Wild elephants vary in their attraction to novelty across an anthropogenic landscape gradient
Data files
Jul 11, 2025 version files 41.78 KB
-
baseline_socialmatch.csv
7.10 KB
-
baselinecomp.csv
7.66 KB
-
NOdata.csv
14.57 KB
-
objectcomp.csv
4.62 KB
-
README.md
7.83 KB
Abstract
Research on how wild animals respond to novelty is becoming more relevant as the overlap between natural habitats and human-dominated landscapes increases. Wild Asian elephants spend more time in anthropogenic landscapes as their habitat is converted to agriculture. Greater neophilia and exploration may allow elephants to successfully access agricultural resources, which may cause negative interactions with people. We compared wild elephant reactions to novel objects in two different landscapes in Thailand (near agriculture and deep inside a protected sanctuary). We also assessed consistency in measures for individuals exposed to different objects to determine whether their reactions could be considered personality traits. Elephants tested near agriculture were more neophilic and exploratory than those inside the sanctuary. However, the limited sample of elephants exposed to both novel objects did not demonstrate consistency in their reactions, and thus, we could not determine whether neophilia or exploration were personality traits in this population. Neophilic and exploratory elephants likely benefit from high-quality agricultural resources, but at a potential cost to both elephants and humans. Knowledge about the elephants’ behavior and attraction to particular landscapes could aid in human-elephant conflict mitigation efforts that consider the needs of both species and aim for more stable coexistence.
https://doi.org/10.5061/dryad.gtht76hxc
Description of the data and file structure
These data describe wild elephant reactions to novel objects in two different landscapes in Thailand (near agriculture and inside a protected sanctuary). Videos of elephant movement within a 10 m perimeter of the objects and interactions with the objects were recorded by camera traps. Elephant behaviors were coded from the videos to assess measures of neophilia and exploration.
Files and variables
File: baselinecomp.csv
Description: This dataset was used to compare elephant behavior between baseline (only cameras and perimeter markers present in environment) and novel object conditions (either cattle brush or firehose present) for the 43 elephants who were observed in both baseline and novel object conditions at the same site.
Variables
- ObsID: Unique Observation identity
- Subject: Elephant identity
- Condition: baseline (only cameras and perimeter markers present in environment) or novel object conditions (either cattle brush or firehose present)
- durinrad: duration elephant spent within the 10m perimeter of the tree or object
- durinprox: duration elephant spent within one body's length of the tree or object
- durinteract: duration elephant spent in contact with the object, not applicable (NA) for baseline conditions when the object was not present to interact with
- ov: duration of observation when elephant is out of view due to cameras not recording
- latapproach2: latency for elephant to approach within a trunk's length of the object or tree after stepping within the 10 m perimeter, adjusted so that if they never approached they were given a maximum value of 150 seconds
- lattouch2: latency for elephant touch object after stepping within the 10m perimeter, not applicable (NA) for baseline conditions when the object was not present, adjusted so that if elephant never touched the object, they were given a maximum value of 150 seconds
- Location: Whether the elephant was observed in the protected area (PA) or near human-dominated landscapes (HDL)
File: baseline_socialmatch.csv
Description: This dataset was used to compare elephant behavior between baseline (only cameras and perimeter markers present in environment) and novel object conditions (either cattle brush or firehose present) when 4 elephants were removed who had different social conditions between their baseline and novel object exposure. This removal did not change the results, so the complete baseline comparison dataset was used for final analysis presented in the manuscript.
Variables
- ObsID: Unique Observation identity
- Subject: Elephant identity
- Condition: baseline (only cameras and perimeter markers present in environment) or novel object conditions (either cattle brush or firehose present)
- durinrad: duration elephant spent within the 10m perimeter of the tree or object
- durinprox: duration elephant spent within one body's length of the tree or object
- durinteract: duration elephant spent in contact with the object, not applicable (NA) for baseline conditions when the object was not present to interact with
- ov: duration of observation when elephant is out of view due to cameras not recording
- latapproach2: latency for elephant to approach within a trunk's length of the object or tree after stepping within the 10 m perimeter, adjusted so that if they never approached they were given a maximum value of 150 seconds
- lattouch2: latency for elephant touch object after stepping within the 10m perimeter, not applicable (NA) for baseline conditions when the object was not present, adjusted so that if elephant never touched the object, they were given a maximum value of 150 seconds
- Location: Whether the elephant was observed in the protected area (PA) or near human-dominated landscapes (HDL)
- Social?: Whether more than one elephant was within the 10m perimeter with the subject, yes (Y) or no (N).
File: NOdata.csv
Description: The dataset of all 125 elephant exposures to novel objects used to determine whether behavior was associated with elephant location.
Variables
- Observation.id: Unique observation identity
- Date: Date of observation
- Time: Time of observation
- Tree: the code for the specific tree where the object or puzzle box was installed
- Subject: Elephant identity
- Sex: male (M) or female (F) or NA (unable to determine the sex)
- AgeClass: Adult (A), Subadult (B), or Calf (C)
- Condition: Novel object type, either firehose or cattle brush
- durinrad: duration elephant spent within the 10m perimeter of the tree or object
- durinprox: duration elephant spent within one body's length of the tree or object
- durinteract: duration elephant spent in contact with the object, not applicable for baseline conditions
- ov: duration of observation when elephant is out of view due to cameras not recording
- latapproach2: latency for elephant to approach within a trunk's length of the object after stepping within the 10 m perimeter, adjusted so that if they never approached they were given a maximum value of 150 seconds
- lattouch2: latency for elephant touch object after stepping within the 10m perimeter, adjusted so that if elephant never touched the object, they were given a maximum value of 150 seconds
- Location: Location where elephant was exposed to the object, either protected area (PA) or near human-dominated landscape (HDL)
- PrevEleInt: Whether there was an elephant interaction with the object prior to the subject's exposure, yes (Y) or no (N).
- Social: Whether more than one elephant was within the 10m perimeter with the subject, yes (Y) or no (N).
File: objectcomp.csv
Description: This dataset was used to assess whether the 19 elephants who were exposed to both types of novel object, reacted consistently towards the objects.
Variables
- Observation.id: Unique observation identity
- Date: Date of observation
- Time: Time of observation
- Exposure: Whether the elephant was exposed to this novel object type first (1) or second (2)
- Tree: the code for the specific tree where the object or puzzle box was installed
- Subject: Elephant identity
- Sex: male (M) or female (F) or NA (unable to determine the sex)
- AgeClass: Adult (A), Subadult (B), or Calf (C)
- Condition: Novel object type, either firehose or cattle brush
- durinrad: duration elephant spent within the 10m perimeter of the tree or object
- durinprox: duration elephant spent within one body's length of the tree or object
- durinteract: duration elephant spent in contact with the object, not applicable for baseline conditions
- ov: duration of observation when elephant is out of view due to cameras not recording
- latapproach2: latency for elephant to approach within a trunk's length of the object after stepping within the 10 m perimeter, adjusted so that if they never approached they were given a maximum value of 150 seconds
- lattouch2: latency for elephant touch object after stepping within the 10m perimeter, adjusted so that if elephant never touched the object, they were given a maximum value of 150 seconds
- Location: Location where elephant was exposed to the object, either protected area (PA) or near human-dominated landscape (HDL)
- PrevEleInt: Whether there was an elephant interaction with the object prior to the subject's exposure, yes (Y) or no (N).
- Social: Whether more than one elephant was within the 10m perimeter with the subject, yes (Y) or no (N).
Code/software
The R markdown file used to analyze the data in RStudio, R v 4.3.2 (R core team, 2022). All packages used are listed in the markdown file.
Behavioral data were coded using Behavioral Observation Research Interactive Software from camera trap videos, and measures were extracted from the codes.