246 Restoration and Rehabilitation of Degraded Ecosystems in Arid and Semi Arid Lands i

8/16/2019 246 Restoration and Rehabilitation of Degraded Ecosystems in Arid and Semi Arid Lands i

1/10

Restoration andRehabilitation of Degraded Ecosystems inArid and Semi-Arid Lands.I. A View from the South

J ronsod

C. le

E. Leloc’hl

C. deR Pontanier

Abstract

A general model is presented describing ecosystemdegradation to help decide when restoration, rehabili-tation, or reallocation should be the preferred re-sponse. The latter two pathways are suggested whenone or more “thresholds of irreversibility” have beencrossed in the course of ecosystem degradation, and

when “passive” restoration to a presumed predisturb-ante

condition is deemed impossible. The young butburgeoning field of ecological restoration, and theolder field of rehabilitation and sustainable rangemanagement of arid and semiarid lands

ASAL ,

arefound to have much in common, especially comparedwith the reallocation of lands, which is often carriedout without reference to pre-existing ecosystems. Af-ter clarifying some basic terminology, we present 18vital ecosystem attributes for evaluating stages of deg-radation and planning experiments in the restorationor rehabilitation of degraded ecosystems. Finally, we

offer 10 hypotheses concerning ecological restorationand rehabilitation as they apply to ASAL and perhapsto all terrestrial ecosystems.

ICentre d’Ecologie Fonctionnelle et Evolutive L. Emberger,C.N.R.S., B.P. 5051 34033 Montpellier Cedex 01, France2Estaci6n Experimental Quilamapu, I.N.I.A., Casilla 426,ChillAn, Chile30RSTOM, B.P. 434, 1004 El Menzah 1, Tunisia

1993 Society for Ecological Restoration

In the nations of the “North,” where restoration ecol-ogy has mostly been pursued thus far, aesthetic or

intrinsic values (Naess 1986) have motivated most ef-forts to date. The primary goal has been to create “liv-ing museums,” or to put things back as they oncewere. On the other hand, the vast literature of ecologyapplied to agriculture, forestry, and range manage-

ment in the “South,” particularly in arid and semiaridlands ASAL , takes for granted that people will con-tinue to be the dominant force in both natural andagro-ecosystems. The main issues are, first, whetherprimary and secondary productivity can be increasedor sustained by new management techniques and, sec-ond, what effects these techniques might have onbiodiversity and ecosystem stability. The models andideas guiding these applied fields have mostly comefrom sources other than restoration ecology and con-servation biology. Nevertheless, we believe that rangemanagers, agronomic engineers, conservation biolo-

gists, and restoration/rehabilitation ecologists could allbenefit from greater exchange of ideas and methodol-ogy, both in the rich North and the poorer South (seeDyksterhuis 1949; Bradshaw 1983; Jordan et al. 1987).

At a higher level, it is the fragmentation and degra-dation of entire landscapes that both restorationistsand rehabilitators must combat. When economic andcultural practices are modified in the direction of eco-logical sustainability and conservation of biodiversity,and when restoration or rehabilitation is applied to allpartially degraded ecosystems, with the help of all thenecessary and appropriate scientific disciplines, the

result would be-to borrow a phrase from Hobbs andSaunders 1991 -an attempt at “reintegration of frag-mented landscapes” (Fig. 1).

After clarifying some basic terminology, we will dis-cuss 18 vital ecosystem attributes for evaluating degra-dation and planning experiments in the restoration orrehabilitation of degraded ecosystems. Finally, we of-fer 10 hypotheses concerning ecological restorationand rehabilitation.

Basic Terminology

Restoration Sensu Strict0 and Sensu Late Restoration ofdegraded ecosystems can be likened to the restorationof a Renaissance painting that has deteriorated overtime but still reveals its initial lines and colors suffi-ciently for the fine arts restorator to do his or her work.Analogously, ecological restorationists seek a completeor near-complete return of a site to a pre-existing state.The Society for Ecological Restoration (SER) definesrestoration as “the intentional alteration of a site toestablish a defined indigenous, historic ecosystem.The goal of this process is to emulate the structure,functioning, diversity, and dynamics of the specified

Restoration Ecology MARCH 99 3


2/10

Restoration and Rehabilitation. I.

Figure 1. Relationships and potentialpathways for exchange among ecologi-cal restoration and rehabilitation ofdegraded ecosystems and variousbranches of ecology, conservationbiology, and sustainable land andbioresource management, includingsoil science.

ecosystem.” Implicit in this definition is the notionthat restoration seeks to reassemble, insofar as possi-ble, some predefined species inventory.

wever, since it is rarely possible to determine ex-what historic or prehistoric ecosystems looked

or how they functioned, let alone establish the fullspecies list of indigenous communities, restoration ef-forts may be plagued by ambiguities in both their goals

nd criteria of success (Cairns 1989, 1991; Simberloff990). We suggest using the term “restoration sense

stricto” to describe endeavors corresponding to the

SER definition, as opposed to restoration sense Z o ,which seeks simply to halt degradation and to redirecta disturbed ecosystem in a trajectory resembling thatpresumed to have prevailed prior to the onset of dis-turbance.

:’ Despite this difference, the primary goal of bothsense stricto and senstl late restoration is the conserva-tion of indigenous biodiversity and ecosystem struc-ture and dynamics. They thus differ from a third possi-

\

ble response to ecosystem degradation, which we callrehabilitation.

Rehabilitation. Rehabilitation, in our sense, seeks to re-pair damaged or blocked ecosystem functions, with theprimary goal of raising ecosystem productivity for thebenefit of local people. Moreover, it attempts toachieve such changes as rapidly as possible. However,a rehabilitation project resembles a restoration attemptin adopting the indigenous ecosystem’s structure andfunctioning as the principal models to be followed,insofar as they can be determined or guessed. That is,they both aim at recreating autonomous or self-sus-

taining ecosystems, which are characterized by bioticchange or succession in plant and animal communi-ties, and the ability to repair themselves following nat-ural or moderate human perturbations. Thus, restora-tion and rehabilitation projects must also share asexplicit or implicit working goals the return to formerpaths of energy flow and nutrient cycling, and the rep-aration of conditions necessary for effective water infil-tration and cycling throughout the ecosystem’s rhizo-sphere (Allen 1988, 1989; DePuit Redente 1988).However, whereas restoration sensu stvicto invariably

seeks a direct and full return to the indigenous, his-toric ecosystem, restoration sense late and, particularly,

rehabilitation may settle on one of many possible alter-native steady states, or a synthetic “simplified ecosys-tem” as an intermediate step in their long-term goals(Fig. 2). The alternative steady states might or mightnot have occurred in the process of degradation of theoriginal, predisturbance ecosystem. In any case, theyare-like our so-called “simplified ecosystems”-apractical method for designing, managing, and evalu-ating ecosystem-level experiments (Fig. 2).

What we call rehabilitation has often been called rec-lamation, particularly in conjunction with mine-tailingrevegetation (Bengson 1977; Bradshaw 1987). But rec-lamation has also been used synonymously with bothrestoration and with some examples of what we callreallocation.

Reallocation. A general term is needed to describewhat happens when part of a landscape, in any state,is assigned a new use that does not necessarily bear anintrinsic relationship with the predisturbance ecosys-

MARCH 99 Restoration Ecology 9


3/10


Thrcrhold o f

Figure 2. General model of ecosystemdegradation and the possible re:sponses to it. The original predisturb-ante ecosystem is pictured in acloudlike space to indicate that it isfrequently ambiguous.

tern’s structure or functioning. We call this realloca-tion. For simplicity’s sake, reallocation pathways inFigure 2 are indicated only after various stages of eco-system degradation have taken place and one or morethresholds of irreversibility have been crossed. In real-ity, it can occur in the case of a slightly disturbed eco-system or even an undisturbed (predisturbed) one.Problems arise when reallocated sites sprawl over

landscapes in a more or less anarchic fashion.In contrast with restoration and rehabilitation, real-location assumes a permanent managerial role for peo-ple and normally requires ongoing subsidies in theform of energy, water, and fertilizers. The huge plan-tations of the fodder shrubs in North Africa, such asAtviplex, Acacia, and Opuntiu spp., are examples of real-location. By contrast,. the native perennial grasses wehave introduced or reintroduced there Cha’ieb 1990;Chai’eb et al. 1992a, 1992b) survive direct grazing, re-produce sexually, and eventually become naturalized.They form part of a rehabilitation experiment.

cosystem of Reference. As Cairns 1991), Simberloff 1990), and Sprugel l991 point out, it’s often not clearwhat ecological yardstick is being used when a restora-tion or rehabilitation experiment is being set up. Yetfor purposes of project design and evaluation, it is de-sirable to establish ahead of time some standard ofcomparison and evaluation, even if it is arbitrary. Wecall this the ecosystem of reference. In restoration, thiswill normally correspond to the SER’s “indigenous,historic ecosystem,” but in some rehabilitation (and, of

course, reallocation) projects, it may be something en-tirely different, depending on the state of advance-ment of the ecosystem’s degradation and on the needsof landowners or local people.

/Alternate Steady-States. It is often assumed that if eco-system degradation has not progressed too far, a re-turn to an “indigenous, historic” state is possible sim-

ply by removing anthropic stresses (intentional fires,wood clearing, overgrazing, etc.) and allowing naturalprocesses to do the rest. Yet in ASAL, this is rarely thecase (Smeins et al. 1976; West et al. 1984; Westoby et al.1989; Omar 1991). In the ASAL, including Mediterra-nean climate regions (Naveh 1988), it seems more rea-sonable to seek a return to an intermediate or alterna-tive steady-state, such as a quasi-metastability that canbe created and maintained by continual but relativelylight human disturbance. Some disturbances are, ofcourse, created in the absence of people, such as natu-ral fires, hurricanes, volcanos, epidemics, but theseappear to be rare in the ASAL. This still implies intro-

ducing new attitudes, insights, and-above all-en-lightened management techniques to local people(Goodloe 1969; Lange 1969; Janzen 1986), particularly ifindigenous ecosystems have already crossed one ormore thresholds in the course of their degradation.

Threshold of Irreversibility. The concept of “thresholds”of environmental change is well established in ecology(Holling 1973; May 1977; Wissel1984 , and has recentlybeen applied to range management as well (Friedel

10

Restoration EcologyMARCH 1993


4/10


1991; Laycock 1991). Still, “thresholds of irrevers-ibility” are not easily detected or quantified, but oncehaving passed through one, most ASAL ecosystemsapparently cannot cross back without interventionsdesigned to correct the specific changes that led to ther threshold being crossed. For example, reconstitutionof seed banks might be needed, or the restocking of

soil organic matter and microorganisms that promotehigher plant establishment and growth. Where trunca-tion of upper soil layers, sedimentation, salinization,or other processes have drastically modified surfaceand sub-soils, reconditioning of soils or reactivation oftheir hydrological functioning may be required.

Restating one difference between rehabilitation andrestoration, we can now say that the former oftenneeds to jumpstart a new ecosystem trajectory andcombat the conditions that constitute any existingthresholds of irreversibility. By contrast, most restora-

of

Resistance and Resilience. Resistance is an ecosystem’sinertia in the face of change (Margalef 1969; Holling1973), and resiliency is its ability to return to a formersuccessional trajectory after being degraded or de-flected by outside disturbances Connell Slatyer1977). Resilience may be the best indicator of ecosys-tem health or integrity (Leopold 1948). Stability is thegeneral concept embracing both resistance and resil-ience Westman 1978).

M a y 1977), Westman 1978), W i l l i s 1963), a n d

others have proposed mathematical models of dy-namic resistance and resilience for a number of theo-

retical systems, but practical ways of measuring it areseldom discussed. Walker et al. (1981) proposed as twopossible measures the coefficient of variation of pro-ductivity in different trajectory phases and the rate ofreturn to a former level of productivity following spe-cific interventions (see Noy-Meir Walker 1986;Westoby et al. 1989). Clearly, these measures are morerelevant to rehabilitation than to restoration. For thelatter, some measure of ecosystem structure, includingits biodiversity, will be needed as well.

I. Vital Ecosystem Attributes as Related to EcosystemStrut

ure Given the dramatic changes in many ecosys-tems’ physiognomies from summer to winter, or, inmany ASAL, from wet to dry seasons, this first groupof VEA should be measured or calculated at the end ofthe main growing season-typically in late spring forarid, temperate ecosystems, or the end of the mainrainy season in the tropical and subtropical aridzones-when maximum expression of alpha, beta,and gamma diversity is found. Ideally, they should be

recorded for several successive years. These attributesare as follows:

1

perennial species richness,

2

annual species richness,

3

total plant cover,

4

aboveground phytomass,

5

beta diversity,

6

life form spectrum,

7

keystone species,

8

microbial biomass, and

9

soil biota diversity.

One point worth emphasizing is that resilience and Both (1) perennial species richness and (2) anntlal speciesresistance are often unconnected (Noy-Meir 1974; richness are relatively easy to obtain through replicatedNaveh Lieberman 1984). Moreover, we argue that releves and, when combined, reveal structural differ-no generalization or equation about either one can ap- ences among phases in an ecosystem’s succession orply to the entire diachronic process of an ecosystem retrogression. Perennials appear to occupy a dominanttrajectory. Instead, they should be employed only in position in most terrestrial and aquatic ecosystems un-the context of a specific phase of retrogressive or pro- der relatively stable conditions (Frank 1968). By con-gressive succession bounded by a “ceiling” and a trast, some successional stages of many semiarid eco-“floor.” We suggest that within a given phase of deg- systems are characterized by very large numbers ofradation, increase of resistance and of resilience can be annual species. Thus, where possible, it is useful toinversely related. have data on the species composition of the ecosystem

Vital Ecosystem Attributes

Noble and Slayter (1980) defined several categories ofvital attributes of life history useful in determining theresponse of a species to recurrent disturbance. Herewe modify Noble and Slayter’s concept. We define asvital ecosystem attributes (VEA) those characteristicsor attributes that are correlated with and can serve as

indicators of ecosystem structure and function. Theyshould help in formulating predictions and designingexperiments in both restoration and rehabilitation.One may object that many of these VEA are not readilyobtainable under normal project conditions. Many ofthem are interrelated however, and determining onemay allow an accurate estimation of other more diffi-cult ones. Quite some time ago, Odum (1969) also con-structed a list of ecosystem attributes to compare “de-velopmental” and “mature” stages of ecologicalsuccession. Disappointingly little has been done sincethen to test his ideas of the “strategy of ecosystem

development.”

MARCH 1993 Restoration Ecology 1 1


5/10


of reference. Similarly, low soil fertility is sometimesassociated with high annual species diversity-in cer-tain temperate grasslands (Gough Marrs 1990) andsavannas (Scholes 1989), for example. Finally, someperennials become abundant, pesty invaders underconditions of prolonged disturbance.

Among perennials, it is useful to distinguish be-

tween herbaceous and woody species, since the inter-actions of woody and herbaceous layers are importantfactors in many of the ecosystems most subject to pro-longed use and anthropic degradation (such as savan-nas, semi-arid grasslands, arid shrublands) (WalkerNoy-Meir 1982; Ovalle Avendano 1987; Archer et al.1988). The ratio of annuals to perennials can be reveal-ing as well.

(3) Total plant cover is a good integrator of VEA 1 and2, but may vary significantly within and among years(Whittaker 1972). In some cases, such as Sahelian sa-vannas, total ligneous plant cover is a more useful in-dicator. Moreover, absolute cover alone, even whencombined with species richness, does not providemuch insight into ecosystem productivity. Thus, twoimportant corollaries are (4) Aboveground Phytomass (kgdry matter per 10 or 10,000 m2) which is best measuredat the end of the main growing season, and biomassproductivity, which will be described below.

(5) Beta diversity is defined as the “extent of speciesreplacement or biotic change along environmental gra-dients” (Whittaker 1972). MacArthur (1965) andWilson and Shmida (1984) have established the impor-tance of determining beta diversity in addition to alphadiversity (the number of species within community

samples) as components of overall diversity. In onestudy (Frank McNaughton 1991), plant communitydiversity was found to be positively correlated withresistance to change in species composition when per-turbed by drought.

(6) Life form spectrum, first defined by Raunkiaer 1934), is an additional indicator of ecosystem structureand, probably, of functioning as well. As for beta di-versity, the range of life forms in an ecosystem usuallydecreases with degradation.

(7) Keystone species are those species critical to ecosys-tem structure and functioning. The concept is known

primarily from the literature of food webs and conser-vation biology (such as Paine 1969; Gilbert IPSO), yet itseems appropriate in restoration ecology as well.Thus, an attempt to reorient the trajectory of a dis-turbed ecosystem may be facilitated by carefully rein-troducing or increasing the density of keystone speciesand, where necessary, by the de-emphasis of the sur-vival of other species (Simberloff 1990).

Given the growing body of empirical evidence of theimportance of perennial, nitrogen-fixing legumes inundisturbed ASAL ecosystems (Jenkins et al. 1987; Jar-

rell Virginia 1990) as well as in alternative steady-state systems found in Mediterranean-climate regionsand savannas (Knoop Walker 1985; Ovalle Aven-dano 1987; and others), we assume that perennial,nitrogen-fixing legumes are among the keystone spe-cies in many ecosystems, including those in our studyareas in central Chile, southern Tunisia, and northern

Cameroon (Aronson et al. 1992).Once identified or at least specified in a workinghypothesis, keystone species may logically be amongthe first candidates for experimental reintroduction todisturbed ecosystems. However, since plant reintro-duction is an expensive, risky option that commits per-sonnel to long-term monitoring and management, it iswell to assess the methods to be employed and therisks they present (Hughes Styles 1987). Differentprovenances of introduced nitrogen-fixing legumescan vary dramatically in their impact on herbaceousplants and soils (Reetu Sharma Dakshini 1991).

If nitrogen-fixing trees and shrubs are keystone spe-cies, then their associated rhizobial and other micro-symbionts must be so considered as well Dommer-gues Krupa 1978). Although they are less welldocumented and understood than higher plants, itseems likely that many of the spatial and temporalconsiderations mentioned for keystone plant specieswill be found to apply to microsymbiont keystones aswell.

Accordingly, (8) microbial biomass and, particularly,(9) soil biotu diversity should both be estimated whenpossible. Soil microorganisms have a tremendous im-pact on vegetation in ASAL and other terrestrial eco-

systems (Whitford Elkins 1986; Carpenter Allen1988; Virginia 1986). For example, soil bacteria, espe-cially Rhizobium play a critical role in plant competitionin a grass-legume community (Turkington et al. 1988).Mycorrhizae can regulate competition between plantsof different successional stages (Allen Allen 1984;also see Nelson Allen in this volume).

The most probable number technique allows esti-mates of microbial biomass. Soil biota diversity is alsoamenable to quantification; just as information onaboveground abundance and diversity is critical, it isimportant to carry out inventories of heterotrophic

nitrogen-fixing bacteria and other microorganismssuch as rhizobia and endo- and vesicular-arbuscularmycorrhizae .

II. Vital Ecosystem Attributes Related to Ecosystem Func

tion. The second group of attributes are as follows:

(1) biomass productivity,(2) soil organic matter,(3) maximum available soil water reserves,(4) coefficient of rainfall efficiency,

12 Restoration Ecology MARCH 9


6/10

Restoration and Rehabilitation.

(5) rain use efficiency,(6) length of water availability period,(7) nitrogen use efficiency,(8) microsymbiont effectiveness, and

(9) cycling indices.

(1) Biomass productivity (kg biomass ha-iyr-‘ , a com-plement to total cover, is of concern to rehabilitation

projects. These indicators are nonetheless insufficientto give all the information required by an ecosystemmanager, despite Margalef’s (1969) generalization thatthe ratio of standing biomass to annual productivityincreases with increased maturity of an ecosystem.Some ecosystems are actually more productive (interms of kg per unit area per unit time) during earlystages of degradation (Odum et al. 1979) than in thepredisturbance state. This is due to the rapid coloniza-tion by nitrophilous weeds, typically including annualAsteraceae and grasses, but occasionally some woodylegume species as well. In advanced stages of degrada-

tion, productivity invariably declines drastically. Thus,additional vital ecosystem attributes are needed to de-scribe ecosystem function.

(2) So d organic matter content is a readily accessibleand highly revealing attribute that complements vital

ecosystem attributes (5) and (6). For example, there is apositive correlation between organic carbon contentand aboveground phytomass in some subtropical soils Lug0 et al. 1986). Low levels of organic matter alsodirectly influence soil features critical to seedling estab-lishment and to water and root infiltration in arid andsemiarid lands (Le Houerou 1969; Cesar 1989). Accord-ingly, there is great interest in the study of leaf litter,

detritivores, and other potential contributors to or-ganic matter in these systems Schaeffer Whitford1981)

(3) Maximum available soil water reserves have greatimportance in ASAL where rainfall is irregular, even ifthey are not always readily obtainable. For example, inthe Tunisian and Cameroonian case studies of FloretPontanier (1982) and Seiny-Boukar et al. 1992), soilprofiles are shallow and water reserves can be easilymeasured and have been correlated with productivity.The differential influence of water reserves on theaboveground biomass productivity of several native

shrub and perennial grass species of the steppes ofsouthern Tunisia is quite dramatic (Chai’eb et al. 1990),and their measure has been used in the design of eco-system projects in that region (Chai’eb et al. 1992,1992b). Similar data on North American grassland spe-cies have been used in the experimental restorationand management of prairies (Burton et al. 1988).

(4) ff i . t foe czen o rainfall efficiency is defined as theamount of water infiltrating to middle and deep soillayers, and it is thus an indicator of soil surface condi-

tions and of the absorption capacity of soils ChaIeb etal. 1992a). All water infiltrating past the soil surface isof course not necessarily used by plants. Yet, as formaximum available soil water reserves, coefficient ofrainfall efficiency is a useful indicator of soil conditionsin ASAL and elsewhere. It is closely tied to the pres-ence or absence of surface crusts, which form in de-graded systems and tend to seal off soils against effi-

cient infiltration of rainfall (Skujins 1991).(5) Rain tlse efficiency equals the slope of the relation-

ship between annual rainfall and aboveground phyto-mass production (Le Houerou 1984). In dryland sys-tems and elsewhere, it serves as an excellent indicatorof soils and, hence, of ecosystem productivity. Wateruse efficiency, expressed in terms of kg of above-ground biomass produced per mm of

evapov

transpirated water, is more accurate than rain use effi-ciency (Floret et al. 1983; Seiny-Boukar et al. 1992).However, rain use efficiency is easier to obtain in mostsituations and more useful for inter-regional compari-

sons.(6) Length of water availability period in the soil is easilymeasured by successive tensiometer readings at differ-ent soil depths. When evaluated in conjunction withrain use efficiency, water availability period allowspredictions of the seasonal@, duration, and extent ofplant growth, and helps guide species selection inearly stages of restoration or rehabilitation projects(Chai’eb et al. 1992a, 199227).

(7) Nitrogen use efficiency is a vital attribute since,even in arid environments, available nitrogen andphosphorus (and other nutrients) can limit plant andanimal growth as much as deficiencies in water can

(Penning de Vries et al. 1980). There may be an inverserelationship between the amount of standing biomassin a system at a given moment and the amount ofnitrogen in that biomass (Penning De Vries Djiteye1982; Vitousek 1982).

As currently defined for individual plants or popula-tions, nitrogen use efficiency (NUE) combines (1) theinstantaneous rate of carbon fixation per unit of nitro-gen, and (2) the mean residence time of nitrogen in theplant (Berendse Aerts 1987). Thus, NUE = AIL,,where A is nitrogen productivity and l/L, is the meanresidence time of nitrogen. The first component is es-

sential, since a close linear relationship exists betweenrelative growth rate and nitrogen concentration inplants (Ingestad 1979). The second component is im-portant because nitrogen can reside in a plant for vary-ing lengths of time before being discarded through theshedding of organs (Berendse Aerts 1987). Undernutrient-poor conditions, a long mean, residence timemay be favorable to the plant (Berendse 1985), whereasunder nitrogen-rich conditions, high nitrogen uptakeand rapid circulation are favored (Vitousek 1982).

MARCH 1 9 9 3 Restoration Ecology 13


7/10


However, clear differences in both nitrogen conver-sion rates and mean residence times exist in co-occurring species (Brown 1978; Berendse Aerts 1987;Berendse et al. 1987). Although Vitousek (1982) sug-gested that nutrient use efficiency for a forest is theinverse of nutrient concentration in the abovegroundlitterfall, it remains to be clarified how exactly to apply

the concept of nutrient use efficiency generally, to acomplex ecosystem. One factor to consider is clearlythe relationship between nutrient use efficiency andlife form. For example, Muller and Garnier (1990) and

Joffre (1990) have shown that some perennial grassesmake,-more efficient use of nitrogen than congenericannual grasses.

(8) Microsymbiont effectiveness is no less critical thanbiomass or diversity. Legume-Xhizobium couples repre-sent the most widely used plant-bacterium association.However, introduced rhizobial strains often fail to sur-vive in competition with indigenous strains, particu-

larly under adverse soil conditions (van ElsasHeijnen 1990; Asad et al. 1991). Thus, the successfuluse of a rhizobial inoculant requires knowledge of theabundance, diversity, and competitive ability of theindigenous strains as well as of the growth limitationsof the inoculants selected for introduction.

Jarrell and Virginia (1990) suggest that soil cationaccumulation in the soil root zone can be used to calcu-late both total water use during the lifetime of a givenecosystem and cumulative symbiotic A&-fixation. Thissuggests that empirical variables could be identifiedthat would allow estimates of soil biota effectiveness,given a fixed amount and diversity of standing vegeta-

tion over a period of constant climate conditions.(9) Cycling indices measure the ratio of the amount of

energy or an element recycled in an ecosystem to theamount of energy or elements moving straightthrough the system (Finn 1976). Species richness andmany other structural and functional attributes are cor-related with soil nutrient levels and cycling (Willis1963; Gough Marrs 1990). For example, soil fertilityhas been found to have a controlling influence on rainuse efficiency in savannas (Penning de Vries et al.1980; Scholes 1989). Furthermore, since the degree towhich a nutrient circulates freely in an ecosystem de-

pends partly on its physical state, and is thus closelylinked with hydrological conditions, cycling indices ofN, P or other nutrients will also reflect on such vitalecosystem attributes as water reserves, water availabil-ity period, and coefficient of rainfall efficiency. Com-paring the cycling index of nutrients (especially phos-phorus and nitrogen) at different phases of anecosystem’s trajectory may thus be useful for evaluat-ing relative perturbation or attempts at restoration orrehabilitation. Since ecosystems in a “mature” stageare thought to have a greater capacity to entrap and

hold nutrients for internal cycling than in less mature,developing stages (Odum 1969), the achievement oftighter mineral cycles and reduced nutrient exchangerates between organisms and the environment shouldreveal that restoration or rehabilitation is beingachieved. Furthermore, in efforts aimed at the reinte-gration of fragmented landscapes, cycling indices

could measure interactiveness among interlocking eco-systems.

Discussion

As Bradshaw (1987) noted,the ecology of ecosystems.

the time has come to studyBoth restoration and reha-

bilitation projects offer good opportunities to do so. Itis also time to study the evolution or the trajectories ofecosystems in the process of degradation, restorationor rehabilitation. We have argued that restorationsenstl stricto seeks to re-establish a full inventory of pre-existing species, while restoration serzsu late is satisfied

to reorient a disturbed ecosystem’s trajectory in a di-rection ressembling its predisturbance state. Rehabili-tation seeks the repair of ecosystem function above all,in the search for sustained productivity. Reallocation,when applied to agriculture or silviculture, assumesdependence on external subsidies and represents anunnatural situation of “perpetually interrupted succes-sion” (Jarrell 1990).

If it is to distinguish itself from traditional forest orrange management and rehabilitation agronomy(Malcolm 1990) on the one hand, and conservation bi-ology on the other, restoration and rehabilitation ecol-

ogy must show itself to be a maturing science by gen-erating its own models and by testing hypotheses onecosystems in retrogression or renewed succession. Tobegin, below are ten hypotheses.

Ten Hypotheses for Restoration and Rehabilitation Ecology.

l Beyond one or more thresholds of irreversibility, eco-system degradation is irreversible without structuralinterventions combined with revised managementtechniques.

l The more thresholds passed, the more time and en-ergy will be required for an ecosystem’s restoration

or rehabilitation.

Without massive intervention, restoration will pro-ceed only as far as the next highest threshold in theprocess of vegetation change or succession.

Beta diversity and life form ranges decline with eco-system degradation, while alpha diversity temporar-ily increases.

l The loss of keystone species speeds degradationmore than the loss of other species. Such losses tendto occur simultaneously with the crossing of thresh-olds of irreversibility.

14 Restoration Ecology MARCH 99 3


8/10


9/10


10/10

246 Restoration and Rehabilitation of Degraded Ecosystems in Arid and Semi Arid Lands i

Documents

Transcript of 246 Restoration and Rehabilitation of Degraded Ecosystems in Arid and Semi Arid Lands i