Heightened risk of infection does not alter macronutrient-specific feeding behavior.
Because shifts in behavior and pathology associated with infection could act as social cues of heightened infection risk to uninfected conspecifics, we also examined if perceived risk of infection (seeing sick conspecifics) could alter feeding behavior and macronutrient selection. Separate lines of evidence indicate that social cues can influence feeding behavior and alter physiological responses relevant to immune function (Cornelius et al. 2018, Schaller et al. 2010, Stevenson et al. 2011, 2012; Love et al. 2021, Gormally and Lopes 2023). Thus, we predicted that birds exposed to a heightened risk of infection would have physiological responses relevant to responding to an immune threat and shift feeding behavior in a way that maximizes immunity. Contrary to our predictions, we found no evidence for shifts in feeding behavior, macronutrient intake, gut microbiota, or physiological responses in birds exposed to a cue of heightened infection risk. It is possible that the cue of infection was not sufficient to stimulate physiological changes or alter the feeding behavior of focal birds because we used a simulated infection (injection with LPS) in this study. The behavioral effects of LPS typically only last between 2-4 days following injection (Sköld-Chiriac et al. 2014, Love et al. 2023), thus the cue of infection elicited by LPS-injection is temporally limited. It is also possible that we failed to capture the relevant timing of shifts in physiology following exposure to sick conspecifics. Recent work in Japanese quail (Coturnix japonica ) found that females exposed to LPS-challenged males for 3 hours had an upregulation of immune genes in the blood, suggesting that physiological responses to LPS-challenged individuals might happen rapidly post-cue (Gormally and Lopes 2023). Additionally, it might be less costly for organisms to employ behavioral defenses (such as avoidance behavior) rather than physiological immune defenses in response to an immune threat in some settings. For example, we previously documented that zebra finches housed across from LPS-challenged conspecifics reduce flight activity and increase preening behavior (Love et al. 2023), and there might be trade-offs in how birds invest in behavioral versus physiological immune defenses (Zylberberg et al. 2012). Future work should explore whether an immune challenge or infection that elicits stronger and longer-lasting behavioral and physiological signs of disease is capable of influencing feeding behavior, gut microbiota, and physiological responses in healthy individuals, as this could have implications for host disease susceptibility and disease transmission potential.
Although we did not detect an effect of perceived infection risk on any of the physiological parameters examined in this study, we did observe changes in testosterone concentrations in male zebra finches over the course of the experiment. Specifically, we observed a decrease in testosterone levels in males 3 days after the injection of stimulus birds, and this decrease occurred in both LPS-focal and saline-focal males. Physiological stress is known to have inhibitory effects on the release of testosterone (Da Silva 1999), so it is possible that the observed decrease in testosterone concentrations in males is related to handling stress associated with repeated blood sampling. However, we did not observe a corresponding increase in circulating plasma corticosterone in birds over the course of the experiment that might be indicative of acute or chronic stress.
While a heightened risk of infection did not influence feeding behavior or gut microbiota composition, higher fat consumption was associated with lower observed richness and lower Shannon diversity across both treatment groups (LPS-focal, saline-focal). Consistent with this result, high fat diets in mice result in lower gut microbiota richness and diversity (Zhang et al. 2012). Similarly, house sparrows experimentally fed a high-fat urban diet have decreased gut microbiota diversity (Teyssier et al. 2020). Conversely, protein consumption by focal birds was unrelated to microbial richness and diversity. Interestingly, neither fat nor protein consumption predicted richness or Shannon diversity in the infected birds (LPS-injected, saline-injected). It is unclear why fat consumption predicted microbial alpha diversity in focal birds but not injected birds, however larger sample sizes might be required to confirm this trend and determine the biological underpinnings of these differences.
The present study extends our understanding of how immune activation can influence feeding behavior and diet selection in vertebrates. We did not detect any shifts in feeding behavior, nor any immune or endocrine changes in response to social cues of infection in the present study. However, we did detect macronutrient-specific illness-induced anorexia in LPS-challenged birds, where birds decreased protein but not lipid intake. Shifts in feeding behavior in sick individuals can affect both host and parasite fitness and ultimately influence disease severity, which is inherently related to disease transmission (Hite and Cressler 2019, Povey et al. 2013). Models indicate that sickness-induced anorexia, like the reduction in caloric intake observed in LPS-birds in the present study, is capable of enhancing or diminishing disease severity depending on dietary context (Hite and Cressler 2019), suggesting that the interactions between infection, resource availability, and host macronutrient selection can have important consequences for disease dynamics and deserve further attention.