Stress. At Transfer (1d), both telencephalic crf paralogs (α and β) were significantly elevated in post–smolts transferred to 13°C compared to the 10°C group ( Figures 2A, B ), whereas only crfα ( Figure 2A ) showed elevation compared to the 16°C and 18°C groups. Conversely, only post–smolts transferred to 16°C and 18°C displayed significantly higher crfbp compared to the 10°C group ( Figure 2C ). Both hsd11b2 (α and β) paralogs were significantly lower in post–smolts transferred to 18°C compared to the 10°C and 16°C treatments ( Figures 2D, E ). Conversely, post–smolts transferred to 13°C had significantly higher mr ( Figure 2F ) and gr2α ( Figure 2I ) compared to fish at 10°C and 18°C, whereas gr1α ( Figure 2G ) was elevated in the 16°C group compared to the 10°C group. This differential regulation in CRs significantly elevated the mr/gr1β ratio ( Figure 3B ) in post–smolts transferred 13°C compared to the 10°C, 16°C, and 18°C groups but lowered the mr/gr1α ( Figure 3A ) ratio in fish transferred to 16°C compared to the 10°C group. However, the magnitude of differences revealed to be larger for the mr/gr1β ratio increased at 13°C compared to the low mr/gr1α ratio seen at 16°C (Supplemental Table 2A)

Neuroplasticity. Neural markers were also affected, with neurod ( Figure 4A ), showing significantly lower values in post–smolts transferred to 18°C compared to the 10°C and 13°C groups, whereas transfer to 16°C significantly lowered neurod levels compared to the 13°C group. Conversely, bdnf ( Figure 4C ) was significantly elevated in post–smolts transferred to 13°C compared to the 10°C group, whereas no differential regulation was observed in post–smolts exposed to the higher temperatures, 16°C and 18°C. The induction of c–fos ( Figure 4D ) was significantly elevated in post–smolts transferred to 13°C and 16°C, but not for fish at 18°C, compared to the 10°C group. Moreover, post–smolts transferred to 13°C showed a significantly higher c–fos response compared to the 16°C and 18°C groups. The expression of pcna was still unaffected at 1 day following the temperature transfer ( Figure 4B ). The magnitude of these differences seen in neural markers showed to be largest for neurod and c–fos (Supplemental Table 2A).

Stress. Following Acclimation (45d), post–smolts at 13°C showed significantly higher crf (α and β) ( Figures 2A, B ) and crfbp ( Figure 2C ) compared to fish at 18°C. Further, crfα ( Figure 2A ) was also higher in post–smolts acclimated to 13°C compared to the 16°C group, whereas the other paralog, crfβ ( Figure 2B ), showed elevation against the 10°C group. Conversely, hsd11b2 (α and β) was significantly lower in post–smolts acclimated to 16°C and 18°C compared to the 10°C group ( Figures 2D, E ), with hsd11b2α ( Figure 2D ) also showing lower levels in the 13°C group.

The CRs were also differentially regulated, with mr ( Figure 2F ) being significantly lower in post–smolts acclimated to 16°C and 18°C compared to the 10°C group, with both gr2 paralogs ( Figures 2I, J ) also showing lower levels at 16°C and 18°C compared to the 13°C group. Further, gr2α ( Figure 2I ) was also lower at 16°C compared to the 10°C group. In contrast, the expression of both gr1 paralogs ( Figures 2G, H ) was unaffected by the prevailing acclimation temperature at a steady state. This suppression in mr decreased both paralog–specific mr/gr1 ratios ( Figures 3A, B ) in post–smolts acclimated to 16°C and 18°C compared to the 10°C group, whereas the mr/gr1α ratio ( Figure 3A ) also displayed lower values in the 18°C group compared to the 13°C group. The magnitude of these differences seen in stress–related genes showed to be strong for gr2α, whereas a moderate difference prevailed for mr and both paralog–specific mr/gr1 ratios (Supplemental Table 2B).

Neuroplasticity. Post–smolts acclimated to 16°C and 18°C had significantly lower neurod ( Figure 4A ) and pcna ( Figure 4B ) compared to the 10°C and 13°C groups, whose genes showed a moderate and strong difference in magnitude, respectively (Supplemental Table 2B). Further, bdnf ( Figure 4C ) was significantly lower in post–smolts acclimated to 16°C compared to the 13°C group, whereas c–fos ( Figure 4D ) was elevated in post–smolts acclimated 18°C compared to the 16°C group.

Stress. The telencephalic expression of crf, hsd11b2, mr, gr2, and gr1 did not respond to the ACT when compared to the initial pre–ACT state [Acclimation (45d)] ( Figures 2A, B, D–J ). crfbp significantly decreased only in the stressed 13°C group compared to their respective pre–ACT state, resulting in similar levels between all stressed temperature treatments ( Figure 2C ). The crfα showed significantly lower expression in stressed post–smolts at 18°C compared to 13°C ( Figure 2A ). Conversely, hsd11b2, mr, and gr2 showed significantly lower expression in stressed post–smolts at 16°C and 18°C compared to the 10°C group ( Figures 2D–F, I, J ).

Further, although gr1 ( Figures 2G, H ) did not respond to the ACT, a differential regulation was still observed when comparing between the stressed temperature treatments. Both gr1 paralogs (α and β) showed significantly lower expression in the 16°C group compared to the 10°C group ( Figures 2G, H ), whereas gr1β was additionally lower compared to the 18°C group ( Figure 2H ). Moreover, a low mr/gr1 ratio was maintained in the stressed 18°C group compared to the 10°C group ( Figures 3A, B ), whereas the mr/gr1β was also lower in stressed post–smolts at 18°C compared to the stressed 13°C and 16°C groups ( Figure 3B ). The magnitude of these differences seen for stress–related genes when challenged by the ACTs was strongest for mr, gr2α, and hsd11b2 (α and β), whereas a moderate difference prevailed for gr2β (Supplemental Table 2C).

Neuroplasticity. In response to ACTs, no differences in neurod and pcna ( Figures 4A, B ) were observed between the pre–ACT and the stressed state. Therein, neurod ( Figure 4A ) was maintained low in stressed post–smolt at 16°C and 18°C compared to the 10°C group, with stressed post–smolts at 16°C also showing lower levels compared to 13°C group. The suppression in pcna ( Figure 4B ) was also maintained in stressed post–smolts at 16°C and 18°C compared to 10°C and 13°C. In contrast, bdnf ( Figure 4C ) significantly increased in stressed post–smolts at 18°C compared to both the respective pre–ACT state and stressed temperature groups at 10°C, 13°C, and 16°C. The c–fos levels were significantly increased in all stressed acclimation treatments compared to their initial pre–ACT state ( Figure 4D ). Moreover, c–fos levels were significantly and sequentially higher at each temperature increase from 10°C and 13°C, with 16°C and 18°C groups having the highest levels ( Figure 4D ). The magnitude of these differences in neural markers seen in stressed post–smolts was strong for neurod, pcna, and c–fos, whereas a moderate response prevailed for bdnf (Supplemental Table 2C).

We performed PCA separately for samplings Transfer (1d), Acclimation (45d), and the ACT (1h) to investigate the inter–relationship between the different stress– and neural–related genes and whole–body cortisol. The KMO test was moderate for Transfer (1d) (KMO = 0.831), Acclimation (45d) (KMO = 0.683), and ACT (1h) (KMO = 0.757), with the Bartlett’s test of sphericity indicating sufficiently large correlation between items for PCA [Transfer (1d): χ² = 1,101.169, df = 105, P < 0.001, Acclimation (45d): χ² = 544.605, df = 105, P < 0.001, ACT (45d): χ² = 496.217, df = 105, P < 0.001). For the PCA on the Transfer (1d), dataset 3 components were extracted, explaining 82.9% of the variance ( Table 2A ), with neural plasticity markers neurod and pcna loading with stress–related genes hsd11b2 (α and β), gr1 (α and β), gr2 (α and β), mr, and crfbp into component 1, crf (α and β) and bdnf into component 2, and cortisol and c–fos loading into component 3. When temperature acclimated (45d), five components were extracted, making up 80.4% of the variance that showed greater gene separation among the PCs ( Table 2B ). Neural markers neurod and pcna loaded with stress–related genes gr2β, mr, and crfβ into component 1, bdnf showed to negatively load with hsd11b2 (α and β) into component 2, c–fos, gr1 (α and β), and gr2α loaded more strongly into component 3 compared to that seen for mr, gr2β, and bdnf, cortisol and crf (α and β) that loaded into component 4, while crfβ loaded together with crfbp into component 5. Following exposure to the added confinement stress, ACT (1h) ( Table 2C ), four components were extracted, making up 76.9% of the variance, with neurod, bdnf, gr1 (α and β), gr2 (α and β), and mr loading together into component 1, whereas neurod, pcna, and hsd11b2 (α and β) showed negative loading with c–fos and bdnf into component 2, crf and crfbp loading together into component 3, and cortisol loading alone in component 4.

Temperature modifications to behavior and circulating cortisol profiles have been evidenced in various teleost species (Lankford et al., 2003; Davis, 2004; Madaro et al., 2018), however, few studies have addressed the potential mechanisms behind these divergences and their implication on key neural processes of the telencephalon. Our findings show that post–smolts abruptly transferred from 10°C to warmer temperatures, 13°C and 16°C, maintained elevated plasma cortisol concentrations, whereas fish transferred to 18°C lacked a cortisol response altogether ( Figure 1 ). Similarly, telencephalic c–fos ( Figure 4D ) was elevated in the 13°C and 16°C groups, with post–smolts transferred to 13°C having the strongest response. This implies that the stress response is still well retained when transferred to 13°C, while a suppression begins to manifest at 16°C and becomes absent at 18°C. Similar cortisol responses are demonstrated in chronically stressed Atlantic salmon (Madaro et al., 2015). Consistent with this notion, post–smolts transferred to 16°C and 18°C upregulated telencephalic crfbp ( Figure 2C ), suggesting a downregulation in the crf system, a typical indicator of a suppressed stress response (Manuel et al., 2014b; Gorissen and Flik, 2016; Samaras et al., 2018). In contrast, crf ( Figures 2A, B ) was elevated in post–smolts transferred to 13°C, supporting a still active crf system that aligns to the largest stress responses seen at this transfer temperature (Gorissen and Flik, 2016). Taken together, these findings suggest impacted stress response capacities in post–smolts when transferred to 16°C and 18°C, with the latter group resembling physiological indices coherent to chronically stressed fish (Madaro et al., 2015; Samaras et al., 2018), that could suggest allostatic loads are beginning to surpass the animal’s reactive scope (Romero et al., 2009).

This differential regulation in stress responses also coincided with alterations to CR profiles, with the 13°C group displaying elevated mr and gr2α ( Figures 2F, I ), the 16°C group showing an upregulation in gr1α ( Figure 2G ), whereas, surprisingly, CRs were less affected in post–smolts transferred to 18°C. Moreover, this differential regulation in CRs was associated with an elevated mr/gr1β ratio ( Figure 3B ) in post–smolts transferred to 13°C, whereas fish at 16°C had a lower mr/gr1α ratio ( Figure 3A ). Collectively, these CR changes could cue an upregulation in mr– and gr1–related functions when transferred to 13°C and 16°C, respectively. Elevated mr at 13°C may also imply that stress conditions are still tolerable, considering the sensitivity of this receptor to cortisol stimulation (Stolte et al., 2008), suggesting that the beneficial physiological effects of mr–related signaling may still be retained to cope with stress loads at this transfer temperature (Karst et al., 2005). Conversely, larger gr1 activation, as suggested by elevated gr1α at 16°C ( Figure 2G ), is a common indice of elevated cortisol stimulation that could represent a larger demand to suppress tissue functions and prevent overstimulation (Stolte et al., 2008), reasoning elevated challenges. Alternatively, the decrease in mr/gr1 ratio ( Figure 3A ) seen at 16°C may also propose an attempt to reduce mr bioavailability, perhaps to limit larger disparities in receptor stimulation as stress loads accumulate. In contrast, fish transferred to 18°C, having a suppressed CRF system (increased crfbp) ( Figure 2C ) and absent stress responses (cortisol and c–fos) ( Figures 1 , 4D ), may explain the limited modulations in CRs as cortisol–related stimulation is readily retained low, thereby lowering the demand to regulate receptor functions. In support, the downregulation in hsd11b2 at 18°C ( Figures 2D, E ) could propose a lower need to inhibit cortisol action (Alikhani–Koopaei et al., 2004; Alderman and Vijayan, 2012). In summary, these telencephalic indices in stress–related genes at 18°C also potentially suggest a convergence toward a reduced reactive scope, with the 16°C group showing an intermediate response (Romero et al., 2009). Contrastingly, post–smolts going from 10°C to 13°C seemingly still retain proper stress responses and telencephalic capacities, cueing manageable stress loads.

Cognition plays key roles in fish stress resilience, in which mild or severe challenges can evoke rapid changes in neural plasticity that modulate cognitive functions either detrimentally or beneficially (Ebbesson and Braithwaite, 2012). Post–smolts elevated bdnf when rapidly transferred to 13°C, in line with our recent findings observing a similar response in fish when exposed from 13°C down to 7°C and 4°C (Tang et al., 2022). This supports the importance of BDNF in managing rapid onsets in stress (Vindas et al., 2017). Elevated bdnf may help resist the adverse effects of cortisol overstimulation on neural networks to preserve neuronal integrity, thus presenting a surprisingly robust response (Vindas et al., 2017). Consistently, bdnf has also shown to promote rapid shifts in behavior to benefit stress resilience under large allostatic loads, as reported for inhibitory behavioral phenotypes (Soliman et al., 2010; Manuel et al., 2014a). Nevertheless, although alteration of bdnf is robust, it also implies a larger demand to resist more adverse stress loads, logically suggesting that rapid temperature elevations are stressful for post–smolts. Surprisingly, in line with the stress response and CR patterns (see Section 4.1.1.), bdnf showed a limited differential regulation in post–smolts transferred to 16°C and 18°C ( Figure 4C ) despite facing a greater thermal challenge. This could suggest greater challenges in line with the downregulation in stress response capacities seen to begin at 16°C and worse at 18°C (see Section 4.1.1.). Alternatively, it cannot be ruled out that suppressed stress functions at 16°C and 18°C may lessen the need to regulate bdnf to protect neuronal circuits, e.g., if cortisol–related stimulation is retained low. However, because allostatic overloads are well recognized to hampering neural functions (Korte et al., 2007; Romero et al., 2009), this still remains a viable alternative for this limited bdnf regulation seen at 16°C and 18°C ( Figure 4C ). In support, post–smolts transferred to 16°C and, especially, 18°C also displayed a suppression in neurod ( Figure 4A ), suggesting a declining pool of differentiating neurons and reduced neural plasticity. Because of the importance of neurod in memory and learning (Salvanes et al., 2013; Grassie et al., 2013), this could argue an inferior cognitive capacity in post–smolts to manage the prevailing stress loads when transferred to 16°C and 18°C. This starkly contrasts post–smolts transferred to 13°C, which still maintained or elevated these neural plasticity markers ( Figures 4A–C ). Taken together, these findings suggest that the transfer of post–smolts to 13°C shows indices of superior cognitive capacities compared to 16°C and 18°C, which, in part, may be due to larger impacts of stress loads on telencephalic neural processes at these higher temperatures.

Stress acts through the actions of CRs to modify neural plasticity, with prior reports demonstrating stimulatory effects of mr, such as for neurod (Van Weert et al., 2017; van Weert et al., 2019) and bdnf (Suri and Vaidya, 2013), whereas, in contrast, gr1 being inhibitory (Suri and Vaidya, 2013; Pavlidis et al., 2015; Chen et al., 2017). In line with this, our findings reveal stress–related elevation in mr and the mr/gr1β ratio ( Figures 2F , 3B ) when transferring post–smolts to 13°C that coincidently maintained neurod ( Figure 4A ) and elevated bdnf ( Figure 4C ). However, it should not be ruled out that crf may also play roles in neural regulation, for instance, bdnf grouped strongly with crf in PC2 of the PCA ( Table 2A ). Nevertheless, these effects inversely mirrored that seen in neural plasticity for post–smolts transferred to 16°C, having elevated gr1α ( Figure 2G ), and suppressed hsd11b2 at 18°C ( Figures 2D, E ). In particular, the downregulation in hsd11b2 ( Figures 2D, E ) at 18°C coincided with the strongest suppression in neurod ( Figure 4A ), supporting prior evidence that hsd11b2 plays important functions in maintaining proper neural development (Alderman and Vijayan, 2012). Conforming to the inter–relationship between stress–related genes and neural plasticity, the PCA results of Transfer (1d) revealed a strong loading of neural plasticity markers with these regulatory genes in the same component (PC1) that explained a large portion of the variance (59.97%) ( Table 2A ). Conversely, cortisol and c–fos grouped together in component 3, revealing a negative loading compared to the other components, perhaps supporting a similar but opposing trend when compared neural plasticity and their regulatory genes ( Table 2A ). Still, although these findings point to the importance of these stress–related genes in regulating neural plasticity, further mechanistic studies are required, considering the complexity of these neural pathways. In summary, these data suggest that post–smolt transfer from 10°C to 16°C and 18°C may impede vital telencephalic functions, whereas transfer to 13°C revealed to be more favorable in sustaining stress and telencephalic responses needed to promote stress resilience and adaptation to the challenge.

When acclimated, cortisol concentrations were basal, indicating recovery from the initial temperature transfer period ( Figure 1 ). Further, in response to the ACTs, all treatments responded with elevated cortisol levels, however, no effect of temperature was observed on the circulating levels ( Figure 1 ). This starkly contrasts our previous work (Tang et al., 2022) and others (Madaro et al., 2018), demonstrating a suppression in cortisol responses when acclimating post–smolts to temperatures below 13°C. However, the present findings are consistent with a prior study in Atlantic salmon, similarly reporting limited modulation in stress–related cortisol release above 12°C (Madaro et al., 2018). Contrasting the cortisol responses, the telencephalic induction of c–fos ( Figure 4D ) to the ACTs increased with increasing acclimation temperature, suggesting larger neural activation to stress at elevated temperatures. As the induction of c–fos is noted for its dependence on stress/stimulus strength (Kovács, 1998) and elevation during both acute and chronic stresses (Pavlidis et al., 2015), this could imply that post–smolts experience greater stress loads when challenged at elevated temperatures, 16°C and 18°C ( Figure 4D ), despite not translating to greater circulating cortisol levels. Therefore, these findings demonstrate c–fos as a relevant stress marker, which, when used in tandem with cortisol, should be considered in future stress studies.

A dual elevation in both crf and crfbp ( Figures 2A–C ) was observed in post–smolts acclimated to 13°C. However, in line with the limited differences in cortisol levels at a steady state, no clear differential regulation between the acclimation treatments was detected for crf and crfbp, suggesting that the CRF system at least within the telencephalon was unaffected by prevailing acclimation temperature at a steady state. Conversely, in response to the ACTs, crfbp ( Figure 2C ) was downregulated in post–smolts acclimated to 13°C, whereas the elevated levels of crfα ( Figure 2A ) were still maintained. This lower abundance of crfbp when stressed at 13°C could serve to enhance the telencephalic activation of the CRF system to manage neural capacities (Gorissen and Flik, 2016). However, as no variation in the cortisol levels ( Figure 1 ) were observed when stressed, the reason behind upregulating this system in the telencephalon requires further study. In contrast, post–smolts acclimated to 16°C and 18°C showed a direct suppression in gr2 ( Figures 2I, J ), mr ( Figure 2F ), and lower mr/gr1 ratio ( Figures 3A, B ). This may suggest that fish acclimated to 16°C and 18°C retain perturbations and disparities in stimulation of mr– and gr2–related functions, which are suppressed as a protective mechanism (Madaro et al., 2015). Moreover, the lower mr/gr1 ratio at 16°C and 18°C ( Figures 3A, B ) may be seen as attempt to downplay mr–related functions or alternatively enhance the bioavailability of gr1 for stimulation to suppress and preserve telencephalic neural processes from elevated stress loads (Stolte et al., 2008). Further, the hsd11b2 ( Figures 2D, E ) became co–suppressed with mr ( Figure 2F ) at 16°C and 18°C, which supports prior work describing a tight regulatory link of hsd11b2 over mr–related action (Alikhani–Koopaei et al., 2004). Thus, the suppression in hsd11b2 at 16°C and 18°C may reflect a repercussion to the stress–related suppression in mr, inherently reducing the demand to maintain elevated hsd11b2 (Alderman and Vijayan, 2012). Taken together, these perturbations in stress–related genes, mr, gr2, and hsd11b2, when acclimated to 16°C and 18°C, cue greater disparities in telencephalic regulation, suggesting a continuation of challenges even after acclimation. Further, considering the importance of mr and gr2 in stimulating neural capacities (Karst et al., 2005), perturbations in these may inadvertently inflict disruptions to telencephalic processes, such as neural plasticity and neurogenesis, and implicate welfare.

Consistent with the seen perturbations in CRs (Section 4.2.1.), acclimating post–smolts to 16°C and 18°C suppressed telencephalic neurod ( Figure 4A ) and pcna ( Figure 4B ), suggesting stress–related impacts on neural differentiation and neurogenesis (Sørensen et al., 2011; Sørensen et al., 2012). Such inhibition of neural functions can impair cognition, due to the well noted roles of these processes in memory and learning (Grassie et al., 2013; Alfonso et al., 2019). Further, these hampered indices of neural plasticity at 16°C and 18°C aligned with the largest c–fos induction when stressed by the ACTs ( Figure 4D ) (Section 4.2.1.) (Kovács, 1998), which may suggest thermal–related disruptions in telencephalic functions. In support, post–smolts acclimated to 18°C also elevated bdnf ( Figure 4C ) when challenged by the ACT, cueing a larger demand to protect neuronal circuits from cortisol or shift behavioral phenotypes to adjust to added stress loads, suggesting elevated stress loads to added challenges (Manuel et al., 2014a; Vindas et al., 2017; Linz et al., 2019).

It is noteworthy that the levels of neurod and pcna ran anti–parallel to the stress–induced telencephalic induction of c–fos to ACTs at 16°C and 18°C ( Figures 4A, B, D ). Although still warranting further study, prior reports in mammals demonstrate the importance of c–fos for synaptic plasticity of learning and memory processes (Guzowski et al., 2001; Fleischmann et al., 2003). Therefore, it is tempting to speculate whether the larger activation of c–fos could relate to attempts to support neural plasticity if dysfunctions in these processes begin emerging in post–smolts at elevated temperatures. Still, it should not be ruled out that the larger induction of telencephalic c–fos may also represent alternative mechanisms to activate other neural pathways to satisfy telencephalic requirements and abilities when challenged at these more stressful temperatures. For example, the importance of c–fos activation is also noted for hippocampal–dependent spatial and fear memory (Katche and Medina, 2017) and could align to the elevation in bndf when stressed at 18°C, which is specifically noted for roles in fear avoidance in fish (Manuel et al., 2014a). Consistently, when stressed, both neurod and pcna negatively loaded with c–fos and bdnf into PC2, suggesting opposing response effects ( Table 2C ). Still, because of our rudimentary understanding over IEG pathways in fish, further mechanistic study is required.

Reminiscent to the initial transfer phase (Section 4.1.2.), the suppression in pcna and neurod ( Figures 4A, B ) when acclimated to 16°C and 18°C coincided with the suppression in gr2 (α and β) and mr and lower mr/gr1 ratio ( Figures 2F, I ,J , 3A, B ), providing further support to the consensus that stress acts through CRs to modify neural processes (Sørensen et al., 2013; Sadoul et al., 2018). This is also supported by the PCAs showing that mr and gr2β loaded strongly with neural markers neurod and pcna in PC1 ( Table 2B ). This is consistent with prior evidence in mammals and fish, demonstrating a tight coupling of mr over neurod potentiation, with this study also pointing to the importance of gr2 (van Weert et al., 2019; Alfonso et al., 2019). Moreover, it is not inconceivable that similar regulatory mechanisms may also exist for pcna in the fishes (Le Menuet and Lombès, 2014). The importance of hsd11b2 for neural processes should also not be undermined as these findings reveal post–smolts acclimated to 16°C and 18°C to having suppressed telencephalic hsd11b2 (α and β) ( Figures 2D, E ), coinciding with both suppressed mr and gr2 ( Figures 2F, I, J ) and neural plasticity (neurod and pcna) ( Figures 4A, B ). This supports reports localizing this enzyme in the neurogenic regions of the telencephalon that argues an importance in limiting discrepancies in CR activity to prevent untimely neural differentiation (Alderman and Vijayan, 2012). In support, hsd11b2, pcna, and neurod loaded negatively together in PC2 of stressed post–smolts ( Table 2C ). Nevertheless, it still should not be ruled out that the lower mr/gr1 ratio ( Figures 3A, B ) at elevated temperatures may also progress the physiological bioavailability GR1 over neural functions (De Kloet et al., 2005; Sørensen et al., 2013; Odaka et al., 2017). In line, studies support the importance of full GR activation for IEG induction (Gutièrrez–Mecinas et al., 2011; Mifsud and Reul, 2018), which may alternatively account for the larger induction of c–fos ( Figure 4D ) at elevated temperatures. In line with this, although not seen when stressed, c–fos loaded strongly with gr1 (α and β) and gr2α in PC3 in non–stressed acclimated fish ( Table 2B ). Still, due to the complexity of these regulatory pathways, further studies are warranted.

In summary, acclimation to 16°C and 18°C lowered the telencephalic expression of genes related to the CRF system (crf αβ and crfbp) while suppressing mr and gr2 (α and β) and lowered the mr/gr1 ratio. These coincided with downregulation in neural plasticity markers neurod and pcna while elevating c–fos induction to added stress at 16°C and 18°C. Together, these indices suggest that challenges are sustained in post–smolts when acclimated to 16°C and 18°C, which may impede proper neural development and have consequences for welfare. Therefore, further investigation is warranted on these temperature–related functional and neuro–anatomical effects and their consequences for cognitive functions.