Brood parasitism and coevolution in weaverbirds


As a new graduate student with three or four unfruitful dissertation ideas already down the drain, I sat in the Bird Library flipping through a book on Introduced Birds of the World, looking for an example where a bird had been introduced to a place where natural selection would act very differently compared to its ancestral range. My latest idea for a dissertation was to discover a case of rapid evolution in a wild vertebrate, having been inspired by the findings by Johnston & Selander in the early 1960s that house sparrow plumage had evolved after they were introduced to the United States. Since Robert B. Payne, my advisor, was an expert on African birds and on brood parasitism (a situation where a bird lays eggs in the nest of another bird), I especially kept an eye out for birds that suffer from brood parasitism that were introduced to places without any brood parasites. An estrildid finch was introduced to Bermuda and Hawaii... nice field sites... but its viduid brood parasite in its African native range is not a very nasty enemy; often the parasite and host young grow up together nicely. Weavers, on the other hand, are parasitized by cuckoos, which completely destroy the reproductive attempt of the host if they get their way. Bob showed me a weaver in the Caribbean that was thought to have changed since it was introduced from Africa, though nobody had studied its responses to brood parasitism in Africa. I had also found the species in the introduced birds book, and it was a prime candidate: (1) it had been introduced to two different places, making independent tests possible; (2) it had a documented history of introduction (this is not as common as one would think); (3) it was subject to nasty brood parasitism in Africa, against which it was thought to have evolved a range of defenses; and luckily (4) there was a lot known about its life history and breeding biology, thanks to UCLA biologists Nicholas and Elsie Collias who had studied them for years in the 1950s-1970s. The fact that the ensuing project succeeded is the main reason why I developed interests and pursued further questions in trait evolution, bird egg coloration, "selection knock-outs", and coevolution. Here I describe this project, briefly mention some follow-up studies, and then summarize my future plans for a long-term multi-species project on coevolutionary dynamics and trait evolution.


Weaverbird ecology

Related to sparrows and finches are a group of mostly African and Asian birds called weaverbirds (family Ploceidae). In most birds, females build the nests. In weavers, however, such as the African village weaver Ploceus cucullatus pictured here, males build the nests. The South African form is on the left, and the West African on the right. The globular weaver nests are remarkable feats of engineering, hence the name of these birds. Males usually defend a territory of several nests on a single branch or palm frond (middle picture).

Then they attempt to attract females by hanging upside-down from the bottom of a nest, and flapping their wings as they swing side to side. When a female (pictured right) accepts a male and his nest, which is at least partly on the basis of its quality and newness of construction, she lines the interior of the nest and mates with the male, and then lays her eggs. The male ceases to enter or hang from the nest, but instead guards it from intruders, and continues to build other nests and attempt to attract females to them.

The diederik cuckoo (Chrysococcyx caprius), however, lurks nearby to wait for the female to leave her nest of eggs. When the weaver mother does take a break from incubating, the female cuckoo stealthily approaches and enters the nest. She removes a weaver egg, lays one of her own, and departs. Unless the weaver female throws it out, the cuckoo egg will hatch first, and before the cuckoo nestling even opens its eyes it will dump any weaver eggs out of the nest by means of a specialized depression on its back. Then the weaver mother, who never rejects a nestling, will end up wasting another month raising the cuckoo to independence. Moreover, cuckoo eggs have evolved to mimic weaver eggs, so it is not necessarily an easy task for a weaver female to discover that she has been parasitized.


Egg appearance and egg rejection as defenses against brood parasitism

Village weavers are known to lay eggs of among the broadest variety of colors and spotting patterns of any bird, but an individual female appears to lay eggs that are consistent in appearance throughout her life (see picture to right; each letter indicates a clutch of an individual female in The Gambia). With my wife April I measured weaver eggs in West and South Africa and confirmed that egg color and spotting pattern vary widely between females, but the eggs in an individual's clutch are surprisingly similar to each other compared to those of other birds. Perhaps weavers have evolved such that an individual's eggs are distinctive enough that she can stand a chance against the cuckoo. The more similar a female's eggs are to each other, and the more different they are from her neighbors', the narrower the window of opportunity for a would-be egg counterfeiter. This is because cuckoo eggs mimic weaver eggs in general, but they cannot mimic a particular individual weaver's egg: mimicry is honed through evolution, and egg appearance changes only between generations and not within a cuckoo's lifetime.

Thirty years before, captive weavers had been found to reject foreign eggs placed into their nest, so we played the part of the cuckoo in Africa and replaced weaver eggs with foreign eggs of known difference in color, spotting, size, and shape. The result was that a weaver female will eject a foreign egg from her nest in proportion to the difference in color and spotting, but not size and shape, from her own eggs. This supports the hypothesis that egg appearance functions as a defense against brood parasitism. I did not find the size or shape of weaver eggs to be consistent within a female or very different between females, so these would have been unreliable cues of identity. In fact, egg spotting in this species would not function in camouflaging the egg from predators since the nests are completely enclosed, so even the existence of spotting in this group of birds may have originally spread because of egg mimicry by cuckoos.


Species introduction provides a test of adaptation

Introductions of species into new habitats provide near-natural experiments in which to document the effects of ecological change on trait evolution. The village weaver population I studied in eastern South Africa was the source of an introduction to the island of Mauritius a century earlier. Another population we studied in West Africa was the source of an introduction over two centuries earlier to the Caribbean Island of Hispaniola. There are no cuckoos or other egg-mimicking brood parasites on either island. Therefore if the traits of egg color and spotting pattern similarity within a clutch, distinctiveness from the eggs of one's neighbors, egg spotting in general, and egg recognition and rejection are all defenses against brood parasitism, they would not serve these functions in the absence of the cuckoo. Weavers might parasitize each other, of course, although theoretical and empirical considerations (such as the fact that the same individual could be both parasite and host) render defenses unlikely to arise in response to other members of the same species. To test this and the importance of cuckoo parasitism to maintaining these traits of the weaver, we measured eggs in the two introduced populations as well as the source populations, and experimentally parasitized nests in all four populations as well. I found that even after such brief spans of time (approximately 75 and 150 generations), egg color patterns evolved rapidly:  both population-level variation and within-clutch consistency decreased in the absence of parasitic cuckoos, resulting in poorer egg recognition in experimental tests of those populations.  However, perceptual and cognitive abilities required to recognize eggs have remained intact in the introduced populations.  This rapid evolution provides strong support for the hypothesis that these traits are counteradaptations to brood parasitism, and illuminates an evolutionary arms race in Africa between cuckoos and weavers.

Introduction history of the village weaver. The yellow is the ancestral range, where the diederik cuckoo is an egg-mimicking brood parasite. The red is the range to which humans have introduced these birds. The islands do not contain egg-mimicking brood parasites. Population variation in village weaver egg color declined following escape from cuckoos (left) and variation in egg color within a clutch increased dramatically (right), both in proportion to the time since the birds were introduced. Spotting variation changed in the same way (not shown). The arrows represent introductions, and the asterisks above the bars indicate the degree of statistical significance.

In follow-up research I have conducted further studies of the ecology of invasive species, the functional significance of egg appearance (see my egg color research page), patterns of behavioral evolution and trait loss, and the genetic signature of these introductions including the role of founder effects. I look forward to the discovery of candidate genes for egg color so that I can perform additional, more detailed analyses of the evolutionary processes I have inferred in the village weaver.  I anticipate that my results will provide further insights into short-term evolution, adaptation, and the interaction between founder events and selection during evolution on islands.


Dynamics of brood parasite-host coevolution

Species that are coevolving with each other provide some of nature’s most exquisite examples of adaptation, and the important sources of selection are often easily identified.  Brood parasite - host interactions in birds provide particular benefits to the study of reciprocal adaptations in that the host and parasite are readily observable and have similar life-histories and generation times, significant selection pressures are imposed by the parasite on the host, and evolutionary responses between the species are multilevel and ongoing.

The same cuckoo species that parasitizes the village weaver also parasitizes several other members of the weaver family (Ploceidae) in Africa.  These weavers appear to vary in the traits that determine whether cuckoo parasitism succeeds, including egg color and spotting; they probably vary in the tendency to reject eggs as well. Thus each host is at a different point in coevolution with the diederik cuckoo-- some might be recently parasitized and rapidly gaining defenses, whereas others might be seasoned hosts with well-developed defenses. Still other weaver species might have been hosts in the past but were abandoned by the cuckoo at some point.

We know very little about this system so far. The figure to the left plots a provisional molecular phylogeny of the true weavers, subfamily Ploceinae, which contains several known or potential hosts of the diederik cuckoo (four species which were not involved in this phylogeny are set off to the upper right). The color or gradient represented by each bar roughly approximates the egg colors of that species, and the presence of a dot before the species name indicates that at least some of that species' eggs are spotted. If the bar has a cuckoo head attached to it, it is a known significant diederik cuckoo host. Obviously a great deal of empirical and phylogenetic work is needed to fill this figure out, and we can't be certain even about the species for which we have some information. So far, however, it looks like all cuckoo hosts in the subfamily lay variable eggs with spots. But why do P. bicolor, xanthopeterus, and rubiginosus have variable eggs? Why has P. xanthopterus ventured into color space that the other species have not? How good are all of these species at rejecting eggs? And what about all of the other weaver species?

I plan to apply both experimental and comparative methods to investigate phylogenetic relatedness, cuckoo parasitism rates, and antiparasite adaptations in all weavers that are known or potential diederik cuckoo hosts. I can then treat host traits as multiple data points to track patterns in host evolution. Most studies of evolution can only examine traits at one slice in time, but with a system like this with one parasite and several hosts, we can infer long-term evolutionary trajectories.

To proceed with this project I am presently planning a collaboration with Mike Sorenson of Boston University, Staffan Andersson at the University of Gotheborg in Sweden, and Robert B. Payne of the University of Michigan, with a threefold objective:

(1) Produce a complete molecular phylogeny of the weaver family Ploceidae

(2) Acquire data on brood parasitism frequency, host egg variability, and behavioral responses to parasitism. During my next field season with weavers (planned for 2010, NSF willing), I will expand my field methods to extract albumen from live eggs in order to assess maternity by protein electrophoresis.  These data will help me to calculate rates not only of cuckoo parasitism but also of intraspecific parasitism. 

(3) Incorporate these results along with existing experimental data into a comprehensive picture of host adaptation and evolution in this group, and also a fully parameterized and generalizable model of coevolution.