Dynamic N-occupancy models: estimating demographic rates and local abundance from detection-nondetection data

Area of Science:

• Ecology
• Population Biology
• Wildlife Management

Background:

• Occupancy modeling is standard for species distribution but often overlooks abundance variation.
• Abundance influences key demographic parameters like extinction, colonization, and detection probability.
• Existing methods fail to integrate abundance information with occupancy data.

Purpose of the Study:

• Introduce a novel "dynamic N-occupancy" model.
• Estimate local abundance, population gains, and apparent survival using only detection/nondetection data.
• Account for imperfect detection and abundance heterogeneity.

Main Methods:

• Utilize a dynamic N-mixture modeling framework.
• Leverage detection heterogeneity driven by site abundances to estimate demographic rates.
• Validate model performance through simulations and assess data requirements (years, sites).

Main Results:

• The dynamic N-occupancy model provides accurate estimates of abundance, population gains, and apparent survival.
• Simulations confirm model validity across various parameter ranges.
• Application to barred owls in Oregon revealed significant spatiotemporal abundance increases.

Conclusions:

• The dynamic N-occupancy model enhances ecological inference from occupancy data.
• Explicitly modeling latent population structure improves understanding of demographic processes.
• This method offers a powerful tool for wildlife abundance and distribution studies.