CASE STUDY

Ecological Restoration of Lantana-InvadedLandscapes in Corbett Tiger Reserve, India

Suresh Babu, Amit Love and Cherukuri Raghavendra Babu

ABSTRACT

Lantana {Lantana cámara), one of the world's most troublesome invasive weeds, has become a menace in most of theprotected areas located in tropical and subtropical belt of India. The lantana-infested landscapes not only are impov-erished as habitats of wildlife but also contribute to human-wildlife conflicts owing to diminished ecosystem services.This paper is a case study of successful eradication and restoration of two lantana-invaded sites in Corbett Tiger Reserve,India. A method for eradicating lantana was developed using knowledge about its ecology, and, subsequently, weed-free landscapes were restored to productive grasslands and mixed woodlands using native species. The restoration ofthese areas to grassland communities has successfully prevented secondary invasions by lantana and other weeds andhas enhanced the habitat quaiity for herbivores whose populations are vital for the survival of top carnivores such astiger {Pantheta tigris corbetti).

Keywords: Corbett Tiger Reserve, ecological restoration, grasslands, lantana (Lantana camera), protected areas

India's first national park, CorbettNational Park, was declared in 1936.

Ever since, there has been a steadyrise in India's Protected Areas (PAs),which include National Parks, Wild-life Sanctuaries, and Tiger Reserves,especially after the enactment of theWildlife Protection Act in 1972. In1988, there were 54 national parksand 372 sanctuaries covering a totalarea of 109,652 km^. There are cur-rently about 597 national parks andsanctuaries in India, and these cover134,572 km-', or 4.74 percent ofthecountry's geographical area. Launchedin 1973 with nine reserves coveringan area of 16,339 km^, the ProjectTiger program has been extended to28 reserves in 18 states, encompassing37,761 km- of tiger habitat. However,the management of these protectedareas is a major challenge with prob-lems of habitat fragmentation, man-wildlife conflicts, poaching, forestfires, and invasive species. In most of

Ecological Restoration Vol. 27. No. 4, 2009

ISSN 1522-4740 E-ISSN 1543-4079

©2009 by the Board of Regents of the

University of Wisconsin System.

these PAs, the peripheral buffer zones(which act as buffers from develop-mental and other human activity) andforest gaps (firelines, forest guard quar-ters, etc.) in the core areas are invadedby non-native invasive species. Tlieworst invasive weeds of PAs includelantana (Lantana cámara)^ bittervine[Mikania micrantha), thoroughwort[Chromolaena sp.), tropical whiteweed{Ageratnm conyzoides), Santa Mariafeverfew {Parthenium hysterophorus),and pignut [Hyptis suaveolem).

Lantana is one ofthe worlds 100worst weeds and is invasive in over 60countries, particularly in tropical andsubtropical regions outside its nativerange in the neotropics (Walton 2006).It is a low, erect or subscandent peren-nial shrub growing to a height of 2.5m or more and forming dense mono-cultures. It reproduces primarily bymeans of copious seed production andalso by vegetative propagation fromcut stems. The juicy pericarp attractsa large number of dispersal agents,among which birds play a majorrole. Tbe plants can grow individu-ally as clumps or form dense thicketscrowding out the native species. It

has demonstrated allelopathic poten-tial and can considerably reduce tbeseedling vigor of native species (Gentleand Duggin 1997, Day et al. 2003).It infests disturbed forest ecosystems,where it dominates understory species,blocks natural succession, reduces bio-diversity, outcompetes native pasturespecies, reduces productivity, increasessoil erosion, and causes illness anddeath of livestock (Sankaran 2007,Walton 2006). The dense under-story of lantana inhibits the seedlingrecruitment of native tree species intbe forest, leading to the depletion ofnative trees (Singh er al. 1995, Sharmaet al. 2005, Bhatt et al. 1994, Kohli etal. 2004). It has invaded the majorityof Indian pasturelands (13.2 millionha) beside other areas, and the cost ofits control is estimated to be USS70per hectare (Singh 1996). Tbercforethe damage from lantana to pasturesalone is estimated to be US$924 mil-lion per year (Pimentel et a!. 2001),and this does not include vast tracts ofdisturbed forest colonized by lantana.

It is not clear how it arrived in India;neither is it known when it reachedCorbett National Park, although there

December 2009 ECOLOGICAL RESTORATION 27:4 467

are anecdotal references to introduc-tion of lantana as an ornamental inthe early 20th century near Nainital.Lantana might have been introducedin India on multiple occasions in late1800s—as an ornamental in Coorg(Hiremath and Sundaram 2005) andas biohedge in Calcutta in the early19th century (Hakimuddin 1929,Pereira 1919). Hakimuddin (1929)also gives an account of its introduc-tion and spread in Nainital Districtaround 1905. Per this account, it wasconfined to hedges and bad sparsedistribution until 1911, but by 1929,lantana had spread from its point ofintroduction to about 40 km in alldirections, forming dense thicketscovering farms, pastures, fallows, andforest areas.

Frugivorous birds are known to playa major role in the dispersal of fleshy-fruited invasives (Gosper et al. 2005,Buckley et al. 2006), and they mayhave dispersed lanrana from villagesand townships, where it was grown asa hedge plant, to the disturbed foresttracLs in tbe periphery of CorbettTiger Reserve (CTR), particularlyintensely grazed habitats. Today, ithas invaded all the plant communi-ties, ranging from gra.sslands to moistdeciduous forests, of CTR—in boththe buffer areas and several patches inthe core area.

Despire a fairly large amount ofpublished literature on the taxonomy,biology, ecology, and control andmanagement of lantana. no successfulstrategy for control and managementis available. The landscapes colonizedby lantana are degraded ecosystemsand pose several challenges for resto-ration, including means of removinginvasives and augmenting ecosystemredevelopment. We have attempted totackle the problem of lantana in CTRwitb a three-step approach: 1) under-standing the biology of lantana withthe objective to develop control mea-sures; 2) developing simple and effec-tive control measures; and 3) evolv-ing a suitable restoration program,keeping in view the requirements ofthe stakeholders and the reference

ecosystems. In this paper, we presentour observations and experiences onthe eradication of lantana, its control,and ecological restoration of lantana-eradicated areas in CTR.

Study AreaCorbett National Park is the oldestNational Park in India, locatedbetween 7S°5' E and 79''5' E longitudeand 29"25' N and 29^40' N latitude(Figure 1) in the state of Uttarakhand.Corbett Tiger Reserve was createdwith the National Park and additionalReserve Forest area as part of the nineTiger Reserves created at the launchofProject Tiger in 1973, with the spe-cific aim of saving the dwindling tigerpopulation. The CTR spreads over anarea of 1318.5 km^ in the foothills(Shivaliks) of northwestern Himalayain the Siwalik-Terai biotic zone of theIndian Himalayan ranges. The reserveis adorned with a unique assemblageof Himalayan flora and fauna and wellknown for a healthy population oftigers. In 1991, an area of 797.72 km^was added as buffer for the CorbettTiger Reserve and includes the wholeof Kalagarh Forest Division and 96.70km^ of Ramnagar Forest Division. TheCTR management is headed by theField Director with headquarters atRamnagar.

The area was declared as a NationalPark not only becatise of its biologicalrichness and as prime tiger habitat butalso because of its hydrological func-tion as a principal catchment area ofthe Ramganga River, a tributary ofthe Ganges. The landscape of Corbettencompasses a network of monsoonaland perennial first-, second-, andthird-order streams having wide riv-erbeds and riverine stabilized islands,floodplains, valleys, gentle hillslopes,hilltops, and marshlands. These dif-ferent habitats support a wide rangeof communities, ranging from seraigrassland communities, shrublands,and moist and dry deciduous forestswith sal {Shorea robusta) as the domi-nant tree to patches of mixed decidu-ous forests, bamboo thickets, and

grasslands. The diversified vegetationtypes support a healthy population ofherbivores and carnivores includingthe riger.

Ihe buffer area of the park includesseveral villages where lantana wasintroduced as a hedge plant to protectthe crop fields from herbivores. Overtime since its establishment, there hasbeen an increase in human-wildlifeconflicts, due mainly to large livestockmaintained by the few villages locatedin rhe buffer area. There were frequentinstances of catde grazing within tbepark areas and thus competing withother herbivore populations. Iherewere also several instances of wild-life raiding the agricultural fields andcattle takings by tigers and leopards.As per government policy at the time,there was pressure to expand the PAand to relocate the villages with theirgrowing livestock herds. As villageswere relocated from within the PA toelsewhere, large tracts of lands oncecultivated or settled were rapidlyinvaded by lantana (Sahu and Singh2008) and other weeds. In 1990-93,Dhara and Jhlrna, two villages .situ-ated on the southern boundary of thereserve, were relocated to the Firoz-pur-Manpur area and Ampokhra,situated on the Ramnagar-Kashipurhighway. Another village, Laldhang,located along the southern boundaryof the park is presently being relo-cated. These villages have been relo-cated outside the park to aid conser-vation programs, and the areas thusvacated have been included in CTR asbuffer zones. If was expected that theforest would regenerate; but as yearspassed, no significant forest regen-eration occurred, as the native spe-cies were outcompeted by weeds. Atpresent no villages remain inside theTiger Reserve, but there are 92 villageswithin 2 to 3 km of its boundary. Thehuman population of these villages isaround 65,982, and the livestock pop-ulation around the reserve is 44,416(Ministry of Environment and Forests2001). ITierefore there is persistentpre.ssure on the Park's boundary fromthese villages. The need is strong to

468 December 2009 ECOLOGICAL RESTORATION 27:4

To Lansdowne

To Marchula

CORBETT NATIONALPARK (CORE ZONE)RESERVE FOREST(BUFFER ZONE)

SONANADI WILDUFESANCTUARY

RESERVOIR AND RIVER

ROAD NETWORK

U\LDHANG

Figure 1. Schematic map of Corbett Tiger Reserve (A) and satellite Image of the inset from Google Earth (B) showing the experimental plots atJhlrna and Laldhang.

December 2009 ECOLOGICAL RESTORATION 27:4 469

restore even the buffer zones that areimpoverished by lantana, to meet thepurpose for which they were created.

Dwindling prey base and human-wildlite conflicts are considered theprime causal factors for the dwindlingpopulation of tigers in the wild acrossthe country (Seidenstickerei al. 1999,Johnsingh and Negi 2003, Johnsinghet al. 2004). For a long time the man-agement was not serious about theinfestation of che park with lantanaon the grounds that it serves as a foodbase for some species of birds. How-ever, the park managers have now real-ized that the spread of lantana in theentire park is changing the landscape,threatening its biodiversity, preventingnatural succession, and enhancing thefrequency of human-wildlife conflictsowing to increased crop raiding byherbivores in nearby villages.

Traditional ControiMethods

A number of methods, includingphysical, chemical, and biologicalmethods, have been used to controland manage lantana invasions in dif-ferent countries. Day and colleagues(2003) summarized the status of cur-rent control methods as inadequateand infeasible owing to the sheer sizeof the infestations, low land values,and lack of incentives for control. Thebiology of lantana and lack of site-specific information on its ecologyare major bottlenecks in developingeffective tools for ¡cs miinagemenc. Ithas yet to be recognized, for instance,that habitat degradation has triggeredthe invasion of lantana and that a finalsolution has to come in the form ofhabitat restoration. In India, forestmanagers recognize lantana infestationin PAs as a major problem, and theyare currendy using a variety of controlmethods. The physical methods diathave been in practice for controllinglantana include 1) chopping the mainstem at the base; 2) clipping aerialshoots 10 to 30 cm above the base; 3)burning the clumps; and 4) grubbing(total uprooting).

Some of these methods do bringabout a short-term reduction in lan-tana cover. But with the onset ofmonsoons, profuse coppicing frommultiple growing points triggered bythe suppression of the apical meri-stems takes place, and soon lantanaovershadows the gtound, preventingthe germination and establishmentof other species, particularly natives.Clipping is totally ineffective andonly causes the clumps to spread andbecome hardier. Burning results incoppicing from the buried meriste-matic zone and also increases soilerosion. Burning on a large scale isnormally not resorted to in PAs, as thefire may spread to adjoining areas andbecause some varieties have a climbinghabit and can cause crown fires. Grub-bing, besides being labor intensive,disturbs the soil, which exposes theburied seeds of lantana and leads totheir rapid germination. The conse-quence of most of these methods hasgenerally been vigorous regenerationof lantana.

In another attempt to controilantana in CTR, giant reed {Amndodonax), a perennial grass, was plantedin one of the lantana-infested areas inthe core area of the park after burning.Within flve years, the giant reed-Ian-tana comtnunity became dominantin the area, with both species spread-ing to new areas. It was realized sub-sequently chat giant reed itself is aninvasive species and is not preferredby herbivores. The current spread oflantana all over the CTR demonstratesthe ineffectiveness of the traditionalconcrol methods in use for over twodecades. Our field observations of tra-ditional control measures raised chefollowing cwo questions: Where andhow does proliferation occur after clip-ping, chopping, or burning? What areche modes of seed dispersal in lantanaand how does ic colonize new areas?

To address che cwin challengesposed by lancana invasion—those ofits control and habicat degradation,we selected two relocated village areas,Jhirna and Laldhang, in the bufferzone of CTR as our study sites. At

both sites, 2 ha rectangular plots weredemarcated and initial conditions wererecorded (Figure 2A) and subsequentlyused to develop a control and manage-ment strategy. We studied the fieldbiology and ecology of lantana wich aview to developing an effective controland management strategy. The aspectsinvestigated include 1) magnitude ofthe invasion; 2) distribution of growthpoints (vegetative reproduction); 3)characteristics of the root system; and4) seed production, modes of seeddispersal, and seed germination.

Field Biology andEcology of Lantana

Magnitude of Invasionof Lantana in CTR

A survey of core and buffer zones ofthe park was carried out using strati-fied random sampling to assess themagnitude of invasion and to under-stand the vulnerability of differenthabitat types to iantana. In each habi-tat type studied, a total area of 0.5ha was sampled using 10 m x 10 mquadrats (« = 50) spread randomlyacross the habitat.

We observed that lancana was foundeverywhere except in dense, closedforests and the percent relative den-sity (the ratio of density of lantana tothe density of all plants in the 0.5 hasampled) varied from 0.1 % to 82.3%(Table 1). The stands along forestroadsides comprised only old clumps,and no new saplings or young plantswere observed. Abandoned arable vil-lage land fields also had a high relativedensity (22.9%) of lantana. Except fora few perching trees, no other nativespecies were found in densely infested

areas.

Distribution of Growing Points

We examined the growth partern of25 individual lantana three monthsafter each of the following treatments:1) chopping at about three-fourthsof the height of the plant (leading toabout 25% removal at about 1.5 mfrom the ground); 2) chopping down

470 I* December 2009 ECOLOGICAL RESTORATION 27:4

Figure 2. Photo series of treatment plots: A) massive original colonization by lantana in Jhirna; B) secondary invasion of the same plot after removalof lantana: C) the same plot ot jhirna, two years after restoration with selected legumes and grasses; and D) and after three years of restoration;E) restoration plot at Laldhang showing development of grasslands two years after lantana removal; and F) same plot with grasslands and scatteredshrubby legumes after tbree years. Photos by Suresh

to the base; 3) burning; and 4) conrrol.Growing points are located from rhelowest node to the uppermost node ofthe main stem and lateral branches.The meristematic zone, located at thetransition between stem and root, wasidentified as the active coppicing zonein lantana based on observations sum-marized in Table 2. This zone is usu-ally buried in the soil (up to 10 cmdepth). Coppicing was profuse iromthis zone after burning and choppingat the ba.se, both of which remove al!

the aerial growing points. It is interest-ing to note that the subterranean cop-picing zone was most active when theaerial growing points were removedby complete chopping or burning ofthe aerial biomass and least active innaturally grown plants (Table 2).

The auxiliary shoots arising fromthe lowest nodes of the main stemare usually prostrate in their lowerhalves, and rooting occurs at thenodes on the prostrate portion of thestem. These shoots sometimes become

independent plants when connectionwith the main plant is severed. This isthe only method of vegetative repro-duction we observed. No root suck-ers were observed, and no fragmentsof aerial branches were found to becapable of rooting and developing intoplants.

Root System

Lantana has a shallow root system. Thefirst-formed taproot is stout, tapersgradually, and penetrates up to a depth

December 2009 ECOLOGICAL RESTORATION 27:4 I* 471

Table 1. Density and percent cover of lantana {Lantana cámara) acrossdifferent habitat types. Seedlings were not included in the survey.

LandscapeRiverbedsStabilized riverine islandsFloodplain without grasslandsFloodplain grasslandsForest roadsidesGrasslands with perching treesArable land abandoned by villagersOpen forests with patches of grassesNatural GapsDense closed forests with understoryMarshy flatlandsSurface drainage channels

Density (ptants/m')0.00770.00210.01440.00270.09090.00720.01260.00260.00210.00120.00030.0048

% Cover6.1

2.216.842.26

82.237.18

22.052.82.171.20.1181.12

Table 2. Growth pattern in lantana (expressed as mean number (± SD) ofshoots per coppicing region of the plant (n = 25) under different traditionalcontrol methods applied in mid-]une 2005; observations recorded in mid-September 2005 after three months of monsoon.

Coppicing Region Chopped Choppedfrom base

Burned Control

Nodes in top one-thirdNodes in middle one-thirdNodes in lower one-thirdRoot-shoot transition

18.4 ±8.2 —28.5 ±7.1 —12.5 ±3.8 61.8 ±10.2

— 38.1 ±16.7— 12.4 ±7.2

18.3±5.1 4.3±2.242.7±15.4 1.8±0.6

of 0.2 m. From the upper portion ofthe main taproot, stout lateral rootsarise and some of these spread hori-zontally in the topsoil. All along theselaterals, fine rootlets with copious roothairs are produced. The high invest-ment in shoots rather than the rootsystem suggests that the species utilizesits energy more for vegetative growthand reproduction.

Seed Bioiogy

Lantana reproduces primarily by seedand is a copious seed producer. Theseed is a pyrene (seed with stony endo-carp) enclosed within the sugary, juicypericarp ofthe fruit, v̂ fhich is usuallyblack in color and eaten by birds. Thefruit, which ranges from 4 to 6 mm,is available nearly throughout the year.

The percent fruit set of plants, esti-mated from the average number offruits produced out of the averagenumber of fiowers, was 68.95% inthe Jhirna area. On average, about1,660 individuals per hectare werepresent in a pure stand of lantana.From these empirical observations,

we estimated that over 1.2 millionfruits are produced per hectare bypure stands of lantana in Corbett.This reproductive output highlightsthe tremendous potential that this spe-cies has for spreading into newer areasofthe Park and the propagule pressureit exerts on nearby habitats.

We collected baseline data on dis-persal vectors of lantana because itwas crucial to understand the com-plex patterns that may result from theinteraction between lantana and itsdispersal agents, as shown by earlierstudies on Heshy-fruired invasive plantspecies (Glyphis et al. 1981, Stansbury2001, Renne et al. 2002, Buckley etal. 2006).

So far, we have identified 18 spe-cies of birds foraging on the fruits oflantana. These include common birds,such as the red-vented bulbul {Pycnon-otus cafer), magpie robin (Copsychmsaularis), common myna [Acridotherestristis), and Indian cuckoo {Cuculusmicropterus). All these are generalistfrugivores, and their roie in dispersalof lantana seeds is implicated in some

other studies also (Bhatt and Kumar2001). Therefore lantana has been ableto utilize the generalized mutualisticinteractions in the invaded habitat forits spread and colonization.

Tliese birds disperse the seedsbeneath perching trees, which aremostly trees that provide shelter andhave fruits of dimensions similar tolantana. We prepared a tentative list ofperching trees for the study area andfound large concentrations of lantanaseedlings underneath these trees. Themost common perching tree was ber{Zizyphus niimmulâria), a commonfruit-bearing tree native to this region.The dispersed lantana seeds are alsocarried by water along surface drain-age channels in the rainy season. Seedsdeposited under perching trees andthose carried by water currents usuallygerminate and establish seedlings. Inother words, the initial sites of lan-tana colonization are around perch-ing trees, surface drainage channels,and riverbeds, all of which are usuallyopen habitats. It has been shown thatpassage of lantana seeds through thegut of dispersers increases germinationrates (Mandon-Dalger et al. 2004),as is also the case with several otherinvasive species employing mutualistsfor their spread (Panetta and McKee1997).

Several attributes ofthe species thatenable it to colonize disturbed habi-tats have come to light. Of particularimportance are the ability for vegeta-tive propagation and employmentof generalist frugivores that use theperching trees in the chaurs (grasslandsalong floodpiains) for dispersal. Wealso observed that seeds of lantanafailed to germinate in the absence ofadequate sunlight and moisture. Anyattempt to restore these invaded land-scapes necessitates effective controlmeasures that address the regenera-tion ability ofthe species. The lantanadensity in different landscapes is suchthat any intervention limited to asmall area of the park is likely to beineffective, as propagules would soonarrive from nearby areas. It must alsobe followed up witb restoration using

472 December 2009 ECOLOGICAL RESTORATION 27:4

Figure 3. Stump of lantana removed by cut-root-stock method: A) root-shoot transition and emerging young shoots; and B) inverted clumps oflantana removed by this method, kept for drying. Photos by Suresh Babu

competitive native species that canresist reinvasion by lantana and otherweeds. Following these insights fromfield biology and from other experi-ences in controlling lantana (Day et al.2003), we developed a two-prongedstrategy—eradication and restora-tion—for the problem of lantana InCTR and put this strategy to test atthe 2 ha plots at Laldhang and Jhirna.

Development of a ControlStrategy and the Cut-Root-Stock Method

Based on our studies of lantana inCTR, a simple, cost-eifective physi-cal method has been developed forthe removal and for long-term man-agement of lantana infestation. Themethod was designated as Cut Root-Stock (CRS) method, as it involveda cut below the root-stetn transitionzone (Figure 3A) identified as theactive zone of proliferation for lantana(CEMDE 2007, Pathak et al. 2006).This method involved choppingthe main taproot 3-5 cm below theground surface with a few hits, usinga kudaal (a locally available heavydigger) with a long handle. Then theclumps were lifted and placed upsidedown to ensure that all the vegeta-tive parts dry up (a minimum of 2-3

months). These dried clumps werecollected in a small area for burning(Figure 3B). Subsequendy, perchingtrees were located and the saplingswere removed from under theiir can-opies and along the nearby surfacerunoff zone. Continuous monitoringwas required in the area where lantanawas eradicated, including manuallyuprooting seedlings of lantana, if any,for a year. The best time to eradicatelancana was in peak summer seasonwhen lantana clumps bear few or nomature fruits. We realized that untillantana is removed from all areas con-tiguous with the experimental plots,the chances that the propagules wouldsoon arrive was very high. Encour-aged by the initial results of the CRSmethod, and with the objective tocompletely remove lantana from allareas of CTR, the management under-took massive scaling up, during whichlantana was removed in different partsof CTR including Dhikala, Gairal,Sarpadhuli, and Bijrani Ranges.

The main advantage of the CRSmethod is that it does not involvepruning of aerial branches or dig-ging and thus is less labor intensive.ITiis method also results in very littleregeneration of lantana by vegetativepropagation and fVom seed banks. Asingle laborer can remove, on average.

about 20-50 clumps per day and theexpenditure per hectare is about INR2,000-4,000 (US$40-$80)—incontrast to grubbing, where a singlelaborer removes only 8-10 clumps perday and the expenditure range is INR8.000-10,000 (US$170-$200) perhectare. However, the cost of removalper hectare varies according to the ageof the clumps, the terrain, and thedensity of lantana.

As we observed in Corbett, simpleremoval of lantana by CRS methodwas not enough. If cbe land wasleft vacant after eradication, it wassoon subjected to secondary invasionby invasive weeds like Santa Mariafeverfew [Parthenium hysterophorus).Senna tora, S. occidentalism llima {Sidacordifolia), wireweed (5. acuta), andalso lantana (Figure 2B). Thereforewe concluded that merely removinglantana will not solve the problem ofecological degradation of these areasand it is necessary to develop a suitablerestoration program for these weed-free areas.

Restoration of Weed-Free Landscapes

Our restoration technique consistedof four essential steps. The firstinvolved the identification of reference

December 2009 ECOLOGICAL RESTORATION 27:4 473

ecosystems by field ecological surveysin relatively undisturbed areas of thePark with similar ecological settings.After transect surveys (Krebs 1989) ofthe areas around Jhirna and Laldhang,v/e concluded that both areas are simi-lar to existing chaurs nearby and thatit is most appropriate to develop these"lantana eradicated landscapes" intograsslands and mixed woodlands. Afterreferring to literature on the flora ofthe foothills (Babu 1977, Pant 1976),we identified 28 grass species and 15legume species that have high for^evalue to herbivores, like chital {Axisaxis) and sambar {Cervm unicolor),that form the prey base for carni-vores and also play a role in the earlysuccession In chaurs.

The second step was to define targetconditions, and for this, the identi-fication of the requirements of thestakeholders (in this case the parkmanagement) was a prerequisite. "Theprime concern of the park manage-ment was to augment the dwindlingprey base for top carnivores like tigers.This could be achieved through thedevelopment of grasslands and mixedwoodlands that would supplement theforaging requirements of herbivores.Since both Jhirna and Laldliang arein the tourist zone of the Park, anindirect benefit was the ecotourismpotential of these areas.

The third step was the developmentof field nurseries and a sequentialintroduction of selected species intothe lantana-free landscapes. A fieldnursery was established 3 km awayat Dhela, which had perennial watersupply from a stream. Selected grassspecies and their seeds were collectedfrom the buíFer areas within 5 km ofthe plots, along with several legumi-nous species that are found locally.Propagules were transferred from thisnursery to both the sites. Transplantsof grasses, consisting of a portionof the rhizome having one to threeculms, were planted at random witha spacing of 8 to 10 cm. Saplings oflegumes were raised in poly-bags andwere planted on the plots. In the sub-sequent year, the seeds of grasses and

legumes were mixed with farmyardmanure and soil, made into pellets,and broadcast directly on the plots. Amix of annual and perennial grasseswas used, avoiding several species ofgrasses with invasive tendencies likecogongrass {Imperata cylindrica) andgiant reed {Arundo donax). The den-sity of tall grasses was kept low andhigh densities of nutritious legumes,like several species of indigo {Indigo-

fera spp.), flemingia {Flemingia spp.),Aiylosia, and ticktrefoil {Desmodiumspp.), were maintained in the initialstages. Subsequently, saplings of sev-eral native trees, particularly legumi-nous trees like albizias [Alhizia spp.),acacias {Acacias\i^.), Indian rosewood{Dalhergia spp.), fruit-bearing shrubssuch as Zizyphus, and trees such asalangium [Alangiurn javanicum),Assyrian plum {Cordia myxa), cham-ror {Ehretia laevis), and Syzygitim wereintroduced in scattered patches. Plant-ings were carried out between July andOctober (during the monsoon) everyyear. Legumes were introduced at six-month intervals, as initial mortalitieswere high.

The fourth and the most importantstep involved monitoring and follow-up activities such as manual weedingunder perching trees and along drain-age channels. A list of the perchingtrees (candidate lantana dispersers)based on field surveys was provided tothe field staff of the forest department.The most common perching trees inthe area are alangium, Assyrian plum,sacred garlic pear {Crateva religiosa)^chamror, emblic {Phyllanthns einhlica)yfigs {Ficus spp.), gardenia {Gardeniatúrgida), glycosmis {Glycosmis pen-taphylU), Miliusa velutina, Java plum{Syzygium cumini), common jujube{Zizyphus jujuha [= Z. zizyphus]),and jhar beri {Zizyphus nummuiaria).This gready facilitated monitoring andfollow-up activities, even without thepresence of experts on-site to identifythe species. The park management cir-culated this list to all the forest rangeswhere the scaling up ot lantana con-trol and management was undertaken.Lantana seeds buried under clumps

removed by the CRS method failedto germinate because the restoredcommunity strongly competed forlight, and also owing to the crowd-ing effect. Continuous monitoring ofthe natural and manmade forest gapsand landscapes that harbor perchingtrees, riverbeds, and surface drainagechannels for seedlings of lantana andtheir uprooting are essential for thesuccessfiil eradication of lantana.

Results and Outreach

The two plots established at Jhirnaand Laldhang, using the integratedcontrol and restoration strategy out-lined above, are now used for dem-onstration of the technique. At bothsites, the original relative density oflantana clumps varied from 80% to100% and were devoid of native spe-cies except for a few scattered perchingtrees. Within two years of the restora-tion program at CTR (June 2005 toJune 2007), both demonstration plotsharbored grassland communities withannual and perennial native grassesand legumes. As of August 2008, seed-lings of several other native species arealso found as a part of the develop-ing community, suggesting that thehabitat is improving (Figure 2C-F).The recorded increase in the numberof species and the greater proportionof grasses and legumes in both plotscompared to the initial conditions areof particular importance (Figure 4).Tlie control areas had fewer species,having experienced some harvestingas they are near the relocated villages.It seems plausible that some of thegrass species may not persist over thelong term on account of competitiveexclusion.

As of August 2008, more than150,000 grass clumps and morethan 5.000 saplings and seedlings oflegumes were planted in CTR in areaswhere lantana had been removed,covering an area of more than 1,600hectares. Much of this effort was madepossible only by active involvement ofCorbett management. At present thefollow-up activities involve monthly

474 December 2009 ECOLOGICAL RESTORATION 27:4

removal of seedlings, if any. Alreadythere are very few seedlings arrivingon these plots, and we believe that asnearby sources of seeds are removedand effective compecitors like grassesand legumes are present in adequatequantities; a reinvasion by lantana isunlikely.

Ihe tree saplings have high survivalrates (measured annually) because ofmodification of soil conditions andthe habiciu by short grasses such asbuffelgrass {Pennisetum ciliaris), sabai-grass [EuLiliopsis binata), tanglehead[Heteropogoti contorttts), watercrowngrass {Pííspalidium scrobiculûris), Ber-mudagrass (Cynodon dactylon), andsmutgrass {Sporobolus indicits), as wellas leguminous species like salpani{Desmodium gangeticttm), threeflowerticktrefoil {D. triflorum)., Flemtngiahracteata, R fruticosa, F. macrophylla,Indigofera astragalina., and rough-hairy Indigo (/. hirsuta). The plotswere browsed by chita! and there wassome mortaliuj'. but it was not a .seriousimpediment because herbivore num-bers were small, as most of the wildlifehad moved away from this area since ithad recently been a settlement.

Both experimental plocs are nowfully covered with a carpet of sev-eral grass species and legumes, withscattered trees (Figure 2C—F). AtLaldbang, the CTR managementhas scaled up the restoration plot to64 ha. In order to prevent repeatedencroachment and cattle grazing fromnearby villages, the management haserected an electric fence around theentire plot until the grassland—mixedwoodland community is established(at least another two years). Despitethis, droppings of chital and wild boar{Sus scrofa) have been observed in thefenced plot. Tiie Jhirna plot was leftunfenced because there was no threatof cattle grazing, lliis plot is now fre-quented by large herds of browserslike chital and sambar, which form themajor prey base for carnivores. So farthere have been three recorded tigersightings at Jhirna in the vicinity ofthe restoration plot.

Initial After 3 years

Laldhang

Figure 4. Comparison of species richness at the the two study sites following restoration. Controlsites wer« forested areas in the proximity of the experimental plots. Data on species occurencewas collected using ten 100 m- quadrats at each of the sites.

Several workshops and hands-ontraining programs were also held toenable the foresters and park manag-ers CO scale up the lantana control andrestoration measures, with field visitsto the plots and demonstration of thetechnique employed. Various aspectsof this model, including the costs andbenefits of clearing lantana and thenconverting che landscape into produc-tive grasslands and woodlands, werehighlighted to gain wider acceptancefor the model. A simple manual wasdeveloped and widely circulated toaid the foresters in removing lantanaby the CRS method and to identifyche grasses and legumes used in tbeCorberc model.

Subsequently, noc oniy the CTRmanagement buc also other PA manag-ers of Utcarakhand State have adoptedthis model and are actively engagedin this program at Rajaji NationalPark, Landsdowne Forest Division,and Nainital Forest Division. Forestdepartments of other states of Indiahave also been convinced by the effec-tiveness of rhe approacK and havealready removed lantana from severalhundred hectares of forest land, partic-ularly around Panchmarhi BiosphereReserve in Madhya Pradesh andKalesar National Park in Haryana.

One of the major reasons for thepopularity of the CRS method hasbeen economics and the fact that

the restored patches do noc requireaftercare like watering, since all thespecies used are local. In CTR, theapproximate cost of removal of lancanafollowed by restoration as describedranged from INR 6,000-9.000(US$125-$ 187) per hectare. How-ever, across different PAs of Utta-ranchal where this strategy was sub-sequently followed, this method ofrestoration cost between INR 4,000-10.000 (US$83-$208) per hectare.The range in cost is explained by thekind of terrain (rocky, slopes, alluvialsoils, etc.) and the age of the clumps.

Implications forManagement andLessons Learned

The PA network in India is the lastrefuge for a large number of threatenedand endangered flora and fauna. Itprovides a variety of ecosystem serviceschar support the enormous livelihoodrequirements of the country. One ofthe major problems faced by manyPAs in India today is habitat degrada-tion and fragmentation and subse-quent colonization of these disturbedhabitats by invasive species. The vasttracts of forest and community landscovered by invasive weeds pose twinchallenges: control and eradication ofweeds and restoration of the weed-freelandscapes to prevent fiarther invasions

December 2009 ECOLOGICAL RESTORATION 17-A 475

and improve habitat conditions. In

this study, we examined the biology

and ecology of lantana in order to

understand its invasive nature, and

based on that understanding, we

developed methods to eradicate lan-

tana infestation and restore the land-

scape. However, as we learned with

our experience in CTR, the ultimate

success of such a program banks heav-

ily on the conviction and motivation

of stakeholders. Landscapes degraded

by invasive species are ofi:en extensive,

and even as weeds are eradicated from

one area, the propagule pressure can

persist irom other areas. A well-coor-

dinated removal program followed by

restoration at suitably large scales is

imperative to tilt the scales in favor of

native species (D'Antonio and Meyer-

son 2002). A strong determination by

the PA managers was crucial for the

success of such a program.

The benefits of lantana eradication

and subsequent restoration to grass-

lands in CTR are many: enhanced

biological productivity, particularly in

terms of palatable species of grasses

and legumes; greater retention of soil

moisture; prevention of soil erosion;

enrichment of native biodiversity;

increased frequency of wildlife sight-

ings; and enhanced recreational values

since CTR is a favorite place for eco-

tourists and ornithologists (Sankaran

2007).

We anticipate that our observations

on lantana management in CTR and

the success of the Corbett model will

help in developing a similar manage-

ment strategy for other PAs, which

are presently under threat of lantana

and weeds like Santa Maria fever-

few, bittervine, and tropical white-

weed. Any restoration program for

a similar setting would also need to

be informed by a variety of factors

such as the dispersal mechanism of

the invader, soil, climate, the pres-

ent ecosystem condition, condition

of neighboring ecosystems, and most

importantly, the requirements of the

stakeholders. As is the case with our

earlier experiences on degraded land-

scapes in the Himalayan region (Babu

et al. 1990, Jha et al. 1994, jha et

al. 1995), the present study further

emphasizes that although the broader

theme of restoration is applicable for

any ecosystem, the approaches for res-

toration of degraded ecosystems are

inherently site and case specific. With

the mounting pressure on our natural

resources, more and more ingenious

methods will have to be developed to

restore the ecologically depauperate

landscapes created by invasive species.

AcknowledgmentsThis work has been carried out under theCentre of Excellence Programme of theMinistry of Environment and Fotests, Gov-ernment of India. We would like to thankMr. K.B. 'Iliampi, ICF, and Mr. DebasisJana, AIGF, of the National Afforestationand Eco-development Board, Ministry ofEnvironment and Forests. Government ofIndia, for constant support and encouragc-menr. We would also like to acknowledgethe extensive suppott we received from theUttarakhand Forest Department, the GhiefWildlife Warden, Dr. Srikant Chandola;then Director of Corbett Nationai Patk andTiger Reserve Mr. Rajiv Bhartari and histeam; and the present Director, Mr.VinodSinghal. Tlie assistance provided by Mr.Pankaj Kumar and Mr. Harkesh Singh isalso acknowledged.

ReferencesBabu, C.R. 1977. Herbaceous flora of Deh-

radun. New Delhi: Publications andInformation Directorace, Council ofScientific Ô£ Industrial Research.

Babu C.R., P.K. Jha and S. Nair. 1990.Biological reclamation of derelict lands:Problems and prospects. Pages 327—.335 in B.B. Dhar (ed). EnvironmentalManagement of Mining Operations, NewDelhi: Ashish Pubiishing House.

Bhatt, D, and A. Kumar. 2001. Foragingecology of red-vented bulbul Pycnon-otus cafer in Haridwar, India. Eorktail17:109-110.

Bhatt, Y.D.. Y.S. Rawat and S.P. Singh.1994. Changes in ecosystem function-ing after replacement of forest by lan-tana shrubland in Kumaon Himalaya.journal ofVegetation Science 5:67-70.

Buckley. Y.M., S. Anderson, C.P. Catterali,R.T. Corlett, T. Engel et al. 2006. Man-agement of plant invasions mediated byfrugivore interactions,Ecology 43:848-857.

Centre for Environmental Managementof Degraded Ecosystems (CEMDE).2007. Annual Report 2006-2007. NewDelhi: Centre of Excellence Programme.Ministry of Environment and Forests,Government of India.

D'Antonio, C. and L.A. Meyerson. 2002.Exotic plant species as problems andsolutions in ecological restoration: a syn-thesis. Restoration Ecology 10:703-713.

Day, M., CJ. Wiley. J. Playford and M.PZaiucki. 2003. Lanrana: Current man-agement status and future prospects.Australian Centre for InternationalAgricultural Research Monograph 102,

Gentle, C.B. and J.A. Duggin. 1997. Alle-lopathy as a competitive strategy in per-sistent thickets of Lantana catnara L.in three Australian forest communities.Plant Ecology \32:%5~^5.

Glyphis, J.P., S.J. Milton and W.R. Sieg-fried. 1981. Dispersal oí Acacia cyclopsby birds. Oci-o%tf 48:138-141.

Gosper. C.R., C D . Stansbury and G. Viv-ian-Smith. 2005. Seed dispersal offleshy-fruited invasive plants by birds:contributing factors and managementoptions. Diversity & Distributions 11:549-558.

Hakimuddin, M. 1929. Lantana in northernIndia as a pest and its probable utilityin solving the cowdung problem. IndianForester 5C>-AQ'>-A\Q.

Hiremath, A.J. and B. Sundaram. 2005. Thefire-lantana cycle hypothesis in Indianforests. Cofiservation Society 3:26—42.

Jha. P.K., S. Nair and C.R. Babu 1994.Development of an inexpensive legume-Rhizobium inoculation technologywhich may be used in aerial seed-ing, journal of Basic Microbiology 34:231-243.

Jha, PK.. S. Nair, M.C. Gopinathan andC.R. Babu. 1995. Suitability of rhizo-bia-inoculated wild legumes Argyrolo-bium jiaccidum.. Astragalus paveolens,Indigofera gangetica d.nd Lespedeza steno-carpa in providing a vegetational coverin an unreclaimed limestone quarry.Plant and Soil 177:139-149.

Johnsingh. A.J.T. and A.S. Negi. 2003.Status of tiger and leopard in Rajaji-Cor-bctt Conservation Unit, northern India.Biological Conservation 111:385-393.

Johnsingh, A.J.T, K. Ramesh, Q. Quresh),A. David, S.P. Goyal et al. 2004. Con-servation status of tiger and associatedspecies in the Terai Arc Landscape.India. Wildlife Institute of India ReportRR-04/001.

Kohli, R.K., K.S. Dogra. D.R. Batish andH. Pal Singh. 2004. Impact of invasive

476 December 2009 ECOLOGICAL RESTORATION 27:4

plants on ihc struccure and composi-tion of natural vegetation of northwest-ern Indian Himalayas. Weed Technology18:1296-1300.

Krebs, C.I.1989. Ecological Methodology. 2nded. Menlo Park CA: Addison-Wesiey.

Mandon-Dalger, I., R Clergeau, J. Tassin,l.N. Rivière and S. Gatti. 2004. Rela-cionships between alien plants and analien bird species on Reunion Island.Journal of Tropical Ecology 20:6^5-642.

Ministry of Environment and Forests. 2001.ProjectTiger: Status report. New Delhi:Government oflndia.

Panctta, ED. and J. McKee. 1997. Recruit-ment of the invasive ornamental Schi'nus terebinthifolius is dependent uponfrugivores. Australian Journal of Ecology22:432-438.

Pant, P.C. 1976. Plants of Corbett NationalPark, Uttar Pradesh. Journal of theBombay Natural History Society 73:287-295.

Pathak, N., S. Basak. A. Ij)ve and S. Babu.2006. Workshop manual for eradicationof invasive weeds and habitat restora-tion. Centre of Excellence Programme,Centre for Environmental Managementof Degraded Ecosystems, University ofDelhi.

Pereira, W.E. 1919. Lanrana in the Mathworking circle in Savantavadi stateforest. Indian /wrcíííT 46:188-193.

Pimentel, D., S. McNair,j.Janecka,J. Wight-man, C. Simmonds et al. 2001. Eco-nomic and envi ron men till threats of alienplant, animal, and microbe invasions.

Agriculture, Ecosystems & Environment84:1-20.

Renne, I.J., W.C. Barrow, L.A.J. Randalland W.C. Bridges. 2002. Generalizedavian dispersal syndrome contributes toChinese tallow tree (Sapium sehifentm,Euphorblaceae) invasiveness. Diversityand Distributions 8:285-295.

Saliu. P.K. and J.S. Singh. 2008. Structuralattributes of lancana-invaded forest plotsin Achanakmar-Amarkantak BiosphereReserve, Central India. Current Science94:494-500.

Sankaran, K.V. 2007. Lantana cámara. Inva-sive Species Fact Sheet. Asia-PacificForest Invasive Species Network, Foodand Agricultural Organization of theUnited Nations (PAO). wviAV.fao.org/forestry/media/13375/1/0

Seidensticker, J., S. Christie and P. jack-son. 1999. Tiger ecology: Understand-ing and encouraging landscape pacrernsand conditions where tiger.s can per-sist. Pages 55-60 in J. Seidensticker,S. Christie and P Jackson (eds). Ridingthe Tiger: Tiger Conservation in Human-Dominated Landscapes. Cambridge:Cambridge University Pres.s.

Sharma, G.P., A.S. Raghubanshi and J.S.Singh. 2005. Lantana invasion: An over-view. Weed Biology and Management5:157-167.

Singh, A., V.S. Reddy and J.S. Singh. 1995.Analysis of woody vegetation of Cor-bett National Park. India. Plant Ecology120:69-79.

Singh, S.P. 1996. Biological control. Pages88-116 in R.S. Paroda, and K.L.Chadha (eds), 50 Years of Crop ScienceResearch in India. New Delhi: IndianCouncil of Agricultural Research.

Stansbury, C D . 2001. Dispersal ofthe envi-ronmental wcc¿ bridal creeper Aspar-agus asparagoides by silvereyes Zoster-ops lateralis in south-western Australia,fma 101:39-45.

Walton, C. 2006. Lantana cámara (shrub).Global Invasive Species Database. Inva-sive Species Specialist Group (ISSG) ofthe IUCN Species Survival Commission.www.i.«g.org/database

Suresh Babu, Research Scientist, Centnof Excellence Progratn. Center for Envi-ronmerttal Management of DegradedEcosystems, University of Delhi, India—110007, sbabu.cemde@grnail.com

Amit Love, Research Scientist, Centre ofExcellence Program. Center for Environ-mental Management of Degraded Ecosys-tems, University of Delhi, India—110007

Cherukuri Raghavendra Babu, EmeritusProfessor, Centre of Excellence Program,Center for Environmental Management ofDegraded Ecosystems, University of Delhi,India—110007

December 2009 ECOLOGICAL RESTORATION 21-A Ml

Copyright of Ecological Restoration is the property of University of Wisconsin Press and its content may not be

copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.