Comparability and repeatability of candling and floating methods for estimating embryo age of waterbird eggs

Determining embryo age is important for predicting hatch dates, planning nest visitation schedules, and as a model covariate examining factors affecting nest survival and behavior. Two common methods of estimating embryo age are egg candling and egg flotation (floating). Despite the reliance on these methods, there is little information regarding the repeatability of these techniques between multiple observers, or whether these techniques produce equivalent estimates. To determine how these methods compare and if precision of each method is biased by individual observers, paired observers candled and floated the same Common Eider (Somateria mollissima) eggs. We compared differences in embryo-age estimates across incubation between (1) candling and floating, (2) paired observers candling eggs, and (3) paired observers floating eggs. Floating and candling produced different embryo-age estimates during some periods of incubation. Although most estimates deviated by a few days, inconsistencies could impact results and statistical power when methods are interchanged between projects or across years. Observers derived different embryo-age estimates for candling, but not for floating, suggesting flotation may be more reliable. Our results suggest caution when combining candling and floating-derived data when collected across incubation stages. Investigators should consider how data collected by multiple observers may affect their research question. Received 22 Apr 2021, accepted 21 Sept 2022.

Automated summary

Determining embryo age is crucial for predicting hatch dates and studying nest behavior of birds. Two common methods, egg candling and egg flotation, are used for estimating embryo age, but there is limited information on their repeatability and equivalence between different observers. This study compared the differences in embryo-age estimates between candling and floating methods, as well as between different observers. The results showed that floating and candling produced different estimates during certain incubation periods, and observers had different estimates for candling but not for floating, suggesting that flotation may be more reliable. Therefore, caution should be exercised when combining data collected using different methods and from multiple observers across incubation stages.