Bacterial killing assays in ecoimmunology require cross-validation by agreement statistics

Over the last 15 years, an optical density (OD-based) technique to quantify bacterial killing assays (BKAs) has been steadily gaining in popularity. This technique uses spectrophotometry to quantify bacterial growth, rather than the colony counts (CFUs) used previously, and reduces the time, resources, and variability inherent to the assay. However, we argue that the OD-based method relies on assumptions that are not true of all immune components, such as leukocytes, and that methods may not be interchangeable. We performed a targeted literature review focused on the methodology of BKAs across vertebrate taxa in ecoimmunology. We then compared the CFU and OD-based methods using leukocytes isolated from Mojave desert tortoises (Gopherus agassizii) and analyzed the quantification method and bactericidal ability using correlation and agreement statistics. Our results suggest poor agreement between techniques, and that immunological processes in cell-based BKAs are likely changing the optical properties of the cultures.