Earlier events :: 2014 :: STUDENT Poster

Unmanned aerial vehicle (UAV) as a field tool for monitoring waterfowl at Kaliveli Wetland, Tamil Nadu, India

Presented by
Mythri S
Pondicherry University
Authors
Mythri S, Victor Mukherjee, Priya Davidar

Introduction: What conservation problem or question does your study address?

Aerial surveys are effective and valuable for waterfowl studies. Visual occlusion and inaccessibility hinder ground surveys at wetlands. Manned aerial surveys are limited by costs, logistics, safety and statistical integrity. In this study, to identify and estimate populations of waterfowl, we test the efficacy of using an off-the shelf, radio controlled (2.4 GHz), fixed-wing plane (UAV) of 0.8 meter wing-span with an on-board 5 MP video camera and SD card storage.

Methods: What were the main research methods you used?

The UAV was flown over two adjacent locations – a roost (for population estimation) and a waterway (for identification) for 3 consecutive days (April 2nd-4th) with one flight-run per day (N=3) for each site. The UAV was hand launched and performed acclimatization sorties before decreasing height to 15-25 meters. Total flight time was about 20 minutes including approach sorties. For the roost, the UAV was flown to cover all angles. A linear flight path of 232 meters was followed to cover the waterway. Birds detected and identified from the UAV were large (>50 cm) or white-plumaged. Grids were created against a set of random images from each day’s reading for ease of enumeration. To compare efficacy of UAV survey, a ground based point-count survey of the same roost was performed over the following 3 consecutive days (April 5th-7th).

Results: What are your most important results?

The UAV proved to be a better field investigator for survey of birds at the roost (mean=332) when compared to point-counts of the same (mean=46) which grossly underestimated the population. The UAV was able to provide a near-total count of birds at the roost. Birds detected from the waterway were identified to 7 families based on color, size, shape and shadow. Small birds that were identified by the ground survey was either not detected or identified from the aerial imagery.

Discussion: What are your important discussion points and what is the relevance of your results to conservation (if any)?

The study shows that a UAV is an effective and minimally intrusive tool for studies on waterfowl. Waterway results demonstrate detection and identification of birds from the UAV with a minimalist camera. To augment information from UAV survey, appropriate ground truthing is required for identifying small species. For UAVs to be impeccable wildlife management tools, they need to be equipped with high-definition cameras, autopilot systems, GPS and data loggers.