From Pixels to Patterns: Trait Plasticity and Species Overlap of <i>Calanus</i> spp. in Arctic Fjords

The marine ecosystems of the Svalbard archipelago are now propelled towards a new climatic state owing to the ongoing "Atlantification" process. Fjords represent valuable natural laboratories hosting both local and advected populations of zooplankton. Our goal was to study life-history traits of two <i>Calanus</i> species (<i>C. glacialis</i> and <i>C. finmarchicus</i>) across slightly different environments in order to better understand their functioning within Arctic ecosystems undergoing substantial transformations. We hypothesized that life-history traits of <i>Calanus</i> copepods (CV life stage) such as size, pigmentation, lipid content, diet, parasite presence, and stage structure would differ across four hydrographically distinct fjords (Hornsund, Isfjorden, Kongsfjorden, van Mijenfjorden) of Spitsbergen. Morphological size-based species identification via stereomicroscopy was supported by molecular methods. Manual image-based measurements of body size and lipid sack area were augmented by machine learning image analyses. Visual color intensity estimations were assisted by HPLC quantification of astaxanthin concentrations. Trophic variability was assessed via stable isotope analyses. The observed substantial variability in the life-history traits highlights their high plasticity and suggests that the traditional morphological distinctions between two <i>Calanus</i> species are becoming increasingly ambiguous. This underscores the need to incorporate genetic tools in ecological studies. The observed variability likely results from the coexistence of several cohorts of both species, including a mixture of local and advected populations. Moreover, each generation could be characterized by different traits depending on their source location and recruitment timing. These findings imply that under progressing "Atlantification" multiple adaptive responses may be expected, including reduced body size, accelerated development, mixed reproductive strategies, decreased pigmentation, shifts in diet, diversified lipid storage strategies, and increased parasite prevalence. Additionally, we introduce a machine learning-based tool for automatic assessment of key traits, such as body size and lipid content from images.