“…To improve regression algorithms, future studies could make use of additional sampling efforts to increase sample size and test multivariate analysis based on chemometrics or machine learning methods that can handle collinearity between predictors (i.e., different wavelengths from a spectrum), enabling the use of the whole spectral reflectance signature to measure different biochemical traits (Burnett et al., 2021; Che et al., 2023). Integration with multiple sensors, such as pulse amplitude modulation fluorometry (Burdett et al., 2012; Schwarz et al., 2005; Summers et al., 2023), along with measurements of different CCA response categories like calcification and primary production (Cornwall et al., 2019; Page et al., 2022), could further refine noninvasive assessments of their photo‐physiology response to different irradiance regimes. A critical aspect, beyond this study's scope, involves linking CCA phenotypic traits with environmental variables, such as irradiance regimes, sea‐ice cover duration, ice thickness, and underwater light composition (Clark et al., 2013; Huovinen & Gómez, 2020; Irving et al., 2005).…”