Evaluation of Climate and Habitat Interactions Affecting the Conservation

and Management of Asian Elephants in Southeast Sri Lanka

 

Work Plan:

          The work plan is divided into sections for A) Climate, B) Hydrology, Vegetation and Habitat, C) Elephant Management and D) Integration and Synthesis

 

Section A: Climate

In May 2000, the IRI entered into a partnership with the Mahaweli Authority of Sri Lanka to provide climate and hydrological forecasts for the Mahaweli and Walawe River Basins. The methodology we propose will draw upon these high quality, weather, climate and hydrological model output already developed for this region by IRI. 

Major Steps

§         Data collection, entry and arrangement in IRI data library

§         Construct climatologies and inter-annual variation patterns from the extensive network of station data in Sri Lanka.

§         Using interpolation techniques, develop climatologies and inter-annual variation patterns of precipitation and temperature for South-Eastern Sri Lanka at a resolution of 1-km

§         Downscale GCM seasonal climate predictions available at the IRI to the available meteorological stations and the high resolution grided data using statistical techniques to relate the retrospective seasonal predictions to past climate data.

 

The principal sections of this portion of the project will be:

1.      Data Collection: Gaps in the IRI archives of climate data for Sri Lanka will be filled with data collected from the Sri Lankan Departments of Meteorology and Irrigation. Printed format data will be entered by hand by a research assistant. Thereafter these data will be incorporated into the IRI data library for ready access, manipulation and visualization. 

2.      Construction of High-Resolution Climatology : Based on the data that is available for Sri Lanka for 400 stations, interpolation techniques will be used to construct climatologies of rainfall and temperature at a 1-km grid resolution for use in GIS.

3.      Downscaling of Seasonal Climate Forecasts : The  GCM seasonal climate forecasts available at IRI are typically at a scale of about 250 km. These forecasts will be downscaled to the available stations and the high resolution grid in the South-East of the island using relationships between IRI retrospective forecasts and actual observed climate

 

 Section B: Hydrology, Vegetation, Habitat

Among the fundamental questions underlying our understanding of the systems supporting elephants for the proposed study area are:

-          What is the location, composition, structure, extent and use of the natural resources (vegetation and surface water) and built systems present within the region

-          How sensitive are these systems to variability in climate (temperature and precipitation) and can these key variables be forecast using downscaling systems?

-          How best can these patterns and processes be captured using digital geospatial technologies

Major Steps

§         Search for and compile existing digital hydrology, land cover, habitat, NDVI history, elevation, geology and soils data

§         Generate digital layers for detailed cover type, vegetation, and biophysical (elevation, slope, aspect, soil moisture) and potential elephant habitat

§         Compare historic patterns and correlations between vegetation (NDVI) data and climate

§         Perform downscaling of climate predictions to variables related to the key features that impact elephant distribution, namely water and vegetation. This will include downscaling to in-situ streamflow and hydrological analysis of streamflows, and NDVI.

 

Sub-Section 1: Climate impact on available water for elephants:

Hydrological Analysis of Surface and Sub-surface Water

 

The main focus will be:

1.      Data Collection: A thorough hydrological data search will be conducted, with the assistance of t he Irrigation Department, the International Water Management Institute (IWMI), and the Mahaweli Authority of Sri Lanka.

2.      Statistical downscaling of climate predictions for streamflow estimation. The critical in-situ variable in hydrological modeling is streamflow. Techniques similar to that used for downscaling climate predictions to station climate will be used to downscale climate forecasts to streamflow forecasts in key sites on the rivers of the project area. (Mahaweli, Menik, Walawe).

3.      Hydrological Analysis:  Observations and downscaled forecasts of streamflow will be extended to other variables such as soil moisture (plant available) and availability of water (elephant available) in surficial features in parks, using hydrologic models, statistical models and subjective inference.

 

Sub-Section 2: Climate impact on available fodder for elephants: Vegetation,

Habitat and Biogeographic Pattern Assessments

 

The main focus will be:

1.      Data Collection: a thorough geospatial data search will be conducted, with the assistance of t he Open University of Sri Lanka, the International Water Management Institute (IWMI), and the Mahaweli Authority of Sri Lanka.

2.      Data Generation: layers lacking adequate spatial, thematic or temporal resolution for land cover, vegetation type, NDVI, elevation, slope, aspect, soil moisture and geology will be augmented with output from satellite image classification work, acquired and tailored for this project. Construction of the final units will be derived to provide a best fit of the size, productivity, composition, structure, contiguity, temporal and geomorphology context of that range of cover types both preferred and utilized by elephants. A hierarchical classification approach will be taken, covering multiple scales, to assess the relevance and rank of importance of system components at the watershed and regional levels.

3.      Statistical downscaling of climate predictions for vegetation related variables: Drawing on the data generation, the impact of climate on variables relevant for elephant fodder will be evaluated, and downscaled climate forecasts will be attempted. For example, it is anticipated that NDVI will be a relevant variable to make downscaled forecasts for.


Section C: Elephant Management

The main limiting factors for elephants in Sri Lanka are the availability of fodder and surface water.  The flux of these resources assume much greater significance than usual when elephants are restricted to conservation areas and are prevented from   tracking changing resources. Human and elephant conflicts increase around such zones, in the face of intensified resource competition. The ability to assess seasonal variabilities of climate impact on habitat conditions may allow remedial action to be taken that could pre-empt or mitigate crisis situations. For instance, over most of it’s range in Sri Lanka, dry seasons brings a time of limitations for the elephant, with water and fodder availability increasing during the main rainy season (October to January).   If, through assessing short-term seasonal climate outlooks, (such as associated with El Nino) rainfall is expected to be below average in a given year, a greater effort can be expended in activities that will address shortfalls in the coming term. eg. additional surface water can be collected by modifying water holes and reservoir systems, and strategies such as habitat modification can be adopted to increase browse availability within reserves. If the rains are projected to fail entirely, plans can be made to provide water and fodder artificially.  While such management interventions would promote the conservation of elephants and other fauna in protected areas, they would also mitigate human-elephant conflict. The presence of sufficient resources for elephants within conservation areas can stem the likelihood of their preferentially moving into human dominated areas.  

 

Major Steps

1.      Data Collection: a meeting of Asian elephant experts will be assembled, (with a possible teleconference with Sri Lankan partners) where existing knowledge bases and knowledge gaps will be compiled outlining elephant ecology within the study area and quantifying the underlying assumptions driving elephant\habitat dynamics within the study.

2.      Data Generation: our understanding of elephant ecology, (range maps, habitat affinity and sensitivity) will be transferred to geospatial data layers, for consideration and use within the GIS.

 

Section D.  Integration and Synthesis: Climate, Habitat, Elephant Ecology

          Calibration between and integration of sections A, B, and C will be ongoing throughout the project. For examples,   downscaling of forecasts to streamflow and vegetation in Section B will draw on climate findings in Section A, and consideration of habitat affinity and intervention strategies in section C will draw on the physical findings in Sections A and B. Additional, output-oriented integration activities and considerations will conclude with:

The main focus will be:

1.      Integration of Sub-Sections- Formats: The component sub-sections   will be integrated and their relationships assessed using the combined IRI data-library, ESRI GIS software and database systems and ERDAS Imagine image processing programs. Output will be constructed to fit current geo-spatial data standards and will be fully augmented with appropriate metadata. Compatibility between sub-sections, data formats, coordinate systems and datums will be addressed in early meetings to assure seamless integration of the assessment of climate influences on elephant habitat.

2.      Scenario Analysis: Current and historical integrated trends and correlations between vegetation, climate and elephant habitat will be assessed

§         Viability of elephant affinity and sensitivity to extreme, short-term climatic stress will be assessed

§         A range of water resource, habitat management and wildlife conservation strategies that can be effective in different climatic regimes will be assessed with an array of what-if scenario outputs

3.  Validation of Integrated Model

§         An expert group on Asian elephant management (Sukumar, Fernando, Manthrithillake) will be assembled to evaluate our validation and scenario analyses. Possible interventions given climate outlooks for coming seasons will be considered.

§         The integrated scenario outputs will be compared with expert systems assessments to evaluate reproducibility of projected biophysical, habitat, elephant, ecological and management response to the seasonal climate cycle in SE Sri Lanka.

§         Retrospective simulations of the past few decades will be carried out to assess the model output against ancillary biophysical data and collected elephant response knowledge bases.