Use of time‐lapse photography and digital image analysis to estimate breeding success of a cliff‐nesting seabird. Issue 1 (4th February 2016)

• Record Type: Journal Article
• Authors: Merkel, Flemming R.; Johansen, Kasper L.; Kristensen, Allan J.
• Is Part Of: Journal of field ornithology. Volume 87:Issue 1(2016)
• Page Start: 84
• Page End: 95
• Publication Date: 2016-02-04
• Subjects: cost‐effective monitoring technique -- digital image analysis -- Greenland -- long‐term monitoring -- seabird reproduction -- Thick‐billed Murre -- time‐lapse photography; Birds -- Periodicals; Ornithology -- Periodicals; Ornithologie -- Périodiques; Oiseaux -- Baguage -- Périodiques; Oiseaux -- Périodiques; 598
• DOI: 10.1111/jofo.12143
• Languages: English
• ISSNs: 0273-8570

Abstract

ABSTRACT: Development of new methods for obtaining basic demographic data from difficult‐to‐access breeding colonies and easily disturbed species is an important challenge in studies of seabirds. We describe a method that can generate data concerning annual breeding success of cliff‐nesting seabirds or other colonial birds with open nests. Our method requires only a single visit to a colony every second or third year, and is based on the use of automated time‐lapse photography. To test our method, we used time‐lapse photos to examine the breeding success of Thick‐billed Murres ( Uria lomvia ) in two breeding colonies in Greenland during the years 2011, 2012, and 2014. Based on the analysis of time‐lapse photos, we found that hatching success during the 3 yr of our study ranged from 60% to 81%, fledging success from 89% to 95%, and breeding success from 53% to 74% (Table 1). Use of digital image analysis made it possible to differentiate between breeding and non‐breeding birds and determine if and when breeding attempts failed or succeeded. The key to making our method a realistic long‐term monitoring technique is the use of an automated, formal image analysis to process the thousands of photos from the time‐lapse cameras and, more specifically, to reduce a large number of photos to a manageable number. Using our method, we needed 12–22 h per study plot, depending on the number of breeding sites per plot (range = 47–127) and whether it was the first or the second time the plot was analyzed, to obtain our estimates of hatching, fledging, and breeding success. This included time for data preparation, image analyses, visual inspections, and summarizing data in a spreadsheet. We found that our estimates of breeding success were comparable to those obtained by direct observation in the field. An important aspect of using time‐lapse technology is to foresee potential reasons why time‐lapse cameras might stop taking pictures, for example, equipment failure (camera, timer, or battery) or interference by visitors (e.g., vandalism or theft). As such, thorough testing of time‐lapse systems and selecting camera locations less likely to be disturbed are most important. We believe that use of time‐lapse photography in combination with digital image analysis to estimate breeding success can be useful for determining the breeding success of other cliff‐nesting seabirds and, more generally, other birds that breed in colonies, especially those located in remote areas and where direct observation may disturb birds.