Attention modulates value normalization in human reinforcement learning by shaping reward encoding.

Correlated evolution refers to the phenomenon in which two or more traits exhibit statistically significant coordinated changes during the species’ evolutionary history [, ]. For example, in the lizard genus , habitat-specific ecomorphs show correlated body length, leg length, and tail length evolution [–]; in the cichlid fishes (Cichlidae), the body shape, caudal fin, dorsal fin, and anal fin shapes all exhibited strong correlated evolution consistent with locomotor ecomorphs []; the oral jaw and pharyngeal jaw exhibited correlated evolution consistent with trophic levels []. Fish skin serves as a protective barrier for underlying tissues against the external environment. Typically, it comprises three distinct layers: the epidermis, scales and dermis. The fish epidermis comprises metabolically active epithelial cells with proliferative capacity, participating in barrier formation and wound repair, alongside epidermal mucous cells (EMCs) that secrete mucus for antimicrobial defense and skin lubrication, and epidermal club cells (ECCs) potentially executing additional secretory roles []. Beneath it, the dermis’s dense stratum compactum consists of fibrous connective tissue providing structural support, anchoring scales, and housing vascular/neural networks. Scales are interposed between the epidermis and dermal compact layer, serving dual functions: offering armor-like physical protection [] and enhancing hydrodynamic efficiency by reducing hydrodynamic drag []. This stratified architecture synergistically facilitates defense, osmoregulation, and locomotory adaptability [, ]. Fish have evolved diverse skin modifications in response to varying environmental pressures, but the correlated evolution of these core structural components remains systematically unexplored. ECCs have been extensively investigated in predator-prey ecology due to their hypothesized role as the source of chemical alarm cues released during predatory attacks [, ]. Initial documentation of conspecific-derived antipredator responses dates back to Von Frisch, who observed minnows reacting behaviorally to water-soluble compounds emanating from injured conspecifics’ damaged tissues [, ]; crucially, only epidermal tissue injury elicited these responses []. These findings spurred broader taxonomic surveys for analogous behaviors. Pfeiffer’s subsequent review [], consolidating significant contributions from his own research, demonstrated that alarm responses are widespread among fish species within the superorder ostariophysans but absent in tested non-ostariophysans. He further noted that ECCs are exclusive to ostariophysans and proposed them as the primary candidate for the alarm cue source. Designating these structures “alarm substance cells”, Pfeiffer reasoned that their superficial location, thin-walled morphology, and lack of ducts facilitating content release render ECCs highly susceptible to rupture during predatory strikes. Consequently, the discharge of ECC contents would signal the presence of an actively foraging predator. This established the prevailing view that ECCs contain a chemical alarm cue, or pheromone, alerting conspecifics to imminent danger. Nevertheless, this rupture-triggered alarm mechanism remains largely hypothetical and lacks conclusive experimental verification []. Nearly half a century since Pfeiffer’s review [], the diversity of fishes tested for alarm responses to conspecific skin has been broadened significantly. Subsequent research has consistently revealed flaws in this prevailing view, mainly from the following aspects. (1) Alarm responses to conspecific skin have been documented in non-ostariophysan fish lacking ECCs, such as tilapia () [] and medaka () [], undermining the requirement for specialized structures to produce alarm signals. Tilapia can produce blood-borne alarm cues that function independently of ECCs []. (2) Some ostariophysan species possessing well-developed ECCs exhibit no alarm responses. For instance, both surface-ecomorphs and cave-ecomorphs of the Mexican tetra () possess ECCs, yet only the surface-ecomorphs display pronounced alarm responses []. This suggests ECCs likely serve alternative functions. (3) Alarm responses persist during the larval stage prior to ECCs development. Larval fathead minnows (), for example, possess internal chemical alarm cues or exhibit antipredator responses before the formation of ECCs []. (4) The efficacy of the alarm responses was not compromised when ECCs were suppressed by UV radiation or Cd exposure [, ]. These discrepancies prompted us to re-examine the relationship between ECCs and alarm responses through a comprehensive comparative analysis of skin structure and behavioral assays across diverse freshwater fish species.