‘Social distancing’ in fish makes their parasites less deadly

Infectious diseases can transmit during social interactions. How do changes in host sociality change how diseases adapt?

Like Comment
Read the paper

Many infectious diseases transmit when hosts are in close proximity. For the parasites (and pathogens) that cause these diseases in animals, social interactions are a golden opportunity to transmit. As a result, many animal hosts "social distance" in response to infectious disease outbreaks. In turn, how should parasites adapt in response to such changes in contact rate?

In theory, evolving parasites face a tradeoff. If parasites exploit their host more, then they can benefit from faster transmission. However, this exploitation can make parasites more lethal (virulent), causing infection to end prematurely with host death. Should "social distancing" in hosts drive parasites toward the more exploitative and lethal end of this tradeoff or the more conservative and benign end?

View from atop a waterfall in Trinidad.
Photo credit: Mary Janecka.

These questions are hard to answer empirically. Host social behavior is difficult to manipulate, experimentally, and may change quickly before parasites have had an opportunity to adapt to it. But at the mixer for the Disease section of the Ecological Society of America, I struck up a conversation with Dr. Jessica Stephenson about a unique opportunity with guppies.

Waterfalls divide the Trinidadian guppy into low predation and high predation populations. This long-term difference in predation leads to a long-term difference in how much guppies shoal to defend themselves against predators. So this "natural laboratory" offered us populations with consistent differences in social contact rates to test how parasites had evolved. Further, Dr. Stephenson introduced me to a rich literature characterizing these guppy populations and a worm ectoparasite that transmits during guppy-guppy contact.

Trinidadian guppy populations lie above (upper course) or below (lower course) waterfalls. This causes sustained differences in predation. In the lower course, the high predation regime has led to much higher guppy shoaling rates as an effective defense against predation. This “natural laboratory” is replicated across many rivers in Trinidad, allowing us and others to characterize many aspects of this system.
Reproduced with permission from Walsman et al.

Thanks to this rich literature, I was able to develop and parametrize a theoretical model of how predation would shift the coevolution of host contact rate and parasite virulence. The model predicted that predation would have little direct impact on virulence evolution but would drive hosts toward much higher contact rate as they grouped for defense against predators. Higher contact rate, in turn, would greatly increase the rate of multiple infections, causing multiple parasite strains to compete on one hosts, selecting for higher virulence. Further, the model gave quantitative predictions for how virulence parasites should evolve in low or high predation populations.

While I joined the Stephenson Lab and worked on the model, the lab's other postdoc, Dr. Mary Janecka, led a team to Trinidad. They sampled right before Trinidad's borders closed in 2020, collecting parasitic worms from low predation populations and high predation populations. We brought these worms back to Pittsburgh and grew them in common garden conditions, measuring their virulence from the death rate of infected guppies.

Dr. Mary Janecka sampling a wild guppy population in Trinidad.
Photo credit: David Clark

As predicted, worms from high predation populations were deadlier than those from low predation populations. Further, the empirical data quantitatively followed model predictions! This match strengthened our confidence in the model's theoretical insights into how predation drives increased social contact and social contact drives evolution of higher virulence. Thus, the guppies in low predation populations were able to "social distance", leading to parasite evolution of lower virulence.

By predicting parasite virulence, the model predicts infected host death
rate imposed by worms from low or high predation regimes. Worms
from high predation regimes imposed significantly higher death rate
in common garden conditions, closely following model predictions.
Reproduced with permission from Walsman et al.

 Beyond guppies, many host-parasite systems have the same core ingredients.  A massive diversity of animal hosts, vertebrate or invertebrate, group for defense against predation and a massive diversity of parasites in these hosts require close proximity for transmission. Further, the fundamental mechanism of increased social contact → increased multiple infections → increased virulence should apply more broadly, perhaps even to some human infections? Hopefully, our model-data synthesis inspires further theory on and empirical tests of vital questions about sociality and parasite evolution.


Walsman, J. et al. Social guppies evade predation but have deadlier parasites. Nature Ecology & Evolution. 10.1038/s41559-022-01772-5 (2022).

Jason Cosens Walsman

Postdoctoral Research Fellow, University of Pittsburgh