Discussion
We found that experimental nest parasitism does not elevate circulating plasma corticosterone in a typically egg-rejecting host of an obligate avian brood parasite 2 hours following the addition of the model eggs. However, at the same time point, the expression of the HPA axis-relevant POMC gene and the stress-related transcription factor ATF3 in the pituitary was elevated in birds exposed to the non-mimetic model egg compared to the individuals exposed to the mimetic egg, although these differences were not significant after false discovery rate correction. FDR limits the type-I error in datasets with a large number of comparisons. However, given that we were mainly interested in a few HPA-axis candidate genes in the pituitary, we also report the uncorrected p-value, while still acknowledging the possibility of this result being a false positive. We detected no changes in the immediate orienting heart rate responses of incubating robins between our model egg-color treatments.
These results are consistent with a delayed activation of the HPA axis in response to brood parasitism. In this study, we measured POMC mRNA levels in the pituitary, not ACTH itself. Although POMC gene expression can be upregulated within 30 minutes of CRH stimulation (Levin et al. 1989), translation and posttranslational modification of the POMC peptide into ACTH likely take time, as POMC needs to be packaged in vesicles where it is cleaved into daughter peptides (Pritchard & White 2007). Thus, it is possible that a POMC-related corticosterone increase occurs after our 2-hour mark. Indeed, we found no association between POMC expression and corticosterone levels at the time of sampling our subjects. Alternatively, it is possible that at 2 hrs we already missed a rapid elevation of circulating glucocorticoids following exposure to the non-mimetic model egg and the p-value based detection of the increased POMC gene expression data is simply a statistical artifact. ATF3, which was upregulated in birds exposed to non-mimetic eggs, is typically upregulated in cells experiencing physiological stress (Haiet al. 1999), although ATF3 expression can also be induced by psychological restraint stressor (Green et al. 2008). The upregulation of ATF3 in the pituitary in response to a brood-parasitic egg thus supports the hypothesis that parasitic eggs are perceived as stressful by the hosts, although the role of ATF3 expression in the pituitary is unclear.
In a different study investigating changes in corticosterone in response to parasitic eggs in a congeneric species (European blackbird) an increase in baseline corticosterone was detected 24+ hrs following experimental parasitism with a non-mimetic egg relative to control treatment (Ruiz-Raya et al. 2018). This suggests that distinctive parasitic eggs may cause a slow but continuous increase in HPA activity over multiple hours. Consistent with this interpretation, we found no difference in the heart rate between the non-mimetic and mimetic egg treatments immediately following the arrival of the female at the nest. Heart rate can serve as a rapid, real-time indicator of acute stress response (Cyr et al. 2009), and the lack of differences in the heart rate immediately after encountering the egg suggests that parasitic stimuli do not cause an acute stress response.
A low, but long-term increase in HPA activity may have functional significance in the context of modulating egg rejection, even if it is not statistically at a single time point of sampling. For example, (Abolins-Abols & Hauber 2020a) found that long-term (overnight) suppression of glucocorticoid synthesis increases the acceptance rate of parasitic eggs by American robins. If natural glucocorticoid release in response to parasitic eggs stimulates egg rejection, then minimal but long-term upregulation of the HPA axis in response to parasitic egg stimuli is consistent with the timeframe of rejection of parasitic eggs by American robins. For example, >90% of robins reject non-mimetic beige cowbird-sized eggs within 5 days (Luro et al.2018), 2 days (Hauber et al. 2019), and even 1 day (Hauberet al. 2020b), but only 33% (7 out of 21) birds rejected the same non-mimetic egg type within 2 hours in this study (also see Scharf et al., 2019). However, individual variation in glucocorticoid levels in this study did not predict egg rejection: circulating corticosterone at our single time-point of sampling did not differ between females who accepted or rejected the non-mimetic beige eggs (Fig. 3).
The few steroid-focused endocrine studies on physiological responses to parasitic egg stimuli so far thus paint a complicated picture. On one hand, experimental studies now show that glucocorticoids (Abolins-Abols & Hauber 2020a) can mediate egg rejection. On the other hand, the HPA axis may only show weak gradual activation in response to exposure to brood parasitic eggs and may have no discernible effect on egg rejection. It is possible that egg rejection by brood parasite hosts may be affected only by pronounced changes in the HPA activity (such as those due to experimental manipulation of hormone levels or intense stressors) and not by weak HPA activation in response to brood parasitism. Furthermore, the relationship between stress physiology and egg rejection may be species-specific. For example, in incubating prothonotary warblers (Protonotaria citrea ), a non-rejecter host species, experimental parasitism with either non-mimetic or mimetic eggs had no effect on glucocorticoid levels relative to non-parasitized controls (Scharf et al. 2021). On the other hand, in yellow warblers (Setophaga petechia ), a species that frequently but variably abandons its nest if parasitized by brown-headed cowbirds, individuals with higher circulating corticosterone levels were more likely to desert an experimentally parasitized nest relative to non-parasitized controls (Turcotte-van de Rydt et al. 2022).
Methodologically, especially when studying freely behaving wild animals, we are still limited by our inability to measure the minute-by minute dynamics in glucocorticoid synthesis and release, as well as our inability to accurately assess long-term subtle changes in glucocorticoids (and other hormones) that could cause a behavioral change. To integrate our findings into a unified framework, future experiments should characterize the full-time course and magnitude of HPA axis activation in response to brood parasitism across acceptor and rejector species. Additionally, dose-response studies, such as those inducing small and short-term changes in hormone levels (e.g., Vitousek et al. 2018) are necessary to test the sensitivity of host behavior to hormones. Finally, future studies must expand beyond the corticosterone paradigm and test the effect of other hormones, such as prolactin, on host behaviors (Abolins-Abols & Hauber 2018; Ruiz-Raya et al.2021), as well as stress-response at the cellular level (Ruiz-Rayaet al. 2022). Together, these studies will allow us to more fully understand the significance of stress signaling in the ecology and evolution of host defenses.