Determining Occurrence Dynamics when False Positives Occur: Estimating the Range Dynamics of Wolves from Public...

Area of Science:

• Wildlife ecology and conservation
• Population dynamics modeling
• Statistical ecology

Background:

• Large-scale monitoring programs are crucial for conservation but can be limited by observational errors, particularly with public-collected data.
• Existing methods primarily address non-detection (false negatives), while misclassification errors (false positives) are less explored but common.
• Accurate estimation of species' occupancy and population dynamics is vital for effective conservation strategies.

Purpose of the Study:

• To develop and demonstrate estimators for dynamic occupancy parameters (colonization and extinction) that simultaneously account for non-detection and misclassification errors.
• To improve the accuracy of inferences from large-scale public surveys by addressing both false negatives and false positives.
• To assess gray wolf (Canis lupus) occupancy dynamics in northern Montana from 2007-2010 using a novel statistical approach.

Main Methods:

• Derived estimators for dynamic occupancy parameters, incorporating a subset of reliably detected observations.
• Simultaneously modeled non-detection (false negatives) and misclassification (false positives) in presence-absence data.
• Utilized hunter survey data (deer and elk hunters) supplemented with locations of radio-collared wolves for analysis.

Main Results:

• Gray wolf occupancy in northern Montana remained relatively stable between 2007 and 2010.
• Wolves were primarily located in high-quality habitats within the study area.
• Site occupancy transitions (colonization/extinction) were infrequent; occupied sites generally remained occupied, and unoccupied sites remained unoccupied.
• Failure to account for false positives led to overestimation of both the area occupied by wolves and the rate of population turnover.

Conclusions:

• Accounting for both false negatives and false positives is essential for improving conservation inferences from public survey data.
• The proposed methodology enhances the understanding of species' status and population dynamics, particularly for species like gray wolves.
• This approach is broadly applicable to various data types susceptible to misclassification errors, improving ecological monitoring and conservation efforts.