How does animal susceptibility to pathogens vary across landscapes? In this shortlisted paper for the Sidnie Manton Award “Macroimmunology: The drivers and consequences of spatial patterns in wildlife immune defence“, Daniel Becker and colleagues discuss the expanding efforts to apply large-scale approaches to ecological immunology. Daniel is a postdoctoral fellow at Indiana University and will be starting as an Assistant Professor at the University of Oklahoma in Fall 2021. Here, he discusses motivations and future directions for “macroimmunology”.
Many emerging diseases that threaten humans, domestic animals, and wildlife stem from environmental changes that alter infection dynamics in natural reservoir hosts. In many cases, such changes increase the likelihood that reservoir hosts will contact recipient hosts and facilitate cross-species transmission. However, the host immune system plays a critical role in determining whether these new interactions actually result in infection. Accordingly, there has been a growing interest in better understanding how environmental change shapes wildlife immune defense, especially in terms of how susceptible wildlife are to infections, and its implications for better forecasting pathogen spillover.
However, studying host immune systems across landscapes poses new challenges for fields such as ecological immunology, which have traditionally studied individual-level drivers of host defense. Many different aspects of an animal’s environment can shape immunity (e.g., pathogen diversity, food resources, habitat fragmentation), and these can manifest at multiple spatial scales (e.g., microclimate, latitudinal gradients). Many of these environmental factors also likely act in concert to shape immune defense, whereas others, such as urbanization, defy simple categorization into few key environmental conditions. Determining the relative importance of these different environmental factors is only possible by taking a macroecological approach and sampling many host populations across broad landscapes.
In our recently published review in Journal of Animal Ecology, we provide a synthesis of how macroecological approaches could help identify the spatial drivers of wildlife immune defense. Because most work on large-scale patterns in infectious disease are inherently biased by sampling effort (and which pathogens are prioritized for surveillance), a “macroimmunology” approach could provide more general insights and predictions into what sorts of environments may drive wildlife to be most vulnerable to novel infections or act as sources of zoonotic pathogens.
To characterize current spatial studies in wildlife immunology, we critically appraised a random sample of the ecological immunology literature. We identified an overall interest in identifying spatial patterns in immunity, as approximately half of studies sampled multiple populations. Many of these studies examined spatial gradients of anthropogenic influence (e.g., urbanization, contaminants) or abiotic and biotic conditions such as temperature, altitude, and host population size. Yet although spatial approaches to host defense have been fairly common, we found that spatial replication (the number of sampled populations) was heavily right skewed, with a median of four populations per study. We also found that studies with larger overall sampling windows were more likely to sample more populations; however, this followed a weak power law relationship. Importantly, this result implies that studies covering a broad geographic area do not show corresponding increases in the number of populations sampled, which could limit the ability of these studies to capture sufficient environmental variation and differentiate spatial drivers.
Methodologically, we also found that very few wildlife immunology studies quantified spatial dependence or controlled for this in their analyses. However, characterizing spatial autocorrelation from pilot data or from similar systems could identify the spatial scales at which sampling should occur to obtain sufficient immunological variation. Better accounting for spatial dependence can also limit biasing inference when studying environmental change and immunity. Several flexible statistical frameworks, such as generalized additive models and integrated nested laplace approximation, offer promising avenues to account for spatial dependence and nonlinearity while also identifying hot- and cold spots of immunity. In one recent application by the authors with red deer (Cervus elaphus), we showed how such an approach can highlight spatial agreement and discordance between immune metrics and parasitism.
What are some future avenues for macroimmunology? Although spatial approaches to ecological immunology are common, spatial replication is generally low and could limit the ability of such studies to capture the environmental variation required for macroecological analyses. Our review lastly highlights what we see as several priority topics. We focus particularly on rodents and songbirds, given that these taxa are generally easy to live capture, can facilitate sampling many replicate populations owing to small home ranges, are common in anthropogenic habitats, and can function as reservoir hosts for many zoonotic pathogens, including, but not limited, to hantaviruses and flaviviruses.
We suggest future large-scale ecological immunology work focus on spatial questions around three topics in particular that generally deal with large spatial scales and integrate multiple environmental factors: (i) does immunity follow biogeographic patterns, (ii) how do range expansions affect defence, and (iii) does urbanization have consistent impacts on immune phenotypes? Field studies focused on these topics could provide new insights into the environmental drivers of immune defense while also facilitating novel opportunities to predict infection risks in the context of climate change, range expansion and biological invasion, and land conversion. By generating large-scale datasets, these endeavors could also facilitate more classic macroecological analyses to identify general laws in how immunity varies over space. In particular, advances in ecological applications of machine learning algorithms and hierarchical meta-analyses could test how particular immune branches (e.g., innate or adaptive) generally respond to environmental drivers like urbanization and thus what kinds of pathogens may have greater impacts on urban wildlife.
Ultimately, macroecology holds promise as an approach to identify the drivers of spatial variation in wildlife immune defence. We hope that our synthesis can help guide future studies of how environmental variation shapes immunity and that such applications can help derive more general insights into where wildlife are most vulnerable to infection.