Point transect sampling along linear features

Area of Science:

• Wildlife ecology
• Population assessment
• Statistical ecology

Background:

• Distance sampling is a standard method for estimating animal abundance.
• A core assumption is randomized sampler placement to ensure independence from animal distribution.
• Surveys often use linear features (e.g., roads) for sampler placement, risking bias if animals are unevenly distributed relative to these features.

Purpose of the Study:

• To present an analytical approach for distance sampling data when point samplers are used along linear features.
• To evaluate the impact of non-random sampler placement on abundance estimates.
• To assess the necessity of accounting for non-uniform animal distributions in analyses.

Main Methods:

• Development of an analysis method for distance data from point samplers on linear features.
• Conducting a simulation study to test the method under various conditions.
• Performing a field survey of Irish hares in Northern Ireland using roads as linear features.

Main Results:

• Simulation results indicated potential for substantial bias when sampler placement is not randomized.
• Field survey data corroborated simulation findings, highlighting bias in abundance estimates.
• Failure to account for non-uniform animal distribution significantly impacts accuracy.

Conclusions:

• Non-random sampler placement in distance sampling surveys can lead to considerable bias.
• Analytical methods must address potential non-uniformity in animal distribution relative to linear features.
• Randomized designs and appropriate analytical adjustments are essential for reliable abundance estimates.