As many people who follow the blog know, there is a lot of work being done on Trinidadian guppies. Just in case you don't know about guppies, a quick overview. Guppies are native to Venezuela and Trinidad, and in the northern mountain range of Trinidad, it so happens that there is a lovely playground for evolutionary biologists. Starting with Ben Seegher
's PhD thesis and continued by very cool scientists like John Endler
, David Reznick
, Helen Rodd
, and Anne Magurran
, the guppy became a model system for rapid adaptation. Guppies in low predation areas are more colourful and differ in life history and behavioural traits, among many other things, as compared to guppies that co-habit with dangerous predators. For the purposes of our paper, we focus on male colour, a well known adaptive trait in the guppy.
The spatial variation in predator composition means guppies have to adapt to their environment to be able to reproduce, and thus, a lot of work has focused on the effects of predation or surveyed differences among populations in relation to the predation level. However, we know life for guppies is very complicated, and that there are other selective pressures acting on them, including parasitism. If we think of adaptation beyond a single causal force, what happens? Will it alter our perceptions of the original causal force? Will the causes interact? Could it be a stronger force than the original one? To address these questions of uni- vs. multi-factorial causes of variation in phenotype, we looked at the effects of both predation and parasitism on male guppy colour.
A pike cichlid, one of the predators
While there are many ways to consider the effects of parasitism, we chose to focus on the ectoparasite Gyrodactylus spp
because previous work has shown that Gyros (as we affectionately call them) can affect a number of traits that also differ between high predation (HP) populations and low predation (LP) populations. The parasite is transferred through fish contact, and we also know that parasitism levels vary considerably in natural populations. We can see Gyros through a stereo-microscope, which means we can assess parasitism levels in the field.
Based on previous work, we know Gyros can impose selection by affecting survival, growth rate, and reproductive success. We also know Gyros can have an effect on behaviour, body size, and colouration, so it is not unreasonable for one to think parasitism can impose selection on phenotypic traits. However, relatively few of these studies are field studies, so we seek to fill that lack of information by building on previous field-surveys.
Specifically, the question we sought to answer was "whether or not parasitism by Gyrodactylus
leaves a signature on guppy phenotypes in nature," and by phenotypes, we mean things like male colour. As stated above, we know predation is an important selective agent driving guppy colour where HP fish have less colour than LP fish. Based on previous work, we would expect parasitism to have a negative
effect on guppy colour. High rates of parasitism would be expected to
lead to lower levels of colour. Thus, if we jointly consider predation and parasitism, is parasitism as important in driving colour differences as predation? Can parasitism modify our perceptions of the effect of predation on guppy colour? To determine this, we surveyed 26 natural populations of guppies in Trinidad. As cool as the science was, I think everyone enjoyed the break from Canadian winters the most... We sample with butterfly nets. How cool is that?
Here guppy, guppy, guppy
Photo by Kiyoko Gotanda
If you want to know our methods, you can check us out online
for the details. It involves equipment, chemicals, a digital camera, and a lot of audiobooks. Or you can ask in the comments or email me. I won't bite. I promise.
So what did we find? Well, first, we found that when we sample around the same time in two subsequent years, the parasitism levels are pretty consistent, that parasitism levels among populations varied quite a bit, and that parasitism levels were generally higher in HP sites than in LP sites. This is consistent with previous surveys, and we speculate this could be because (1) flooding might sweep infected fish downstream from LP to HP populations, (2) HP guppies like to shoal more than LP guppies, possibly increasing transmission, (3) susceptibility might be higher in HP populations, and (4) HP sites might differ ecologically in ways conducive to parasite infection.
But what about colour? Did parasitism have an effect on male guppy colour? Well, interestingly, it appeared not to. Lab studies have shown an effect of parasitism on guppy colour, but we didn't find any direct effects of parasitism on male colour. In fact, a previous field survey didn't find an association with orange colour and parasitism either. So what gives? We have a couple of ideas on why things might be detected in the lab, but not in nature. First, our field survey was a snapshot survey, so we don't know a given guppy's infection history or what the population parasitism levels were like before or after the survey. We know that Gyro epidemics increase and decrease quite rapidly, so it's possible our measures of parasitism are not a true representation of parasitism levels. Second, lab studies can control for a lot of things such as light and resources. We can't control nature, so there might be variation in other selective agents that prevented us from detecting a parasite signature on colour patterning. Third, mortality of infected or uninfected fish can differ, potentially altering parasite and colour distributions.
So we didn't find a significant effect of parasitism on guppy colour. But what if we throw predation back into the story? Does parasitism modify our conclusions about the effect of predation on colour? When considering predation, our results were congruent with what we expected a priori
: males in LP sites had more colour than HP sites in general. But we had a lot of variation. It wasn't a clear cut story. Some HP sites had more colour than some LP sites. Previous work has also found similar differences and nuances, and from this, we can conclude that predation IS an important selective agent on guppies, but it's a lot more complicated than just predation alone.
We now ask "does the predation regime story benefit from a simultaneous consideration of parasitism?" Well, we didn't find an effect of parasitism on colour, and adding a parasitism term to our models almost never affected our conclusion about the effect of predation. We still must consider the potential drawbacks of field surveys, but based on our data, we find that parasitism does not modify our interpretation of the effects of predation.However, we are not saying parasitism has no effect, but rather that the
signature of an effect of parasitism on male guppy colour might be
swamped by other selective agents (e.g. predation) leaving us unable to
detect a significant effect of parasitism on colour.
So, in a nutshell, we found that parasites vary between populations and were relatively consistent when assessed over two years, higher infection levels are often found at HP sites as compared to LP sites, parasitism doesn't appear to have an effect on male guppy colour, and considering parasites does not alter our current conclusions about predation.
That's all she wrote!
photo by Kiyoko Gotanda
: Gotanda K.M., Delaire L.C., Raeymaekers J.A.M., Pérez-Jvostov F., Dargent F., Bentzen P., Scott M.E., Fussmann G.F. & Hendry A.P. (2013). Adding parasites to the guppy-predation story: insights from field surveys. Oecologia.172(1): 155-166. DOI: 10.1007/s00442-012-2485-7