Abstract Background Correlated evolution refers to the coordinated changes of multiple traits during species evolution. Fish skin exhibits diverse structures comprising the epidermis, dermis, scales, epidermal mucous cells (EMCs), and epidermal club cells (ECCs), yet the correlated evolution among these components remains unexplored. ECCs are classically hypothesized to be the source of chemical alarm cues in ostariophysan fishes, which are passively released during predator-inflicted skin damage to alert nearby conspecifics of active predation threats, this process known as the alarm response. However, we identified two critical exceptions from 53 freshwater fishes: naked carp ( Gymnocypris , ostariophysans) have lost ECCs, while mandarin fish ( Siniperca , non-ostariophysans) have acquired them. The established link between ECCs and alarm responses in ostariophysans may be confounded by phylogeny. Their consistent co-occurrence does not establish a causal relationship. The two evolutionary exceptions serve as ideal models to test this relationship. Results We investigated correlated evolution across seven core skin structures, including epidermal thickness, dermal thickness, scale thickness, EMCs size/number, ECCs size/number in 53 freshwater fishes. Analysis revealed significant correlated evolution among skin structures. Epidermal thickness showed positive correlations with all other metrics, except for the relationship between scale thickness and EMCs number. Scale thickness exhibited antagonistic correlations with both epidermal and dermal thickness. Notably, scale thickness positively correlated with EMCs number, reflecting functional adaptation to reduce inter-scale friction. Critically, behavioral experiments demonstrated that naked carp, despite lacking ECCs, exhibited robust alarm responses. Furthermore, we tested six close relatives of the naked carp, all of which possessed ECCs and exhibited robust alarm responses. Conversely, mandarin fish, despite possessing ECCs, showed no alarm responses. This evidence does not support ECCs as the necessary source of chemical alarm cues. Conclusions In conclusion, our study demonstrates significant correlated evolution among core skin structures in freshwater fishes. Crucially, we challenge the prevailing paradigm by showing that chemical alarm cues can originate from integumentary components other than ECCs, evidenced by the robust alarm response in ECCs-lacking naked carp. Secondly, ECCs may play a multifunctional role evidenced by the absence of alarm response in ECCs-possessing mandarin fish. These insights advance our understanding of fish skin functional morphology and ecology.
Li et al. (Sat,) studied this question.