Mapping Elephant Distribution with Remote Sensing and GIS

AUGUST 12, 2018BY CLAUDIA WINDECKGIS LEARNING

Ad:

Share this article:

FacebookTwitterRedditSubscribe
Understanding the drivers of elephant abundance and distribution, both in Asia and Africa, is
crucial for efficient elephant and landscape conservation since both species are running out of
living space. Their enclosure in reserves in order to avoid elephant-human conflicts and the
related changes in population led to significant issues regarding living space as well as habitat
degradation due to unsustainable proportions. A thorough understanding of their migratory
patterns is thus a prerequisite for successful management and conservation of elephant
populations as well as habitat health. As long-ranging vertebrates, elephants play an essential
role in ecosystems and, nowadays, the local economy due to tourism. However, reported
increases in populations and their long-ranging activity has a significant effect on vegetation
abundance and composition, other species biodiversity and ecosystem productivity. Spatio-
temporal information regarding migratory patterns is thus crucial for conservation
management to secure both sides of the coin: elephants and habitats. A variety of applications
using GIS, remote sensing and GPS tracking of individuals per collars have advanced in the
last years and are increasingly used in this context.

Spatial information at the landscape scale provides crucial information about the elephants’
ranging behaviour as a result of precipitation-driven dynamics of vegetation [1]. Mapping of
elephant habitats, corridors and vegetation patterns has been successfully applied in this
context using normalised difference vegetation index time series (NDVI) in combination with
digital elevation models (DEM) derived from Landsat satellites. Especially Landsat 8’s
Operational Land Imager (OLI) datasets from the USGS Earth Explorer provided valuable
information regarding wet as well as dry seasons [2]. However, a variety of aspects need to
be considered when using satellite imagery to map elephant habitat maps. Not only
vegetation composition and abundance plays a critical role but in particular their temporal
changes in seasonal terms and under current climate change issues. Furthermore, mapping of
total biomass, forest canopy sizes and thicket patches in relation to time is essential to
provide robust baseline information, as also is the identification of waterholes and potential
refugia sites [3]. Remote sensing delivers the tool for a cost-effective provision of the
baseline prerequisites when considering the spatial dimensions of elephant habitats and the
remoteness of their location since sufficient funding for conservation is often more than rare.
 FIGURE 1: THE MARSABIT STUDY AREA AND RECORDED ELEPHANT MOVEMENTS. (A) WITH
AGRICULTURAL AREAS (SHADED AND OUTLINED IN RED), MAJOR ROADS (BOLD LINES), MINOR
ROADS (THIN LINES), AND ELEPHANT LOCATIONS RECORDED IN (B) DECEMBER 2005–2008 ON A
BACKGROUND OF NDVI MEASURED BETWEEN DECEMBER 19 AND 31 DECEMBER 2006, AND (C) JULY
2006–2008 ON A BACKGROUND OF NDVI MEASURED BETWEEN 12 AND 28 JULY 2006, DRAPED OVER A
DIGITAL ELEVATION MODEL. [1]

However, the use of remote sensing alone as a single instrument to determine elephant impact
and habitat alteration can also be misleading due to the influence of ruminants and other
ungulates on vegetation structure [3]. Field assessments of certain variables are thus crucial.
This is in particular valid for the interaction in African study sites with antelopes, since these
often use openings created by elephants as foraging grounds, thereby in many cases
destroying undergrowth vegetation without affecting tree canopy. This alteration of
underground vegetation which affects the elephants’ foraging grounds cannot be detected by
satellite imagery but might have a severe impact on habitat suitability due to decreased
biomass availability. Furthermore, especially in vast areas or those not intensely covered by
ranger patrols, population density as a critical aspect has to be determined in the field,
through for instance the dung count method using line transects [4], direct observations or
trails and footprints, allowing for upscaling of population per hectare. Also, spectral scans of
preferred elephant forage vegetation types using multispectral radiometers improve the
accuracy of large-scale mapping using Landsat imagery for NDVI.
Tagging elephants with satellite collars equipped with Global Positioning Systems
(GPS) allows pinpointing of precise elephant locations more efficiently compared to the
challenging direct observations in the wild, in particular in forested habitats with dense
canopy cover and undergrowth vegetation. Advancements in technology in the last 20 years
allows in these days to collect high quality spatial and temporal data due to more long-lasting
collars which log location points at hourly intervals. Satellite tracking of individuals,
especially if carefully selected, give valuable insights about range metrics and migration
patterns as well as habitat use. Furthermore, the data collected from spatial datasets also
provide information about elephant behaviour and identify patterns, that might potentially
remain undetected by field observations and remote sensing alone [5]. An innovation in the
utilisation of GPS tags can be found in the context of illegal poaching of elephants by
deterring poachers from killing and transportation of animal parts due to tracking options by
satellites. Furthermore, researchers now recognised the use of GPS tagged elephants for
further research in conservation and ecology, due to their significance as sensible living
sensors that provide information about changes in their environment.
 FIGURE 2: MAP OF THE ELEPHANT INDIVIDUAL “WILLOW’S” ANNUAL RANGES IN THE AMBOSELI
NATIONAL PARK KENYA, DURING THE TWO-YEAR TRACKING PERIOD. IT SHOWS THAT SHE MADE A
PRONOUNCED SHIFT TO A NEW DISPERSAL AREA IN THE SECOND YEAR. [5]

Summarizing all these collected data into GIS applications allows for the final modelling of
habitat suitability maps, their analysis as well as visualisation of migratory patterns including
corridors – a so-called ‘bottleneck’ allowing elephants to migrate between more substantial
habitat expenses [6,7]. Furthermore, potential conflict zones regarding anthropogenic
interaction can be determined and their visualisation used in communication with local
stakeholders. The ranges of elephants within specific research areas have in a variety of
studies been successfully determined, as also have been habitat fragmentation of study sites
due to loss of corridors and the need for conservation management to address these issues [8].
With all this information gathered, remote sensing and GIS applications allowed researchers
to understand the complex ecological requirements for elephant conservation and gave them
insights into drivers that otherwise might have remained undetected. They now have the tools
to accurately assess potential landscape-related issues affecting elephants and habitats,
allowing them to co-work more efficiently with local authorities and wildlife management.
All in order to develop conservation strategies that offer the best potential output to mitigate
impacts to elephants, their natural environment and human neighbors.

FIGURE 3: SEASONAL HABITAT USE BY ELEPHANTS (LOXODONTA AFRICANA) IN THE MOLE

NATIONAL PARK OF GHANA. [2]

References
[1] Bohrer, G., Beck, P. S., Ngene, S. M., Skidmore, A. K., & Douglas-Hamilton, I. (2014).
Elephant movement closely tracks precipitation-driven vegetation dynamics in a Kenyan
forest-savanna landscape. Movement Ecology, 2(1), 2.
[2] Ashiagbor, G., & Danquah, E. (2017). Seasonal habitat use by Elephants (Loxodonta
africana) in the Mole National Park of Ghana. Ecology and evolution, 7(11), 3784-3795.
[3] Jordaan, M. (2012). The role of remote sensing for sustainable elephant management in
South Africa. Four medium sized game reserves as case studies. ISPRS-International
Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 259-264.
[4] Baskaran, N., Kannan, G., Anbarasan, U., Thapa, A., & Sukumar, R. (2013). A
landscape-level assessment of Asian elephant habitat, its population and elephant–human
conflict in the Anamalai hill ranges of southern Western Ghats, India. Mammalian Biology-
Zeitschrift für Säugetierkunde, 78(6), 470-481.
[5] Sowers, M., Fishlock, V. & Manor, T. (2015). Mapping a future for Kenya’s Amboseli
elephants. Available online: http://www.esri.com/esri-
news/arcnews/summer15articles/mapping-a-future-for-kenyas-amboseli-elephants. Accessed:
05.08.2018
[6] Sukumar, R., Venkataraman, A., Cheeran, J. V., Mujumdar, P. P., Baskaran, N.,
Dharmarajan, G., … & Narendran, K. (2003). Study of elephants in Buxa Tiger Reserve and
adjoining areas in northern West Bengal and preparation of conservation action plan. Final
Report. Centre for Ecological Sciences, Indian Institute of Science, Bangalore.
[7] Venkataraman, A. (2005). What is an Asian elephant (Elephas maximus) corridor. Right
of Passage: Elephant Corridors of India. Wildlife Trust of India, New Delhi, 24-33.
[8] Zhang, L., Dong, L., Lin, L., Feng, L., Yan, F., Wang, L., … & Luo, A. (2015). Asian
elephants in China: estimating population size and evaluating habitat suitability. PloS
one, 10(5), e0124834.

Mapping a Future for Kenya's Amboseli Elephants

By Mark Sowers, Vicki Fishlock, and Tal Manor, Amboseli Trust for Elephants
6 159 227

An elephant family roams freely in Kenya's Amboseli National Park. (Photo courtesy of the Amboseli Trust for
Elephants.)

An organism with the sheer magnitude of an elephant—weighing up to roughly six tons—requires

proportional amounts of space and resources for survival.
The elephants of Kenya's Amboseli National Park ecosystem have always needed far more
space than the relatively small, 242-square-mile (390-square-kilometer) reserve can provide.
Amboseli elephants roam freely inside and outside the park, covering a broad swath of habitat at
the foot of Mount Kilimanjaro that spans southern Kenya and traverses what, to them, is an
invisible international border with Tanzania.
This landscape is changing rapidly, however. Increasing human populations and expanding
development are putting pressure on available wildlife habitat. With humans consuming ever-
larger shares of resources, elephant conservation is facing new challenges.

The Importance of Amboseli Elephants

Researchers with the Amboseli Trust for Elephants (ATE) have continuously monitored the
Amboseli elephant population since 1972, when Dr. Cynthia Moss and Dr. Harvey Croze founded
the Amboseli Elephant Research Project. Amboseli is unusual in that it is home to one of the few
relatively undisturbed elephant populations left in Africa. As poaching to fuel the illegal ivory trade
has made a comeback across much of the rest of the continent, the Amboseli elephants have
been largely spared.
With an estimated 35,000 African elephants being lost to poaching annually, Amboseli's elephant
population has become increasingly valuable for understanding how these animals can navigate
a changing landscape and adapt to increased human presence within their ranges.

A Rich History of Spatial Research

Documenting the spatial distribution of Amboseli elephants has been key since the project
began. Every elephant sighting in the project database is associated with a one-kilometer square
within a grid covering the study area, and sightings records since 1999 include a corresponding
GPS point. In the early days, to define the full extent of elephant ranges and their movements in
and out of the national park, researchers also fitted several elephants with radio collars; however,
that technology provided such a limited amount of data relative to the risk of traumatizing the
elephants while putting the collars on that their use was curbed.
In 2002, researchers used the GPS points, satellite imagery, and Esri software to perform GIS
analysis on the study's first 30 years of elephant location records. The results provided a clear,
long-term picture of how elephant family units and independent adult males use the national park
and how they change the location and size of their home ranges (the extent to which they move
during the course of daily activities) over time. Fascinating patterns on both annual and seasonal
scales emerged, but the vast majority of these sightings fell within the core study area inside and
immediately adjacent to the national park itself—just a small fraction of the elephants' total range.

This map of Willow's annual ranges during the two-year tracking period shows that she made a pronounced
shift to a new dispersal area in the second year.

Resurrecting Collar Tracking

To conduct research that addresses current knowledge gaps in elephant behavior, which could
aid in creating informed conservation decisions for elephant populations that are less fortunate
than Amboseli's, ATE has increasingly turned to innovative geospatial tools. Modern satellite
tracking collars can collect fine-scale temporal and spatial data, so investing in them is more
practical now. These more accurate, longer-lasting collars can allow researchers to extrapolate
major movement patterns for much of the elephant population by deploying tracking collars on
just a few carefully selected individuals. That is because, although each elephant family has a
unique strategy and pattern in how it utilizes the Amboseli ecosystem, the patchiness of
resources, particularly water, means that different subsections of the population tend to make
use of the same food and water sources, moving in similar geographic spaces.
In 2011, ATE deployed five satellite collars on adult female elephants from different families
around Amboseli. These individuals—known to researchers as Ida, Lobelia, Maureen, Vicky, and
Willow—were chosen because, while their families roam in different areas from each other inside
and outside the park, they are still illustrative of where other elephant families in their subgroups
tend to go.
Over the following two years, these elephants' collars recorded more than 78,000 GPS location
points logged at hourly intervals. This wealth of data was analyzed using ArcGIS Spatial Analyst
tools to investigate a variety of range metrics and movement parameters for the tracked
individuals.

Developing a Deeper Understanding of Elephants' Movements

The results provided valuable additions to the already detailed understanding of the population.
We now know, for example, that elephants make seasonal movements to and from the
permanent swamps in the core of Amboseli's ecosystem in response to rainfall and available
vegetation in outlying areas—a pattern that was detected in field observations and confirmed
using the satellite collars. We also now have clear answers to questions that could not be easily
explained using direct observations, such as which areas Willow and her family utilize during
their long absences from the park; how far Maureen ranges into Tanzania; and what route Vicky
uses to reach her feeding area north of the park.
The satellite collars also identified new patterns that may have otherwise gone undetected,
suggesting new directions for further research. For instance, we now want to investigate why Ida
and Lobelia, in the eastern part of the ecosystem, have smaller home ranges and spend more
time inside the park than the individuals that use other areas. We are also curious about what
caused Maureen to decrease the size of her home range and shift out of Tanzania in the second
year.
The kinds of spatial datasets obtained with satellite tracking collars can provide insights into
elephant behavior as well. One important focus of our current research is understanding how
elephants perceive and respond to the risks associated with living in close proximity to humans.
Analyses done using ArcGIS have shown that Amboseli elephants increase their travel speed
while moving through human-populated areas and that they avoid these places altogether during
daylight hours. The data has also confirmed the locations of frequently used habitat areas and
corridors where elephants move long distances, which can help with conservation planning by
identifying places that need additional monitoring resources and protection.
The results of this project, along with the outcomes of future collaring operations, will help guide
ATE and other stakeholders in securing space for all of Amboseli's wildlife. Ultimately, we want
the future of the Amboseli ecosystem to be one of peaceful coexistence between humans and
elephants for generations to come.
Visit ATE at elephanttrust.org or email info@elephanttrust.org to get more information.