Ecology and Evolution of Infectious Disease

Pathogens are a ubiquitous component of ecological communities, where they strongly impact numerous aspects of host ecology and evolution. Understanding these ecological and evolutionary dynamics is made urgent by the emergence of new diseases and a need to understand the role of biotic interactions in the dynamics of natural ecosystems. Projects in the lab focus on a range of important questions in disease evolutionary ecology, and do so by combining mathematical models with empirical studies of anther-smut disease (Microbotryum) on wild carnations.  


Anther-smut is a model system for disease ecology because its biology and transmission have strong analogies to vector-borne and sexually transmitted diseases in animals and humans. It is easily detected in nature by the dark brown spores that take the place of pollen on the anthers, and prevalence estimates are straightforward since the fungus sterilizes but does not kill the host.  Moreover, demographic studies, manipulative field experiments, and controlled inoculation studies are much easier in plants than animals.

Ecology and evolution of transmission mode

Evolution of new transmission modes can play an essential role in disease emergence, and pose serious challenges for disease control (e.g. sexual transmission of zika virus).  However, concerns about evolution in emerging infectious diseases has largely focused on increases in virulence and we know much less about the ecological factors driving the evolution of transmission mode. I am using anther-smut disease on alpine carnations as a model system to study the impacts of host density and host resistance on the evolution of transmission mode. Previously, We have shown that the disease is maintained in nature through two different modes: frequency-dependent vector-transmission to adult flowering plants and density-dependent aerial-transmission to highly susceptible seedlings. 

Current research is focused on determining how variation in host density and age-structure affects the success of transmission through these two different routes. This work is part of a collaborative NIH-EEID funded project with Janis Antonovics, Michael Hood and Mike Boots. 

Grey lines: predicted disease prevalence in the Rifugio Garelli population under either a pure-vector transmission model or a mixed vector and aerial transmission model.  Black lines show observed prevalence. From Bruns et al. 2017. Transmission and temporal dynamics of anther-smut disease (Microbotryum) on alpine carnation (Dianthus pavonius). Journal of Ecology, 105: 1413-1424.

We use arrays of cut flowers in tubes to manipulate density and study spatial patterns of spore deposition through vector and aerial transmission routes

Evolution of age-dependent resistance

Juvenile susceptibility is wide-spread among plants, wildlife, and humans, and plays an essential role in disease dynamics. For example, many of the diseases that have shaped the course of human history, such as measles and smallpox are primarily ‘childhood diseases.’ Yet we still lack a basic evolutionary explanation for why juvenile hosts retain such high susceptibility.  We are using a combination of mathematical modeling and comparative inoculation studies to understand the costs, benefits and numerical feedbacks that drive the evolution of age-specific resistance.

Resistance to anther-smut disease in Dianthus pavonius has strong age-specificity.  Seedlings are ten times more susceptible to the disease than adults, and we have shown that this high level of juvenile susceptibility carries significant epidemiological consequences in the field. The pattern cannot be explained by invoking physiological constraints, since our inoculation studies have shown that there is genetic variation for juvenile resistance. Together with Ben Ashby, I have been developing theory to understand the evolution of juvenile susceptibility, and its relationship to host lifespan. Our recent paper shows that juvenile susceptibility should be greatest in long-lived hosts.  We are working on testing this prediction by comparing age-specific susceptibility in anther-smut hosts with different lifespans. This work has just been funded by a joint NSF-NERC program on 'Bridging Ecology and Evolution"!!

Susceptibility of adult and seedling D. pavonius plants

Disease and the evolution of separate sexes

Sex-specific differences in disease exposure have been widely observed throughout the animal and plant kingdoms, yet, whether these differences in disease risk play a role in the incipient evolution of separate sexes (dioecy) from hermaphroditism has been largely overlooked. Dianthus pavonius has a gynodioecious mating system composed of hermaphrodites and female plants. Female flowers have significantly lower levels of spore deposition than hermaphrodite flowers. The mechanism for this difference in exposure appears to be floral closing behavior: female flowers close shortly after receiving pollen reducing their duration of disease exposure, while hermaphrodites remain open.

Spore deposition on hermaphrodite and female flowers.

The strength of selection for females is only half the equation: the evolutionary response to sex-specific differences in disease exposure will also depend on the genetic basis of sex.  In gynodioecious species sex is commonly determined by epistatic interactions between cytoplasmic male sterility (CMS) genes located in the mitochondria and ‘restorer’ genes located in the nucleus. We’ve shown that strong selection from disease can drive the fixation of CMS alleles (Miller and Bruns 2015), resulting in an increase in female frequency, and the evolution of nuclear sex determination (a critical pre-cursor to the evolution of dioecy). We are currently carrying out crossing studies to identify D. pavonius families with different CMS cytotypes, and developing mitochondrial markers. We will use these tools to search for signatures of past CMS selection in natural plant populations with high and low levels of anther-smut disease.

Evolution of CMS (red line), restorer (blue line), and female frequency (black line) in response to the introduction of disease (dashed black line).  

Previous projects

Disease can drive numerical fluctuations in host populations and rapid evolution.  Fitness is at the very heart of these dynamics: yet for many natural host-pathogen interactions we still lack basic information on the fitness impacts of disease. This is especially true for long-lived perennial hosts, where single point measures of survival, growth, or reproduction are insufficient for predicting either numerical dynamics or selection gradients.  ‘Aster’ is a relatively developed statistical method for estimating lifetime fitness from repeat measures of survival and reproduction (Shaw et al. 2008). I have been working with Helen Alexander (University of Kansas) and Carolyn Malmstrom (Michigan State University) to apply these methods to a grass-virus system. Increased habitat agricultural development in the Great Plains has allowed crop-associated viruses to ‘spillover’ into native prairie grasses, yet we know very little about their fitness impact. Our recent paper shows that that BYDV-PAV, a wheat-derived Luteovirus has significant impacts on the lifetime fitness of switchgrass (Panicum virgatum).

Fitness effects of virus on a native grass

Fitness impacts of virus on two cultivars of switchgrass.  Cumulative fitness was estimated by integrating survival, flowering, and inflorescence data in aster models. Fig. from Alexander et al. 2017.

Rapid pathogen evolution in response to new drugs or, in the case of agriculture, new disease-resistance genes, poses a major threat to human health and food security. My PhD work focused on understanding the nature of evolutionary constraints in the fungal pathogen Puccinia coronata, which causes crown rust disease in oats (Avena sativa). Rapid evolution of infectivity in many agricultural pathogens, including crown rust, has raised the possibility that these pathogens are successful because they lack genetic constraints. My results show otherwise; I found evidence of limits to pathogen variation in some traits (Bruns et al. 2012) and a trade-off between infectivity and reproduction. Pathogen genotypes that were able to defeat a larger number of host resistance genes suffered a cost of delayed reproduction (Bruns et al. 2014). These results demonstrate that, in agricultural settings where resistance is widespread, pathogen fitness costs are not a significant barrier to the rapid evolution of multi-locus infectivity.

Evolution of pathogen infectivity 

Left) Oat seedling infected with P. coronata.  Right) Relationship between multilocus virulence level and latent period (time between infection and sporulation) for 29 P. coronata genotypes. Virulence level was defined as the number of host resistance genes a pathogen strain could defeat.  

Can disease effect host distributions? A classic ecological model of species distributions is that abundance is highest in the range center and declines towards the range limits. Many diseases have a threshold host density for disease persistence, leading to the expectation of a disease-free halo at species range margins. However, diseases with frequency-dependent transmission (e.g. vector and sexually transmitted diseases) are not restricted by thresholds because their transmission is relatively independent of density.  Theory predicts that frequency-dependent diseases should be able to persist at host range margins.  We recently tested this prediction by mapping the elevational distribution of four different alpine plants and their anther-smut pathogens.  We found that disease was present at host elevational range margins and occurred at high prevalence (Bruns et al. 2018).  We have also shown that the disease can persist at high prevalence and reduce host population size (Bruns et al. 2017).  These results indicate that disease could be impacting on host distribution.

Disease at species range limits

Distribution of anther-smut disease on D. pavonius. Host distribution is shown in light green for reference. Top: Proportion of populations with disease. Bottom: Disease prevalence within populations.

This site was designed with the
website builder. Create your website today.
Start Now