Europeans became obsessed with the Paradisaeidae after skins of the Lesser and Greater Birds-of-Paradise were brought home from the Moluccas in 1522, by the surviving crew of Ferdinand Magellan’s expedition.1 Indeed, by the end of the 1540s, mounted birds-of-paradise were a ‘common site in the cabinets of Europe and Turkey’, according to Pierre Belon’s L’Histoire de la Nature des Oyseaux (1555).2 Naturalists and collectors of the day were beguiled by the bird’s strange anatomy, for the legs had been removed by local traders who mistakenly believed that Westerners were only interested in the birds’ colourful plumages. The bizarre conjecture that the birds lacked feet resulted in the notion that the birds spent their entire lives in perpetual flight, living off the ether’s dew, only to grace terra firma on death. It was their supposed ethereal existence, between heaven and earth, that earned the birds their befitting name, ‘birds-of-paradise’. Others, only slightly less gullible, knew they were being fooled and believed that the specimens had been cobbled together from different animal parts by unscrupulous dealers.3 Today, of course, we know better.
Plate tectonics and vicariant speciation
The Paradisaeidae originated approximately 24 million years ago among the string of small islands and archipelagos that emerged in the seas to the north of Australia, in the area of Papua-Melanesia.4 The 41 currently recognised species are core members of a large assemblage of crow-like passerines. Remarkably, despite being so utterly different, they have all descended from a common ancestor, one that probably resembled the drably plumaged Paradise-crow, a species restricted today to Halmahera and the islands of the North Moluccas. The fact that such a nondescript ancestor could have undergone such unparalleled diversity is the result of New Guinea’s unique combination of tectonic activity and abundant food supplies.
New Guinea’s landmass, the second-largest island in the world after Greenland, was never part of any continent but was created in recent geological times by the relentless forces of plate tectonics. As the northerly-moving Australian plate collided with the Pacific plate, its leading edge buckled, rising out of the sea to form the southern half of New Guinea, an area that includes most of the island’s central mountains. To the north, a sequence of volcanic island arcs, formed by rising magma between the two subducting plates, was carried southwards as if on a geological conveyor belt. Eventually, these islands collided and docked (‘accreted’ in geological jargon) with the newly formed New Guinean landmass to form the northern part of the present-day island. Once fused, the 30 or so fragments of continental crust added to the rising central mountains, as well as forming the Huon Peninsula, and the Foja and Adelbert ranges in the north. This convergence process is ongoing, and several million years into the future the island of New Britain will have been added to New Guinea’s mainland as the Australian plate bulldozes its way ever northwards.
It was against the backdrop of extreme geological turmoil, amid the jostling of volcanic island arcs and fast-growing mountains, that the birds-of-paradise first made their appearance. However, in contrast to their closest relatives, the monarch flycatchers, crows and drongos which dispersed to other continents and remote oceanic islands, the birds-of-paradise did not venture far and remained in the vicinity of New Guinea. The most distantly scattered birds were the forebears of the Paradise-crow, the oldest of the bird-of-paradise lineage, one that diverged 17 million years ago.4 Today the Paradise-crow is found on the islands of the North Moluccas, some 300 kilometres to the west of New Guinea, although, interestingly, they did not fly there, but were carried. It is now known that Halmahera and the adjacent islands used to be part of mainland New Guinea. As the Pacific plate sheared past the Australian plate, fragments of the continental margin detached, creating islands that spilled out into the surrounding ocean,5 carrying with them their cargo of ancestral birds-of-paradise.
Similar geological processes underpin the evolution of two highly ornate sickletails from the genus Diphyllodes, the Wilson’s Bird-of-Paradise and the Magnificent Bird-of-Paradise (Plate 26). Currently, the two species are separated by the narrowest of barriers, the 3-kilometre-wide Sagewin Strait, although this was not always the case. Millions of years ago, the clash of tectonic forces prised a fragment of land, together with its population of early sickletails, from the western tip of New Guinea. As the newly formed island drifted away, the isolated population evolved into one of the most striking species on Earth, the Wilson’s Bird-of-Paradise. The male presents a riotous palette: vivid yellow nape, blue skullcap, crimson back and coverts, iridescent emerald-green breast and shiny curlicue tail feathers. Not surprisingly, Wilson’s Bird-of-Paradise is regarded by many birders as their most sought-after species. Indeed, the ornithologists Tim Laman and Edwin Scholes have described the species as ‘a paragon of the endless forms for which the birds-of-paradise are renowned. Their extremes in size, shape, color and behavior seem to have been literally embodied in the males of this one species.’6 Later, as a result of continuing geological forces, the island split and a 450-square-kilometre splinter, the Indonesian island of Batanta, headed back towards the mainland. Despite this splinter approaching to within 3 kilometres of the larger island of Salawati, the population of Wilson’s Bird-of-Paradise has not been able to cross the strait. As a result, Batanta’s birds remain isolated from their closest relatives on Salawati and the New Guinea mainland, the Magnificent Bird-of-Paradise. The Sagewin Strait must surely be one of the smallest geographical barriers in the avian world to have resulted in vicariant speciation.
According to the most recent phylogeny, the Paradise-crow’s nearest kin is the bizarrely vocal Trumpet Manucode (genus Phonygammus), a species that evolved approximately 14 million years ago.4 Four million years later, the closely related genus Manucodia diverged, forming an assemblage of medium-sized birds, with glossy black to purple-green plumages. Manucodes, in general, are widespread and inhabit the mainland’s lowland forests, although two species, the Trumpet Manucode and Glossy-mantled Manucode, are found in the Aru Islands, located to the southwest. Both species may have flown there, but it is now thought more likely that they reached the islands during the Pleistocene ice age when the sea levels fell, and a land bridge formed connecting Aru to New Guinea. The adjacent Kai and Tanimbar Islands are surrounded by much deeper water and, since no land bridges ever formed, birds-of-paradise have been unable to gain a foothold on these otherwise ecologically similar islands. Another example of a land-bridge effect is seen in the Magnificent Riflebird, which inhabits both New Guinea and the rainforests of Cape York in Australia. When sea levels rose and the Torres Strait formed, 50 thousand years ago, Magnificent Riflebirds were left on either side of the divide. Though still considered a single species, the two populations are now evolving separately, since the birds are reluctant to fly across the open water.
The formation of New Guinea’s fragmented and fast-growing cordilleras, which now stretch from east to west, not only divided the ancestral birds-of-paradise but also provided the opportunity for altitudinal speciation. The central mountains, for example, form a well-defined boundary between the low-lying habitats of the Lesser Bird-of-Paradise to the north, and those of the Greater Bird-of-Paradise to the south. In contrast, the sicklebills, which diverged around 12 million years ago, became restricted by altitude in the central mountains – the Brown Sicklebill living above 2,000 metres while the Black Sicklebill inhabits the forests below. Other birds-of-paradise, species that evolved on the volcanic islands after they had fused to the mainland, have highly restricted distributions since they are surrounded by vast lowland basins of unsuitable habitat. The rarely seen Bronze Parotia, for example, is confined to the remote Foja Mountains in West Papua, a fact that delayed its identification as a new species until 2005. The accreted mountains of the Huon Peninsula, lying to the northeast, possess a number of endemic species, including the Emperor Bird-of-Paradise (Plate 27). This species, adorned with unusual flank plumes and iridescent green head and throat, diverged 6 million years ago from the rest of its genus, all of which remain in New Guinea’s lowland areas.
But geology alone cannot account for the bewildering diversity in appearance and behaviour that characterise the Paradisaeidae. Their extravagant plumages, complexity of colours, piercing calls and minutely choreographed displays all scream one thing: female choice, or ‘sexual selection’.
Female choice, male handicap
It was Charles Darwin’s realisation that nature does not select, individuals do, which led to the concept of sexual selection.7 He argued that while natural selection was concerned with staying alive, sexual selection was about attracting mates, a preference unconstrained by the functional laws that govern evolutionary form. Although his classic example was the flamboyant tail of the peacock, Darwin’s theory applies equally well to other species, including the birds-of-paradise. For, like the peahen, the female bird-of-paradise’s choice of mate appears to be quite arbitrary, limited only by the individual’s sensory and cognitive abilities. So, while the males’ long tails, wiry end-feathers or elongated head-plumes may seem arbitrary or just unimportant to us, they are everything to the female. And it is because such preferences can operate outside the limitations imposed by Darwin’s ‘survival of the fittest’ that the absurd ornamentations and behaviours of male birds-of-paradise have evolved. This view, however, was not widely accepted at first.
In the years following his seminal publication The Descent of Man, and Selection in Relation to Sex (1871), Darwin’s ideas were deemed too controversial, and his theory fell into disrepute. Indeed, most biologists, including Alfred Russel Wallace, hated the apparent randomness of Darwinian sexual selection and believed that female choice of costly ornamentation was incompatible with natural selection. Instead, the Wallacean view emphasised that any ornamentation or dexterity in display must co-exist with superiority in utilitarian qualities. A notable exception in the early twentieth century, however, was the British geneticist and statistician Ronald Fisher, who used mathematical reasoning to resolve the paradox and showed for the first time how ‘beauty for beauty’s sake’ might come about.8 In essence, Fisher realised that there must be a genetic basis for both female preference and male ornamentation, and that if both coevolved then runaway selection for a particular trait could result. The increase in both the selected trait and the female preference for it would continue ‘exponentially, or in geometric progression’ until practical physical constraints limited any further exaggeration.
To better understand the concept of ‘Fisherian runaway’, let us consider the case of the ornamental head-wires of the King of Saxony Bird-of-Paradise (Plate 28). This iconic highland species is one of the most bizarrely adorned on Earth, with the male boasting a pair of half-metre-long ornamental head-plumes – head-wires, or flags – that are unlike any other known feather. Each one, emblazoned with 40–50 sky-blue pennants, projects from its head and can be waved seductively in front of prospective females.
Now imagine, if you will, the ancestral population of King of Saxony Birds-of-Paradise, somewhere in the central highlands of New Guinea, and suppose that one of the males has developed two small protruding head feathers, rooted behind each eye. Since this aberrant trait is the result of a random mutation, possibly one base change in a billion, it will be passed on to the next generation. If its possession is neutral and does not affect the bird’s survival, it may, in time, become more widespread within the population. Now imagine that the females of the species, at some later point, develop a genetic preference for males with small protruding head feathers. The reasons can be arbitrary or even whimsical – maybe the ornamentation just looks attractive or ‘sexy’ to them. Since there will be a cohort of males with head feathers, the females are likely to choose the ones with the most prominent plumes. Crucially, and this is the nub of Fisher’s theory, the resulting progeny will possess two sets of genes, one for longer head feathers and one for their preference. Of course, only the genes coding for head feathers will be expressed in males, while the preference genes will only manifest in females. Furthermore, since males with the longest head feathers will have the greatest success in attracting females, they will father more offspring, and the long plumes will become more common in subsequent generations. The important point is that female King of Saxony birds will tend to favour head-wires that are slightly longer than average, however long that average has become. But this raises the stakes and leads to a phenotypic ‘arms race’. In the future, to stand out from the competition, a male would have to have even longer feathers, or some additional feature to gain the attention of a mate. The feathers, for example, could develop unilateral vanes or be adorned with odd looking projections or be wafted enticingly about the head.
I think you can see where this discussion is heading. If females evolved a preference for feathers with all three modifications, then over evolutionary time male King of Saxony Birds-of-Paradise would sport long, mobile and highly ornate head-wires, just as they do today. In the words of Richard Dawkins, the male’s bizarre ‘flags’ can be viewed as the ‘end-product of explosive, spiralling evolution by positive feedback.’9 But female choice can be fickle. The preference for today’s head-wires could just as easily change in the future for no other reason than it’s what females have come to prefer.
However, Darwin’s idea of ‘aesthetic evolution’ by mate choice remains, even today, an anathema to most scientists.10 Instead, many theorists prefer alternative explanations: neo-Wallacean ideas in which female choice has a beneficial genetic effect for her progeny. It may be, for example, that the burdensome appendages are a male’s means of advertising his health, one free of disease and infection. In effect, this means that females favour traits that are disadvantageous for males. Such an idea was championed by the charismatic and energetic Israeli biologist Amotz Zahavi, who proposed what has become known as the ‘handicap principle’.11 Essentially, Zahavi, who died in May 2017, regarded all sexual signals as having a cost, a downside for their owner. In the case of the male King of Saxony Bird-of-Paradise, the head-wires indicate not only underlying vigour but also that their possession is consuming considerable resources. In other words, the male birds are saying ‘I have survived in spite of these enormous head-wires, and I am, for this reason, fitter and a better catch than all the other males.’ As long as the advantages of ‘healthy’ genes outweigh the cost of the handicap, then the benefit to the offspring will be greater than for the progeny that result from random mating.
I can fully appreciate Zahavi’s reasoning, after observing the remarkable display flight of Africa’s Long-tailed Widowbird, a member of the weaver family (Ploceidae). During the breeding season, the male undergoes a remarkable Cinderella-like transformation, changing from an inconspicuous brown individual to an astounding bird with jet-black plumage and up to eight tail feathers of nearly half a metre in length (more than twice its body length). He also develops orange and white epaulettes (shoulders) and a bluish bill. Once the male’s new outfit is complete, he displays during the breeding season by flying just above the bushveld’s grasses, keeping his wings above his body, beating them slowly and erratically. At the same time, he trails his absurdly long tail feathers vertically, fluttering in the wind, in the form of a deep, long keel. This performance, which can be seen by females from a great distance, is clearly energy-sapping, and the male gives the impression of relief as he finally drops to the ground or clings to a prominent vantage point. The possession of such long feathers is obviously not beneficial for the male, as it makes flying more difficult while making it easier for predators such as falcons to pick him off. Even on the ground, he appears encumbered and drags his appendages through the grass while feeding. Indeed, just to produce the new complement of feathers every year involves a substantial investment of metabolic resources.
Nevertheless, despite these obvious disadvantages, there appears to be no limit to the female’s desire for long tails. In a classic experiment by Malte Andersson, the tail length of several males was manipulated by cutting off tail feathers from some individuals and gluing them onto others. Amazingly, Andersson’s results revealed that females preferred males with longer tails than those that occur in the natural setting.12 These findings are often quoted in support of Zahavi’s ‘handicap principle’, because only the fittest males will be able to produce the largest tails and survive long enough to pass on their ‘good’ genes. However, Zahavi’s idea, as originally presented, was not just vague but flawed, and it took the input of Oxford mathematician Alan Grafen to rescue the theory by showing that such systems can be evolutionarily stable.13 As a result, the aesthetic versus good-gene debate – the Darwinian versus neo-Wallacean polarisation – continues to this day.
The courtship displays of birds-of-paradise can be astonishingly complex, and for the male King of Saxony Bird-of-Paradise, head-wire waving is merely the opening act of a highly ritualised performance.14 If a female shows interest, the male then reveals his mantle cape and breast shield feathers, after which he makes a descending invitation flight to a nearby springy vine, where an even more complex display takes place. Here, in the understorey, he bounces vigorously to vibrate the vine, which in turn bounces the female perched above. This unique bouncing show, with all feathers displayed, is followed by a dance consisting of upper-body rotation and side-to-side waggling of the head to cause the head-wires to swirl about. Each element of the male’s elaborate and bizarre courtship has coevolved as the result of the arbitrary whims of female choice, a fact best understood in terms of the Fisherian runaway model, discussed above. But can such courtship displays tell us anything about the evolutionary relationships of the birds-of-paradise?
According to the American ornithologist Edwin Scholes, they can. Scholes believes that the complex dance routines can be broken down into a series of smaller individual units, and that these can be combined to create a single choreographed piece. As a result, males can create novel performances by just reorganising and modifying the individual steps. Scholes also believes that it is their modularity that underscores courtship evolution, since by simply rearranging existing components, new dances can evolve quickly. If females prefer these new combinations, then these courtships will eventually become standard behaviour for that species. The dances’ modularity may also help chart the evolution of the Paradisaeidae, since, once the relationship between the dance elements is known, then the phylogeny of the various species could be deduced. Indeed, this approach suggested that the King of Saxony Bird-of-Paradise is most closely related to the genus Parotia, a fact confirmed recently by molecular genetics.4
At first sight, manucodes appear to be the most normal of all the birds-of-paradise, given that both sexes look the same and that they live together in pairs with the males helping to raise the young. But sexual selection has created a hidden secret, one that cannot be seen, but which can only be heard. For male manucodes have evolved a greatly elongated windpipe that coils around the abdomen, just beneath the skin, before curling back to the lungs (Plate 29). Their resulting low-pitched calls penetrate the dense forest, helping to keep pairs in contact and to defend their territories from rivals. The most unusual vocalisations belong to the Curl-crested Manucode, a species restricted to the forests of several islands lying off the east coast of New Guinea. Their evocative, mellow, fluting calls are a characteristic sound of the islands’ interiors, a vocalisation that has been honed by aeons of female selection
But why have the world’s most extraordinary birds evolved in New Guinea, and why have they not dispersed to other nearby regions? In part, it relates to the island’s lack of large predators. New Guinea was never connected to Asia, and its mammals have either had to fly across or make their way from Australia when sea levels were much lower. Over time, the island’s dense forests have favoured smaller mammals – echidna, tree kangaroos and bats – most of which are insectivorous or herbivorous and pose no threat to birds. However, the basic explanation relates to the birds’ diet of fruit. Unlike the manucodes, most of the ‘showy’ Paradisaeidae feed on complex fruits, berries and capsular fruits that are widespread, highly nutritious and available throughout the year in the lush rainforests. As a result, the birds do not have to spend much time and energy foraging and can devote more effort to matters relating to sex. In The Bowerbird’s Story we saw how a frugivorous lifestyle is linked to elaborate bower-building and courtship rituals. A similar diet, in the same part of the world, has enabled the birds-of-paradise also to evolve a ‘many-female’, or polygynous, breeding system. For polygyny requires females to spend lengthy periods observing intricate courtship displays to ensure they pick the ‘best’ male. Females also have to raise their families by themselves, as the males spend most of their day tending display sites and courting other females. Even adolescent males practise their complex courtship routines to prepare themselves for the rigours of the female selection process to come. So, without New Guinea’s abundance of all-year-round high-energy fruit, the evolution of the birds-of-paradise’s promiscuous breeding systems and the males’ absurd exaggerations, could never have happened. As David Attenborough aptly concluded, ‘just as fruit plays a significant role in the biblical view of paradise, so it has created a paradise on earth for the birds.’15
However, it is because of their complex and unique relationship with the environment that the birds-of-paradise are restricted to New Guinea. Their strong attachments to display sites and the fact that males and females spend long spells apart make successful long-distance dispersal and the establishment of new breeding populations highly unlikely.