Optimal sampling design and the accuracy of occupancy models.

Area of Science:

• Ecology
• Wildlife Biology
• Conservation Science

Background:

• Occupancy estimation is crucial for wildlife management and conservation.
• Imperfect detection rates in surveys complicate accurate occupancy estimates.
• Existing survey designs (standard, removal, conditional) have limitations.

Purpose of the Study:

• To establish theoretical limits on occupancy estimation accuracy (mean squared error).
• To identify key design factors influencing estimation accuracy.
• To introduce and evaluate a novel hybrid survey design.

Main Methods:

• Developed general theoretical limits for occupancy estimation accuracy.
• Conducted simulation studies to validate theoretical findings.
• Introduced a hybrid survey design and applied it to a real-world dataset.

Main Results:

• Survey effort, fraction of visits to occupied sites, and visits to unknown sites determine accuracy.
• Optimal implementations of standard, removal, and conditional designs offer limited room for improvement.
• The proposed hybrid design achieves accuracy closer to theoretical limits.

Conclusions:

• Occupancy study design is critically influenced by site visit strategies.
• A new hybrid design offers enhanced accuracy in specific parameter spaces.
• Findings provide practical insights for improving wildlife occupancy estimation.