Community epidemiology

• mult-host, multi-parasite interactions in a community context

Multi-host parasites

• One parasite that infects multiple host species (recall previous lectures)

• Can be either at species (rabies) or individual (mosquitos) level

Host-predator-prey

• Predator eats prey, prey gets infected by parasite

• Important in a couple ways:

  • trophic transmission
  • healthy herds

Trophic transmission

• consumption of prey is necessary for the parasite to advance life cycle
• the parasite infects the predator as well

Healthy herds

• Predators might eat prey non-randomly
• Sickness behaviors leave prey weaker or more isolated from groups
• The consumption of infected prey by the predator could reduce disease burden in the host population

Co-infection

• The same host individual is infected by more than one parasite species

• Extremely common

Co-infection

• Do parasites interact?

• Through what mechanisms?

  • bottom up (e.g, competition)
  • top down (e.g, immune-mediated)

• How do interactions within infected hosts affect population-level patterns?

Coinfection can result in non-linear changes to

• disease severity
• host survival
• parasite replication and shedding
• design of control programs

Detecting parasite-parasite associations

_{Dallas et al. 2019 PRSB}

Types of interactions

Bottom-up + Within-host competition for resources or space + Ecological interference via reduced contact or density

Top-down + Facilitation via host defenses (HIV and opportunistic infections) + Indirect competition via host defenses (e.g., immune priming or cross protection)

How do we estimate the effect of coinfection?

• Either parasite species richness (treats infection by parasite \(i\) as binary) or an estimate of infection intensity for each coinfecting parasite

• Hosts with lots of parasites are not necessarily the ones with highest burdens

  • If parasites compete or that host is mounting a strong but failing immune response

• The effect of host traits, geography, etc. need to be considered

  • We’ll get into a conceptual framework for thinking about coinfection later

Some definitions before we start to conceptualize how this works

• Infracommunity : all the parasite species within a single infected host

• Component community : all the parasite species within a host population (basically a local estimate of parasite species richness for a species)

Let’s start with a null model

• Null models incorporate no ecological process

• They make really simple assumptions that provide good benchmarks

What would the distribution of coinfection look like if all hosts were equally probable of getting infected and all parasites were distributed independently? (5 minute small group discussion)

So we’ll start with assumptions

• All hosts equally likely to be infected
• All parasites are distributed independently

Uses of null models to understand host-parasite communities

_{Canard et al. 2014 Am Nat}

Incorporating non-random processes could lead to expected differences

• So what could we incorporate into our null to address this?

• Covariance between parasite distributions (some hosts are more likely)
• Incorporate host traits like body size (big hosts more parasites)
• other stuff

Examples of coinfection and how we study it

• Observational field studies

• Experimental field studies

• Meta-analysis

• Models

Observational field studies

_{Ezenwa & Jolles 2011 Int & Compar Biology}

Tuberculosis and helminth coinfection in wild buffalo

_{Ezenwa & Jolles 2011 Int & Compar Biology}

Infection intensity of intestinal helminths

• Positive relationship between infection intensities of coinfecting parasites

• What else could be driving this apart from joint parasite effects?

_{Lello et al. 2004 Nature}

Experimental field studies

• Wood mice have many parasites

• Are they interacting?

• Tested using nematode parasite knock-downs by treating wild populations with ivermection

_{Knowles et al. 2013 PRSB}

No entire community response, but found something

• A protozoan parasite significantly increased in abundance when nematode abundance was knocked down

• This parasite is located in the same tissue, suggesting a potential role of parasite-parasite competition

_{Knowles et al. 2013 PRSB}

Meta-analysis

_{Griffiths et al. 2011 J of Infection}

Models

_{Gorsich et al. 2018 PNAS}

Models

_{Gorsich et al. 2018 PNAS}

Models

_{Gorsich et al. 2018 PNAS}

Models

_{Seabloom et al. 2015}

Side by side

Seabloom

Gorsch

Models

_{Seabloom et al. 2015}

A conceptual framework of coinfection

• Exposure and susceptibility as drivers of coinfection

• This mirrors how we think about transmission as being encounter * susceptibility

_{Viney & Graham 2013 chapter in book Advances in Parasitology}

Exposure

• geography: host and parasite must occupy the (micro)habitat

• behavior: host behavior modifies exposure to parasites

• host traits: host traits modify exposure to parasites (e.g., home range size)

Susceptibility

• genetics: immune or defense traits

• diet: can control exposure (through contact) and susceptibility (through host condition)

• behavior: can control exposure (through contact) and susceptibility (if behavior is costly)

End of lecture 1

What have we learned?

• Coinfection is incredibly common

• Generally has negative consequences to host fitness

• A bit on null models and some fun coinfection examples

Models of coinfection

Pathogen coexistence, with dashed line being immunosuppression and dotted line being cross-protection.

_{Seabloom et al. 2015 Ecol Letters}

Small group activity

3 minutes to provide an interpretation of this figure

• explain the figure to your neighbor

• have your neighbor explain the figure to you

• come up with a composite explanation

Parasite community assembly processes

• Parasite exposure can occur at the same time, but is this likely? (not really)

• The timing of parasite exposure matters to the resulting parasite community!

Priority effects

• Priority effects describe the process of community assembly, in which the order of arrival determines the resulting community structure

• This is related to coinfection, as being exposed to two parasites at the same time might have very different outcomes from being exposed to two parasites at different times

Infection trials at the same time

Increasing parasite richness decreased parasite persistence

_{Johnson & Hoverman 2012 PNAS}

Infection trials at the same time

But even if parasites did not persist, host survival declined with increasing parasite richness

_{Johnson & Hoverman 2012 PNAS}

Infection trials at the same time

And parasite load increased

_{Johnson & Hoverman 2012 PNAS}

How can we get at parasite community assembly?

• mark-recapture studies in wild populations

• experimental lab studies

• models

mark-recapture studies in wild populations

• Work from folks that we’ve shown

  • Amy Pedersen (mouse-helminth system)
  • Pieter Johnson (amphibian-helminth system)

• Idea to is to longitudinally monitor individuals, which allows parasite community assembly to be estimated directly

• pros and cons?

experimental lab studies

• Work from folks that we’ve shown

  • Daniel Benesh (fish-helminth system)
  • Pieter Johnson (amphibian-helminth system)

• Idea is to experimental control timing and dose of parasites to explore how parasite communities could assemble.

• pros and cons?

models

• Work from folks that we’ve shown: Gorsich, Seabloom, Ezenwa, etc.

• Idea is to generate testable hypotheses by leveraging both disease theory and community assembly theory from free-living organisms.

• Pros and cons?

coinfection models can get a bit wild

_{Marchetto and Power 2018 Am Nat}

Why is it so difficult to model the impact of coinfections on resulting disease dynamics?

5-10 minutes in small groups

So coinfections influence host fitness non-linearly

• explorations of parasite community assembly and coinfection suffer from feasibility issues and chonky models

  • but tons of great work here

• What if we simplify the approach?

  • What is the expected distribution of parasites in a host population if coinfection just happens randomly?

  • i.e., there is no facilitation, there is no host trait variation that promotes coinfection, coinfecting parasites do not interact.

One parasite

Two parasites

Two parasites

To what extent does high burden of one parasite correspond to high burden of another?

Is this a good approach?

• Doesn’t really get at costs of coinfection

• Can’t tease apart role of host condition

• Assumes that priority effects don’t matter

• It is extensible to

  • a lot of parasite species
  • variation in host infection probability
  • different distributions of infection intensity
  • host and parasite communities combined

So that was an odd note to go out on, but I wanted you to link some of the work of infectious disease in community models with some of how we previously talked about aggregated burdens to this lecture on coinfection dynamics

Next time we’ll talk about large-scale patterns of parasite diversity and some other fun bits