This blog was taken from a recent university study I conducted, using a case study on Cambodian wetland management to investigate using a Geographical Information System for management purposes. It therefore takes a more formal format than regular blogs but I didn’t alter it as I wanted to keep the research based tone.
Image Source: TYTO Wetlands
Wetlands are the classified as areas of land which are covered by water at least part of the year (DEE 2018). They can be considered transitional areas between the aquatic and marine, where the water table is frequently interacting with the surface. The floodplain is the alluvial plain of a river, the extent of which is often considered the maximum reach of flooding, they include wetlands as areas which are mostly wet throughout the year. Wetlands are ecologically significant and play a crucial role in providing ecosystem services or benefits to human populations on local and global scales. One key role is the regulation of water movement (Richardson 1994, Mitsch & Gosselink 1993). As such, wetlands provide flood control, increase groundwater recharge, reduce sediment load, contribute to carbon sequestering, reduce pollution by absorbing nutrients and provide habitats for plant and animal species. This is on top of benefits to industry such as providing nurseries for fish or supporting fertile soils for agriculture (DEE 2018, Datta & Gosh 2015). Therefore, the protection and management of these unique ecosystems is crucial to continue the economic, social and environmental benefits. This is particularly true with the growing pressures such as agriculture, rural and urban development and increased groundwater usage, which the Ramsar Wetland Convention estimates have contributed to the decline in the extent of global wetlands between 64-71% (Gardner et al. 2015).
The Ramsar Convention, an intergovernmental treaty, aims to counter this global trend of wetland loss and degradation through the designation of sites considered unique or important for conserving biological diversity. Once a set of criteria is met, sites can be designated as Wetlands of International Importance (Ramsar 2018). That Ramsar site is then required to be managed by that country with the express aim of conserving and ensuring its “wise use”, or the “maintenance of ecological character… within the context of sustainable development” (Ramsar 2010). Loosely there are two sets of criteria for classifying a site with a Ramsar status. One for unique wetlands and one for sites important in conserving biological diversity (Ramsar 1971).
Cambodia, a country of 181, 035 , has an estimated 20-30% of its total area as wetlands on which the country is both culturally and economically reliant (Sophal 2004). The Mekong River and its tributaries drive the hydrology of the area. Though the overall population density of Cambodia is relatively low for a southeast Asian country (89.3 people per square kilometre in 2016) the areas adjacent to the Mekong and Lake Tonle Sap are estimated at 500 people per square kilometre (Knoema 2018). Tonle Sap Lake, the heart of the system, is uniquely known for its pulsing flushes. From November to May the lake drains into the Mekong, shrinking substantially and then undergoes significant flooding during the monsoon season as water from the Mekong reverses the flow of the Tonle Sap river into the Lake. The Lake absorbs this influx of water, increasing in size anywhere between 4 to 6 times (Hoskin & Hopkins 1991), releasing slowly post flood (DeClerck et al. 2013). This flushing system results in momentous changes between wet and dry seasons on top of significant variations between wet and dry years. This variation has historically been difficult to predict and map (MacAlister & Mahaxay 2009).
Image Source: Mekong Heritage Travel
This pulsing is the life support of the system, driving major biogeochemical cycles (Junk et al. 1989). Of the many communities which live and rely upon the system, an estimated 80% of the population rely on the production of wetlands resources (Quan et al. 2015). The pulses provide the necessary nutrients and water to sustain the productive freshwater and agricultural systems in the area (Davies et al. 2014, Flower & Fortnam 2015). When the Mekong inflows into the Lake an estimated 100 distinct species of migratory fish follow the waters to spawn in the Lake (Lovgren 2017), this increased activity is crucial to communities heavily reliant on fishing (MacAlister & Mahaxay 2009). Freshwater fisheries account for 82% of the total fishery production in Cambodia (Sithirith 2015). In general, communities within the floodplain are known to divide their time between rice cultivation and fishing (DeClerck et al. 2013). A sizable portion of the predominantly rural communities (80% of the total population) rely on subsistence farming, with the cycle of vulnerability of these people driven by high rates of poverty in combination with a country particularly susceptible to natural disasters, preventing producers reaching a state where they can sell large surplus’ (New Agriculturists 2018).
The Royal Government of Cambodia recently converted the “fishing lot” system or fishing territory allocation to open fishing on the Tonle Sap once again, a concern for fish conservationists. There are four Ramsar recognised wetland areas (sites), two of which are found within the vicinity of Tonle Sap. However, it has long been recognised that outside of these efforts the lack of coordination between various government agencies, the reliance on international aid, the neglect on the provincial level to maintain conservation efforts, the limited budget allocated to wetland protection, the issues with government regulation enforcement and the lack of information available on wetland management, are all significant barriers to the proper protection and management of wetland areas in Cambodia (Sophal 2004, Flower & Fortnam 2015). Particularly within the keeping of “wise use” for the prolonged benefits of ecosystem services.
The limitations to the Ramsar wetland sites of importance is in the criteria itself. Particularly when applied to a country such as Cambodia, where the reliance on wetlands and the associated ecosystem services is clear, but there is great uncertainty in relation to management of natural resources within the floodplain areas. Along with disorganised and segmented management, a lack of legal definition of wetland areas, defined by Cambodian Law, is preventing areas outside of the Ramsar status from obtaining appropriate protection and management (Sophal 2004, MacAlister & Mahaxay 2009).
Though flooding is vital to the systems overall function the Mekong River Commission estimates that the total damage to agriculture and infrastructure caused by flooding is 60-70 million USD annually (MRC 2018). The 2013 flood, the magnitude of which is not considered anywhere within the scale of a typical 1 in 100-year flood event for the region (MRC 2014), affected 1.8 million people across 20 provinces (Flower & Fortnam 2015). Floods with this level of impact have been increasing in severity and frequency. Previously major disruptive flood events would occur every 5 years, floods considered to have reached disaster level have occurred every 1-3 years since 1990 (Flower & Fortnam 2015). Climate change is expected to increase the magnitude, volume and duration of floods, particularly during the August-November flood pulse (ADB 2014, Keskinen et al. 2010). Cambodia was ranked eighth on a list of 193 countries, in terms of a Climate Change Vulnerability, an index rating the exposure, sensitivity and potential adaptability of countries. This vulnerability driven by the frequent exposure to droughts and floods as well as the countries reliance on agriculture and the lack of capacity for adaptation due to low incomes and limited support infrastructure (Maplecroft 2014, Flower & Forntam 2015). Water resources infrastructure development is predicted to change the flood regime (Xu et al. 2009, Grumbine & Xu 2011, Lauri et al. 2012, Flower & Fortnam 2015), with an estimated 126 hydropower projects currently under consideration between nations within the Mekong Basin (ADB 2014). This development is expected to “dampen” the flood pulse as well as shift spatial habitat patterns and interrupt migratory fish routes (Flower & Forntam 2015). This is coupled with the growing influence of expanding agricultural areas, increasing soil erosion driven by deforestation, in turn increasing sedimentation (Torell et al. 2004) which will influence the system’s capacity for storage of monsoon flood water (DeClerk et al. 2004).
Previous studies have been conducted into mapping and classifying wetland areas for management purposes, to ensure the continued sustainable use of their ecosystem services. In New Zealand, for example Ausseil et al. (2007) used a relatively simple digital terrain model with self-created landscape indicators to assign a score to a wetland based on the relative importance of each indicator. The City of Quebec used a Multi Criteria Analysis (MCA) method to access the ecological value of wetlands within its bounds (Lavoie et al. 2016). A method for wetland identification by Hunter et al. (2012) attempted to address the issues relating to poor mitigation site selection and the subsequent failure to recover full wetland function, by using a complex patchwork identification model coupled with an entropy based predictive model. Though effective, some of these techniques would be difficult to apply a location such as Cambodia, where the large flood plain area undergoes such drastic changes over the course of even just a year and there is limited data available. There is a need for a simple, cost effective method that can be easily adapted for site specific needs which can include the cultural and social reliance of areas, such as the floodplains of Cambodia, on the ecosystem functioning of wetlands.
The purpose of this report is to investigate the possible use of a simple Boolean Overlay (BO) through a Geographical Information System (GIS) in determining areas for mitigation, management and protection of wetlands. This is with the overall aim of determining the effectiveness of a simple method in accounting for the various forms of reliance on wetlands, as well as contributing to increasing the awareness of the importance of wetlands and the ecosystem services they provide.
With the current strains on resources, the lack of official government guidance, the vulnerability to natural disasters and climate change, as well as the heavy cultural and economical reliance on wetlands and the floodplain, the Kingdom of Cambodia is a suitable example study area to conduct a MCA to determine areas of the floodplain which could potentially provide the greatest benefit. The secondary purpose of this investigation is to evaluate whether an approach that considers the services provided by wetlands will differ to an approach that primarily considers vulnerability in light of future changes to the system.
This analysis has been separated into three parts:
Part 1: The Study Area – The Floodplain
The determination the extent of the floodplain in relation to the link between possible wetland areas and the comparison to established Ramsar sites.
Part 2: The Site-Specific Ecosystem Services
The identification of sites of wetland areas for priority management, based on specific ecosystem services provided.
Part 3: The Future Vulnerability
The identification of areas with the greatest vulnerability and how this knowledge could influence the identification of sites for possible wetland management. A comparison will then be conducted between these areas and those identified based on the ecosystem services provided.
Materials and Methods
Data and Software
It was decided that a model builder approach within ArcGIS would be an efficient method to prepare data for pre-processing before applying an MCA. Figure 1, Model Builder 1, shows the steps undertaken to prepare the data for analysis and the outputs necessary for Part 1. Model Builder 2, Figure 2, shows the steps taken for Parts 2 and 3 of the analysis. A list of the initial data sets (and sources) modified for the purposes of this analysis can be found in at the end.
A BO is the intersection of data layers that have been selected via binary methods, such as AND or OR. A simple BO form of an MCA was chosen due to its simplicity. This was partly the case because of limitations on data type, quality and detail. Rather than change site for this analysis it was considered part of the process to determine if a BO approach would be useful with such limitations on data. For example, ideally one of the criteria for selection would be based on the extent of a 1 in 100-year flood, however this data was not available for the region. BO is used in to create the proposed sites based on the binary separation of data created in Parts 1, 2 and 3. This output final layer was a result of the attempt to prioritise various ecosystem services and was then compared to an output map created from considering areas of varying future vulnerability.
Part 1: Study Area
As discussed in previously, Part 1 of this analysis was the identification of the specific study area for further steps of the analysis. There was no data available with predesignated areas classified as wetlands, as this has not been undertaken on the national government scale. Therefore, a floodplain area was determined to be sufficient for the purposes of this analysis. To determine the floodplain area, data was collated with land that has historically been subjected to inundation. Those areas which were hydrologically defined as being either non-perennial, intermittent or fluctuating, were all considered to be within the definition of a floodplain and therefore likely to contain areas of wetland.
A “typical” flood data set was obtained, to compare and validate this area allocated as floodplain. The 2013, September/November flood was chosen as the conditions of this flood have been shown to be within historical norms, with peak discharge ranging across the extent of the flood from average to below average (Adamson et al. 2014). This is in conjunction with limitations on data availability. A 2013 flood extent dataset was obtained and modified (e.g Clip tools) as shown in Figure 1, to obtain an estimation of the area affected by the flood that was not in the designated floodplain of this analysis.
Established Ramsar sites were pulled from data containing the natural protected areas of Cambodia. The area adjacent to Tonle Sap within this data set had no formal protection status allocated and was in fact designated as “multi-use” and therefore would be of no further assistance in this analysis and in fact further confirmed the need to conduct an investigation such as this one.
All datasets were clipped to only include areas within the administrative bounds of Cambodia. Population data was associated with certain layers when necessary, using the Add Field tool. The floodplain layer created in this part would then become the bounds of the study area for the rest of the analysis.
Part 2: Ecosystem Services
To include the benefits of ecosystem services in consideration for site selections for wetland management, key services within the context of the Cambodian floodplains needed to be identified. Arguably, as discussed in previously, flood protection is clearly an dominant ecosystem service for Cambodia. Therefore, to represent this reliance on flood protection within the floodplain the following aspects were investigated.
Regions reliant on agriculture
As discussed previously, subsistence farming is an integral part of the culture in Cambodia which has historically in this form been vulnerable to flooding. Therefore, areas that were designated for agricultural land use were identified and accessed. Around 90% of the agricultural production of the entire country is rice (New Agriculturists 2018). Hence, for a more in-depth analysis, the percentage of land cover by rice was found at the smallest administration level that data was available for (communes).
Regions (by commune) reliant on fish farming
It is estimated that 80% of the human consumed protein in Cambodia is sourced from freshwater fish (Arias et al. 2013). Data representing communes reliant on fish farming was also included, which was based on the results of the 2013 Census.
Regions with higher poverty rates
A study in 2014 (Davies et al. 2014), found that the incidences and risks associated with water borne diseases increases with flooding in Cambodia and are related to the inaccessibility of health service as well as inability, without access to power, for people to treat water. It is estimated that a loss of around 1,200 riels (equivalent of 30 cents US) per day of income would plunge 3 million Cambodians back into poverty, which would double the rate of poverty in Cambodia (World Bank 2014). Considering this, poverty rates per commune was considered a useful method for demonstrating the reliance on flood protection within the floodplain as it could be considered that those areas of higher poverty rates would likely be more susceptible to flooding.
After an initial assessment of the results from parts 2a, b and c, the following criteria was established in order to create a proposed area for wetland protection within the floodplain of Cambodia:
- Communes which have more than 75% of the land use as predominately for rice farming
- Communes deemed dependent on fish farming
- Communes where more than 30% of the population are below the national poverty line
By binary selection and then exporting from the layers established in Model Builder 2 the criteria above were individually segregated. Then through the use of the Union tool (BO “AND” method) the combination of these three criteria were used to create an entirely new layer of the proposed area for wetland protection and management.
Part 3: Future Vulnerability
Cambodia is considered a Kingdom relatively vulnerable to climate change, as discussed previously. Data was available from the Economy and Environment Program for Southeast Asia, a program administrated by WorldFish (World Fish 2013), that visualised the vulnerability of areas in Cambodia to climate change. The data represented the level of exposure to climate variability and sensitivity to climatic stresses, as well as relative adaptability through an index value from 0 to 1, 1 being the highest vulnerability rating.
Another key future vulnerability of the system is the expected influence on the flood pulse system of growing hydrological development. To represent this, one predicted change to the system was spatially applied to show the area that would be affected by this change. Expected changes to the Mekong are that the wet season on the river system will lower by 40-50cm (Lauri et al. 2012). This was found to relate to a horizontal shift of around 500m, creating a reduction across the system in the floodplain area that is annually flooded (Arias et al. 2013). A Buffer tool was used to apply this reduction of 500m to the floodplain area to find the remaining area that would be affected.
Results and Discussion
Part 1: Study Area
The result of Part 1 of this analysis are shown in Map 1. It is clear, as expected, that the areas designated as the floodplain, the study area for this analysis, are those adjacent to Lake Tonle Sap and other major water bodies of Cambodia. As can be seen in Map 1, all those areas defined as Ramsar sites are within the floodplain. These only account for 3% of this total area of floodplain. In comparison the floodplain itself is a total of 13% of the surface area of the country. The 2013 flood extent was also compared to this floodplain. 24% of the area affected by this “typical” flood was outside the areas classified as regularly inundated; the floodplain.
Part 2: Ecosystem Services
As can be seen in Map 2, accessing agricultural areas based on land use appears to have given a substantial underestimation, that does not match what is known about the prevalence of subsistence farming. The areas allocated for agricultural use are also predominantly isolated from the adjacent areas of the lake, also outside expectations in terms what has previously been discussed as the cultural/social trends within the floodplain. Map 3 however shows some clear patterns in terms of a separation between communes and the reliance on rice farming and fishing. Approximately 25% of the floodplain area contains communes with greater than 75% by area of rice farming and 23% of the floodplain contains areas that are classified, based on the 2013 Census as being dependent on fish farming. Map 4 indicates that 11% of the floodplain area has 30% or more of its population living below the national poverty line.
The resultant area found through the combination all the three criteria given in he Methodology section was approximately 50% of the total floodplain area, as shown in Map 5.
Part 3: Future Vulnerability
Map 6 shows the result of the combination of the two vulnerabilities under consideration. Of the total floodplain area of Cambodia, approximately 27% has a climate change vulnerability rating of greater than 0.5. This is classified as a “high” level of vulnerability to climate change, as given by the Economy and Environment Program, Southeast Asia by Worldfish (Worldfish 2013). The area found to be affected by a 500m horizontal reduction in the floodplain was approximately 32% of the total floodplain area.
A simple visual inspection of Map 1 indicates that, for the purposes of this study, it was justified to only consider the floodplain extent and not the extent of the 2013 flood. The flood data appears incomplete in certain areas and where it is complete however, the data set serves to validate the determined floodplain area. It may be a slight underestimation of the potential flood area but was deemed suitable for the purposes of this analysis. The inability to consider the results from Map 2 as part of the final site selection for the potential area for wetland protection, highlights the issues in data reliability. The assumption is for the rest of the analysis is that with such a transient system the data available would still be sufficient to represent patterns, particularly on a large, preliminary scale.
The areas found to be reliant on fish farming and rice cultivation in this analysis are, to a certain extent, within expectations based on previous studies. Perhaps the difficulties in mapping and recording such a transient system has led to a possible slight underestimation of these two factors. It was estimated in 2004 that 84% of cultivated land was rice crops (Yu & Fan 2009), and it is commonly understood that those who live within the floodplain are self-reliant on a combination of fishing and rice farming. The area of the floodplain (27%) found to have a climate change vulnerability index greater than 5 was smaller than expected, considering that the overall country rating for Cambodia has regularly been given as above 7 over the last decade. The area found to be affected by a 500m horizontal reduction in floodplain was 7,531 , which is approximately 4% of the total land surface of Cambodia. This 32% reduction of floodplain is greater than that estimated by Arias et al. (2013), of 22% of the seasonal floodplain.
Despite these relatively small deviations from expected patterns, the BO has shown to be useful, particularly as a cost effective and simplified method for a broad form of site identification. The results of Map 5, if nothing else, stress the importance of the floodplain areas surrounding Tonle Sap and qualitatively highlight the extent of the reliance of the floodplain communities on the wetland system. Though the final proposed area for possible protection was a sizable portion of the study area, this analysis has shown the usefulness of the method in a preliminary assessment in site identification. It is popular opinion that to overcome some of the governmental barriers in designating wetland areas for management and protection, a deviation from the top down approach is necessary. Specifically, that more community participation is necessary, coupled with an improvement to the general understanding of wetland functions (Sophal 2004, Sithirith 2015). The method could be an effective support tool in identifying which communes and communities could become a focus within management strategies for wetland protection.
It is clear from comparing Maps 5 and 6 that the two approaches for identifying wetlands areas for priority management will give differing results. Though there are some similarities in areas adjacent to the southwest of Tonle Sap, which could be considered important for protection and management if considering both ecosystem services as well as relative future vulnerability.
Limitations to this method include the need for on the ground, community engagement to confirm the feasibility of sites selected for management. Further analysis could be conducted into potential applications of this method in establishing buffer zones around wetland areas as protections from agriculture fields and communities, particularly considering the results found applying reductions to wetland areas from hydrological influences. Depending on data availability, potential next steps could be to apply a weighted analysis and compare the level of detail and area that would result.
The purpose of this study was to investigate the potential for a simple method such as a BO in providing guidance for site selection for wetland management. This study was successful in establishing the feasibility of this method, particularly when applied to creating criteria based upon the site-specific ecosystem benefits of wetlands and the potential vulnerabilities of the system. A clear takeaway from this analysis is that the allocated Ramsar wetlands are a small fraction of the area that could potentially be prioritised for management, when considering wetlands of “importance”. Under the Ramsar Convention the “wise use” of wetlands includes not only the long-term maintenance but also “human well-being” and the “alleviation of poverty”. Within these definitions, and the results of this analysis implies that there is a need to closely examine how wetland areas are related to wellbeing and the alleviation of poverty, within a specific site or within the context of a country’s social, economic and environmental reliance on wetlands.
Adamson, P, Hak, S, Phethany, B, Bauasuna, B & Pham, T 2014. ‘Annual Mekong Flood Report 2013.’ Mekong River Commission. Available from: http://www.mrcmekong.org/assets/Publications/basin-reports/Annual-Mekong-Flood-Report-2013.pdf
Arias, M, Cochrane, T, Norton, D, Kileen, T & Khon, P 2013. ‘The Flood Pulse as the Underlying Driver of Vegetation in the Largest Wetland and Fishery of the Mekong Basin’. Ambio vol42 pp864-874. Available from: http://www.jstor.org.ezproxy.library.uwa.edu.au/stable/pdf/24708988.pdf?refreqid=excelsior:9305021a5380325f0fd6bd3b551a3725
Ausseil, A, Dymond, E, Shepard, D & James, D 2007. Rapid Mapping and Prioritisation of Wetland Sites in the Manawatu-Wanganui Region, New Zealand’. Environmental Management vol 39 pp316-325. Available from: https://search-proquest-com.ezproxy.library.uwa.edu.au/docview/729721945?OpenUrlRefId=info:xri/sid:primo&accountid=14681
Datta, D & Gosh, P 2015. ‘Evaluating sustainability of community endeavours in an Indian floodplain wetland using multi‐criteria decision analysis.’ Singapore Journal of Tropical Geography vol 36 pp38-56. Available from: https://onlinelibrary-wiley com.ezproxy.library.uwa.edu.au/doi/full/10.1111/sjtg.12092
Davies, G, McIver, L, Kim, Y, Hashizume, MIddings, S & Chan, V 2014. ‘Water-borne diseases and extreme weather events in Cambodia: review of impacts and implications of climate change.’ Envir Res Public Health vol 23 pp191-203. Available from: https://www.ncbi.nlm.nih.gov/pubmed/25546280
DeClerck, F, Kosal, M & Johnstone, G 2013. ‘CGIAR: Preventing Cardiac Arrest for Cambodia’s Heart’. CGIAR. Available from: https://18.104.22.168/thrive/2013/05/22/preventing-cardiac-arrest-cambodia%E2%80%99s-heart
DEE 2018. Department of Energy and Environment 2018. http://www.environment.gov.au/wetlands
Flower, B & Fortnam, M 2015. ‘Urbanising Disaster Risk’.Available from: https://reliefweb.int/sites/reliefweb.int/files/resources/47109_urbanisingdisasterriskreportinteractive.pdf
Gardner, R.C., Barchiesi, S., Beltrame, C., Finlayson, C.M., Galewski, T., Harrison, I., Paganini, M., Perennou, C., Pritchard, D.E., Rosenqvist, A., and Walpole, M. 2015. State of the World’s Wetlands and their Services to People: A compilation of recent analyses. Ramsar Briefing Note no. 7. Gland, Switzerland: Ramsar Convention Secretariat. Available from: https://www.ramsar.org/sites/default/files/documents/library/bn7e_0.pdf
Grumbine & Xu 2011. ‘Environment and development. Mekong hydropower development.’ Science. Vol 332 pp178-179. Available from: https://www.ncbi.nlm.nih.gov/pubmed/21474740
Hoskin J & Hopkins A 1991. ‘The Mekong River and its people’. Published Bangkok. Available from: https://catalogue.nla.gov.au/Record/1285243
Junk, W. J., P. B. Bayley, and R. E. Sparks. 1989. The flood pulse concept in river- floodplain systems. In Proceedings of the International Large River Symposium (LARS), ed. by D. P. Dodge, pp. 110–127. Canadian Special Publication of Fisheries and Aquatic Sciences, Ottawa, Canada.
Keskinen, M, Kummu, M, Kakonen M & Varis, O 2010. ‘Mekong at the Crossroads: Next Steps for Impact Assessment of Large Dams’. Ambio vol 41 pp319-324. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3357849/
Knoema 2018. ‘Knoema: Cambodia Population Density’, 15 May 2018. Available from: https://knoema.com/atlas/Cambodia/Population-density
Lovgren, S 2017. “Enourmous fish make one of the world’s largest migrations’. National Geographic Website. Available from:https://news.nationalgeographic.com/2017/03/mekong-river-fish-migrations/
MacAlister, C & Mahaxay, M 2009. ‘Mapping wetlands in the Lower Mekong Basin for wetland resource and conservation management using Landsat ETM images and field survey data.’ Journal of Environmental Management. Vol 90 pp2130-2137. Available from:https://www-sciencedirect-com.ezproxy.library.uwa.edu.au/science/article/pii/S0301479708000339?via%3Dihub
Maplecroft 2014. ‘Climate Change and lack of food security: Climate Change and Risk Atlas.” Available from: https://maplecroft.com/portfolio/new-analysis/2014/10/29/climate-change-and-lack-food-security-multiply-risks-conflict-and-civil-unrest-32-countries-maplecroft/
Mitsch, W & Gosselink, J 1993. ‘The value of wetlands: importance of scale and landscape setting.’ Ecological Economics, vol 35 pp25-33. Available from: https://pdfs.semanticscholar.org/6640/bcf8b1997a4effe595cfe451fe71262a46ee.pdf
MRC 2018. ‘Flood and Drought’. Mekong River Commission. Available from: http://www.mrcmekong.org/topics/flood-and-drought/
New Agriculturists 2018. ‘Country profile: Cambodia’. New agriculturists website. Available from: http://www.new-ag.info/en/country/profile.php?a=860
Pheakdey, S & Naval, C 2018. ‘Strengthening local communities in Lake Tonle Sap’. Ramsar website. Available from: http://www.worldwetlandsday.org/stories/-/asset_publisher/tKO0MQoSsBTr/content/angrove-restoration-to-ensure-sustainable-livelihoods-in-senegal
Ramsar 1971. ‘Criteria for sites of international importance.’ Available from: https://www.ramsar.org/sites/default/files/documents/library/ramsarsites_criteria_eng.pdf
Ramsar 2010. ‘Handbook 1: Wise Use of Wetlands’ Available from: https://www.ramsar.org/sites/default/files/documents/library/hbk4-01.pdf
Ramsar 2018. Ramsar Wetlands Homepage, 20 May 2018, Available from: https://www.ramsar.org/
Richardson, C 1994. ‘Ecological functions and human values in wetlands: A framework for assessing forestry impacts.’ Wetlands. Vol 14 pp 1-9. Available from: https://link.springer.com/article/10.1007/BF03160616
Sithirith, M 2015. ‘The Governance of Wetlands in the Tonle Sap Lake, Cambodia’. ReasearchGate. Available from: https://www.researchgate.net/publication/282978187_The_Governance_of_Wetlands_in_the_Tonle_Sap_Lake_Cambodia
Sophal, M 2004. ‘Sustainable utilization and management of Cambodia wetland.’ Available from: https://www.researchgate.net/publication/23550923_Wetlands_management_in_Cambodia_socioeconomic_ecological_and_policy_perspectives
Torell, M, Salamanaca, A & Ratner, R 2004. ‘Wetlands management in Cambodia: socioeconomic, ecological, and policy perspectives’. Reasearch Gate. Available from: https://www.researchgate.net/publication/23550923_Wetlands_management_in_Cambodia_socioeconomic_ecological_and_policy_perspectives
World Bank 2014. ‘Poverty has fallen, yet many Cambodians are still at risk of slipping back into poverty, new report finds’. World Bank Press Release. Available from: http://www.worldbank.org/en/news/press-release/2014/02/20/poverty-has-fallen-yet-many-cambodians-are-still-at-risk-of-slipping-back-into-poverty
World Fish 2013. Worldfish Homepage. Available from: https://www.worldfishcenter.org