Evaluating Bird Population Estimation Methods
Author Information
Author(s): Sandra L. Taylor, Katherine S. Pollard
Primary Institution: Department of Statistics, University of California Davis
Hypothesis
How do the double observer and double sampling methods compare in estimating bird populations?
Conclusion
The double sampling method generally provides more accurate population estimates than the double observer method, especially at lower detection probabilities.
Supporting Evidence
- The double sampling method provided unbiased estimates across various conditions.
- The double observer method underestimated populations at low detection probabilities.
- Confidence interval coverage was closer to the nominal level for the double sampling method.
Takeaway
This study looked at two ways to count birds and found that one method is usually better, especially when it's hard to see the birds.
Methodology
Simulations were used to compare the performance of double observer and double sampling methods under various conditions.
Potential Biases
Potential bias exists if the detection probabilities are not accurately estimated or if the population is not fully observable.
Limitations
The study's findings may not apply to all bird species or environments, and the methods depend on accurate detection probabilities.
Statistical Information
Confidence Interval
95%
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website