Impacts of Rubber Expansion in SE Asia

Rubber (Hevea brasiliensis) is the major commercial crop replacing traditional agriculture and secondary forests in upland areas of mainland Southeast Asia (Ziegler et al., 2009, Science 324).  Our preliminary research suggested significant plausible impacts of rubber cultivation on watershed hydrological processes (Guardiola-Claramonte et al. 2008).  Hydrologic change within these upland watersheds could have serious consequences for the approximately 200 million inhabitants of the region’s lowlands.  The overarching science questions being addressed by our project are three-fold: (1) how does the conversion from existing land covers to rubber affect local energy, water, and carbon fluxes; (2) how extensive will rubber become; and (3) what are the consequences of those changes for regional hydrology and carbon sequestration? 



The Rubber Juggernaut

Ziegler, AD, JM Fox, J Xu.  2009.  The rubber juggernaut.  Science 324: 1024-1025.  

Ziegler et al 2009_Science324(1024-1025)
Adobe Acrobat Document 322.2 KB




Rubber is native to the humid tropics and has traditionally been cropped in the equatorial zone between 10°N and 10° S. In mainland Southeast Asia this included portions of southern Thailand, southeastern Vietnam, and southern Myanmar. In the early 1950s, China invested in research on growing rubber in marginal environments and eventually established state rubber plantations in Hainan and Yunnan provinces in areas that lie as far north as 22° north latitude. China’s success in growing rubber in these ‘non-traditional’ environments greatly expanded the habitat in which rubber was perceived to be productive.

            Fox et al (2014) explain that investors from China, Vietnam, Malaysia, and Thailand are investing heavily in rubber plantations in non-traditional rubber growing areas of Laos, Cambodia, and Myanmar, as well as portions of their own countries—e.g., northwest Vietnam and northeast Thailand. In Laos more than 140,000 ha of rubber have been planted in the last decade; and the plantation area may reach 300,000 ha during the next decade. In Cambodia, the Ministry of Agriculture plans to expand the area under rubber cultivation from 100,000 ha to as much as 800,000 ha by 2015. In Myanmar, rubber is expanding into border areas in Kachin and Shan States. In Thailand, rubber has expanded to include over 64,000 ha in the north and 348,000 ha in the northeast. The rubber growing area in Vietnam increased from 395,000 ha in 1999 to 550,000 ha in 2007, with 4,500 ha planted in the northwest region. The government has a target of 700,000 ha of rubber by 2020. Collectively more than 1,000,000 ha of rubber have been planted in the last several decades in non-traditional rubber growing areas of China, Laos, Thailand, Vietnam, Cambodia, and Myanmar. By 2050, the area of land dedicated to rubber (and/or other monoculture plantation crops) in these areas could quadruple, largely by replacing lands now occupied by evergreen broadleaf trees and swidden-related secondary vegetation (Fox et al. 2012).

        In prior works we mapped rubber at two sites (Nam Ken, China; Ratanakiri Cambodia)  using a multi-date, multi-sensor approach.  In the present project we are expanding our efforts by investigating emerging, rubber-growing hotspots in the region to characterize the current extent of rubber and changes in the recent past. Specifically, we are characterizing and mapping rubber in emerging rubber-growing regions from 2000-present and producing a weighted suitability model of rubber in non-traditional rubber growing areas. Outputs will be used in the regional land-cover/land-use change (LCLUC) simulations.

       Two sets of simulations are being implemented to project LCLUC (including rubber expansion) over the next 50 years and to derive annual simulations for a 65-year period (1985 - 2050).  The first is a set of “historical” simulations (1985–2000) that will fill in LCLU knowledge gaps between existing regional and global LCLU datasets.  Existing global LCLU datasets will be used to help calibrate CLUE-s, a dynamic, spatially-explicit LCLUC simulation model, prior to running model the second set of simulations (Fox et al., 2012).  The second set are future LCLUC simulations (2001–2050) that incorporate rubber as a distinct LCLU type and use a scenario reflecting the current observed trend towards a plantation economy with substantial expansion and intensification of rubber and other commercial crops (Sen et al., 2012).

         To simulate the transient carbon and water dynamics and estimate the contribution of rubber expansion to changes in water and CO2 budgets, a mechanistic dynamic vegetation model is needed that can simulate short- and long-term ecological processes and their responses and feedbacks to LCLUC and other natural and anthropogenic disturbances.  We are utilizing the Ecosystem Demography (ED) model, a mechanistic terrestrial biosphere model that can simulate ecological processes on time scales from hours to centuries and on spatial scales from meters to continents, including leaf-level physiological processes (such as photosynthesis, respiration, and ET), individual-based vegetation dynamics (growth, allocation, mortality, phenology, and reproduction), landscape-scale disturbance events (fire, wind-throw), and belowground decomposition and soil hydrology.  To examine the impact of rubber expansion, other LCLUC, and climate change on water and carbon dynamics, we are using ED to conduct the following four experiments: (1) current rubber and land use (2000) and current climate (2000); (2) projected rubber and land use (2001-2050) and current climate (2001-2050); (3) current rubber and land use (2001-2050) and projected climate (2001-2050); and (4) projected rubber and land use (2001-2050) and projected climate (2001-2050).

        For most of various plant functional types (PFTs) included in the model, some field measurement of their ecosystem processes have been done.  The water and carbon exchange characteristics of rubber are less well understood. We are, therefore, focusing our field observations on water and carbon fluxes of rubber.  We are currently conducting a program of field measurements at representative rubber plantation sites within MMSEA to: (1) obtain direct measurements of stand-level fluxes of energy, water, and carbon for rubber; (2) measure rubber tree size and ecophysiological characteristics needed to parameterize the ED model; and (3) provide time-dependent data sets of water and carbon fluxes and related micrometeorological and hydrological variables for use in validating ED model predictions for rubber.  We have established two eddy covariance tower sites, one in Nong Khai Provice, NE Thailand, and one in the Cambodia Rubber Research Institute plantation at Kampong Cham.  Each of these towers is currently measuring stand-level fluxes of water, carbon, and energy, and will greatly improve understanding of ecosystem processes of rubber plantations.



Recent Findings


Evapotranspiration of rubber (Hevea brasiliensis) cultivated at two plantation sites in Southeast Asia


Our new paper lead by Tom Giambelluca (U Hawaii, USA) investigates the effects of expanding rubber (Hevea brasiliensis) cultivation on water cycling in Mainland Southeast Asia (MSEA),. Evapotranspiration (ET) was measured within rubber plantations at Bueng Kan, Thailand, and Kampong Cham, Cambodia (Giambelluca et al., 2016, see below). After energy closure adjustment, mean annual rubber ET was 1211 and 1459 mm mm/y at the Thailand and Cambodia sites, respectively, higher than that of other tree-dominated land covers in the region, including tropical seasonal forest (812–1140 mm mm/y), and savanna (538–1060 mm/y). The mean proportion of net radiation used for ET by rubber (0.725) is similar to that of tropical rainforest (0.729) and much higher than that of tropical seasonal forest (0.595) and savanna (0.548). Plant area index (varies with leaf area changes), explains 88.2% and 73.1% of the variance in the ratio of latent energy flux (energy equivalent of ET) to potential latent energy flux (LE/LEpot) for midday rain-free periods at the Thailand and Cambodia sites, respectively. High annual rubber ET results from high late dry season water use, associated with rapid refoliation by this brevideciduous species, facilitated by tapping of deep soil water, and by very high wet season ET, a characteristic of deciduous trees. Spatially, mean annual rubber ET increases strongly with increasing net radiation (Rn) across the three available rubber plantation observation sites, unlike non-rubber tropical ecosystems, which reduce canopy conductance at high Rn sites. High water use by rubber raises concerns about potential effects of continued expansion of tree plantations on water and food security in MSEA.

Giambellcua et al., 2016, Water Resources Research
Giambelluca_etal_2016_WRR_rubber ET.pdf
Adobe Acrobat Document 4.1 MB

Mean annual energy-closure-adjusted evapotranspiration (ET) at tower sites from this study and a site in Xishuangbanna reported by Tan et al. [2011] shown against a map of estimated regional ET from the LandFlux-EVAL merged benchmark synthesis products of ETH Zurich produced under the aegis of the GEWEX and ILEAPS projects (Figure 9, Giambelluca et al., 2016).

Do current trends of rubber plantation expansion threaten biodiversity and livelihoods?


The first decade of the new millennium saw a boom in rubber prices that led to rapid and widespread land conversion to monoculture rubber plantations in continental SE Asia, where natural rubber production has increased >50% since 2000. Here, we analyze the subsequent spread of rubber between 2005 and 2010 in combination with environmental data and reports on rubber plantation performance. We show that rubber has been planted into increasingly sub-optimal environments. Currently, 72% of plantation area is in environmentally marginal zones where reduced yields are likely. An estimated 57% of the area is susceptible to insufficient water availability, erosion, frost, or wind damage, all of which may make long-term rubber latex production plausibly unsustainable. Future climate change is likely to lead to a net exacerbation of “environmental marginality” for both current and predicted future rubber plantation area (based on analysis of data from 39 models from the Coupled Model Intercomparison Project Phase 5 across four Representative Concentration Pathways for 2050). New rubber plantations are also frequently placed on lands that are important for biodiversity conservation and other ecological functions.  For example, between 2005 and 2010, more than 2,500 km2 of natural tree cover and 610 km2 of protected areas were converted to plantations. Overall, expansion into marginal areas creates the potential for the lose-lose situation of clearing high-biodiversity value land for economically unstainable plantations that are poorly adapted to local conditions and alter landscape functions (e.g. hydrology, erosion), ending in compromised livelihoods, particularly when rubber prices fall.


Caption: Environmental stress map. (a) Areas where environmental stresses are so severe that there is a risk of unsustainability. To generate a composite map of primary risks we first delineated the typhoon and topographic risk zones, and then assigned remaining risk areas to the drought zone or frost zone depending on which was furthest from its optimum (median value within the natural rubber range). (b) Sub-optimal areas with dry stress. (c) Sub-optimal areas with cold stress (Figure 3, Ahrends et al., 2015)


Featured paper:

Ahrends, A., PM Hollingsworth, AD Ziegler, JM Fox, H Chen, Y Su, J Xu.  2015. Current trends of rubber plantation expansion may threaten biodiversity and livelihoods.  Global Environmental Change 34, 48-58

ahrends+2015-GEC-34-currents trends in r
Adobe Acrobat Document 3.9 MB

More of our Publications


 Kumagai, T., RG Mudd, TW Giambelluca, N Kobayashi, Y Miyazawa, TK Lim, W Lin, M Huang, JM Fox, AD Ziegler, S Yin, SV Mak, P Kasemsap. 2015. How do rubber (Hevea brasiliensis) plantations cope with seasonal drought in northern Thailand and central Cambodia?  Agricultural and Forest Meteorology 213: 10-22.

kumagai et al 2015 - AFM213- how do rubb
Adobe Acrobat Document 1.8 MB

Fox, J, Castella, J-C, AD Ziegler, SB Westley.  2014. Expansion of rubber mono-cropping and its implications for the resilience of ecosystems in the face of climate change in Montane Mainland Southeast Asia.  Global Environmental Change 29: 318-326.

fox et al 2014 GEC rubber and redd.pdf
Adobe Acrobat Document 457.6 KB

Kobayashi, N., T Kumagai, Y Miyazawa, K Matsumoto, M Tateishi, TK Lim, RG Mudd, AD Ziegler, TW Giambelluca, S Yin. 2014.  Transpiration characteristics of a rubber plantation in central Cambodia. Tree physiology 34 (3): 285-301.

Kobayashi et al Tree Physiol-2014-285-30
Adobe Acrobat Document 2.0 MB

Fox, JM., JC Castella, AD Ziegler, SB Westley. 2014.  Rubber plantations expand in mountainous Southeast Asia: What are the consequences for the environment. Asia Pacific Issues 114: 1-8.

fox et al 2014 API - Rubber Plantaitons
Adobe Acrobat Document 481.7 KB

Sen, OL, JM Fox, JB Vogler, TW Giambelluca, AD Ziegler 2012. Hydro-climatic effects of future land-cover/land-use change in montane mainland southeast Asia. Climatic Change. DOI 10.1007/s10584-012-0632-0.

sen et al 2013 climate change 118-hydrod
Adobe Acrobat Document 2.4 MB

Kumagai, T., RG Mudd, Y Miyazawa, W Liu, TW Giambelluca, K Kobayashi, TK Lim, M Jomura, K Matsumoto, M Huang, C Qi, AD Ziegler. 2013. Simulation of canopy CO2/H2O fluxes for a rubber (Hevea brasiliensis) plantation in central Cambodia: the effect of the regular spacing of planted trees. Ecological Modelling 265: 124-135.

kumagai et al 2013 - em1301 - simulation
Adobe Acrobat Document 2.0 MB

Fox, FM, JB Vogler, OL Sen, TW Giambelluca, AD Ziegler. 2012.  Simulating land-cover change in Montane Mainland Southeast Asia. Environmental Management 49(5): 968-979. 

fox et al 2012 EM49-simulating landcover
Adobe Acrobat Document 981.9 KB

Guardiola-Claramonte, M, PA Troch, AD Ziegler, TW Giambelluca, M Durcik, JB Vogler, MA Nullet. 2010. Modeling basin-scale hydrologic effects of rubber (Hevea brasiliensis) in a tropical catchment. Ecohydrology. 3(3): 306-314. 

Guardiola-Claramonte et al 2010_Ecohydro
Adobe Acrobat Document 695.1 KB

Guardiola-Claramonte, M, PA Troch, AD Ziegler, TW Giambelluca, JB Vogler, MA Nullet. 2008. Local hydrologic effects of introducing non-native vegetation in a tropical catchment.  Ecohydrology 1: 13-2

Guardiola-Claramonte et al 2008_Ecohydr
Adobe Acrobat Document 1.9 MB