Journal of Threatened
Taxa | www.threatenedtaxa.org | 26 March 2025 | 17(3): 26704–26714
ISSN 0974-7907 (Online)
| ISSN 0974-7893 (Print)
https://doi.org/10.11609/jott.9251.17.3.26704-26714
#9251 | Received 30 June
2024 | Final received 03 March 2025 | Finally accepted 14 March 2025
Implementation
strategy and performance analysis of a novel ground vibration-based elephant
deterrent system
Sanjoy Deb 1,
Ramkumar Ravindran 2
& Saravana
Kumar Radhakrishnan 3
1,2 Department of ECE, Bannari Amman Institute of Technology, Sathyamangalam,
Tamil Nadu 638401, India.
3 School of Electronics
Engineering, Vellore Institute of Technology, Chennai, Tamil Nadu 600127,
India.
1 sanjoydeb@bitsathy.ac.in,
2 ramkumarr@bitsathy.ac.in, 3 r.saravanakumar@vit.ac.in
(corresponding author)
Editor: Heidi Riddle, Riddle’s Elephant and Wildlife
Sanctuary, Arkansas, USA. Date of publication: 26 March 2025
(online & print)
Citation: Deb, S., R. Ravindran & S.K. Radhakrishnan (2025). Implementation strategy and performance
analysis of a novel ground vibration-based elephant deterrent system. Journal of Threatened Taxa 17(3): 26704–26714. https://doi.org/10.11609/jott.9251.17.3.26704-26714
Copyright: © Deb et al. 2025. Creative Commons Attribution 4.0 International License. JoTT allows unrestricted use, reproduction, and
distribution of this article in any medium by providing adequate credit to the
author(s) and the source of publication.
Funding: We received funding from ‘DST SERB Core Research Grant’ CRG/2023/005596 on 16 January 2024.
Competing interests: The authors declare no competing interests.
Author details: Dr. Sanjoy Deb, from Kolkata, holds a BSc in Physics, an MSc in Electronics, and an MTech in nanoscience from Jadavpur University. He earned his PhD in 2012 and is now a professor at Bannari Amman Institute of Technology, Sathyamangalam, Tamilnadu,
India. He has published more than 40 papers, and his current research focuses on mitigating human-animal conflict. Dr. R. Ramkumar, from Tamil Nadu, earned his B.E. and M.E. from Anna University and a PhD in Electrical Engineering in 2024. With 14 years of teaching and research experience, he specializes in embedded systems and wireless networks. He has published 14 journal articles and currently serves as an assistant professor at Bannari Amman Institute of Technology, Sathyamangalam, Tamilnadu,
India. Dr. R. Saravana Kumar, from Tamil Nadu, earned his B.E., M.E., and Ph.D. from Madras and Anna University. With 17 years of experience, he has published 43 research articles, authored three books, and six book chapters. He is an associate professor at Vellore Institute of Technology, Chennai, Tamil Nadu, India, specializing in VLSI and embedded system design.
Author contributions: SD—played a pivotal role in the development of hardware design, algorithm formulation, programming, and verification of field systems. RR-—was responsible for conducting the field survey, overseeing the installation, and managing data collection. SKR-—contributed significantly to the field installation process and the drafting of the manuscript.
Acknowledgements: The authors sincerely appreciate the support provided by the management team and ground staff of the Nilgiri Biosphere Nature Park in facilitating this research throughout the study period. The authors also gratefully acknowledge the invaluable assistance of the forest officials and ground staff of Sathyamangalam Tiger Reserve and Coimbatore Forest Division, in the successful execution of this research. We acknowledge the financial support for technology development and system maintenance from DST SERB CRG (Ref No.
CRG/2023/005596-G).
Abstract: The establishment of
human habitations, expansion of cultivation lands, and constant degradation of
forest areas have intensified human-elephant negative interactions over the
years in the Anaikatti area located at Coimbatore and
Periyanaickenpalayam forest range in southern India.
A few nature parks have been established in this interaction-prone area and are
also affected by frequent elephant presence. To safeguard one such park, Nilgiri Biosphere Nature Park, from elephant and other
wildlife intrusions, 13 units of a ground vibration-based ‘elephant deterrent
system’ have been installed along its periphery. The system is a
field-deployable version of our ground vibration-based ‘elephant early warning
system’, designed to deter elephants using sound units upon detection. It
analyzes the frequency of footstep vibrations to initially differentiate
between elephant and non-elephant footsteps. The cumulative vibration data from
sensors is then used to identify elephants more precisely. Furthermore, for
certain system units, the system’s algorithm has been adjusted via on-the-fly
software updates to detect all animal footstep vibrations, activating deterrent
sound effects tailored to the specific requirements of the current application.
Insights from location surveys and discussions with local residents have
contributed to the development of innovative implementation strategies and the
careful selection of installation sites, which are detailed in this paper. The
paper also outlines the system’s installation layout, case-specific algorithms,
and hardware architecture. Performance was monitored over an eight-month
period, with the results analyzed alongside feedback from field observations.
Notably, the system trial phase showed a reduction in elephant intrusions
within the park. This report is the first detailed account of a trial field
performance, making it a valuable reference for replicating similar solutions
in other conflict locations.
Keywords: Human-elephant
negative interaction, microcontroller, sensor string integration, signal
conditioning unit, vibration sensor, warning system.
Introduction
Over the years,
several technologies and systems have emerged for human-elephant negative
interaction management, but they come with their advantages and limitations
(Shaffer et al. 2019; Vogel et al. 2020; Tiller et al. 2021). If categorized
broadly, the technologies come in two categories: first, the elephant early
warning system, and second, the elephant deterrent system (Choudhury 2010;
Rohini et al. 2016; Tripathy et al. 2021). Although
there has been notable progress in the domain of early warning technologies,
very few successful non-contact elephant deterrent systems have been reported
so far. (Nayak & Swain 2020; Feuerbacher et al.
2021). The high intelligence and adaptive learning capability of the elephant
have restricted technologists from designing a long-lasting elephant deterrent
system (Locke et al. 2016; MoEF 2020). The few
reported short-term successful systems also had a lack of range, element of
surprise, and have terrain-specific limitations (MoEF
2020).
Considering those
technological ambiguities and urgent needs, our ground vibration-based
‘elephant early warning system’ (EEWS) was reconfigured into an elephant
deterrent system, with a re-engineered system design, operational algorithm, control
circuit, and addition of a high-volume hooter/siren. The EEWS was designed over
the years with national and international funding (Ramkumar & Deb 2021).
The EEWS was tested through simulated experiments, as well as with field
implementation at Sathyamangalam Tiger Reserve in
2020 (Ramkumar & Deb 2021). With the feedback data from field-installed
EEWS units, the technology was refined. With all those added attributes, the
EEWS was re-configured into a ground vibration-based ‘elephant deterrent
system’ (EDS). Under this work, a total of 13 units were installed to cover the
3.5 km periphery of the Nilgiri Biosphere Nature Park
(NBNP).
Location Survey
Anaikatti is a small township near Coimbatore, located in
the Western Ghats at the Tamil Nadu-Kerala border in southern India. Human
activities such as agriculture, urbanization, and tourism are disrupting the
traditional migratory routes of elephants. Additionally, the depletion of
forest resources has forced elephants to explore new migration paths, making Anaikatti a key interaction hotspot (Karthick et al. 2016;
A Times of India Report 2019; Deivanayaki et al.
2019). The intensity of the conflict is so severe that the area frequently
makes news headlines and has been the subject of several research articles
(Ramkumar et al. 2013; Natarajan et al. 2024). Being situated in this area, the
Nilgiri Biosphere Nature Park (NBNP) has elephants
visiting the site over the years. The NBNP is a nature-based organization
designed to introduce and educate young minds about the unique flora and fauna
of the Western Ghats, boasting a large collection of these species. The
availability of food and water, especially during summer, has made the park an
attractive entry point for the elephants.
To assess the
elephant visitation scenario at NBNP, a detailed field survey was carried out
on foot to accurately map the elephant movement paths. Additionally, key
terrain factors such as soil conditions, ground slope, sunlight availability,
and other parameters relevant to system installation were also surveyed. On the
northern side of the NBNP, a hillside is covered in forest. To the east of the
park, there is open land extending for about 1.5 km. This area features small
patches of forest, scattered agricultural fields, and a few houses, as
illustrated in Figure 1. Meanwhile, the southern and western sides of the park
are covered by cultivation land and human habitation. There is a narrow
footpath covering the three sides of the park, except for the western side,
which is covered by a motorable road. According to local reports, the narrow
path is utilized by cattle grazers, wood collectors, and farmers during the
day, while at night, it becomes a route for deer, pigs, leopards, and other
wildlife, including elephants. We interviewed a group of 50 individuals in and
around the park, local forest officials, including park workers, to understand
the status of interactions, map the movement paths of elephants & other
wildlife, and analyze the intentions behind these intrusions, their frequency,
distribution across seasons, and times of day. The survey was conducted during
the first two weeks of August 2022, and the results are presented in Table 1.
The information from
the general survey, presented in Table 1, indicates that over the past three
years, a sub-adult male resident elephant and a mature migrating bull have
frequented the site. The survey also reveals that the bull enters the area from
November to April each year. During the day, elephants settle on the eastern
side of the hill forest and visit the park and nearby villages after sunset.
Despite the entire park perimeter being secured by an electric fence, it has
proven insufficient to prevent elephant intrusions over the years.
Implementation
Strategy
All potential entry
and exit paths of the elephants have been marked on the map by analyzing ground
conditions, gathering residents’ feedback, and reviewing the survey report, as
illustrated in Figure 1. It has been identified that most elephant paths from
the northern and eastern sides of the park terminate at the boundary, which is
secured by an electric fence. According to feedback from local residents and
park workers during the field survey, once the elephants reach the fence, they
walk along it in search of a weak point to breach the fence. Alternatively,
they may continue their journey to reach the river and agricultural areas on
the southern or western side. The survey also revealed that a narrow monsoon
river runs through the southern section of the park, and during the dry months,
this path is frequently used by elephants to access those destinations.
A comprehensive
analysis of terrain conditions, vegetation, local infrastructure, animal
species, the nature and direction of the visit, and other localized factors is
crucial for designing and implementing an effective system to minimize visits.
For instance, while we specialize in laser fence-based animal early warning
systems, the steep slopes, dense vegetation, and the elephant movement paths
along the park’s electrical fence make such a solution impractical (Ramkumar
& Deb 2022). Based on our survey and feedback from other project
stakeholders, we have concluded that to effectively manage the human-elephant
interactions in this area, it is essential to prevent elephant movement along
the paths surrounding the park’s perimeter. Therefore, we decided to install
footstep vibration-based EDS units at the junctions where elephant paths
intersect with the park’s boundary. This solution is anticipated to be highly
effective, as illustrated in Figure 2.
System Details
The EDS is a modified
variant of EEWS with few added features, as described in the following sections
with Figure 3.
System Hardware
Architecture
The EDS is a
two-sensor strings-based design, where one sensor string takes reference input
from the other string to reject any common vibration. With two separate sensor
strings, only one string captures footstep vibrations during a visit, while
vibrations from rain, landslides, and vehicle movement are detected
simultaneously by both strings. This allows the system to effectively
distinguish and eliminate noise vibrations, responding only to footstep
vibrations. The sensor string is designed with piezoelectric sensors in
successive series and parallel combinations to optimize sensor string output in
terms of both current and voltage. Two sensor strings are connected with the
‘signal conditioning unit’ (SCU), as shown in Figure 3. The signal conditioning
unit is the combination of two identical ‘pre-amplifier and filter sections’
connected with each sensor data line separately. The signal conditioning
circuit of EDS is a design with few instantly configurable pot resistors, and
thus its vibration sensitivity can be adjusted in real-time as per the terrain
conditions and the target vibration. In a nutshell, the EDS can be configured
into a highly sensitive mode to capture footstep vibration even from a house
cat or extremely less sensitive, where it will sense the footstep vibration of
large animals only. The authors have already analyzed the signal parameters for
different animals footsteps and reported in (Ramkumar
& Sanjoy 2021).
The control unit
functions based on a microcontroller circuit. In this work, we utilized an
Arduino-based microcontroller unit for decision-making, which is an open-source
hardware and software component. The vibration patterns of various animals and
humans are stored in the microcontroller. When the control unit receives
processed signals from the SCU, it runs an identification algorithm and
compares the input with pre-saved reference signal patterns. Upon detecting a
match, the control unit activates the hooter to repel intruding animals. The
basic identification algorithm has already been analyzed and documented by (Ramkumar
& Sanjoy 2021), and the
modified version used in the preset application is presented in
detail in the following sections. The EDS operates on a 12-volt power supply
and includes a stand-alone unit featuring solar panels (12V, 20W), charge
controllers (12V, 6A), and batteries (12V, 2.5Ah). A daylight sensor is
integrated into the system, allowing it to activate at dusk and automatically
turn off at twilight.
System Implementation
Design
In the current EDS
design, each sensor string consists of four sensors, with each sensor spaced 1
m apart. The sensor string is buried at a depth of 20 cm and follows a zigzag
pattern, providing a cumulative physical coverage area of 3 m² (calculated as 2
× 1.5 m²), as shown in Figure 4. However, once buried, each sensor has a
vibration detection radius of approximately 2 m, making the effective sensing
coverage area 2–3 times larger than the physical coverage area. When the sensor
string is placed underground, it creates a detection field similar to an
underground sensor carpet. The sensor string can be placed at a long distance
from the hooter pole, providing a long detection range. The system is versatile
and can be placed in various terrain conditions, except for waterlogged areas.
Placing the sensor
string too deep can reduce its sensitivity but also help minimize background
noise vibrations, so the depth must be optimized based on the terrain
conditions and target species. The separation between two sensor strings
(denoted as ‘x’ in Figure 4) must also be adjusted according to specific unit
requirements. For this project, the maximum separation ‘x’ is 20 m for EDS
Unit—10, while for EDS Unit—2, the separation between the two strings is 5 m.
In the current
application, five types of 12-volt hooters are used across different system
units in a random pattern, each producing a distinct sound to ensure sound
diversity. The positioning of the hooter poles, the number of hooters, and
their orientation are tailored to the specific requirements of each case.
System Algorithm
The system monitors
three key parameters: ‘signal frequency’, ‘signal amplitude’, and the
cumulative ‘volume of vibration’. The EDS operates on a 10-second ‘detection
loop’, controlled by a microcontroller (which aligns with the verified time an
elephant typically takes to cross the sensor string). The flowchart shown in
Figure 5 outlines the basic process for detecting and identifying elephants and
other animals in the EDS. Previous simulated studies have indicated that
elephant footsteps generate low-frequency vibrations, in contrast to animals
with hooves, which produce higher-frequency vibrations above 100 Hz (Ramkumar
& Deb 2021). This distinction is especially noticeable on rocky ground.
After the signal is pre-amplified and filtered, the system algorithm checks the
frequency input. If the frequency is identified as less than 100 Hz, it
proceeds along the “elephant line”.
Following frequency
determination, the signal values are accumulated over a 10-second period,
referred to as the “detection loop”, and the resulting value is recorded as the
‘cumulative vibration’ (Vc). During each loop, the
system checks for vibration peaks above a pre-set threshold. All amplitude
values exceeding this threshold are accumulated within the loop to calculate
the Vc. If the Vc is
greater than or equal to the ‘voltage elephant threshold’ (Vte),
the sound deterrent unit is activated to repel the elephant. This Vte has been determined from a previous simulated
experiment with an elephant but also needs slight adjustment to counter the
background noise of the implementing site. While humans and other soft-toed
animals also generate low-frequency vibrations, previous observations show that
their cumulative vibration values are significantly lower than the Vte, allowing them to be excluded when targeting elephants
specifically.
This unique approach
has been shown to achieve over 80% accuracy in detecting elephants through
footstep vibrations, as confirmed by previous simulated experiments (Ramkumar
& Deb 2021). The remaining 20% discrepancy in accuracy arises from system
limitations in detecting elephants under certain conditions, such as muddy soil
or loose sand, where sensor sensitivity is significantly reduced. In these
situations, the system may incorrectly identify elephants and other animals.
Additionally, high-volume vibrations from overlapping frequencies generated by
a group of other animals crossing the sensor field could cause the system to
misinterpret the detection as an elephant, leading to potential confusion.
To further
distinguish elephant detections from those of other animals, the EDS employs
distinct sound patterns. For example, when an elephant is detected, the hooter
will sound continuously for five minutes to maximize the deterrent effect. This
distinct sound pattern serves as an alert to park security personnel, prompting
them to verify the potential elephant intrusion. In contrast, detections of
other animals will trigger a one-minute sound with a five-second on-off
pattern, ensuring different responses based on the type of detection.
Considering our practical experience, the system algorithm is designed to
trigger a maximum of 20 times per day, ensuring that contentious sound
generation is avoided throughout the night, even in the event of a system
malfunction.
System Performance
Analysis
In October and
November 2022, 13 EDS units were installed around the park perimeter. In
addition to elephants, wild pigs, and spotted deer frequently visit the park,
predating upon or uprooting plants including flower & vegetable gardens.
Visitations are not limited to animals, as wood poachers have occasionally
entered the park and poached valuable trees. To address these safety concerns,
all units except Unit—1, Unit—5, and Unit—13 were configured in “all-animal
detection” mode to reduce animal and human activity along the park’s perimeter
pathways at night.
As outlined in the
algorithm flowchart, the EDS operates in two modes: ‘elephant line’, which
detects and responds exclusively to elephant footsteps, and ‘other animal
line’, which detects and responds to vibrations caused by various animals,
including elephants. This enables the EDS to function either as an ‘animal
deterrent system’ or an ‘elephant deterrent system’. For trial purposes, Units
1, 5, and 13 were configured in elephant deterrent mode to evaluate their
effectiveness, while the remaining units were set to animal deterrent mode to
meet practical needs.
To create distinct
sound effects, five types of horns and hooters were used with varying on-off
patterns, ensuring unique sound signatures for each unit. Park security
personnel monitored the system for eight months, recording unit-specific
detections based on these unique sound patterns. During this period, the system
was most frequently triggered by pigs, spotted deer, leopards, and humans, with
elephants triggering the system only rarely.
The unit-wise EDS
detection report for the eight-month period of December 2022–July 2023 is shown
in Figure 6. According to our field observation report, based on input from
local stakeholders, most detections were caused by wild animals and human
activities, with only two instances involving elephants. Animal activity was
found to vary seasonally; during peak summer, the scarcity of natural water and
food sources attracted more animals to the park, where pump water holes are
available at several locations. Consequently, most EDS units reported higher
animal intrusions during late winter and peak summer.
A discrepancy was
noted between the number of animal detections (total count of sound alarm) by
the system and the actual number of animal intrusions into the park. This
mismatch occurs because many animals bypass the park, using paths that lead to
nearby villages instead. Notably, detections by Units 11 and 12 remained
consistent throughout all months, later identified as being primarily due to
human footsteps. To understand this pattern, time-wise detection data for all
units was analyzed and is presented in Figure 7.
The survey revealed
that most human outdoor activities around the park completely cease after 2000
h and resume after 0500 h. Except for three units, all other EDS units are
configured to detect all animal modes. Thus, it can be inferred that detections
occurring before 2000 h and after 0500 h are predominantly due to human
activities. Most of the detections from units 11 and 12, located along human
movement paths, occurred during these times, confirming them as human activity.
Field investigations further revealed that several houses on the eastern side
of the NBNP (marked with red circles in Figure 1) have residents who frequently
use pathways near these units during those hours.
In contrast, other
units primarily captured animal movements, which peaked before 2100 h,
gradually decreased by 2300 h, increased again around 0300 h, and settled after
0500 h. This pattern may be due to animals moving towards nearby cultivation
areas, villages, and rivers in search of food and water, especially as human
activity is high in the evening and early morning. This aligns with the
well-known pattern of animals raiding crops during late evening and early
morning hours. Some units, like Unit 7, which are far from regular human
pathways, recorded consistent animal activity during early evening, late night,
and intermittently throughout the night.
The specially
configured units (1, 5, and 13) did not detect any elephants during their
runtime but did register a few false elephant alarms. The exact cause of these
false alarms remains unclear, although no major technical malfunctions were
identified. Elephant footsteps were detected on two occasions—at Unit 3 in
January and Unit 8 in March. However, since these units were not configured in
elephant detection mode, they produced sounds associated with other animals.
While no systematic
statistical data exists on the exact number of elephant intrusions in the park
over the years, discussions with staff and other relevant individuals indicate
approximately nine visitations occurred in the three years prior to system
installation. In contrast, following the system’s installation, only one
intrusion was recorded. This incident occurred during the peak north-east
monsoon when many units were struggling with low battery issues, and the system
failed to trigger an alarm.
According to park
staff, elephants typically follow their habitual paths at night, testing the
fence for weak points to enter. It is believed that the loud sounds triggered
by their footsteps, or the frequent sounds triggered by other animal movements,
have discouraged them from using their regular paths along the park’s
periphery. Although the EDS has demonstrated a significant impact, its
long-term effectiveness requires further validation, additional installations
at other high-risk locations, and a detailed investigation into the underlying
factors contributing to its success.
EDS Pictorial
Representation
Figures 8–13 show
some system-relevant pictures to help us better understand the EDS actual field
architecture, infield performance, and other notable issues.
Conclusion
The ground
vibration-based elephant deterrent system presented in this paper represents a
pioneering approach and serves as the first trial report from India. This
system is an advanced, field-implementable adaptation of the previously
field-validated elephant early warning system technology. The paper provides a
comprehensive description of the EDS hardware, field implementation strategy,
and its innovative operational algorithm. This study documents the deployment
of 13 EDS units in
NBNP nature park and evaluates their performance over eight
months. Additionally, it includes a field survey and subsequent analysis of
conflict scenarios at the project site, accompanied by an accurate map of
elephant movement paths. Such surveys and precise mapping are crucial for
designing a strategic insulation plan, and the details shared in this paper
offer valuable insights for similar projects. While the EDS is intended
primarily to detect elephant footstep vibrations with precision, it has been
optimized using a modified algorithm to enhance sensitivity, enabling the
detection, and deterrence of other animals. This capability has been
implemented and thoroughly reported in the current project. The system’s performance
analysis, which considers detection data across different times and seasonal
variations, demonstrating that the EDS units effectively mitigate animal
activities in the operational areas. By addressing the fundamental limitations
of earlier animal-deterrent systems, the innovative EDS design has proven
successful. The insights detailed in this paper provide a foundation for
replicating this solution at other human-wildlife conflict hotspots,
contributing significantly to the field once published.
Table 1. The conflict
status survey. It involved a selected group of 50 individuals from NBNP Park
and surrounding areas, including local villagers, park security, staff, and
local forest officials. The sample was composed of 70% adult males, 20%
children aged 7–13, and 10% females.
|
Questions |
People Response |
|
How many times has
he/she seen an elephant in the past two years? |
40% have not seen
an elephant, 20% have seen one 1–2 times, 10% have seen it more than
twice, and 30% have not seen it but felt its close presence. |
|
What size was the
elephant observed (adult, semi-adult, juvenile)? |
70% reported seeing
adults, 20% observed semi-adults, and 10% were unable to distinguish due to
darkness. |
|
During which season
did he/she see the elephant? |
80% of sightings
occurred in summer, 5% in the monsoon, 10% in winter, and 5% could not recall
the season. |
|
At what time of day
did he/she observe the elephant? |
60% saw elephants
during late evening, 30% in early morning, and 10% at midnight. |
|
What was the likely
path or track of the elephant's movement? |
Did it bypass the park area and move toward the
riverside? 40% of the time. Did it go to the crop fields on the southern and
western sides of the park? 25% of the time. Did it intrude into the park area? 10% of the time. The remaining 25%
were unsure. |
|
What might be the
cause of the elephant's intrusion? |
Did it go to the river for water? 30% of respondents
answered yes. Did it raid crops in the agricultural land? 40% of respondents
answered yes. Did it go to the park area for food and water? 10% of respondents
answered yes. The remaining 20%
were unsure. |
FOR
FIGURES & IMAGES - - CLICK HERE FOR FULL PDF
References
A Times of India Report (2019). Booming real estate
increases conflicts. https://timesofindia.indiatimes.com/city/coimbatore/booming-real-estate-increases-conflicts/articleshow/67648848.cms.
Accessed 23 January 2019.
Choudhury, A. (2010). Human-elephant
conflicts in northeast India. Human Dimensions of Wildlife 9(4): 261–270.
https://doi.org/10.1080/10871200490505693
Deivanayaki, M., N. Ezhilarasi & B. Ramakrishnan (2019). Fatal Human and
Elephant Conflicts, 2000–2017: Anamalai Tiger
Reserve, Southern Western Ghats. The Indian Forester 145(10): 927–934. https://www.indianforester.co.in/index.php/indianforester/article/view/149189
Feuerbacher, A., C. Lippert, J. Kuenzang & K. Subedi (2021). Low-cost electric
fencing for peaceful coexistence: An analysis of human-wildlife conflict
mitigation strategies in smallholder agriculture. Biological Conservation
255: 108919. https://doi.org/10.1016/j.biocon.2020.108919
Karthick, S., B. Ramakrishnan & M. Illakia (2016). Human-elephant
conflict issues with special reference to crop damage and people’s perception
in and around Coimbatore Forest Division, southern India. The Indian
Forester 142(10): 1010–1018.
Locke, P. & J. Buckingham (eds.) (2016). Conflict,
Negotiation, and Coexistence: Rethinking Human–Elephant Relations in South Asia.
Online edn., Oxford Academic https://doi.org/10.1093/acprof:oso/9780199467228.001.0001
MoEF (2020). Best Practices of
Human Elephant Conflict Management in India. Ministry of Environment &
Forests, Government of India https://moef.gov.in/wp-content/uploads/2020/08/Best-Practice-Man-Animal-Conflict.pdf
Natarajan, M. (2024). Rapid assessment of
human-elephant conflict: a crime science approach. Crime Science 13:
24. https://doi.org/10.1186/s40163-024-00223-9
Nayak, N. & P.K. Swain (2020). From fear to
festivity: Multi-stakeholder perspectives on human-elephant conflict and
coexistence in India. Journal of Public Affairs e-2496. https://doi.org/10.1002/pa.2496
Ramkumar, R. & S. Deb (2021). Real-Time System
Design for Sensing, Recording and Analyzing Elephant Seismic Waves through
Ground Vibration Algorithm. Journal of Circuits, Systems and Computers
31(03): 1–24. https://doi.org/10.1142/S0218126622500487
Ramkumar, R. & S. Deb (2022). Design,
Implementation of a Generic Roadkill Prevention System (RPS) using Laser Beams
to Reduce Human-Animal Conflict in Forest Boundaries. Lasers In Engineering 53(5–6): 285–298.
Rohini, C.K., T. Aravindan, K.S.A. Das & P.A. Vinayan (2016). Human-elephant conflict
around North and South Forest Divisions of Nilambur, Kerala, India. Gajah
45: 20–27.
Shaffer, J.L., K.K. Khadka, J.V.D. Hoek & K.J. Naithani (2019). Human-elephant
conflict: a review of current management strategies and future directions. Frontiers
in Ecology and Evolution 6: 235. https://doi.org/10.3389/fevo.2018.00235
Tiller, L.N., T. Humle, R.
Amin, N.J. Deere, B.O. Lago, N. Leader-Williams, F.K. Sinoni,
N. Sitati, M. Walpole & R.J. Smit (2021). Changing seasonal,
temporal and spatial crop-raiding trends over 15 years in a human-elephant
conflict hotspot. Biological Conservation 254: 108941. https://doi.org/10.1016/j.biocon.2020.108941
Tripathy, B. R., X. Liu, M. Songer, L. Kumar, S. Kaliraj, N.D.
Chatterjee, W.M. Wickramasinghe
& K.K. Mahanta (2021). Descriptive spatial analysis of human-elephant
conflict (HEC) distribution and mapping HEC hotspots in Keonjhar
Forest Division, India. Frontiers in Ecology and Evolution 9: 640624. https://doi.org/10.3389/fevo.2021.640624
Vogel, S.M., S.A.
Blumenthal, W.F. de Boer, M. Masake, I. Newton, A.C. Songhurst, G. McCulloch, A. Stronza,
M.D. Henley & T. Coulson (2020). Timing of dietary
switching by savannah elephants in relation to crop consumption. Biological
Conservation 249: 108703. https://doi.org/10.1016/j.biocon.2020.108703