Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 11
The Scientific Bases for Preservation of the Mariana Crow 2 History The aga occurs only on 2 islands in the Mariana archipelago: Guam and Rota. Although these islands are similar in their origins and overall ecology, they differ considerably in some of their physical characteristics and in their ecological and political histories. To place the current status of the Guam and Rota populations of the aga within a historical framework, the initial sections of this chapter describe the physical and climatic characteristics of both islands, provide an overview of the ecological changes that have occurred from human colonization to the present, and summarize recent avian population trends and extinctions on each island. Later sections focus specifically on the aga, providing an overview of its natural history, probable causes of decline, population demography, and past and present conservation measures. DESCRIPTION OF GUAM AND ROTA Micronesia encompasses an area of 7.5 million square kilometers in the western Pacific that stretches from the equator to 20° N and from the International Date Line to 130° E. Although more than 2,000 islands are in Micronesia in the Gilbert, Marshall, Caroline, and Mariana islands, the total combined land area of all the islands is less than 2,000 km2 (Engbring and Pratt 1985; Engbring and others 1986), which is smaller than the state of Rhode Island (3,144 km2). The Mariana archipelago is the most northerly island group in Micronesia, lying midway between Japan and New Guinea and about 2,600 km east of the Philippines. Volcanic in origin, the 15-island archipelago has a total land area of about 640 km2 (Baker 1951) and extends over 675 km from Guam (13° 30´ N, 144° 45´ E) in the
OCR for page 12
The Scientific Bases for Preservation of the Mariana Crow south to Uracas (20° 31´ N, 144° 54´ E) in the north (Figure 2-1). The islands generally decrease in size from south to north, and some of the northern islands are actively volcanic (Reichel and Glass 1991). Guam is the most southerly and populous of the Mariana Islands and is administered as a US territory. All the other islands in the archipelago are part of the Commonwealth of the Northern FIGURE 2-1 Mariana Archipelago. Source: Adapted from Engbring and others 1986.
OCR for page 13
The Scientific Bases for Preservation of the Mariana Crow FIGURE 2-2 Guam. Dashed line indicates the approximate area of transition from the northern plateau to the southern mountainous area. Source: Adapted from Jenkins 1983. Mariana Islands (CNMI). Saipan is the most populous island of the CNMI, followed by Rota and Tinian. Agiguan and the ten islands north of Saipan are small and now uninhabited (Engbring and Pratt 1985; Engbring and others 1986). Guam (Figure 2-2) is the largest island (550 km2) in the Mariana archipelago, and the most developed island in Micronesia. About 45 km long and 6–13 km wide, it is separated into a northern, uplifted limestone plateau and a southern mountainous region of volcanic origin (FWS Figure 2-1, Figure 2-2 1990). The northern plateau averages 100–200 m in elevation, whereas the highest mountain (Mount Lamlam) in the south is 405 m. The island is largely surrounded by fringing reefs except for small areas along the steep limestone cliffs of the northern plateau. Because of the porous coralline limestone soil, the northern plateau lacks permanent streams and marshes. In contrast, streams, ponds, and marshes are common in the south, where the volcanic soil is nonporous. Guam is believed to have been covered by tropical broadleaf forests before human colonization (Fosberg 1960). Today, however, the island is greatly dissected by fields, savannas, urban and suburban areas, commercial facilities, military installations, and roads. Although the northern plateau still supports relatively diverse secondary "limestone" and scrub forest, primary forest is largely restricted to the cliffline and immediately adjacent areas (Conry 1988; Engbring and Ramsey 1984; Jenkins 1983; Stone 1970). In the south, the uplands are dominated by savannas and the sheltered valleys by ravine forest (Stone 1970). The extensive savannas probably resulted from clearing by the island's first inhabitants and a burning regime (Fosberg 1960; Mueller-Dumbois 1981).
OCR for page 14
The Scientific Bases for Preservation of the Mariana Crow FIGURE 2-3 Island of Rota, Commonwealth of the Northern Mariana Islands. The majority of the island's inhabitants reside in the villages of Songsong and Sinapalo. Source: Committee-generated from USGS topographical maps. Rota lies about 49 km north of Guam (Figure 2-1). It is about 20 km long and 4–8 km wide, with an area of about 85 km2 (FWS 1990). Much of the island's coastline is rocky and steep, although there is a substantial coral sand beach inside the fringing reef along most of the northern shoreline. The village of Songsong and two commercial ports are on the isthmus of a prominent, narrow peninsula that juts to the southwest from the western side of the island (Figure 2-3). The topography of the western half of the island is dominated by a large uplifted plateau (12 km2) known as the Sabana at an elevation of about 450 m. The plateau is mostly open grassland as a result of past agricultural and phosphate-mining activities, and it is still used extensively for agricultural production. Steep cliffs drop from the north, west, and south sides of the Sabana to narrow coastal shelves and then to the ocean. The northeastern side of the Sabana slopes more gently to a large shelf with an elevation of about 150 m, which covers the eastern half of the island. This area and adjacent coastal shelves were cultivated for sugar cane during the Japanese administration of the island, in 1914–1944 (Engbring and others 1986). Although many of those areas have reverted to scrubby second-growth forest since World War II, they are now being progressively cleared for agricultural homesteads, grazing, and tourist development. The native vegetation remains less disturbed on Rota than on Saipan or Tinian because much of Rota's terrain is precipitous and unsuitable for large-scale agriculture and because it suffered less damage during World War II. Vegetation and habitat types on Rota are similar to those on Guam (Engbring and others 1986; Fosberg 1960). The Mariana Islands are warm and humid with only moderate seasonal or daily variation in temperature. Temperatures usually range from 25 to 32°C during
OCR for page 15
The Scientific Bases for Preservation of the Mariana Crow the day and from 24 to 27°C at night. The mean annual temperature is about 27° C on Guam and slightly lower on Rota (Eldredge 1983). Annual rainfall is about 220 cm; three-fourths of the total falls during the wet season (July–November) (Engbring and Pratt 1985; Engbring and Ramsey 1984; Engbring and others 1986; Stone 1970). Both temperature and rainfall decrease from south to north within the Mariana archipelago (Butler 1991). Winds can blow from any direction during the wet season but are generally light. The dry season (January–May), which can vary considerably in length and intensity, is characterized by steady easterly and northeasterly trade winds (Engbring and Ramsey 1984). Major annual variations in weather patterns are related primarily to the distribution of rainfall and the frequency of typhoons. Typhoons are common in this region of the Pacific during the wet season, and there is a 1 in 3 chance that a typhoon will pass close enough to affect the inhabited islands of the Marianas in any particular year (NOAA 1982). ECOLOGICAL HISTORY AND FAUNAL CHANGE As in other isolated and small island groups of the Pacific Basin, the prehuman fauna of the Mariana Islands was shaped primarily by infrequent immigration and colonization events (Diamond and Mayr 1976; MacArthur and Wilson 1967). Most of the endemic land birds appear to have been derived from colonizations from the Philippine and New Guinea regions (Baker 1948, 1951). Endemic mammals included only fruit bats (Pteropus mariannus and P. tokudae) and the sheath-tailed bat (Emballonura semicaudata). Knowledge of the historical reptilian and invertebrate fauna is poor, but it seems likely that the endemic fauna was restricted to the blind snake (Rantophotylops brahminus), several species of skinks and geckos, and a number of partulid snails, and numerous insects (Rodda and Fritts 1992; Wiles and others 1990, 1995). The Chamorros colonized the Mariana Islands from the Indo-Malaysian region about 4,000 years ago (Craib 1983). A variety of plant and animal commensals were introduced at that time, including such food plants as rice (Solenberger 1967) and such animals as the red jungle fowl (Gallus gallus) and rats (Rattus exulans and possibly R. rattus ). The southern "high islands" of Guam, Rota, Agiguan, Tinian, and Saipan are believed to have been completely forested at the time of colonization (Fosberg 1960). As the human population increased, slash-and-burn methods of cultivation of food crops led progressively to the development of open savannas in upland areas (Mueller-Dumbois 1981). Recent paleontologic excavations on Rota, Agiguan, and Tinian have revealed that, as elsewhere on oceanic islands, numerous species of land birds and resident breeding populations of sea birds were extirpated in prehistoric times as a result of overharvesting by humans and predation by introduced predators (Atkinson 1984; Steadman 1995a). Excavations on Rota have yielded bones of 22 resident avian species in the late Holocene (last 5,000 years), of which 13 are no longer on the island
OCR for page 16
The Scientific Bases for Preservation of the Mariana Crow (Steadman 1992, 1995b). It is expected that the number of extinct and extirpated species will increase as additional data are obtained. Magellan's "discovery" of the Mariana Islands in 1521 ushered in an era of ecological change that continues today. In 1521, at least 13 islands were inhabited by the native Chamorro people, and their population is estimated to have been between 30,000 and 40,000. By the early 1700s, the Chamorro population had been reduced to less than 4,000 by epidemics, war, and their forced relocation by the Spanish from northern islands to Guam and other Spanish colonies (Carano and Sanchez 1964; Hezel and Berg 1981). Although the Spanish presence in the Marianas was sporadic until a permanent mission was established on Guam in 1668, the Marianas were visited regularly for water and provisions by galleons traveling between the Americas and the Philippines after 1565 (Hezel 1982). A number of plants and animals were introduced to the islands from the Philippines and Mexico (Butler 1991; Engbring and others 1986; Hezel 1982; Seale 1901) as a result of the galleon trade (for example, maize, Zea; goat, Capra sp.; cattle, Bos sp.; pig, Sus scrofa; and deer, Cervus unicolor) and later increased commerce with the Philippines (for example, Philippine turtle dove, Streptopelia bitorquata; blue-breasted quail, Coturnix chinensis; water buffalo, Bubalus bubalis; Norway rat, Rattus norvegicus ; house mouse, Mus musculus; monitor lizard, Varanus indicus; domestic cat, Felis cattus; and domestic dog, Canis familiaris). The introduced ungulates had an especially dramatic effect on the native forest and understory, and introduced predators undoubtedly had an adverse effect on the native vertebrate fauna. After the Spanish-American War (1898), Spain ceded Guam to the United States, except for a brief period of Japanese occupation in World War II (1942–1944), Guam has remained a US territory. In 1899, Spain sold all the other Mariana Islands to Germany. The period of German administration of the northern islands, 1899–1914, saw relatively few ecological changes, although there was a substantial immigration from the Carolines to some islands north of Guam and an increase in the development of coconut plantations and orchards (Engbring and others 1986). After World War I, Germany relinquished its control of the northern islands to Japan under a mandate of the League of Nations. The period of Japanese administration of the northern islands, 1914–1944, witnessed extensive selective logging, phosphate mining, and agricultural development and an associated increase in immigration of Japanese, Korean, and Okinawan workers. On Saipan, Tinian, Agiguan, and Rota, native forest was cleared on nearly all flat, tillable areas for sugar-cane production (Baker 1946; Hezel and Berg 1981; Mayr 1945). For example, on Rota, a railroad was built that ringed the island, a sugar-cane mill was constructed in Songsong, and extensive phosphate mining was undertaken on the Sabana. Several conspicuous introductions of plants and animals occurred during this period (Baker 1951; Engbring and others 1986), including the ironwood (Casuarina equisetifolia), Formosan koa (Acacia confusa), flame
OCR for page 17
The Scientific Bases for Preservation of the Mariana Crow tree (Delonix regia), black drongo (Dicrurus macrocercus), marine toad (Bufo marinus), and African land snail (Achatina fulica). The drongo, introduced to Rota from Formosa in 1935, colonized Guam in 1960 and is now well established on both islands (Jenkins 1983). When US troops invaded the Marianas during World War II, the fighting, military construction, and influx of large numbers of men and material had many ecological effects on the southern islands (Baker 1946; Mayr 1945). On Saipan and Tinian, much of the remaining forest was damaged by the fighting. On Guam, habitat destruction associated with the fighting was less because organized Japanese resistance did not last as long. However, as on Tinian, large portions of Guam's native limestone forest were cleared for military construction (Fosberg 1960). Neither Rota nor Agiguan was invaded, but large areas of both islands were heavily bombed before Japanese surrender (Baker 1946; Butler 1991). The military seeded large areas with tangantangan to prevent erosion after invasion, and large areas are now overgrown with this tree and an assorted mixture of native and exotic vegetation (Engbring and others 1986). Since the war, a number of additional species have been either accidentally or intentionally introduced to Guam (Engbring and others 1986), including the brown tree snake (Boiga irregularis), musk shrew (Suncus murinus), black francolin (Francolinus francolinus), Eurasian tree sparrow (Passer montanus), and chestnut mannikin (Lonchura molacca). The brown tree snake apparently arrived from Melanesia in the late 1940s or early 1950s in war material that was being consolidated on Guam by the military (Rodda and others 1992). The brown tree snake, tree sparrow, and musk shrew have since spread to other islands in the archipelago. Economic development on Guam has continued steadily since World War II as a result of a continued military presence and increasing tourism. Development on the other islands slowed after the war but has increased in recent years because of economic prosperity in Japan. Farming and cattle ranching have expanded on both Tinian and Rota. On Rota, a program of providing free homesteads to natives of the islands has stimulated agricultural activities. As on Guam, tourism and associated development have expanded on Saipan and Rota because they are relatively close to Japan. Despite development, large areas of native and secondary forest remain on both Guam and Rota (Engbring and Ramsey 1984; Engbring and others 1986; Savidge 1984, 1988). Although the Micronesian megapode (Megapodius laparouse) was extirpated from Guam at the turn of the century, the native avifauna of both islands that survived until the century remained largely intact through the 20th century until the 1960s. The resident avifauna of Guam included 28 species (17 land and water birds, 4 sea birds, and 7 introduced land birds). The resident avifauna of Rota included 23 species (11 native land and water birds, 6 sea birds, and 6 introduced land birds) as shown in Table 2-1. Since the middle of this century, Guam's avifauna has undergone a catastrophic decline, with the loss of 15 resident species (Table 2-1). The extinction of the Mariana mallard (Anas oustaleti) and
OCR for page 18
The Scientific Bases for Preservation of the Mariana Crow TABLE 2-1 Recent Resident Birds of Guam and Rota Common Name (Scientific Name), Chamorro Namea Guam Rota Mariana crow (Corvus kubaryi), aga Rb Rb Black drongo (Dicrurus macrocercus) I I Black francolin (Francolinus francolinus) I — Blue-breasted quail (Coturnix chinensis) I — Bridled white-eye (Zosterops conspicillata), nossa Rx Rb Brown booby (Sula leucogaster) Rx Rb Brown noddy (Anous stolidus) Rx Rb Cardinal honeyeater (Myzomela cardinalis), egigi Rx Rb Chestnut mannikin (Lonchura molacca) I I Collared kingfisher (Halcyon chloris) — Rb Common moorhen (Gallinula chloropus) Rx —b Eurasian tree sparrow (Passer montanus) I I Guam flycatcher (Myiagra freycineti) Re — Guam rail (Gallirallus owstoni), koko Reb — Island swiftlet (Collocalia vanikorensis), yayaguak Rb Rx Mariana fruit-dove (Ptilinopus roseicapilla), totot Rx Rb Marianas mallard (Anas platyrhynchos oustaleti) Rec — Micronesian kingfisher (Halcyon cinnamomina), sihek Red — Micronesian megapod (Megapodius laparouse) Re R Micronesian starling (Aplonis opaca), sali Rb Rb Nightingale reed warbler (Acrocephalus luscinia) Rx — Pacific reef heron (Egretta sacra) Rb Rb Philippine turtle dove (Streptopelia bitorquata) I I Red jungle fowl (Gallus gallus) — I Red-footed booby (Sula sula) — Rb Red-tailed tropic bird (Phaethon rubricauda) — Rb Rock dove (Columba livia) I I Rufous fantail (Rhipidura rufifrons), chichirika Rx Rb White tern (Gygis alba) Rb Rb White-browed rail (Porzana cinerea) Rx — White-tailed tropic bird (Phaethon lepturus) Rx Rb White-throated ground dove (Gallicolumba xanthonura),paluman apaka Rx Rb Yellow bittern (Ixobrychus sinensis) Rb Rb NOTES: R = resident I = introduced — = absent b = breeding e = extinct x = extirpated a When known, the Chamorro name for the bird is included. b Species extirpated from Rota during prehistoric times. c Species has been seen on Saipan. d Species survives in captivity. Source: Information assembled from Engbring and others 1986; FWS 1990; and Reichel and Glass 1991.
OCR for page 19
The Scientific Bases for Preservation of the Mariana Crow extirpation of the white-browed rail (Porzana cinereus) during the 1960s were apparently caused by excessive hunting and the drainage and development of the island's wetlands (Engbring and Ramsey 1984; Savidge 1984). The loss of the remaining species during the late 1970s and 1980s has been related primarily to predation by the brown tree snake (Engbring and Fritts 1988; Reichel and others 1992; Savidge 1987, 1988). Both introduced and native birds have been affected (Corny 1987), and most species survive only in reduced numbers. In contrast, Rota's resident avifauna has remained relatively stable, with the loss of only a single species—the island swiftlet (Collocalia vanikorensis)—in the late 1970s (Engbring and others 1986). Even so, declines in the Rota populations of the bridled white-eye (Zosterops conspicillata) and rufous fantail (Rhipidura rufifrons) have been noted and are of special concern (Engbring and Pratt 1985; Engbring and Ramsey 1984; FWS 1981; Pratt and others 1979; Ralph and Sakai 1979). Predation by the introduced black drongo has been suggested as a principal cause of the decline of these small-bodied passerines (Craig and Taisacan 1994), but definitive studies have not been completed. The causes of this decline remain unknown. The apparent decline of the aga population on Rota is presumably associated with habitat loss related to development and perhaps other causes. NATURAL HISTORY OF THE AGA The aga is coal black, weighs about 250 g, and is about 38 cm long; males are slightly larger than females (Baker 1951). This crow has the lax plumage characteristic of island corvids (Goodwin 1986), with a slight greenish-black gloss to the head and back, bluish-black tail, and dull underparts (Baker 1951). Immature birds resemble adults but have less gloss and browner wings and tail. The evolutionary relationship of the aga to the corvids is poorly understood. No genetic studies have been conducted on which to base a phylogeny. Possible founding ancestors could have come from Asia, Indonesia, Australia, New Guinea, or the Solomon Islands. On the basis of proximity and morphology, the slender-billed crow (Corvus enca) from Indonesia is probably its closest relative (Baker 1951; Goodwin 1986). Aga are most abundant in mature forest (Baker 1951; Engbring and others 1986). Emergent trees are especially important in perch and nest sites on Guam (Morton 1996a). Avoidance of human habitation was noted by Baker (1951) and confirmed by recent studies (FWS 1990; Morton 1996a); crow sightings decline with increasing proximity to roads and aircraft runways (Morton 1996a). Aga eat a variety of plants and animals. Principal foods are insects (grasshoppers, mole crickets, praying mantis, earwigs, seasonally abundant caterpillars [Pericyma crugeri], and lepidopteran larvae), small vertebrates (including immature rats), and invertebrates (hermit crabs, lizards, bird eggs, and nestlings), fruit, seeds, flowers, buds, foliage, and bark (Beaty 1967; Jenkins 1983; Lusk 1996, Michael 1987; Tomback 1986). They forage in the canopy and under-
OCR for page 20
The Scientific Bases for Preservation of the Mariana Crow growth and on the ground (Jenkins 1983; Tomback 1986). Information is lacking on seasonal variation in diet because most observations of foraging have been made during the breeding season. Aga typically occur in pairs or groups of 2–5 (Jenkins 1983; Tomback 1986). Groups are most likely families consisting of a pair and their recently fledged offspring (Jenkins 1983; Tomback 1986). Occasionally, larger groups (of up to 14) are observed (Jenkins 1983). They appear to form monogamous pairs, but detailed studies of marked individuals have not been conducted. A variety of vocalizations are used, including contact calls among group members during foraging (Jenkins 1983; Michael 1987; Tomback 1986). Pairs vocalize quietly at their nest with rambling ''monologues" (Tomback 1986). Alarm calls are given in response to human activity and the presence of potential predators. Juveniles beg from their parents with squalling vocalizations (Morton 1996a; Tomback 1986). Breeding can occur throughout the year, but a distinct peak occurs in the beginning of the dry season from October to December. Both parents construct the nest and care for the young (Jenkins 1983; Morton 1996a: Tomback 1986). Females do most of the incubation and brooding; males provision the females and nestlings. Incubation lasts 21–23 days (FWS 1990), and fledging occurs when nestlings are 36–39 days old (Morton 1996a). Clutch size is one to three eggs, but usually only one or two fledgings are produced (Morton 1996a). Reasons for nest failure (Table 2-2) include predation by brown tree snakes and monitor lizards (Aguon and others in press; Engbring and Fritts 1988; Jenkins 1983; Savidge 1987, nest abandonment due to unknown causes, egg infertility or nonviability (Aguon 1993; DAWR 1991, 1992; DFW 1990; Grout 1993; Wiles and others 1994), and disruption of breeding behavior by black drongos, other crows, and human activity (Grout 1993; Morton 1996a). A typical aga nest consists of an outer platform of sticks (mean diameter, 37 cm) and an inner cup of vines, rootlets, and fiber (mean diameter and depth, 13 and 7 cm, respectively [Lusk and Taisacan 1996]). As with most crows, aga TABLE 2-2 Fate of Nests on Guam After Development and Installation of Electric Barriers Number of Nests 1992–93 1993–94 1994–95 1995–96 Found 28 40 27 8 Successful 0 2 0 2 Apparently abandoned 13 26 17 3 Apparently predated by snakes 4 0 1 0 Failed to hatch 3 4 7 1 Destroyed by researchers 0 2 0 1 Apparently failed for other reasons 3 1 2 1 Fate not known 5 5 0 0
OCR for page 21
The Scientific Bases for Preservation of the Mariana Crow commonly build and abandon multiple nest platforms before constructing a full nest, which can be constructed in as little as a week. New nests are constructed by a pair throughout a season after nest failure (Morton 1996a). Renesting 5 times within a season is not uncommon, and 7–10 successive nests have been documented (Morton 1996a). Prolific renesting by aga is likely an adaptation to their variable environment, where typhoons destroy many nests. It is an important characteristic that bodes well for restoration, in that more eggs could be harvested from a single pair of aga than from a pair of 'alala. COMPARISON WITH THE 'ALALA The general habits of the 'alala and aga are similar. Both occupy mature forest and forage on a variety of introduced and native plants, vertebrates, and invertebrates (Jenkins 1983; Sakai and others 1986, 1990; Giffin and others 1987; Morton 1996a). Both appear to be monogamous, but also form small family groups. Large groups have been documented in the aga (Jenkins 1983), and reported in the 'alala (Unger 1996). As is the rule for corvids, male and female 'alalas and aga cooperate to raise a brood. Development to independence is slow in both species; parental care is extensive and can last up to a year (Banko 1996). Reproductive potential is much higher in the aga than in the 'alala. The 'alala rarely fledges more than 1 young from a nest and usually renests only once per year after nest failure (Banko and Banko 1980). The aga also typically fledges one young, but a pair will renest 5–10 times after nest failure (Morton 1996a). Although both species are threatened with extinction, the reasons for their endangerment differ in important ways. Predation appears much more important as a limiting for the aga, whereas disease might be the primary limiting factor for the 'alala (Atkinson 1993; Jenkins and others 1989; vanRiper and others 1986). Food does not appear to limit their species, whereas senescence, inbreeding, and infertility might be affecting remnant populations of both. Aga appear slightly more gregarious than 'alalas. DEMOGRAPHY OF THE AGA This section describes what is known about the demography of the aga on the basis of analyses of data collected during field studies by the DAWR and FWS. Estimates of nesting success and adult survivorship are developed to help infer the causes of the population decline and for use in population projections. Nesting success of the aga is different on Rota than on Guam. Although little field work has been conducted on Rota, John Morton (unpublished data) found 24 nests from 1993–1996. Of those 24, at least ten nests were known to fledge 14 young, for an average of 1.4 young per successful nest. One nest was destroyed by an unknown predator, but studies were not sufficient intensity to be sure of
OCR for page 28
The Scientific Bases for Preservation of the Mariana Crow there is no indication that the crows on either island have been adversely affected by pesticides, no thorough studies have been completed. Competition As noted above, seven species of land birds have been introduced to Guam, and six have been introduced to Rota (Reichel and Glass 1991). Of these, the only species that could be considered to be a potential competitor is the largely insectivorous black drongo. The drongo was introduced to Rota in the 1930s and spread naturally to Guam in 1960 (Savidge 1987). The density of black drongos has remained lower on Guam than on Rota, presumably because of the high rate of nest predation by the brown tree snake on Guam (Beck 1996). Maben (1982) concluded that drongos were not competing directly with native species, including the crow, because of differences in habitat use and foraging techniques. However, drongos have been seen harassing nesting crows (Grout 1996). Differences in nest placement by crows on Rota (in the subcanopy) and Guam (in the canopy) have been ascribed to differences in drongo densities, and hence harassment rates, between islands (Tomback 1986). Additional effects of such interactions on breeding and incubating crows remain unknown. Disease Extensive disease and sentinel studies (studies where data on animals exposed to contaminants in the environment are regularly and systematically collected and analyzed to identify potential health hazards to other animals or humans [NRC 1991]) conducted on Guam in 1982–1984 did not reveal any evidence that the declines in the native avifauna were caused by infectious disease (Savidge 1987). Although no similar detailed studies have been conducted on Rota, health examinations on captured crows, bridled white-eyes, and Mariana fruit doves have not revealed pathogenic organisms or parasites (Olsen 1996). Postmortem examinations have been done opportunistically in the past. A more concerted effort needs to be made to recover all dead embryos, chicks, or fledgings for complete postmortem examinations. Pathology protocols need to be developed in advance and applied consistently in order to maximize the usefulness of the data. Having accurate, comprehensive pathology data will help answer questions regarding causes of mortality in various age classes and the significance of any subclinical infections or nutritional problems. Predation Various predators of birds have been introduced to the Mariana Islands, including dogs, cats, rats, monitor lizards, and the brown tree snake. All except the brown tree snake are present on all the major islands of the Mariana archipelago
OCR for page 29
The Scientific Bases for Preservation of the Mariana Crow and have coexisted with the native avifauna for a long time without causing important population declines or extinctions. As discussed below, the population declines of Guam's native avifauna, including the crow, appear to be caused directly by predation by the extremely numerous brown tree snake (Engbring and Fritts 1988; Reichel and others 1992; Savidge 1987). Because the brown tree snake has now been detected repeatedly on Saipan and other Pacific islands (Fritts 1988), a series of unintended experiments are now under way. Food Resources Recent increases in the number of nonlaying pairs and in the production of nonviable eggs in the Guam aga population could possibly be related to food or micronutrient resources and foraging energetics. Clearly, the brown tree snake has dramatically altered the fauna of Guam. Some food resources previously used by the aga such as the eggs and nestlings of other forest birds, have been eliminated by the snake. Other important resources, such as skinks and geckos, survive, but at substantially reduced population levels because of snake predation. A recent study of lizard populations on 102-m plots with snakes (36 snakes/ ha) and without snakes indicated that overall skink and gecko densities and biomass were about 33% less with predation by snakes (density, 13,210 vs 19,650 lizards/ha; biomass, 35 vs 53 kg/ha). Although those differences are substantial, there still seem to be sufficient prey for the aga. Furthermore, given the wide variety of food items regularly consumed by crows, it seems unlikely that food resources could be limiting the population, except perhaps for brief periods after typhoons. However, because the brown tree snake does not appear to occur on Rota, the food- shortage hypothesis could easily be tested by comparing the rate at which males feed their mates on the nest on the 2 islands. Inbreeding Because the Guam population of aga is small and has been declining for years, it might be hypothesized that reduced productivity is the result of inbreeding depression (Ralls and others 1980, 1986; Schoenwald-Cox and others 1983; Soulé 1987). That possibility seems slight, however, considering that although the genetic diversity of the Guam population is greater than that of the Rota population, the Rota population does not appear to be experiencing any reductions in egg fertility, egg hatchability, or chick survival. Again, comparative information is needed to test this hypothesis. Senescence In the last few years, breeding pairs of crows on Guam have produced a number of infertile or nonviable eggs (nonviability of eggs can be caused by, for
OCR for page 30
The Scientific Bases for Preservation of the Mariana Crow example, infertility, inbreeding depression, thin or cracked egg shells, poor parental nutrition, or inadequate incubation) or have constructed nests that lacked eggs when examined. That has resulted in the suggestion that many of the crows in the Guam population are exhibiting reproductive senescence—a suggestion based on the premises that the population has been progressively declining for a number of years, that recruitment to the population has been extremely low, and that most of the birds in the population are reaching the end of their life span. Declines in reproductive performance in older birds have been documented in a number of species (Clutton-Brock 1988; Holmes and Austad 1995; Martin 1995; Saether 1990). However, senescence can be manifested in a variety of forms (such as egg fertility, clutch size, laying dates, and renesting potential), and these effects usually vary greatly between individuals (Clum 1995; Fowler 1995). Comparative information on the black-billed magpie (Pica pica) (Reese and Kadlec 1985) and the rook (Corvus frugilegus) (Røskaft and others 1983) indicates that older females typically are more vigorous breeders, laying larger clutches earlier in the breeding season, hatching more nestlings, and raising more fledglings than younger females. However, both male and female Florida scrub jays show a drop in survival and reproductive output with extreme age, although a great deal of variation exists between individuals. Some females continue laying until their death; others lay smaller clutches, are less likely to renest, experience slightly decreased hatching success, lay mutant eggs, or cease laying altogether but act territorial and build nests (Fitzpatrick 1996). Similar patterns (egg infertility, reduced renesting, and nest construction without egg-laying) have been observed in the small remnant wild population of 'alala. However, as noted by Clum (1995) and Cézilly and Nager (1996), in species with long-lasting pair bonds, individuals might enhance breeding performance by repeatedly breeding with the same partner; hence, an apparent effect of senescence could result partly from an overrepresentation of ''new" pairs in older age classes. Few individuals in the Guam population are banded, so it is difficult to discriminate between the alternative hypotheses. Because of the difficulties in monitoring nests in the field, it is also possible that some cases of "egg infertility" actually represent early embryo deaths brought on by poor incubation by adults as a result of disturbance by snakes. Similarly, cases of "nonlaying" might simply represent snake predation of eggs soon after laying. Summary In summary, the principal cause of the decline of the aga population on Guam appears to be predation by the brown tree snake, although the reproductive effects of advanced age, frequent remating due to natural deaths, nest abandonment, or some as-yet-unidentified factor cannot be ruled out as a more recent problem. On Rota, the decline of the crow population appears to be related primarily to habitat modification and secondarily to human-caused deaths associated with land development.
OCR for page 31
The Scientific Bases for Preservation of the Mariana Crow The brown tree snake might be present on Rota, but as yet it has not been reported there. POPULATION PROJECTION FOR THE AGA ON GUAM The committee was asked to conduct a population viability analysis (PVA) for the aga. Such analyses usually use stochastic models to make long-term projections of populations and to make probabilistic estimates of the rate of extinction under different management options (Burgman and others 1993). A PVA can predict the size of a population for 50–200 years by allowing the rates of fecundity and survivorship to vary stochastically for each year, based on estimates of the variance in these parameters. This requires good estimates of the annual variation in survival and fecundity that results from demographic stochasticity, environmental stochasticity and catastrophic events (Boyce 1992). At least 10 years of demographic data are required to calculate valid variances in fecundity and survivorship for a long-lived bird like the aga. Such data are not available for the aga, as the review of its demography in the previous section indicates. Even when parameter values are available, the structure of PVA models may be inappropriate and hence the results may be unreliable (Caughley 1994; Caughley and Gunn 1995). For example, PVA models hardly even consider the interaction between the target population of interest and the population dynamics of its predator. For the aga, this would be a serious shortcoming, since changes in the dynamics of the brown tree snake could have a large effect on the population dynamics of the aga. Thus, the committee did not construct a PVA model for the aga because it was inappropriate. The committee considered using a simplified Leslie-Lefkovitch matrix to predict lambda, the short-term rate of annual change in the population. The Leslie-Lefkovitch matrix has been a standard tool to evaluate population dynamics for more than 50 years (Leslie 1945; Lefkovitch 1965). However, the same concerns that led to the realization that a PVA analysis is inappropriate apply equally well to lambda. Furthermore, it is doubtful that such an analysis would lead to a different conclusion than that which is strongly indicated by the evidence, namely, that the aga population on Guam is in extreme danger of extinction and that the major cause is adult mortality. While more sophisticated analyses might be useful in the future, once recovery is underway, they would require much better, more reliable data. PAST AND CURRENT CONSERVATION AND MANAGEMENT The major government organizations that have been involved with conservation and management of the aga on Guam and Rota have been DAWR, DFW of the CNMI Department of Lands and Natural Resources, FWS, and the National
OCR for page 32
The Scientific Bases for Preservation of the Mariana Crow Biological Service (NBS) of the US Department of Interior (DOI). The US Air Force and Navy, the US Department of Agriculture, and the DOI Office of Insular Affairs (formerly the Office of Territorial and International Affairs) have played smaller roles. Since 1993, the American Zoo and Aquarium Association and 3 member zoos have participated in the Marianas Archipelago Rescue and Survey (MARS) program's efforts to add the aga to their captive-propagation activities. With the implementation of an intensive management plan for the aga by DAWR in 1994, there has been increasing interaction with the 'alala recovery program in Hawaii. All those participants share one or more of the following objectives: to diagnose the causes of the population declines in the aga on Guam and Rota, to ensure that it survives in the wild, and to eradicate or at least control the brown tree snake on Guam and ensure that it does not spread to other islands. In 1981, the aga received legal protection by the government of Guam for the first time; and in 1984, it was federally designated as endangered (FWS 1984). Of the 6 other endangered species listed at that time, 2—the Mariana moorhen and the island swiftlet—still survive. The Guam broadbill and the Guam bridled white-eye were extirpated from Guam in the 1980s, and the Guam rail and the Guam Micronesian kingfisher exist now only in captivity. A recovery plan for endangered birds of Guam and Rota prepared by Robert Beck and Julie Savidge (FWS 1990) recommended a captive-breeding program for the crow and study of the feasibility of translocating birds from Rota to Guam. In 1991, the crow was among the Guam birds proposed for the designation of critical habitat by FWS (1991). The proposal was withdrawn (FWS 1994) because, with the inclusion of these lands in the newly established Guam National Wildlife Refuge, most of the lands proposed for critical habitat became protected. The refuge consists of about 150 ha of newly protected areas and more than 9,000 ha of current military lands at the northern end of Guam. The Pacific Island Recovery Team (PIRT) in FWS is charged with developing a recovery plan for the aga and the other endangered birds of the Pacific islands. Guam The decline in the aga population on Guam has been documented with stationary counts, roadside counts, and sampling along transects. Roadside counts conducted by DAWR since the 1960s showed that the aga had disappeared from southern and central Guam by the early 1970s (Aguon 1983; Jenkins 1983). In 1981, Engbring and Ramsey designated randomly selected starting points and parallel transects and used a circular-plot-survey method (Reynolds and others 1980) to estimate the number of crows on Guam. After correction for detection distance and differences among observers, their record of 241 birds led to a total population estimate of 357 (Engbring and Ramsey 1984). By the middle 1980s, it was becoming apparent to researchers that although there were many threats to the persistence of the aga and other native birds on Guam, the overriding threat
OCR for page 33
The Scientific Bases for Preservation of the Mariana Crow was predation on eggs and young birds by extremely high populations of the brown tree snake (Conry 1988; Engbring and Fritts 1988; Savidge 1986, 1987). A repetition in 1990 of Engbring and Ramsey's 1981 transect survey by DAWR recorded only 107 birds. This survey used recorded-call playbacks, which are about twice as effective a detection method as the standard circular-plot detections used in 1981. The last record of unassisted fledging of juvenile crows was in 1985 (Michael 1987). DAWR estimates of the number of crows on Guam fell from 100 in the breeding season of 1990–1991 to 18 territorial birds and a few nonterritorial birds in 1995–1996 (Anderson 1996). Only 1 of the pairs of the crows remaining in 1995–1996 produced eggs. DAWR has developed an electric and mechanical barrier, tested it on trees, and shown that it is potentially effective in excluding snakes and monitor lizards from nest trees of the aga (Aguon and others in press) and improving the ability of the birds to nest (Aguon and others 1992). The development of the tree-barrier technology began in 1985; by 1989, it was combined with trapping snakes in and around the barriered trees. In 1989, a chick fledged from a nest protected by a tree barrier but died soon thereafter. Improvements in tree-barrier technology led to the fledging of three chicks in 1991–1992. The finding that females were laying larger clutches (3 versus 1–2 eggs) in nests in protected trees during the 1992–1993 breeding season strongly supported the suggestion that in previous years many eggs were lost to brown tree snakes during the laying period. In 1993–1994, DAWR researchers discovered that a high percentage of eggs laid were not viable, so an intensive aviculture intervention program was begun to study this phenomenon, to incubate eggs artificially, and to hand-rear chicks. The crows readily renest, so taking eggs ("double clutching") encourages them to produce more. One chick was successfully hatched in 1994–1995 and was returned to its nest 2 days after hatching, but it died before fledging. In 1995–1996, although only 1 pair of crows on Guam laid fertile eggs, 1 male and 1 female aga were successfully reared from their eggs. These young birds remain (August 1996) in captivity on Guam, but their wild male parent has disappeared. Simultaneously with the research efforts for the aga on Guam in the 1980s, an intensive research program on the brown tree snake was begun by Savidge (1986, 1987). From necropsies of snakes, she determined that in the period 1982–1986, a time when birds had virtually disappeared from much of Guam, the percentage of birds and bird eggs in the diets of snakes was lower than it had been in the 1970s, and the percentage of lizards was higher. By using sentinel birds (coturnix quail, domestic chickens, domestic canaries, and wild-caught bridled white-eyes from Saipan) placed in the forest, she determined that predation on eggs and young birds was high. There was no evidence that either disease or chemical contamination was causing declines in the birds (Grue 1985; Savidge 1986). Continuing efforts by NBS (efforts formerly housed in FWS) have greatly extended the research on snakes (Fritts 1988; Rodda and others 1992, in press a). Field studies in the early 1990s of the behavior of the aga on Guam during the
OCR for page 34
The Scientific Bases for Preservation of the Mariana Crow breeding season showed that many nests were abandoned before they were completed and that many eggs failed to hatch. One must take into account that corvids readily renest. But other factors might contribute to the inability of the birds to complete the nesting cycle, such as harassment by unmated aga, disturbance during low overflights by aircraft, disturbance by snakes, and infertility of eggs (DAWR 1993; Grout 1993; Morton 1996a; Wiles and others 1994). Even full protection from the brown tree snake might be insufficient to keep the aga from being extirpated from Guam. In 1994, DAWR implemented an aviculture intervention plan in which eggs laid in the wild would be taken for artificial incubation. Captive hatchlings would be hand-reared and then either placed back in nests that had been held in the incubation phase with dummy eggs or released as juveniles. Protocols for this project were developed in consultation with the personnel of the 'alala recovery program in Hawaii. After a series of problems with infertility and incomplete development, the 2 captive birds mentioned above fledged in captivity in March 1996. In late 1995, DAWR requested a FWS permit to translocate eggs and young aga from Rota to Guam. Because FWS anticipated public and scientific interest in the proposal to introduce birds from Rota to Guam, and because comments it received showed concern about such actions, FWS requested this NRC study. Fulfillment of the permit request awaits the findings of this committee. The need for herpetologic expertise and snake-control activities applicable to the protection of the aga within the DAWR program is clear. Even so, there are vacant positions in the organization and sizable amounts of unexpended funds earmarked for such work. The staff herpetology position, which was vacated in 1992, remains unfilled. In summary, conservation and management efforts for the aga on Guam have emphasized surveys, behavioral observations of unbanded birds, and placement of tree barriers to protect nests from predation by brown tree snakes and monitor lizards. Most recently, they have included intensive management efforts involving the removal of eggs from wild nests for artificial incubation and hand-rearing of the young. In spite of those efforts and the protection of crucial forested habitat as part of the Guam National Wildlife Refuge, the status of the current wild population of the aga on Guam must be regarded as desperate. Rota Field work by DFW on Rota has been limited by a lack of permanent staff. Field observations on Rota indicate that the aga there forages and nests within the forest canopy and is less conspicuous than on Guam. There is some information on the omnivorous diet (which includes fruits and insects) and behavior of the aga on Rota (Tomback 1986). Evidence of an inability to complete construction of nests or the nesting cycle has been described in annual CNMI reports in the 1990s and is reminiscent of similar observations of Guam. However, one must take into account the tendency for corvids to build multiple platforms before completely
OCR for page 35
The Scientific Bases for Preservation of the Mariana Crow constructing a nest. Several chicks on Rota have been banded, but there have been no consistent followup efforts (Grout 1996; Morton 1996b). There have been differences of opinion about the size and status of the Rota population of crows. Pratt and others (1979) considered the species to be common in spite of persecution by hunters, whereas Ralph and Sakai (1979) considered it to be uncommon. In 1982, Engbring and others (1986), using randomly selected parallel transects and a circular-plot-survey method, recorded 454 crows on Rota. After correcting for an estimated detection distance of 123 m and for observer effects, they estimated that there were about 1,300 crows on Rota. Because this survey method involves estimating detection distances and extrapolating numbers for unsurveyed areas, it undoubtedly is less reliable for crows than other potential methods. Similar procedures in 1994, as part of a contracted 5-year population survey, produced a population estimate of 592 birds (95% confidence interval, 474–720). The reliability of those estimates is difficult to judge; the aga on Rota is probably declining together with the bridled white-eye and rufous fantail. The magnitude and causes of these declines are not understood, although obvious evidence of predators is lacking. The most recent population estimate of 592 crows on Rota is based on a variable circular-plot survey in October and November 1995. The survey used the same methods and many of the same transects as were used by Engbring and others (1986) and was conducted in the same fashion. Grout and others (1996) reanalyzed the 1982 survey data of Engbring and others and revised the total estimate from 1,318 to 1,348. Then, comparing that estimate with the estimate for 1995, Grout and others (1996) estimated that the total population had declined by 50% between 1982 and 1995. If one assumes that the 1982 transects that were not repeated had no crows, a direct comparison of the numbers of crows observed on the same 11 transects that were repeated plus the 2 that were not shows a much greater decline—about 71%—in 13 years (from 224 to 64). DFW is convinced that there have been substantial declines in the Rota population of crows in the last decade (Palacios 1996). Seventeen known territories of crows have been mapped in 1994—1996 (Grout and others 1996). In 1994, FWS opened an office on Rota. A contractor (Resources Northwest, Inc., Bellevue, WA) has been hired on Rota to develop a habitat-conservation plan for the island (see chapter 5, "Habitat Protection"). The 2 major concerns on Rota are the danger that the brown tree snake will or may already have reached the island in cargo arriving by ship or air from Guam and that insufficient forest habitat will be protected from development to sustain populations of forest birds. A captive population of aga was founded with wild birds captured on Rota in 1993–1995 as part of the MARS program. The MARS program includes 7 zoologic institutions (Houston Zoological Gardens, National Zoological Park, Louisville Zoological Garden, Memphis Zoological Garden and Aquarium, Honolulu Zoo, North Carolina Zoological Park, and Philadelphia Zoological Garden). The
OCR for page 36
The Scientific Bases for Preservation of the Mariana Crow program was initiated largely to develop techniques for the capture, acclimation, transport, and propagation of aga, Rota bridled white-eyes, and Mariana fruit doves; to provide training to FWS, DAWR, and DFW biologists in field capture and acclimation techniques; and to conduct basic research on the past and present distribution of Mariana birds. To date, 10 wild aga have been collected from Rota: 7 in July 1993 as part of the first MARS collection, 1 in July 1994 by CNMI biologist David Worthington, and 2 in January 1995 as part of the second MARS collection. Of the 7 birds collected in 1993, 2 (1 male and 1 female) were sent to the Philadelphia Zoological Garden for quarantine and then were transferred to the Houston Zoological Gardens (HZG) for breeding. The remaining 5 (3 males and 2 females) were sent to the National Zoological Park's Conservation and Research Center (NZP-CRC in Front Royal, VA. The single female salvaged in July 1993 was in ill health at the time of capture and was given antibiotic therapy on Rota before being transferred to NZP-CRC later in the same month. She died on January 9, 1996, despite continuous medical treatment throughout her captivity. The necropsy revealed that death resulted from renal failure. Although the cause of the condition was not determined, the persistent high blood concentrations of uric acid obtained from the time of capture suggest that poor renal function was the underlying cause of her original illness in the wild. The last 2 birds (1 male and 1 female) in the captive population were collected on Rota in January 1995 and transferred to NZP-CRC. Two offspring were hatched by the pair of crows at the HZG on April 21 and 22, 1995. The second nestling disappeared from the nest at the age of 5 days and presumably was eaten by the adults, inasmuch as no remains could be found in the nest or elsewhere in the breeding enclosure. The first nestling (a female) is still housed at the HZG. Although 3 adult pairs held at NZP-CRC have laid eggs, no offspring have as yet been produced. All clutches have been either infertile or destroyed by the adults during incubation. Habitat Protection As mentioned above, about 9,000 ha of forested habitat on Guam was recently incorporated into the Guam National Wildlife Refuge for the preservation, protection, and management of endemic endangered birds (FWS 1994). In addition to these refuge lands and private holdings, the government of Guam owns and manages some 1,620 ha of suitable forest habitat (FWS 1994). This amount of land—if kept in its current condition—should be adequate for maintaining a viable population of aga on Guam. Ideally, however, additional forest lands at various locations throughout the island should be acquired to minimize the potential risks associated with catastrophic events, such as typhoons. On Rota, Engbring and others (1986) reported that native forests made up about 60% (5,100 ha) of the island's area in 1982. The total has probably decreased
OCR for page 37
The Scientific Bases for Preservation of the Mariana Crow by about 5–10% in the intervening years (Grout and Lusk 1996). With increasing development pressures on Rota, the loss and fragmentation of native forest habitat is almost certain to accelerate in the immediate future. Because most of the land on Rota is privately owned, a habitat-conservation plan for the island is being developed to preserve essential forested habitat for the aga and other native wildlife. FWS has assigned a full-time biologist on Rota to assist in information-gathering and planning and is providing geographic information system (GIS) support to CNMI planners (Grout and Lusk 1996). Habitat conservation plans offer a mechanism for creative partnerships between the private sector and local, state, and federal agencies in protecting and preserving endangered species and their ecosystems under the Endangered Species Act. However, this approach has been used infrequently. It has sometimes been impossible to achieve consensus on a plan, but where such consensus has been achieved and plans have been implemented, very promising results have been seen (NRC 1995). Because development of the Rota habitat-conservation plan has only recently begun to be implemented, it is difficult to evaluate its eventual role in the conservation of the Rota population of the aga. Ultimately, however, the success or failure of the plan will depend on whether enough native forest is set aside for the maintenance of a viable crow population. Several wildlife-conservation areas have already been established on Rota (FWS 1996), but they are not inhabited by crows. The biologists on both Guam and Rota and the various workers from FWS and academe who have worked there are dedicated to protection of the endangered native biota, but the higher levels of government and the general public consider economic development to have a higher priority. This poses a potential conflict with crow-habitat preservation (Kuhlman 1996.) An approved islandwide habitat-conservation plan for Rota might protect 3 areas as forest but allow development of resorts and homesteads in some areas that are now occupied by crows. The objective is to allow economic development in concert with protection of native plants and animals. In spite of previous work on both Guam and Rota (Jenkins 1983; Marshall 1949; Tomback 1986; annual reports of DAWR and DFW), the life history, social structure, and ecology of the aga are not fully understood, so a full diagnosis of all the factors that might be contributing to its decline is not yet possible. There is an urgent need for new intensive programs of research and management for both the aga and the brown tree snake.
OCR for page 38
The Scientific Bases for Preservation of the Mariana Crow
Representative terms from entire chapter: