Modeling Systematic Change in Stopover Duration Does Not Improve Bias in Trends Estimated from Migration Counts

Area of Science:

• Ecology
• Wildlife Population Dynamics
• Conservation Biology

Background:

• Monitoring long-term population trends of migrating animals often relies on counts of unmarked individuals.
• A key assumption is the independence of daily counts and a constant detection rate.
• However, extended migratory stopovers violate count independence and can affect detection probabilities.

Purpose of the Study:

• To investigate how variations in migratory stopover duration impact the accuracy and precision of population trend estimations.
• To assess methods for mitigating bias in population trend data caused by changing stopover durations.

Main Methods:

• Simulated migration count data with known population changes.
• Varied daily probability of survival (as an index of stopover duration) to be constant, random, cyclical, or linearly increasing.
• Tested bias reduction through covariate modeling and data subsampling.

Main Results:

• Systematic increases in stopover duration significantly increased bias and the detection of false population trends.
• Increased stopover duration led to compounding effects on counts due to higher detection and recounting probabilities.
• Bias could not be solely corrected by a stopover duration covariate; sampling modifications were crucial.

Conclusions:

• Standard methods for estimating population trends from unmarked migrating animals are unreliable when stopover durations vary systematically.
• Accurate long-term population monitoring requires incorporating stopover duration as a covariate and employing sampling strategies like subsampling to minimize recounting.