Parasites and evolution - Andrew MacColl

Ecology has been very successful at exploring certain types of interactions between species within natural systems, while generally ignoring the possibility that such interactions result in natural selection and can evolve. Evolutionary biologists have been very good at demonstrating natural selection and exploring the genetic basis of selected phenotypes, but have tended to do so in isolation from their ecological settings. In the past decade there has been a resurgence of interest in reconnecting these closely related but disjointed perspectives, with a view to better understanding how mechanism (ecology) and process (natural selection) lead to pattern (evolution). The main theme of Andrew MacColl's research is to understand the role of ecology in driving natural selection and how this can produce the divergent evolution that builds into variation between individuals, populations and that ultimately accumulates into speciation. 

                                                        A Schistocephalus solidus tapeworm dissected out of the body cavity of a dead stickleback. These worms can be almost the same weight as their host!

Our approach is based on the integration of theory, observation and experiment. We are particularly interested in the relative importance of different selective agents in directing evolution. In general little is known about whether evolution is driven mainly by the abiotic environment or by ecological interactions such as competition, predation and parasitism. Classical ecology has focussed on the part played by competition between organisms in determining individual success (and hence evolution). Other ecological interactions have received less attention. At present we are particularly interested in the role of parasitism as a driver of host evolution, because parasitism has been poorly studied from this perspective.

Stickleback as a model for the study of evolution

Three-spined sticklebacks (Gasterosteus aculeatus) are a good model species because they are common, widely distributed and easy to keep in the lab. A great deal is known about their natural history and genetics (their genome has been sequenced). They are particularly interesting because they exhibit a great deal of phenotypic and genetic diversity between populations. The photographs below show sticklebacks from different populations on the island of North Uist, Outer Hebrides. These are all lab raised fish of a similar age (7 months). The fish in the middle (a male in breeding condition) is from an anadromous (sea-going) population. Others are from different freshwater populations that have been established on North Uist in the last 10 – 20,000 years. Note that some freshwater populations have shown considerable morphological evolution in this time. Many have lost the (eponymous) dorsal spines and/or the pelvis spines that are present in the anadromous (and probable ancestral) population, as well as exhibiting rather different overall body shapes.

Figure 1. Morphological variation between five populations of three-spined sticklebacks on North Uist in the Outer Hebrides (photographs by Job de Roij).

We use three-spined sticklebacks to address the following kinds of general questions:

Do parasites vary between host populations in ways that are sufficiently substantial and consistent to contribute to host evolution?

Figure 2. Variation in the prevalence of the three commonest macroparasites of sticklebacks (N=20 per loch) in 15 lochs in a 1000 km2 area of North Uist. 

This graph clearly shows that there are large differences in parasitism between host populations even in small geographical areas. Initial results show that these differences are fairly stable across years.

Is the change in selection caused by parasites important when hosts invade novel environments? 

This is equivalent to asking whether hosts become adapted to the parasites in their own population. One way to look at this is to simulate invasions of novel environments, by carrying out transplant experiments. We have introduced lab raised fish into enclosures in lochs and monitored their growth and parasite burdens after a month. The parents of these fish came either from that loch (sympatric) or from a different loch (allopatric). Some results are shown in Figure 3.


                    Introducing sticklebacks into enclosures in a freshwater loch on North Uist.


Figure 3. The relationship between the burden of trematode parasites in the eyes of fish and their length at the end of the experiment. Fish were 17-18mm when the experiment began. The figure shows two things: (i) Allopatric sticklebacks suffer higher parasite burdens than sympatric ones. (ii) The growth of infected allopatric fish is reduced compared to uninfected allopatric fish or to sympatric fish. Taken together these suggest that parasites can be an important agent of selection on hosts.

What kind of host traits are affected by parasite imposed selection?

The data in Figure 3 suggest that parasites could cause selection on life history traits. This possibility has been discussed at length theoretically, but as yet there are few relevant data. There is certainly substantial variation in life history characters between stickleback populations, and this is associated with variation in parasite prevalence (Fig. 4). Some of our current research is directed at determining whether such relationships have a causal basis.


Figure 4. The relationship between prevalence of Gyrodactylus arcuatus and the reproductive investment of female sticklebacks (gonadosomatic index, GSI) across freshwater populations on North Uist (P<0.025).

It is also likely that variation in parasitism between populations should lead to variation in the ability of hosts to resist or tolerate their parasites. This can be investigated by exposing lab raised sticklebacks from different populations to infection with standard dose of parasites in the lab. Limnetic and benthic sticklebacks that occur together in lakes in British Columbia exhibit very different parasite loads in the wild and also have very different resistance to artificial infection with the parasite Diplostomum scudderi (Fig. 5). Crosses between these two species result in hybrids that have the same susceptibility to infection as the limnetic grandparents, suggesting that susceptibility may be dominant and contribute to the poor performance of hybrids in the wild.

We are beginning to explore differences in resistance traits between Scottish populations of sticklebacks, and whether such differences are are involved in trade-offs with life history and mate choice traits.

Figure 5. Numbers of parasites recovered from sticklebacks artificially infected with one or two doses of twenty Diplostomum scudderi. ‘ben’ = benthic, ‘lim’ = limnetic and F2 = second generation hybrids, all lab reared fish originally from Paxton lake, British Columbia.

Can the interaction between hosts and parasites lead to divergence between host populations that results in reproductive isolation between those populations? 

The results shown in figure 5 (above) suggest that hybrids between divergent host populations may have poor resistance to parasites. This is a good example of how parasites could contribute to ecologically determined post-mating reproductive isolation. We are beginning to explore other ways in which parasite imposed selection could contribute to accumulation of reproductive isolation between host populations.

What next?

We are also beginning to explore whether the traits of parasites and hosts coevolve in our study systems, including such questions as:

Does the selection that parasites impose on their hosts result in coevolutionary feedback to the parasites, because of changes in host density or mean host phenotypes?

Does coevolutionary dynamism result in arms races between parasites and hosts that can lead in fundamentally unpredictable directions?