Feature

The Louse that Missed the Boat

Lice can be surprisingly informative...

Dr Adrian Paterson

If there was ever a biological Olympics then the gold medal for most successful lifestyle would surely go to parasitism. We know that there are a lot of free-living species, but for every bug, beetle, bat and barley species there are always several parasite species (like nematodes, flukes, fleas and even copepods) willing to make a living off that host.

Perhaps the most basic question that we can ask about a parasite species is why it is present on a particular host species and absent from another?

For example, the chewing louse genus Halipeurus, a group of insects that live and forage amongst birds' feathers, is hosted by bird species of the order Procellariiformes (seabirds, such as storm petrels, petrels, shearwaters and diving petrels). But Halipeurus species are not found on all Procellariiformes, such as albatrosses, fulmarines and Procellaria petrels.

How has this particular distribution of louse species come about? It is important to realise that the absence of Halipeurus from a given bird species also requires some explanation. Either the given bird species or its ancestral lineage has never been colonised by Halipeurus, or it has been colonised but the parasite has been subsequently lost by what are termed "sorting events".

There are two basic scenarios. First, the ancestor of the seabirds possessed a Halipeurus species and passed it on to its offspring (along with its genes, habitat, behaviour and so on) where it continued to speciate as the host speciated. Second, Halipeurus was present on the procellariid petrels and colonised or host-switched to the storm petrels some time in the past.

It appears that the apparently straightforward looking question of why we find a parasite species on some hosts and not others is more complex than first thought. How, then, do we go about answering this question?

Evolutionary Hints

We begin by assuming that the evolutionary history of host and parasite match; as hosts become new species their parasites will also become new species (cospeciation). Of course the path of evolution seldom runs this smoothly and a number of events will affect this pattern.

First, parasite species can speciate within the host (intrahost speciation). Lice appear able to speciate in response to new feather niches developed by their host. For instance, a species may specialise for life in short head feathers while another species may adapt to long wing feathers. Second, the host may speciate without the parasites speciating. Third, parasites may host-switch. Fourth, parasites may be lost from a host species by two types of sorting events. When a host speciates there may be such a small louse population present that it is unsustainable and goes extinct. Alternatively, the few hosts that found a new species may not contain a particular louse species (they "miss the boat" ).

So how do we find out about the evolutionary history of Halipeurus and the seabirds? We begin by obtaining the evolutionary relationships of the seabird and Halipeurus species based on genetic data. Seabird and louse evolutionary trees do not exactly match. A TreeMap analysis maps the louse tree onto the host tree and indicates the likely evolutionary events that have occurred. The analysis indicates that there have been three cospeciation events, one host-switching event and one sorting event. It is all very well predicting that these events have occurred, but do we have any evidence that they actually have occurred?

To answer this question we have to make a detour through the idea of a molecular clock. Over time there are random mutations of DNA sequence which may occur at a particular rate. By looking at how much change we observe between the DNA sequences of two species we obtain a relative estimate of their time of divergence from one another. For example, if we know that a stretch of 12S ribosomal RNA sequence changes at one base pair every million years and we know that there are three base pair changes between species one and two, then they probably diverged about three million years ago.

Testing History

Our predictions of the evolutionary history of seabirds and Halipeurus are tested by calculating the genetic distance between the seabird species and also between the louse species. The three cospeciation events are supported, as the genetic distances were similar for lice and their seabird hosts, indicating simultaneous speciation. The host-switching episode is supported, as the genetic distance between the two Halipeurus species is much less than that of their storm petrel and mottled petrel hosts, indicating that the two louse species diverged much more recently than their hosts.

Finally there is the sorting event. How can we test something that is absent? Well we can not, not directly. The best that we can do is to look at the likelihood of sorting events occurring.

There are at least two sources of data which indicate that sorting events are likely. We assume that lice are distributed within a species such that most birds have a few parasites and a few birds have a lot of lice. This is known as a negative binomial distribution and is common in parasite populations. It is likely, therefore, that a small group of founding birds, drawn randomly from a larger population, that go on to establish a new species will either not have some parasite species or only have them in very low population numbers. The limited studies on seabirds have shown a negative binomial distribution of lice.

The second type of data on sorting events is obtained by comparing the numbers of parasite taxa present on daughter taxa with those present on their parent taxa. For example, many ectoparasite species inhabit house sparrows (Passer domesticus). Sparrows were introduced into North America from Europe and have only 35 of the original 69 ectoparasite species present in the European population. The house sparrow was also released into New Zealand from Britain, and New Zealand sparrows possess only two of the five louse species found in their parent population.

Another example is that of the Chatham Island Oystercatcher, which has two louse species, and its probable parent species the South Island Pied Oystercatcher, which has three louse species. In fact, of 48 New Zealand species pairs that can be clearly identified as parent and daughter, 36 of the daughter species have fewer louse species than their parent species, indicating that extinction and "missing the boat" events are reasonably common.

Overall, there is good support for our predicted events. The evolutionary history of the Halipeurus has been one of cospeciation, with its absence from Westland Petrels due to a sorting event.

The superficially simple question of why are some parasites present on one host and not on another has been shown to complex in the extreme. We have had to unravel the routes that louse and seabirds have taken through a complex, entwined evolutionary history. It has required the use of two centuries' worth of distribution information, molecular data, complex programs on powerful computers and several new concepts.

We have not even asked ecological questions like which conditions are necessary for successful host colonisation, why some parasite lifecycles provide more opportunities for speciation or which host defences remove parasite species? That, as they say, is another story.

Adrian Paterson and Brent Emerson are in the Department of Entomology and Animal Ecology at Lincoln University.