Birdwatchers Across North America Tend to Survey Birds in the Morning for No Apparent Reason
College
College of Sciences
Department
Department of Biological Sciences
Graduate Level
Doctoral
Graduate Program/Concentration
Ecological Sciences
Presentation Type
Oral Presentation
Abstract
Bird surveys are typically conducted in the early morning hours. This time of day is chosen since it presumably reflects the greatest activity of birds: many species are most active around sunrise, making them easily detected. While most bird research calls for strict and standardized protocols for when to survey for birds, such approaches are often labor-intensive and limited to very small spatial scales. A community-based approach (often termed “citizen science”) offers a data-intensive alternative to conventional data collection. This community approach involves gathering data from volunteers who submit observations of birds that they encounter at any point, along with information that describes their sampling effort. Such volunteers are not given any specific instructions as to how and when to collect bird data. In this study, I used observations submitted to eBird—a popular web-based platform where more than 800 thousand birdwatchers from Canada and the U.S. have contributed bird sightings between 2010 and 2023. I tested whether observers were biased on when they were birdwatching. I estimated a time-of-day bias as a deviation of estimated kernel density of solar time for >4 million observations from >30,000 locations across Canada and the U.S. relative to a simulated uniform timing distribution. I found a substantial time-of-day bias across observations wherein a large proportion were submitted immediately after local sunrise. Night observations, however, were scarce and represented only a negligible part of the dataset. In fact,
Keywords
eBird, Avian diversity, Time-of-day bias, Community-based science, Ecoinformatics
Birdwatchers Across North America Tend to Survey Birds in the Morning for No Apparent Reason
Bird surveys are typically conducted in the early morning hours. This time of day is chosen since it presumably reflects the greatest activity of birds: many species are most active around sunrise, making them easily detected. While most bird research calls for strict and standardized protocols for when to survey for birds, such approaches are often labor-intensive and limited to very small spatial scales. A community-based approach (often termed “citizen science”) offers a data-intensive alternative to conventional data collection. This community approach involves gathering data from volunteers who submit observations of birds that they encounter at any point, along with information that describes their sampling effort. Such volunteers are not given any specific instructions as to how and when to collect bird data. In this study, I used observations submitted to eBird—a popular web-based platform where more than 800 thousand birdwatchers from Canada and the U.S. have contributed bird sightings between 2010 and 2023. I tested whether observers were biased on when they were birdwatching. I estimated a time-of-day bias as a deviation of estimated kernel density of solar time for >4 million observations from >30,000 locations across Canada and the U.S. relative to a simulated uniform timing distribution. I found a substantial time-of-day bias across observations wherein a large proportion were submitted immediately after local sunrise. Night observations, however, were scarce and represented only a negligible part of the dataset. In fact,