Yongyut Trisurat

Projecting land-use change and its consequences for biodiversity in northern Thailand.

Rapid deforestation has occurred in northern Thailand over the last few decades and it is expected to continue. The government has implemented conservation policies aimed at maintaining forest cover of 50% or more and promoting agribusiness, forestry, and tourism development in the region. The goal of this paper was to analyze the likely effects of various directions of development on the region. Specific objectives were (1) to forecast land-use change and land-use patterns across the region based on three scenarios, (2) to analyze the consequences for biodiversity, and (3) to identify areas most susceptible to future deforestation and high biodiversity loss. The study combined a dynamic land-use change model (Dyna-CLUE) with a model for biodiversity assessment (GLOBIO3). The Dyna-CLUE model was used to determine the spatial patterns of land-use change for the three scenarios. The methodology developed for the Global Biodiversity Assessment Model framework (GLOBIO 3) was used to estimate biodiversity intactness expressed as the remaining relative mean species abundance (MSA) of the original species relative to their abundance in the primary vegetation. The results revealed that forest cover in 2050 would mainly persist in the west and upper north of the region, which is rugged and not easily accessible. In contrast, the highest deforestation was expected to occur in the lower north. MSA values decreased from 0.52 in 2002 to 0.45, 0.46, and 0.48, respectively, for the three scenarios in 2050. In addition, the estimated area with a high threat to biodiversity (an MSA decrease >0.5) derived from the simulated land-use maps in 2050 was approximately 2.8% of the region for the trend scenario. In contrast, the high-threat areas covered 1.6 and 0.3% of the region for the integrated-management and conservation-oriented scenarios, respectively. Based on the model outcomes, conservation measures were recommended to minimize the impacts of deforestation on biodiversity. The model results indicated that only establishing a fixed percentage of forest was not efficient in conserving biodiversity. Measures aimed at the conservation of locations with high biodiversity values, limited fragmentation, and careful consideration of road expansion in pristine forest areas may be more efficient to achieve biodiversity conservation.

Integrating land use and climate change scenarios and models into assessment of forested watershed services in Southern Thailand

The Thadee watershed, covering 112km(2), is the main source of water for agriculture and household consumption in the Nakhon Srithammarat Province in Southern Thailand. As the natural forests upstream have been largely degraded and transformed to fruit tree and rubber plantations, problems with landslides and flooding have resulted. This research attempts to predict how further land-use/land-cover changes during 2009-2020 and conceivable changes in rainfall may influence the future levels of water yield and sediment load in the Thadee River. Three different land use scenarios (trend, development and conservation) were defined in collaboration with the local stakeholders, and three different rainfall scenarios (average rainfall, climate change and extreme wet) were determined on the basis of literature sources. Spatially explicit empirical modelling was employed to allocate future land demands and to assess the contributions of land use and rainfall changes, considering both their separate and combined effects. The results suggest that substantial land use changes may occur from a large expansion of rubber plantations in the upper sub-watersheds, especially under the development land use scenario. The reduction of the current annual rainfall by approximately 30% would decrease the predicted water yields by 38% from 2009. According to the extreme rainfall scenario (an increase of 36% with respect to current rainfall), an amplification of 50% of the current runoff could result. Sensitivity analyses showed that the predicted soil loss is more responsive to changes in rainfall than to the compared land use scenarios alone. However, very high sediment load and runoff levels were predicted on the basis of combined intensified land use and extreme rainfall scenarios. Three conservation activities-protection, reforestation and a mixed-cropping system-are proposed to maintain the functional watershed services of the Thadee watershed region.

Improving the viability of large-mammal populations by using habitat and landscape models to focus conservation planning

Context. Assessing the viability of animal populations in the wild is difficult or impossible, primarily because of limited data. However, there is an urgent need to develop methods for estimating population sizes and improving the viability of target species. Aims. To define suitable habitat for sambar (Cervus unicolor), banteng (Bos javanicus), gaur (Bos gaurus), Asian elephant (Elephas maximus) and tiger (Panthera tigris) in the Western Forest Complex, Thailand, and to assess their current status as well as estimate how the landscape needs to be managed to maintain viable populations. Methods. The present paper demonstrates a method for combining a rapid ecological assessment, landscape indices, GIS-based wildlife-habitat models, and knowledge of minimum viable population sizes to guide landscape-management decisions and improve conservation outcomes through habitat restoration. Key results. The current viabilities for gaur and elephant are fair, whereas they are poor for tiger and banteng. However, landscape quality outside the current distributions was relatively intact for all species, ranging from moderate to high levels of connectivity. In addition, the population viability for sambar is very good under the current and desired conditions. Conclusions. If managers in this complex wish to upgrade the viabilities of gaur, elephant, tiger and banteng within the next 10 years, park rangers and stakeholders should aim to increase the amount of usable habitat by ~2170 km2 or 17% of existing suitable habitats. The key strategies are to reduce human pressures, enhance ungulate habitats and increase connectivity of suitable habitats outside the current distributions. Implications. The present paper provides a particularly useful method for managers and forest-policy planners for assessing and managing habitat suitability for target wildlife and their population viability in protected-area networks where knowledge of the demographic attributes (e.g. birth and death rates) of wildlife populations are too limited to perform population viability analysis.

Consequences of land use change on bird distribution at Sakaerat Environmental Research Station

The objectives of this research were to predict land-use/land-cover change at the Sakaerat Environmental Research Station (SERS) and to analyze its consequences on the distribution for Black-crested Bulbul (Pycnonotus melanicterus), which is a popular species for bird-watching activity. The Dyna-CLUE model was used to determine land-use allocation between 2008 and 2020 under two scenarios. Trend scenario was a continuation of recent land-use change (2002-2008), while the integrated land-use management scenario aimed to protect 45% of study area under intact forest, rehabilitated forest and reforestation for renewable energy. The maximum entropy model (Maxent), Geographic Information System (GIS) and FRAGSTATS package were used to predict bird occurrence and assess landscape fragmentation indices, respectively. The results revealed that parts of secondary growth, agriculture areas and dry dipterocarp forest close to road networks would be converted to other land use classes, especially eucalyptus plantation. Distance to dry evergreen forest, distance to secondary growth and distance to road were important factors for Black-crested Bulbul distribution because this species prefers to inhabit ecotones between dense forest and open woodland. The predicted for occurrence of Black-crested Bulbul in 2008 covers an area of 3,802 ha and relatively reduces to 3,342 ha in 2020 for trend scenario and to 3,627 ha for integrated-land use management scenario. However, intact habitats would be severely fragmented, which can be noticed by total habitat area, largest patch index and total core area indices, especially under the trend scenario. These consequences are likely to diminish the recreation and education values of the SERS to the public.

Basin-wide impacts of climate change on ecosystem services in the Lower Mekong Basin

Water resources support more than 60 million people in the Lower Mekong Basin (LMB) and are important for food security—especially rice production—and economic security. This study aims to quantify water yield under near- and long-term climate scenarios and assess the potential impacts on rice cultivation. The InVEST model (Integrated Valuation of Ecosystem Services and Tradeoffs) forecasted water yield, and land evaluation was used to delineate suitability classes. Pattern-downscaled climate data were specially generated for the LMB. Predicted annual water yields for 2030 and 2060, derived from a drier overall scenario in combination with medium and high greenhouse gas emissions, indicated a reduction of 9–24% from baseline (average 1986–2005) runoff. In contrast, increased seasonality and wetter rainfall scenarios increased annual runoff by 6–26%. Extreme drought decreased suitability of transplanted rice cultivation by 3%, and rice production would be reduced by 4.2 and 4%, with and without irrigation projects, relative to baseline. Greatest rice reduction was predicted for Thailand, followed by Lao PDR and Cambodia, and was stable for Vietnam. Rice production in the LMB appears sufficient to feed the LMB population in 2030, while rice production in Lao PDR and Cambodia are not expected to be sufficient for domestic consumption, largely due to steep topography and sandy soils as well as drought. Four adaptation measures to minimize climate impacts (i.e., irrigation, changing the planting calendar, new rice varieties, and alternative crops) are discussed.

The International Long-Term Ecological Research–East Asia–Pacific Regional Network (ILTER-EAP): history, development, and perspectives

There are growing needs to broaden and deepen our multi-faceted understanding of the ecosystems, and the networks of Long-Term Ecological Research (LTER) can play significant roles in fostering and applying ecosystem studies at regional and global scales. The International LTER Network (ILTER) is organized as a global network of field research sites and scientists to address current ecological issues such as biodiversity loss and ecosystem degradation within a globally changing environment. The ILTER East Asia–Pacific Regional Network (ILTER-EAP) is one of the four constituent ILTER regional networks. Since 1995, ILTER-EAP has been developed to promote data sharing, research collaborations and capability building in the science and to bridge gaps between societal needs and scientific imperatives on concerns in the Asia–Pacific Region. Currently, ILTER-EAP comprises nine formal ILTER members and two associate networks. Their activities involve long-term and multiple-site observations of structural, functional and developmental aspects of ecosystems, data sharing, and bridging society and ecological science. This paper presents a review of the activities of ILTER-EAP, focusing on its: (1) vision and the development following its inception, (2) scientific activities and major outputs related to selected thematic areas, (3) contributions from ILTER-EAP to the international initiatives, and (4) future challenges and opportunities relating to its development and role in facilitating regional and global research collaborations. Accordingly, regional research questions were identified that could be most effectively addressed by opening up a common research platform, integrated data management system and the network science, which is open to all interested parties.

Assessing potential effects of land use and climate change on mammal distributions in northern Thailand

Abstract Context. Tropical ecosystems are widely recognised for their high species richness and outstanding concentrations of rare and endemic species. Previous studies either focussed on the effects of deforestation or climate change, whereas studies on the combined effects of these two major threats are limited. Aims. This research aimed to model current and future distributions of medium- to large-sized mammal species on the basis of different land-use and climate-change scenarios in 2050 and to assess whether the predicted effects of land-use change are greater than those of climate change and whether the combined effects of these drivers are greater than those of either individual driver. Methods. The present article demonstrates a method for combining nationwide wildlife-inventory data, spatially explicit species-distribution models, current and predicted future bioclimatic variables, other biophysical factors and human disturbance to map distributions of mammal species on the basis of different land-use and climate-change scenarios and to assess the role of protected areas in conservation planning. Key results. Seventeen medium- to large-sized mammal species were selected for modelling. Most selected species were predicted to lose suitable habitat if the remaining forest cover declines from the current level of 57% to 50% in 2050. The predicted effects of deforestation were stronger than the effects of climate change. When climate and land-use change were combined, the predicted impacts were more severe. Most species would lose suitable habitat and the average shift in species distribution was greater than 40%. Conclusions. The predicted effects were positive for only a few species and negative for most species. Current and future centres of mammal-species richness were predicted in large and contiguous protected forests and the average contribution of existing and proposed protected areas in protecting the focal species will increase from 73% to 80% across all scenarios. Implications. The present research advances the current understanding of the ecology of 17 medium- to large-sized mammal species with conservation relevance and the factors that affect their distributions at the landscape scale. In addition, the research demonstrated that spatially explicit models and protected areas are effective means to contribute to protection of mammal species in current and future land-use and climate-change scenarios.

Long-Term Monitoring and Prediction of Ecosystem Using Remote Sensing and the CLUE-S Model: Sakaerat Environmental Research Station

Deforestation has been given much attention in land use and landscape changes because of the high rate of forest change and the ecological importance of the forest ecosystem. Forest loss in Thailand was ranked the highest of all countries in the Greater Mekong subregion and as fourth in the “top 10 of tropical countries in terms of annual rate of loss in 1995 (CFAN 2005). The Royal Forest Department in Thailand has had long-term monitored forest cover using satellite images for the last four decades. The results revealed that in 1961 the forest cover was 53.3% of the country area whereas approximately 25.2% remained in 1998. The average annual loss was approximately 400,000 ha, or 2.0%, during this 37-year period (RFD 2005). Deforestation in Thailand is mainly caused by commercial logging of primary forest, by agribusiness, and by urban development, driven by ongoing population growth and the national development strategy to gain foreign income (TRisurat 2007).

Effects of Land Use and Climate Change on Siamese Eld's Deer (Rucervus eldii siamensis) Distribution in the Transboundary Conservation Area in Thailand, Cambodia, and Lao PDR

The Emerald Triangle Protected Forests Complex (ETFC) is recognized as a globally outstanding area for transboundary biodiversity conservation. Eld’s deer (Rucervus eldii) include three subspecies; R. eldii siamensis, R. eldii eldii, and R. eldii thamin. This research focused on Siamese Eld’s deer (R. eldii siamense), which is one of the 10 critically endangered vertebrates found in the ETFC. Its habitats are threatened by forest conversion to agriculture and human settlements, as well as by future climate change. The objectives of this article were to predict Siamese Eld’s deer distribution and to determine potential shifts in its suitable habitat as the results of different land use and climate change scenarios in 2030. Occurrence data for Siamese Eld’s deer were gathered from literature and field surveys. The Maximum Entropy (Maxent) and logistic regression models were used to generate suitable habitats. The model that generated the greatest accuracy was selected for distribution mapping. The results showed 142 records of Siamese Eld’s deer. The predicted distribution map generated from the logistic regression model provided greater accuracy (90%) than the Maxent (80%). The predicted habitats of Siamese Eld’s deer covered 6.0% of the ETFC landscape by 2013. They were concentrated in the protected areas of the lowland forests of Cambodia and Lao PDR. The land use change only did not affect the distribution of Siamese Elds Deer, but climate change would impact the distribution substantially. In addition, the combination of all future land use and climate changes would significantly reduce the current habitat to approximately 2.6% of the ETFC landscape, which is a decline 58.0%. The lowest extent of 2.5% was predicted for the combination of unsustainable land use and climate change scenario. Recommendations on biodiversity conservation cooperation among the three countries, habitat protection and ex-situ conservation were proposed.

Calibration of Hydrological Streamflow Modeling Using MODIS

LULC (land use and land cover) plays an important role in mathematical hydrological modeling. As many countries, available LULC are not always updated to reflect the most current situation. In this regard, the objective of this study was to investigate the potential capability of moderate resolution satellite imagery such as MODIS (Moderate Resolution Imaging Spectroradiometer), acquired in 2010 for updated LULC. This issue was illustrated through the application of the most current LULC as one of the data inputs of the SWAT (Soil and Water Assessment Tool) model in the Tonle Sap Lake Basin, a sub-basin of the Mekong River. The streamflow was tested using moderate resolution LULC of 500 meters. The statistical evaluation results at a monitoring station for model calibration and validation showed that the R 2 for daily and monthly values range from 0.76 to 0.88 and 0.86 to 0.89 respectively, whereas the Nash-Sutcliffe efficiency daily and monthly values range between 0.75 to 0.85 and 0.76 to 0.87 respectively. The simulation result based on MODIS imagery demonstrates LULC at moderate resolution holds considerable potential as an effective hydrological modeling tool. An additional level of confidence is provided by the notion that the methods described here could be applied in similar watershed conditions.