Hosts use resources to combat infections and limited resources may force tradeoffs between immunity and growth or reproduction (Lochmiller and Deerenberg 2000, Rauw 2012). Pathogens also use host resources making it unclear whether host-consumed resources should be beneficial or costly to the pathogen. Resource quantity or quality and host traits might determine whether increased resources are more beneficial for the host or pathogen (Cressler et al. 2014a). In this experiment, we assessed the level of evidence in support of the hypothesis that pathogen traits can alter competition for resources between host and pathogen. We found that resource level was positively associated with pathogen load for Ranavirus, a rapidly replicating endoparasite, but negatively associated with pathogen load for Bd, a more slowly replicating ectoparasite (Fig. 2).
In both experiments, increasing food availability increased frog growth, regardless of whether the frogs were exposed to pathogens or not (Fig. 1A, B), suggesting that hosts in the high resource groups had sufficient resources to combat pathogens. However, neither Ranavirus nor Bd infections affected host growth rate (g/week) or survival, regardless of resource availability. In contrast to growth rates, when hosts were exposed to Ranavirus and had limited resources, hosts had significantly decreased development rates, whereas Ranavirus exposure did not alter development rates when hosts were in the medium or high resource treatment groups (Fig. 1C). Amphibian larvae might increase development to escape a resource-limited environment, as has been seen with other inhospitable conditions, such as with predators (Relyea 2007), pesticides (Rohr et al. 2004), and in rapidly drying freshwater environments (Bekhet et al. 2014).
While higher resource levels did not alter IgY levels, they did decrease Bd intensity in hosts, suggesting that these increases in resources allowed hosts to more effectively combat Bd (Fig. 2A). Increased resources could have led to increases in immunity other than IgY, such as increases in antimicrobial peptides used to combat Bd, or increases in non-immunological defenses, such as increased skin shedding rates, which have been shown to decrease Bd intensity (Rollins-Smith et al. 2002, Cramp et al. 2014). Our results suggest that any increases in Bd replication from increased resources was not sufficient to counteract subsequent increases in immunity, leading to the highest Bd loads when the host had the least food. Bd is an ectoparasite and has a ~ 4 day replication rate, which could make it challenging for Bd to use host-consumed resources effectively (Piotrowski et al. 2004). This pattern of increased food resources decreasing Bd load was shown in a past study on larval southern leopard frogs (Venesky et al. 2012). However, other studies found that the quality of resources do not alter Bd loads in three different species of larval frogs (Cothran et al. 2015, Buck et al. 2016). Unlike these previous studies, our study tested resource quantity and not quality, which could account for these differences. Additionally, amphibian immunity in the larval stage is less complex than in the adult life stage; thus, how hosts use resources to combat Bd in the larval life stage might be different than in the adult life stage that we tested (Rollins-Smith 1998).
In contrast to Bd, abundant food increased Ranaviral intensity (Fig. 2B) while decreasing Ranavirus prevalence. Ranavirus grows best in hematopoietic tissues (Chinchar 2002), suggesting that Ranavirus may benefit from the increased growth we found in hosts with access to more resources. Also, Ranavirus replicates rapidly and is an endoparasite, which may allow this pathogen to more effectively use host resources for replication than some more slowly replicating ectoparasites, like Bd. In a past study on wood frogs (Rana sylvatica), reduced resources did not affect host susceptibility to Ranavirus (Reeve et al. 2013). However, in this study frogs had reduced access to resources only before individuals were infected, which means that Ranavirus likely had similar levels of host-consumed resources to utilize (Reeve et al. 2013). In our study, Ranavirus prevalence decreased with increasing resources, the opposite of the intensity pattern. These results together suggest that increased resources may help the host mount an initial immune response to combat Ranaviral infections, but once established, Ranavirus can use host consumed resources to rapidly replicate.
Our experiments did not test the same host life stage or resource type, and thus we cannot rule out these differences as explanations for our findings. We chose to study different host life stages because we wanted to test particularly harmful (i.e., deadly) pathogens during the host life stage when they are most susceptible (Gray et al. 2009, McMahon and Rohr 2015). Otherwise, differences among the pathogens in their responses might simply be due to low growth rates and thus low statistical variance on the host life stage where the pathogen performs most poorly. Importantly, Bd has much lower growth rates and is less virulent in tadpoles than postmetamorphic frogs, not because of superior immunity at this stage, but because of a lack of substantial keratin, the resource for the pathogen. In fact, although there are differences between the immune systems of tadpoles and adults, both life stages have functioning immune responses to combat and clear the tested pathogens (Rollins-Smith 1998, Gantress et al. 2003, Grayfer et al. 2012, Fites 2014). Previous work has demonstrated that protein quantity in food is an important predictor in how hosts respond to infection (Sandland and Minchella 2003, Venesky et al. 2012). However, both spirulina (fed to the tadpoles) and crickets (fed to the adult frogs) have similarly high protein levels (60–65%; NOW Spirulina, Bloomingdale, IL, USA, Mariod et al. 2017), suggesting that protein differences are unlikely to account for our results. Despite not being able to rule out life stage or diet differences as explanations for our results, we believe that the most parsimonious explanation is that the observed patterns reflect differences in host and pathogen resource use. We encourage future studies to test the hypothesis that rapidly replicating endoparasites and slowly replicating ectoparasites benefit more than and less than hosts, respectively, from host-consumed nutrients using a single host life stage and virulent pathogens.
Organisms in natural ecosystems often experience resource limitation from natural or anthropogenic factors, driving the importance of understanding how resources alter disease progression in hosts. In our study, higher resources increased amphibian growth but altered pathogen abundance positively or negatively depending on the pathogen. These results support the hypothesis that pathogen traits may be an important factor in determining whether the host or pathogen benefit more from increased resource availability. Other studies have found that resource availability is positively associated with pathogen replication in viruses (Arsnoe et al. 2011, Hall et al. 2012) and in endoparasites (Pulkkinen and Ebert 2004, Tylianakis et al. 2004), consistent with our results and the notion that rapidly replicating endoparasites might benefit from increases in host-consumed resources. Future studies should further examine the role that pathogen traits play in determining whether the host or pathogen benefit more from increased resource availability.