Copyright copy 2013 by the author(s) Published here under license by the Resilience AllianceBoissiegravere M B Locatelli D Sheil M Padmanaba and E Sadjudin 2013 Local perceptions of climatevariability and change in tropical forests of Papua Indonesia Ecology and Society 18(4) 13 httpdxdoiorg105751ES-05822-180413

Research part of a Special Feature on Traditional Ecological Knowledge and Global Environmental Change North andSouth Perspectives

Local Perceptions of Climate Variability and Change in Tropical Forestsof Papua IndonesiaManuel Boissiegravere 12 Bruno Locatelli 12 Douglas Sheil 234 Michael Padmanaba 2 and Ermayanti Sadjudin 5

ABSTRACT People everywhere experience changes and events that impact their lives Knowing how they perceive react andadapt to climatic changes and events is helpful in developing strategies to support adaptation to climate change Mamberamoin Papua Indonesia is a sparsely populated watershed of 78 million hectares possessing rich tropical forests Our study comparesscientific and traditional ecological knowledge (TEK) on climate and analyzes how local people in Mamberamo perceive andreact to climatic variations We compared meteorological data for the region with local views gathered through focus groupdiscussions and interviews in six villages We explored the local significance of seasonality climate variability and climatechange Mamberamo is subject to strikingly low levels of climatic variation nonetheless local people highlighted certainproblematic climate-related events such as floods and droughts As our results illustrate the implications vary markedly amongvillages People currently consider climate variation to have little impact on their livelihoods when contrasted with other factorseg logging mining infrastructure development and political decentralization Nonetheless increased salinity of water suppliescrop loss due to floods and reduced hunting success are concerns in specific villages To gain local engagement adaptationstrategies should initially focus on factors that local people already judge important Based on our results we demonstrate thatTEK and an assessment of local needs and concerns provide practical insights for the development and promotion of locallyrelevant adaptation strategies These insights offer a foundation for further engagement

Key Words adaptive strategy coping mechanism deforestation ecosystem services gender seasonality traditional ecologicalknowledge

INTRODUCTIONPeople who depend on natural resources especially thepoorest are often particularly vulnerable to climate variabilityand change (Morton 2007) We need to understand how peopleexperience and respond to such variability to guide climatechange adaptation strategies Rural societies already have in-depth knowledge of local climate variability and change aspart of their traditional ecological knowledge (TEK) ie theirknowledge acquired and transferred through generations(Berkes et al 1995 2000) Here ldquolocal perceptionsrdquo refers tothe way local people identify and interpret observations andconcepts (Byg and Salick 2009 Vignola et al 2010) Eventhough climate change may bring conditions beyond previousexperience local knowledge and perceptions remain thefoundation for any local response

Most studies on perceptions of climate deal with temperatureand rainfall ie amount annual distribution start and enddates (Deressa et al 2009 Fisher et al 2010) Meteorologicaldata are often used to confirm villagersrsquo assessments (Orloveet al 2000 Vedwan and Rhoades 2001 Deressa et al 2009Fisher et al 2010) or refute them (Maddison 2007Bandyopadhyay et al 2011 for long-term perceptions)Previous studies have dealt with perceptions of seasonality(Bryan et al 2009 Bandyopadhyay et al 2011) perceptionsof risks and threats related to climate variability (Grothmann

and Patt 2005 Thomas et al 2007 Adger et al 2009 McCarthy2011 Saroar and Routray 2012) and local knowledge inforecasting weather and adapting to climate (Orlove et al2000) Some authors highlight the need to consider climate ina wider context such as health or policies (Mubaya et al 2012Shackleton and Shackleton 2012) and emphasize theimportance of these broader interactions when developingadaptation projects and policies (Reid and Vogel 2006OrsquoBrien et al 2009) this is also our approach

Our aim is to examine how local knowledge can be used toinform and thus improve projects and policies for adaptationto climate change This work is unusual in examining remotetropical forest communities We contrast scientific and localknowledge about different aspects of climatic variationseasonality climate variability and climate change in sixvillages in Mamberamo a forested area in Papua IndonesianNew Guinea We are particularly concerned with how thesevariations may have caused problems or hardships what theresponses and constraints have been and what can be learnedfrom these that may assist in future adaptation Seasonality isthe cycle of conditions seen over a year when long-termrecords are averaged and climate variability is the deviationfrom the average We focus on interannual variability iehow temperature or precipitation changes from one year toanother including extreme events such as drought or heat

1Centre de coopeacuteration Internationale en Recherche Agronomique pour le Deacuteveloppement (CIRAD) 2Center for International Forestry Research (CIFOR)3Institute of Tropical Forest Conservation (ITFC) 4School of Environmental Science and Management Southern Cross University 5ConservationInternational

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

waves Climate change involves longer term trends fromdecades to centuries

We consider two different categories of local responses toclimatic variations coping and adaptive responses Thedifference between them has been widely discussed in theliterature on development food security and more recentlyclimate change (Davies 1993 Roncoli et al 2001 Berkes andJolly 2002 Eriksen et al 2005 Smit and Wandel 2006Fabricius et al 2007 Eriksen and Silva 2009 Ravera et al2011 Brockhaus et al 2013) Coping responses ormechanisms are unplanned reactive and short-term responsesto immediate threats whereas adaptive responses or strategiesrefer to proactive and anticipatory changes over long periodsto reduce the impacts of recurrent threats or gradual changes(Davies 1993 Berkes and Jolly 2002) The dichotomy betweenshort-term and long-term activities is seldom clear becauseshort-term activities affect and depend on long-term onesFor example households may prioritize their strategies tostabilize future incomes over their immediate needs (Eriksenand Silva 2009) In addition Davies (1993) differentiatescoping mechanisms which do not modify prevailing systems(eg production systems social or economic structures) andadaptation which implies changes in rule systems or the moraleconomy Coping mechanisms can also be anticipatory Forexample Cooper et al (2008) consider diversified croppingin which failure of any one is tolerable and failure of all isunlikely Coping mechanisms are developed by individuals orhouseholds and adaptive strategies occur at the communitylevel or above (Berkes and Jolly 2002 Osbahr et al 2008)

In the context of a broader multifaceted study on naturalresource management we researched how people perceiveclimate variations and their impacts Specifically wecompared scientific data and local perceptions of climateseasonality variability and change We characterized whatpast climatic events people recognized and how these hadimpacted peoplersquos lives and how people had adapted andidentified associated constraints We also assessed perceptionsof other nonclimatic events threatening local livelihoods andthe forests on which people depend We discuss how scientificand local perceptions can be combined to inform and improveadaptation programs and policies

STUDY SITEMamberamo is located in the Indonesian province of PapuaThe watershed covers 78000 kmsup2 and is dominated by threemajor rivers the Tariku (or Rouffaer) and Taritatu (orIdenburg) rivers that join the Mamberamo River (Fig 1) The20000 kmsup2 Mamberamo-Foja Wildlife Reserve is located inthe southern and eastern parts of the watershed MamberamoRaya is a new 31000 kmsup2 regency ie a political subdivisionof the province administered by an elected head and localparliament The population density is less than 1 inhabitantper square kilometer Approximately 23000 people live in 59

villages most of which are located along the major riversBecause there are no paved roads settlements are accessedvia waterways trails and small airstrips Ecosystems inMamberamo are mostly relatively undisturbed natural forestsin swamps lowlands hills and mountains and riparian areasThe rich fauna and flora remain little studied (De Fretes 2007Polhemus and Allen 2007 Johns et al 2007 van Heist et al2010)

Fig 1 Map of Mamberamo watershed and the surveyedvillages (Kwerba Burmeso Metaweja Yoke Papasena 1and 2)

Based on our knowledge of the region and a cluster analysisof 40 villages with existing data ie on topographyecosystems population density distance to market andservices and official zoning we identified five groups ofvillage types These groups were called in short swampsaccessible hills in protected forests accessible productionforests remote hills in protected forests and coasts Weselected one village randomly in each group but in onelocation (Papasena) the people insisted on differentiatingbetween two villages (Papasena 1 amp 2) and we finally workedin a total of six villages Papasena 1 amp 2 (hereafter referred toas Papasena) Kwerba Burmeso Metaweja and Yoke (Table1) Although these villages are separated from each other byless than 200 km differences in elevation and wind exposureaccount for differences in local climates for exampletemperature and orographic precipitation Burmeso has thelargest population (145 households) and Metaweja thesmallest (44 households) Common cultural traits are sharedby the different local communities such as similar taboossacred sites and rituals (De Vries 1988 Oosterwal 2007)

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

Table 1 Description of the surveyed villages The column lsquoarearsquo is for customary land

Name Area(kmsup2)

Main ecosystems Status of thesurrounding forest

Main livelihoods Important vegetalresources

Important animalresources

Papasena1 amp 2

1700 Permanent andsemipermanentswamp forests hilland mountain forests

Protected area Fishing wild andplanted sagoexploitation huntinghome garden cultivation(banana red pandanusbetel)

Sago Fijian longanand a lot of wild fruits

Freshwater fishcrocodiles wild pigsbirds of Paradisecassowaries cockatooshornbills maleo

Kwerba 1100 Hill and mountainforests

Protected area Wild and planted sagoexploitation huntinghome garden cultivation(banana sweet potatocassava red pandanusbetel)

Sago DipterocarpsFijian longan ironwood

Crocodiles freshwaterfish wild pigs birds ofParadise cassowariescockatoos hornbillsmaleo

Burmeso (futurecapital of theregency)

1500 Lowland and hillforests

Production forestdegraded because oflogging activities

Fishing planted sagoexploitation huntinghome garden cultivation(cocoa banana sweetpotato red pandanusbetel)

Sago Iron woodFijian longan Agathis

Wild pigs cassowariesphalangers birds ofParadise freshwaterfish

Metaweja 300 Hill mountain andriparian forests onsteep slopes

Protected area Wild and planted sagoexploitation huntinghome garden cultivation(banana sweet potatocassava betel)

Sago Agathis Fijianlongan iron wood

Wild pigs cassowariesbirds of paradisecrowned pigeons

Yoke 1400 Swamp forest lakesand mangrove

Protected area Fishing wild andplanted sagoexploitation coconutplanting cocoa plantinghunting home gardenswith betel cassavabanana

Sago WatercressFijian longan salty-leaf trees

Salt and fresh waterfish molluscscrustaceans crocodileswild pigs cassowaries

Local livelihoods vary according to ecological factors egtopography and the ecosystems in which villages are locatedHowever people in all the villages rely on fishing huntingand the collection of various forest products including sago(Metroxylon sagu) Vegetables are cultivated close tosettlements and camps

METHODSThis research was built on a broader multidisciplinary studyon land-use planning and natural resource managementoriginally developed for studying the importance ofbiodiversity and landscapes for local people (Sheil et al 2002)Our research used focus group discussions householdinterviews and climate data analysis

Most young men in these communities are fluent inIndonesian and most other people have at least a reasonableworking knowledge of the language Thus we were able toperform most of our interviews in Indonesian but in somecases eg talking with some elders men or women we usedlocal translators all villages had people able to play this roleNot all respondents were literate but the facilitator ensuredthat literacy was not necessary An important question waswhether local people make strict use of the conventional

western calendar Based on our discussions with them overmany months of fieldwork we are confident that they do Thisreflects their education the influence of the church and churchcalendar events and other official schedules

Focus group discussionsSemistructured questionnaires were used with four groups ofvillagers defined by gender and age in each village young(less than 30 years old) men young women mature men andmature women We worked with a total of 23 groupsinsufficient young women were present in Papasena 2 at thetime of the survey so this group was omitted During ourdiscussions with villagers we used ldquoseasonsrdquo (musim) ratherthan ldquoclimaterdquo (iklim) We discussed ldquonormal seasonsrdquo (musimbiasa) ie weather people expect during a normal year andldquoabnormal seasonsrdquo (musim tidak biasa) ie seasonalpatterns that differ from a normal year and that can occur everyseveral years or exceptionally We explained extreme climaticevents as ldquounexpected seasonsrdquo (keadaan musim yang luarbiasa) to distinguish them from nonclimatic natural disastersWe discussed how ldquoabnormal seasonsrdquo had changed Weasked how variability and change had impacted them forexample their gardens health or physical assets and howthey had responded eg moving away or asking for

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

government aid To address climate change issues we onlyasked the groups how they perceived climate (using the termmusim) would be in the future The groups were not presentedany scenario of global climate change ie scenariosdeveloped by climate modelers contrary to other studies inwhich climate scenarios are used to analyze futurevulnerability with local stakeholders (Ravera et al 2011)

InterviewsHousehold surveys were conducted in parallel with the focusgroup discussions in 30 of the households and in at least 30households when villages had less than 90 households Forty-four households were surveyed in Burmeso and 30 in each ofthe other villages with a total of 164 In each village a quicksurvey was first conducted in all the households with simplequestions ie respondentrsquos name family members clan andethnic group The name of each household was put in a bagand we randomly withdrew names for a more detailed surveyIf a selected household was absent we picked a new namefrom the bag We then asked the age of each member of thehousehold level of education main occupation and secondaryoccupation We conducted semistructured interviews withheads of households some of them being women about localperceptions of threats to natural resources (eg what factorsinfluenced forest cover or water) and to peoplersquos livelihoods(eg what factors influenced changes in cultivation andsettlement areas) During the interviews we let the villagerstalk about any perceived threat for example privatecompanies diseases unsustainable extractive activities andlandslides Threat was defined as an external phenomenon thatcan potentially negatively impact peoplersquos livelihoods andnatural resources We asked people to rank threats accordingto their perceived impact and frequency We organized theresponses in categories during the data analysis

Data analysisIn each village the four group discussions resulted in differentviews on seasonality variability and change A month wasconsidered as ldquowetrdquo (ldquodryrdquo) in the village if three or moregroups perceived it as ldquowetrdquo (ldquodryrdquo two or more groups in thecase of Papasena 2 where there were only three groups)Qualitative data from focus group discussions were processedwith MaxQDA software to organize and cluster them basedon keywords (MAX Qualitative Data Analysis version 22007) For quantitative data from the demographic survey (ie age education job of each member of the households) andhousehold surveys (ie perceived importance of the forest forlocal livelihoods perceived threats to the forest and localpeople) we used cross tabulation with Chi Square (SPSSversion 17 2008)

Climatic dataWe searched for local climate data but the only local weatherstation belonging to the logging company PT MamberamoAlas Mandiri provided only three years of observations

(2009-2011) too few for analysis of variability or trends Wesearched global climatic databases for information on meanclimate past time-series and future climate scenariosBecause of the coarse spatial resolution of some datasets (eg 05 arc-degree or around 50 km) data were selected for asingle point inside our study area using the coordinates 230degS and 13803degW For datasets with a finer resolution (eg 25arc-minutes or 5 km) we retrieved the information specific tothe six villages For each village we compared meteorologicaldata with local perceptions of seasonality climate variabilityand future climate change looking for agreement ordiscrepancy

For the mean climate we used the WorldClimCL25 datasetwhich provides the average monthly temperature andprecipitation over the 1950 to 2000 period Data from weatherstations are spatially interpolated at a resolution of 25 arc-minutes or 5 km (Hijmans et al 2005) To analyze theseasonality of an average year we differentiated wet months(at least 5 more rainfall than the monthly average) dry (atleast 5 less rainfall than the monthly average) and average(within 5) Before choosing this 5 value for the thresholdwe tested several other values ie higher values resulting inless months identified as dry or wet but with similardistribution of dry and wet seasons within the year We chosethe 5 threshold because it resulted in a similar number ofwet and dry months according to local perceptions and climatedata

For the temperature and precipitation time series from 1960to 2009 we used the CRUTS31 dataset which is based oninterpolation of weather station data and has a resolution of05 arc-degree or 50 km (Mitchell and Jones 2005) Thisdataset was selected because it is the most frequently useddataset in agricultural studies (Ramirez-Villegas andChallinor 2012) We extracted climate data for our study sitebut the validity of these data is questionable because they areinterpolated from sparse climate records in a 10degx10deg square(around 1000 km x 1000 km) centered on Mamberamo theCRUTS31 dataset references only 11 stations

With the CRUTS31 dataset we also compared the seasonalityof an average year and the interannual variability of time seriesbetween our site and sites with similar climates (within 10of the mean annual precipitation and 1 degree of the meanannual temperature) To indicate the future climate we usedscenario data from the TYNSC20 dataset (Mitchell et al2004) The 16 scenarios of this dataset represent the possiblefuture climate in a grid with a resolution of 05 arc-degreeThe scenarios were developed with four scenarios of globalgreenhouse gas emissions coupled with four different modelsof the global climate system to predict how climatic variablescould possibly evolve in space and time

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

RESULTS

Seasonality in Mamberamo according to differentperspectives

From meteorological recordsRecords show that Mamberamo experiences only minorseasonality (Fig 2) with monthly precipitation from 215 to319 mm in an average year Around 97 of the total area ofthe world with similar climates for example in the Amazonthe Philippines Sumatra and Borneo has a higher intra-annual range of precipitation ie higher seasonalitySimilarly monthly temperature varies from 263deg to 271degCand 97 of the global sites with similar climates experiencea higher intra-annual range of temperature

Fig 2 Monthly mean temperatures and precipitation inMamberamo based on WorldClimCL25 dataset (Hijmans etal 2005)

According to local perceptionsIn the six villages people perceived the principal seasonssomewhat differently (Table 2) For example in Papasena onedry season was perceived to be from July to September and inKwerba from August to December In Metaweja and Kwerbatwo dry seasons were perceived In the coastal village of Yokethe dry season was perceived as the longest of the six villages

Villagers from Yoke noted how the wet season is associatedwith winds from the west and large waves and tides whereasthe dry season is associated with winds from the east andsmaller waves and tides Wind tides and waves affectactivities such as fishing

In Papasena and Kwerba during the dry season crocodilesare hunted because they are more visible on the riverbankswhen the water is low This is also the season that land iscleared for new gardens but crops are planted in all the villages

in the wet season utilizing the rain and river for watering Mostother activities such as sago harvesting occur all year roundIn general villagers reported that they make decisionsregarding fishing hunting or preparing and planting gardensbased on their experiences and perceptions They plant theirvegetables during the wet season but they need dry days toprepare the land They decide when to plant their gardens basedon the rain They also pay attention to heavy rains for fear offlash floods

Table 2 Seasonality perceived in the surveyed villages(focus group discussions with women young and matureand men young and mature) and analyzed withWorldClimCL25 (Hijmans et al 2005) Note black = wetmonth grey = dry month white = unclear or average

Comparisons between local perceptions and meteorologicalrecordsDifferences in seasons in the north of the watershed (Yoke)in the center (Burmeso Metaweja) and in the south (KwerbaPapasena) are revealed by the interpolated climate data andthe results from group discussions (Table 2) The comparisonof climate data with local perceptions of seasonality shows agood match in Papasena because both show a dry season fromJune or July to September or November In Kwerba andMetaweja the two dry seasons (February-March and August-December in Kwerba and in March-May and September-October in Metaweja) perceived by local people do not appearin the climate data analysis only one dry season was revealedin these two cases In Burmeso and Yoke the dry seasonappears later in the year in the climate data (August-November) than according to group discussions (June-Augustin Burmeso April-September in Yoke)

Past climate variability and trends in Mamberamoaccording to different perspectives

From meteorological recordsThe interannual variability of recorded time-series is relativelyminor For annual precipitation the coefficient of variation(ratio of standard deviation and mean) is 120 between 1960and 2009 Comparison with other climatically similar placesin the world shows that 72 have a higher coefficient ofvariation Similarly the standard deviation of annualtemperature is 02degC between 1960 and 2009 which is lower

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

Table 3 Extreme events recent occurrences and impacts according to focus group discussions in each village with 4 separategroups of men young and mature and women young and mature (combined summary) Dates in parentheses are years of eventsaccording to local people

Name of village Noticed extreme events Impacts on villagers Impacts on agriculture and natural resourcesPapasena1 amp 2

Strong rains and floods (19982009 2010)

Malaria influenza headaches Crops rot Cassowary and wild pigs cannot find dry land

Long dry season and heatwaves (2009)

Reduced drinking water Sweet potato cassava sago affected Fishes in the lakesdie

Strong winds (not dated) No impact No impactKwerba Heat wave and drought (1992

1997 2007 2010)Diarrhea Crocodiles move from dry tributaries Tree kangaroos

cassowaries wild pigs move near rivers Fishes and prawnsdie

Strong winds (2011) Houses blown away or roof tiles blownoff

No impact

Long period of rain and floods(1996 2005)

No impact Sago betel taro banana wild pigs crocodiles fishaffected

Burmeso Prolonged dry season and heatwave (1994)

No impact Sago annual plants fishes and wild pigs affected Water insmall streams gets warmer and dries up Roads becomedusty

Reversed seasons irregularrains and floods (2009)

Influenza malaria No impact

Metaweja Long rains and floods (19902009 onward)

Diarrhea influenza malaria housesdamaged

Gardens damaged dogs drowned

Strong winds (2008 2011) Houses damaged Coconuts breadfruit betel cocoa blown overLong dry season (2009) No impact Fishes and prawns affected by water temperature

Yoke Long dry season (1994 20032010)

Crop failure wells dry up or increase insalinity no drinking water near the village

Crops affected (betel bananas gnemo and sago)Freshwater fish die (saline water)

Long rainy season (19972001 2011)

Diseases mosquito infestation strongtides disturb fishing activities

Crops fail especially long cycle crops Animals die inlowlands

Strong winds (2000 20062010)

No impact Coconut trees betel palm and pine trees blown over

than in 94 of places with similar climate The range of annualmean temperature between 1960 and 2009 is around 12degCAnnual precipitation has not changed significantly over the1960 to 2009 period The annual temperature has increasedsignificantly by 008degC per decade (95 confidence interval004-011) during this period a smaller increase than theobserved global trend of 0128degC per decade over the sameperiod (Trenberth et al 2007)

According to local perceptionsPeople from the six villages do not keep a careful account ofminor climatic variations but they identified three types ofextreme events related to climate variability in the last 10years droughts floods and strong winds Notable events andtheir impacts differed among villages but in general theyaffect health eg malaria during longer wet seasons ordestroy crops (Table 3)

According to villagers located near the Mamberamo River(Burmeso Kwerba and Papasena) significant floods happenevery five years or so and the last one occurred in 2009 InYoke people were less concerned about interannualvariability than in other villages in part because their coastallocation appears less vulnerable to floods In Metaweja peoplereported whirlwinds occurring every year or two with impactson tree crops eg betel coconut and sago and buildings

roofs being especially vulnerable Papasena villagers saidexceptionally large floods have normally occurred about every15 years These floods have an impact on wildlife because pigsand cassowaries seek refuge on the higher ground far from thevillage

When asked about recent climatic events men and womenemphasized different concerns in all the villages exceptKwerba (Table 4) The perceived events differed in theirfrequency and severity and men more often highlighteddroughts whereas women more often referred to extendedrains In Kwerba the principle difference was betweengenerations For older people prolonged rains were of majorconcern whereas youths put more emphasis on extended dryseasons

People did not report changes in mean climate but they didhighlight a change in the frequency of various extreme eventsVillagers from Papasena explained that damaging floods nowoccur at least every 5 years instead of 15 years apart as in thepast Kwerba villagers also said extended rainy seasons areincreasingly frequent In Metaweja and Yoke people saidstrong winds are now more common Villagers in Burmesoand Yoke also felt extended dry seasons were becoming morefrequent

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

Table 4 Most frequent extreme events according to genderbased on focus group discussions in each village with 4separate groups of men young and mature and women youngand mature

Village Perception of seasonal eventsMen Women

Papasena1 amp 2

Prolonged dry seasons andstrong winds

Prolonged rains

Kwerbadagger No clear gender-specificdifferences

No clear gender-specificdifferences

Burmeso Long droughts Prolonged rains and floodsMetaweja Floods Strong winds (but reported

responses to extreme eventswere about floods)

Yoke Prolonged dry seasons andstrong winds

Prolonged rainy seasons(affecting long-term crops)and prolonged dry seasons(affecting short-term crops)

daggerThere are age-specific differences in the perception elders prolongedrains and floods youth prolonged droughts

Comparisons between local perceptions and meteorologicalrecordsIn Burmeso people reported reversed seasons in 2009 (thewet season was dry and vice-versa) and precipitation time-series confirms this observation Rainfall during this dryseason (June to August) was abnormally high (between 10and 19 monthly and 17 overall) whereas this wet season(October to December) was particularly dry (between 2 to26 less rain monthly and 10 less overall) The recordeddrought at the end of 2009 also matched the perceptions ofpeople in Papasena

In the 1961 to 2009 period monthly rainfalls higher than 420mm occurred in 1969 1970 2005 2007 and 2008 (420 mmis the 99 threshold we used to identify extreme events highermonthly rainfalls occur on average only every 100 months or83 years) The presence of three recent years in this listconfirms the villagersrsquo observations (extended rains in 20052007 and 2008) However it was not possible to compareperceptions and climate data about winds and short-termevents eg drought followed by heavy rain during the samemonth which were not recorded in the monthly temperatureand precipitation dataset

Future changes in climate in Mamberamo according todifferent perspectives

From climate modelsOut of the 16 climate scenarios we used 12 predicted anincrease in annual precipitation by 2080 (up to 1370 mmincrease) and 4 predicted a decrease (up to 350 mm decrease)All scenarios predicted an increase in mean annualtemperature with values ranging from +10 to 34degC by 2080

(Fig 3) Twelve scenarios predicted that mean temperaturewill increase by more than one degree by 2050 compared withthe period 1961-1990

Fig 3 Future change (minimum quartiles and maximum)in annual precipitation (left) and temperature (right)according to the 16 scenarios of the TYNSC20 dataset(Mitchell et al 2004)

According to local perceptionsNone of our respondents indicated any knowledge ofinternational climate change concerns When we asked ourrespondents about their perception of future trends intemperature and rainfall they replied that the current trendswould continue or that they had no idea about future changes

Adaptations to climate variabilityVillagers reported different responses to climate variability(Table 5) but we did not note any marked difference in copingstrategy by age or gender Only in Yoke men discussed thestrategies for finding drinking water during long droughtswhereas women described the crops they plant in suchsituations An anticipatory action to avoid flood damage wasto locate gardens outside flood-prone areas although issuesregarding land rights sometimes encourage people to plantwherever they can including on riverbanks eg BurmesoKwerba Papasena Metaweja In Papasena villagers buildtheir houses on stilts to avoid flood damage In the event of amajor flood the people in Burmeso and Kwerba are alwaysready to move the village temporarily to higher grounds

In contrast people in Yoke are unwilling to move for politicalreasons This community is divided into two settlements butis considered as one village by the administration Thegovernment suggested merging the two settlements and

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

Table 5 Coping and adapting strategies according to focus group discussions in each village with four separate groups of menyoung and mature and women young and mature (combined summary)

Village Category of strategy Strategies for flood Strategies for droughtPapasena1 amp 2

Settlements Build houses on higher stilts No strategy

Productive Move gardens to higher ground Hunt in areas not affected andnear gardens where animals look for food Plant fast growing cropsand harvest before the rainy season (eg sweet potato and banana)

Plant cassava which can grow in dryconditions

Kwerba Settlements Relocate temporarily to higher ground Replace roof tiles with grass or Pandanusleaves for air circulation

Productive Move gardens Go back to old flooded gardens afterward Samecultivars are used

Fish more because fish is easier to catch inclearer water

Cultural No strategy Use rituals for calling rainBurmeso Settlements Relocate temporarily to higher ground Stay in the forest in small settlements

shaded by treesProductive Plant only short-cycle crops (string beans ground nuts other green

vegetables)Use spades to break hard soil but thetechniques for cultivation remain the same

Metaweja Settlements Relocate temporarily to higher ground No strategyProductive Move gardens to higher ground Harvest annual crops before rainy

season Look for new locations of natural fishponds when streamchanges direction

Gnemo trees (Gnetum gnemon) decreasingand replaced by fern leaves as wildvegetable for household consumption

Cultural Taboos to prevent gardens from being planted in Nuari Mountain(sacred area) In case of violation important floods may occur Usetaboos against planting near the river to avoid crop damage byflood (now no longer observed)

No strategy

Yoke Settlements Repair damaged houses with whole community (this applies towinds)

No strategy

Productive No specific action because gardens are not moved Look for drinking nonsalty water far fromthe village

moving people far from the coast to avoid damage fromtsunamis and erosion which was rejected by villagers whowant the two settlements to be recognized as separate entities

Beyond climate other perceived changesIn addition to the group discussions on climate we alsodiscussed the general changes that affect their landscapes andlivelihoods No one suggested that climate or climate relatedevents would play a role in influencing their landscape Allvillagers considered local forests vital to their livelihoods (Fig4) for food (wild sago bush meat gnetum leaves and fruits)water (regulating water flow preventing floods and erosionand providing clean water) construction materialsagriculture shelter and as a reserve of products for futuregenerations In addition forests were also important as sacredareas and for spiritual reasons No significant difference wasfound between villages and respondentsrsquo gender

Most people believed that the area of forest will decrease intheir territories in the future (Fig 5) Villagers explained thattheir travels in the region and observations of other villagesmade them aware of the changes that may affect them Theybelieved that new settlements land clearance for new gardensinfrastructure development and private sector activities willreduce the forest area (Fig 6) They foresaw an increase inmining or logging and industrial plantations Some villagersin Papasena and Yoke thought the forest area will not changebecause they protect the forest themselves

Fig 4 Answers from 164 villagers in the surveyed villageswhen asked why forest is important to them Onerespondent could provide more than one answer

In the six villages the main perceived reasons for forest losswere new settlements infrastructure development privatesector activities and agriculture expansion (Fig 6) The mainreason for forest loss was agriculture expansion in Kwerbawhereas it was infrastructure development in Metaweja (Fig6) In all villages but one (Burmeso) villagers believed their

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

gardens will expand because of a growing human populationVillagers from Burmeso suggested that garden areas willactually decrease because the village will become the regencycapital providing job opportunities and reducing the need foragriculture

Fig 5 Perceptions of the surveyed villages on changes in(1) forest resources (2) agricultural production and (3)population in the next decade from a survey conducted in164 households in 6 villages (Papasena 1 amp 2 KwerbaBurmeso Metaweja Yoke)

DISCUSSION

Differences in perceptions and strategiesOur results show that local perceptions of seasonality andclimate variability differ mainly according to village locationsand the surrounding ecosystems which determine locallivelihoods Some differences in perceptions were observedbetween gender and age groups in relation to their differentactivities their TEK and experience about how their activitiesare affected by climate variations For example menunderstand better how events influence the availability of thewildlife they hunt in the forest and women how cultivatedareas are affected

Even though different groups perceived climate variationsdifferently they shared some similar responses For examplewhen threatened by flash floods local people temporarilymove to higher ground until the threat has gone In generalboth men and women fish and share similar knowledge ofwhere to fish after floods All six villages have detailedknowledge of when and where to hunt when seasonal fruitsare scarce and wild pigs difficult to find The observeddifferences in responses can be explained by the agro-ecological conditions of each village such as in the coastalvillage of Yoke where people have developed responses to theintrusion of salty water into wells

Though elsewhere in Papua local farmers are keen innovatorsin terms of adoption of new crops (Boissiegravere and Purwanto2007) local people in the six villages did not mention suchinnovation in the context of coping with climatic events as hadbeen reported in other countries (Maddison 2007 Fisher et al2010) Rather Mamberamo communities focused on theprotection of their settlement (eg building houses on highstilts against floods) and mobility (eg moving to the forestduring drought or changing fishing or hunting areas) Resultsshowed that responses were not limited to settlements andproductive activities rituals are used for example revivedtaboos restrict gardening near the riverbanks Some peopleconsidered the protection of sacred forests as a way to reducefloods Comparable situations have been observed in PapuaNew Guinea (Jacka 2009)

Villagers almost never mentioned that they used more forestproducts when coping with extreme events as has beenobserved elsewhere (Pramova et al 2012) For example aftera flood in Kalimantan Indonesia the most heavily affectedthe poorest and the least-educated people relied the most onforests for their coping strategies (Liswanti et al 2011) Inrural Peru the gathering of forest products is an importantstrategy for coping with floods (Takasaki et al 2004) Thisrole of forests as safety nets does not appear in our case Itmay be explained by the fact that forests are permanently usedin livelihood activities thus not mentioned as a coping strategydiffering from business-as-usual It may also reflect that therole of forest products was not a specific emphasis of our

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

Fig 6 Answers from 164 villagers in the surveyed villages when asked the reasons for forest changes ornonchanges One respondent could provide more than one answer ldquoOthersrdquo includes erosion trade lackof forest rehabilitation hydroelectric dam and donrsquot know

research questions Additional research would be needed tobetter understand the links between forest products and copingwith extreme events

None of the reported strategies seem to have resulted fromrecent climatic events or perceived trends In addition theresponses described by the villagers can be considered ascoping mechanisms either in anticipation or in reaction tothreats because they are short-term responses and do not implychanges in production or social systems (Davies 1993) Indeedlocal livelihoods are already adapted to the environmentincluding normal climate variations and are thus robust underthese conditions

Comparing traditional ecological knowledge and climatedataMost studies on local perceptions of climate variability andchange focus on how local peoplersquos perceptions fit climatedata (West and Vagravesquez-Leoacuten 2003 Gbetibouo 2009) Someauthors encourage collaboration between scientists andindigenous people but highlight the uncertainties andmethodological challenges of eliciting local knowledge (Sheiland Lawrence 2004 Couzin 2007) Inconsistencies betweenlocal perceptions and biophysical data have generally raisedconcerns about the validity of local knowledge (Hansen et al2004 Sagravenchez-Corteacutes and Chavero 2011)

Although the patterns in some climate data matched the localperceptions of seasonality and variability reported by ourinformants climate data must be considered with caution InMamberamo meteorological data has limited reliability andspatial resolution These uncertainties limit our ability to makevillage centric assessments that take account of both actualand perceived local differences

Limited climatic variation reduces peoplersquos experience ofchange Although it might be anticipated that people in suchconditions might develop a highly nuanced appraisal of thesmall variations that do occur this does not appear to be thecase More specifically the differences in the perceptions ofseasons among groups even within a single village appear toreflect the limited seasonal variation upon which thesejudgments are based This limited variation makes it difficultto identify clear dry and wet seasons from climate data as wellas when discussing with villagers The dichotomy betweendry and wet seasons may have only limited local relevanceand more complex or subtle seasonal patterns may be moresignificant eg availability of certain fishes or fruits Oneclear example of such seasonal perceptions was thesignificance of wind direction and strength reported in YokeThe local reports of recent extreme events in Mamberamowere consistent among villagers and between perceptions and

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

climate data Local people mentioned trends such as morefrequent droughts floods and whirlwinds However forcognitive reasons people may interpret trends from a fewrecent events (Grothmann and Patt 2005) For example tworecent floods may create a feeling of increasing floodfrequency even though floods occurred at similar but irregularintervals in the past Because neither climate data nor localperceptions can be considered as perfect sources ofinformation both can complement each other and providevaluable insights However when working in regions with fewweather stations local knowledge is a more detailed and inmany ways more relevant source of information as long as itis considered with caution

Implications for climate change adaptationUnderstanding how local people experience climate variationsand how they anticipate or react to them is likely to becomecrucial in the context of climate change Given the lowinterannual climate variability in Mamberamo sustainedclimate change will eventually create novel conditions forlocal people Local peoplersquos farming behaviors are longestablished and thus well suited to the low interannual climatevariability that prevails in this region Some practices such asmixed cropping and dispersal of sago and other crops indifferent locations may provide some resilience To spurlarger scale adaptations may require climatic events to crosssome threshold eg frequent or marked floods or droughtsthat impact local crops and livelihoods For example with onedegree of temperature increase (the third quartile oftemperature increases in 2050 for the considered scenariossee Fig 3) 98 of the annual mean temperatures would behigher than the maximum annual mean temperatureexperienced in the past 50 years assuming that the variationsof annual temperatures around the mean is similar in the pastand the future The uncertainty of future precipitation hasimportant local implications For example drier and wetterclimates would have different consequences in differentlocations In the hills and mountains such as in Metawejapeople would be more affected by drier climates because thesmall rivers would dry up as reported in the drought of 1997In the swamps such as in Papasena larger floods might evensubmerge land areas that currently are always above waterlevel

None of the informants predicted major changes inMamberamorsquos climate in the next 10 or 20 years Their focuswas on interannual variability and current trends from whatthey had experienced (more frequent droughts floods andwhirlwinds) We can anticipate how people may adapt toclimate change by looking at their coping strategies insituations of climate variability and extreme climatic eventsLocal responses imply different mechanisms such as thetemporary migration of villages or displacement of cultivationareas the use of crops that can better resist climate changeand the solicitation of networks for help in a crisis However

with climate change current strategies may not be effectiveenough and new or improved responses may be neededExternal interventions will probably address climate changeadaptation in the future eg external aid could increase withclimate change but not all interventions are good foradaptation (Eriksen et al 2011)

External interventions for facilitating local adaptation toclimate change are more likely to be successful if they buildon existing knowledge strategies and traditions For examplein Metaweja traditional rules are used by the community toprevent villagers from cultivating gardens on the flood-proneriverbanks These rules could be used in community-basedadaptation plans after people agree collectively on where torestrict cultivation

Local knowledge can also help identify potentialmaladaptation ie ldquoaction taken ostensibly to avoid or reducevulnerability to climate change that impacts adversely on orincreases the vulnerability of other systems sectors or socialgroupsrdquo (Barnett and OrsquoNeill 2010211) In the coastal villageof Yoke droughts increase salinity of the freshwater supplyGovernment interventions have caused the destruction ofcoastal vegetation such as coconut trees for building concretehouses along the shoreline to encourage villagers to move fromtheir over-the-water houses According to villagers this hasincreased shoreline erosion and aggravated the problem ofsalinity in inland wells Even though building concrete housescan be seen as an adaptation option the associated removal ofvegetation can result in maladaptation

One commonly suggested adaptation option is to resettlevillages located in disaster-prone areas However previousexperience with resettlement has not been particularly positive(for example transmigration in Indonesia Fearnside 1997) andforced relocation generates strong opposition and conflicts InMamberamo temporal migration is a common coping strategybut villagers are reluctant to permanently shift the location oftheir villages In Yoke people feel they need to maintain apermanent presence to keep their settlement eligible forldquoofficial villagerdquo status Adaptation programs will need to lookfor solutions that accommodate these concerns

Adaptation interventions should also strive to find localsolutions to current problems For example in Papasenafloods affect hunting and increase crop damage because wildanimals move to the higher grounds where gardens are locatedFood security is threatened by floods and climate change canaggravate this problem Adaptation plans could increase foodsecurity now and under a future climate scenarios withtechniques to reduce crop damage eg traps for the animalsand alternative sources of food in case of shortage

Adaptation interventions should also consider multiplechanges affecting local people and their landscapes ratherthan climatic variations only Villagers said their livelihoods

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

were more affected by political (decentralization) economic(activities from the private sector) and demographic changesthan climatic variations For example decentralization has ledto new institutions eg multiplication of regencies villagesand government representatives and business developmentsthat create new pressures on natural resources These changesmay be as unpredictable as changes in climate but the wayslocal people cope with them provide valuable insight whenlooking at the implications for adaptation to climate changeOne important observation is related to information flows inthe region as people regularly move along the MamberamoRiver they have a good knowledge of changes and impactsover the watershed This information sharing could bevaluable to future adaptation

CONCLUSIONEliciting traditional ecological knowledge and localperceptions can help to analyze extreme events such as floodsand droughts and their consequences By comparing local andtechnical knowledge on climate we have been able to identifygaps and areas of agreement Although meteorological datafrom even a few weather stations can provide information onclimate variability in isolated areas such as Mamberamo itcannot detect all relevant changes at the local village levelLocal people on the other hand can provide more detailedinformation based on their experience

People in Mamberamo have experienced low climaticvariability in general and this may partly explain why localpeople from the six villages of the study consider nonclimaticfactors as having the biggest influence on their livelihoods Ifadaptation programs seek local agreement and participationtheir implementers will need to understand these wider issuesVillagers react to changes that they observe To date this doesnot include ldquoclimate changerdquo per se but in the future the studyof local variations in coping mechanisms to extreme climaticevents and variability will help to determine what adaptationstrategies need to be developed to address climate change

Our examples illustrate a diversity of concerns andimplications arising from climatic events These examplesincluded increased salinity of water supplies crop loss due tofloods and reduced hunting success each occurring in adifferent village Although climate change is a globalphenomenon adaptation strategies must be specific to givenlocations and needs Addressing local specificities will be achallenge but any inflexible one-size-fits-all approach islikely to fail Local engagement and diagnosis of problemsencountered by local people will be fundamental to ensuresuccess Because TEK focuses on elements of significance forlocal people it will play a primordial role in future plans foradaptation to climate change Collaboration between TEKholders and scientists can generate new knowledge of highrelevance to these plans

Responses to this article can be read online at httpwwwecologyandsocietyorgissuesresponsesphp5822

Acknowledgments

This paper was presented during the 13th Congress of theInternational Society of Ethnobiology 21-25 May 2012 Theauthors thank the local community in Mamberamo RayaRegency for their participation to their activities They alsothank Glen Mulcahy the two anonymous reviewers and theorganizers of the session ldquoTraditional Ecological Knowledgeand Resilience in the context of Global EnvironmentalChangerdquo for their valuable comments and editing Theyacknowledge the Agence Franccedilaise de Deacuteveloppement (AFD)the US Agency for International Development (USAID) andthe Australian Agency for International Development(AusAID) for their financial support This research wascarried out by the Center for International Forestry Research(CIFOR) the Centre de coopeacuteration Internationale enRecherche Agronomique pour le Deacuteveloppement (CIRAD) aspart of the CGIAR Research Program on Forests Trees andAgroforestry and Conservation International (CI)

LITERATURE CITEDAdger W N S Dessai M Goulden M Hulme I LorenzoniD R Nelson L O Naess J Wolf and A Wreford 2009Are there social limits to adaptation to climate changeClimatic Change 93335-354

Bandyopadhyay S L Wang and M Wijnen 2011Improving household survey instruments for understandingagricultural household adaptation to climate change waterstress and variability Living Standards Measurement Study- Integrated Surveys on Agriculture Washington DC USA

Barnett J and S OrsquoNeill 2010 Maladaptation GlobalEnvironmental Change 20211-213 httpdxdoiorg101016jgloenvcha200911004

Berkes F J Colding and C Folke 2000 Rediscovery oftraditional ecological knowledge as adaptive managementEcological Applications 101251-1262 httpdxdoiorg1018901051-0761(2000)010[1251ROTEKA]20CO2

Berkes F C Folke and M Gadgil 1995 Traditionalecological knowledge biodiversity resilience and sustainabilityPages 281-299 in C A Perrings K-G Maumller C Folke C SHolling and B O Jansson editors Biodiversity conservationKluwer Academic Dordrecht The Netherlands

Berkes F and D Jolly 2002 Adapting to climate changesocial-ecological resilience in a Canadian western Arcticcommunity Conservation Ecology 5(2) 18

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

Boissiegravere M and Y Purwanto 2007 The agriculturalsystems of Papua Pages 1125-1148 in A J Marshall and BM Beehler editors The ecology of Papua PeriplusSingapore

Brockhaus M H Djoudi and B Locatelli 2013 Envisioningthe future and learning from the past adapting to a changingenvironment in northern Mali Environmental Science ampPolicy 2594-106 httpdxdoiorg101016jenvsci201208008

Bryan E T T Deressa G A Gbetibouo and C Ringler2009 Adaptation to climate change in Ethiopia and SouthAfrica options and constraints Environmental Science ampPolicy 12413-426 httpdxdoiorg101016jenvsci200811002

Byg A and J Salick 2009 Local perspectives on a globalphenomenonmdashclimate change in Eastern Tibetan villagesGlobal Environmental Change 19156-166 httpdxdoiorg101016jgloenvcha200901010

Cooper P J M J Dimes K P C Rao B Shapiro BShiferaw and S Twomlow 2008 Coping better with currentclimatic variability in the rain-fed farming systems of sub-Saharan Africa an essential first step in adapting to futureclimate change Agriculture Ecosystems amp Environment 126(1-2)24-35 httpdxdoiorg101016jagee200801007

Couzin J 2007 Opening doors to native knowledgeScientific and local cultures seek common ground for tacklingclimate-change questions in the Arctic Science 3151518-1519httpdxdoiorg101126science31558181518

Davies S 1993 Are coping strategies a cop out IDS Bulletin 24(4)60-72 httpdxdoiorg101111j1759-54361993mp24004007x

De Fretes Y 2007 Opportunities and challenges for doingconservation in Papua Pages 1311-1326 in A J Marshall andB M Beehler editors The ecology of Papua PeriplusSingaporeDeressa T T R M Hassan C Ringler T Alemu and MYesuf 2009 Determinants of farmersrsquo choice of adaptationmethods to climate change in the Nile Basin of EthiopiaGlobal Environmental Change 19248-255 httpdxdoiorg101016jgloenvcha200901002

De Vries J 1988 Kwerba view of the supernatural worldIRIAN 161-16

Eriksen S P Aldunce C S Bahinipati R D A Martins JI Molefe C Nhemachena K Obrien F Olorunfemi J ParkL Sygna and and K Ulsrud 2011 When not every responseto climate change is a good one identifying principles forsustainable adaptation Climate and Development 3(1)7-20httpdxdoiorg103763cdev20100060

Eriksen S H K Brown and P M Kelly 2005 The dynamicsof vulnerability locating coping strategies in Kenya and

Tanzania Geographical Journal 171(4)287-305 httpdxdoiorg101111j1475-4959200500174x

Eriksen S and J A Silva 2009 The vulnerability contextof a savanna area in Mozambique household drought copingstrategies and responses to economic change EnvironmentalScience amp Policy 12(1)33-52 httpdxdoiorg101016jenvsci200810007

Fabricius C C Folke G Cundill and L Schultz 2007Powerless spectators coping actors and adaptive co-managers a synthesis of the role of communities in ecosystemmanagement Ecology and Society 12(1) 9 [online] URLhttpwwwecologyandsocietyorgvol12iss11art29

Fearnside P M 1997 Transmigration in Indonesia lessonsfrom its environmental and social impacts EnvironmentalManagement 21(4)553-570 httpdxdoiorg101007s002679900049

Fisher M M Chaudhury and B McCusker 2010 Do forestshelp rural households adapt to climate variability Evidencefrom Southern Malawi World Development 381241-1250httpdxdoiorg101016jworlddev201003005

Gbetibouo G A 2009 Understanding farmers perceptionsand adaptations to climate change and variability The caseof the Limpopo Basin South Africa Discussion paperInternational Food Policy Research Institute Washington DC USA

Grothmann T and A Patt 2005 Adaptive capacity andhuman cognition the process of individual adaptation toclimate change Global Environmental Change 15199-213httpdxdoiorg101016jgloenvcha200501002

Hansen J S Marx and E Weber editors 2004 The role ofclimate perceptions expectations and forecasts in farmerdecision making the Argentine Pampas and South FloridaInternational Research Institute for Climate and SocietyPalisades New York USA

Hijmans R J S E Cameron J L Parra P G Jones and AJarvis 2005 Very high resolution interpolated climatesurfaces for global land areas International Journal ofClimatology 251965-1978 httpdxdoiorg101002joc1276

Jacka J 2009 Global averages local extremes the subtletiesand complexities of climate change in Papua New GuineaPages 197-208 in S A Crate and M Nuttall editorsAnthropology and climate change From encounters toactions Left Coast Press Walnut Creek California USA

Johns R J G A Shea and P Puradyatmika 2007 Lowlandswamp and peat vegetation of Papua Pages 910-944 in A JMarshall and B M Beehler editors The ecology of PapuaPeriplus Singapore

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

Liswanti N D Sheil I Basuki M Padmanaba and GMulcahy 2011 Falling back on forests how forest-dwellingpeople cope with catastrophe in a changing landscapeInternational Forestry Review 13442-455 httpdxdoiorg101505146554811798811326

Maddison D 2007 The perception of and adaptation toclimate change in Africa Policy Research Working Paper TheWorld Bank Washington DC USA

McCarthy N 2011 Understanding agricultural householdsadaptation to climate change and implications for mitigationland management and investment options Living StandardsMeasurement Study - Integrated Surveys on AgricultureLEAD Analytics Inc Washington DC USA

Mitchell T D T R Carter P D Jones M Hulme and MNew 2004 A comprehensive set of high-resolution grids ofmonthly climate for Europe and the globe the observed record(1901-2000) and 16 scenarios (2001-2100) Tyndall Centrefor Climate Change Research Norwich UK

Mitchell T D and P D Jones 2005 An improved methodof constructing a database of monthly climate observationsand associated high-resolution grids International Journal ofClimatology 25693-712 httpdxdoiorg101002joc1181

Morton J F 2007 The impact of climate change onsmallholder and subsistence agriculture Proceedings of theNational Academy of Sciences 10419680-19685 httpdxdoiorg101073pnas0701855104

Mubaya C P J Njukib E P Mutsvangwa F T Mugabeand D Nanja 2012 Climate variability and change or multiplestressors Farmer perceptions regarding threats to livelihoodsin Zimbabwe and Zambia Journal of EnvironmentalManagement 1029-17 httpdxdoiorg101016jjenvman201202005

OrsquoBrien K T Quinlan and G Ziervogel 2009 Vulnerabilityinterventions in the context of multiple stressors lessons fromthe Southern Africa Vulnerability Initiative (SAVI)Environmental Science amp Policy 1223-32 httpdxdoiorg101016jenvsci200810008

Oosterwal G 2007 Murumarew a dual organized village onthe Mamberamo West Irian Pages 157-188 in Koentjaraningrat editor Villages in Indonesia First EquinoxEdition Equinox Singapore

Orlove B S J C H Chiang and M A Cane 2000Forecasting Andean rainfall and crop yield from the influenceof El Nino on Pleiades visibility Nature 40368-71 httpdxdoiorg10103847456

Osbahr H C Twyman W N Adger and D S G Thomas2008 Effective livelihood adaptation to climate change

disturbance scale dimensions of practice in MozambiqueGeoforum 391951-1964 httpdxdoiorg101016jgeoforum200807010

Polhemus D A and G R Allen 2007 Freshwaterbiogeography Pages 207-245 in A J Marshall and B MBeehler editors The ecology of Papua Periplus Singapore

Pramova E B Locatelli H Djoudi O A Somorin 2012Forests and trees for social adaptation to climate variabilityand change Wiley Interdisciplinary Reviews ClimateChange 3581-596 httpdxdoiorg101002wcc195

Ramirez-Villegas J and A Challinor 2012 Assessingrelevant climate data for agricultural applicationsAgricultural and Forest Meteorology 16126-45 httpdxdoiorg101016jagrformet201203015

Ravera F D Tarrasoacuten and E Simelton 2011 Envisioningadaptive strategies to change participatory scenarios foragropastoral semiarid systems in Nicaragua Ecology andSociety 16(1) 20 [online] URL httpwwwecologyandsocietyorgvol16iss1art20

Reid P and C Vogel 2006 Living and responding tomultiple stressors in South Africa ndash glimpses from KwaZulu-Natal Global Environmental Change 16195-206 httpdxdoiorg101016jgloenvcha200601003

Roncoli C K Ingram P Kirshen 2001 The costs and risksof coping with drought livelihood impacts and farmersresponses in Burkina Faso Climate Research 19(2)119-132httpdxdoiorg103354cr019119

Saacutenchez-Corteacutes M S and E L Chavero 2011 Indigenousperception of changes in climate variability and its relationshipwith agriculture in a Zoque community of Chiapas MexicoClimatic Change 107363-389 httpdxdoiorg101007s10584-010-9972-9

Saroar M M and J K Routray 2012 Impacts of climaticdisasters in coastal Bangladesh why does private adaptivecapacity differ Regional Environmental Change 12169-190httpdxdoiorg101007s10113-011-0247-4

Shackleton S E and C M Shackleton 2012 Linkingpoverty HIVAIDS and climate change to human andecosystem vulnerability in southern Africa consequences forlivelihoods and sustainable ecosystem managementInternational Journal of Sustainable Development amp WorldEcology 19275-286 httpdxdoiorg1010801350450920-11641039

Sheil D and A Lawrence 2004 Tropical biologists localpeople and conservation new opportunities for collaborationTrends in Ecology and Evolution 19634-638 httpdxdoiorg101016jtree200409019

Ecology and Society 18(4) 13httpwwwecologyandsocietyorgvol18iss4art13

Sheil D R K Puri I Basuki M van Heist M Wan NLiswanti Rukmiyati M A Sardjono I Samsoedin KSidiyasa Chrisandini E Permana E M Angi F GatzweilerB Johnson and A Wijaya 2002 Exploring biologicaldiversity environment and local peoplersquos perspectives in forestlandscapes methods for a multidisciplinary landscapeassessment CIFOR Bogor Indonesia

Smit B and J Wandel 2006 Adaptation adaptive capacityand vulnerability Global Environmental Change 16(3)282-292 httpdxdoiorg101016jgloenvcha200603008

Takasaki Y B L Barham and O T Coomes 2004 Riskcoping strategies in tropical forests floods illnesses andresource extraction Environment and DevelopmentEconomics 9203-224 httpdxdoiorg101017S1355770X03001232

Thomas D S G C Twyman H Osbahr and B Hewitson2007 Adaptation to climate change and variability farmerresponses to intra-seasonal precipitation trends in SouthAfrica Climatic Change 83301-322 httpdxdoiorg101007s10584-006-9205-4

Trenberth K E P D Jones P Ambenje R Bojariu DEasterling A Klein Tank D Parker F Rahimzadeh J ARenwick M Rusticucci B Soden and P Zhai 2007Observations surface and atmospheric climate change Page235-336 in S Solomon D Qin M Manning Z Chen MMarquis K B Averyt M Tignor and H L Miller editorsThe physical science basis Contribution of Working Group Ito the Fourth Assessment Report of the IntergovernmentalPanel on Climate Change Cambridge University PressCambridge UK

van Heist M D Sheil I Rachman P Gusbager C ORaweyai and H S M Yoteni 2010 The forests and relatedvegetation of Kwerba on the Foja Foothills MamberamoPapua (Indonesian New Guinea) Blumea 55153-161 httpdxdoiorg103767000651910X526889

Vedwan N and R E Rhoades 2001 Climate change in theWestern Himalayas of India a study of local perception andresponse Climate Research 19109-117 httpdxdoiorg103354cr019109

Vignola R T Koellner R W Scholz and T L McDaniels2010 Decision-making by farmers regarding ecosystemservices factors affecting soil conservation efforts in CostaRica Land Use Policy 271132-1142 httpdxdoiorg101016jlandusepol201003003

West C T and M Vaacutesquez-Leoacuten 2003 Testing farmersrsquoperceptions of climate variability a case study from theSulphur Springs Valley Arizona Pages 233-250 in S Straussand B Orlove editors Weather climate culture BergOxford UK