Grand Valley State University ScholarWorks@GVSU

Peer Reviewed Publications Biology Department

Fall 11-1998 Deforestation in the Tropics: Reconciling Disparities in Estimates for India Shaily Menon Grand Valley State University, [email protected]

Kamaljit S. Bawa University of Massachusetts - Boston

Follow this and additional works at: https://scholarworks.gvsu.edu/biopeerpubs Part of the Biology Commons, Forest Sciences Commons, and the Terrestrial and Aquatic Ecology Commons

Recommended Citation Menon, Shaily and Bawa, Kamaljit S., "Deforestation in the Tropics: Reconciling Disparities in Estimates for India" (1998). Peer Reviewed Publications. 12. https://scholarworks.gvsu.edu/biopeerpubs/12

This Article is brought to you for free and open access by the Biology Department at ScholarWorks@GVSU. It has been accepted for inclusion in Peer Reviewed Publications by an authorized administrator of ScholarWorks@GVSU. For more information, please contact [email protected]. Royal Swedish Academy of Sciences

Deforestation in the Tropics: Reconciling Disparities in Estimates for India Author(s): Shaily Menon and Kamaljit S. Bawa Source: Ambio, Vol. 27, No. 7 (Nov., 1998), pp. 576-577 Published by: Allen Press on behalf of Royal Swedish Academy of Sciences Stable URL: http://www.jstor.org/stable/4314794 Accessed: 08/07/2009 12:46

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=acg.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

Allen Press and Royal Swedish Academy of Sciences are collaborating with JSTOR to digitize, preserve and extend access to Ambio.

http://www.jstor.org Synopsis Deforestation in the Tropics: Reconciling Disparities in Estimates for India

Widespread deforestation of tropical forests estimation of deforestation in the Legal on forest ecosystems for their subsistence is expected to have profound global Amazon region of Brazil (4-6). (8). However, consistent estimates of de- consequences. If deforestation continues Here we examine recentdisparate estimates forestation rates in India are lacking. unabated, we are likely to experience altered of deforestationobtained for India.We discuss Deforestation is defined here as forest loss patterns of climate and distribution of the sources of disparity and the implications resulting from clearcutting and conversion of biodiversity (1, 2). Agricultural and econom- of inaccurate estimates and suggest ways in forests to other land-use types. Several work- ic sustainability is very much dependent which future attempts at estimating de- ers have attempted to estimate deforestation upon the understanding of the magnitude forestation might reconcile the disparity. rates for the entire country and individual of habitat degradation and the development Despite the importance of deforestation and states, but these estimates are widely dis- of strategies to contain such degradation. its consequences, no attempt has been made parate. It is also important to note that the Two serious consequences of deforestation to reconcile the different estimates obtained rates of deforestation generally do not cap- are the emission of greenhouse gases and for India. turethe widespreadrate of forest degradation, the loss of biodiversity. Deforestation and which is the gradual erosion of naturalvege- degradation of tropical forests can con- tation resulting in reduced canopy cover and tribute to the rise in atmospheric gases such DISPARITYIN DEFORESTATION altered species composition. The Forest as carbon dioxide CO2 (3). Accurate esti- ESTIMATESFOR INDIA Survey of India estimates that 39% of Indian mates of deforestation and forest degradation Nowhere in the world, perhaps, is the forests have a canopy cover between 10% are necessary to provide the basis for the pressure on land so intense as in India, or and 40% (9). inventory and amelioration of CO2and other the consequences of land-use change for Two noteworthy attempts at estimating greenhouse gas emissions. Furthermore, humanity as far reaching as in South Asia. deforestation during the last decade (1981- because tropical forests are storehouses of India's 900 mill. people make tremendous 1990) in India by FAO (10) and NRSA biodiversity and genetic resources and the demands on the approximately 3 287 000 (reported in 11) differ widely. The estimate losses of these resources are largely kM2 of land area, and consequently extract a by FAO (10) was part of a global assess- irreversible,precise measuresof deforestation heavy toll in the form of land degradation, ment of forest resources and deforestation and forest degradationrates are indispensable loss of soil nutrients, and reduction in forest in tropical countries. This study estimated as a basis for the assessment and monitoring cover. Deforestation in India is of particular an annual deforestation rate of 0.6% be- of biodiversity loss. concern because of the subcontinent's specta- tween 1981 and 1990 for India. The second Although there is general agreement that cular wildlife, unique flora, and high con- estimate by NRSA described in Ravindranath deforestation of tropical forests has in- centration of wild relatives of domesticated and Hall ( 11) suggests that India's total forest creased markedly since the early 1970s, ac- plants and animals (7). Moreover, the sus- areadeclined by only 0.04% annuallybetween tual estimates are fraught with uncertainty. tained productivity of India's agriculture, 1982 and 1990. Estimates of deforestation rates for the same plantations of perennial crops, and forest Details pertaining to the two studies, their region often differ widely leading to con- sector is dependent upon forests that con- methodologies, definitions used, and results fusion and disagreement about appropriate serve soil, nutrients, water, and genetic are presented in Table 1 along with a mention responses (4). For example, several authors resources. India's forests also directly sup- of associated caveats. A major source of have commented on discrepancies in the port approximately 50 mill. people that rely discrepancy is the difference in definitions of

Table 1. Comparison of two deforestation estimates. FAO Estimate NRSA Estimate Reference FAO. 1993. Forest Resources Assessment, 1990. Ravindranath and Hall. 1994. Ambio 23:521-523.

Time period 1981-1990 1982-1990

Data Sources Forest cover: Satellite Imagery (Landsat MSS and Forest cover and plantation: Satellite Imagery TM, IRS) at sample locations and extrapolation by (Landsat MSS and TM) modeling Plantation: Country reports (tabular data)

Scale 1:1 million (1981-83) 1:250,000(1985-87,1987-89,1989-91)

Ground-truthing Yes at selected sample locations Yes

Forest definition Natural forest with >10% crown cover Natural forest + tree plantation with >10% crown cover and contiguous over 25 ha Plantation definition Artificial stands for production or wood, Trees >10% crown cover and contiguous over 25 ha fuelwood, and non-wood products Caveats Imagery evaluated only at sample locations then Proportion of tree plantation induded as forests is extrapolated unknown

Total forest area (1981) 56.5 Mha 64.2 Mha

Total forest area (1990) 51.7 Mha 64 Mha 70.6 Mha (forest + plantations)

Annual deforestation 339,000 ha (0.6%) 23,750 ha (0.04%)net deforestation estimate

Annual afforestation 1981-1990:1.44 Mha Not reported reported (plantation)

576 C)Royal Swedish Academy of Sciences 1998 Ambio Vol. 27 No. 7, Nov. 1998 http://www.ambio.kva.se forests and plantations. The FAO study level studies. We recommendthat various Indiaare slowly beginningto capitalizeon definedforests as "ecosystemswith a mini- processes (deforestation,afforestation, de- the opportunityoffered by the high quality, mumof 10 percentcrown cover of treesand/ gradation,regeneration) and vegetation types highresolution imagery from Indian satellites or bamboos,generally associated with wild (natural forest, plantations) should be (IRS-lA, IRS-IB, and IRS-IC). IRS 1-C's flora,fauna and naturalsoil conditions,and consistentlydefined so that estimatesfrom WiFs data has a 5-day repeatcycle which not subjectto agriculturalpractices." (10). differentstudies can be compared.In par- increasesthe probabilityof cloud-freedata. Deforestationwas defined as theprocess that ticular,tree plantations should not be included This coupled with its low processingtime "refersto changeof landuse with depletion with naturalforests because they are very requirementwill allow morefrequent "rapid of treecrown cover to less than 10 percent." differentwith respect to their role in con- changeassessment" (17). The LISS-IIIsen- Forestdegradation was definedas "changes servingbiodiversity and performing various sors with its 23.5 m spatialresolution will withinthe forestclass (fromclosed to open ecosystemfunctions. Data sources used in a provideadequate detail for forest mapping forest)which negativelyaffect the standor study should be consistent to the extent at 1:25000 scale (17). The countryhas the site and, in particular,lower the production possible.Gross countrywide tabular data for technology,human resources, and appropri- capacity"(of forests). Degradationis not one vegetationtype shouldnot be compared ateinstitutions to undertakea morethorough reflectedin thedeforestation estimates in the withdata for anothervegetation type derived assessment and monitoring of its forest FAO study. fromsatellite imagery. Consistent methodo- resources.The consequencesof deforesta- The NRSA study, on the other hand, logy should be used on time series spatial tionare severe enough, both ecologically and definedforests as "areasunder natural forest datasources for the sameregion in orderto economically,to warranta concertedand or tree plantationswith > 10% tree-crown obtainconsistent estimates of ratesof change. acceleratedprogram to detect changes in cover and which is contiguousover 25 ha". Analysisof remote-sensingimagery coup- naturalforest cover. Thus, as pointed out by Ravindranathand led with groundtruthing provides one of the Hall (11), "itis difficultto say whatpropor- most effective ways to rapidly determine References and Notes tionof treeplantations are included as forests forestcover, for relativelylarge areas. Time 1. Myers,N. 1989. Deforestationrates in TropicalForests and their ClimacticImplications. Friends of the Earth, in the NRSA assessment."To obtainan idea series analysesof remotelysensed imagery London, UK. of actualor net loss of naturalforest in the in turnallow moreprecise determination of 2. Woodwell, G.M. 1978. The carbon dioxide question. Sci. Am. 238, 34-43. NRSA studywe must examinethose states changesin forestcover. However, biologists 3. Brown, S., Iverson,L.R. andLugo, A.E. 1994. Landuse (such as Orissa,Maharashtra, and Andhra have yet to exploit the full capabilityof and biomass changes of forests in peninsularMalaysia thathad less forestarea in thesecond remotely-sensed for from 1972 to 1982: A GIS approach.In: The Effect of Pradesh) imagery calculating Land-useChange on AtmosphericCO2 Concentrations. time periodcompared with the first. When deforestationrates as well as for estimating Dale, V.H. (ed.). Springer-Verlag,New York,USA, pp. only these statesare considered we find that the carbonbudget and biodiversitylosses. 117-143. 4. Downton,M.W. 1995.Measuring tropical deforestation: forestedarea in Indiadeclined by 497 800 ha For example, although forest degradation Development of the methods. Environ. Conserv. 22, annuallybetween 1986 and 1988 andby 266 and forest regenerationundoubtedly play 229-240. 5. Neto, R.B. 1989. Disputes about destruction.Nature 700 ha annuallybetween 1986 and 1988. importantroles in carbonloss/sequestration (April 13), 338. Thesefigures are much closer to the 339 000 and in biodiversityloss/conservation, their 6. Fearnside,P.M. 1990.The rate and extent of deforestation haof annualdeforestation estimated by FAO. role is as yet unquantified.In forest in BrazilianAmazonia. Environ. Conserv. 17, 213-226. fact, 7. Khoshoo, T.N. 1990. Conservationof Biodiversity in Thus, the value for annualdeforestation of degradation may account for levels of Biosphere.In: Indian GeosphereBiosphereProgramme: 23 750 ha (0.04%)arrived at by the NRSA atmosphericCO2 above that expectedas a Some Aspects. Khoshoo, T.N. and Sharma,M. (eds). Har-AnandPublications, Vikas Publishing House Pvt. studyseverely underestimates forest loss by consequenceof deforestationalone (13). Ltd., New Delhi, India,pp. 183-233. confoundingdeforestation and afforestation. Unfortunately,however, the process of 8. NCHSE (National Center for Human Settlements and Environment). 1987. Documentation on Forest and Onthe other hand, the FAO study explicitly degradationis not easy to assess and moni- Rights. Volume One. National Center for Human provides figures for afforestation (plan- tor from remotely-sensedimagery because Settlementsand Environment,New Delhi, India. tations). The study reported an annual of the difficultyin 9. FSI (Forest Survey of India). 1994. The State of the remotelyidentifying key ForestReport1993. Govemmentof India,Forest Survey afforestationof 1.44 mill. ha between 1981 components of natural surfaces, such as of India,Dehra Dun. and 1990 in India.This estimateof annual greenvegetation, dry vegetation and woody 10. FAO (Food and AgriculturalOrganization). 1993. For- est ResourcesAssessment 1990. FAO, Rome, Italy. afforestationis almost 3 times greaterthan components,soil, rock, and water.Spectral 11. Ravindranath,N.H. and Hall, D.O. 1994. Indianforest the annualdeforestation estimated by FAO. mixtureanalysis allows a meansto separate conservationand tropical deforestation. Ambio 23, 521- 523. Unfortunately,the FAO study used dif- thesecomponents in eachpixel on a remotely- 12. Hall, C.A.S., Tian,H., Qi, Y., Pontius,G. andComell, J. ferent sources of data for estimation of sensed image. Fundamentalspectral com- 1995.Modelling spatial and temporal pattems oftropical forestcover andfor plantations.The former aredefined in termsof or land use change. J. Biogeogr. 22, 753-757. ponents laboratory 13. Houghton, R.A. 1991. Tropical deforestation and was estimatedfrom satellite imagery and field spectra(end-members) of well charac- atmosphericcarbon dioxide. Climate Change 19,99-118. modelingwhereas values for the latterwere terizedmaterials, and image pixels are mode- 14. Bateson, A. and Curtis,B. 1996. A method for manual endmemberselection and spectral unmixing. Remote obtainedas tabulardata from gross country led as mixturesof these end-members(14). Sens. Environ.55, 229-243. reports.However, highly aggregateddata Spectral mixture modeling appearsto be 15. FSI (Forest Survey of India). 1996. The State of the ForestReport1995. Govemmentof India,Forest Survey at the national level do not accurately superiorto conventionalmethods of image of India,Dehra Dun. representactual patternsof land use and classification and can prove valuable in 16. Rasch, H. 1994. Mappingof vegetation,land cover, and biomasschange at the local/regionallevels assessingand monitoring degradation and its land use by satellite-experience and conclusions for futureprojectapplications.Photogram. Engineer. Remote (12). The lackof consistentdata sources and effects. Sens. 60, 265-271. methodology in estimating extent of dif- ForIndia, in particular,the wide disparity 17. Roy, P.S., Dutt, C.B.S., Jadhav,R.N., Ranganath,B.K., Murthy,M. S. R., Gharai,B., UdayaLakshmi,V., Kandya, ferent vegetationtypes in the FAO study in deforestationrates underscoresthe need A.K. andThakeker, P.S. 1996.IRS- 1C data utilization for gives rise to the problemthat their values for a more accurateassessment of the rate forestryapplications. Curr. Sci. 70, 606-613. 18. This paper representscontribution # 55 of a research for deforestationand afforestationare not at which forestsare being lost. To date,the program in Conservation of Biodiversity and the comparable.In other words, the country two yearlyestimates of forest cover by the Environmentjointly coordinatedby the Tata Energy reportof afforestationshould not be com- ForestSurvey of Indiaremain based largely ResearchInstitute and the University of Massachusetts at Boston. The program is supportedin part by The paredwith an estimateof deforestationderi- on visual interpretationof remotely-sensed MacArthurFoundation. ved fromsatellite imagery. images. These estimates show no defor- estationbetween 1989 and 1995 and do not Shaily Menon distinguish between natural forest cover Department of Biology RECONCILINGDISPARITIES Grand Valley State University and plantations(15). Visual interpretation Allendale, Ml 49401 USA Thus,disparity in estimatesof deforestation can give good results for preliminary e-mail: [email protected] rates arisesfrom severalsources, including classificationand is less expensivethan digi- differencesin methodologyand in thedefini- tal methods(16). However,the reliabilityof Kamaljit S. Bawa Department of Biology tionand classification of vegetationand land- theinterpretation can be affectedby thescale Universityof Massachusetts at Boston use types.Such disparities in estimatesof the and quality of hard copy images and can 100 Morrissey Blvd. rate and extent of deforestationunderscore differdue to observer differences (6). Govern- Boston, MA 02125 USA the need for more carefulregional or local ment and nongovemnmentorganizations in e-mail: bawa umbsky.cc.umb.edu

Ambio Vol. 27 No. 7, Nov. 1998 ? Royal Swedish Academy of Sciences 1998 577 http://www.ambio.kva.se