CHAPTER 20 Communities The wer tropical lowlands are rich in diversity of sp diversity of structure, and in their general aspect of hecuriance. [MacArhur 1972, p. 199] “es, in the local population size for a species but also how many species can coexist locally in one habitat. Coexisting groups of species are called communities. This chapter emphasizes the dynamic nature of bird communities. cluding those on islands where the patterns of losses and gains of species are best documented. Whether the numbers of species reach an equilib- rium helps to define positions in the continuum of community structures. Ac one end of the continuum are “open” assemblages determined largely by geographical history with room tor additional species. At the other end of the continuum are “closed” communities organized by interspeciti petition for limited resources and resistant to invasion by additional species. After a review of the general patterns of species diversity in space and time, this chapter looks at the role of competition in structuring bird com- munities. Patterns of ecological segregation and geographical replacement among similar species point to a role for competition in community struc- ture, But rigorous documentation of actual competition is difficult. The role of interspecific competition as a dominant force that controls the composition of communities is hotly debated. In one situation, the effects of competition among European tit species on their reproduction and for- aging illustrate the costs of coexistence for ecologically similar and closely related species. The local compositions of species, however, also relate «© larger playing fields of geography and history. tC esources, such as food and nest holes, determine not only The Dynamics of Communities ‘Whether it’s a question of tallying the numbers of individual birds of di ferent species while birding or doing a formal census. the result is the604 CHAPTER TWENTY 80 60 Veery llack-capped Chickadee Black-throated Five most common species Green Warbler make up 55% of birds in Ovenbird avian community 40: ‘Cumulative percentage of avifauna i Red-eyed Vireo o ° 30 60 90 vo ‘Most common Least common Species ranked from most to least common FIGURE 20-1 Most species are rare, and a few are abundant. The distributions of species abundance typically compile to a log normal distribution, with a long skewed tall to the right, such as the one shown here. Data that are intrinsically geometric oF ‘multiplicative rather than additive produce log normal distributions. Populations ‘multiply. So, in principal, the distribution of their abundances should be log normal [Graph after Temple 2004] same. Most species are rare or present in small numbers, and a few are common or abundant (Figure 20-1). Their distributions correspond to the dynamic nature of bird species, which multiply, prosper. and full. They expand and contract. They generalize and specialize. Local and assem- blages of species therefore correspond to their combined histories and prosperities. Competition for limited resources also may determine which particular species coexist and in what numbers, Evolution and the availability of resources both play major roles in forming communities. The local compositions of species change from epoch to epoch in accordance with evolutionary history and from season to season in accordance with resource availability. Reflecting regional his- tory, avifaunas (the birds of a particular region or time period) are the grand resule of millions of years of evolution, adaptive radiation, immigra- tion, dispersal, and extinction of bird taxa, Invasions and fusions of species from other regions supplement adaptive radiations of ecological types within a region. Owing to the availability of resources as well as to their histories, local bird communities increase dramatically in both size and di- versity in tropical forests, in comparison with local communities in north- erm temperate forests The species in a community can occupy different trophic levels. or feeding levels, and range from insect-eating warblers to the hawks that eat them and from plankton feeders to fish eaters (Box 20-1). Each speciesTE bx 20-1 STABLE ISOTOPES CLARIFY SEABIRD FOOD WEBS Stable isotope analysis rosy (SIA) helps ornithologists to decipher food webs, es- pecially marine systems that are hard to study di- rectly. The ratio of two isotopes of nitrogen ('5N/'4N) in animal tissues, for example, increases from zooplankton (lcill) to krilleating fish or birds to fish-eating birds (cormorants) and sea lions at the top of the food chain (Sydeman et al.1997; see graph). This ratio increases with each step in the food chain because the metabolic syn- thesis of new protein from food consumes more of the lighter isotope "N, leaving more (3 per- cent) 'SN in the predators’ tissues. The marine- system food web off the coast of California, for example, includes five levels, with seabirds occu- pying the upper levels, from three to five. Offshore "N/M (76) CAA ‘Al The ratio of stable isotopes of carbon (°2C/"2C) in tissues reveals where seabirds feed, on the basis of the kinds of prey that they eat, and whether they change feeding habits at dif- ferent stages of the breeding cycle. Feeding on fish near the coast leaves a different signature of carbon isotopes in the tissues from that left by feeding on krill in the open ocean. SIA of the egg albumen of California seabirds, for exam- ple, revealed that Cassin’s Auklets and Common Murres use krill to produce their eggs. Common Murres then shift to fish to feed their young, as do most of the other seabirds. Rhinoceros Auklets and Pigeon Guillemots rely on fish throughout the breeding cycle, including egg production. inshore ewecu Ly® © @ANCH 'SBRF SABLE@ @SARD unce’ esqui a0 8 BerPC Cr) Trophic structure of a community of seabirds and their prey on the coast of California based on analyses of stable nitrogen and carbon isotopes, expressed in parts per thousand (%.). (The formula for calculating these values in standard delta notation produces a negative number for carbon.) Nitrogen isotopes change with trophic level in the community. Birds and sea lions occupy higher trophic positions than fish, squid, and kel Carbon isotopes vary with inshore versus offshore feeding habits. Abbreviations: EUPH (Keil, or euphausids), LING (ingcod), SQUI (Squid), SABLE (sablefish), SARD (sardine), SALM (salmon), SBRF (short-bellied rockfish), CAAU (Cassin’s Auklet), ANCH (anchovy), WEGL (Western Gull), COMU (Common Murte), PIGU (Pigeon Guillemot), RHAU (Rhinoceros Auklet), PESH (Pelagic Shag), BRCO (Brandt's Cormorant), {MU (northern sea lion). [From Sydeman et al. 1997]606 (CHAPTER TWENTY FIGURE 20-2 According to some ecologists, stable communities of coexisting bird species derive from 3 larger pool of species through the dynamics of population dispersal, through colonization in relation to habitat or other resources, and sometimes by competitive resolution of unstable species combinations. (After Wiens 1983) has specific requirements, called its fundamental ecological niche (Grin nell 1917). For example, we expect to find Pileated Woodpeckers in forests with large trees full of carpenter ants. Pileated Woodpeckers oc- cupy a wide variety of forests with big, ant-ridden trees, but the Red- cockaded Woodpecker of the southeastern United States has a different and highly specialized niche. It requires old pine forests with trees from 80 to 100 years old that have been infected by the red heart fungus. ‘Woodpeckers play a central role in the larger community. They and their dead trees are a resource that supports a healthy diversity of verte brate and invertebrate species that depend on one another. The foraging and nest-excavation activities of woodpeckers determine how dead trees actually decay and become available for use by other species (Farris et al 2004). For example, more than 50 percent of the woodpeckers sampled in the ponderosa pine forests of northern California carried fgal spores on their bills, They transport these spores from tree to tree, causing new trees to decay and to become suitable for excavation of their new nest holes. Many species of hole-nesting birds and small mammals use the di- versity of nest holes that result, building guilds (ets of ecologically simi- lar species) of interdependent nest-cavity species (see Figure 15-5). Resource availability and evolutionary history shape communities. and, together, they give rise to competition—the interactions among species through which one species may directly or indirectly preclude the pres ence of another, Competition itself can play a major role in forming com- munities, From a larger pool of potential candidates for a community, competition selects those species best able to coexist (Figure 20-2). Com- petition typically constrains a species’ habitat or feeding opportunities and thus narrows that species’ realized ecological niche. This outcome is more likely than the exclusion of one species by another. Speciation Extinction —f species pool Populations “Assembly rules Resources Dispersal ability Colonization probability Predictability Competitive ability Habitat requirements: Productivity Exploitation strategy Competitive interactions Variety Permitted species combinations ‘Community Stable coexistence Partitioning of resourcesEcologists debate the relative importance of these three forces—evo- lutionary history, resource availability. and competition—in community ecology. Competition has long been assumed to be the main force in structuring bird communities, but irregular seasonal changes keep many communities in a state of dynamic flux. Habitat disturbance, fluctuating resources, ongoing colonizations, and local extinctions impose short-term changes on avian communities. In addition, deforestation and the rota- tion of farmlands transform woodland habitats and their communities of bird species into patchworks of chronically unstable and unsaturated habi tats. New or unstable habitats invite opportunistic use by birds and inv: sion by exotic or introduced species. Some communities, therefore, lack predictable structure and, instead, may be fortuitous or historical collec- tions of species (Hubbell 2001; but see McGill 2003), Islands Because islands are such discrete places. island birds provide many of our clearest examples of community dynamics as well as the process of spe~ ciation (see Chapter 19). We can deduce and sometimes actually docu- ment the arrivals of new colonists and their subsequent adaptations. expansions of distribution, and disappearances in time. The active dispersal from and colonization of isolated places are twade- marks of bird behavior. The dynamics of colonization are most apparent BOX 20-2 607 ‘COMMUNITIES CHARACTER DISPLACEMENT EVOLVED RAPIDLY IN SUPERTRAMPS Between the large, exotic 7 New Britain in the South Pacific lie several small islands, the Long group. On these little-known islands, Jared Diamond and his colleagues (1989) discovered an example of the rapid evolution of character displacement be- tween two species of small, nectar-feeding hon- eyeaters (genus Myzomela) The Long group's volcanic caldera collapsed 300 years ago in one of the largest known vol- canic explosions. It undoubtedly eliminated all bird life. The current avifauna consists mostly of small island specialists, or supertramps, that have recolonized the island. Among the now estab- lished supertramps are two similar honeyeaters, the Ebony Myzomela from the northern Bismarck Archipelago and the Scarlet-bibbed Myzomela from the southern Bismarck Archipelago. The two honeyeaters are present abundantly together all over the Long islands and often feed in the same flowering trees. ‘These honeyeaters are the only two bird pop- ulations on the Long islands in which the birds have evolved to different sizes since colonization. The Ebony Myzomela is now larger than the birds of its ancestral populations, and the Scarlet bibbed Myzomela is now smaller. The difference in size between the birds of the two colonizing populations has increased from 24 to 43 percent ‘on the Bismarck Archipelago to 52 percent on the Long islands. Elsewhere, closely related species of honeyeaters coexist only if they differ in size by 50 percent or more. Diamond and his colleagues concluded that the changes in size in the two honeyeaters on the Long islands represent character displace- ment in response to each other's presence dur- ing the few centuries since the Long group's devastating eruption,608 CHAPTER TWENTY on oceanic islands, such as the West Indies, which receive periodic a rivals of new visitors dispersing over water from larger source areas. Wa- ter barriers favor colonization by highly mobile species tha cravel in small groups. Bananaquits in the West Indies and white-eyes in the Indian and Pacific Oceans are superb island colonists—or “supertramps” (Diamond 1974). Their extraordinary dispersal abilities enable them to be the most predictable first colonists on newly formed islands (Box 20-2). Successful colonization of one island may be followed by the colonization of adja- cent islands and continued spread throughout a region. A colonis’s ecological flexibility and its competitive ability to fit into the local community increase its chances of establishing a population on a new island, Bananaquits and white-eyes are generalized opportunists, able to take advantage of local situations. They breed readily and repeat- edly. After they are established, their populations thrive and grow rapidly in an environment with few enemies—specialized predators, competivors diseases, or parasites. Population growth under such conditions of so-called ecological re- lease leads to large. dense populations and to the use of a wider variety of habitats than is the case on the mainland. Resident birds of the Pearl islands off western Panama, for example, achieve densities from 20 to 40 percent higher than those reached on the adjacent mainland. They also forage over a greater vertical range and use more habitats than do their mainland counterparts (MacArthur et al. 1972). Both the average num- ber of habitats used by a species and the density of each species in a par~ ticular habitat may double on small islands with few species, such as St. Lucia and St. Kitts of the Caribbean (Table 20-1) Populations on different islands diverge from one another as they adapt to local niches. The divergence of allopatric populations was discussed as PIABLE 20-1 Relative abundance and habitat distribu ition of birds in five tropical localities* Average Number Number of of Species Relative Species per ‘Abundance Observed Habitat_—-Habitats_—_per Species Relative Relative (Regional (Local per per Habitat” Abundance Abundance Locality Diversity) Diversity) Species. _ (Density) __per Species’ _of all Species! Panama 135 302 2.01 2.95 5.93 800 Trinidad 108 28.2 2.35 331 778 340 Jamaica 36 214 3.43 4.97 17.05 955 ‘St. Lucia 33 15.2 415 5.77 23.95 790 St. Kitts 20 19 5.35 5.88 31.45 629 “Based on 10 counting periods in each of nine habitats in each locality. 'The relative abundance of each species in each habitat is the number of counting periods in which the species vas seen (maximum 10); this number times number of habizats gives rel ve abundance per species; this relative abundance times number of species gives relative abundance of all species together. From Cox and Rickles 1977,FIGURE 20-3 White-eyes are excellent island colonists that often occupy unfilled niches on remote istands that have few other birds. Shown here are (A) a typical species, the Bridled White-eye, and (8) a large, thrushlike species, the Giant White-eye. ‘These species are found together on Belau (Palau) in the Caroline Islands. [From Lack 1971] part of the speciation process for Darwin’s finches (see Chapter 19) Generalized colonists, such as white-cyes, may take over the specialized niches of species that are missing from island communities. The increased specialization of the first colonists then contributes to their ability. as well as to that of their descendants, to coexist with later arrivals. On La Réunion in the Indian Ocean, where there is no nectar-feeding sun bird, the Reunion Olive White-eye has become a specialized nectar feeding species in both bill morphology and behavior (Gill 1971). It co exists there with a second, generalized white-eye, the Reunion Gray ‘White-eye (see Chapter 19). In the tropical Pacific, unusually large species of white-eyes have evolved independently on 12 small islands that have few other species, These large white-eyes often coexist with one or more other, smaller species of white-eyes (Figure 20-3) Turnover Tumover—the addition 3 uses the composition of avifaunas to change with the passage of time. The time scales range from the long-term, gradual replacements of ancient taxa by more recently evolved taxa to short-term, yearly changes. The number of species may 1d loss of species 609 ‘COMMUNITIES,610 CHAPTER TWENTY or may not reach an equilibrium where new additions through coloniza~ tion or speciation balance extinctions. The number of species present in northern Scandinavian bird com- munities, for example, Muctuates by 15 percent from year to year (Jirvinen 1980). Farther south, in the rich deciduous forests of Birdsong Valley, Sweden, annual eurnover is about 10 percent. In both regions, the repeated extinction and immigration of small numbers of rare species (from one to two pairs total) are the primary causes of annual variations in com- munity richness, Higher turnover in the north is caused by the greater unpredictability of weather during the breeding season and the greater proportion of rare species, which in turn are more likely to become ex- tirpated or locally extinct. The local turnover of species is a fundamental concept of geographi- cal ecology. Both island and mainland avifaunas are subject to conspi ous species turnover. Turnover rates of | to 20 percent of species a year prevail on small, offshore islands, such as the Channel islands off the coast of California (Diamond 1980). Annual censuses of birds on a small British land, Calf of Man, revealed that Northern Wheatears and Eurasian Stonechats repeatedly became extinct on the island, only to be replaced within a short time by new immigrants of the same species. In this way, regular immigrations may obscure regular extirpations until che popula- tion establishes itself (Figure 20-4). Like human civilizations, bird populations undergo long-term cycles. The expansion of their numbers and their geographical range is followed by diversification, speciation, specialization, and then dectine, local extir~ pation, and the replacement by younger species (Ricklef 2005). The bird population cycles themselves span long intervals of roughly | million years. which allows plenty of time for new traits to evolve Initially formulated to explain the deployment of ants on the islands of the South Pacific (Wilson 1961), these so-called taxon cycles capture the concept of dynamic changes in the composition and structure of bird communities on islands and most likely on continents, too, Current data 2” & pe le 1 €1 = ul LAG Bol S20 BY Foe Te 2 Lad od iso 96s 907 Year FIGURE 20-4 Turnover of Northern Wheatears (top) and Eurasian Stonechats (bottom) as indicated by annual breeding bird censuses on the Calf of Man, a small British island. Regular extinctions and replacements of small populations take piace. Abbreviations: |, immigration; E, extinction. {From Diamond 1980]suggest that the cycles of different species are largely independent and not coordinated, for example, by climate cycles. ‘What really triggers the expansions in birds’ population size We don’t know. A principal hypothesis favors es ‘enemies—parasites, diseases. or predators—by virtue of a new mutation or behavioral trait. Conversely, reproductive success and population size tend to decline asa result of increased susceptibility to enemies—namely. parasites and predators. Aggressive new colonists or competitors also force previously established residents to retreat into a subset of the available habitats, De- clining population sizes increase the probability that some populations will become extinct. Local extinctions then fragment the distribution of the species. The House Wren, for example, was once widespread on islands of the Lesser Antilles of the eastern Caribbean, It then disappeared for no apparent reason trom Guadeloupe and Martinique and is close to extine- tion on St. Vincent. What may appear to be the direct, competitive replacement of older endemic species on islands by new colonists may only be coincidence. For example, Mourning Doves replaced the endemic Socorro Dove on Socorro Island, off Baja California. in the 1970s. The extinction of the Socorro Dove and the colonization by the Mourning Dove turn out to have been independent events related to the establishment of a garrison of the Mexican army. The army's cats ate the tame endemic doves. Its new wells provided the drinking water required by the colonizing Mourn- ing Doves (Jehl and Parkes 1983), Biochemical analyses of ages and relationships among island bird pop- ulations confirm complex. patterns of expansion. specialization, and the fragmentation of distributions. The oldest of the 57 lineages of small land birds occupying the Lesser Antilles dates back roughly 10 million years (Rickles 2005). Expansions and recolonizations by derived species pro- duced a complex mosaic of communities of species on different islands, For example, wo lineages of mimic thrushes (Mimidae), che West Indian thrashers and uemblers, date back roughly 4 million years. Their history features at least 16 chapters of island colonizations and divergences of sub- species or species (Hunt et al, 2001), The three modem-species groups of tremblers derive trom splits at the beginning of the Pleistocene 2 mil- lion years ago. nd range? pe from burdens of Island Biogeography Although island avitiunas may change regularly in the composition of their species, the number of species present theoretically reaches a bal- ance between gains due co immigration and losses due co extinction (MacArthur and Wilson 1967). The predictable balance between gains and losses is known as the equilibrium theory of island biogeography. By this theory, the point of intersection between the immigration curve and the extinction curve defines an equilibrium species number (Figure 20-3) The rate of extinetion increases with the number of species on an island because there are more species with competition-reduced population sizes. eu (COMMUNITIES,12 (CHAPTER TWENTY @ Immigration rate Satinetion rate, “ 5 Number of species on island Number of species on island FIGURE 20-5 (A) The number of species found on an isiand reflects the balance between the rate of immigration (colonization) and the rate of extinction. (B) Immigration rates on islands that are distant from source areas are lower than rates on islands close to source areas. Extinction rates on large islands are lower than those on small islands. Extinction rates increase as the number of species present on an island increases. The point of intersection of the two curves for any particular island defines the expected equilibrium number of species, S. [Fram Ricklefs 1976a] Conversely, the rate of immigration fills as the number of species in- creases, Fewer new species from source areas are possible, and the colo- nization of an island full of competitors is more difficult. Rates of extinction and colonization vary among islands of different sizes and degrees of isolation, In general, the observed relations between the number of species and island size are in accord with the model (Fig ure 20-6), However, they do nor take into account massive prehistoric extinctions caused by humans, such as the Polynesian colonists on the Hawa d elsewhere in the South Pacific (Olson and James 1982). The number of equilibrium species for large land-bridge islands, which were once part of a mainland with a full complement of species, is much greater (often three times) than that for large distant oceanic islands. which depend solely on colonists that cross the seas. Land-bridge islands have lost species steadily since they were isolated by rising sea levels at the end of the Pleistocene epoch (10,000 years ago). Small land-bridge islands have lost a greater proportion of their inital populations of birds than have large land-bridge islands of comparable age (Table 20-2) Barro Colorado Island in Panama is a land-bridge island that was sep- arated from the mainland in recent times by the opening and flooding of the Panama Canal in 1914, The island is home to the prestigious Smith- sonian Tropical Research Institute. Its scientists thoroughly documented the species that were present and how they fared in 85. years of isolation (Robinson 1999). The result: 65 bird species have disyppeared from the Islands a613 4 COMMUNITIES Number of species (5) 10 100 1000 10,000 100,000 Island area (km?) (A) FIGURE 20-6 The number of species, S, found on islands increases in direct relation to island area, A. This graph is plotted for islands of the West Indies. [From Ricklefs and Cox 1972} island. Once gone, sedentary forest-dwelling species in particular have not recolonized the island. Many of them were lost before 1970. At least 14 species disappeared after 1970, and three others (Slate-colored Grosbeak. Speckled Mourner, and Rufous Piha) that were abundant in 1970 now persist only in small numbers. Extinction can result simply from the inevitable Hluctuations in the size of small populations. The probability of extinction, therefore, depends on population size and on the area containing the population, What has been causing this loss of bird species from Barro Colorado Island? Several changes contribute to the loss, but one of them highlights, the portance of large predators, such as jungle cats and eagles, These ist land-bird faunas of five major land-bridge islands. PI VABLE 20-2 Present ana’ provable Number of Species a Island Area (km?) nal Present Extiner Fernando Po 2.036 360 128 232 Trinidad 4.834 350 220 130 Hain. 33.710 198 123 B Ceylon 65,688 239 71 68 Tasmania 67,978 180 x8: 92 From Terborgh and Winter 1980,o14 (CHAPTER TWENTY predators disappeared soon afier isolation, leading to increases in their prey—small forest mammals and other predators of ground-nest birds Similarly, mainland forests are increasingly reduced to small island frag- ments that are subject to loss of species in what is sometimes called the small-island effect (see Chapter 21). Open and Closed Communities Compared Historical forces add and subtract species from local communities over long time periods. Evolutionary changes, on one hand, combined with seasonal changes and daily species interactions on the other hand, mold the actual compositions of local assemblages of species. Whether commu- nities achieve stable equilibria through short-term or evolutionary adjust- ments remains uncertain Classical community ecology was largely an equilibrium science that assumed that both population and community processes tended strongly toward equilibrium states (Ricklefs 2005). Replacing that assumption now isa growing recognition that dynamic evolutionary and ecological adjust- ments may prevail. A continuum of possible community structures indi- cates the diversity of these processes in play. Derived from carly botanical thought are two polar views that em- brace the continuum of community formation and dynamics (Figure 20-7). Open communities are fortuitous, dynamic assemblages of essen- tially noncompeting species. The species in these communities each align themselves independently along environmental gradients according to their own ecological requirements (Gleason 1926; Wiens 1990). Closed communities are stable combinations of species that separate themselves from a pool of possible colonists and competitors. Community efficiency, stability, and resistance co invasion by additional species increase with evo- lutionary adjustment among members of the community. In this view competition for limited resources structures the local composition of species. The closed-community concept has been a dominant theme of modern studies of bird communities. These views are extreme, but evidence exists for both, For example the birds that breed in upland hardwood stands in southern Wisconsin form open-community assemblages of species (Bond 1957). Their habi: tats range from open, dry. deciduous forests dominated by black oak trees to denser, moist forests dominated by sugar maple trees. Certain bird species, such as the Red-eyed Vireo. are more common in wet-climate forests. Others, such as the Bhick-capped Chickadee, are more common in dry-climate forests. Some, such as the American Redstart, are most comimon in intermediate forest types. Each species has specific preferences or needs and chooses its habitat accordingly. These overlapping distribu- tions suggest independent, ecologically related associations, not coincident relations of coadapted species. Disturbance and unpredictable climate changes foster open-community assemblages. Series of unpredictable wet and dry years or of severe and benign winters are the norm worldwide. Pronounced year-to-year variationsClosed communities Ecotone Abundance Open communities Environmental gradient FIGURE 20-7 Open and closed communities are extremes on the continuum of possible community structures along environmental gradients, such as dry forest to wet, forest. In open communities (bottom), species are arrayed independently according 10 their particular ecological needs. In closed communities (top), distinct sets of species occupy particular habitats with breaks at the interfaces between habitats, called ecotones (arrows). [Frum RicklefS 19792] are typical even of tropical rain forests, once thought to be the most sta- ble of ecosystems, Fire, an extreme form of disturbance, naturally and pe- riodically devastates chaparral communities in California and the grasslands of both Africa and the western United States. Regrowth after a bum pro- ceeds through regular patterns of plant succession and associated bird com- munities. Through che years. localized burns and recoveries create an ever-changing mosiic of unstable habitats. The local communities that occupy them are therefore dynamic ones rather than self-perpetuating systems at equilibrium, Similarly, fire controls the shrub-steppe bird communities of the Great Basin of western North America, Interspecific competition does not seem to be an important force in such dynamic communities (Dunning 1986), Instead, these birds appear to be limited principally by winter food avail- ability rather than by summer food availability. They respond opportunis- ically and without competition ¢o locally abundant, nonlimiting resources, on their breeding grounds. 615 ‘COMMUNITIES616 CHAPTER TWENTY Communities of hummingbirds also are dynamic ones centered on current bloonis of favored flowers. The evolutionary history of humming- birds features physiological adaptations to high altitudes, such as chose of the Andes of South America, and specializations of bill structure for feed- ing and pollinating particular flowers (see Figure 1-9). It also features in- tense competition within and among species for nectar. Hummingbirds, as well as other nectar-feeding birds, provide some of the best examples of closed assemblages of species that compete for food Coexisting hummingbirds organize themselves in predictable, but often temporary, sets around one or two large, territorial species that control access to preferred flowers. In the highlands of Costa Rica, for example, the aggressive Fiery-throated Hummingbird and Green Violetear control the clusters of flowers that are rich in nectar (Wolf et al. 1976). Comple- menting such territorial species in each local set of species are (1) a large, nonterritorial, long-billed species. which visits scattered flowers contain- ing large nectar volumes: (2) a small-sized. nonterritorial, short-billed species, which visits small flowers containing small amounts of nectar; and (3) a small-sized, sharp-billed filcher, such as the Tufted Coquette, which steals nectar from undefended big flowers by piercing the base of the corolla (Figure 20-8A) Omichologists frequently discover that the ranges of closely related species are mutually exclusive. Each island in the West Indies, for exam= ple, is inhabited by only 2 or 3 hummingbird species. although 15 hum- mingbird species inhabit the region (Figure 20-8B). Only nwo resident species. a small one and a large one, inhabit low-lying islands. Mountain- ous islands are populated by three types of hummingbirds: a small. wide spread species and ewo large ones, one of them in the lowlands and the other in the highlands, Competition can be inferred to be the process his- torieally responsible for such replacement patterns of ecologically similar species. If communities were truly closed, they would be resistant to invasion by new, introduced species. But so-called invasive species occupy many bird communities worldwide. Think of Common Starlings and Pheasants in North America. The avifaunas of the West Indies now in- clude 23 established exotic bird species. The Hawaiian Islands have more than 45. and Florida has 58 plus more than 100 additional species that es- cape or are deliberately released each year but don't yet breed. What is the distinction, if there is one. between what we call “inva sive.” “alien,” “exotic,” of “nonnative” species and what we call “native species that expand their range or invade a new island naturally? It rests on the role of human transport in their dispersal (Ricklets 2005). Whether introduced species are poor dispersers that advance because they are su perior competitors that can penetrate a closed community is unknown. The many exotic species introduced into the Hawaiian Islands have largely failed to invade the intact native forest there (Scott 1986). Instead. they thrive only in disturbed and human landscapes. much like those of their origins. As with natural range expansions, exotic species may thrive be- cause they temporarily escape from the herbivores, predators, or pathogens that formerly limited their numbers. Common.617 ‘COMMUNITIES FIGURE 20-8 (A) The small Tufted Coquette uses its sharp bil to steal nectar from undefended big flowers by piercing the base of the corolla. (8) Geographical replacements of possible competitors, Different species of small hummingbirds do not coexist in the West indies; instead, they have segregated distributions. [After Lack 1971] ‘What is known is that the success of an introduction corresponds pri- narily to the introduction effort itself—how many many times. Many hali-hearted introductions fil, Species Diversity The number and diversity of bird species in an area vary with habitat, among continents, and from one region of the planet to another. The major patterns of species richness—lowest in the Arctic. highest in the Tropics—have been known for years. The underlying causes of these618 (CHAPTER TWENTY patterns, however, continue to challenge us, Understanding why there are so many more species in tropical communities than in temperate com- munities is especially challenging. The next sections review the patterns of species diversity in time and space, with special emphasis on tropical species diversity, Differences mong communities in numbers and relative abundances of their species relate to regional and historical processes as well as to lo~ cal forces such as productivity and seasonal stability. Diversity in tropical communities. in particular, relates to their long, stable histories of accu- mulation of specialized species (Moreau, 1966). Ancient communities may be the most species rich of all. For example, the forest faunas of Panama are richer than those of Affica, but the grasslands and savannas of Panama are relatively impoverished. The lowland forests in Africa were restricted in extent during the Pleistocene period. which prevented the develop- faunas (Karr 1976). The man-made grasslands in ment of rich forest avi Panama are quite young (15,000 years) relative to the anciens. natural grasslands and savannas of Africa. As a result, grassland communities in Africa are species rich, whereas chose in Central America are species poor, Spatial Components The distributions of most bird species are restricted both globally and lo- cally. Penguins are limited to the Southern Hemisphere, auks co the Northern Hemisphere, curassows (sce Figure 3-14) to tropical South America, mousebirds and turacos to Africa (Figure 20-9), and the Dodo FIGURE 20-9 The turacos, such as this Ross's Turaco, are limited to Africa(once upon a time) to the island of Mauritius. Such boundaries may be due to physical restrictions. The flightless Dodo, for example. was lim- ited co one oceanic island (see Box 21-1). More often, however. the lim- its of the distribution of a bird species, even on continents, are due to intrinsic limits of population growth. competitive replacement by another species, availability of resources, physiological tolerances, or some com- bination of these factors. Even at the local level of a county, parish, or shire perhaps, few birds use the full variety of hi ailable to them, Local diversity, or alpha diversity. corresponds to the scruetural com= plexity of the habitat. Plant species and physical form, amount of vege: tative cover, and details of habitat structure all contribute to habitat complexity and to the local diversity of birds. The physical structure of habitats provides courtship and display stations. nest sites, protection from predators, shelter fiom climatie stress, and food. The vertical distribution of vegetation roughly defines the variety of foraging opportunities and, hence, the variety of species that ean occupy a habitat (Figure 20-10). The Hubbard Brook forests of New Hampshire, for example, contain ground feeders such as thrushes, Ovenbirds, and Dark-eyed Juncos: tree-trunk feeders such as woodpeckers and nuthatches: general canopy feeders stich as Scarlet Tanagers. some vireos. and one species of flycatcher, which search widely in both deciduous and conif- erous trees: and specialized canopy feeders such as Blackburnian Warblers and Black-capped Chickadees, which tend to restrict their searches for food to the outer twigs of conifers (Holmes et al. 1979). A closer look reveals that variations in foraging behavior among insectivorous birds are directly related to variations in foliage height. tats a Bird species diversity 00 025 050 a 700 Foliage height diversity, FIGURE 20-10 The local diversity and relative abundance of bird species are correlated with the relative height and diversity of the foliage, illustrated here for sites in illinois (sites 1 through 4), Texas (sites 5 through 8), and Panama (sites 9 through 15). [After Karr and Roth 1971] 619 ‘COMMUNITIES,620 CHAPTER TWENTY Nest predation can determine local species compositions by favoring bird species that nest at different heights and in different microhabitats. In a 20-year seudy of birds associated with the conversion of old, aban- doned fields into shrublands and then into forest—called old-field eco- logical succession—on Long Island, New York, nine species of birds established nesting territories in a sequence that corresponded to the availability of nest cover (Lanyon 1981). Red-winged Blackbirds were the first to nest in a field, and Eastern Towhees were the last. As an open field converts into shrubland and then into forest, the availability of nest-supporting vegetation and the amount of shade for nests deter~ mine the suitability of the habitat for breeding. In chis example. the pro- cession of species of overlapping tenures caused the average number of territorial species present to peak 20 years after the field was last culti- vated (Figure 20-11), Species that coexist in seemingly homogeneous habitats, such as grass- lands or spruce forests, may segregate their niches more finely with re- spect to foraging station (Box 20-3). Consider the classic study of niche partitioning by warblers in northern spruce forests of Maine (MacArthur 1958). There, the Yellow-rumped Warbler feeds mostly in the under- story below 3 meters, the Black-throated Green Warbler in the middle ‘Average of four censuses in Oak Woodland, 54~62 years after clear-cutting a Succession not arrested territorial species) Succession not arrested Diversity of breeding birds (average number \ Succession arrested \, (shift back to grasses) ” succession arrested (Grasses dominant) 13 46 7-9 12 BIS 16-18 19-21 22-23, 29-31 32-34 35-37 38-40 41-43 44-45 ‘Age of old field (number of years after cultivation) FIGURE 20-11 The diversity of species nesting in old fields changes with the age of the field and the successional change of the vegetation from grass, to bushes, and Ultimately to trees. Successional change was arrested by cultivation. [From Lanyon 1981]HM bx 20-3 DISTANCE TO COVER DEFINES THE NICHES OF SPARROWS Distance to protective seal cover affects the variety ‘of sparrows that can coex- in open, simply structured grassland habitats (Pulliam and Mills 1977). In southeastern Arizona, four species of sparrows inhabit open grasslands that have scattered mesquite trees, which provide some protection from predators such as Prairie Falcons. From the sparrows’ point of view, this habitat offers concentric rings of increasing dis- tance from the nearest cover. The Vesper Spar- row stays closest to the mesquite trees (within 4 meters), the Savannah Sparrow feeds farther out (4-16 meters), the Grasshopper Sparrow still far- collared Longspur feeds far from the trees in the most open grassland. The behavior of these species when flushed corresponds to the risks of flying increasing dis- tances to cover. Vesper Sparrows fly quickly to nearby cover. Savannah Sparrows fly to an ex- posed perch the first time they are flushed and then to full cover if flushed again. Rather than face the risks of a longer flight, Grasshopper Spar- rows usually drop back into the grass when flushed, but they fly for cover if repeatedly flushed. Longspurs, however. either crouch to the ground to hide or fly off in tight flocks that help thwart predators, ther out (8-32 meters), and the Chestnut- story, and the Blackburnian Warbler at the tops of the same spruce trees. Sharing the midsection of the trees with the Black-throated Green War- bler, which searches the foliage for food. is the Cape May Warbler, which feeds on insects attracted to sap on the tree trunk, Sharing the tree tops with the Blackburnian Warbler, which feeds on the outer twigs and sal: lies out after aerial insects, is the Bay-breasted Warbler. which searches for insects close to the trunk, In Europe, different species of tits show parallel choices of their foraging stations (see page 629). Bird species diversity increases from that in a single local habitat {alpha diversity), to that across multiple local habitats (beta diversity), co large-scale regional diversity (gamma diversity). Habitat preferenc the variety of birds to increase as the number of distinct habitats increases. From the West Coast of the United States to the East ast, the num ber of species peaks in che Sierra, Rocky, and Appalachian mountains, where markedly different habitats at various altisudes increase beta diver- sity. Some counties in South America —Colombia, Ecuador, and Peru— owe their extraordinary avifiunas in part to the topographical and ecological diversity of the Andes. ase Tropical Diversity The density of species in an area increases as in the procession southward from North America (Figure 20-12) and peaks in the tropical forests of western Amazonia. Tropical forests support more than 300 resident species in a 3-square-kilometer study plot, compared with only 30 to 50 spi in temperate forests (Marra and Remsen 1997). A 400-square-mile sec~ sion (160,000 square miles in area) of the United States contains from 120 621622 (CHAPTER TWENTY 2s sp galas 5 Ay oben ace gif 128 142 119 190 61 f 48 71 67) 68 gh 42136, 148, 129_125 yo 146 148 139 129 TP ETS, T9136 147 187 119 121-16 aby 162 176 134 109 110. 109 IN wes 122 9391 Y 213 rao fisa 270 Ror FIGURE 20-12 The number of land-bird species that breed in geographical areas of 400 square miles in North and Central America decreases with increasing latitude [From Macarshur 1969] to 150 species of breeding land birds, but the same area in Central Amer- ica contains from 500 to 60K), In western Amazonia, more than 101) species can be found in suich an area, and more than 535 species. inchad- ing as many as 40 species of antbirds, can be found locally in a 100-heetare site (Terborgh et al. 1990). The variety of species also increases downhill from high to low alti- tudes. In the Andean country of Colombia alone, 47 species reside above the timberline, 270 at lower temperate altitudes. more than 480 at even lower. subtropical altitudes, and more than 1000 in the tropical lowlands. ‘The greater diversity of species in the Tropics compared with diver~ sity in the temperate zone is due in part co different and more varied food resources (Ricklefs and Travis 1980). For example, groups of fruiteaters— couicans, hornbills, barbets, trogons, cotingas, manakins, broadbills. and turacos—expand the dimensions of eropical bird communities. Parrotslarge and small consume a wide variety of seeds, fruits, and nectars that are not available in northern forests. Hummingbirds and tanagers abound in New World tropical forests, but only a few species live in the north. The diversity of insect sizes also is greater in the Tropics than in temperate-zone habitats, and che diversity of bill sizes of topical birds in- creases accordingly (Figure 20-13). Some families of stricly topical birds—pufibirds, motmots, antbirds, wood hoopoes. and jacamars—spe- cialize in large insects and small reptiles that are noc present in temperate ecosystems. Foraging specialists, such as ant followers and epiphyte probers, ako add to the diversity of bird communities in tropical regions. For decades, ornithologists thought that greater structural habitat com- plexity was responsible for the high diversity of species in tropical forests but this conclusion may not be so (Marra and Remsen 1997), Despite their obvious differences, tropical and temperate forests do not differ in overall heterogeneity or complexity. Direct comparison of the use of the forest structure by foliage-gleaning insectivorous birds of the underscory: however, exposes more specialized forest-floor foraging behaviors in the tropical species (Figure 20-14). Compared with cemperate species, each tropical species is more selective of particular horizontal and vertical habi- tats and has narrower niche breadth in foraging substrates and foraging heights. In addition, the tropical species overlap less in foraging. stations chan do comparable temperate-zone species. In general, tropical species tend to use a narrower range of habitats. They may be more specialized in their foraging behavior and less tolerant of climatic variation than heir temperate counterparts. Greater ecologi- cal specialization leads co tighter packing of species in local communities, smaller geographical distributions, and greater species richness. “ © 20 20 4-6 12- 20- 2 14 22 30 38 46 Bill ength (mm) Bill length (mm) FIGURE 20-13 ‘The existence of large-billed birds causes an increase in bird species diversity in the Tropics. Shown here are the bill lengths of insectivorous birds that breed in (A) tropical latitudes (8°-10" N) and (B) temperate latitudes (42°-44° N), [From Schoener 1971] 623 ‘COMMUNITIES,624 CHAPTER TWENTY ‘Temperate forest Foraging heights (m) ientucy Carolina Northern White-eyed Wilson's Warbler Wren Cardinal Vireo Warbler Tropical forest Foraging heights (m) Rufousstapped Ringed Black-faced __Chestnut-talled ‘Antthrush Antpipit Antbird Antbird Species FIGURE 20-14 Mean foraging heights of (A) temperate and (8) tropical forest species. [From Marra and Remsen 1997] Temporal Components Local assemblages of species change with time as well as with space or geography. Their composition fluctuates regular regularly with climate and resource availability. Migrants coexist only tem- porarily with residene species and change their community membership with the seasons. Disturbances due to deforestation or fire or to coloniza~ tion or extinction keep many habitats in flux. yy with che season and ir-Virtually all bird communities consist of both resident and nonresident species. Residents stay put and accommodate monthly changes in climate and food availability. Nonresidents are seasonal specialists that take advan- tage of predictable periods of local regional food abundan The mobil- ity of birds and the evolution of the migratory habit have made possible nonresidency and the opportunistic exploitation of variable envi (ee Chapter 10). nments Ephemeral resources attract opportunistic species. Temporary assem= blages of highly mobile birds may last hours, weeks, or years. Flocks of seabirds over shoal of fish, for example, are brief in duration (minutes or hours) and highly variable in species composition, Assemblages of hum- ingbirds and sunbirds at flowers feature high turnover of both individ- ual birds and species during the brief blooming periods of days or weeks (Figure 20-15). The regional diversity of small, short-billed hummingbirds depends ‘on their ability to circulate seasonally among locally blooming flowers 60 v0) 625 ‘COMMUNITIES, 975 (site 2) Bh Colden-winged Sunbird Malachite Sunbird 189 oroncy Sunbird W others 80 Weeks FIGURE 20-15 The composition of sunbird species feeding in fields of flowering mint in the Rift Valley of central Kenya changes weekly as a result of local competitive interactions, and varies through the years and over sites as a result of regional flower availability and colonization. [After Wolf and Gill 1980) July626 (CHAPTER TWENTY 1600 Transients, migrants ae 140 i erie 1200 1000 800 Spring reproduction 2 600 = B soo = Adults 8 40 Permanent residents 200 = ° lan Dee. Month FIGURE 20-16 Model of seasonal composition of the pine-forest bird community of Grand Bahama Island. Local numbers increase with the addition of young birds in the summer and again with the arrival of wintering migrants, which leave in April. [After Emlen 1980) (Feinsinger 1980). Only cwo short-billed species, the Copper-rumped Hummingbird and the Ruby Topaz, inhabit the small island of Tobago. where they must coexist year-round in the principal nonforested habi- tat, Seven similar species coexist on the larger island of Trinidad, where more diverse habitats enable segregations and seasonal patterns of local migration, Seasonal residents form a major part of most birel communities, In sea~ son, migrant shorebirds can dominate the bird life of coastal wetlands. The influx of wintering migrants from the north triples the number of species found in the open pine forests of Grand Bahama Island and in- creases the density of birds from 900 to 1600 per square kilometer (Elen 1980: Figure 20-16). In the tropical evergreen forests of western Mexico, the density of small foliage gleaners increases from an average of 2 to 64 birds per hectare with the arrival of the migrants (Hutto 1980). These extraordinary der sities of wintering birds result when large populations compress them- selves into small areas, Migrant North American land birds from 16 million square kilometers of breeding range squeeze into 2 million square kilo- meters of winter range in northern Central America and the West Indies. Competition Competition arises when one bird's use or defense of a resource mak that resource less available to other birds. Interspecific competition arises when birds of coexisting species require some of the same limited re- sources. The use or defense of those resources by members of one species reduces the availability of resources to members of another species. Re~ call that competition among members of one species reduces the rate ofpopulation growth in that species by limiting survival or reproduction. Competition among members of different species also can alfect popula tion growth, The competitive exclusion principle—a fundamental concept of ecol- ogy also known as Gause’s law after G. F. Gause, a pioneering Russian ecologist —states that two species with identical ecological niches cannot coexist in the same environment, Laboratory results support this concept. One species usually rephices another similar species when the two are forced to share the same environment in a laboratory. Similarly, competitive interactions are expected to be most intense within sets of ecologically similar species, or guilds, that depend on the same set of resources (Root 1967). Interference Competition Competition can be expressed as the overt aggressive displacement of in dividual birds, called interference competition, or as the reduction of the fecundity and survival of one species by another, called exploitative com- petition, In an unambiguous example of interference competition, large, dominant species of hummingbirds aggressively exclude other species trom the densest concentrations of flowers. Forced by dominant species to use other feeding grounds with fewer flowers, subordinate species quickly shit back co the best available Feeding grounds whenever it is posible to do so. Coexisting antbirds that gather at swarms of army ants in tropical forests hibit similar behavior (see Figure 11-8). So do Golden-crowned Spar- rows wintering in California. They aggressively restrict the use by Dark- eyed Juncos of foraging space near shrubs (Davis 1973). The junces increase their use of sites closer to protective cover when Golden-crowned Sparrows are removed experimentally but revert €o infrequent use when the sparrows return. Behavioral dominance forces sexes of the same species, as well as birds of different species, to use different habitats, Peter Marra (2000) tested the habitat prefer adult male redstarts mostly occupy mangrove habitats, where they estab lish and defend winter territories. Female redstarts mostly oceupy nearby low-quality secondary scrub habitats. The females that winter in scrub habitat suffer a measureable decline in body condition, which affects their migration back to breeding grounds in North America and, potentially, their future reproductive success (Marra and Holberton 1998). Before a series of removal experiments, yearling male and female red- starts arrived first and settled in both habitats, Adult males arrived next and increased the competition for territories. They displaced subordinate redstarts, especially females, from the high-quality mangrove study areas to the low-quality scrub habitat, Marra then removed equal numbers of established redstarts from both habitats to test the response of the remain= ing redstarts to the experimental vacancies. No longer excluded from the best habitat by dominant males, subordinate redstars, primarily females, quickly filled the vacancies in the mangroves first. ences of American Redstarts wintering on Jamaica, There, 627 COMMUNITIES,628 CHAPTER TWENTY Exploitative Competition Rather than overt aggression, most interspecific competition subtly de- presses a species’ survival or breeding success by reducing critical resources. Such competition is exploitative competition. Some of the best evidence of the effects of one species on the fee cruitment of another comes from research on Great Tits and Blue Tits. This research is an extension of the work on population regulation of the Great Tit (See Chapter 18). The local assemblages, or guilds, of titmice have been the focus of in~ tense research on the role of interspecific competition in bird communi- ties. The reduction of the food supplies by tits can affect the reproductive success of species outside their guild, such as Collared Flycacchers in Sw: den (Gustalison 1987). However. the details of competitive interactions between Great Tits and Blue Tits are of particular interest. These nwo species negatively affect each other in a variety of ways. both trivial and consequential (Dhondt 1989), Whereas competition between them affects their annual population dynamics, the balance of their interactions leads to coexistence rather than the exclusion of one species by the other. Recall that reproductive success, or fecundity, in Great Tits decreases as population density increases. In addition to being sensitive to the local densities of members of their own species, the fecundity of Great Tits is sensitive to the numbers of coexisting Blue Tits. Even though they tend to use different foraging stations, there is overlap. High densities of Blue Tits during the breeding season reduce food availability. Reduced food availability increases nestling mortality and causes fewer Great Tits to at- tempt second broods. In this way. high densities of Blue Tits reduce the reproductive output of Great Tits, The effects of Blue Tits on the repro- ductive output of Great Tits during the breeding season, however. are only temporary ones and have little final effect on the population density of Great Tits. Instead. their density is controlled primarily by winter sur- vival and the recruitment of juveniles (see Chapter 18). The reproduction of Blue Tits is neither density dependent nor greatly. affected during the breeding season by the local numbers of Great Tis. with one caveat: Great Tits control nest boxes if they are in short sup- ply. They may even kill Blue Tits in the process (Léhr] 1977). This form of interference competition is extreme. Of greater consequence is the reversal of competition between these two species outside the breeding season. Then Great Tits truly limit the number of Blue Tits in a woodlot by controlling the availability of roost holes. When the number of Great Tits in a population that depended on man-made boxes for roosting (as well as for nesting) was halved (by nar- 1g the nest entrances from 32 co 26 mm and thereby excluding the larger Great Tits), many more male juvenile Blue Tits were recruited into the woodlot in the autumn and subsequently joined the breeding popu- lation in the following year (Figure 20-17). The measurable competitive interactions between Great Tits and Blue Tits affect their annual successes. Although it does not result in the ex- ndity, survival. and population re~ rowi20 1c Gontrode a 8 4 & oly, 1 L (oa & olyet— 19671969 9 197319) FIGURE 20-17 Experimental demonstration of interspecific competition. When Great Tits were excluded from nest boxes (from 1976 to 1978, screened area) in the ‘experimental area at Gontrode, Belgium (white citcles), more Blue Tits established themselves there than in a control area at Zevergem (black circles) that had the ‘normal number of Great Tits. {After Dhiondt and Eyckerman 1980) clusion of one species from a woodlot by the other. competition favors different foraging behaviors. body gation by habitat is another potential consequence in some spe es, and nest-site preferences. Segre~ Ecological Segregation A corolla should y of the competitive exclusion principle is that competition nerease directly with overlap in the use of limited resources. Detti- mental ecological overlap may foster changes in behavior or morphology that reduce competition. Observed ecological differences between related species. therefore, may be the “ghosts of competition past” (Connell 1980). Local separation by habitat and feeding stations is typical of titmice. In Europe, the Great Tit, Blue Tit, and Marsh Tit inhabit broadleaf forests. The Crested Tit and Coal Tit live primarily in coniferous forest used by the other three species only as a suboptimal habitat. The species that live together feed in different places: Great Tits on the ground. Marsh Tits on large branches, and Blue Tits on the smaller twigs. Differences among European titmice in their feeding locations are associated with differences in body mass and beak size. Larger species feed at a lower level and on larger insects and harder seeds than do smaller species. Species that live in coniferous forests have longer and narrower beaks than those that |i broadleaf woods. Each species of European tit has a counterpart in North America (Fig- ure 20-18). However, only two of the North American species usually live together in the same habitat. In many areas, a small chickadee coex- ists with a large titmouse. which has different requirements. Where two species of small chickadees coexist, they inhabit different habitats. In New England. the Boreal Chickadee inhabits dark conifer stands, whereas the 629 COMMUNITIES,630 Tufted Tiemouse Great Tit Black-capped Chest backed Chickadee Willow Tit FIGURE 20-18 Certain species of North American (top row) and European (bottom row) chickadees and titmice act as ecological equivalents. [From Lack 1971] Black-capped Chickadee inhabits more open, mixed deciduous and conifer forest. On the West Coast. Chestnut-backed Chickadees and Black-capped Chickadees similarly separate by habitat. If competition actually restricts a species, one would expect shifts in the distribution, habitat use. or foraging behavior of a species when it is ited by a competitor, The absence of a competitor allows ecolog- ical release. On the Si in Islands of the Pacific Northwest, where there are no Black-capped Chickadees. the Chestnut-backed Chickadees in habit broadleaf forests used elsewhere by the Black-capped Chickadees. Shifis in habitat use in the absence of other species are well documented among European tits. Marsh Tits, for example. inhabit pine plantations only in Denmark, where Willow Tits are absent from this habitat, In Ire- land, Coal Tits feed regularly in the understory of evergreen forests in the absence of the Marsh Tits, Willow Tits, and Crested Tits that normally preempt this niche, not li Ecological and Character Displacement Darwin's finches of the Galapagos Islands provide a classic example of the apparent role of competitive exclusion and character displacement (Grant 1986). The adaptive radiation of these finches has propagated species with a variety of bill sizes that relate directly to seed sizes. Ground finches andcactus finches with distinctly different bill sizes inhabit every island, The differences in the average bill size of coexisting species are consistent with the hypothesis of interspecitic competition for food. Species with similar- sized bills replace one another on various isinds, and the bills of various species are more alike when they do not live cogether (Figure 20-19). Simple ecological displacements as a result of competition should lead to evolutionary reinforcement in the form of morphological character dis- placement, or enhanced differences (in size, for example) where nwo species coexist. On the Swedish ishind of Gotland, in the absence of larger competitors—specifically. Crested Tits and Willow Tits—Coal Tits are larger than on the mainland (Alatalo et al. 1986). Their larger size on Po a a Small Medium Large 1 7 . * santa Cruz no | [ i: 404 Daphne Major 20 Los Hermanos 3 5 7 Bill depth (mr FIGURE 20-19 Thiee species of ground finches that coexist on the Galapagos island of Santa Cruz have bills of different depths (top), which enable them to feed on different seeds. Certain islands, such as Daphne Major and Los Hermanas, have only fone species. in the absence of other species, the solo populations on these islands evolve intermediate-sized bills, as the lower two graphs show. [After Grant 1986] 631 COMMUNITIESHB 0x 20-4 COMPETITION AFFECTS THE USE OF FORAGING SITES BY TITS The ornithological litera- sexe ture contains many, often anecdotal, observations of apperent niche shifts in the absence of a com- petitor. Controlled experimental demonstrations with the use of free-living birds in natural popu- lations, however, are few. One exception is the study of foraging tits and Goldcrests in the conif- erous forests of central Sweden (Alatalo et al 1987). These small birds exploit nonrenewable in- sect and seed resources in their group territories during the long, cold winter. Two smaller and so- cially subordinate species, the Coal Tit and the Golderest, forage on the outermost branches and needles, whereas two larger and dominant, species, the Willow Tit and the Crested Tit, forage inside the trees In this experiment, the ornithologists removed the Coal Tits and Goldcrests from three of six flocks to test whether Willow Tits and Crested Tits would change their foraging behavior. They did. In late winter, Crested Tits moved farther out on the spruce branches in experimental flocks than in control flocks. Willow Tits did so in pine trees but not in spruce trees ‘The Swedish team concluded that exploitative competition directly based on food depletion, without any interference, influences the use of foraging sites by tits that coexist in coniferous forests 632 Gotland coincides with a shift in foraging niche from the outside of the tree and on needles, where small size is advantageous, to the inner parts of the pine trees (Box 20-4). The abrupt replacement of one species by another at various altitudes in the Andes and in New Guinea also suggests that competition from one species limits the distribution of another. New Guinea birds with well- defined altitudinal distributions replace one another abruptly at various el- evations (Figure 20-20). The range of elevations occupied by a species seems to depend on the presence or absence of related species. For ex- ample, the Red-fanked Lorikeet is confined co low elevations in regions with one of che two highland species—either the Red-fronted Lorikeet or the Red-chinned Lorikeet. In regions in which the Red-flanked Lori- keet is alone, however, it ascends to high elevations. Conversely. in re~ gions in which either the Red-tronted Lorikeet or the Red-chinned Lorikeet is alone, it descends to sea level. Similar patterns of altitudinal replacement are found in the mountains of eastern North America, where as many as five species of thrushes nest in the high mountain forests of New England (Noon 1981). Veeries share. low elevations with the Wood Thrush and are replaced at higher el tions by the Swainson’s Thrush and the Bicknell’s Thrush. In the Smoky mountains, where the Swainson’s Thrush and the Bicknell’s Thrush are absent, the Very shitis to higher elevations and overlaps only slightly with the Wood Thrush ac low elevations. Such studies of ecological replacement assume weakly that the loca~ tions are identical except for the presence or absence of the purported competitors. At best. the process of competitive replacement must be in- ferred from an observed pattern. The role of interspecific competition is