cafions of deforesfafion, such as greenhouse gas emis-

Rehabilitation of Tropical sions, loss of biodiversify, and global climafe change.

Liffle affenfion is given fo a more fundamenfal and

Lands: A Key to Sustaining immediafe problem of susfaining economic acfivify for

abouf half of fhe world's populafion. In fhis paper we
discuss fhe issue of fropical land-use changes and fheir
Development consequences, and we propose an approach fo curb
rafes of forest loss, namely the rehabilifafion of dam-
aged lands. Land rehabilifafion provides a fundamen-
fal solufion fo currenf problems of fropical foresf loss
Sandra Brown^ and a means fo slow fhe process. We focus on the
Ariel E. Lugo^ tropics but believe fhaf fhe principles apply to all
lands.
We first review the concepts and definitions applied
Abstract to the topic of land conversion and repair. We fhen
Land rehabilitation is proposed as a management discuss the fypes of human activities that produce
strategy to reverse the negative consequences of tropi- damaged or degraded lands in the tropics, discuss fhe
cal deforestation and land degradation. We first de- ecological and socioeconomic sfrafegies for rehabilifat-
fine the concepts associated with ecosystem modifica- ing fhese lands, provide some case studies that are
tjon—conversion, damage, and degradation—and examples of successful rehabilifafions, and conclude
those associated with ecosystem repair—restoration, wifh a discussion of fhe basic requirements for success
rehabilitation, and reclamation. We then present a in any rehabilitation project.
scheme of sustainable land use in the tropics, with
illustrations of how rehabilitation and restoration ac-
Concepts and Definitions
tivities fit into the overall scheme of the use of land.
Because damaged lands cannot contribute effectively The topic of land degradafion and rehabilifafion is
to sustained economic development, land rehabilita- fraughf wifh many concepfs and definifions. Here we
tion is a necessary step for increasing the chances of define most of fhe terms we use in fhis paper based on
attaining sustainability. Approaches for rehabilitat- our own understanding and thaf of Bradshaw (1987),
ing ecosystems are discussed, including the manage- Lamb (1988), Lugo (1988, 1991), Berger (1990), Schreck-
ment of stressors and subsidies in relation to their enberg et al. (1990) and Barrow (1991). The use of fhese
point of interaction in the ecosystem. Finally, we il- definifions and concepfs is illustrated in Figure 1.
lustrate the concepts of ecosystem rehabilitation of Lands damaged by nafural phenomena or human ac-
damaged, degraded, and derelict lands with examples fivifies and fhaf can be repaired by nafural succession
of case studies from dry to humid life zones in island fo fheir original state (boxes and arrows above the line
and continental situations throughout the tropics. The in Figure 1) will be excluded from furfher discussion in
case studies demonstrate that opportunities for suc- fhis paper.
cess exist, even with severely degraded lands, but a The arrows on fhe leff-hand side of Figure 1 repre-
considerable amount of research remains to be done senf fhe ferms we use fo describe fhe changes lands
before we have a full understanding of the complexity undergo generally by human acfivifies. Conversion is
of the task facing us. a more generic ferm fhaf represenfs fhe whole spec-
frum of changes. We view conversion as any modifica-
tion of a mafure nafural ecosystem, from as simple as
Introduction removing bark or branches from frees fo as severe as
complefe deforesfafion and mining of fhe subsfrafe
le increasing worldwide affenfion given fo fhe on which a foresf once sfood. Damage, on fhe ofher
ropical biome usually focuses on fhe global impli- hand, is a more resfricfive ferm fhaf we apply fo con-
verted lands incapable of normal recovery fo original
condifions (mafure ecosysfems) because one or more
1 Department of Forestry, University of Illinois, W-503 of fheir key affribufes (soils, biofa, land morphology,
Turner Hall, 1102 S. Goodwin, Urbana, IL 61801, U.S.A. hydrology, efc.) have been modified. However, dam-
^Infernational Institute of Tropical Forestry, USDA Forest aged lands sfill have fhe capabilify fo produce goods
Service, Call Box 25000, Rio Piedras, Puerto Rico 00928, and services fo safisfy human needs. Lands are de-
U.S.A. scribed as degraded when fheir edaphic conditions and/
© 1994 Society for Ecological Restoration
or biofic richness have been reduced by human aefivity

JUNE 1994 Restoration Ecology Vol. 2 No. 2, pp. 97-111 97

Rehabilitation of Tropical Lands

Types of ecosyslems cess offen becomes arresfed, however, because of se-

vere limifafions in the site or the biota. Rehabilitation is
the return of any converfed ecosysfem, damaged or
degraded, fo a fully functional ecosysfem, irrespective
No human of ifs original or desired final state. The term restoration
Secondary forests/
(oresl fallows
intervention is used when any converfed ecosysfem is refurned fo
required
ifs presumed or relafively original indigenous sfafe.
Time is an imporfanf facfor in the use of fhe ferms
reclamafion and rehabilifafion because some mighf ar-
Fulfy funcfional
ecosystem gue fhaf given enough fime (cenfuries fo millenia), all
damaged and degraded lands may be restored to their
Requires
iiuman
indigenous sfafe. In an ever-changing world, how-
intervention ever, fhis amounf of time is nof likely fo be available,
nor is if likely fhaf condifions of the past will be re-
stored in the fufure. Insfead, we use fhese ferms in
relafion fo shorfer fime periods on fhe order of a hu-
man generafion, which we inferpref fo mean abouf 100
Arrested succession years.

Productivity and Sustainabie and Unsustainabie Tropical Land Uses

biota restored
The highesf rafes of foresf clearing during fhe past
several decades have occurred in the tropics. In the
late 1970s, abouf 11.3 million ha yr"^ of fropical foresfs
Dereiict fands
were deforesfed and converfed fo nonforesfed lands
(Lanly 1982). During fhe period 1980-1990, fhe rate
was 15.4 million ha yr"' (Singh 1993). In addition to
Figure 1. An illustration of the definifions and common outright clearing, more than 4 million ha yr ' of ma-
terms used to discuss the changes lands undergo, gener-
fure foresfs were selecfively logged during fhe 1970s
ally caused by human activities, and their modes of repair
by human-directed succession. (Lanly 1982). As wood producfion increased during
fhe 1980s (Food and Agriculfure Organizafion 1980-
1990), so mosf likely has the area of logged forests.
These are the changes in fropical land use whose mag-
fo such a degree that their ability to satisfy parficular nifude we know something abouf; fhere are ofher
uses has declined. In shorf, degradafion generally oc- changes occurring, such as fhe formafion of damaged
curs when any of fhe ecosysfem's sforages (sfafe vari- and degraded foresf and agriculfural lands, fhe magni-
ables such as soil organic maffer, soil nufrienfs, seed fude of which we know liffle abouf (Grainger 1988;
pools, and biomass) have been reduced fo fhe poinf Brown ef al. 1991; Richards & Flinf 1994). For example,
that nafural inpufs cannof replenish fhem fo fheir orig- a major reason for deforestation in fhe tropics is to
inal state. Lands can also be degraded by contamina- replace agricultural land that has become degraded
tion wifh foxic chemicals fhaf can render fhe ecosystem and unproductive for growing crops (Brown 1993; Dale
useless fo people. The poinf af which lands go from et al. 1993; National Research Council 1993). Much of
being damaged fo degraded is not always clear and the land currently being cleared of foresfs is replacing
would fend fo be sife-specific. In an advanced state, degraded agriculfural land, so fhaf fhe increase in agri-
degraded lands are analogous to derelict lands, or culfural land does nof equal fhe decrease in foresf
lands thaf are of very limifed use to people or develop- land.
menf programs. To undersfand fhe importance of land rehabilifafion,
On fhe righf-hand side of Figure 1 are fhe definifions it is first necessary fo undersfand fhe processes fhaf
of the terms commonly used to describe the modes of lead to degraded and damaged lands and to do so in
ecosystem repair by human-directed succession. By fhe confexf of susfainable land use in the tropical bi-
succession we mean all fhe changes fhaf occur in eco- ome. Land conversion need not be defrimenfal fo for-
systems over time (Odum 1969). Reclamation is the pro- ests or to people, if done properly (Figure 2). In fhe
cess by which derelict or very degraded lands are re- pasf, under low populafion pressure and in fhe ab-
turned fo producfivify and by which some measure of sence of potent technological tools, people extracted
biofic function and productivity is restored. This pro- forest products from complex fropical foresfs or con-

Restoration Ecology JUNE 1994

 Rehabilitation of Tropical Lands

Available for re-use Under mosf of fhe condifions jusf described, if is

(8) possible fo converf a foresfed landscape fo a landscape
mosaic composed of mafure foresfs, logged foresfs,
foresf fallows, shiffing agriculfural plots, and perma-
nent agricultural lands (Brown & Lugo 1990). Humans
derive multiple benefifs, bofh goods and services,
from each of fhese land uses, as well as from fhe land-
use conversion process. However, yields, services,
and cosfs fo humans differ according fo fhe fype of
land use. The use of such a landscape can be sustain-
able because natural successional processes are gener-
ally sufficient fo repair any shorf-ferm damage caused
by human acfivify. There is little or no loss of natural
productivity, and liffle fo no human infervenfion is
needed fo repair fhe landscape.
A cycle of damage fo lands begins wifh overuse of
sifes. Excessive exfraction of maferials involving modi-
ficafion of fopography, compacfion of soils, unchecked
erosion, confinuous cropping wifhouf fallow periods
or crop rofafions, and oufrighf removal of vegefarion
cover are examples of human acfivifies fhat result in
damage and degradafion to the land (pafhways #3,
Figure 2. Conversion of tropical lands showing pathways Figure 2). For example, machine-clearing of foresfs is
that lead to (1) sustainable land uses; (2) damaged, de- more damaging to sites than clearing by hand because
graded or derelict lands; (3) arrested succession; and (4) machines remove or heavily compact soil, thus greatly
repaired lands. See text for further description. refarding the regeneration cycle (Uhl ef al. 1988). Im-
proper or excessive use of poisonous chemicals can
verfed fhem fo femporary agriculfural fields to pro- also damage a sife and make if unsuifable for furfher
duce food and fiber without degrading sifes irrevers- use. In general, fhe greater the intensity of use of a
ibly (pafhways # 1 , Figure 2). Nafural succession site, the greater fhe probabilify fhaf it can be damaged
occurred affer abandonmenf, and fhese sifes reverfed and fhaf precaufions will need fo be faken.
to foresfs (pafhways #2, Figure 2). Af infermediafe
stages of succession, secondary foresfs were produced
that provided many valuable products and services for Table 1. Values of secondary forests for human use.*
humans (Table 1). Provide fruits, medicinal plants, construction materials,
Technological innovafion and improved manage- and animal browse.
menf skills allowed people fo infensify agriculfural ac- Produce valuable timber species because they are often
dominated by a few species.
tivities in certain localities where soil, climafe, and Produce a uniform raw material with respect to wood
moisfure condifions were favorable for confinuous density and species.
use. Vasf areas of fropical lands have been used fhis Wood of secondary species tends to be low in resins,
way for cenfuries, and fhey are sfill producfive areas, waxes, etc., which facilitate their use.
such as fhe rice-growing regions of fropical Asia (Na- Produce biomass at a fast rate.
tional Research Council 1993; Richards & Flinf 1994). Relatively easy to regenerate.
Species tend to have properties that foresters seek as suit-
Logging of foresfs also infensified as global demand able for plantations.
increased and new fechnologies for exfracfing and pro- Have an advantage as a raw material source because they
cessing fimber were developed (Whifmore 1984; are generally more accessible to markets than primary
Tucker 1992). Commercial logging can be a sustainable forests.
activity without long-term damage fo foresfs, if prop- Serve as foster ecosystems for late-secondary to mature
forest species.
erly managed (Uhl & Vieira 1989; Uhl ef al. 1991; Johns Serve as useful templates for designing agroecosystems.
1992). Buf few foresfs in fhe fropics are currently man- Restore site productivity and reduce pest populations.
aged properly (Schmidt 1987; Johnson & Cabarle 1993), May support higher animal production and serve as pro-
with the resulf fhaf many of fhese lands become dam- ductive hunting grounds.
aged or are converfed fo nonforesf uses (as in en- Presence of a greater number of vertebrates may enhance
ecotourism.
croaching nonsusfainable agriculfure) and are in need
of rehabilifation. 'Modified from Brown and Lugo 1990.

JUNE 1994 Restoration Ecology 99

Rehabilitation of Tropical Lands

Sometimes normal successional processes are insuffi- Table 2. Examples of causes of damage and degradation to
cient or too slow (occuring over more than several dec- tropical forest lands.*
ades) to rehabilitate lands fo producfive sysfems, or Cause Type of Stressor
fhey are arresfed at an undesirable end-point; in fhese
cases, infervention by humans is needed. Arresfed Site alteration allowing unplanned inva- 4-5
succession and lands degraded wifh overuse and poor sions of exotics and pests
Continuous use without proper fallow or 4-5
managemenf require repair via rehabilifative actions crop rotation
(pathways, #4, Figure 2) or restoration (pathway #5). Excessive harvesting of any comparfment. 3-5
Intensive damage results in degradation towards a including plant or animal populations
derelict state (pathway #6), and land reclamation is Overgrazing 3-5
needed fo reverse fhis sfafe (pafhway #7). Rehabili- Overfertilization 3-5
Introduction of fires or changing natural fire 3-5
fafed lands resulf in alfernafive foresf ecosysfems, frequencies
which can then be reused to serve human needs (path- Contamination with poisonous/radioactive 2-5
way #8).
Inappropriate agricultural techniques 2-5
Changes to hydroiogic cycle 1
Causes of Damage to Ecosystems Soil salinization 1-3
Soil compaction and erosidn 1-4
Before discussing the specific causes of damage to Mining of soils or geologic substrate 1-5
tropical lands, we first provide a conceptual model fhaf Consequences of poor resource analysis/ 1-5
illustrates how stressors impact ecosystems, and fhaf planning
Poor land-use policy 1-5
shows how fhe severity of impact depends on where Use of inappropriate technology 1-5
in the ecosystem a stressor acts (Lugo 1978). A simpli- War 1-5
fied model of an ecosystem with energy inputs (sun-
light, rain, wind), major comparfmenfs (planfs, con- *The type of stressor is keyed to the numbers in Figure 3.
sumers, and soil organic maffer, nutrients, and
associated decomposers), and funcfions (primary pro-
ducfion, respirafion, and nufrienf cycling) is shown in materials before they can be fully used by the system
Figure 3. Five fypes of ecosystem stressors are illus- or interfere with physiological processes (Lugo 1978,
trated by the encircled numbers; for consistency, we 1988). Sfressors fhaf harvesf planfs or consumers
refain the numbering system used originally by Lugo (fypes 3-5, Figure 3) generally have less impacf be-
(1978, 1988). Stressors of types 1-2 (higher-impact) cause they do not affecf fhe sysfem's ability to process
cause more damage than those of fypes 3-5 (lower- energy and renew itself. On a long-ferm basis, how-
impacf) because fhey remove low-qualify energy and ever, even fhe lower-impacf sfressors can damage and
degrade fhe ecosysfem.
The causes of damage fo land are many, and fhey
resulf from complex inferacfions of cultural and eco-
logic factors (Bainbridge 1990). Examples of specific
causes and how fhey affect the ecosysfem are pre-
senfed in Table 2. Generally, fhese causes affecf fhe
land's physical characferisfics, nutrient status, foxicify,
and sfrucfure of the biofic communify (Bradshaw
1989). Many of fhe causes acf in unison or synergisfi-
cally, making fhe rehabilifafion of such lands an ex-
pensive underfaking. For example, (1) poor land policy
may lead fo use of inappropriafe agriculfural fech-
niques fhat result in unchecked soil erosion, and (2)
over-harvesting of frees may change fhe hydroiogic
cycle, causing fhe wafer fable fo rise and soils fo be-
o
Figure 3. Simplified model of an ecosystem with energy
come salinized. Furthermore, although several causes
affecf fhe same parf of fhe ecosysfem (Table 2), fheir
inputs, major compartments, and flows showing where net effect may differ. Infensity, frequency, seasonalify,
sfressors (numbers in circles) act on an ecosystem to cause and areal exfenf of fhe sfressor are also crifical in defer-
damage. Stressors 3-5 cause the least damage, whereas mining fhe amount of damage.
stressors 1-2 do the most damage (based on Lugo 1978, The degree of damage to forest lands can also range
1988; see text for further explanation). widely depending on the causal human action. A sub-

100 Rcstoration Ecology JUNE 1994

 Rehabilitation of Tropical Lands

FORESTED LANDS

High

Low
Low

Low
atives

ssion
Mature stands
S Damaged secondary forests

900
CO
0)
Biomass reduction
o tn -wood/bark
o -litter
E
DC Z -understory
-peat
CO
to Species impoverishment
r—
a> * Frequently burned
O
8 Exotic/pathogen-invaded
ra
cu CO
labiiitative res

ehabilitat:ioiT actioiIS requ

NON-FORESTED LANDS
rehab ilitative Frequently burned

iabilitation
selectiion
Overgrazed

'dama
repair
Eroded
Nutrient depleted soils
£ o o "o o Waterlogged soils
robab ility

.2 o "o OJ
•D
d)
Salinized soils
\leed'

Deg
d.
Cost

CO o NON-VEGETATED LANDS
Highly eroded slopes
CL tr Poisoned lands
-chemical spills
-industrial wastes
aID gO air pollution
O) *~
War ravaged
Arti ficial

Strip-mined lands
Exo tics

j= x: Derelict lands
enei
Cha

Hig
Hig

.31
Figure 4. Subjective ranking of tropical lands according to of damage based on intensity of the stressor and natural
the degree of damage or degradation by human activities, environmental factors (clitnate and soils, for example). (See
from least damaged (relatively mature) to most degraded Fig. 5 for types of actions needed).
(derelict). Within each state there are gradients or degrees

jective ranking of fropical lands according fo degree of degree of damage or land degradafion are needed fo
damage is illusfrafed in Figure 4. Lands are grouped facilifafe research, mifigafion, and collecfion of sfafisfi-
into three categories (forested, nonforested, and non- cal dafa. A criferion based on fhe fype and magnifude
vegetated), with lands subject to increasing degrees of of sfressor fhaf causes fhe damage (Figure 3) is, we
damage ranked under fhese fhree cafegories. We as- believe, a step in the right direcfion.
sume fhat each land sfafe is subject to the same envi-
ronmenfal condifions as fhe mafure sysfem. Therefore,
The Role of Rehabilitation
fhe listing generally reflecfs progressive damage or
degradafion of the original system. Rehabilifation is a special form of succession characfer-
Trends in mode of repair, types, costs, and probabil- ized by human intervention and expenditure of energy
ity of success of rehabilifafion are shown by fhe arrows (Figure 5). Rehabilifafion of damaged or degraded eco-
on the left-hand side of the diagram. Damaged but sysfems involves a reversal of fhe acfions in Figure 3,
still-forested lands are subject fo lower-impacf sfres- shown by a reversal of fhe numbering system. If con-
sors (fypes 3-5 at low to moderate intensity. Figure 3). sisfs of five types of acfions whose cosfs increase as fhe
Lands in fhis cafegory can be rehabilifafed wifh nafive numbers in Figure 5 become higher. The simplest and
species af low cosf and wifh a high probabilify of suc- cheapest action is (1) fo remove or confrol fhe sfressors
cess. Nonforesfed lands resulf when sfressors of fype 2 acfing on fhe sysfem, such as high fire frequency,
and high-infensify 3 are added. Rehabilifafing fhese over-grazing, or biomass removal (poinfs 3-5, Figure
lands requires a mix of nafive and exofic species and 3). A second acf ion is (2) fo add species (planfs or
some human infervenfion; cosfs will be higher and fhe animals) or maferials (ferfilizer, organic residues, or
probabilify of success lower fhan in rehabilifafion of wafer). More expense and efforf are involved in regu-
lands sfill foresfed. When sfressors of fype 1 are infro- lafing fhe speed of ecosysfem processes (3), such as
duced, lands become mosf degraded. The cosf fo reha- reversing soil compacfion or managing fhe quanfity
bilitate or reclaim fhese lands is the highest, has the and quality of soil organic maffer inpufs fo confrol syn-
lowest probability of success, and requires fhe mosf chronizafion of nutrient release and plant uptake. The
research and technological efforf. Objecfive criferia of most difficulf rehabilifative strategies (poinfs 4 and 5 in

JUNE 1994 Restoration Ecology 101

Rehabilitation of Tropical Lands

fits: (1) it converts unproductive lands fo self-perpefu-

afing ecosysfems; (2) if prevenfs furfher damage fo
downsfream ecosysfems; (3) it reverses a worldwide
negative trend of land degradation; and (4) if removes
pressure on undisfurbed lands and thus contributes to
a reduction in furfher deforesfation. Beyond these ben-
efits, land rehabilitation helps people manage change
and correct or adjust the path of developmenf so that it
remains on a positive track relative fo both people and
the environment.

Chalienges in Tropical Forest Rehabilitation

Figure 5. Simplified model of an ecosystem (see legend to The literature in tropical forest rehabilitation is sparse,
Figure 3) that illustrates the actions (numbers in circles) and it provides more ideas and suggestions than actual
needed to rehabilitate damaged, degraded, or derelict examples of successful l-ehabilitafions of complex for-
lands. The increasing numbers reflect the increasing cost esf ecosystems (see Bradshaw 1987; Lamb 1988; Berger
of rehabilitation (based on Lugo 1988; see text for further 1990; Schreckenberg et al. 1990; Barrow 1991). The rea-
explanation).
sons for fhis lack of success are many. For example, it
is only recently that the issue of rehabilifafing fropical
forests became a goal of managers. Furthermore, fropi-
Figure 5) involve either removing fhe higher-impact cal foresfs are complex ecosystems; we undersfand
stressors (points 1-2, Figure 3) or regulating fhe en- fheir structure and function insufficienfly fo be able to
ergy inpufs info a sysfem to control fhe condifions for puf one together starting with an array of isolafed
ecosysfem developmenf. Because higher-impact stres- parts. In addition, because most fropical foresfs are
sors are the most damaging to an ecosysfem, high already in some less-than-mature state (Brown & Lugo
costs are generally incurred just to remove them. 1990; Wood 1993), the target we are frying fo reach is
Changing the hydrology of a foresfed wefland uncerfain. Such a primifive state of knowledge is not
(through reversing draining or damming) or modify- exclusive to tropical forests. A critical examination of
ing topography on hillsides (by terracing to control the rehabilitation and resforafion liferature reveals that
erosion) are examples of fhese sfrategies. most successes have been with simple, highly-stressed
Rehabilifation has several drawbacks: (1) it is a lasf- ecosysfems such as mangroves, marshes, and sand
difch effort to reverse a negative sifuafion caused by dunes, or on derelicf lands where the establishment of
poor managemenf of nafural resources; (2) if has a high any vegetation is considered a success (Lugo 1988).
energy and economic cosf—if is not free like nafural Highly-stressed ecosystems are relatively easy fo reha-
ecosysfem development; (3) it demands knowledge bilifafe because of their low sfrucfural complexify and
and understanding of ecosysfem repair processes, so relafively hardy biofa. Moreover, these systems usu-
research is needed; and (4) fhe resulfs of rehabilifafion ally respond rapidly to changes in environmental con-
are unpredicfable because fhe ecosysfem can recover ditions. For example, prairies and wetlands respond
in a mulfiplicify of ways, and if is offen impossible fo rapidly fo fire and fo changes in its frequency or inten-
direcf fhe rehabilifation to a specific final sfafe (Fig- sify.
ure 2). In confrasf, complex foresfs fax our abilify and
The costs of careless use of fhe land are very high knowledge because fheir successful rehabilifafion re-
and are offen paid back in ferms of fime, fhus generaf- quires fhe building of highly diverse sfrucfures and
ing a "fime fax." The fime tax is the time thaf society interactions. Furthermore, it is easy for fhe sysfem fo
must spend waiting for a resource fo heal, during evolve in directions fhaf are fofally unanficipafed by
which fhe resource cannof be used and musf be fhe manager. Therefore, fhe preferred sfrafegy of reha-
nursed. This implies cosfs in energy and resources bilifafing complex ecosysfems involves whaf Lugo
without immediafe refurn on invesfment. The return is (1988) calls "rolling with the punch," in fhe sense fhaf
the rehabilitation itself. The fime fax is proportional fo the manager provides a set of condifions for fhe devel-
the initial damage inflicted upon the resource. The opmenf of fhe ecosysfem, allows natural processes to
fime fax—like income fax, deafh, and enfropy—is un- do most of fhe work, and is willing fo accepf what the
avoidable and must be paid! natural succession offers. Humans become directors or
Land rehabilifafion has at least four immediate bene- care-fakers of fhe nafural process of succession, nudg-

102 Restoration Ecology JUNE 1994

 Rehabilitation of Tropical Lands

ing if one way or anofher af crifical fimes in fhe pro- Table 3. Ecological requirements for rehabilitating damaged
and degraded tropical forest lands.*
cess.
Despife fhe lack of know-how for rehabilifafing General Strategies
damaged and degraded lands, some general sfrafegies • Maintain flexibility in the approach.
for improving chances of success have been proposed. • Be alert to local environmental conditions.
These can be divided info ecological and so- • Avoid specificify of the ultimate goal.
cioeconomical sfrafegies (Lamb 1988; Lugo 1988); sfraf- • Couple systems that maximize their value and accelerate
rehabilitation (for example, use treated sewage effluent
egies from both groups must be considered for any for irrigafion when possible fo qrcelerafe planf growfh).
rehabilitative project fo be successful. • Create nuclei of biotic activity.
Fundamenfal ecological requiremenfs (Table 3) are • Maximize ecosystem complexity to optimize use of site
fo improve fhe sfrucfure, fauna, organic maffer, and resources.
ferfilify of fhe soil, and fo assure an ample supply of • Maximize protection from pests, fires, etc., and minimize
risks.
genefic maferial (Lugo 1988; Bainbridge 1990; Har- • Use stressors to arrest succession at desirable end-point.
rington & Howell 1990; Lavelle et al. 1993). Moreover,
Strategies Targeted at the Soil
fhe selecfion of any suife of ecological sfrafegies musf
' Manipulate the soil as needed, either physically or bioti-
consider the climatic, edaphic, and geomorphologic cally (for example rip the soil, introduce soil fauna, my-
factors fhat characterize a given site because they regu- corrhizae, etc.).
late the speed of recovery (fhe fime fax) as well as • Fertilize and irrigate when absolutely necessary (for ex-
defermine fhe amounf of human infervenfion needed. ample, on derelict lands or in arid areas).
For example, fhe use of fallows may be appropriate in • Keep topsoil moist, cool, and shaded.
humid climates where plants offen re-esfablish • Be aware that nutrient cycling strategies may change
during the rehabilitation process.
quickly, assuming fhaf sufficienf seed sources are
available fo fhe area fo be rehabilifated. Under very Strategies Targeted at the Flora and Fauna
arid climates, however, planfing and irrigafion may be • Maximize vegetation cover.
needed fo ensure fhaf vegefafion cover is restored. • Restore tree cover.
• Manipulate the existing vegetation before attempting
Some of the strategies in Table 3 may be controversial, substitution.
particularly the careful use of exotic species. This is • Use multiple seeding techniques when in doubt as to
recommended because many exofic species are gener- what to plant.
ally adapfed fo poor soil condifions, and fhey have • Let natural selection decide the best species combination
been shown fo fosfer nafive species (Lugo 1992fl; Par- for a site.
• Develop species mixtures based on their ecological com-
rotta 1992, 1993; Lugo et al. 1993). bining ability.
Although rehabilitation is often thought to be an ec- • Use fallows to do most of the forest rehabilitation.
ological problem, socioeconomic impediments also ex- • Use exotic species to foster native species.
ist, which may be more important fhan fhe ecological •Based on Lugo 1988.
obstacles (Lamb 1988). There are a variety of reasons
why social and economic facfors are so infracfable; as a
result, no universal solutions exist. Based on a variety
of case studies, however, the most important socioeco- ecfs, as well as fo femper expectations from any reha-
nomic requirements appear to be stable land-use paf- bilitation program.
ferns, equifable land-fenure sysfems, homogeneous
human populafion (wifh respecf fo efhnicify, eco- Rehabilitation Dominated by Removal of Lower-Impact Stres-
nomics, and so forfh), local public involvemenf, sfrong sors. The best examples of fropical foresf rehabilifa-
local leadership (fo enforce collecfive rules), and parfic- fion are fhose where fhe main human infervention is
ipation by governmental insfifufions, parficularly for the removal of fhe lower-impacf sfressors (sfrategy 1,
large projects (Lamb 1988; Karki 1991). Figure 5). In this situation, ecosystems repair them-
selves mostly by natural succession affer fhe stressors
have been removed. Buf fhese rehabilifafions are nof
Examples of Rehabilitation Projects enfirely nafural because fhey offen involve exofic spe-
The examples of rehabilitation projects that we discuss cies (producfs of pasf human acfivify) or nafive species
below are generally site-specific in response fo local seiecfed and favored by people fo fhe exclusion of
environmenfal condifions and socioeconomic needs. ofher nafive species. The final producfs are ecosysfems
They are presenfed not only to illustrate the diversity wifh differenf species composifion and producfivify
of problems and solutions in fhe field of rehabilifafion than those existing prior to human activity.
and resforation ecology, but also to serve as a basis on Guanica Forest, Puerto Rico. For 123 years, people lived
which to build and provide guidance for future proj- inside the Guanica Forest, now a Man and the Bio-

jUNE 1994 Restoration Ecology

Rehabilitation of Tropical Lands

sphere (MAB) Reserve. This MAB Reserve encom-

passes a dry forest fhaf has been exfensively studied
by ecoiogists for many years (Murphy & Lugo 1986).
Human activities involved wood and fuelwood collec-
tion, grazing animals, farming, and construction of
houses, schools, and recreational areas (stressor types
3-5, Figure 3). No machines were used, and deforesfa-
tion occurred by manual means. Studies are docu-
menting what happened to the foresf lands used infen-
sively by people for 128 years and lafer abandoned for
42-52 years.
The resulfs fo dafe show fhaf: (1) previous land use
made a difference in fhe direcfion of recovery (Figure
6); (2) fhe most intensively used lands (the recreation
areas) recovered fhe leasf, while less infensively used
foresf sfands (fhose used for fuelwood collecfion) had
fhe most advanced recovery; (3) fhe harsh sife condi-
fions (low wafer availabilify) seiecfed for adapfed na-
tive species rather fhan exofics, which in Guanica were
more common in pioneer sfages of developmenf; and
(4) human preference for cerfain free species was very
imporfanf in defermining biomass, basal area, and
wood volume of fhe foresf because those trees favored
by people became the largest frees in fhe recovering
areas (Figure 6b). For example, mahogany trees
{Swietenia mahogani) on abandoned farmland had an
importance value of only 6.3% buf confributed abouf
15% of fhe sfand's basal area.
Human use of fhe Guanica forest was based on fhe
slow consumption of fhe long-ferm sfores of fhe foresf,
namely its rich soil organic horizons and high under-
ground biomass, which supported rapid resproufing.
This use slowly depleted site fertility, although if did
so over a long fime (128 years). Half a cenfury later, the
sifes are still in recovery phases in spite of nof being
sfressed any furfher. This example underscores the
time elemenf in rehabilifafion. Time fo complefe recov-
ery can be very long, depending upon fwo indepen-
dently operating facfors: fhe exfent of damage and the
intrinsic capacify of a site to recover.
Guanacaste National Park, Costa Rica. The goal of fhis
project is fo rehabilitate degraded farmland in fhe dry
forest zone of Gosta Rica to create a new Guanacaste
National Park (Janzen 1986, 1988). The lands have
been subject to burning, overgrazing, over-cutting, or
continuous farming over more fhan 400 years (sfressor
fypes 3-5, Figure 3). Much of fhe area is now grassland
Figure 6. {a) Foresf recovery on both sides of the main buf sfill subjecf fo frequenf fires. Buf pafches of near
street (still used as a forest trail) of a small town in the original vegefafion fhaf confain original planf and ani-
Guanica forest after about 50 years of abandonment. Here,
the rehabilitation strategy was to remove the stressors mal species sfill exisf (Janzen 1986).
(sfrategy 1, Figure 5). {b) Lands used for collection of wood Several steps (strategies 1-2, Figure 5) are involved:
producfs had the mosf advanced recovery because many (1) excluding wildfires, anthropogenic in origin, by
species, such as Pisonia albida, are capable of coppicing; constructing fire lanes or confroUed burns fo reduce
fherefore, a closed canopy quickly forms (photos by A. E. fuel loads; (2) helping fosfer recolonizafion by woody
Lugo). planfs eifher fhrough exclusion of fire and grazing or

104 Restoration Ecology JUNE I99<

 Rehabilitation of Tropical Lands

by confroUed grazing; and (3) sfopping hunting. Socio- region. Three kinds of managemenf practices for the
economic strategies are also used, including acquiring pastures were used: (1) light use (20% of the region),
ownership of fhe land and making efforfs fo ensure involving cutting and burning the foresf, seeding wifh
fhaf the local populafion views fhe projecf as being pasfure grasses, no weeding, light to no grazing, and
economically and culfurally beneficial fo fhem. early abandonment; (2) moderate use (70% of fhe re-
The presence of remnanf foresf pafches helps wifh gion), involving fhe same inifial sife pracfices as (1)
revegetafion because many seeds are dispersed by and also including weeding by machefe, additional
wind info the grassland areas. But dispersal distances burning at 1-3 year intervals, moderate grazing, and
are relafively shorf, and dispersal over larger disfances abandonmenf affer 6-12 years; and (3) heavy use (10%
may have fo be done by hand-seeding or planfing of of the region), involving fhe same sife pracfices as (2),
seedlings info groves or sfrips (sfrafegy 2, Figure 5). followed evenfually by mechanical clearing wifh bull-
Trees with seeds thaf are dispersed by animals, parfic- dozers, soil disking and levelling, replanfing of pas-
ularly birds, produce a more diverse inpuf info fhe fure grasses, mechanical mowing for weed confrol sev-
degraded areas, depending upon fhe behavior of fhe eral times a year, moderate grazing, and abandonment
disperser. Trees or rock oufcrops fhaf provide perches affer 3-6 years (Uhl ef al. 1988). On lands subjecf fo
are imporfanf here. managemenf pracfices (1) and (2), foresfs generally re-
Efforts to educate the local populafion abouf fhe eco- generafed well fhrough nafural succession affer fhey
nomic and cultural benefifs of fhe projecf fo fhem- were abandoned (sfrategy 1, Figure 5) and required
selves include explaining fhe nafure and value of fhe little additional infervenfion (Uhl ef al. 1988; Nepsfad
resource as well as offering employmenf. Those em- ef al. 1991).
ployed nof only help prevenf fires and poaching buf On abandoned pasfure lands subjecf to heavy use
also planf seeds. (stressors 3-5 over large areal extent. Figure 3), natural
regeneration is uncertain because of many ecological
Rehabilitation by Tree Plantings. Somefimes, fhe cycle of barriers fo planf esfablishmenf, survival, and growfh
conversion, over-use, and abandonment of land (Nepsfad ef al. 1991). These required more human in-
results in such damage to foresfs fhat rehabilitation fervenfion. The main ecological barriers fo tree inva-
strategies such as fhose described in fhe previous ex- sion that need fo be overcome fo rehabilitate these
amples are impossible. Succession is usually arrested lands are low propagule availability, seed and seedling
in some undesirable condifion, or fhe land simply con- predation, seasonal drought, and root competition
finues to erode and lose fertility. In fhese cases, reha- with vegetafion of fhe abandoned land. Any frees fhaf
bilitation requires a more intensive human effort (strat- do invade fhese sifes are resfricfed fo those f^ew species
egy 2-3, Figure 5). that can overcome all these barriers. These trees are
Forests can be rehabilitated, but only fhrough active imporfanf, however, as fhey provide nuclei for furfher
planting programs that start wifh a selecfion of free regenerafion of new species—for example, by provid-
species known fo be able fo survive condifions in fhe ing perches for seed-carrying mammals, shading ouf
damaged sife. Offen fhese are exotic species adapted compefing grasses, changing fhe microclimafe, and, as
to harsh condifions. The ecological basis for fhe use of fhe area grows larger, reducing fhe fhreaf of damaging
free planfings or planfafions for rehabilifafion of native fires. In fhis way new free islands form (Nepstad et al.
foresfs in damaged fropical lands has been described 1991).
by Lugo (1992a, 1992^), Lugo ef al. (1993), and Parroffa Alfhough fhe potential for nafural regeneration ex-
(1992, 1993). Tree planfafions are nufrienf and organic- isfs, fhe process is slow, and alfernafives were sought.
mafter sinks, a condifion fhat leads fo fhe re-establish- To facilitate forest recovery, it will be necessary fo cata-
ment of soil fertility and soil organic matter and of lyze natural processes of tree-island formafion by
forest conditions within which native species can pros- planfing trees, parficularly as parf of an agroforesfry
per (Lugo 1992b). As long as the planted species are sysfem. Uhl (1988) considered fhe following fech-
carefully matched to site condifions, bofh exotic and niques for rehabilifafing fhese degraded pasfures: hu-
native species can be used. When conditions are equal, man dispersal of seeds, fransporf of foresf soils to the
native species should be used, but exotic species musf sifes, use of new types of seed vecfors, and free planf-
be used when fhere are no nafive ones fo do fhe job. ing (sfrafegies 2 and 3, Figure 5).
Degraded pastures in the Amazon. During fhe lasf 20 Degraded lands in southern China. A hisfory of more fhan
years or so, large areas of fhe Brazilian Amazon foresfs 100 years of poor land use (sfressors of type 3-5 over
have been converfed fo pasfures (Uhl ef al. 1988; Nep- long duration. Figure 3) in western Guangdong Prov-
stad et al. 1991). These pastures were generally pro- ince (Xiaoliang, Dianbai Gounfy) produced large areas
ductive for 4-8 years before they were abandoned, re- of highly degraded, pracfically derelicf land (Figure
sulting in large areas of degraded pastures in the 7a): barren, eroded, unproductive, wifh fhe remaining

JUNE 1994 Restoration Ecology 105

Rehabilitation of Tropical Lands

agricultural areas threatened by floods and lack of irri-

gafion wafer. Two projects of note have been imple- j
menfed in fhe region to return these lands to produc-
tive uses.
The firsf one, esfablished in 1959, planted fhe se- ;
verely degraded lands wifh eucalypfs. Acacia confusa,
and fhe nafive Pinus massoniana, species fhaf do well
on poor soils (strategy 2, Figure 5) (Yu & Pi 1985; Yu &
Wang 1989). After 14 years or so, this foresf was infer-
planfed wifh several native broadleaf free species, fruit
frees, and undersfory planfs of economic value (mostly
medicinal) fo form a mulfi-sforied, mixed forest (Figure
7b). Over about a 23-year period, this managed succes-
sional pattern improved the soil fhrough increases in ,
soil organic maffer (from 0.6% fo 1.7%), pH (from 4.5
fo 5.7), soil moisfure (from 10% fo more than 13%) and
biomass of soil fauna (from 1.3 to 72 g m"^), and de-
creased erosional losses (from 20,000 kg ha"^ fo 3 kg
ha"') (Yu & Pi 1985). Furfhermore, increases in plant
species richness occurred in some areas containing
more fhan 100 species, 50% of which colonized natu-
rally (Yu & Pi 1985). Economic invesfmenf greafly ex-
ceeded income in fhe early sfages of fhe projecf, but
over the 23-year duration fhe income-fo-invesfmenf ra-
fio was 3.2 (Yu & Pi 1985). This paffern of sequenfial
planting of increasingly economically valuable crops
has now been adopted by a number of local villages
(Jiangming Mo, South Ghina Insfifufe of Bofany, 1993,
personal communication).
The second project, esfablished in 1983 on 143 ha,
was to transform the severely degraded lowlands to a
complex, human-made system of forests, fruit-tree or-
chards, grasslands, and fish ponds (many aspects of
this project are described in Dinghu Shan Forest Eco-
system Station, Academia Sinica 1990) (Figure 7c). The
plan was to plant forests on the upper part of fhe
slopes, followed by fhe orchards, fhen the grasslands
to provide food for the fish ponds, and finally the fish
ponds af fhe bottom of fhe slope. Liffer from fhe forest
was fo be used fo fertilize the orchard and grassland.
Any nutrient loss from the forest would be captured by
the orchard and fish pond, and nutrient-rich muck col-
lected in fhe pond was to be returned fo fhe forest,
orchard, and grassland. This system requires intensive
human management (strategies 2-3, Figure 5), but it
provides a variety of producfs (fuelwood, fimber, fruit,
and fish) that are in great demand by the local popula-
Figure 7. Rehabilitation of severely degraded lands of
southern China: (a) before, (b) more than 15 years after tion and thus a more sustainable source of income.
planting with Acacia spp. and interplanted with native Establishment of fhe forest was the first phase of the
understory species, and (c) after transformation to a com- project. Different forest species were planted initially,
plex system of forests (background), orchards (young including pure stands of Acacia mangium, A. auricu-
plantings in fronf of foresf), grasses, and a fish pond. laeformis, Pinus massoniana (a native), and P. elliotii,
Photos provided by South China Institute of Botany {a and mixed stands of several Acacia spp., and nafive broad-
b) and A. E. Lugo (c). leaf species. After seven years of growth, the mixed

106 Restoration Ecology JUNE 1994

 Rehabilitation of Tropical Lands

forest was structurally and functionally better than the

single-species stands (Yu 1990). This forest contained
more biomass (136 Mg ha^^), was more productive (9
Mg ha"Vyr~'), had a higher leaf-area index (4.6), con-
tained more understory species (18 species, many of
which were typical of the native mature evergreen for-
ests of the area), and had larger increases in soil or-
ganic matter, soil nitrogen, available phosphorus, and
soil pH than did the single species forest.
Six years after establishment, the soil fertility and
microbial activity of the orchard and grassland were
higher than those of the original barren land. Further-
more, both the orchard and fish pond were producing
products within 4-5 years after the start of the project.
The results from this initial phase of the project have
been expanded to an area of 4000 ha as of 1989, and the
project is being supported by the local government as
an ecological and socioeconomical means of rehabili-
tating the degraded lands (Yu 1990).
Damaged lands under continued human use. In some situa-
tions, rehabilitation is initiated but, because of press-
ing needs by local people for fuel supplies, the speed
and direction of recovery is altered or arrested. An
example is the harvest and collection of understory
and litter from replanted forest lands in southern
China (Figure 8).
The present buffer zone of Ding Hu Shan Biosphere
Reserve, southern China, was planted with native
pine trees {Piniis massoniana) in 1944 by the local people
to reforest an area that was covered with degraded Eigure 8. {a) A pine forest growing on previously degraded
grasslands resulting from intense over-use (stressors lands in Guangdong Province, southern China, where
of type 3-5 of long duration. Figure 3) of the previous understory and litter are collected for human fuel needs.
forest lands. Since 1947, the local villagers have been The darker polygonal shapes on the slopes are where the
prevented from cutting the trees but are allowed to understory has recently been harvested, {b) A closer view
of the pine forest in the Ding Hu Shan Biosphere Reserve,
harvest other forms of biomass to meet their fuel
showing the type of litter being removed from the forest
needs. The standard practice, at various times of the by the local women villagers. Note the bamboo rakes with
year, is to rake the litter off the ground, cut the under- widely spaced tines that are used by the women for the
story plants down to the roots, and trim dead branches raking (photos by S. Brown).
from the trees (low-intensity stressors of type 3-5, Fig-
ure 3) (Figure 8). This removal of organic matter, a
critical link in nutrient cycling, has the potential to
retard recovery of the landscape to productive forest that succession has been arrested (Lenart 1992; S.
lands because it interferes with the rehabilitation of Brown, unpublished research results).
soil fertility. After almost 50 years of growth in a tropical seasonal
A cooperative project was initiated (between Sandra climate with high rainfall, the frees were still small
Brown of the University of Illinois and Kong Guo Hui (mean diameters of about 14 cm and mean height of 7
of the South China Institute of Botany) whose overall meters), and total biomass (above and below ground)
goals were to determine the effect of biomass and nu- was 94 Mg ha~^ Characteristics of the soil—including
trient removal in the harvested materials on the pro- organic matter, nitrogen, and fauna—were well below
ductivity, nutrient cycling, soil fertility, and biological expected values based on neighboring mature or late-
diversity of the pine forest, and the pattern of recovery secondary forests. However, the soil characteristics
when this local practice was stopped. In terms of forest were similar to those for the 6-year-old single-species
structure, biomass production, and soil fertility, forests in the above example. The total number of
results suggest that the forest is not doing well and plant species in the understory was 43, indicating that

JUNE 1994 Restoration Ecology 107

Rehabilitation of Tropical Lands

despife continued human impact, a relatively rich flora menf plan for fhe mining operation. Because of fhis,
developed. rehabilitation sfarfs before fhe mining. The firsf 30 cm
The foresf had low annual rafes of producfion—8.9 of fopsoil—and fhus soil nufrienfs, seeds, and mycor-
Mg/ha~Vyr \ with wood and litter producfion ac- rhiza—are removed and stockpiled. Following min-
counfing for 27% each and undersfory for 46%. About ing, fhis maferial is spread over fhe reconstituted land-
50% of fhe understory and litter producfion was har- scape. In the next stage, brush maffing and sorghum
vested per year, removing about 44-73% of N, P, K, seeds are spread abouf. This provides a quick cover
Ca, and Mg in fhese maferials (Mo 1994). The amounf crop fhaf lasfs for abouf one year. Affer fhis, seeds are
of nufrients removed from fhe harvesfed maferials ap- sown of species originally on fhe sife buf nof present in
peared fo exceed inpufs of N, P, and K from afmo- the brush maffing or seed bank. In addifion, various
spheric deposition. In addition to the direct removal of indigenous, fasf-growing Acacia species are also sown
nufrienfs, fhe harvest practice also reduced the active fo provide profecfion from wind erosion and fo fix ni-
pool of soil organic maffer, thus reducing nutrient trogen. The sowing program is accompanied by a one-
mineralization and soil nufrienf availability. time applicafion of NPK ferfilizer.
Sfopping fhe harvesting practice produced a signifi- Wifhin a year or so fhe sifes were reasonably well '
cant response in undersfory producfion and N cycling profected from wind erosion by a diverse cover of na-
wifhin two years (S. Brown, unpublished research five species (Lamb 1988). Affer six years, however,
results). Understory production increased by a 1.6-fold modificafion was badly needed. The Acacia species
factor, N concentration in free foliage increased, and were sown foo densely and, insfead of facilifafing veg-
leaching losses of N below fhe roofing zone decreased. efafive recovery, fhey essenfially arrested if by shading ^
These resulfs supporf the contention fhaf organic-maf- ouf and excluding ofher species. Through furfher frial
fer harvesting is responsible for fhe presenf state of the and error, applicafion of seeds of Acacia spp. has been
forest, rather than the argument fhaf fhe foresf is in an reduced fo less fhan 5% of fhe original amounf. Affer
arrested state from previous degradation. abouf 15 years, fhe Acacia species died out and the
To continue meefing fuel needs and af fhe same fime mafure woodland species were able fo affain fheir nor-
fo reduce nutrient losses, an alternative practice was mal role as oversfory species.
proposed (Mo 1994). A combination of liffer raking Malaysia. Tin mining acfivify in Peninsular Malaysia
(only half of fhe litter production was removed by fhis has produced abouf 200,00 ha of degraded or derelicf
process) and harvesfing fhe nef production of wood lands (Hoe 1991). Several projecfs have been inifiafed,
would reduce the current loss of nutrients by 40-65%. including afforesfafion and agroforesfry, fo find fhe
To harvest the net wood production would entail cut- besf mefhods of refurning soil ferfilify and producfiv-
ting about 3% of fhe free density, which of course ity to fhese lands (Awang 1991; Hashim 1991; Hoe
would need replanting to ensure a steady supply of 1991). All these projecfs had somefhing in common, as
frees. Not only would nutrient losses be reduced and is expecfed when dealing wifh derelicf lands: emphasis
the recovery of the foresf allowed fo confinue, buf fhis was given firsf fo fhe building of soil organic maffer,
alternative pracfice would provide higher qualify fuel nufrienfs, and biofa fo accelerafe fhe recovery. This
(wood has a higher heaf confenf than litter) and reduce was accomplished by a variefy of acfions, including
fhe fime and energy expended by fhe women fo coUecf inoculafion wifh mycorrhiza, fhe addifion of agricul-
the material. fural wastes, and addition of mulch and fertilizers. It
was found thaf suifable species must be able to fix
Rehabilitation Dominated by Human iVIanagement. When nitrogen on fhese very inferfile soils and fo wifhsfand
the inifial condifions are derelicf lands (subjecf to all heat and drought (for example. Acacia spp.). Because
stressors in Figure 3), great human effort is needed for these rehabilitation projects are costly, their feasibility
rehabilitation because ecosystem recovery processes depends on economic viability—as short-rofafion tim-
are extremely slow. These rehabilitations are the mosf ber plantations, for example. Several free management
cosfly because managers musf alfer fhe fopography, pracfices, such as pollarding or pruning, and their ef-
hydrology, and edaphic condifions, and supply all fhe fecf on survival and growfh were invesfigafed; results
biofic maferial (all sfrafegies in Figure 5). High in cosf, indicafed fhaf shorf-rotation planfafions were biologi-
these rehabilitations may result in complex forests, but cally feasible.
they generally do not. Examples are fhe rehabilitations In the case of mining, it is clear thaf rehabilifafion
affer mining in Ausfralia and Malaysia. should be parf of fhe mining process. Governments
Australia. A large subfropical island off fhe east coast of need to realize that the cost and fime involved in re-
Ausfralia is being sfripped-mined for heavy minerals claiming or rehabilifating derelicf lands affer mining is
(Lamb 1988). Here, plans for rehabilifafion and restora- finished are much higher fhan if rehabilifafion were
tion of the area must be part of fhe overall manage- included as parf of fhe mining process (as in Austra-

108 Restoration Ecology JUNE 1994

 Rehabilitation of Tropical Lands

lia). Not only are fhe direcf cosfs more because soil Table 5. Criteria for measuring the socioeconomic success
strucfure, ferfilify, and seed sources have fo be re-es- of land rehabilitation.*
fablished, buf cosfs of losf opporfunify (fhe fime tax) Issue Criteria
are high as well because fhe land sits idle and is likely
Sfabilify of fhe Sysfem Stable human population densify
to exert a negative impact on adjoining systems. (growth, migration rates).
Sfable market prices.
Stable and equitable land-use
Measuring the Success of Rehabilitation Projects and pafferns and land-fenure sys-
the Need for Technology and Increased Management tems.
Balance between subsistence and
A key task in measuring the success of any rehabilifa- cash crops.
tion projecf is defining ifs main objecfive; fhis will de- Stable rate of consumption of
fermine whaf should be measured. If is imporfanf fhaf other nafural resources (such as
both change and ifs direcfion be measured (Lamb fuelwood, water).
1988). Criferia for measuring fhe success of rehabilifa- Efficiency Degree of public involvement and
participation in the project (all
tion projecfs can be divided into those fhat measure members of coinmunity, not
ecological facfors (Table 4) and fhose fhaf measure so- only fhose who benefit the
cioeconomic facfors (Table 5). Many of fhese criferia most).
are difficult to apply in practice, and our inability to Flexibility Extent fo which economic flexibil-
use them routinely underscores the developmenf of ity is possible (such as through
production of products that iead
fhe rehabilifafion field and fhe need fo couple rehabili- to accumulation of wealth or
fation programs wifh research. surplus food).
While if is unavoidable fo pay the fime fax on poor
judgemenf and pasf misfakes wifh fhe use of re- *From Lamb 1988.
sources, and while fhe secref of success in rehabilifa-
fion is fo use nafural processes as much as possible These fwo imperafives require concerfed social
("self design" or "rolling wifh the punch"), there re- acf ion of various fypes. Firsf, we musf recognize fhe
main two additional requirements for making rehabili- imporfance of fhe economic role of a healfhy landscape
fafion of damaged lands a success and assuring a and be willing fo invest our best technology and efforf
straighf pafh fowards susfainable use of resources and fo mainfain fhe nafural healfh of lands and ecosys-
sustainable developmenf. These are (1) use of the best fems. Second, we musf recognize fhaf as human influ-
technology possible, and (2) increased pro-acfive man- ence increases on fhe landscape, if is the correct appli-
agemenf of fhe landscape (Lugo 1991). cation of human ingenuity and understanding fhaf will
reduce fhe negafive effecfs of fhe human presence.
Third, we musf value resource management as a criti-
Table 4. Criteria for measuring fhe ecological success of cal human activity and, as such, we must train our besf
land rehabilitation.* minds and provide fhe required social supporf for fhe
Issue Criteria insfifutions responsible for resource managemenf. In-
stifufions musf be renovafed fo assure thaf fhey acf in
Sustainabilify of the Sysfem Must be self-perpetuating coordinafion and af the appropriate scales of time and
and not need subsidies
(biocides, fefilizers, or space. The development of new technologies fo fackle
irrigation) to maintain problems of global and regional scale is anofher re-
itself. quirement for successful managemenf of lands. Tech-
Vulnerability fo Invasions Must resist invasion by ofher nologies such as geographic informafion systems, re-
species. source inventories, and modeling are some of fhe new
Biological Diversity All the key species or func- fools fhaf need fo be developed fo implemenf rafional
tional groups (plant, ani-
mal, and microbiota) must ecological engineering fechniques (Mifsch & Jorgensen
be present. 1989; Lugo 1991) in a world dominafed by humans and
Productivity Should be as productive as lands under judicious human confrol.
the original.
Nutrient Retention Nutrients should be recycled
efficiently. LITERATURE CITED
Hydroiogic Cycle Acceptable levels of wafer
quality and quantify (sur- Awang, K. 1991. Growth of three multipurpose tree species on
face and groundwater) tin tailing in Malaysia. Page 4 in D. Lamb, editor. Proceed-
must be reestablished. ings of the workshop on the restoration of degraded tropical
lands, November, 11-15, 1991. Botany Department, Univer-
•Based on Ewel 1987 and Lamb 1988. sity of Queensland, Brisbane, Australia.

JUNE 1994 Restoration Ecology 109

Rehabilitation of Tropical Lands

Bainbridge, D. A. 1990. The restoration of agricultural lands and Johnson, N., and B. Caharle. 1993. Surviving the cut: Natural
drylands. Pages 4-13 in ]. J. Berger, editor. Environmental forest management in the humid tropics. World Resources
restoration: Science and strategies for restoring the Earth. Institute, Washington, D.C.
Island Press, Washington, D.C. Karki, M. B. 1991. The rehabilitation of forest land in Nepal.
Barrow, C. J. 1991. Land degradation: Development and break- Nature and Resources 27(4):38-46.
down of terrestrial environments. Cambridge University Lamb, D., editor. 1988. Restoration of degraded ecosystems.
Press, Cambridge, England. International Union for the Conservation of Nature-World
Berger, J. J., editor. 1990. Environmental restoration: Science Conservation Union, Gland, Switzerland.
and strategies for restoring the earth. Island Press, Wash- Lanly, J. P. 1982. Tropical forest resources. Forestry Paper 30.
ington, D.C. Food and Agriculture Organization, Rome, Italy.
Bradshaw, A. D. 1987. Reclamation of derelict land and the ecol- Lavelle, P., E. Blanchart, A. Martin, S. Martin, A. Spain, F.
ogy of ecosystems. Pages 53-74 in W. R. Jordan, III, M. E. Toutain, I. Barois, and R. Schaefer. 1993. A hierarchical
Gilpin, and J. D. Aber, editors. Restoration ecology: A syn- model for decomposition in terrestrial ecosystems: Applica-
thetic approach to ecological research. Cambridge Univer- tion to soils of the humid tropics. Biotropica 25:130-150.
sity Press, Cambridge, England. Lenart, M. T. 1992. Organic matter dynamics in a human-im-
Bradshaw, A. D. 1989. Wasteland management and restoration pacted pine forest in subtropical China. Masters thesis. Uni-
of western Europe. Journal of Applied Ecology 26:775-786. versity of Illinois, Urbana-Champaign, Illinois.
Brown, S. 1993. Tropical forests and the global carbon cycle: The Lugo, A. E. 1978. Stress and ecosystems. Pages 62-101 in J. H.
need for sustainable land-use patterns. Agriculture, Ecosys- Thorp and J. W. Gibbons, editors. Energy and environmen-
tems, and Environment 46:31-44. tal stress in aquatic systems. U.S. Department of Energy
Brown, S., and A. E. Lugo. 1990. Tropical secondary forests. Symposium Series CONF-77114. National Technical Infor-
Journal of Tropical Ecology 6:1-32. mation Service, Springfield, Virginia.
Brown, S., A. J. R. Gillespie, and A. E. Lugo. 1991. Biomass of Lugo, A. E. 1988. The future of the forest: Ecosystem rehabilita-
tropical forests of South and Southeast Asia. Canadian Jour- tion in the tropics. Environment 30(7):17-20, 41-45.
nal of Forest Research 21:111-117. Lugo, A. E. 1991. Cities in the sustainable development of tropi-
Dale, V. H., R. A. Houghton, A. Grainger, A. E. Lugo, and S. cal landscapes. Nature and Resources 27(2):27-35.
Brown. 1993. Emissions of greenhouse gases from tropical Lugo, A. E. 1992fl. Tree plantations for rehabilitating damaged
deforestation and subsequent uses of the land. Pages 215- lands in the tropics. Pages 247-255 in M. K. Wali, editor.
260 in Sustainable agriculture and the environment in the Environmental rehabilitation, vol. 2. SPB Academic Publish-
humid tropics. National Academy Press, Washington, D.C. ing, The Hague, The Netherlands.
Dinghu Shan Forest Ecosystem Station, Academia Sinica, editor, Lugo, A. E. 1992/;. Comparison of tropical tree plantations with
1990. Tropical and Subtropical Forest Ecosystem 7:1-175. secondary forests of similar age. Ecological Monographs
Ewel, J. J. 1987. Restoration is the ultimate test of ecological 62:1-41.
theory. Pages 31-33 in W. R. Jordan, IH, M. E. Gilpin, and Lugo, A. E., J. Parrotta, and S. Brown. 1993. Loss in species
J. D. Aber, editors. Restoration ecology: A synthetic ap- caused by tropical deforestation and their recovery through-
proach to ecological research. Cambridge University Press, management. Ambio 22:106-109.
Cambridge, England. Mitsch, W. J., and S. E. Jorgensen, editors. 1989. Ecological
Food and Agriculture Organization. 1980-1990. Yearbook of for- engineering, an introduction to ecotechnology. John Wiley
est products. Rome, Italy. & Sons, New York.
Grainger, A. 1988. Estimating areas of degraded tropical .lands Mo Jiang Ming. 1994. Nutrient dynamics of a human-impacted
requiring replenishment of forest cover. International Tree pine forest in a MAB reserve of subtropical China. Masters
Crops Journal 5:31-61. thesis. University of Illinois, Urbana-Champaign, Illinois.
Harrington, J., and E. Howell. 1990. Pest plants in woodland Murphy, P. G., and A. E. Lugo. 1986. Ecology of tropical dry
restoration. Pages 61-69 in J. J. Berger, editor. Environmen- forests. Annual Review of Ecology and Systematies 17:67-
tal restoration: Science and strategies for restoring the 88.
Earth. Island Press, Washington, D.C. National Research Council. 1993. Sustainable agriculture and the
Hashim, A. 1991. Rehabilitation of ex-tin mining land hy agro- environment in the humid tropics. National Academy
forestry practices. Page 2 in D. Lamb, editor. Proceedings of Press, Washington, D.C.
the workshop on the restoration of degraded tropical lands, Nepstad, D. C , C. Uhl, and E. A. S. Serrao. 1991. Recuperation
November 11-15, 1991. Botany Department, University of of a degraded Amazonian landscape: Forest recovery and
Queensland, Brisbane, Australia. agriculture restoration. Ambio 20:248-255.
Hoe, A. L. 1991. Problems and prospects of afforestation on Odum, E. P. 1969. The strategy of ecosystem development. Sci-
sandy tin tailings in Peninsular Malaysia. Page 2 in D. ence 164:262-270.
Lamb, editor. Proceedings of the workshop on the restora- Parrotta, J. A. 1992. The role of plantation forests in rehabilitat-
tion of degraded tropical lands, 11-15 November 1991. Bot- ing degraded tropical ecosystems. Agriculture, Ecosystems
any Department, University of Queensland, Brisbane, Aus- and Environment 41:115-133.
tralia. Parrotta, J. A. 1993. Secondary forest regeneration on degraded
Janzen, D. H. 1986. Guanacaste National Park: Tropical, ecologi- tropical lands. The role of plantations as "foster ecosys-
cal, and cultural restoration. Editorial Universidad Estatal a tems." Pages 63-73 in H. Lieth and M. Lohmann, editors.
Distancia, San Jose, Costa Rica. Restoration of tropical forest ecosystems. Kluwer Academic
Janzen, D. H. 1988 Tropical ecological and biocultural restora- Publishers, The Netherlands.
tion. Science 239:243-244. Richards, J. F., and E. R. Flint. 1994. A century of land use
Johns, A. D. 1992. Species conservation in managed tropical change in South and Southeast Asia. Ch. 2 in V. Dale, edi-
forests. Pages 15-54 in T. C. Whitmore and J. A. Sayer, tor. Effects of land use change on atmospheric CO2 concen-
editors. Tropical deforestation and species extinction. Chap- trations: Southeast Asia as a case study. Springer Verlag,
man and Hall, London, England. New York.

110 Restoration Ecology JUNE 1994

 Rehabilitation of TrofJical Lands

Schmidt, R. 1987. Tropical rain forest management: A status Uhl, C, A. Verissimo, M. Mattos, Z. Brandino, and I. C. G.
report. Unasylva 39:2-17. Vieira. 1991. Social, economic, and ecological consequences
Scbreckenberg, K., M. Hadley, and M. I. Dyer, editors. 1990. of selective logging in an Amazon frontier: The case of
Management and restoration of human-impacted resources: Tailandia. Forest Ecology and Management 46:243-273.
Approaches to ecosystem rehabilitation. Man and the Bio- Whitmore, T. C. 1984. Tropical rainforests of the Far East. 2nd
sphere Digest 5. United Nations Economic and Social Coun- edition. Oxford University Press, Oxford, England.
cil, Paris, France. Wood, D. 1993. Forests to fields, restoring tropical lands to agri-
Singh, K. D. 1993. The 1990 tropical forest resources assessment. culture. Land Use Policy, April: 91-107.
Unasylva 174(44):10-19. Yu Zuo Yue. 1990. Recovery of vegetation on southern subtropi-
Tucker, R. P. 1992. Foreign investors, timber extraction, and cal degraded downlands in Guangdong and establishment
forest depletion in Central America before 1941. Pages 265- of model eco-agriculture. Tropical and Subtropical Forest
276 in H. K. Steen and R. P. Tucker, editors. Changing Ecosystem 7:1-9.
tropical forests: Historical perspectives on today's chal- Yu Zuo Yue and Pi Yong Feng. 1985. The path and its effects to
lenges in Central and South America. Forest History Soci- rehabilitating vegetation on the tropical littoral eroded land
ety, Duke University, North Carolina. in Guangdong Province. Tropical and Subtropical Forest
Uhl, C. 1988. Restoration of degraded lands in the Amazon Ba- Ecosystem 3:97-108.
sin. Pages 326-332 in E. O. Wilson and F. M. Peter, editors. Yu Zuo Yue and Wang Zhu Hao. 1989. The method of recovery
Biodiversity. National Academy Press, Washington, D.C. of forest vegetation in degraded ecosj'stems in tropical and
Ubl, C , and I. C. G. Vieira. 1989. Ecological impacts of selective subtropical waste lowlands in Guangdong. Pages 18-22 in
logging in the Brazilian Amazon: A case study from the S. Adisoemarto, editor. Proceedings of regional workshop
Paragominas region of the state of Para. Biotropica 21:98- on ecodevelopment process for degraded land resources in
106. southeast Asia. Indonesian Man and the Biosphere National
Uhl, C , R. Buschbacher, and E. A. S. Serrao. 1988. Abandoned Committee and Regional Office for Science and Terluic)ogy,
pastures in eastern Amazonia. I. Patterns of plant succes- United Nations Economic and Social Council, Southeast
sion. Journal of Ecology 76:663-681. Asia, Jakarta.

JUNE 1994 Restoration Ecology 111