| The Magazine of the CALIFORNIA ACADEMY OF SCIENCES |
|
|
|
feature Instinct and Imitation The black-headed duck (Heteronetta atricapilla) of South America is a brood parasite; the female builds no nests of her own but lays her eggs in the nests of other species. When the hatchling black-headed duck abandons its foster parents' nest, it is immediately capable of swimming, diving, and foraging. It is completely capable of fending for itself, performing activities that could not have arisen through imitation--it has never seen its parents! Caribbean waters abound in fish, but are also the abode of poisonous sea snakes. Juvenile herons of three species were presented with various eels and sea snakes. The eels were eagerly consumed. However, the birds avoided the sea snakes having never seen them before, indicating an innate ability to recognize these poisonous creatures. These behavior patterns develop independently of learning by experience, but this is by no means true of all bird activity. Many of the most intriguing bird behaviors, often fundamental to survival, are clearly learned by imitation, so that, within the same species, behaviors in one locale may differ markedly from those in another. Some European oystercatchers (Haematopus ostralegus), for example, venture into shallow water where they thrust their bills between the shells of mussels and cockles, then saw through the adductor muscles and pick out the flesh. Other members of this species feed, primarily, by hammering a hole through the shell and removing its contents. Eggs taken from a pair that "hammered" and placed under an incubating pair that "opened" mussels produced chicks that later developed into openers, and vice-versa. Although it took about a year for a young oystercatcher to perfect either technique, the foraging behavior must have developed from observation- learning. Though there has been considerable discussion over the last few decades about learned traditions among primates, until recently there has been little attention given to comparable behavior among birds. Green-backed herons (Butorides striatus) often exhibit the remarkable behavior of fishing with bait or a "lure." Small fish are soon attracted to nibble at the bait placed on the water's surface, and with lightning speed the heron's head shoots forward and captures a fish. The prey is swallowed and thereafter the bird may pick up its lure and place it in a different spot to attract yet another fish. The heron will move the lure if a site proves unproductive, concentrating its efforts where it sees fish breaking the water's surface. Although green-backed herons are widespread, this behavior has been observed only in the southern United States, western Africa, and Japan. A population of green herons studied in southern Japan used a variety of lures including insects (flies, small dragonflies, cicadas), earthworms, berries, twigs, and leaves. More fish were captured when the birds used live bait rather than twigs or berries. Herons were more successful when they fished with bait. Juveniles were less successful at fishing than adults. This was partly because, unlike adults, young herons did not crouch after throwing the bait and must have been more conspicuous. Juveniles also tended to use lures--sticks and feathers--that were too long, while an adult was seen to hold six-to-seven-centimeter twigs under its foot, then break off small pieces with its bill before using them as lures. Truly a case of tool using! While detailed studies on lure-fishing have only been conducted on green-backed herons, there are anecdotal observations of other water birds fishing with lures: a captive squacco heron (Ardeola ralloides) in England, African pied kingfishers (Ceryle rudis), and a captive sun bittern (Eurypyga helias). Perhaps the most celebrated example of tool use in birds is that of the Gal~pagos tool-using or woodpecker finch (Cactospiza pallidus), first described by Academy ornithologist Edward Gifford in 1919. The tool-using finch feeds mainly on arthropods such as beetles, larvae, spiders, and their eggs. It flies from tree to tree in search of small creepy-crawlies. It may chip off bark to reveal hiding prey, or it may use its bill as a wedge to remove bark or grasp a piece of protruding bark by one end and pull. However, if the insect or larva is situated too deeply in a narrow crevice, the finch will break off a twig or leaf stalk and, holding it with its bill, use it as a tool to probe and pry out its prey. When the prey is captured, the tool may be dropped or it may be held under a foot, while the insect is eaten, and then reused. In the xeric lowlands, the tool-using finch makes use of the spines of the endemic Opuntia cactus. Academy fellow Robert Bowman noted tool using most often in the dry season when almost every individual was seen with a spine in its bill. Bowman and his student George Millikan studied these birds in the laboratory, where they also observed individuals using sticks as weapons to fend off other birds competing for the same food. They were also keeping a large cactus ground finch (Geospiza conirostris) in a cage next to four tool-using finches. Two tool-using neighbors often passed sticks back and forth to the cactus finch, who eventually developed, albeit poorly, the skill to probe with sticks for insects. The investigators tried to teach tool using to a few songbirds (e.g. plain titmice, Parus inornatus) that are known to hold items with their feet, but were unsuccessful. However, there are anecdotal reports of related ground finches (Geospiza spp.) and a warbler finch (Certhidea olivacea) using tools in the wild. This suggests that although tool use is learned, the disposition to learn is controlled by special genes which are shared by many, perhaps all, of Darwin's finches (the geospizines) but not by unrelated songbird species. This behavior is an example of nature (genes) interacting with nurture (observational learning) in the development of a behavior pattern. Tool use may be only a local phenomenon. In Tangipahoa Parish in Louisiana, brown-headed nuthatches (Sitta pusilla) use scales of bark as wedges or levers to remove other scales from longleaf pine and expose insects and other prey items. In a forest of loblolly pine and spruce pine only 30 miles away the nuthatches used no tools, for bark from these pines was not flaky enough to be easily detached. A quite different form of tool use is seen in Egyptian vultures (Neophron percnopterus). Ostrich eggs comprise a significant part of their diet although these eggs are too large for the vultures to carry and the shells are probably too thick to be cracked with the bill. So the vultures open the eggs by throwing rocks at them. On one occasion an investigator observed a palm cockatoo display by banging on the trunk with its foot. The bird then flew to a nearby tree, bit off a branch about two centimeters thick, and removed the leaves to make a stick about ten centimeters long. He then chewed the two ends until he produced his preferred shape. The drumstick produced a much louder sound than the foot alone. Parrots are very imitative species, and as with primates much of their behavior tends to be learned. Australian galah cockatoos (Eolophus roseicapillus) raised by Major Mitchell's cockatoos (Cacatua leadbeateri) imprinted on their foster parents' species and completely ignored their own kind. They even modified their wingbeats during flight to match those of their foster parents, learned to eat the same foods, and acquired their calls. Africa is the home of many lark species, members of the family Alaudidae. One of these, the flappet lark (Mirafra rufocinnamomea), may sing during its aerial territorial display, but it relies mainly on mechanical burring sounds made with its wing feathers to advertise its territory. During this display it ascends steeply to some 50 to 100 meters above the ground. Both wings are outstretched and quivered rapidly and the ascent may be followed by a circling flight over its territory covering a diameter of some two hundred meters. During flight the bird produces several bursts of sound, the number differing between localities. In the western section of Lochinvar National Park in Zambia, these larks produce "wing-songs" or "flappet songs" consisting of three phrases--"prrt-prrt-prrt." In the eastern section of the park birds produce flappet songs with only two bursts--"prrt- prrt." At nearby Kundaburika songs may consist of four bursts. Examination of the flappet songs on sonograms reveals that the number of notes per burst ("phrase") may vary from 24 to 25 per second. Individuals vary not only in the number of phrases per song but also in the number of notes per phrase and in the intervals of silence between the phrases. A male delivering a flappet song usually stimulates a neighbor to do the same so that bouts of counterflapping may ensue. The fact that nearest neighbors have more similar flappet songs than non-neighbors, unlike birds of many species, suggests that flappet songs are learned. In addition to sound, birds also use "visual" signals in advertising. Male bowerbirds of Australasia compensate for their drab appearance by constructing elaborate and often colorful display arenas, or "bowers," to attract females. Depending on the species, these bowers may be a "maypole" of sticks woven around an erect central pole, an avenue with woven stick walls, or an elaborate hut. These structures are decorated with hundreds or thousands of colorful objects such as flower petals, beetle elytra, feathers, colored fruit, and even human artifacts such as colored ribbons or cigarette boxes. Some species, such as the satin bowerbird (Ptilonorhynchus violaceus) of Australia, may use a "brush" consisting of a sponge-like wad of fibrous bark to "paint" their bowers with colorful fruit pulp or chewed charcoal. Females are attracted to these bowers, where they mate and then leave to raise their young without assistance from the male. When ethologist Gerald Borgia removed decorations from intact bowers, the number of matings was reduced by 75 percent. Jared Diamond, a longtime student of bowerbird behavior, has discovered that Vogelkop gardener bowerbirds (Amblyornis inornatus) also have local traditions. In various parts of Irian Jaya, the shape and structure of their bowers differ, as do the colors used in decorating them. At South Kumawa bowers consisted of a two-to-six-meter-high tower of sticks, all glued together. The tower rested on a mat of dead brown moss which was painted shiny black. Excreta tend to be white in other populations of these bowerbirds. However individuals at this site produced a unique oily black substance in their excrement, the source of the "paint." The bower was decorated with hundreds of gray and brown snails, brown beetles, dark brown acorns, and black stones. These birds were truly Puritan in their color taste. At a site only eight kilometers away, Vogelkop bowerbirds decorated their bowers with piles of fruit of six different colors: red, yellow, green, blue, white, or black. Additionally, brown pandanus leaves were propped against the maypole. An entirely different kind of bower is built by this species on the Wandammen Mountains. The maypole is concealed in a hut 0.9 to two meters wide and 0.4 to 0.8 meters high. In contrast to Kumawa bowers, sticks in Wandammen bowers were not glued but woven together. The bower rested on a mat of unpainted green moss. Black and brown beetle elytra and black bracket fungi were piled along with orange, red, and blue fruit and orange flowers. Although some decorations were placed inside and outside the hut, orange flowers were always piled outside. To test whether colors used to decorate bowers reflected real preferences for certain hues or nearby availability of objects, Diamond presented bower owners with poker chips of various colors: blue, purple, orange, red, lavender, yellow, and white. He either placed single chips of a certain color or three chips of each of the seven colors on the bower mat. These were later checked to see if the bowerbird had incorporated them into the bower ("harvested") or had removed and discarded them ("weeded"). At the Kumawa site where bower decorations were black or brown, all poker chips were weeded. At Wandammen, all individuals rejected white chips. However, all the other hues were incorporated into the bowers, with a preference for blue followed by a consistent hierarchy of preferences for the other colors. Although natural colored objects (fruit, insects, etc.) were available at both sites, only at the Wandammen site were colored poker chips harvested, indicating a real difference in color preference between the two localities. Diamond also found that within a population, individuals differed in decoration styles. Some individuals used butterfly wings and others not. Although beetle elytra were a popular item of decoration among all these bowerbirds, some removed beetle heads prior to use whereas others preferred the heads on. Are color preferences and bower styles learned or inherited? Diamond believes that they are learned, although his evidence is indirect. Various bowerbird species may take from four to seven years to build a bower, and young males spend a lot of time watching older males as they are building. Items are tested, discarded, or repositioned at various sites. Items of the same color tend to be placed in groups. Diamond also noted that nearest neighbors tended to use more similar decorations than non-neighbors. For example, only two out of ten individuals at Wandammen used butterfly wings, and those built adjacent bowers. This indicates that neighbors tend to imitate each other in choosing items or style (maypole versus hut) of bower. Most songbirds, however, convey their courtship intentions and their territoriality through their vocal (as opposed to mechanical) songs. Here, too, there appears to be ample evidence of learned expression. The baron found that a young chaffinch learned its song during a period preceding its first molt, which ended in August. As the short daylight hours of winter lengthened the following February, young chaffinches acquired still more songs and sometimes continued to do so until April. Thereafter their vocal repertoire remained fixed for life. The time window during which young birds are able to learn vocalizations is called the "critical period" or "sensitive phase." Vocal tradition, he discovered, gave rise to regional dialects; chaffinch songs from different localities differed from each other as Cockney English differs from Oxford English. All this the baron accomplished by ear, without the benefit of instrumentation. One of the great breakthroughs in the study of sound communication in humans and other animals was the invention of the sound spectrograph or "sonograph" machine by Bell Laboratories prior to World War II. We can now quantify duration and rhythm (in seconds), pitch or frequency (in kilohertz), and loudness (in decibels) in animal vocalizations. We can also describe tonal quality by the harmonic structure of the sounds as seen on the sonograph. In the early 1960s, Academy Fellow Peter Marler and his associates at the University of California at Berkeley described two regional song dialects for the white-crowned sparrows (Zonotrichia leucophrys) living in Berkeley. Birds recorded in the city typically sang a song beginning with two pure whistles, followed by two trills that descended in frequency (pitch). In neighboring Tilden Park, birds sang songs beginning with two downward-inflected whistles, followed by a vibrato or buzz, followed by a single trill. Sonograms of the two dialects revealed that although syllables in the first portions of their trills were similar, syllables at the ends of the trills differed between the two populations. In the late 1960s and early 70s I expanded MarlerUs studies and discovered that white-crowned sparrows in San Francisco could be divided into three dialectal populations. One song dialect was restricted to the Presidio. A second was sung by birds throughout the city, and a third was sung by birds inhabiting the shores of Lake Merced and in Daly City. From these results, and similar ones from transect studies along the California coast, I concluded that dialects in these non-migratory sparrows are very local in distribution, changing every few kilometers. I then wanted to know what occurred in areas where two dialects met. At one of these boundaries, in Strawberry Canyon in the Berkeley hills, I found that some birds sang the Tilden Park dialect, some sang the Berkeley city dialect, and a small subset sang songs containing syllables borrowed from both dialects. One individual was "bilingual," singing both the Berkeley and Tilden dialects. In mid-September, white-crowned sparrows from Oregon, Washington, and British Columbia arrive in the Bay Area to overwinter. They move about in flocks and remain until April, when they return north to breed. I wished to know if these visiting migrants would teach the local birds their alien dialects. But first I needed to know what sounds comprised dialects in songs of migrants, and I embarked on a camping trip up the Pacific coast. I found that in contrast to the local non-migratory white- crowned sparrows, song dialects of the migratory populations were distributed over large areas covering 50 to two hundred kilometers, perhaps because migratory birds may settle miles away from their hatching sites. Happily (for me), in those hundreds of kilometers comprising my transects only seven dialects could be recognized. In 1968, I recorded a white-crowned sparrow on the U.C. Berkeley campus singing a strange song which did not match any of the Bay Area dialects. By matching his sonograms to those in my catalog I identified it as typical of birds living in southern Oregon. Was he an Oregon sparrow that had settled in Berkeley or a local bird that had learned an alien dialect from a winter visitor? By placing a tape recorder playing his own song next to a trap, I lured him to it. I could now compare him with sparrows found in the north and others found in the Bay Area. Bay Area birds, belonging to the subspecies nuttalli, are darker and larger than migratory birds assigned the name pugetensis. This individual matched the local birds. Since then I've found a dozen local birds singing songs learned from migrants, including one living at Pacheco and Sunset Boulevard, which sings a dialect from San Francisco and a dialect typical of Alaskan birds. If local individuals learn dialects from migrants,
why are there so few resident birds singing alien dialects? Why are
these borrowed songs not passed on to more birds? The answer lies in
the phenomenon of match-countersinging, where birds tend to respond
to their neighbor's singing using the same dialect. When I visited the
Alaskan bird a month later, I found that he was mated and only sang
San Francisco songs. When I played him recordings of his Alaskan song,
it took 70 repetitions before he responded in kind. He still remembered
the alien dialect but did not use it. Just as students of dialect geography may use language studies to trace movements of human races or tribes, ornithologists may use dialects to trace origins of colonizing populations of birds. For example, human dialect geographers have studied Polynesian dialects and traced migration routes of Polynesian peoples across oceans. The southernmost breeding population of Sierran white- crowned sparrows (belonging to the subspecies oriantha) is at southern Sequoia National Park. In 1956, white-crowned sparrows were first recorded breeding in the San Bernardino Mountains, 300 kilometers to the south of Sequoia. Some years later, when I recorded their songs and compared their sonograms with my lexicon of white-crowned sparrow syllables I was able to trace the origin of these populations to Mammoth Lakes, some 450 kilometers to the north. Sparrows on Treasure Island, in San Francisco Bay, sing a dialect similar in structure to those sung by San Francisco birds. Sparrows on Angel Island and Alcatraz Island sing songs similar to those recorded in Marin County. Based on song similarities I have concluded that these islands were colonized from different mainland populations. Evolution of species results from accumulation of micromutations. Evolution of dialects and languages results from accumulation of "cultural micromutations" or copy errors. By comparing related languages one can sometimes find intermediate steps or cultural homologues going from a mother language to an evolved language. For example, from the Proto-Indo-European word "mater," meaning mother, the following cultural homologues were derived: matar (Sanskrit), mathir (Old Irish), mat (Russian), mere (French), Mutter (German), madre (Spanish), mae (Portuguese), mother (English). Just as human languages are continually changing and evolving, so also are notes and syllables in birdsong constantly changing, and the intermediate forms enable investigators to trace these cultural changes. The syllables present in white-crowned sparrow songs that I recorded in front of the Academy 20 years ago no longer exist and have been replaced by others. There are also parallels in birds and humans in the modes whereby vocal traditions are passed on. Traditions are said to be vertical when passed from parents to progeny. They are termed horizontal when passed between age peers and are said to be oblique when passed from adults to non-related juveniles. While conducting field work on the island of Karkar off the coast of south-eastern Papua New Guinea, I was once invited to participate in a special singing ceremony. One of the senior villagers had purchased a song. He had paid three pigs, the first of several payment installments, to a famous composer for a beautiful song. I was informed that his oldest son would inherit his land, but his second son, Dalos, would inherit this song, his father's next most precious possession. Each evening father would teach Dalos the verses of his song, and they together led the village in singing. Someday it would be Dalos alone who would be leading the singing. Father was passing his song to his son, an example of vertical tradition. A parallel may be found in several species of Darwin's finches, among which sons learn songs directly from their fathers. Any idiosyncrasy found in a father's song will be found in those sung by the sons. Juvenile male Australian zebra finches also learn songs directly from their fathers. The white-crowned sparrow, in contrast, follows an oblique tradition, learning songs from adults at its nesting site. White-crowned sparrows do have an innate preference to learn their own species song. A juvenile, untutored white-crown, when played a tape containing junco, green-tailed towhee, Bewick's wren, and white-crowned sparrow song, will select white-crown song to imitate. However, juvenile white-crowned sparrows played tapes of song sparrow songs rejected them as models altogether and sang simple songs similar to those typical of birds deprived of any tutoring. But white-crowns exposed to living song sparrows produced good copies of their alien tutors' songs. The barrier to learn an alien song was surmounted by interaction with a live tutor. Here again is a remarkable parallel with humans. Teachers have noted that children exposed to tapes or videos designed to teach a second language acquired only rudimentary knowledge of the language, even if the tapes were prepared in a tutorial style. Children interacting with live teachers, however, readily learned new languages. Although human speech and birdsong are products of two analogous systems, many parallels exist between the production, acquisition, and dispersion of human and bird dialects. By studying one system we may discover still more insights to better understand the other. But as I hurry to document the vocalizations of my subjects, I worry that they may go the way of the dusky seaside sparrow, whose hopeful voice is on record, but whose behavior and vitality are gone forever. As anthropogenic extinctions claim bird species around the world, it is a small consolation to have recordings of creatures who will never sing their own songs again.
Luis F. Baptista is Curator of Ornithology and Mammalogy at the California Academy of Sciences. |
Fall 1994
|
