ) is a traditional form of agriculture used by at least 240 million people in the humid tropics (Padoch and Vayda, 1983 ). A small area of forest (0.5-2 ha) is felled and the debris is burnt providing an ash bed. A variety of plants is then grown until their productivity declines after 2-4 years; that area is then abandoned to the encroaching forest, and agriculture shifts to a new forest clearing. In the Amazon Territory of Venezuela, for example, shifting cultivation has been practised for more than 3000 years (Jordan, 1987). The Venezuelan studies (Jordan, 1987, 1989; Buschbacher et al., 1988 ) support the long-standing work of Nye and Greenland (1960) that 'the soil under shifting cultivation' suffers

278 P.M. Auiwill / Forest Ecology and Management 63 (1994) 247-300

little, nutritionally or physically, under traditional practices. Jordan (1987) concludes that 'in contrast to crop plants, native successional species invading an old conuco do not appear to suffer from low availability of nutrients. The wild species are adapted to take up nutrients relatively unavailable to crop species'.

Nor does the general pattern of recovery of forest structure and composition suffer under traditional shifting cultivation; Saldarriaga et al. (1988) state that shifting agriculture has influenced the Amazon basin for millenia and that, together with wind, lightning and fire, is part of the disturbance regime - presumably a regime of endogenous disturbance. Similarly, there is no clear distinction between patterns of recovery of tropical forest in South- east Asia after disturbance by shifting cultivation or by other natural disturbances (Kartawinata et al., 1989). The studies based on Venezuelan Ama- zonia are outstanding (see Fig. 1; Herrera et al., 1981; Uhl et al., 1981, 1982, 1988a,b; Jordan, 1985, 1987; Saldarriaga, 1987; Uhl, 1987; Buschbacher et al. 1987, 1988; Saldarriaga et al., 1988; Saldarriaga and Uhl, 1991 ) in that they clearly define, in terms of biomass, productivity, nutrient cycling and species composition, the decreasing capacity of the tropical forest to recover with increasing intensity of disturbance. With low-level disturbance (shifting cultivation or slash-and-burn, Fig. 1 ) return to the mature primary forest is possible with a time-scale of about 190 years. The greater the input into maintaining the clearing in a forest-free state (bulldozing, ploughing, applying fertilizers and herbicides, sowing grasses for pasture) the less certain it becomes that forest will re-establish (Fig. 1 ) and the time-scale for possible recovery increases to a thousand or more years. Under traditional resource-use the tropical forests recover from disturbance, whereas under modern, fossil fuel- using humans, the forests are destroyed (Padoch and Vayda, 1983 ). The threat of catastrophic and exogenous disturbance of tropical forest for agriculture has been eloquently summarized: 'The period of the great cattle ranches was only the beginning of the exploitation of the Amazonian nature for the benefit of foreign, non-Amazonian regions. With it started a true 'second wave of the conquista' for the Amazonian part of South America... The second wave now rolls unhindered over the interior of Amazonia and devours one of the very last reserves on earth of still almost intact nature. The aim is the exportation of all natural resources of Amazonia that can be turned to money in other countries, other continents, i.e. to the so-called 'highly developed', industrial- ized countries in North America and Europe, as well as Japan, without any benefit for the Amazonian region and the native inhabitants of its interior' (Sioli, 1991 ).

It is not only the large-scale operations of forest clearing and pasture development which are the cause for great concern; forestry operations, the increasing intensity of shifting cultivation, and the increasing risk of fire are all contributing to serious changes in forest structure and composition.


The rate of logging without any substantial measures for regeneration in the tropics has been estimated as 12 X l 0 6 ha year-~ (G6mez-Pompa and Burley, 1991 ). Bruenig ( 1985 ) makes the valid point that it is often implied that logging invariably causes deforestation, which of course is not true. How- ever, management is the key to successful logging and subsequent regeneration of the forest; ' the arguments for intensive management (of the tropical forest) have never been stronger' (Wadsworth, 1983). Bruenig (1989) concluded that the obstacles to conservation and rational, sustained-use management of tropical forests are political: 'applications of knowledge, implemen- tation of plans and improvement of forestry practice are not so much problems of scientific knowledge and technical practicability, as of public awareness, and political responsibility and goodwill'. Whitmore ( 1991 ) makes the same point: 'Nor should we forget that rain forest management fails, as it usually does, not because the silvicultural system is biologically unsound, but because of social, economic and political forces'. The deficiencies of management are vividly described for the Paragominas area of the Eastern Amazon: 'Taken alone, selective forest cutting is not a severe disturbance or cause for concern, but in the Paragominas region, several factors interact to make timber harvesting activities much more detrimental than they first appear. Since cutting is being done by people who do not own the land, there is no interest in doing the work carefully. Many more trees are cut than are actually harvested. Dur- ing skidding operations when boles with good shape are dragged to spur roads, many young saplings are crushed, and dead trees and branches are scattered about. The end result is thousands of square kilometres of cut-up forest criss- crossed with soil exposed by machinery, and laden with dead slash. The extensive canopy openings and the dry slash on the forest floor turn a normally fire-resistant ecosystem into a fire-susceptible ecosystem. Fires that are set in nearby pastures to control weeds easily spread into the logged forests. The logging roads provide a route for rapid spread of fire into the forest interior. Few forests in the Paragominas area have escaped fire damage.' (Busch- bacher et al., 1987).

Traditional shifting agriculture can be viewed as an endogenous disturbance, as outlined above. However, as the intensity of shifting agriculture increases, degradation of the ecosystem follows (Fig. 1; and see Herrera et al., 1981 ). The problem is intensified with greater accessibility, and this often follows the construction of roads for logging. Clear-felling from intensive agriculture is then associated with t imber removal, and with major disturbance by machinery as the openings become larger and site preparation becomes more intensive. This pattern of intensified shifting agriculture with increased accessibility has been recorded in Costa Rica (Ewel et al., 1981 ), in Brazil (Malingreau and Tucker, 1988; Uhl et al., 1991 ), in Thailand (Kunstadter, 1990) and in South-east Asia (Kartawinata et al., 1989).

Fire in tropical rainforest has been widespread (as shown by the common


occurrence of charcoal in Amazonian soils) but it is relatively rare (Uhl and Kauffman, 1990). When the canopy is opened, however, there is an increase in the amount of debris, an increase in temperature and a decrease in moisture content of potential fuel. The risk of spread of fire therefore increases substantially, and evidence for the Amazon Basin is that fire is now occurring at an unprecedented rate (Kauffman and Uhl, 1990).

Major disturbance by insects

Much of the ecological literature on major disturbance by insects is from North American studies on the eastern spruce budworm (Choristoneura fu- miferana) and the spruce beetle (Dendroctonus rufipennis) in north-eastern forests of balsam fir (Abies balsamea) and spruce (Picea spp. ), on the western spruce budworm (Choritoneura occidentalis) in north-western forests of Douglas-fir (Pseudotsuga menziesii) and other conifers, and on the spruce beetle and the mountain pine beetle (Dendroctonous ponderosae) in north- western forests of Engelmann spruce (Picea engelmannii), subalpine fir (Abies lasiocarpa ) and lodgepole pine ( Pinus contorta ).

The eastern spruce budworm attacks balsam fir causing high mortality. Os- ta f fand MacLean (1989) cite up to 23 million ha of dead and dying trees in a severe outbreak in 1983. In an unusually severe spruce budworm outbreak on Cape Breton Island, Nova Scotia, from 1974 to 1981, attack was first most prominent in less vigorous trees, but later was distributed equally over all size classes (MacLean and Ostaff, 1989). Budworm caused the death of 78-89% of the balsam fir and 27% of the spruce. A further 39% of the spruce was killed by spruce beetle (Ostaff and MacLean, 1989). MacLean and Ostaff (1989) found that mortality of fir due to defoliation by budworm was highly correlated with basal area. In contrast, Piene (1989a,b) found that mortality was significantly greater in spaced stands of fir than in unspaced stands.

There has been a great deal of debate over the years on the eastern spruce budworm (see reviews by Ghent et al., 1957; Baskerville, 1975; MacLean, 1980). Has it become worse with post-settlement logging and fire control? Blais ( 1983 ) describes spruce budworm outbreaks as 'a natural phenomenon - the result of the coevolution of a plant-insect system'. However, outbreaks are becoming more frequent and larger in size, the 1970 outbreak covering some 55 million ha. Blais (1983) links these increases with fire suppression, clear-felling for pulpwood and the application of insecticides, all practices which favour the growth and survival of fir, the species most preferred by the budworm. After budworm attack, there is 'an extraordinary accumulation of debris.., which offers no obstacle to the rooting of balsam' (de Gryse, 1944, cited in Ghent et al., 1957). Ostaff and MacLean (1989) concluded that 'spruce budworm outbreaks are periodic natural occurrences that.., recycle older fir stands to regenerating fir stands'.

P.M. A ttiwill / Forest Ecology and Management 63 (1994) 247-300 2 81

In the sub-alpine forest of the Rocky Mountains, succession is dominated by disturbance (Veblen et al., 1991 ). After severe fire, Engelmann spruce and lodgepole pine dominate the seedling regeneration. The shade-tolerant subalpine fir (Abies lasiocarpa) establishes on forest litter and slowly increases in abundance. Severe spruce beetle attacks on pine and spruce then shift dominance to sub-alpine fir (a non-host species). As the fir matures, it is attacked by fungi and dies, releasing sub-canopy fir and spruce; the stand then becomes increasingly susceptible to beetle attack and blowdown in wave-like oscilla- tions. 'Disturbance by spruce beetle outbreaks is wide-spread in the subalpine forests of the southern Rockies, and perhaps is as significant as fire in forest development ' (Veblen et al., 1991 ).

The westem spruce budworm attacks Douglas-fir (Pseudotsuga rnenziesii) forests of the Pacific north-western USA and British Columbia (Van Sickle et al., 1983 ) and other conifers. McCune ( 1983 ) estimated that fire suppression in mesic coniferous forest of the Bitterroot Range of the Rocky Moun- tains has increased the fire return time from 60 years pre-settlement to 7000 years between 1910 and 1980. McCune (1983 ) suggests that recent increases in intensity of budworm outbreaks are linked to fire exclusion; fire exclusion favours the shade-tolerant and budworm-susceptible species (Abies grandis, A. lasiocarpa and Picea engelmannii).

The mountain pine beetle kills extensive areas of lodgepole pine with a frequency of build-up of infestation within a given area of forest from 20 to 40 years (Amman and Schmitz, 1988): 'Mountain pine beetle-lodgepole pine interactions appear to favour survival of both species... The beetles' contin- ued role in the seral stands depends upon the presence of fire... The large buildup of fuel as a result of mountain pine beetle infestation sets the stage for a stand replacement fire that eliminates competitive tree species such as Douglas-fir, the true firs, and spruces. Following fire, lodgepole pine usually seeds-in abundantly. Mountain pine beetle outbreaks occur again close to the t ime when lodgepole pine matures, i.e. when adequate seed is available to regenerate the site to lodgepole, but before excessive amounts of seed are available that would cause an overstocked stagnated stand'.

Episodic regeneration patterns of lodgepole pine are associated with intensity and frequency of disturbances by both fire and the mountain pine beetle (Stuart et al., 1989). The forests therefore range from even-aged (regeneration after fire) to multi-aged (regeneration following tree mortality caused by endemic beetles and low fire intensity). It is of great interest to note that beetle outbreaks follow a period of decrease in the rate of growth of lodgepole pine, and that growth increases following disturbance (Stuart et al., 1989 ).

Aber and Melillo (1991) sum up these important disturbances: 'Distur- bance by fire, defoliation, or other agents is an intrinsic and necessary part of the function of most terrestrial ecosystems - a mechanism for reversing de- clining rates of nutrient cycling or relieving stand stagnation. The require-


ment for fire to reverse soil organic matter accumulation and increase nutrient cycling has been known for some time... In the cases of the budworm and the mountain pine beetle, stand break up, the reinitiation of succession, and the reversal of stand stagnation are facilitated by herbivory rather than fire'.

Other natural disturbances in forests

Disturbance other than that caused by windfall, catastrophic wind, fire and insects has not been extensively reported. A brief account of some recently reported disturbances is included here.

Ice storms Despite a plea for ' improved and standardized glaze observations' some 30

years ago (Lemon, 1961 ), there is little work on the effects of ice. Lemon ( 1961 ) suggested that the more susceptible species belong to earlier successional stages, so that the effect of ice storms is to accelerate forest succession. Such a result was found in a forest in south-eastern Wisconsin (De Steven et al., 1991 ) dominated by sugar maple (Acer saccherum) with beech (Fagus grandiflora), basswood (Tilia americana) and others as secondary dominants. Ice storms causing heavy ice accretions up to 12 cm diameter caused some mortality of beech and other dominants and sub-dominants, promoting the release of saplings of the shade-tolerant sugar maple. De Steven et al. ( 1991 ) concluded that disturbance by ice-storms accelerated succession toward increased cominance by sugar maple, together with the persistence of beech and basswood through re-sprouting.

The pattern of disturbance by ice storm in the northern hardwoods appears to be heterogeneous, and intensity varies largely with topography (Boerner et al., 1988; De Steven et al., 1991 ). In contrast, ice storms in spruce-fir forests in the Black Mountains, North Carolina, and in the Southern Appalachians have a frequency of 2-4 years and severe ice storms 'may be frequent natural disasters' (Nicholas and Zedaker, 1989). There is currently concern about decline in red spruce (Picea rubens) forests. Nicholas and Zedaker (1989) suggest that 'ice storms play a considerable role in southern high elevation forest growth and dynamics and should be evaluated before anthropogenic factors are implicated in a possible forest decline'.

Snow avalanche There is very little work on the effects of avalanches. The probability that

trees will overtop shrubs obviously increases with increasing return time. Johnson ( 1987 ) showed that return times were both shorter and less variable at the top of the avalanche path, increasing as a double exponential function


with distance down the path. Avalanches therefore are an important local determinant of the timberline (Veblen et al., 1977).

Erosional landslides and earthflows Landslides and earthflows have provided a range of age and size of distur-

bance which have been used in studies of plant succession (e.g. in the White Mountains of New Hampshire (Flaccus, 1959); the large (1.7 × 0.4 km) Gothic Earthflow in Colorado (Langenheim, 1956)). In a recent study in subtropical lower montane wet forest of Puerto Rico, Guariguata (1990) found that species composition resembled pre-disturbance condition within 52 years, and suggested that landslides may be important in maintaining diversity of communi ty structure.

Earthquake landslides and volcanic activity Garwood et al. (1979) commented on the occasional, extensive distur-

bance of tropical forest by earthquakes; they cite denudation of 54 km 2 of tropical forest in Panama and of 135 km 2 of tropical forest in New Guinea in 1935. Garwood et al. ( 1979 ) presented estimates of the rate of treefall and of erosional landslides in tropical forest generally, and of earthquake landslides in tropical forest in Panama and New Guinea. Return times are 62-125 years for treefall, 3300 years for erosional landslide, and 620-1250 years for earthquake landslide in New Guinea and 5000 years for earthquake landslide in Panama. If we allow 20 years for recovery after treefall, 50 years for recovery after erosional landslide, and 100 years for recovery after earthquake (some- what shorter times than those used by Garwood et al., 1979), then the total disturbed area is 12-25% by treefall, 1.5% by erosional landslide, 8-16% by earthquake landslide in New Guinea and 2% by earthquake landslide in Pan- ama. Earthquakes produce a patchiness of landslides and are therefore an important determinant of vegetation mosaic and species diversity in tropical forests.

Catastrophic disturbance (mass movement and volcanic eruptions) are the major determinants of temperate rainforests dominated by Nothofagus spp. in the Valdivian Andes of Chile, 'one of the most seismically-active parts of the world' (Veblen and Ashton, 1978; Veblen, 1979, 1989; Veblen et al., 1981 ). The mid-elevation forests are dominated by the evergreen and shade-intolerant N. dombeyi, together with sub-dominants of the deciduous species N. al- pina and N. obliqua. There is no seedling regeneration in the old-growth stands. Nothofagus is long-lived ( > 500 years) and the frequency of periodic catastrophic disturbance by earthquake landslides, by deposition of volcanic ash, and by large-scale windthrows ensures that 'extensive Nothofagus-free stands do not occur in the Valdivian Andes' (Veblen et al., 1981; Veblen, 1985a,b, 1989).

In the Antillanca area of Chile, the evergreen N. betuloides and the deci-


duous N. pumilio grow together on the lower slopes ( 1000 m to 1100 m or 1300 m, depending on aspect ). Above this, N. pumilio forms a narrow band, giving way to krummholz of N. antarctica and N. pumilio. The former, because of its prostrate form and its ability to produce adventitious roots from its branches, is maintained by its ability to survive the effects of scoria deposition and avalanche damage (Veblen et al., 1977). The timberline is dynamic and is determined by catastrophic events including strong winds and snow avalanches and their interactions with topographical features, such as aspect, steepness and stability of slope (Veblen et al., 1977 ).

Temperate rainforest dominated by Nothofagus in New Zealand has a similar regeneration strategy to that described above for mid-altitude Nothofagus forests in Chile, with even-aged stands regenerating after major disturbances, such as mass earth movement , snow damage or windthrow (Wardle, 1984; Stewart, 1986 ). Nothofagus solandri var. cliffortioides (mountain beech) grows in essentially pure, even-aged stands on the South Island. Regeneration is ab- sent beneath the mature forest canopy but is prolific after disturbance. Windthrows 10-100 ha are common, however: 'Wind damage is little cause for concern as it must be viewed in a framework of short-term forest stability. Periodic mortality in mountain beech forests can be seen as a regeneration strategy of a light-demanding species, since it produces ideal conditions for forest perpetuation. Forest collapse, followed by mass regeneration is thus an effective competitive mechanism against a more shade-tolerant canopy species' (Jane, 1986).

This paper does not cover the extensive ecological literature on coloniza- tion and succession of new areas emerging from lava flows or glaciers. How- ever, there is an interesting study of the effects of tephra (Antos and Zobel, 1985 ). The Mount St. Helens eruption, May 1980, destroyed some 500 km 2 of forest and deposited tephra over a much larger area. Antos and Zobel ( 1985 ) studied the effects of tephra deposits from 23 to 150 cm deep in Tsuga heterophylla and Abies amibilis forests to the north-east of Mount St. Helens, the area of max imum tephra deposition. Shrubs and trees were not greatly damaged, even where the depth of tephra was 150 cm. Bryophytes ad herbs were more affected, but few species were eliminated by any depth of tephra. Antos and Zobel ( 1985 ) place great importance on refugia such as fallen logs from which recolonization of the devastated area can spread.

The managed utilization of forests: perspectives in conclusion

Disturbance has always been recognized by ecologists and toresters, but it has been traditionally treated as a superfluous influence (Oliver and Larson, 1990). In the last 20 or so years, however, world-wide, explicit and instant communications increasingly document the magnitude, frequency and uni- versality of disturbance. Oliver and Larson (1990) make the very strong


comment that 'natural disturbances are so common and so devastating that there are probably few human disturbances for which a counterpart cannot be found in nature'.

The literature on experimentation and observation of disturbance as a natural force in forest ecology reviewed here (in part) and elsewhere (Pickett and White, 1985a; Oliver and Larson, 1990) is enormous, and the conclusions from this literature that development of most forests of the world is moulded by disturbance and that their sustainability depends on disturbance seem incontrovertible.

Since diversity, structure and function of forests are moulded by natural disturbances, it follows (as Whitmore (1982) proposed) that management of natural forests should be based on an ecological understanding of the processes of natural disturbance. I accept S.G. Taylor's (1990) concept of a 'natural' forest as one which is recognized as supporting native vegetation by simple field observation and the application of conventional phytosociological criteria. In dealing with the term 'natural forest' in Australia, Taylor (1990) states that: 'the present equilibrium vegetation (in Australia) has not been 'isolated in time' from the pre-Aboriginal native vegetation of the late Pleis- tocene. It has descended from this late Pleistocene native vegetation through an unbroken sequence of autogenic and allogenic successional responses to human-generated disturbance and other natural agents of landscape change'.

We are therefore part of the natural forest; all of our options such as fire control policies, t imber utilization plans, protection of old-growth values and conservation of biodiversity involve management decisions. But we can go little further since nature and its conservation are controversial. As examples of this controversy, Passmore' s (1974) humanism, on the one hand, treats conservation as the wise husbanding of nature in the interests of social traditions and harmony; on the other, Sheldrake's (1990) animism leads to a 'deep ecology' which is 'life-centered rather than human-centered, biocentric rather than anthropocentric; an ecology which recognizes the interconnectedness of all life and sees humanity as part of a larger living whole'.

This paper concludes with some comments on conservation of tropical forests; the argument for planned management of forests is nowhere more pow- erful than for the forests of the tropics, nor is the destruction of forest anywhere greater.

It is clear that the greater the intensity of disturbance, the less likely it is that the forest will recover (Fig. 1 ). For example, whereas traditional shifting agriculture is sustainable, the intensive and extensive clearing of tropical forest for non-sustainable agriculture which followed the Green Revolution leads to degradation (e.g. G6mez-Pompa et al., 1972). Selective logging which is not properly planned and supervised results in only the best t imber being ex- tracted, leaving behind a fragmented and altered forest; the spirit of coloni- zation (of opening up new frontiers, of building new roads) together with the


perception that the land-use value of forests is small, leads further to the clearing of the selectively-logged forests for farms and pastures (Uhl et al., 1988a, 1991; Dykstra and Heinrich, 1992).

If the high-value timber of the tropical forests is to be utilized in a way which is ecologically compatible and sustainable, selective silvicultural systems must be developed based on knowledge of size and frequency of natural gap disturbance (e.g. Kasenene and Murphy, 1991 ). The appropriate silvicultural systems must be accompanied by detailed and intensive planning, management and supervision and by appropriate revenue structures (Re- petto, 1987; Kemp, 1992; Mok, 1992; Brown and Press, 1992; and see Heal- ey's (1992) report on a conference preceding the Earth Summit (Rio de Ja- neiro, June 1992) on Wise Management of Tropical Forests). Sustainable utilization of wood must be planned in concert with sustainable utilization of non-wood products, the rich ethnobotanical yield of the tropical forest (Boom, 1985; Myers, 1988; Bennett, 1992). Utilization must be planned within appropriate systems of reservation (e.g. Nations and Komer, 1983; Katzman and Cale, 1990) and supplemented by appropriate areas of plantation (Glad- stone and Ledig, 1990). The planned and sustainable utilization of tropical forests for all of their benefits demands that short-term, larger but non-sustainable profits are forsaken for the protection of the tropical forests, its peoples and their cultures.

Whitmore ( 1990 ) concluded his book on tropical forests with words which are generally applicable to the management of forested lands and which are therefore appropriate as a summary of this paper: '...tropical forests are a renewable resource which can be utilized and still retain their diversity and richness for mankind's continuing benefit; but only if we care to learn enough about how they work, and also if, as has been repeatedly stated, utilization takes place within the limits of the forest's inherent dynamic processes'.

Acknowledgements

This review was funded by the Forest Protection Society and I am most grateful to Robyn Loydell and Dominique LaFontaine for their support and encouragement. Dr. Pauline Grierson and Isabelle Strachan assisted with li- brary research. My interest and concern in this topic have been sustained over recent years by research on the recovery of mountain ash forest following the disastrous bushfire of Ash Wednesday 1983. Australian Research Council, Rural Credits Development Fund, Ian Potter Foundation, William Buckland Foundation, Department of Conservation and Natural Resources, Victoria, and Victorian Association of Forests and Forest Industries funded various aspects of that project. My interests in tropical forests were greatly enhanced by time spent with Drs. Ernesto Medina and Elvira Cuevas, Instituto Vene-


zolano de Investigaciones Cientificas, Caracas, Venezuela, and with Dr. Ariel Lugo, Institute of Tropical Forestry, San Juan, Puerto Rico.

References

Aber, J.D. and Melillo, J.M., 1991. Terrestrial Ecosystems. Saunders College Publishing, Holt, Rinehart and Winston, Orlando, FL, 429 pp.

Abrams, M.D., 1992. Fire and the development of oak forests. Vegetatio, 42: 346-353. Abrams, M.D. and Nowacki, G.J., 1992. Historical variation in fire, oak recruitment, and post-

logging accelerated succession in central Pennsylvania. Bull. Torrey Bot. Club, 119:19-28. Adams, M.A. and Atkinson, P.I., 1991. Nitrogen supply and insect herbivory in eucalypts: the

role of nitrogen assimilation and transport processes. In: P.J. Ryan (Editor), Productivity in Perspective. Proceedings, Third Australian Forest Soils and Nutrition Conference, For- estry Commission of NSW, Sydney, pp. 239-241.

Agee, J.K., 1991. Fire history of Douglas-fir forests in the Pacific Northwest. In: Wildlife and Vegetation of Unmanaged Douglas-fir Forests. U.S. Department of Agriculture, Forest Serv- ice, Pacific Northwest Research Station, Portland, OR, pp. 25-33.

Agee, J.K. and Smith, L., 1984. Subalpine tree re-establishment after fire in the Olympic Moun- tains, Washington. Ecology, 65:810-819.

Ahlgren, I.F. and Ahlgren, C.E., 1960. Ecological effects of forest fires. Bot. Rev., 26: 483-533. Alvarez-Buylla, E.R. and Martinez-Ramos, M., 1990. Seed bank versus seed rain in the regen-

eration of tropical pioneer trees. Oecologia, 84:314-325. Amman, G.D. and Schmitz, R.F., 1988. Mountain pine beetle - lodgepole pine interactions and

strategies for reducing tree losses. Ambio, 17: 62-68. Antos, J.A. and Zobel, D.B., 1985. Recovery of forest understories buried by tephra from Mount

St. Helens. Vegetatio, 63:103-111. Armson, K.A. and Fessenden, R.J., 1973. Forest windthrows and their influence on soil mor-

phology. Soil Sci. Soc. Am. Proc., 37: 781-783. Arriaga, L., 1988. Natural disturbance and treefalls in a pine-oak forest on the Peninsula of

Baja California, Mexico. Vegetatio, 78: 73-79. Ashton, D.H., 198 la. Tall open-forest. In: R.H. Groves (Editor), Australian Vegetation. Cam-

bridge University Press, Cambridge, pp. 121-151. Ashton, D.H., 1981b. Fire in tall open-forests (wet sclerophyll forests). In: A.M. Gill, R.H.

Groves and J.R. Noble (Editors), Fire and the Australian Biota. Australian Academy of Science, Canberra, pp. 339-366.

Attiwill, P.M., 1979. Nutrient cycling in a Eucalyptus obliqua (L'Herit.) forest. III. Growth, biomass and net primary production. Aust. J. Bot., 27: 439-458.

Attiwill, P.M., 1985. Effects of fire on forest ecosystems. In: J.J. Landsberg and W. Parsons (Editors), Research for Forest Management. CSIRO, Melbourne, pp. 249-268.

Attiwill, P.M., 1991. The disturbance of forested watersheds. In: H.A. Mooney, E. Medina, D.W. Schindler, E.-D. Schulze and B.H. Walker (Editors), Ecosystem Experiments. Scope 45. Wiley, Chichester, pp. 193-214.

Attiwill, P.M., 1994. Ecological disturbance and the conservative management of eucalypt forests in Australia. For. Ecol. Manage., 63: 303-348.

Aubr6ville, A. (1938). La f6ret coloniale: les f6rets de l'Afrique occidentale frangaise. Ann. Acad. Sci. Colon. Paris, 9: 1-245.

Barrett, S.W., Arno, S.F. and Key, C.H., 1991. Fire regimes of western larch-lodgepole pine forests in Glacier National Park, Montana. Can. J. For. Res., 21:1711-1720.

Baskerville, G.L., 1975. Spruce budworm: super siviculturalist. For. Chron., 51 : 138-140.

288 e.M. Attiwill / Forest Ecology and Management 63 (1994) 247-300

Basnet, K., Likens, G.E., Scatena, F.N. and Lugo, A.E., 1992. Hurricane Hugo: damage to a tropical rain forest in Puerto Rico. Oecologia, 8: 47-55.

Beadle, N.C.W. and Costin, A.B., 1952. Ecological classification and nomenclature. Proc. Linn. Soc. NSW, 77: 61-82.

Bennett, B.C., 1992. Plants and people of the Amazonian ralnforests: The role of ethnobotany in sustainable development. BioScience, 42: 599-607.

Biswell, H.H., 1974. Effects of fire on chaparral. In: T.T. Kozlowski and C.E. Ahlgren (Editors), Fire and Ecosystems. Academic Press, New York, pp. 321-364.

Blais, J.R., 1983. Trends in the frequency, extent, and severity of spruce budworm outbreaks in eastern Canada. Can. J. For. Res., 13: 539-547.

Boerner, R.E.J., Runge, S.D., Cho, D. and Kooser, J.G., 1988. Localized ice storm damage in an Appalachian Plateau watershed. Am. Midl. Nat., 119:199-208.

Bonan, G.B., 1990. Carbon and nitrogen cycling in North American boreal forests. I. Litter quality and soil thermal effects in interior Alaska. Biogeochemistry, 10:1-28.

Boom, B.M., 1985. Amazonian indians and the forest environment. Nature, 314: 324. Bormann, F.H. and Likens, G.E., 1979a. Catastrophic disturbance and the steady state in north-

ern hardwood forests. Am. Sci., 67: 660-669. Bormann, F.H. and Likens, G.E., 1979b. Pattern and Process in a Forested Ecosystem: Distur-

bance, Development, and the Steady State based on the Hubbard Brook Ecosystem Study. Springer-Verlag, New York, 253 pp.

Boucher, D.H., 1990. Growing back after hurricanes. BioScience, 40:163-166. Brandani, A., Hartshorn, G.S. and Orians, G.H., 1988. Internal heterogeneity of gaps and spe-

cies richness in Costa Rican tropical wet forest. J. Trop. Ecol., 4: 99-119. Brewer, R. and Meritt, P.G., 1978. Wind throw and tree replacement in a climax beech-maple

forest. Oikos, 30: 149-152. Brokaw, N.V.L., 1985. Treefalls, regrowth, and community structure in tropical forests. In: S.T.A.

Pickett and P.S. White (Editors), The Ecology Of Natural Disturbance and Patch Dynam- ics. Academic Press, Orlando, FL, pp. 53-69.

Brokaw, N.V.L. and Scheiner, S.M., 1989. Species composition in gaps and structure of a tropical rainforest. Ecology, 70: 538-541.

Brown, N. and Press, M., 1992. Logging rainforests the natural way. New Sci., 1812:21-25. Bruenig, E.F., 1985. Deforestation and its ecological implications for the rain forests in south

east Asia. Environmentalist, 5:17-26. Bruenig, E.F., 1989. Use and misuse of tropical rain forests. In: H. Leigh and M.J.A. Werger

(Editors), Ecosystems of the World 14B, Tropical Rain Forest Ecosystems: Biogeographical and Ecological Studies. Elsevier, Amsterdam, pp. 611-636.

Burrell, J.P., 1981. The invasion of coastal heaths of Victoria by Leptospermum laevigatum (J. Gaerth.) F. Muell. Aust. J. Bot., 29: 747-764.

Buschbacher, R.J., Uhl, C. and Serr~o, E.A.S., 1987. Large-scale development in Eastern Ama- zonia. Pasture management and environmental effects near Paragominas, Para. In: C.F. Jor- dan (Editor), Amazonian Rain Forests Ecosystem Disturbance and Recovery. Springer-Ver- lag, New York, pp. 90-99.

Buschbacher, R, Uhl, C. and Serr~o, E.A.S., 1988. Abandoned pastures in eastern Amazonia. If. Nutrient stocks in the soil and vegetation. J. Ecol., 76: 682-699.

Canham, C.D., 1989. Different responses to gaps among shade tolerant tree species. Ecology, 70: 548-550.

Canham, C.D. and Loucks, O.L., 1984. Catastrophic windthrow in the presettlement forests of Wisconsin. Ecology, 65: 803-809.

Canham, C.D., Denslow, J.S., Platt, W.J., Runkle, J.R., Spies, T.A. and White, P.S., 1990. Light regimes beneath closed canopies in temperate and tropical forests. Can. J. For. Res., 20: 620-631.


Christensen, N.L. and Muller, C.H., 1975. Effects of fire on factors controlling plant growth in Adenostoma chaparral. Ecol. Monogr., 45: 29-55.

Clark, J.S., 1990a. Twentieth-century climate change, fire suppression, and forest production and decomposition in northwestern Minnesota. Can. J. For. Res., 20:219-232.

Clark, J.S., 1990b. Fire and climate change during the last 750 yr in northwestern Minnesota. Ecol. Monog., 60:135-159.

Cockfield, S.D., 1988. Relative availability of nitrogen in host plants of invertebrate herbivores: three possible nutritional and physiological definitions. Oecologia, 77:91-94.

Connell, J.H. and Slatyer, R.O., 1977. Mechanisms of succession in natural communities and their role in community stability and organization. Am. Nat., 111:1119-1144.

Cooper, C.F., 1960. Changes in vegetation, structure, and growth of southwestern pine forests since white settlement. Ecol. Monogr., 30:129-164.

Cooper, C.F., 1961. The ecology of fire. Sci. Am., 204:150-160. De Steven, D., Kline, J. and Matthiae, P.E., 1991. Long-term changes in a Wisconsin Fagus-

Acer forest in relation to glaze storm disturbance. J. Veg. Sci., 2: 201-208. Denslow, J.S., 1980. Patterns of plant species diversity during succession under different dis-

turbance regimes. Oecologia, 46:18-21. Denslow, J.S., 1985. Disturbance-mediated coexistence of species. In: S.T.A. Pickett and P.S.

White (Editors), The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, Orlando, FL, pp. 307-323.

Denslow, J.S. and G6mez Diaz, A.E., 1990. Seed rain to tree-fall gaps in a Neotropical rain forest. Can. J. For. Res., 20: 642-648.

Denslow, J.S. and Spies, T.A., 1990. Canopy gaps in forest ecosystems: an introduction. Can. J. For. Res., 20:619.

Denslow, J.S., Schultz, J.C., Vitousek, P.M. and Strain, B.R., 1990. Growth responses of tropical shrubs to treefall gap environments. Ecology, 71:165-179.

DiCastri, F., 1981. Mediterranean-type shrublands of the world. In: F. DiCastri, D.W. Goodall and R.L. Specht (Editors), Ecosytems of the World. Vol. 11. Mediterranean-type Shrub- lands. Elsevier, Amsterdam, pp. 1-52.

Dix, R.L. and Swan, J.M.A., 1971. The roles of disturbance and succession in upland forest at Candle Lake, Saskatchewan. Can. J. Bot., 49: 657-676.

Douglas, G.W. and Ballard, T.M., 1971. Effects of fire on alpine plant communities in the north Cascades, Washington. Ecology, 52: 1058-1064.

Dreistadt, S.H., Dahlsten, D.L. and Frankie, G.W., 1990. Urban forests and insect ecology. BioScience, 40:192-198.

Drury, W.H. and Nisbet, I.C.T., 1973. Succession. J. Arnold Arbor. Harv. Univ., 54:331-368. Dunn, G.M. and Connor, D.J., 1991. Management of Transpiration Loss and Water Yield in

Mountain Ash. Rep. Fac. Agric. Univ. Melbourne, 61 pp. Dykstra, D.P. and Heinrich, R., 1992. Sustaining tropical forests through environmentally sound

harvesting practices. Unasylva 169, 43: 9-15. Dyrness, C.T., Viereck, L.A. and Van Cleve, K., 1986. Fire in Taiga communities of interior

Alaska. In: K. Van Cleve, F.S. Chapin III, P.W. Flanagan, L.A. Viereck and C.T. Dyrness (Editors), Forest Ecosystems in the Alaskan Taiga: A Synthesis of Structure and Function. Springer-Verlag, New York, pp. 74-86.

Egler, F.E., 1954. Vegetation science concepts I. Initial floristic composition, a factor in old- field vegetation development. Vegetatio, 4:412-417.

Elfring, C., 1989. Yellowstone: fire storm over fire management. BioScience, 39: 667-672. Ewel, J., 1983. Succession. In: F.B. Golley (Editor), Ecosystems of the World. 14A. Tropical

Rain Forest Ecosystems: Structure and Function. Elsevier, Amsterdam, pp. 217-223.


Ewel, J., Berish, C., Brown, B., Price, N. and Raich, J., 1981. Slash and burn impacts on a Costa Rican wet forest site. Ecology, 62:816-829.

Eyre, L.A., 1987. Jamaica: test case for tropical deforestation. Ambio, 16: 338-343. Flaccus, E., 1959. Revegetation of landslides in the White Mountains of New Hampshire. Ecol-

ogy, 40: 692-703. Force, D.C., 1981. Postfire insect succession in southern California chaparral. Am. Nat., 117:

575-582. Forman, R.T.T. and Boerner, R.E., 1981. Fire frequency and the pine barrens of New Jersey.

Bull. Torrey Bot. Club, 108: 34-50. Foster, D.R., 1988a. Disturbance history, community organization and vegetation dynamics of

the old-growth Pisgah forest, south-western New Hampshire, U.S.A.J. Ecol., 76:105-134. Foster, D.R., 1988b. Species and stand response to catastrophic wind in central New England,

U.S.A.J. Ecol., 76: 135-151. Fox, B.J., 1983. Mammal species diversity in Australian heathlands: the importance of pyric

succession and habitat diversity. In: F.J. Kruger, D.T. Mitchell and J.U.M. Jarvis (Editors), Mediterranean-type Ecosystems: The Role of Nutrients. Springer-Verlag, Berlin, pp. 473- 489.

Frank, D.A. and McNaughton, S.J., 1991. Stability increases with diversity in plant communities: empirical evidence from the 1988 Yellowstone drought. Oikos, 62: 360-362.

Frelich, L.E. and Lorimer, C.G., 1991. Natural disturbance regimes in hemlock-hardwood forests of the upper Great Lakes region. Ecol. Monogr., 61:145-164.

Garren, K.H., 1943. Effects of fire on vegetation of the southeastern United States. Bot. Rev., 9: 617-654.

Garwood, N.C., Janos, D.P. and Brokaw, N., 1979. Earthquake-caused landslides: a major disturbance to tropical forests. Science, 205: 997-999.

Ghent, A.W., Fraser, D.A. and Thomas, J.B., 1957. Studies of regeneration in forest stands devastated by the spruce budworm. For. Sci., 3: 184-208.

Gill, A.M., Groves, R.H. and Noble, I.R. (Editors), 1981. Fire and the Australian Biota. Aus- tralian Academy of Science, Canberra, 582 pp.

Gladstone, W.T. and Ledig, F.T., 1990. Reducing pressure on natural forests through high-yield forestry. For. Ecol. Manage., 35: 69-78.

G6mez-Pompa, A. and Vfizquez-Yanes, C., 1981. Successional studies of a rain forest in Mex- ico. In: D.C. West, H.H. Shugart and D.B. Botkin (Editors), Forest Succession: Concepts and Application. Springer-Verlag, Berlin, pp. 246-266.

G6mez-Pompa, A. and Burley, F.W., 1991. The management of natural tropical forests. In: A. G6mez-Pompa, T.C. Whitmore and M. Hadley (Editors), Rain Forest Regeneration and Management. Man and the Biosphere Series Volume 6. UNESCO and The Parthenon Pub- lishing Group, N J, pp. 3-18.

G6mez-Pompa, A., V~zquez-Yanes, C. and Guevara, S., 1972. The tropical rain forest: a non- renewable resource. Science, 177: 762-765.

Grime, J.P., 1979. Plant Strategies and Vegetation Processes. Wiley, New York, 222 pp. Grubb, P.J., 1977. The maintenance of species-richness in plant communities: the importance

of the regeneration niche. Biol. Rev., 52: 107-145. Guariguata, M.R., 1990. Landslide disturbance and forest regeneration in the upper Luquillo

Mountains of Puerto Rico. J. Ecol., 78:814-832. Habeck, J.R. and Mutch, R.W., 1973. Fire-dependent forests in the Northern Rocky Moun-

tains. Quat. Res. (NY), 3: 408-424. Halpern, C.B., 1988. Early successional pathways and the resistance and resilience of forst com-

munities. Ecology, 69:1703-1715. Halpern, C.B., 1989. Early successional patterns of forest species: interactions of life history

traits and disturbance. Ecology, 70: 704-720. Hanes, T.L., 1971. Succession after fire in the chaparral of southern California. Ecol. Monogr.,

41: 27-52.

P.M. Attiwill /Forest Ecology and Management 63 (1994) 247-300 291

Hara, M., 1987. Analysis of seedling banks of a climax beech forest: ecological importance of seedling sprouts. Vegetatio, 71: 67-74.

Harborne, J.B., 1988. Introduction to Ecological Biochemistry. 3rd edn. Academic Press, Lon- don, 356 pp.

Harmon, M.E., Bratton, S.P. and White, P.S., 1983. Disturbance and vegetation response in relation to environmental gradients in the Great Smoky Mountains. Vegetatio, 55:129-139.

Hartshorn, G.S., 1978. Tree falls and tropical forest dynamics. In: P.B. Tomlinson and M.H. Zimmerman (Editors), Tropical Trees as Living Systems. Cambridge University Press, Cambridge, pp. 617-638.

Harwell, M.A., Cropper, W.P. and Ragsdale, H.L., 1977. Nutrient recycling and stability: a reevaluation. Ecology, 58: 660-666.

Healey, J., 1992. The role of scientists and foresters in the wise management of tropical forests. Trends Ecol. Evol., 7: 249-250.

Heinselman, M.L., 1973. Fire in the virgin forests of the Boundary Waters Canoe Area, Min- nesota. Quat. Res. (NY), 3: 329-382.

Heinselman, M.L., 1981. Fire and succession in the conifer forests of northern North America. In: D.C. West, H.H. Shugart, and D.B. Botkin (Editors), Forest Succession: Concepts and Application. Springer-Verlag, Berlin, pp. 374-405.

Herrera, R., Jordan, C.F., Medina, E. and KIinge, H., 1981. How human activities disturb the nutrient cycles of a tropical rainforest in Amazonia. Ambio, 10:109-114.

Heyward, F., 1939. The relation of fire to stand composition of longleaf pine forests. Ecology, 20: 287-304.

Hilbert, D.W., 1987. A model of life history of chaparral shrubs in relation to fire frequency. In: J.D. Tenhunen, F.M. Catarino, O.L. Lange and W.C. Oechel (Editors), Plant Response to Stress: Functional Analysis in Mediterranean Ecosystems. Springer-Verlag, Berlin, pp. 597- 606.

Holling, C.S., 1973. Resilience and stability of ecological systems. Annu. Rev. Ecol. Syst., 4: 1- 23.

Hopkins, M.S., 1990. Disturbance - the forest transformer. In: L.J. Webb and J. Kikkawa (Ed- itors), Australian Tropical Rainforests. Science-Values-Meaning. CSIRO, Melbourne, pp. 40-52.

Hopkins, M.S. and Graham, A.W., 1987. Gregarious flowering in a lowland tropical rainforest: A possible response to disturbance by Cyclone Winifred. Aust. J. Ecol., 12: 25-29.

Horn, H.S., 1974. The ecology of secondary succession. Annu. Rev. Ecol. Syst., 5: 25-37. Horton, J.S. and Kraebel, C.J., 1955. Development of vegetation after fire in the chamise cha-

parral of southern California. Ecology, 36: 244-262. Houghton, R.A., 1990. The global effects of tropical deforestation. Environ. Sci. Technol., 24:

414-422. Houghton, R.A., Lelkowitz, D.S. and Skole, D.L., 1991. Changes in the landscape of Latin

America between 1850 and 1985. Progressive loss of forests. For. Ecol. Manage., 38: 143- 172.

Houston, D.B., 1973. Wildfires in northern Yellowstone National Park. Ecology, 54:1111-1117. Howe, H.F., 1990. Habitat implications of gap geometry in tropical forests. Oikos, 59:14 I-144. Hutnik, R.J., 1952. Reproduction on windfalls in a northern hardwood stand. J. For., 50: 693-

694. Jane, G.T., 1986. Wind as an ecological process in mountain beech forests of Canterbury, New

Zealand. N.Z.J. Ecol., 9: 25-39. Johnson, E.A., 1987. The relative importance of snow avalanche disturbance and thinning of

canopy plant populations. Ecology, 68: 43-53. Johnson, E.A., Fryer, G.I. and Heathcott, M.J., 1990. The influence of man and climate on

frequency of fire in the interior wet belt forest, British Columbia. J. Ecol., 78: 403-412.


Jonsson, B.G. and Esseen, P-A., 1990. Treefall disturbance maintains high bryophyte diversity in a boreal spruce forest. J. Ecol., 78: 924-936.

Jordan, C.F., 1985. Nutrient Cycling in Tropical Forest Ecosystems: Principles and their Appli- cation in Management and Conservation. Wiley, Chichester, 190 pp.

Jordan, C.F., 1987. Shifting cultivation. Slash and burn agriculture near San Carlos de Rio Ne- gro. In: C.F. Jordan (Editor), Amazonian Rain Forests Ecosystem Disturbance and Recov- ery. Springer-Verlag, New York, pp. 9-23.

Jordan, C.F., 1989. An Amazonian Rain Forest. The Structure and Function of a Nutrient Stressed Ecosystem and the Impact of Slash-and-Burn Agriculture. Man and the Biosphere Series Volume 2. UNESCO and The Parthenon Publishing Group, N J, 176 pp.

Kartawinata, K., Jessup, T.C. and Vayda, A.P., 1989. Exploitation in Southeast Asia. In: H. Leigh and M.J.A. Werger (Editors), Ecosystems of the World 14B, Tropical Rain Forest Ecosystems: Biogeographical and Ecological Studies. Elsevier, Amsterdam, pp. 591-610.

Kasenene, J.M. and Murphy, P.G., 1991. Post-logging tree mortality and major branch losses in Kbale Forest, Uganda. For. Ecol. Manage., 46: 295-307.

Katzman, M.T. and Cale, W.G., 1990. Tropical forest preservation using economic incentives. BioScience, 40: 827-832.

Kauffman, J.B. and Uhl, C., 1990. Interactions of anthropogenic activities, fire, and rain forests in the Amazon Basin. In: J.G. Goldammer (Editor), Fire in the Tropical Biota. Ecological Studies 84. Springer-Verlag, Berlin, pp. 117-134.

Keane, R.E., Arno, S.F. and Brown, J.K., 1990. Simulating cumulative fire effects in Ponderosa Pine/Douglas Fir forests. Ecology, 71:189-203.

Keeley, J.E., 1986. Resilience of mediterranean shrub communities to fires. In: B. Dell, A.J.M. Hopkins and B.B. Lamont (Editors), Resilience in Mediterranean-type ecosytems. W. Junk, Dordrecht, pp. 95-112.

Keeley, J.E., 1987. Role of fire in seed germination of woody taxa in California chaparral. Ecol- ogy, 68: 434-443.

Keeley, S.C., Keeley, J.E., Hutchinson, S.M. and Johnson, A.W., 1981. Postfire succession of the herbaceous flora in southern California chaparral. Ecology, 62:1608-1621.

Keeley, J.E., Morton, B.A., Pedrosa, A. and Trotter, P., 1985. Role of allelopathy, heat and charred wood in the germination of chaparral herbs and suffrutescents. J. Ecol., 73: 445- 458.

Kemp, R.H., 1992. The conservation of genetic resources in managed tropical forests. Unasylva 169, 43: 34-40.

Kilgore, B.M., 1973. The ecological role of fire in Sierran conifer forests: Its application to Na- tional Park management. Quat. Res. (NY), 3:496-513.

Kilgore, B.M. and Taylor, D., 1979. Fire history of a sequoia-mixed conifer forest. Ecology, 60: 129-142.

Komarek, E.V., 1974. Effects of fire on temperate forests and related ecosystems. In: T.T. Ko- zlowski and C.E. Ahlgren (Editors), Fire and Ecosystems. Academic Press, New York, pp. 251-277.

Kozlowski, T.T. and Ahlgren, C.E. (Editors), 1974. Fire and Ecosystems. 1 st edn. Academic Press, New York, 542 pp.

Kruger, F.J., 1983. Plant community diversity and dynamics in relation to fire. In: F.J. Kruger, D.T. Mitchell and J.U.M. Jarvis (Editors), Mediterranean-type Ecosystems: The Role of Nutrients. Springer-Verlag, Berlin, pp. 446-472.

Kuczera, G., 1987. Prediction of water yield reductions following a bushfire in ash-mixed species eucalypt forest. J. Hydrol., 94:215-236.

Kunstadter, P., 1990. Impacts of economic development and population change on Thailand's forests. In: J.I. Furtado, W.B. Morgan, J.R. Pfaffiin, K. Ruddle (Editors), Tropical Re-


sources: Ecology and Development. Harwood Academic Publishers, Chur, Switzerland, pp. 171-190.

Lamb, D., 1985. The influence of insects on nutrient cycling in eucalypt forests: a beneficial role? Aust. J. Ecol., 10: 1-5.

Landsberg, J., Morse, J. and Khanna, P., 1990. Tree dieback and insect dynamics in remnants of native woodlands on farms. In: D.A. Saunders, A.J.M. Hopkins and R.A. How (Editors), Australian Ecosystems: 200 Years of Utilization, Degradation and Reconstruction. Surrey Beatty, Chipping Norton, NSW, Proc. Ecol. Soc. Aust., 16: 149-165.

Lang, G.E. and Knight, D.H., 1983. Tree growth, mortality, recruitment, and canopy gap formation during a 10-year period in a tropical moist forest. Ecology, 64: 1075-1080.

Langenheim, J.H., 1956. Plant succession on a subalpine earthflow in Colorado. Ecology, 37: 301-317.

Lanly, J.-P., 1982. Tropical Forest Resources, FAO Forestry Paper 30, 106 pp. Larsen, J.A., 1980. The Boreal Ecosystem. Academic Press, New York, 500 pp. Lawton, R.O., 1990. Canopy gaps and light penetration into a wind-exposed tropical lower

montane forest. Can. J. For. Res., 20: 659-667. Lawton, R.O. and Putz, F.E., 1988. Natural disturbance and gap-phase regeneration in a wind-

exposed tropical cloud forest. Ecology, 69: 764-777. Lemon, P.C., 1961. Forest ecology of ice storms. Bull. Torrey Bot. Club, 88:21-29. Levey, D.J., 1988. Tropical wet forest treefall gaps and distributions of understory birds and

plants. Ecology, 69: 1076-1089. Lewis, L.A. and Coffey, W.J., 1985. The continuing deforestation of Haiti. Ambio, 14:158-

160. Lewis, C.L., Swindel, B.F. and Tanner, G.W., 1988. Species diversity and diversity profiles:

concept, measurement, and application to timber and range management. J. Range Manage., 41: 466-469.

Lieberman, M., Lieberman, D. and Peralta, R., 1989. Forests are not just Swiss cheese: canopy stereogeometry of non-gaps in tropical forests. Ecology, 70:550-552.

Loope, L.L. and Gruell, G.E., 1973. The ecological role of fire in the Jackson Hole Area, North- western Wyoming. Quat. Res. (NY), 3: 425-443.

Loucks, O.L., 1970. Evolution of diversity, efficiency, and community stability. Am. Zool., 10: 17-25.

Lugo, A.E., Schmidt, R. and Brown, S., 1981. Tropical forests in the Caribbean. Ambio, 10: 318-324.

Lugo, A.E., Applefield, M., Pool, D.J. and McDonald, R.B., 1983. The impact of Hurricane David on the forests of Dominica. Can. J. For. Res., 13:201-211.

Lunan, J.S. and Habeck, J.R., 1973. The effects of fire exclusion on ponderosa pine communities in Glacier National Park, Montana. Can. J. For. Res., 3: 574-579.

MacLean, D.A., 1980. Vulnerability of fir-spruce stands during uncontrolled spruce budworm outbreaks: a review and discussion. For. Chron., 56:213-221.

MacLean, D.A. and Ostaff, D.P., 1989. Patterns of Balsalm-fir mortality caused by an uncontrolled spruce budworm outbreak. Can. J. For. Res., 19:1087-1095.

Malingreau, J.-P. and Tucker, C.J., 1988. Large-scale deforestation in the southeastern Amazon Basin of Brazil. Ambio, 17: 49-55.

Margaris, N.S., 1981. Adaptive strategies in plants dominating Mediterranean-type ecosystems. In: F. di Castri, D.W. Goodall and R.L. Specht (Editors), Ecosystems of the World 11, Med- iterranean-Type Shrublands. Elsevier, Amsterdam, pp. 309-315.

Martinez-Ramos, M., Alvarez-Buylla, E., Sarukh~n, J. and Pifiero, D., 1988. Treefall age deter- mination and gap dynamics in a tropical forest. J. Ecol., 76:700-716.

McCune, B., 1983. Fire frequency reduced two orders of magnitude in the Bitterroott Canyon, Montana Can. J. For. Res., 13: 212-218.


Mclntosh, R.P., 1961. Windfall in forest ecology. Ecology, 42:834. McNaughton, S.J., 1977. Diversity and stability of ecological communities: A comment on the

role of empiricism in ecology. Am. Nat., 111:515-525. Mok, S.T., 1992. Potential for sustainable tropical forest managment in Malaysia. Unasylva

169, 43: 28-40. Mooney, H.A. and Godron, M. (Editors), 1983. Disturbance and Ecosystems: Components of

Response. 1st edn. Springer-Verlag, Berlin, 292 pp. Myers, N., 1988. Tropical forest: much more than stocks of wood. J. Trop. Ecol., 4:209-221. Myers, N., 1991. Tropical forests: present status and future outlook. Clim. Change, 19: 3-32. Nakashizuka, T., 1989. Role of uprooting in composition and dynamics of an old-growth forest

in Japan. Ecology, 70: 1273-1278. Nations, J.D. and Komer, D., 1983. Central America's tropical rainforests: positive steps for

survival. Ambio, 12: 232-238. Naveh, Z., 1974. Effects of fire in the Mediterranean region. In: T.T. Kozlowski and C.E. Ahl-

gren (Editors), Fire and Ecosystems. Academic Press, New York, pp. 401-434. Naveh, Z., 1975. The evolutionary significance of fire in the Mediterranean region. Vegetatio,

29: 199-208. Nicholas, N.S. and Zedaker, S.M., 1989. Ice damage in spruce-fir forests of the Black moun-

tains, North Carolina. Can. J. For. Res., 19: 1487-1491. Noble, I.R. and Slatyer, R.O., 1980. The use of vital attributes to predict successional changes

in plant communities subject to recurrent disturbances. Vegetatio, 43:5-21. Nye, P.H. and Greenland, D.J., 1960. The Soil Under Shifting Cultivation. Technical Bulletin

No. 51, Commonwealth Agricultural Bureau, Harpenden, 156 pp. Ohmart, C.P., Stewart, L.G. and Thomas, J.R., 1983. Leaf consumption by insects in three

Eucalyptus forest types in southeastern Australia and their role in short term nutrient cycling. Oecologia, 59: 322-330.

Ohmart, C.P., Stewart, L.G. and Thomas, J.R., 1985. Effects of nitrogen concentrations of Eu- calyptus blakelyi foliage on the fecundity of Paropsis atomaria (Coleoptera: Chrysomeli- dae). Oecologia, 68: 41-44.

Oliver, C.D., 1981. Forest development in North America following major disturbances. For. Ecol. Manage., 3: 153-168.

Oliver, C.D. and Larson, B.C., 1990. Forests Stand Dynamics. McGraw-Hill, New York, 467 PP.

Olsen, M. and Lamb, D., 1988. Recovery of subtropical rainforest following storm damage. Proc. Ecol. Soc. Aust., 15: 297-301.

O'Shaughnessy, P.J. and Jayasuriya, M.D.A., 1991. Managing the ash type forests for water production in Victoria. In: F.H. McKinnell, E.R. Hopkins and J.E.D. Fox (Editors), Forest Management in Australia. Surrey Beatty, Chipping Norton, NSW, pp. 341-363.

Ostaff, D.P. and MacLean, D.A., 1989. Spruce budworm populations, defoliation, and changes in stand condition during an uncontrolled spruce budworm outbreak on Cape Breton Island, Nova Scotia. Can. J. For. Res., 19: 1077-1086.

Padoch, C. and Vayda, A.P., 1983. Patterns of resource use and human settlement in tropical forests. In: E.B. Golley (Editor), Ecosystems of the World 14A, Tropical Rain Forest Eco- systems: Structure and Function. Elsevier, Amsterdam, pp. 301-313.

Parsons, D.J., 1976. The role of fire in natural communities: an example from the southern Sierra Nevada, California. Environ. Conserv., 3:91-99.

Passmore, J., 1974. Man's Responsibility for Nature. Ecological Problems and Western Tradi- tions. Gerald Duckworth, London, 213 pp.

Peart, D.R., Cogbill, C.V. and Palmiotto, P.A., 1992. Effects of logging history and hurricane damage on canopy structure in a northern hardwoods forest. Bull. Torrey Bot. Club, 119: 29-38.


Peterson, C.J. and Pickett, S.T.A., 1991. Treefall and resprouting following catastrophic windthrow in an old-growth hemlock-hardwoods forest. For. Ecol. Manage., 42:205-217.

Peterson, C.J., Carson, W.P., McCarthy, B.C. and Pickett, S.T.A., 1990. Microsite variation and soil dynamics within newly created treefall pits and mounds. Oikos, 58: 39-46.

Pickett, S.T.A., 1983. Differential adaptation of tropical tree species to canopy gaps and its role in community dynamics. Trop. Ecol., 24: 68-84.

Pickett, S.T.A. and White, P.S. (Editors), 1985a. The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, Orlando, FL, 472 pp.

Pickett, S.T.A. and White, P.S., 1985b. Patch dynamics: a synthesis. In: S.T.A. Pickett and P.S. White (Editors), The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, Orlando, FL, pp. 371-384.

Piene, H., 1989a. Spruce budworm defoliation and growth loss in young balsam fir: defoliation in spaced and unspaced stands and individual tree survival. Can. J. For. Res., 19:1211- 1217.

Piene, H., 1989b. Spruce budworm defoliation and growth loss in young balsam fir: recovery of growth in spaced stands. Can. J. For. Res., 19: 1616-1624.

Platt, W.J. and Strong, D.R. (Editors), 1989. Special Feature - Gaps in forest ecology. Ecology, 70: 535.

Polglase, P.J., Attiwill, P.M. and Adams, M.A., 1986. Immobilization of soil nitrogen following wildfire in two eucalypt forests of south-eastern Australia. Oecol. Plant., 7:261-271.

Poljakoff-Mayber, A., Symon, D.E., Jones, G.P., Naidu, B.P. and Paleg, L.G., 1987. Nitrogen- ous compatible solutes in native South Australian plants. Aust. J. Plant Physiol., 14: 341- 350.

Poulson, T.L. and Platt, W.J., 1989. Gap light regimes influence canopy tree diversity. Ecology, 70: 553-555.

Putz, F.E., 1983. Treefall pits and mounds, buried seeds, and the importance of soil disturbance to pioneer trees on Barro Colorado Island, Panama. Ecology, 64: 1069-1074.

Putz, F.E. and Brokaw, N.V., 1989. Sprouting of broken trees on Barro Colorado Island, Pan- ama. Ecology, 70:508-511.

Putz, F.E., Coley, P.D., Lu, K., Montalvo, A. and Aiello, A., 1983. Uprooting and snapping of trees: structural determinants and ecological consequences. Can. J. For. Res., 13:1011-1020.

Raup, H.M., 1981. Physical disturbance in the life of plants. In: M.H. Nitecki (Editor), Biotic Crises in Ecological and Evolutionary Time. Academic Press, New York, pp. 39-52.

Repetto, R., 1987. Creating incentives for sustainable forest development. Ambio, 16: 94-99. Richards, P.W., 1952. The Tropical Rainforest. Cambridge University Press, Cambridge, 450

pp. Romme, W.H., 1982. Fire and landscape diversity in subalpine forests of Yellowstone National

Park. Ecol. Monogr., 52: 199-221. Romme, W.H. and Knight, D.H., 1981. Fire frequency and subalpine forest succession along a

topographic gradient in Wyoming. Ecology, 62:319-326. Romme, W.H. and Despain, D.G., 1989. Historical perspectives on the Yellowstone fires of

1988. BioScience, 39: 695-699. Rowe, J.S., 1961. Critique of some vegetational concepts as applied to forests of northwestern

Alberta. Can. J. Bot., 39: 1007-1017. Rowe, J.S. and Scotter, G.W., 1973. Fire in the boreal forest. Quat. Res. (NY), 3: 444-464. Rundel, P.W., 1981. Fire as an ecological factor. In: O.L. Lange, P.S. Noble, C.B. Osmond and

H. Zeigler (Editors), Encyclopedia of Plant Physiol., Vol. 12A, Physiological Plant Ecology I. New Series. Springer-Verlag, Berlin, pp. 501-538.

Rundel, P.W. and Parsons, D.J., 1979. Structural changes in chamise (Adenostoma fascicula- turn) along a fire-induced age gradient. J. Range Manage., 32: 462-466.

Rundcl, P.W. and Parsons, D.J., 1980. Nutrient changes in two chaparral shrubs along a fire- induced age gradient. Am. J. Bot., 67:51-58.

Rundel, P.W., Baker, G.A., Parsons, D.J. and Stohlgren, T.J., 1987. Postfire demography of


resprouting and seedling establishment by Adenostoma fasciculatum in the California chaparral. In: J.D. Tenhunen, F.M. Catarino, O.L. Lange and W.C. Oechel (Editors), Plant Response to Stress: Functional Analysis in Mediterranean Ecosystems. Springer-Verlag, Ber- lin, pp. 575-596.

Runkle, J.R., 1985. Disturbance regimes in temperate forests. In: S.T.A. Pickett and P.S. White (Editors), The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, Or- lando, FL, pp. 17-33.

Runkle, J.R., 1990. Gap dynamics in an Ohio Acer-Fagus forest and speculations on the geography of disturbance. Can. J. For. Res., 20:632-641.

Russell, E.W.B., 1983. Indian-set fires in the forests of the northeastern United States. Ecology, 64: 78-88.

Rykiel, E.J., 1985. Towards a definition of ecological disturbance. Aust. J. Ecol., 10:361-365. Salati, E. and Vose, P.B., 1983. Depletion of tropical rainforests. Ambio, 12:67-71. Saldarriaga, J.G., 1987. Recovery following shifting cultivation. A century of succession in the

Upper Rio Negro. In: C.F. Jordan (Editor), Amazonian Rain Forests Ecosystem Distur- bance and Recovery. Springer-Verlag, New York, pp. 24-33.

Saldarriaga, J.G. and Uhl, C., 1991. Recovery of forest vegetation following slash-and-burn agriculture in the Upper Rio Negro. In: A. G6mez-Pompa, T.C. Whitmore and M. Hadley (Editors), Rain Forest Regeneration and Management. Man and the Biosphere Series Vol- ume 6. UNESCO and The Parthenon Publishing Group, NJ, pp. 303-312.

Saldarriaga, J.G., West, D.C., Tharp, M.L. and Uhl, C., 1988. Long-term chronosequences of forest succession in the upper Rio Negro of Colombia and Venezuela. J. Ecol., 76: 938-958.

Sanford, R.L., 1990. Fine root biomass under light gap openings in an Amazon rain forest. Oecologia, 83: 541-545.

Schaetzl, R.J., Burns, S.F., Johnson, D.L. and Small, T.W., 1989. Tree uprooting: review of impacts on forest ecology. Vegetatio, 79:165-176.

Schindler, D.W., Newbury, R.W., Beaty, K.G., Prokopowich, J., Ruszczynski, T. and Dalton, J.A., 1980. Effects of a windstorm and forest fire on chemical losses from forested watersheds and on the quality of receiving streams. Can. J. Fish. Aquat. Sci., 37: 328-334.

Schule, W., 1990. Landscapes and climate in prehistory: interactions of wildlife, man and fire. In: J.G. Goldammer (Editor), Fire in the Tropical Biota. Ecological Studies 84. Springer- Verlag, Berlin, pp. 273-318.

Schullery, P., 1989. The fires and fire policy. BioScience, 39: 686-694. Seischab, F.K. and Bernard, J.M., 1991. Pitch pine (Pinus rigida Mill. ) communities in central

and western New York. Bull. Torrey Bot. Club, 118:412-423. Seischab, F.K. and Orwig, D., 1991. Catastrophic disturbances in the presettlement forests of

western New York. Bull. Torrey Bot. Club, 118:117-122. Shaft, M.I. and Yarranton, G.A., 1973. Diversity, floristic richness, and species eveness during

a secondary (post-fire) succession. Ecology, 54: 897-902. Sheldrake, R., 1990. The Rebirth of Nature. The Greening of Science and God. Rider, London,

215 pp. Shugart, H.H., 1984. A Theory of Forest Dynamics: The Ecological Implications of Forest

Succession Models. Springer-Verlag, New York, 278 pp. Shugart, H.H. and West, D.C., 1981. Long term dynamics of forest ecosystems. Am. Sci., 69:

647-652. Singh, G. and Geissler, E.A., 1985. Late Cenozoic history of vegetation, fire, lake levels and

climate at Lake George, New South Wales, Australia. Philos. Trans. R. Soc. Lond. B Biol. Sci., 311: 379-447.

Singh, G., Kershaw, A.P. and Clark, R., 1981. Quaternary vegetation and fire history in Aus- tralia. In: A.M. Gill, R.H. Groves and I.R. Noble (Editors), Fire and the Australian Biota. Australian Academy of Science, Canberra, pp. 23-54.


Sioli, H., 1991. Introduction to the Symposium: Amazonia - deforestation and possible effects. For. Ecol. Manage., 38:123-132.

Sirois, L. and Payette, S., 1989. Postfire black spruce establishment in subarctic and boreal Quebec. Can. J. For. Res., 19: 1571-1580.

Sirois, L. and Payette, S., 1991. Reduced postfire regeneration along a boreal forest-forest-tundra transect in northern Quebec. Ecology, 72:619-627.

Specht, R.L., 1970. Vegetation. In: G.W. Leeper (Editors), The Australian Environment. CSIRO and Melbourne University Press, Melbourne, pp. 44-67.

Specht, R.L. (Editor), 1979. Ecosystems of the World. Vol. 9A. Heathlands and Related Shrub- lands. Descriptive Studies. Elsevier, Amsterdam, 497 pp.

Specht, R.L., 198 la. Heathlands. In: R.H. Groves (Editors), Australian Vegetation. Cambridge University Press, Cambridge, pp. 253-275.

Specht, R.L., 1981b. Responses to fires in heathlands and related shrublands. In: A.M. Gill, R.H. Groves and I.R. Noble (Editors), Fire and the Australian Biota. Surrey Beatty, Chip- ping Norton, NSW, pp. 395-415.

Specht, R.L. (Editor), 1981c. Ecosystems of the World. Vol. 9B. Heathlands and Related Shrublands. Analytical Studies. Elsevier, Amsterdam, 385 pp.

Specht, R.L. and Moll, E.J., 1983. Mediterranean-type heathlands and sclerophyllous shrublands of the world: An overview. In: F.J. Kruger, D.T. Mitchell and J.U.M. Jarvis (Editors), Mediterranean-type Ecosystems: The Role of Nutrients. Springer-Verlag, Berlin, pp. 41-65.

Spies, T.A., Franklin, J.F. and Klopsch, M., 1990. Canopy gaps in Douglas-fir forests of the Cascade Mountains. Can. J. For. Res., 20: 649-658.

Sprugel, D.G. and Bormann, F.H., 1981. Natural disturbance and the steady state in high-altitude balsam fir forests. Science, 211: 390-393.

Spurr, S.H., 1956. Natural restocking of forests following the 1938 hurricane in central New England. Ecology, 37:443-451.

Spurt, S.H. and Barnes, B.V., 1980. Forest Ecology. 3rd edn. Wiley, New York, 687 pp. Stearns, F.W., 1949. Ninety years change in a northern hardwood forest in Wisconsin. Ecology,

30: 350-358. Stephens, E.P., 1956. The uprooting of trees: a forest process. Soil Sci. Soc. Am. Proc., 20: I 13-

116. Stewart, G.H., 1986. Forest dynamics and disturbance in a beech/hardwood forest, Fiordland,

New Zealand. Vegetatio, 68:115-126. Strong, D.R., 1977. Epiphyte loads, tree falls, and perennial forest disruption: a mechanism for

maintaining higher tree species richness in the tropics without animals. J. Biogeogr., 4 :215- 218.

Stuart, J.D., Agee, J.K. and Gara, R.I., 1989. Lodgepole pine regeneration in an old, self-perpetuating forest in south central Oregon. Can. J. For. Res., 19:1096-1104.

Swain, A.M., 1973. A history of fire and vegetation in Northeastern Minnesota as recorded in lake sediments. Quat. Res. (NY), 3: 383-396.

Swank, W.T. and Waide, J.B., 1980. Interpretation of nutrient cycling research in a management context: Evaluating potential effects of alternative management strategies on site productivity. In: R.H. Waring (Editor), Forests: Fresh Perspectives from Ecosystem Analysis. Oregon State University Press, Corvallis, OR, pp. 137-158.

Taylor, A.H., 1990. Disturbance and persistence of sitka spruce (Picea sitchensis (Bong) Cart. ) in coastal forests of the Pacific Northwest, North America. J. Biogeogr., 17: 47-58.

Taylor, D.L., 1973. Some ecological implications of forest fire control in Yellowstone National Park. Ecology, 54:1394-1396.

Taylor, S.G., 1990. Naturalness: the concept and its application to Australian ecosystems. In: D.A. Saunders, A.J.M. Hopkins and R.A. How (Editors), Australian Ecosystems: 200 Years


of Utilization, Degradation and Reconstruction. Surrey Beatty, Chipping Norton, NSW, Proc. Ecol. Soc. Aust., 16:411-418.

Trabaud, L., 1981. Man and fire: impacts on Mediterranean vegetation. In: F. di Castri, D.W. Goodali and R.L. Specht (Editors), Ecosystems of the World 11, Mediterranean-Type Shrublands. Elsevier, Amsterdam, pp. 523-537.

Uhl, C., 1987. Factors controlling succession following slash-and-burn agriculture in the Ama- zon. J. Ecol., 75: 377-407.

Uhl, C. and Kauffman, J.B., 1990. Deforestation, fire susceptibility, and potential tree responses to fire in the eastern Amazon. Ecology, 71: 437-449.

Uhl, C., Clark, K., Clark, H. and Murphy, P., 1981. Early plant succession after cutting and burning in the upper Rio Negro region of the Amazon Basin. J. Ecol., 69:631-649.

Uhl, C., Jordan, C., Clark, K., Clark, H. and Herrera, R., 1982. Ecosystem recovery in Amazon caatinga forest after cutting, cutting and burning, and bulldozer clearing treatments. Oikos, 38: 313-320.

Uhl, C., Buschbacher, R. and Serrao, E.A.S., 1988a. Abandoned pastures in eastern Amazonia. I. Patterns of plant succession. J. Ecol., 76: 663-681.

Uhl, C., Clark, K., Dezzeo, N. and Maquirino, P., 1988b. Vegetation dynamics in Amazonian treefall gaps. Ecology, 69:751-763.

Uhl, C., Verissimo, A., Mattos, M.M., Brandino, Z. and Vieira, I.C.G., 1991. Social, economic, and ecological consequences of selective logging in an Amazon frontier: the case of Tailfin- dia. For. Ecol. Manage., 46: 243-273.

Unwin, G.L., Applegate, G.B., Stocker, G.C. and Nicholson, D.I., 1988. Initial effects of tropical cyclone Winifred on forests in northern Queensland. In: R. Kitching (Editor), Ecology of Australia's Wet Tropics. Surrey Beatty, Chipping Norton, NSW, Proc. Ecol. Soc. Aust., 15: 283-296.

Van Cleve, K. and Dyrness, C.T., 1983. Introduction and overview of a multidisciplinary research project: the structure and function of a black spruce (Picea marina) forest in relation to other fire-affected ecosystems. Can. J. For. Res., 13: 695-702.

Van Sickle, G.A., Alfaro, R.I. and Thomson, A.J., 1983. Douglas-fir height growth affected by western spruce budworm. Can. J. For. Res., 13: 445-450.

Van Wilgen, B.W., 1982. Some effects of post-fire age on the above-ground plant biomass of fynbos (Macchia) vegetation in South Africa. J. Ecol., 70: 217-225.

Vandermeer, J., Zarnora, N., Yih, K. and Boucher, D., 1990. Regeneracion inicial en una selva tropical en la costa caribena de Nicaragua despues del huracan Juana. Rev. Biol. Trop., 38: 347-359.

Veblen, T.T., 1979. Structure and dynamics of Nothofagus forests near timberline in south- central Chile. Ecology, 60: 937-945.

Veblen, T.T., 1985a. Stand dynamics in Chilean Nothofagus forests. In: S.T.A. Pickett and P.S. White (Editors), The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, Orlando, FL, pp. 35-51.

Veblen, T.T., 1985b. Forest development in tree-fall gaps in the temperate rain forests of Chile. Nat. Geogr. Res., 1: 162-183.

Veblen, T.T., 1986. Treefalls and the coexistence of conifers in subalpine forests of the central Rockies. Ecology, 67: 644-649.

Veblen, T.T., 1989. Tree regeneration responses to gaps along a transandean gradient. Ecology, 70: 541-543.

Veblen, T.T. and Ashton, D.H., 1978. Catastrophic influences on the vegetation of the Valdi- vian Andes, Chile. Vegetatio, 36:149-167.

Veblen, T.T., Ashton, D.H., Schlegel, F.M. and Veblen, A.T., 1977. Plant succession in a timberline depressed by vulcanism in south-central Chile. J. Biogeogr., 4: 275-294.


Veblen, T.T, Donoso Z, C., Schlegel, F.M. and Escobar R.B., 1981. Forest dynamics in south- central Chile. J. Biogeogr., 8:211-247.

Veblen, T.T., Hadley, K.S., Reid, M.S. and Rebertus, A.J., 1989. Blowdown and stand development in a Colorado subalpine forest. Can. J. For. Res., 19:1218-1225.

Veblen, T.T., Hadley, K.S., Reid, M.S. and Rebertus, A.J., 1991. The response of subalpine forests to spruce beetle outbreak in Colorado. Ecology, 72:213-231.

Viereck, L.A., 1973. Wildfire in the taiga of Alaska. Quat. Res. (NY), 3: 465-495. Vitousek, P.M., Gosz, J.R., Grier, C.C., Melillo, J.M. and Reiners, W.A., 1982. A comparative

analysis of potential nitrification and nitrate mobility in forest ecosystems. Ecol. Monogr., 52: 155-177.

Wadsworth, F.H., 1983. Production of usable wood from tropical forests. In: F.B. Golley (Edi- tor), Ecosystems of the World 14A, Tropical Rain Forest Ecosystems: Structure and Func- tion. Elsevier, Amsterdam, pp. 279-288.

Walker, D., 1982. The development of resilience in burned vegetation. In: E.I. Newman (Edi- tor), The Plant Community as a Working Mechanism. Blackwell Scientific Publications, Oxford, pp. 27-43.

Walters, C.J. and Holling, C.S., 1990. Large-scale management experiments and learning by doing. Ecology, 71: 2060-2068.

Wardle, J., 1984. The New Zealand Beeches - Ecology, Utilization and Management. New Zea- land Forest Service and Caxton Press, Christchurch, 447 pp.

Waring, R.H. and Schlesinger, W.H., 1985. Susceptibility and response of forests to natural agents of disturbance. In: Forest Ecosystems: Concepts and Management. Academic Press, Orlando, FL, pp. 211-238.

Watt, A.S., 1947. Pattern and process in the plant community. J. Ecol., 35: 1-22. Weaver, H., 1951. Fire as an ecological factor in the southwestern pine forests. J. For., 49: 93-

98. Weaver, H., 1974. Effects of fire on temperate forests: Western United States. In: T.T. Ko-

zlowski and C.E. Ahlgren (Editors), Fire and Ecosystems. Academic Press, New York, pp. 279-319.

Webb, L.J., 1958. Cyclones as an ecological factor in tropical lowland rainforest, North Queens- land. Aust. J. Bot., 6: 220-228.

Webb, S.L., 1989. Contrasting windstorm consequences in two forests, Itasca State Park, Min- nesota. Ecology, 70:1167-1180.

Webb, L.J., Traeey, L.G. and Williams, W.T., 1972. Regeneration and pattern in the subtropical rain forest. J. Ecol., 60: 675-695.

Weber, M.G. and Taylor, S.W., 1992. The use of prescribed fire in the management of Canada's forested lands. For. Chron., 68: 324-334.

Webster, J.R., Waide, J.B. and Patten, B.C., 1975. Nutrient cycling and the stability of ecosystems. In: F.G. Howell, J.B. Gentry and M.H. Smith (Editors), Mineral Cycling in South- eastern Ecosystems. ERDA Conference 740513, National Technical Information Service, US Department of Commerce, Springfield, VA, pp. 1-27.

Wells, P.V., 1962. Vegetation in relation to geological substratum and fire in the San Luis Ob- ispo Quadrangle, California. Ecol. Monogr., 32: 79-103.

Westman, W.E., 1986. Resilience: concepts and measures. In: B. Dell, A.J.M. Hopkins and B.B. Lamont (Editors), Resilience in Mediterranean-type ecosytems. W. Junk, Dordrecht, pp. 5- 19.

Westman, W.E. and O'Leary, J.F., 1986. Measures of resilience: the response of coastal sage scrub to fire. Vegetatio, 65:170-189.

Weston C.J. and Attiwill, P.M., 1990. Effects of fire and harvesting on nitrogen transformations and ionic mobility in soils of Eucalyptus regnans forests of south-eastern Australia. Oecolo- gia, 83: 20-26.

3 O0 P.M. Attiwill / Forest Ecology and Management 63 (1994) 247-300

Whelan, R.J. and Main, A.R., 1979. Insect grazing and post-fire plant succession in south-west Australian woodland. Aust. J. Ecol., 4: 387-398.

White, P.S., 1979. Pattern, process, and natural disturbance in vegetation. Bot. Rev., 45: 229- 299.

White, T.C.R., 1984. The abundance of invertebrate herbivores in relation to the availability of nitrogen in stressed food plants. Oecologia, 63: 90-105.

White, P.S. and Pickett, S.T.A., 1985. Natural disturbance and patch dynamics: an introduction. In: S.T.A. Pickett and P.S. White (Editors), The Ecology Of Natural Disturbance and Patch Dynamics. Academic Press, Orlando, FL, pp. 3-13.

Whitmore, T.C., 1982. On pattern and process in forests. In: E.I. Newman (Editor), The Plant Community as a Working Mechanism. Blackwell Scientific Publications, Oxford, pp. 45-59.

Whitmore, T.C., 1989. Canopy gaps and the two major groups of forest trees. Ecology, 70: 536- 538.

Whitmore, T.C., 1990. An Introduction to Tropical Rain Forests. Clarendon Press, Oxford, 226 PP.

Whitmore, T.C., 1991. Tropical rain forest dynamics and its implications for management. In: A. G6mez-Pompa, T.C. Whitmore and M. Hadley (Editors), Rain Forest Regeneration and Management. Man and the Biosphere Series Volume 6. UNESCO and The Parthenon Pub- lishing Group, NJ, pp. 67-89.

Whittaker, R.H., 1953. A consideration of climax theory: the climax as a population and pattern. Ecol. Monogr., 23: 41-78.

Whittaker, R.H., 1956. Vegetation of the Great Smoky Mountains. Ecol. Monogr., 26: 1-80. Yih, K., Boucher, D.H., Vandermeer, J.H. and Zamora, N., 1991. Recovery of the rain forest of

southeastern Nicaragua after destruction by Hurricane Joan. Biotropica, 23:106-113. Young, K.R., Ewel, J.J. and Brown, B.J., 1987. Seed dynamics during forest succession in Costa

Rica. Vegetatio, 71: 157-173. Zackrisson, O., 1977. Influence of forest fires on the north Swedish boreal forest. Oikos, 29: 22-

32. Zammit, C.A. and Zedler, P.H., 1988. The influence of dominant shrubs, fire, and time since

fire on soil seed banks in mixed chaparral. Vegetatio, 75:175-187. Zasada, J., 1986. Natural regeneration of trees and tall shrubs on forest sites in interior Alaska.

In: K. Van Cleve, F.S. Chapin III, P.W. Flanagan, L.A. Viereck and C.T. Dyrness (Editors), Forest Ecosystems in the Alaskan Taiga: A Synthesis of Structure and Function. Springer- Verlag, New York, pp. 44-73.

Zasada, J.C., Norum, R.A., Van Veldhuizen, R.M. and Teutsch, C.E., 1983. Artificial regeneration of trees and tall shrubs in experimentally burned upland black spruce/feather moss stands in Alaska. Can. J. For. Res., 13: 903-913.

Zedler, P.H., Gautier, C.R. and McMaster, G.S., 1983. Vegetation change in response to extreme events: the effect of a short interval between fires in California chaparral and coastal scrub. Ecology, 64:809-818.

Documents

The disturbance of forest ecosystems: the ecological basis ... · PDF fileThe disturbance of forest ecosystems: ... ecological basis for conservative management ... The extensive literature