FragmentationinMangroveForestecosystems

NRS534:ECOLOGYOFFRAGMENTEDLANDSCAPES
JusticeCamillusMensah
April29,2013
Fragmentation has been termed as one of the greatest worldwide threats to biodiversity and the
primary cause of species extinction (Laverty and Gibbs, 2007). In generally, fragmentation is the
subdivision of a formerly contiguous landscape into smaller units. It reduces continuity of habitat
andinterfereswithspeciesdispersalandmigration,therebyisolatingpopulationsanddisruptingthe
flowofindividualplantsandanimalsacrossalandscape.Fragmentationisanissueforbiodiversityin
both terrestrial and aquatic (including marine) environments. In marine environment, mangrove
forestsareamongthemostvulnerableecosystems(LavertyandGibbs,2007)
Mangroves are an assemblage of tropical and subtropical halophytes (i.e., salt tolerant) woody
plantswhichgrowinloosewetsoilsofbrackishtosalineestuariesandshorelinesinthetropicsand
subtropics.Globally,theyareknowntobethemostproductiveanduniquecoastalecosystemsthat
support a wide range of goods and services. The ecosystem goods and services that mangroves
provide include protecting the coastline from tidal waves and storm surges; acting as biological
filtersinpollutedcoastalareas;supportingaquaticfoodchains;shieldingalargenumberofjuvenile
aquaticorganisms;andsequestrationofgreenhousegases
Despite their ecological benefits, the health and persistence of mangroves continue to be
threatened.Ithasbeenestimatedthattheworldhadlostatotalofabout35%oftheareaonce
covered by mangroves since the 1940s, and these important ecosystems are still reducing at a
currentlossratesofabout1to3%peryear.Themaincausesoffragmentationinmangrovescould
be natural or anthropogenic. Natural causes of fragmentation in mangroves include natural
disturbances like hurricanes and tsunamis, however most of the disturbance to mangroves are
humaninduced.Anthropogenicfactorsthatcontributetomangrovefragmentationandhabitatloss
includeagriculturalconversion,deforestation,redistributionofwaterresourcesandthegrowthof
humansettlements(SetoandFragkias,2007;LavertyandGibbs,2007).
Similar to many other plants, mangroves have strong relationships with the surrounding
environment.Theoccurrenceofmangrovespeciesatacertainlocationisrelatedtothesurrounding

ecologicalgradientssuchaselevation,tidalinundation,watersalinity,andsoilpH.Inotherwords,
mangrove species are likely to grow within their own niches. This phenomenon, in many cases,
causes striplike patterns (i.e. mangrove zonation) that are usually found in tropical mangrove
forests.Thusmangrovesdependheavilyonthesurroundinghydrologyandthesubstratetherefore
the fragmentation of this ecosystem could have great impacts. It is important to note that
fragmentation of mangroves mimics terrestrial fragmentation more closely than that of other
marinesystems.Asstatedabove,mangrovespeciescreatetheirownnichesinpartsoftheintertidal
landscape where better opportunities than others are provided to fulfill their ecological
requirements. Fragmentation and degradation can however increase the patchiness of the
landscapeintermsofmeetingaspeciesneeds.Likemanyotherecosystems,fragmentationimpacts
speciesprincipallybyreducingavailableresourcesandmicroenvironments.Fragmentedmangroves
also affect the movement and dispersal of propagules and eventually lead to modified species
behavior. Laverty and Gibbs (2007) indicate that three major consequences are apparent as
fragmentation progresses in a landscape. These are decreasing patch size; increased edge effects;
andincreasedpatchisolation.
The consequences of mangrove fragmentation and loss are well documented in the literature.
Mangrovefragmentationissaidtoberesponsiblefordecreasesinabundanceanddiversityoffish.It
increases light penetration, visibility and proximity to open water, reducing protection from
predators.Mangrovelossandfragmentationalsoreducestheresilienceandincreasesthesensitivity
ofthenatural,aswellasthehuman,environmenttostochasticoceanicorclimaticeventssuchas
cyclonesandtsunamis.Ofthe418villageshitbytheDecember2004tsunamialongtheAndaman
coast, only 30, or 7%, were severely devastated. Whereas areas where mangroves have been
fragmentedandlosttoaquacultureorthetouristindustries,thispercentagereachesanestimated
80to100%.Mangrovefragmentationandlosscouldalsoresultinlargescaleerosionandsiltationof
surroundingenvironment.
Anumberofmetricshavebeendevelopedtocharacterizehabitatlossandfragmentation,mostlyin
the terrestrial environment. Some of these metrics have been successfully applied for mangrove
ecosystems due to their similarity to terrestrial systems. This characterization is often based on
categorical or thematic or choropleth maps derived from remotely sensed data. Mangrove
fragmentation is a landscapelevel process in which a contiguous mangrove forest stand is
progressively subdivided into smaller, more dispersed and/or disjunct (i.e., subdivided),

geometricallymorecomplex(initially,butnotnecessarilyultimately),andmoreisolatedpatchesasa
result of both natural processes and human land use activities. This process involves changes in
landscapecomposition,structure,andfunctionandoccursonabackdropofanaturalpatchmosaic
created by changing landforms and natural disturbances. In FRAGSTATS for instance, different
metrics that focus on subdivision of the landscape, isolation, dispersion, shape indices and many
otherscanbecomputedtounderstandtheextentoffragmentationofthemangroveforestinthe
landscape. Though some of these suites of metrics have been criticized for some inherent
limitations, many of them have been used successfully in the mangrove environment to inform
managementandconservationplanninginseveralareas.SetoandFragkias(2007)reportsonhow
they successfully used metrics like the Contiguity index and the Euclidean Nearest Neighbor
distributionstocharacterizemangrovefragmentationinRamsarsitesinVietnam.
Inmanycases,thecharacterizationofmangrovefragmentationandlossisnotanend,butrequired
to help managers make appropriate management decisions. These management decisions often
border around conservation and restoration. In terms of mangrove restoration efforts, either a
natural recovery method or mangrove replanting have been implemented. Lewis and Marshall
(1997) have suggested five critical steps necessary to achieve successful mangrove restoration.
These are to understand the autecology of the mangrove species at the site, in particular the
patternsofreproduction,propaguledistributionandsuccessfulseedlingestablishment;understand
the normal hydrologic patterns that control the distribution and successful establishment and
growth of targeted mangrove species; assess the modifications of the previous mangrove
environment that occurred and currently prevents natural secondary succession; design the
restoration program to initially restore the appropriate hydrology and utilize natural volunteer
mangrove propagule recruitment for plant establishment; and finally, replanting when natural
restorationisnotfeasible.

AnnotatedBibliography

Laverty,M.F.andJ.P.Gibbs.2007.EcosystemLossandFragmentation.Synthesis.American
MuseumofNaturalHistory,LessonsinConservation.Availableathttp://ncep.amnh.org/linc.

Inthissynthesispaper,LavertyandGibbs(2007)dissectthephenomenonofecosystemfragmentation
andlossingeneral.Thoughtheemphasisisnotdirectlyonmangroveecosystems,thepaper
undoubtedlysetsthestageforourdiscussiononfragmentationinmangroves.LavertyandGibbs(2007)
establishedtherelationshipbetweenecosystemfragmentationandloss,andconcludedthat
fragmentationisusuallyaproductofecosystemloss.Habitatlosesintheterrestrial(ForestsTropical
andTemperateand;GrasslandsTropical,Temperate,andTundra)andaquatic(Wetlandsandriverine
systems)environmentwerediscussedwithametaanalysisofearlierworksontheextentofthese
biomesovertime.Theyalsohighlightedsomeofthechallengesassociatedwitheffortstoquantifythe
extentandrateoflossoftheworldsmajorbiomesatvariousscalesandfordifferenttimeperiods.
Otherimportantaspectofecosystemfragmentationthatwerediscussedincludethecausesof
fragmentation(naturalandanthropogenic);theeffectsoffragmentationandrecommendationsforthe
managementoffragmentedlandscapes.Interestingly,thepaperconfirmedthatthoughmangrovesare
intheaquaticenvironment,theymimictheterrestrialfragmentationmorecloselythanthatofother
marinesystems,makingitpossibletoapplymostofthemanagementtechniquesoftheterrestrial
ecosystems.Inmyestimationthisisawellorganizedpaperandisdefinitelyagoodsourceofgeneral
informationaboutthegeneralissueofecosystemfragmentation.

JeriLynnParrent,MatteoGarbelottoandGregoryS.Gilbert.2004.Populationgenetic
structureofthepolyporeDatroniacaperatainfragmentedmangroveforests.Mycol.Res.108
(4):403410(April2004).
ThisoverlyscientificstudyexaminedthegeneticstructureofDatroniacaperatapopulationsfromfour
Panamanianmangroveforestsusingamplifiedfragmentlengthpolymorphism(AFLP)markers.D.
caperata,abasidiomycetefungus,isoneofthedominantpolyporespeciesfoundinneotropical
mangroveforestfragments,whereitislocallyspecializedonLagunculariaracemosa.Theobjectiveof
thestudywastoassessthescaleofgeneticstructureofD.caperataonLagunculariaracemosaamong
fragmentedstandsofmangroves.Thoughverycomplex,thispaperdemonstratedthatdespite
productionofcopiousbasidiosporescapableoflongdistancedispersal,somehomobasidiomycetefungi
maybesusceptibletogeneticisolationduetohabitatfragmentation.

Marlow,J.A.,Harrison,T.,Rastogi,A.,Vu,H.L.,andPost,J.R.Fish.Communitiesin
continuousandpatchmangrovehabitatsonlittleCaymanIsland.

Thispaperisareportofastudythatwasconductedintheredmangrove(Rhizophoramangle)stands
alongthecoastofSouthHoleSoundonLittleCaymanIsland(Anthilles).Inthisstudy,Marlowetal
comparedtheabundance,speciesrichness,andcommunitycompositionoffishcommunitiesinfour
habitattypes:continuousmangrove,fragmentedmangrove,stretchesofsandyshorebetween
mangrovepatchesandboatdocks.Theyobservedthatfishabundanceandspeciesrichnessofbothdock
andsandyshorehabitatswerelowerthaninmangrovehabitats.Althoughtheydidn'tseedifferencesin
thefragmentedandpatchymangroveswithrespecttoabundance,speciesrichness,andcommunity
composition,theyobservedlargefishindividualsincontinuousmangroves,butnotinpatchmangroves.
Theinabilityoftheiranalysistodetectdifferences,inmyopinion,couldbeattributedtofactthatthe
continuousandpatchmangrovehabitatsdifferedinthespatialscaleoftheirmangroverootcoverage.

CsarAlejandroBerlangaRobles,ArturoRuizLunaandRafaelHernndezGuzmn.2011.
ImpactofShrimpFarmingonMangroveForestandOtherCoastalWetlands:TheCaseof
Mexico,AquacultureandtheEnvironmentASharedDestiny,Dr.BarbaraSladonja(Ed.),
ISBN:9789533077499,InTech,
Oneofthemajorthreatstomangrovefragmentationandlossisaquacultureincludinglargescaleshrimp
farming.Theobjectiveofthepresentstudywasanalyzethelandusechangescausedbyshrimpfarming
inthecoastallandscapeofMexico,oneofthemainproducersworldwide,usingremotesensingand
(GIS)toolswithinalandscapechangeframeworktocontributetoabetterunderstandingoftheimpacts
ofshrimpfarmingoncoastalwetlands.ToaccomplishthisBerlangaRoblesetal(2011)performeda
changedetectionanalysisinthreemainsteps.Thefirststepwastoconductashrimpfarmlocationand
inventoryinthestudyarea.ThiswasfollowedbyathematicdataextractionofLandsatTMimagesfrom
1986to1999.Finallyachangeanalysisofthe1986andthe1999thematicmapswereconducted.They
concludedthatthoughshrimpaquacultureinMexicoisnotthemaincauseofmangrovefragmentation,
theindustryisstillfarfromsustainablebecausealmosthalfofthepondareahasresultedinthedirect
removalofothernaturalwetlands.Theyreiteratethat,theentireassociatedinfrastructureinterrupts
localandregionalecologicalprocessbyfragmentationoftheintertidalzone.

SaudaminiDasandJeffreyR.Vincent2009.Mangrovesprotectedvillagesandreduceddeath
tollduringIndiansupercyclone.PNASvol.106no.18.73577360

Thisreiteratestheabilityof"unfragmented"mangrovestoreducethedamagedamagecausedby
tsunamisandtropicalstorms.Inthiscontroversialstudy,DasandJeffrey(2009)showedskepticsthat
mangrovesarenotonlycapableofprotectingcommunitiesagainststormsurgesbutalsotropicalstorms

andeventsunamis.Theobjectivewastoshowthatmangroveswereassociatedwithstatistically
significantreductionsinhumandeathsduringasupercyclonethatstrucktheeasterncoastofIndiain
October1999.Theyanalyzedthenumberofstormrelateddeathsinthe4administrativeunits(tahasils)
inthestateofOrissainIndia.Theytestedthenullhypothesisthat,conditionalonpopulationandother
relevantfactors,villageswithwiderremainingmangrovesbetweenthemandthecoasthadthesame
averagenumberofdeathsduringthe1999stormasvillageswithnarrowerornomangroves.Using
socioeconomicdataofthearea,GISspatialcharacteristicsandcontrollingforpotentiallyconfounding
variables,DasandJeffreyconcludedthatvillageswithwider(contiguous)mangrovesbetweenthemand
thecoastexperiencedsignificantlyfewerdeathsthanoneswithnarrower(fragmented)orno
mangrovesandthathumanimpactsonthemangroveecosystems(i.e.,deforestation)thusaffectedthe
deathtoll.Fewstudiesonmangrovesareconductedtoincorporatenotonlytheecologicaldimensions
withremotesensingGIStechnologyaswellasthehumandimensionsofthearea.Thispaperhappensto
beoneofthefew.

EmilA.Cherrington,BetzyE.Hernandez,NoelA.Trejos,OctavioA.Smith,EricR.Anderson,
AfricaI.Flores,andBessyC.Garcia.2010.TechnicalReport:Identificationofthreatenedand
resilientmangrovesintheBelizebarrierreefsystem.WaterCenterfortheHumidTropicsof
LatinAmericaandtheCaribbean(CATHALAC)
TheobjectivesofthisstudywastomapthecurrentextentofthemangroveecosystemsoftheBelize
BarrierReefSystem,identifyareasofchangeandidentifymangroveareaspotentiallyatriskofbeing
cleared.Takingadvantageofavailableexistingmangroveextentdata,Cherringtonetal(2010)usedthis
pastmangrovedataasbaselinetoconductmultitemporalremotesensingbasedchangedetectionon
Landsatsatelliteimageryfortheyears1980,1989,1994,2000,2004,and2010.Theyanalyzeddifferent
mangroveecosystemparametersanddynamics,includingthemangrovefragmentation.Theydidnot
onlyevaluatemangrovefragmentation,butalsocomputedthefragmentationpotentialofthe
mangroves,whichisoutstanding.TheyusedtheEcosystemPatchIrregularityIndex(EPII)toindicate
howeasyitisforamangrovepatchtobecomefragmentedduetoitsinherentgeometry.

KarenC.Seto,MichailFragkias.2007.Mangroveconversionandaquaculturedevelopmentin
Vietnam:AremotesensingbasedapproachforevaluatingtheRamsarConventionon
Wetlands.GlobalEnvironmentalChange17(2007)486500
ThisworkbySetoandFragkias(2007)emphasizestheroleofremotesensinginevaluatingwetland
habitatsandassessingtheimpactofRamsarconventiononwetlands.Thestudywasconductedintwo
sitesinVietnam;XuanThuyNaturalWetlandReserve(theonlyRamsarsiteinVietnam)andTienHai

NatureReserve.Theobjectiveofthestudywastousetimeseriesanalysisoflandscapepatternmetrics
toassesslandcoverconditionsbeforeandafterdesignationofRamsarstatustomonitorcompliance
withtheConvention.Theyusedfourmetrics(totalmangroveextent;mangrovefragmentation;
mangrovedensity;andaquacultureextent)toevaluatethesuccessoftheRamsarconventioninthe
country.Intermsofmangrovefragmentation,theyusedFRAGSTATStocalculateavarietyoflandscape
metricsassociatedwithhabitatlossandfragmentation.SetoandFragkiassuccessfullyappliedthese
metrics(suchasaveragecontiguityindex,meanandmedianpatchsize,thelargestpatchindexand
EuclideanNearestneighbordistributions)toevaluatethemangrovefragmentationinthe2sitesover
time.

AllanM.StrongandG.ThomasBancroft.1994.Patternsofdeforestationandfragmentationof
mangroveanddeciduousseasonalforestsintheUpperFloridaKeys.BulletinofMarine
Science.54(3):795804.

Inthisstudy,StrongandBancroft(1994)usedtheavailabletechnologyoftheirtimetodeterminethe
changesinmangrovesanddeciduousforestcoverageintheUpperFloridaKeys,fromtheoriginal
condition.WithintheATLASdesktopGISpackage,1991aerialphotographsweredigitizedtomapthe
remainingmangroveanddeciduousseasonalforestsusingminimumsizecriteriaof24mX24m.They
madeassumptionsfortheoriginalmapofthesehabitats.Toestimatefragmentation,theyusedcanopy
closureandpotentialforvegetationregenerationascriteriaforseparationoffragments.Forexample,
adjacentfragmentswithcontinuouscanopycoverageseparatedbyapavedroadwereconsidered
differentfragments.Theyconcludedthatdeforestationhascauseda41%decreaseintheareaof
deciduousseasonalforestanda15%decreaseintheareaofmangroveforest.Fragmentationhas
resultedinmuchofthecurrentforestremaininginsmallparcelswithahighedgetoarearatio.This
paper,asImentionedearlier,tookadvantageoftheavailabletechnologythenwhichisimpressive,but
thereisalotmorethatcouldbedone.Inthefirstplace,theassumptionoftheoriginalmapofthe
habitatsandthechoiceofcanopyclosureasaproxyforfragmentationwilldefinitelyaffectthefigures.

SaifulArifAbdullahandNobukazuNakagoshi.2007.Forestfragmentationanditscorrelation
tohumanlandusechangeinthestateofSelangor,peninsularMalaysia.ForestEcologyand
Management241(2007)3948

ThispaperlooksbroadlyatforestfragmentationinthestateofSelangorinMalaysiawiththeobjective
oftodevelopingasingleforestfragmentationindextodescribethedegreeofforestfragmentationand
tounderstandtherelationshipbetweenforestfragmentationandhumanlandusechange.Thepattern
changeofforestfragmentationanditsrelationshiptohumanlandusewereanalyzedin1966,1981and
1995.Inthisstudy,AbdullahandNakagoshi(2007)developedaforestfragmentationindexbasedona
combinationofthreelandscapemetrics.Thesemetricswerenonforestarea;forestedgeborderedby
humanlanduses;andpatchsizecoefficientofvariation.Theyconcludedthatthisstudyshowedthat,
overthethreetimeperiodsofthisstudy,humanlanduseactivitiesincreasinglyfragmentedtheforestin
thestateofSelangorandthatthisoccurredparticularlyinwetlandlandscape,dominatedbywetland
forestsofpeatswampsandmangroves.TheythereforerecommendthatthestateofSelangorshould
giveprioritytowetlandforestsinconservationstrategiesandsustainablelanduseplanningand
management.Thispaperrevealedthatknowledgeofforestfragmentationthroughasingleindexin
differentnaturallandscapesprovidesapathwayforidentifyingwhichforestedareasarehighly
threatenedformanagementdecisions.

HaiHoaNguyen,CliveMcAlpine,DavidPullar,KasperJohansen,NormanC.Duke.2013The
relationshipofspatialtemporalchangesinfringemangroveextentandadjacentlanduse:
CasestudyofKienGiangcoast,Vietnam.Ocean&CoastalManagement76(2013)12e22

Inthisinterestingstudy,theauthorswereinterestedinthespatialtemporalchangesintheextentand
widthoffringemangroves,andchangesinadjacentlanduse.Thetotalmangroveextentwasestimated
atsevenpointsintimeusingfourLandsatTMimages(1989,1992,2003,and2006)andthreeSPOT
images(1995,2003and2009).Supervisedclassificationwasconductedforimageclassificationand
mappingchangesusingaMaximumLivelihoodClassifierapproach.Changesinmangrovewidthwere
calculatedfor2003and2009usingtheDigitalShorelineAnalysisApplication(DSAS).Itwasinterestingto
seehowtheDSASmodelwasappliedtodetectchangesinextentofmangroves.Thismodelwas
developedbytheUSGSformodelingcoastalerosionandcoastalrelatedchangesovertime.Though
interesting,IthinktheNguyenetal(2013)couldhavebenefitedfromothermodelsothermodelslike
theLandChangemodelerorGeoModtoachievedbetterandmoreresults(ratherthanjustthechanges
inthewidth).

OtherReferences

Lewis,R.R.andM.J.Marshall(1997),Principlesofsuccessfulrestorationofshrimpaquacultureponds
backtomangroveforest,Programa/resumesdeMarcuba97,September15/20,Palaciode
ConvencionesdeLaHabana,Cuba.