227
Satellite telemetry allows animals with radio collars to be tracked from the ground
or air. Animal locations are plotted on base maps from which their home range can
be approximated. This method is particularly useful for detecting the movement of
individuals and understanding tiger ecology and behaviour (Smith et al. 1998).
Camera trapping has long been used to estimate populations of faunal species
(Otis et al. 1978; Balme et al. 2009; Carbone et al. 2001; Harihar et al. 2009) and is
currently the most accepted technique for studying tiger populations in the Indian
Subcontinent (Karanth and Sunquist 1995; Karanth et al. 2002; Karanth et al. 2006;
Royle et al. 2009; Wegge and Storaas 2009). This method works on the a priori
premise that individual tigers have unique biometrics, such as stripe patterns or tail
to body ratios, that can be captured by placing automated camera traps in optimal
field locations. The biometric information can then be analyzed within a robust
capture-recapture statistical framework to arrive at estimates of numbers. The technique has been used for determining the spatial location and movements of carnivores such as tigers (Karanth et al. 2006; O’Brian et al. 2003; Linkie et al. 2006),
leopards (Harihar et al. 2011), pumas (Kelly et al. 2008), jaguars (Davis et al. 2010)
and other large predators. New techniques that have been used in conjunction with
camera traps include the use of scent-matching dogs for Amur tigers (Kerley and
Salkina 2007), fecal genetics to prepare DNA profiles of individual tigers (Samrat
et al. 2009) and large-scale occupancy surveys for tigers and their prey (Jhala et al.
2008).
Use of Remote Sensing Data for Tiger Habitat Mapping
To address the mapping of tigers distributions, a handful of studies (e.g., Sanderson
et al. 2006; Imam et al. 2009) have used thematic or continuous layers (e.g., representing vegetation, land use and elevation data) derived from remote sensing data as
input to Species Distribution Models (SDM) (e.g., occupancy) or population models. Most have focused on using freely available pre-processed optical satellite sensor data, pixel-based image classifications and geometry-based geospatial analysis
(Karanth et al. 2009; Ranganathan et al. 2008; Jhala et al. 2008). However, their
generation and use has often been compromised by uncertainty and errors (positional and thematic) related to geometric correction accuracy, poor atmospheric and
topographic correction (particularly in the mountainous areas that tigers commonly
occupy), and inconsistent use of classification and accuracy assessment techniques.
There is, however, considerable interest in the use of remote sensing data, including
as input to SDMs and population models, and the increased quality, diversity, availability and accessibility of data and derived products is contributing to greater
uptake. A summary of information that can be obtained from remotely sensed data
and which is relevant to understanding and/or modelling tiger distributions, movement and behaviour is provided in Table 2.
Coupling Field Sampling with Earth Observation Increases Understanding of Tiger…
Satellite telemetry allows animals with radio collars to be tracked from the ground
or air. Animal locations are plotted on base maps from which their home range can
be approximated. This method is particularly useful for detecting the movement of
individuals and understanding tiger ecology and behaviour (Smith et al. 1998).
Camera trapping has long been used to estimate populations of faunal species
(Otis et al. 1978; Balme et al. 2009; Carbone et al. 2001; Harihar et al. 2009) and is
currently the most accepted technique for studying tiger populations in the Indian
Subcontinent (Karanth and Sunquist 1995; Karanth et al. 2002; Karanth et al. 2006;
Royle et al. 2009; Wegge and Storaas 2009). This method works on the a priori
premise that individual tigers have unique biometrics, such as stripe patterns or tail
to body ratios, that can be captured by placing automated camera traps in optimal
field locations. The biometric information can then be analyzed within a robust
capture-recapture statistical framework to arrive at estimates of numbers. The technique has been used for determining the spatial location and movements of carnivores such as tigers (Karanth et al. 2006; O’Brian et al. 2003; Linkie et al. 2006),
leopards (Harihar et al. 2011), pumas (Kelly et al. 2008), jaguars (Davis et al. 2010)
and other large predators. New techniques that have been used in conjunction with
camera traps include the use of scent-matching dogs for Amur tigers (Kerley and
Salkina 2007), fecal genetics to prepare DNA profiles of individual tigers (Samrat
et al. 2009) and large-scale occupancy surveys for tigers and their prey (Jhala et al.
2008).
Use of Remote Sensing Data for Tiger Habitat Mapping
To address the mapping of tigers distributions, a handful of studies (e.g., Sanderson
et al. 2006; Imam et al. 2009) have used thematic or continuous layers (e.g., representing vegetation, land use and elevation data) derived from remote sensing data as
input to Species Distribution Models (SDM) (e.g., occupancy) or population models. Most have focused on using freely available pre-processed optical satellite sensor data, pixel-based image classifications and geometry-based geospatial analysis
(Karanth et al. 2009; Ranganathan et al. 2008; Jhala et al. 2008). However, their
generation and use has often been compromised by uncertainty and errors (positional and thematic) related to geometric correction accuracy, poor atmospheric and
topographic correction (particularly in the mountainous areas that tigers commonly
occupy), and inconsistent use of classification and accuracy assessment techniques.
There is, however, considerable interest in the use of remote sensing data, including
as input to SDMs and population models, and the increased quality, diversity, availability and accessibility of data and derived products is contributing to greater
uptake. A summary of information that can be obtained from remotely sensed data
and which is relevant to understanding and/or modelling tiger distributions, movement and behaviour is provided in Table 2.
Coupling Field Sampling with Earth Observation Increases Understanding of Tiger…