Restoration Ecology: practice and policy.

Restoration ecology is likely to be

one of the most important fields of the coming cen-
tury.

Repairing the Earth’s Key words: ecosystem repair, conceptual framework,

dynamic ecosystems, restoration goals, thresholds, suc-
Ecosystems in the cess criteria.

New Millennium Introduction

R. J. Hobbs1 T he start of the new millennium is a useful time to

reflect and take stock of where we are and where
we think we should be going. The latter part of the last
J. A. Harris2 millennium saw unprecedented changes in all aspects
of human existence on Earth, not the least of which
Abstract were the increasing numbers of humans on the planet
and increasing impacts of humanity on Earth and all its
The extent of human-induced change and damage to ecosystems. In the twilight of the second millennium
Earth’s ecosystems renders ecosystem repair an essen- we switched to a new relationship with our planet, one
tial part of our future survival strategy, and this de- in which humanity dominates all other living things, se-
mands that restoration ecology provide effective con- questers the majority of the products of photosynthesis
ceptual and practical tools for this task. We argue that and most of the available freshwater, and increasing
restoration ecology has to be an integral component of proportions of Earth’s fish stocks for its own use (Vi-
land management in today’s world, and to be broadly tousek et al. 1997). In addition, humanity is collectively
applicable, has to have a clearly articulated conceptual changing the composition of the atmosphere and trans-
basis. This needs to recognize that most ecosystems forming the Earth’s ecosystems at an unprecedented
are dynamic and hence restoration goals cannot be rate, and in the process causing widespread damage to
based on static attributes. Setting clear and achiev- the life-support systems upon which we, and every
able goals is essential, and these should focus on the other living thing, depend.
desired characteristics for the system in the future, The new millennium is thus something of a nexus for
rather than in relation to what these were in the past. humanity, at which we need to decide whether we wish
Goal setting requires that there is a clear understand- to proceed with this huge transformation of our planet,
ing of the restoration options available (and the rela- and in so doing, put our continued existence at increas-
tive costs of different options). The concept of restora- ing risk. Or whether we want to seek alternatives in
tion thresholds suggests that options are determined which we aim to protect the resources, both living and
by the current state of the system in relation to biotic abiotic, that we have left, and set about repairing some
and abiotic thresholds. A further important task is the of the damage we have inflicted in the past. It is our
development of effective and easily measured success hope that we have the collective wisdom to choose the
criteria. Many parameters could be considered for in- latter course, and it is in this context that we consider
clusion in restoration success criteria, but these are of- the rapidly developing field of restoration ecology. If
ten ambiguous or hard to measure. Success criteria we are to persist on our planet, repair of Earth’s ecosys-
need to relate clearly back to specific restoration tems and the services they provide will be an essential
goals. If restoration ecology is to be successfully prac- component of our survival strategy. How well placed is
ticed as part of humanity’s response to continued eco- the science of restoration ecology to meet this chal-
system change and degradation, restoration ecologists lenge? Does it have a sufficiently well-developed con-
need to rise to the challenges of meshing science, ceptual or theoretical base to be applied broadly? Does
it have a suitable arsenal of strategies and tactics to
tackle the often intractable problems it encounters?
1 Address correspondence to R. J. Hobbs, School of Environ-
Moreover, does it have sufficient pathways into policy
mental Science, Murdoch University, Murdoch WA 6150, and practice to enable it to be applied effectively and
Australia
2 Department of Environmental Sciences, University of East quickly?
London, Romford Road, London E15 4LZ, United Kingdom
In this paper we examine these questions, and pro-
vide what we hope will provide pointers for the way
© 2001 Society for Ecological Restoration ahead.

JUNE 2001 Restoration Ecology Vol. 9 No. 2, pp. 239–246 239

Repairing Ecosystems

Restoration as Part of Land Management On the other hand, practitioners have identified the
need for a much firmer ecological foundation for devel-
Much of the practice of ecological restoration is carried
oping and implementing restoration projects (Clewell &
out after the devastation of land has occurred, which is
Rieger 1997). In addition, it is becoming increasingly ap-
best summed up by the phrase “I wouldn’t start from
here, if I were you.” It is useful to identify where resto- parent that the assumptions underlying many restora-
ration ecology fits in with current practices in land use. tion projects have their roots in outdated concepts of
An apparent dichotomy is often erected between con- how ecological systems function. This has led to much
servation and restoration, indicating that they are con- angst over questions which are now largely irrelevant or
sidered to be alternative options. Certainly, where funds unanswerable (Pickett & Parker 1994; Wyant et al. 1995;
are short for natural resource management, priorities Parker & Pickett 1997; Middleton 1999). This is particu-
have to be decided, and choices made (for instance, be- larly true in relation to assumptions on the stability of
tween land acquisition for nature reserves and restora- ecological systems and their ability to return to particu-
tion of degraded habitat areas). However, the dichot- lar equilibrium states following disturbance (Hobbs &
omy is a false one, since restoration activities should Morton 1999) If we are to train restoration ecologists ef-
ideally be placed within a broader context of sustain- fectively for the future and equip them with skills that
able land use and conservation. In terms of nature con- are transportable from one system to another, we need
servation, there is no substitute for preserving good to have an up-to-date and comprehensive conceptual
quality habitat, and the maintenance and management framework to provide a context for their activities.
of this is a number one priority. However, in many It seems apparent then, that some attention needs to
parts of the world, this is either no longer an option be- be paid to the conceptual basis for restoration, but that
cause few areas of unaltered habitat remain, or it is no this be related back to features that are of importance in
longer sufficient since the remaining habitat on its own the practical realm (Fig. 1). As has been discussed more
cannot sustain the biota, and hence needs to be im- widely concerning the relationship between theoretical
proved or expanded. Hence restoration is an integral and applied ecology (Lawton 1996), there needs to be
part of conservation in many areas, and restoration an ongoing dialog between the conceptual and on-
ecology and conservation biology have much to gain ground aspects of restoration ecology. The conceptual
from closer interaction with each other (Young 2000). framework aims to provide general understanding of
Similarly, restoration has an integral part to play in the how ecosystems work and the factors involved in sys-
development and maintenance of sustainable production tem restoration, while on-the-ground application re-
systems. Virtually nowhere in the world can we claim to quires methodologies which can be applied in specific
have truly sustainable production systems, since all pro- situations. Ideally, there should be ongoing interaction
duction systems inevitably degrade the natural resources between the general and the specific, so that the concep-
on which they depend, or rely heavily on external sub- tual basis guides specific actions, while on-the-ground
sidies of energy, nutrients and/or water. If we are to
develop sustainable systems, we have first to repair the
damage that has been done by past and current systems.
We thus argue that ecological restoration is an inte-
gral component of land management in today’s world.
Hence, restoration ecology needs to ensure that it de-
velops and maintains links with other disciplines relat-
ing to land management.

Conceptual Base

Why should restoration ecology bother about having a

conceptual base to work from? It has been pointed out
repeatedly that ecological restoration has been, and
continues to be, practiced widely without apparent re-
course to any background conceptual framework (Allen
Figure 1. The relationship between a conceptual framework,
et al. 1997; Palmer et al. 1997). It has recently been sug- which aims to provide general understanding of how systems
gested that, while we might need to develop conceptual operate, and on-the-ground application, which requires meth-
bases to satisfy our academic need for basic research, odologies relating to particular sites and situations. The arrow
we shouldn’t let this get in the way of the huge opera- indicates the need for strong interaction and feedback be-
tional restoration tasks which are required (Young 2000). tween the two (adapted from Lawton 1996).

240 Restoration Ecology JUNE 2001

 Repairing Ecosystems

experiences feed back to refine the overall conceptual (Pickett & Parker 1994; Aronson et al. 1995). Often, past
framework (Hobbs & Yates 1997). system composition or structure are unknown or par-
We have both been involved with the development of tially known, and past data provide only snapshots of
the conceptual basis of restoration ecology (Harris & system parameters. An alternative is to use nearby ex-
Birch 1992; Harris et al. 1996; Hobbs & Norton 1996; isting systems as a model or reference; this certainly can
Hobbs 1999), as have many others recently (Aronson et al. be used to advantage in inferring the likely manage-
1993a, 1993b; Allen et al. 1997; Pfadenhauer & Grootjans ment interventions needed to restore degraded systems
1999; Urbanska et al. 1997; Whisenant 1999), and we (Yates et al. 1994; Noss 1996). Current undegraded sys-
will not reiterate much of this material here. Instead, tems at least have the advantage that their structure
we will highlight a few key points, which seem to be and dynamics can be studied in detail. These can, there-
emerging as central components of this conceptual fore, act as potential reference systems against which
framework. We will not also dwell on the definition of the success of restoration efforts in degraded systems
restoration, since this has been aired repeatedly in the can be measured. This approach is also not without
recent literature. While some argue that preciseness of problems, however, since apparent matching of the re-
definition is essential (Aronson & Le Floc’h 1996a; Higgs stored system with the reference system in terms of
1997), we take the view that goal definition is more im- composition may mask continued underlying differ-
portant than definition of terms. Whatever a particular ences in function (Zedler 1995, 1996).
activity is called (restoration, rehabilitation, repair or An alternative approach is to explicitly recognize the
other re- words), the clear enunciation of goals is essen- dynamic nature of ecosystems, and to accept that there
tial for its success, and the ability to assess the progress is a range of potential short- and long-term outcomes of
toward success. restoration projects. The aim should be to have a trans-
parent and defensible method of setting goals for resto-
ration which focus on the desired characteristics for the
Dynamic Systems and Restoration Goals
system in the future, rather than in relation to what these
Numerous attributes can be considered when we aim were in the past (Pfadenhauer & Grootjans 1999). As
to set restoration goals. For instance, Hobbs & Norton Captain Kirk on the USS Enterprise said, “What binds
(1996) identified ecosystem composition, structure, func- us to the past prevents us from embracing the future.”
tion, heterogeneity and resilience as attributes which If we change the focus of restoration from trying to re-
might be considered. Higgs (1997) similarly suggested create something from the past to trying to repair dam-
that restoration goals should focus on “ecological fidel- age and creating systems which fulfill sensible goals,
ity,” which comprises three elements; namely, struc- we will go a long way to solving many of the conun-
tural/compositional replication, functional success and drums facing the science and practice of restoration
durability. In addition, recent discussions of ecosystem ecology. Of course, the goals set for a particular area
health have put forward system vigor, organization might still include the retention or restoration of partic-
and resilience as properties which can be assessed (Rap- ular compositional or structural elements, but this
port et al. 1998), and hence could be used to develop should be only one of a number of potential goals.
goals for restoration projects. These are all fine in gen- Where it is impossible or extremely expensive to restore
eral terms, but how do we turn these into effective goals composition and structure, alternative goals are appro-
for specific projects? Which attributes should we con- priate. These may aim to repair damage to ecological
centrate on? Do we aim for the whole lot, or are some function or ecosystem services (which may be a more
more appropriate than others depending on the circum- appropriate goal in some situations, in any case – see
stances? We suggest that we need a clear rationale for below), or to create a novel system using species not na-
setting goals, which takes into account the nature of the tive to the region or suited to particular physico-chemi-
systems being restored, the factors leading to degrada- cal constraints (Wheeler et al 1995). These novel sys-
tion and the types of action required to achieve restora- tems will be appropriate in some situations and not in
tion of different attributes. others, depending on the pre-defined goals of the resto-
Ecosystems are naturally dynamic entities, and hence ration activities.
the setting of restoration goals in terms of static compo- Goal setting thus becomes an extremely important
sitional or structural attributes is problematic. Much of component of the restoration process. Goals for a par-
restoration ecology is backward looking, seeking to re- ticular site, or more broadly for a landscape, will need
create ecosystems with properties which were charac- to be determined iteratively by considering the ecologi-
teristic of the system at some time in the past. There has cal potential for restoration and matching this against
been increasing debate as to whether this is either desir- societal desires. Higgs (1997) has suggested that “Good
able or possible, due to the dynamic nature of ecosys- ecological restoration entails negotiating the best possi-
tems, and the irreversibility of some system changes ble outcome for a specific site based on ecological

JUNE 2001 Restoration Ecology 241

Repairing Ecosystems

knowledge and the diverse perspectives of interested weed species frequently miss the point that the weed
stakeholders: to this end it is as much process as prod- invasion is merely a symptom of more fundamental
uct oriented.” This argues for an adaptive approach to system change (Hobbs & Humphries 1995). Hence, res-
restoration (Fig. 2), which garners ecological knowl- toration activities need to be prefaced by a rigorous as-
edge from as many sources as possible (including on- sessment of the current state of the particular system or
the-ground practitioners), and uses this knowledge to landscape, and the underlying factors leading to that
develop ecological response models which can indicate state. Once this has been achieved, a clearer picture of
the likely outcomes of restoration activities. Which res- the necessary restoration activities is possible, and a
toration option is taken up is decided on the basis of range of restoration options can be arrived at.
stakeholder expectations and goals, and the extent to This is where the requirement for ecological response
which it is implemented depends on the degree of fi- models becomes apparent. These models can be simple
nancial and resource input from various sectors, includ- or complex, quantitative or conceptual, but they need to
ing individual investment and public subsidy or incen- capture the essence of the system and its dynamics.
tives (Hobbs & Saunders 2001). As Higgs (1997) points Here again, there needs to be consideration of both gen-
out, the success of restoration depends greatly on an eral characteristics of ecosystems and more specific ele-
open and effective process of arriving at mutually- ments relating to specific cases. A general feature of
agreed upon restoration goals. many systems seems to be the potential for the system
to exist in a number of different states, and the likeli-
hood that restoration thresholds exist, which prevent
Restoration Options
the system from returning to a less-degraded state
Arriving at clear restoration goals requires that there is without the input of management effort (Hobbs &
a clear picture of the restoration options available for a Norton 1996). Whisenant (1999) has recently suggested
particular site, landscape or region. Often, restoration that two main types of such threshold are likely: one
projects launch full steam ahead into activities which that is caused by biotic interactions, and the other
may either be inappropriate to particular goals or caused by abiotic limitations. Figure 3a illustrates these
which target apparent symptoms without considering two thresholds and indicates that the type of restoration
underlying causes. For instance, in the Western Austra- response needed depends on which, if any, thresholds
lian wheatbelt, fencing out livestock is frequently seen have been crossed. If the system has degraded mainly
to be a primary activity needed for the restoration of na- due to biotic changes (such as grazing-induced changes
tive woodland communities, but this fails to address in vegetation composition), restoration efforts need to
more fundamental changes caused by soil degradation focus on biotic manipulations which remove the de-
(Yates et al. 2000). Similarly, projects that aim at system grading factor (e.g., the grazing animal) and adjust the
restoration through the removal or control of invasive biotic composition (e.g., replant desired species). If, on
the other hand, the system has degraded due to
changes in abiotic features (such as through soil erosion
or contamination), restoration efforts need to focus first
on removing the degrading factor and repairing the
physical and/or chemical environment. In the latter
case, there is little point in focusing on biotic manipula-
tion without first tackling the abiotic problems.
The above argument is akin to ensuring that system
functioning is corrected or maintained before questions
of biotic composition and structure are considered.
Considering system function provides a useful frame-
work for the initial assessment of the state of the system
and the subsequent selection of repair measures (Tong-
way & Ludwig 1996; Ludwig et al. 1997). Where func-
Figure 2. A framework for identifying restoration options
tion is not impaired, restoration can legitimately focus
based on response models developed from a variety of data
on composition and structure as parameters to be con-
sources and in relation to the goals of individual managers
and society at large. Implementation of particular options will sidered when setting goals.
depend on the availability of resources, policy instruments, The same scheme can be considered at a landscape
etc. Monitoring and evaluation is an essential part of the pro- scale. Hobbs and Norton (1996) and others have em-
cess, which not only assesses the success of a project in rela- phasized the need for restoration ecology to develop ef-
tion to the stated goals, but also feeds back to the response fective approaches for broad-scale restoration at land-
model (modified from Hobbs & Saunders 2001). scape and regional scales. At broad scales, however, it

242 Restoration Ecology JUNE 2001

 Repairing Ecosystems

type of threshold relates to the loss of biotic connectiv-

ity as habitat becomes increasingly fragmented and
modified, while another relates to whether landscape
modification has resulted in broad-scale changes in land-
scape physical processes, such as hydrology. Here again,
this schema can assist in setting restoration priorities. If
the landscape has crossed a biotic threshold, restoration
needs to aim at restoring connectivity. If, on the other
hand, a physical threshold has been crossed, this needs
to be treated as a priority. Hence, for instance, in a frag-
mented forested landscape, the primary goal may be
the provision of additional habitat or reestablishing
connectivity for particular target species, whereas in a
modified river or wetland system, the primary need
may be to reestablish water flows (Middleton 1999).
Of course, within these broad categorizations, there
may be numerous sub-categories and thresholds. For
instance, McIntyre and Hobbs (1999, 2001) have re-
cently explored how to categorize landscapes in terms
of the degree of habitat destruction and modification,
and hence how to assign management and restoration
priorities. It may also be the case that the restoration ac-
tivities required to overcome particular physical changes
also act to overcome biotic thresholds. An example of
this would be if extensive revegetation is required to
counteract hydrological imbalances, and at the same
time can have a positive impact on biotic connectivity
(Hobbs 1993; Hobbs et al. 1993).
Once the options for restoration have been derived
from an ecological response model, these then have to
be considered in the broader context of individual and
societal goals. To succeed, restoration activities need
not only to be based on sound ecological principles and
information, but also to be economically possible and
practically achievable. They also have to take their
place amongst other options such as providing more re-
Figure 3. (a) Conceptual model of system transitions between sources to protect existing habitats. There is also always
states of varying levels of function, illustrating the presence of the “do nothing” option, which is often the easiest, but
two types of restoration threshold, one controlled by biotic in- not necessarily the most desirable. Often another pri-
teractions and one controlled by abiotic limitations (adapted mary driver in deciding which options will be pursued
from Whisenant 1999). (b) A similar model applied to land- is the prevailing political climate, which drives govern-
scapes, indicating transition thresholds controlled by loss of ment support and funding for restoration activities. Un-
biotic connectivity and loss of physical landscape function. fortunately, political opportunism often plays more of a
part in setting priorities and deciding on options than
any rational process.
becomes even more difficult to decide what should be
restored, where and how. Attempts to focus on key
Measurements of Success
landscape attributes have so far provided many possi-
ble parameters (Aronson & Le Floc’h 1996b), but not We will not discuss in detail the implementation of res-
much of a framework in which to set priorities and toration projects here, but will consider the need for ad-
goals. We suggest that a start can be made on this by equate measures of progress toward agreed-upon resto-
considering whether restoration thresholds exist at the ration goals. These are important for many reasons, not
landscape scale. It can be hypothesized that similar the least of which are the statutory requirements often
threshold types might exist at this scale as are apparent placed on management agencies, mining companies and
in particular ecosystems or sites (Fig. 3b). Thus, one the like to demonstrate adequate achievement of stated

JUNE 2001 Restoration Ecology 243

Repairing Ecosystems

goals. If we have goals relating to composition, struc- balanced by uptake. These parameters are readily ame-
ture, function and the like, what measures do we use to nable to measurement. This may be achieved by careful
quantify the success, or otherwise, of the restoration measurement of the rates of flux of small molecules
process? from a site, combined with an estimation of how much
There have been numerous attempts to provide cate- material is bound in the living biomass. This then offers
gories of assessment that will contribute to a picture of a true “systems condition” parameter with which to
the “healthy ecosystem,” which have varying degrees of gauge success.
ease of measurement. Biological potential inventory is A potential index for use in tracking restoration is
probably the earliest form of ecosystem assessment, typi- that of “1/f noise.” 1/f noise is the signal that emerges
fied by the species list. This can take the form of a simple when the rate of change in a parameter of a system is
list of plant species, extending to complex descriptions of measured. For example, if the rate of change in the
everything from bacteria to avian guild structures, in- height of the water table in a peat bog is measured, and
cluding abundance measurements. Although this can be the reciprocal of this rate plotted, then the 1/f power re-
extremely useful for assessing conservation status, and is lationship results. This reciprocal signal of rate fluxes
greatly improved by measurements over time, it often can be found in a range of phenomena as diverse as
does not get to the basics of what is causing the degrada- traffic flow, evolutionary extinction rates and stock
tion, rather simply reflecting the magnitude and direc- market price fluctuations (Bak 1997), and indicates that
tion of its effect. We also need to ask what level of struc- the system is fluctuating “efficiently.” We can measure
tural/compositional replication we want to set as a goal. rates of change in water levels, fixation rates, nitrogen
We also need to consider how this relates to normal suc- fluxes and population sizes. In natural systems we
cessional processes (Parker 1997). If the goal is to speed should get 1/f noise signals. Therefore, one concrete
up system development beyond what would happen aim of a restoration would be to “restore” this signal.
without intervention, how fast is fast enough, and can This treatment of how to measure the progress of res-
we compare different trajectories effectively? Can we be toration projects has, like most others in the literature,
sure that a trajectory model for system development is been superficial and poses more questions than it an-
appropriate (Zedler & Callaway 1999)? swers. We suggest that measures of success have to be
More complex measurements of biological integrity linked back to clear definitions of goals for restoration.
can assess food-web complexity and the development Assessment processes can be complicated and expen-
of symbiotic relationships. However, difficulty of as- sive, and if they are too complicated or expensive, they
sessment increases greatly. Measurements relating to will not be carried out. There is no point in assessing
ecosystem function can include measurements of pro- something unless it relates to specific goals. If the resto-
duction, standing crop, mass balance and mineral cy- ration goal is to “reestablish a diverse vegetation cover
cling pools, particularly fixation, mineralization, immo- resembling that present before disturbance,” we do not
bilization and “leakiness.” The problem with all these know how diverse is diverse enough, how closely the
measures lies in determining what the target should be, vegetation needs to resemble the pre-disturbance vege-
in relation to the problems discussed above concerning tation, and in all likelihood do not have a clear picture
reference systems. of what the pre-disturbance vegetation was anyway.
Other more abstract possibilities may be worth pur- Any assessment process will thus produce equivocal re-
suing. For instance, the concept of entropy points to a sults. If, on the other hand, we have as a goal “to rees-
gradual decline in order in all systems over time (Miller tablish vegetation with a woodland structure of 20 trees
1971). All living entities remain “alive” by pumping out per hectare, comprising local provenance native species
disorder, i.e., maintaining themselves against thermo- which attain a height of at least 2 m within 5 years, and
dynamic gradients by taking in energy and locally re- an understory of native shrubs, forbs and herbs achiev-
ducing the production of entropy, by organizing small ing a site diversity of 25 /– 6 species,” we can then
molecules (mineral ions and gases) into large ones (or- start to measure the actual performance of the restora-
ganic molecules and DNA). Similarly, human activities tion in these terms. This goal can be set in relation to
in maintaining production systems aim to impose or- data on the pre-existing vegetation, or to the composi-
der, but frequently succeed in increasing disorder in the tion of adjacent vegetation, or can be settled on by dis-
surrounding environment. Addiscott (1995) has sug- cussion with stakeholders about what may be possible
gested that an audit of small versus large molecules (a and desirable on the site.
ratio) may be useful as a measurement of sustainability,
and hence also as a measure of restoration status. There-
Putting This into Practice
fore, in an efficient system, small molecules should per-
sist for only short periods before being reassimilated by The start of the new millennium is a good time to take
the biomass. In addition, any gaseous losses should be up a challenge. There are plenty of challenges facing

244 Restoration Ecology JUNE 2001

 Repairing Ecosystems

humanity in this new era, and here we have focused on Clewell, A., and J. P. Rieger. 1997. What practitioners need from
the particular challenges facing restoration ecologists. restoration ecologists. Restoration Ecology 5:350–354.
Harris, J. A. 1997. Certification for responsible restoration. Resto-
We present the challenge to restoration practitioners ration and Management Notes 15:5.
and scientists alike to get our act together and devise Harris, J. A., and P. Birch. 1992. Land reclamation and restoration.
and deliver effective restoration strategies and practices Pages 269–291 in J. C. Fry, G. M. Gadd, R. A. Herbert, C. W.
which can help repair the widespread ecological dam- Jones, and I. Watson-Craik, editors. Microbial control of pol-
age left to us from the last millennium. We need effec- lution. Society for General Microbiology, Symposium 48,
Cardiff, March 1992, Cambridge University Press.
tive interaction between scientific analysis, land-user Harris, J. A., P. Birch, and J. Palmer. 1996. Land restoration and
innovation and the development of principles. We need reclamation: principles and practice. Adison Wesley Long-
effective links between academics, practitioners and pol- man, Harlow.
icy makers at all levels. We need the translation of re- Higgs, E. S. 1997. What is good ecological restoration? Conserva-
search findings into action, and continuous feedback tion Biology 11:338–348.
Hobbs, R. J. 1993. Can revegetation assist in the conservation of
between users and researchers. We need to make sure biodiversity in agricultural areas? Pacific Conservation Biol-
that our actions are based on the best knowledge avail- ogy 1:29–38.
able now, and that managers have up-to-date para- Hobbs, R. J. 1999. Restoration of disturbed ecosystems. Pages
digms in their heads when they act. At the same time, 673–687 in L. Walker, editor. Ecosystems of the world. 16.
we need to ensure that researchers ask questions that Disturbed ecosystems. Elsevier, Amsterdam.
Hobbs, R. J., and S. E. Humphries. 1995. An integrated approach
are relevant to the real world. It has been argued that to the ecology and management of plant invasions. Conser-
this could form part of an on-going professional accred- vation Biology 9:761–770.
itation program (Harris 1997). Hobbs, R. J., and D. A. Norton. 1996. Towards a conceptual frame-
Restoration ecologists cannot find all the answers by work for restoration ecology. Restoration Ecology 4:93–110.
themselves. Indeed, it is not our place to answer all the Hobbs, R. J., and S. R. Morton. 1999. Moving from descriptive to
predictive ecology. Agroforestry Systems 45:43–55.
questions relating to what restoration goals should be Hobbs, R. J., and D. A. Saunders. 2001. Nature conservation in ag-
and how they should be achieved. These discussions ricultural landscapes: real progress or moving deckchairs?
need to be held more broadly within society. What we Pages 1–12 in J. Craig, N. Mitchell, and D. Saunders, editors.
can provide, however, is input to this discussion in rela- Nature conservation. 5: Nature conservation in production
tion to the ecological validity, costs and likelihood of landscapes. Surrey Beatty and Sons, Chipping Norton, New
South Wales, Australia.
success of various restoration options. Restoration ecol- Hobbs, R. J., D. A. Saunders, and G. W. Arnold. 1993. Integrated
ogy provides positive hope for the future, and hence landscape ecology: a Western Australian perspective. Bio-
restoration ecologists have a weighty responsibility to logical Conservation 64:231–238.
ensure that our science and practice live up to expecta- Hobbs, R. J., and C. J. Yates. 1997. Moving from the general to the
tions. specific: remnant management in rural Australia. Pages 131–
142 in N. Klomp and I. Lunt, editors. Frontiers in ecology:
building the links. Elsevier, Amsterdam.
LITERATURE CITED Lawton, J. H. 1996. Corncrake pie and prediction in ecology. Oi-
kos 76:3–4.
Addiscott, T. M. 1995. Entropy and sustainability. European Jour- Ludwig, J., D. Tongway, D. Freudenberger, J. Noble, and K.
nal of Soil Science 46:161–168. Hodgkinson, editors. 1997. Landscape ecology, function and
Allen, E. B., W. W. Covington, and D. A. Falk. 1997. Developing management: principles from Australia’s rangelands. CSIRO
the conceptual basis for restoration ecology. Restoration Publishing, Melbourne.
Ecology 5:275–276. McIntyre, S., and R. J. Hobbs. 1999. A framework for conceptualiz-
Aronson, J., S. Dhillion, and E. Le Floc’h. 1995. On the need to se- ing human impacts on landscapes and its relevance to man-
lect an ecosystem of reference, however imperfect: a reply to agement and research. Conservation Biology 13:1282–1292.
Pickett and Parker. Restoration Ecology 3:1–3. McIntyre, S., and R. J. Hobbs. 2001. Human impacts on land-
Aronson, J., C. Floret, E. Le Floc’h, C. Ovalle, and R. Pontanier. scapes: matrix condition and management priorities. Pages
1993a. Restoration and rehabilitation of degraded ecosys- 301–307 in J. Craig, D. A. Saunders, and N. Mitchell, editors.
tems in arid and semiarid regions. I. A view from the South. Nature Conservation 5: Nature Conservation in Production
Restoration Ecology 1:8–17. Environments. Surrey Beatty and Sons, Chipping Norton,
Aronson, J., E. Le Floc’h, C. Floret, C. Ovalle, and R. Pontanier. New South Wales, Australia.
1993b. Restoration and rehabilitation of degraded ecosys- Middleton, B. 1999. Wetland restoration: flood pulsing and dis-
tems in arid and semiarid regions. II. Case studies in Chile, turbance dynamics. John Wiley and Sons, New York.
Tunisia and Cameroon. Restoration Ecology 1:168–187. Miller, G. T. J. 1971. Energetics, kinetics and life: an ecological ap-
Aronson, J., and E. Le Floc’h. 1996 a. Hierarchies and landscape proach. Wadsworth, Belmont, California.
history: dialoging with Hobbs and Norton. Restoration Ecol- Noss, R. F. 1996. Ecosystems as conservation targets. Trends in
ogy 4:327–333. Ecology and Evolution 11:351.
Aronson, J., and E. Le Floc’h. 1996 b. Vital landscape attributes: Palmer, M. A., R. F. Ambrose, and N. L. Poff. 1997. Ecological the-
missing tools for restoration ecology. Restoration Ecology 4: ory and community restoration ecology. Restoration Ecology
377–387. 5:291–300.
Bak, P. 1997. How nature works: the science of self-organised crit- Parker, V. T. 1997. The scale of successional models and restora-
icality. Oxford University Press, Oxford, England. tion objectives. Restoration Ecology 5:301–306.

JUNE 2001 Restoration Ecology 245

Repairing Ecosystems

Parker, V. T., and S. T. A. Pickett. 1997. Restoration as an ecosys- Whisenant, S. G. 1999. Repairing Damaged Wildlands: A process-
tem process: implications of the modern ecological para- oriented, landscape-scale approach. Cambridge University
digm. Pages 17–32 in K. M. Urbanska, N. R. Webb, and P. J. Press, Cambridge.
Edwards, editors. Restoration ecology and sustainable de- Wyant, J. G., R. A. Meganck, and S. H. Ham. 1995. A planning
velopment. Cambridge University Press, Cambridge. and decision-making framework for ecological restoration.
Pfadenhauer, J. and Grootjans, A. 1999. Wetland restoration in Environmental Management 19:789–796.
Central Europe: aims and methods. Applied Vegetation Sci- Yates, C. J., R. J. Hobbs, and R. W. Bell. 1994. Landscape-scale dis-
ence 2:95–106. turbances and regeneration of semi-arid woodlands of south-
Pickett, S. T. A., and V. T. Parker. 1994. Avoiding the old pitfalls: op- western Australia. Pacific Conservation Biology 1:214–221.
portunities in a new discipline. Restoration Ecology 2:75–79. Yates, C. J., D. A. Norton, and R. J. Hobbs. 2000. Grazing effects
Rapport, D. J., R. Costanza, and A. J. McMichael. 1998. Assessing on soil and microclimate in fragmented woodlands in south-
ecosystem health. Trends in Ecology and Evolution 13:397–402. western Australia: implications for restoration. Austral Ecol-
Tongway, D. J., and J. A. Ludwig. 1996. Rehabilitation of semi- ogy 25:36–47.
arid landscapes in Australia. I. Restoring productive soil Young, T. P. 2000. Restoration ecology and conservation biology.
patches. Restoration Ecology 4:388–397. Biological Conservation 92:73–83.
Urbanska, K. M., N. R. Webb, and P. J. Edwards, editors. 1997. Zedler, J. B. 1995. Salt marsh restoration: lessons from California.
Restoration ecology and sustainable development. Cam- Pages 75–95 in J. J. Cairns, editor. Rehabilitating damaged eco-
bridge University Press, Cambridge. systems. 2nd edition. Lewis Publishers, Boca Raton, Florida.
Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. Melillo. 1997. Zedler, J. B. 1996. Ecological issues in wetland mitigation: an in-
Human domination of Earth’s ecosystems. Science 277:494–499. troduction to the forum. Ecological Applications 6:33–37.
Wheeler, B. D., S. C. Shaw, W. J. Fojt, and R. A. Robertson. 1995. Zedler, J. B., and J. C. Callaway. 1999. Tracking wetland restora-
Restoration of temperate wetlands. John Wiley and Sons, tion: do mitigation sites follow desired trajectories? Restora-
Chichester, United Kingdom. tion Ecology 17:69–73.

246 Restoration Ecology JUNE 2001