By Rex Graham
Cheaters never win!
Well, that adage is not true in nature. One avian cheater you may love to hate is the Common Cuckoo.
The cuckoo, whose plaintive call is copied by namesake clockmakers, is itself a clever counterfeiter par excellence.
It has evolved a specialized and highly tunable process that allows it to make egg patterns that trick host bird species into accepting counterfeit eggs. If “brood parasite” cuckoo eggs go undetected and hatch, the foster-parent hosts don’t discriminate against parasitic chicks, rearing them as if they were their own.
However, new research has revealed details of two countermeasures host birds have acquired as part of an escalating co-evolutionary arms race with egg-laying cuckoo cheaters.
- To varying degrees, birds targeted by the Common Cuckoo have evolved egg-patterns that are more difficult for cuckoos to counterfeit.
- Host birds with effective egg patterns also have evolved a human-like, pattern-recognition process to compare their egg patterns with those of counterfeit eggs surreptitiously laid in their nests by the cuckoos. Some species accurately identify and reject almost 100 percent of counterfeit eggs.
“Human brains are quite different from bird brains, but the processes we use for pattern recognition are thought to be similar,” said Mary Caswell Stoddard, a researcher at Harvard University’s Museum of Comparative Zoology. “In that sense, the way in which a bird recognizes its own egg patterns may indeed be similar to the way in which humans recognize faces.”
Computer modeling counterfeit eggs
In a paper published June 18, 2014, in Nature Communications, Stoddard, zoologist Rebecca Kilner and computer scientist Christopher Town, both from the University of Cambridge, described how a computer tool designed for fingerprint and facial pattern recognition enabled them to understand how nesting birds identify counterfeit eggs in their nests laid by the Common Cuckoo.
“This study suggests that, at least in these hosts of the Common Cuckoo, the species that have been most fiercely mimicked have evolved the best egg-pattern signature and are kicking out differing eggs more effectively,” said David Lahti, an assistant professor of biology at Queens College at the City University of New York. “That result suggests that, despite cuckoo evolution to produce eggs that blend in with host eggs, the hosts are keeping up, perhaps even winning.”
Brood parasitism is found in birds, fish and insects. About 59 cuckoo species worldwide are brood parasites. Bronze- and Brown-headed Cowbirds are the only brood parasite bird species in North America, laying eggs in the nests of as many as 200 species of host birds.
Brood parasitism – stealth required
Common Cuckoos are known for their sneaky ways. They approach temporarily unoccupied nest of a potential host and quickly lay an egg. When the host bird returns, they may find a nest full of eggs — all strikingly similar in size, shape, color and pattern.
Individual races of Common Cuckoos specialize on one host species. For example, one race of females parasitizes only the nests of Meadow Pipits. Others seek out only Reed Warbler nests. Still another stalks the nests of Bramblings.
Cuckoos are known to plunder nests too advanced for one more new egg. Often the hosts re-lay, giving cuckoo females another opportunity. Cuckoo eggs hatch before host eggs and the new chicks evict hosts eggs and nestlings.
Size matters – not
In her team’s study, Stoddard focused on 689 total eggs laid by eight host bird species and their respective counterfeiting cuckoos. The eggs from each nest, including the cuckoo eggs, were studied as groups.
“Eggs laid by the Common Cuckoo are sometimes slightly larger than those laid by their hosts,” Stoddard said. “Surprisingly, the size of the egg does not seem to tip off the hosts, who appear to pay more attention to the egg’s color and pattern.”
Stoddard’s team used a computer program to mathematically score the blotches, speckles and other markings on each egg. No two were identical.
Two host birds, the Great Reed Warbler and Pied Wagtail, produce eggs that are densely patterned, seemingly impossible to counterfeit. In a surprise, however the computer program found that those markings yield eggs that don’t stand out against those laid by cuckoos.
“This concept — that increased complexity can actually result in reduced accuracy of a predictive model — is fundamental to information theory and machine learning, yet relatively new to biological systems,” the three authors wrote in the Nature Communications paper.
Meadow Pipit eggs, which have a dense, highly complex sandpaper-like pattern of small speckles, have the lowest spatial dispersion. That dense pattern is little help to those birds, which are poor at recognizing and rejecting similarly patterned cuckoo eggs.
Brambling eggs, with an intermediate patterning of widely separated blotches, had the highest spatial dispersion of markings. Indeed, Bramblings are very proficient at identifying and rejecting cuckoo counterfeits.
Accepted bird pattern-recognition idea challenged
The computer program used in the study found that intermediate pattern complexity was a feature of the most effective egg pattern signatures evolved by hosts.
“Perhaps we should have known this, but we didn’t, and Stoddard, Kilner, and Town show it beautifully,” said Lahti. “There is no Scotland Yard in the bird world, but this study suggests that the hosts of the Common Cuckoo that have been most fiercely mimicked are the ones that have evolved the best signature and are kicking out counterfeit eggs more effectively.”
Lahti, who was not involved in Stoddard’s project, said even with a small chance of tricking its hosts into accepting counterfeit eggs, Common Cuckoos have good reason to keep trying. “If just a slight likelihood of success remains, a counterfeiter can still win by trying again and again,” he said. “In this case, she’s got everything to gain and nothing to lose but the cost of making an egg.”
Avian pattern recognition is poorly understood. So Stoddard’s team used the NaturePatternMatch computer application to evaluate egg patterns in a way that is neurobiologically plausible for primates and birds. Future studies will be aimed at understanding which biologically inspired models of object recognition are best at approaching the capabilities of the avian visual system.