Publications archive - Publications
Key departmental publications, e.g. annual reports, budget papers and program guidelines are available in our online archive.
Much of the material listed on these archived web pages has been superseded, or served a particular purpose at a particular time. It may contain references to activities or policies that have no current application. Many archived documents may link to web pages that have moved or no longer exist, or may refer to other documents that are no longer available.
A Dunn and FAR Hill
Environment Australia, September 1997
Note:This publication has been superseded by the National Recovery Plan for the Abbott's Booby Papasula abbotti
Abbott's Booby Papasula abbotti is a large seabird with a white body and black wings and back. It is one of seven species belonging to the family Sulidae, gannets and boobies. Abbott's Booby is unique amongst the sulids and was recently placed in its own genus (Olson & Warheit 1988, Christidis & Boles 1994). It has some very ancient characters and may have diverged from other sulids prior to the divergence of gannets (Morus) and boobies (Sula). It is unique in its breeding biology and behaviour (Nelson 1971), and in its osteology (Olson & Warheit 1988).
Abbott's Booby is currently listed as nationally Endangered (Schedule 1, ESPA 1992). Its distribution on a world scale has been reduced markedly so that it now only occurs on a single island in the Indian Ocean, Christmas Island. On Christmas Island, clearing of primary rainforest for phosphate mining has resulted in loss of a large percentage of the boobies' nesting habitat. In the mid 1980s, the Abbott's Booby Monitoring Program estimated that 33% of known Abbott's Booby habitat had been cleared (Yorkston & Green 1997). Wind turbulence affects breeding success of birds nesting in forest adjacent to clearings (Reville et al. 1990). Abbott's Booby is also listed in the Japan/Australia Migratory Bird Agreement (1974) as a migratory species in need of protection.
Abbott's Booby was described from a specimen collected from Assumption Is. (NW of Madagascar) by the American naturalist W.L. Abbott in 1892. There has been some debate as to whether Abbott actually collected the specimen from Assumption Is or nearby Glorioso Is (Gibson-Hill 1950, Nelson 1974, 1978, Stoddart 1981), but breeding populations in the western Indian Ocean had disappeared by 1926 (Vesey-Fitzgerald 1941, Stoddard 1981). Its disappearance from these islands was attributed to forest clearing (Nelson 1978). It was first collected from Christmas Island (Indian Ocean) by C.W. Andrews of the British Museum in 1897. Currently, Abbott's Booby breeds only on Christmas Island (Stokes 1988) and forages in the waters surrounding the island (Becking 1976). Sightings of adults and juvenile Abbott's Booby in the Banda Sea south of the Moluccas has led some authors (Cadee 1987) to speculate their exists another breeding population in that region (van Balen 1995) however currently only the Christmas Island breeding population is known.
Within Christmas Island most nests are found in the tall plateau forest on the central and western areas of the island and in the upper terrace forest of the north coast. They were once thought to be restricted to areas above 150m, mostly on the sides of North-West facing slopes (Nelson 1978, Stokes 1988), but a survey in 1991 located them in many new areas (Yorkston & Green 1992). This revised distribution would be due partly to movement of the birds (Reville et al. 1990a), but the survey also discovered previously-unknown nesting areas (Yorkston & Green 1992, Yorkston & Green 1997).
Abbott's Booby nests in tall rainforest trees in the western, central and northern portions of Christmas Island. Most nest trees are associated with uneven terrain created by gullies, hill-sides or cliffs. Within suitable habitat, the location of Abbott's Booby nest sites is determined by the topography and nature of the canopy, resulting in a patchy distribution. Most nest trees are located in uneven canopy containing emergent trees, and sites are densest along crests of gullies and west-facing slopes (Nelson and Powell 1986). The tree species most often used are Syzygium nervosum and Planchonella nitida; Tristiropsis acutangula is not as common but is frequently used when it becomes emergent (Nelson & Powell 1986, Yorkston & Green 1992).
Abbott's Booby prefer nest sites with a clear area below and immediately downwind to facilitate take-off and landing. SE trade winds prevail between April and November and thus emergent trees which can be approached from the NW are most often used as nest sites.
Due to their habit of nesting high up in tall rainforest trees, it has been relatively difficult to gain information on the breeding biology of Abbott's Boobies, and most information has been gathered from unmarked individuals. Nelson & Powell (1986) analysed the results of extensive field studies on the breeding biology of Abbott's Booby carried out between the mid 1970's and 1982. Reville et al. (1991a) summarised the results of an intensive study of breeding biology and population dynamics between 1983 and 1989. These two reports summarise most of the information available to date.
Most Abbott's Boobies can only breed once every two years because of the long post-fledging period of dependence. Although birds were not marked in these studies, there was strong evidence that established pairs return to the same nest site and mates usually remain faithful for successive breeding attempts. Breeding commences in March when established pairs begin returning to nest sites and start collecting nest material (Nelson & Powell 1986, Reville et al. 1987). The pair moves nest sites only if they have been repeatedly unsuccessful or the nest site has been destroyed. When displaced they probably move only a short distance since the pair use the nest site as a focus for re-mating (Nelson & Powell 1986, Reville et al. 1990b). Therefore, it takes a pair many years to move away from an area of disturbance.
Egg-laying may occur as early as 2-3 weeks after return to the nest site, but pre-laying activities usually last c. 2 months. Laying may occur any time between April and October, but most lay between mid-May and mid-July (Nelson and Powell 1986). The eggs are much larger and heavier than other sulids and the 57 day incubation is by far the longest within the Sulidae (Nelson 1971). Some pairs will lay in consecutive years if their young of the previous year becomes independent in June or July, but only about 5% of pairs ever succeed in raising successful fledglings from eggs laid in consecutive years.
Most young hatch between June and November (Reville et al. 1987). The mean period from hatching to fledging is 151 days (range 140-175, n=11). This is also much longer than the 30 to 60 days in other Sulidae. Free-flying juveniles remain dependent on their parents for a further 230 days (162-260). Therefore, the total time from hatching to independence is, on average, 363 days (314-418, n=22). The complete cycle can take between 486 and 504 days to complete (Nelson and Powell 1986).
Abbott's Boobies are thought to be very long lived. Using estimates of breeding success and pre-breeding mortality, Nelson & Powell (1986) estimated that it would take 24 years for a breeding pair to replace itself. Reville et al. (1990a) used different information and concluded that it takes between 25 and 31 years for parents to produce their replacements. They had some evidence that the age of first breeding was probably at eight years of age and suggested that the average life span could be about 40 years.
Results of the monitoring program suggest that the adult mortality rate between 1983 and 1988 averaged 4.5% (Reville et al. 1990a). However, adult mortality rate was found to be influenced by the proximity of their roost or nest site to clearings. Adults in areas that were unaffected by clearings had a mortality rate of 3.2%, whereas those that nested or roosted downwind of clearings had a mortality rate of 5.4%. Downwind of clearings, wind turbulence in the canopy is increased. Reville et al. (1990a) suggested two reasons for increased adult mortality downwind of clearings:
(i) Difficulty in landing in the increased turbulence. If an adult misses its footing and falls to the forest floor then it will starve unless it can climb through hanging vegetation to the canopy. Adults cannot take off from the ground.
(ii) Stress brought about by more frequent nesting attempts. Nests failed more often downwind of clearings and so pairs in these areas were laying three eggs to every two for pairs upwind of clearings to produce the same output of chicks.
Assuming that pairs require 24 years to replace themselves, Nelson and Powell (1986) calculated that in order for the population to remain stable the annual mortality rate must be 4% or less, or the survival rate between independence and breeding age must be greater than that of other sulids. Evidence from a number of sources suggests that in other sulids, 40% of independent young reach breeding age (Nelson & Powell 1986). However, as mentioned above, Reville et al. (1990a) measured an average annual mortality rate of 4.5% from their six year study and calculated that between 28 and 35% of fledglings reach breeding age. These figures take into account different mortality rates between disturbed (breeding areas affected by wind turbulence) and undisturbed areas and suggest that the number of young birds recruited into the population each year cannot replace the number of adults that die. If the whole population resided within undisturbed areas then the mortality rate of 3.2% would permit sufficient recruitment to maintain a stable population. In 1986/87, 42.6% of the known population of breeding pairs were nesting within 305m of clearings (Reville et al. 1990b). The surveys of 1991 found that 36% of the population were in areas affected by clearings (Yorkston & Green, 1997).
Christmas Island is close to a number of cold water upwellings which probably provide an increased availability of food that is seasonal in nature, and upon which a number of the seabirds may depend for raising their young. This may be one reason why Abbott's Booby are found only on Christmas Island. Abbott's Booby goes on very long fishing trips in a NW direction (Nelson 1971, 1972) which is the direction of one of the major upwellings. It is thought that they may travel up to 400 km to feeding grounds when they are breeding (Becking 1976), but the location of fishing areas has not been confirmed. Only a few Abbott's Boobies have been seen out at sea. When not breeding they may travel large distances; Abbott's Boobies have been recorded near the Chagos Archipelago, some 4000 km west of Christmas Island (Hirons et al. 1976) and in the Banda Sea, c. 2000km north-east of Christmas Island. Reville et al. (1987) reported a significant correlation between breeding success and sea-surface temperature in the known upwelling sites south of Java, with longer periods of cool water upwelling in a particular year associated with higher breeding success, however, a causal relationship is speculative without knowledge of where the feeding areas are. Low sea temperatures in the Indonesian area occur when strong El Niņo events occur in the Pacific Ocean (Quinn et al. 1978; Rasmusson & Carpenter 1982).
Determining the degree of decline of Abbott's Booby has been complicated due to the difficulty of surveying their numbers. There were no realistic estimates of numbers of Abbott's Boobies before 1967, but it is known that its global range had been severely reduced. In 1967, the breeding population was estimated at 2300 pairs (Nelson 1971). Powell and Tranter (1981) resurveyed the population and although they did not estimate numbers they found that the distribution of nests had not changed significantly from that of 1967 despite large areas of habitat clearance. The most recent population survey was carried out in 1991 (Yorkston & Green 1992) and the population was estimated at 2500 pairs. Although this is greater than that estimated by Nelson, this survey covered much more of the island and discovered nests in areas not previously known. Within the areas that were previously known and monitored, there has been a documented decline. In 1983, the breeding population estimated to occur within Nelson's 1967 distribution was 1900 pairs (Reville et al. 1990a), a decline of 17.4% over 16 years. Some of the birds that have disappeared may have moved outside Nelson's 1967 distribution, but known movements of nest sites tend to be slow and over very short distances. Furthermore, the breeding success of pairs currently outside Nelson's 1967 distribution tend to be relatively low (see Reville et al. 1990b).
Phosphate was first discovered on Christmas Island in 1887 and subsequently the British government annexed the island for the purposes of phosphate mining and began settlement of the island in 1888. The first shipment of phosphate left the island in 1895. Australia and New Zealand purchased the lease and assets of the island in 1948, and sovereignty was passed to the Australian Government in 1958. Phosphate mining increased in the 1960's and reached peak production in 1973-74.
In 1967, Dr. J.B. Nelson undertook a detailed survey of the Abbott's Booby population which demonstrated the degree to which phosphate rich areas overlapped with Abbot's Booby habitat. In the 1970s, jungle clearing to gain access to phosphate deposits destroyed extensive areas of Abbott's Booby nesting habitat. Clearing also killed many adults and juveniles (D. Powell, personal communication). Concern for the Booby was expressed by the House of Representatives Standing Committee on Environment and Conservation in 1974 and this prompted the Australian Government to restrain mining operations. After these reports, phosphate mining was restricted to areas which would minimise impact on the booby, and in December 1987 mining operations were stopped altogether. Mining began again in 1990 but only involved the reworking of previously-mined areas.
Reville et al. (1990b) showed how clearance of nesting habitat and the resultant minefields had a number of adverse effects on survival of adults and on breeding success. These are described below.
Forested areas up to 300m downwind of cleared areas suffer much greater wind turbulence in the canopy than other forested areas (Brett 1989, Reville et al. 1990b). Abbott's Boobies nesting downwind of clearings had lower breeding success than those residing upwind of clearings. Furthermore, experienced breeders moved nest sites more often and there was a higher proportion of consistently unsuccessful sites downwind of clearings compared with pairs nesting upwind of clearings (Reville et al. 1990b). Since forest clearing, there has been a slow movement of nest sites away from cleared areas to areas that were previously little used. These newer areas are more remote from clearings, but breeding success in these newer areas was lower than that upwind of clearings (Reville et al. 1990b). The edge of the forest downwind of clearings also shows signs of dieback due to increased wind exposure. Targetted re-afforestation of cleared areas is recommended to reduce wind turbulence downwind.
Abbott's Boobies probably rely on a seasonal increase in fish numbers associated with cold water upwellings to raise their young. Reville et al. (1990a) showed that sea-surface temperature data obtained from satellites was strongly correlated with average annual breeding success of Abbott's Boobies. Fluctuations in the duration of the cold water upwellings from year-to-year have the potential to halve the overall breeding success of Abbott's Boobies in a bad year (Reville et al. 1990a). If sea-surface temperatures increase in the future as a result of global warming, this could result in lower breeding success for Abbott's Boobies regardless of nest location. On the other hand, more frequent El Niņo events may lead to more years of higher breeding success for Abbott's Boobies (Reville et al. 1987).
A severe storm in March 1988 damaged significant areas of the island's rainforest. It felled approximately one third of the nest sites monitored by the Abbott's Booby Monitoring Program and also killed one third of the monitored fledglings (Reville et al. 1990a). It is uncertain what effect the storm had on adult mortality; in 1988 and 1989, numbers of adults attempting to nest were significantly lower (Reville et al. 1990a, Yorkston 1992), but there was no evidence of a decline in the adult population in the total population survey in 1991 (Yorkston 1992, Yorkston & Green 1997). It is evident however, that severe storms can have a marked effect on the reproductive rate of Abbott's Boobies in ensuing years.
Currently there is very little information on the areas in which Abbott's Booby feeds. It is unknown whether feeding areas are safe or whether over-fishing is a potential problem. Abbott's Boobies have been sighted off the coast of Java (Becking 1976) and if they regularly feed close to the Java coast then there could be the potential for competition with Indonesian fishermen.
Currently, most remaining Abbott's Booby habitat is protected within the Christmas Island National Park. The National Park was declared in 1980 to provide protection for some Abbott's Booby nesting areas. In 1986, two further extensions were added to the park, and in 1989, further extensions increased the park size to cover approximately 63% of the island. These extensions were designed to encompass most of the remaining Abbott's Booby habitat. There are only a few small areas used by Abbott's Booby for nesting that occur outside the park.
The Abbott's Booby Monitoring Program (ABMP) commenced in 1983 and continued until 1989.
Yorkston (1992) listed five objectives which were proposed by the ABMP monitoring team and an expert panel after completion of the ABMP. These were to:
(i) Detect long-term changes in the breeding population, especially further evidence of a decline.
(ii) Determine effects of the 1988 storm on adult mortality.
(iii) Detect long term changes in nesting distribution, including island-wide patterns and changes.
(iv) Measure differences in breeding success between areas, especially near clearings as reafforestation proceeds.
(v) Increase productivity of the population by hand-rearing of grounded chicks.
This program utilised two methods to address these questions:
(i) A core sample of nests was monitored to determine breeding success. There were 108 core sites and a core sample originally included the core site plus the four nearest nests. Nests subsequently found closer to the core site were incorporated into this sample. Nest activity was checked every fortnight. The total sample included 600-700 nest sites.
(ii) The activity of nests within 108 study blocks was monitored to determine nest density and movement of nest sites. A study block included all nests within a 61m radius of a core site. Nest activity was checked every three months.
This monitoring program required three full time staff for the duration of the study. Additional staff were required occasionally to assist.
A reduced monitoring program, the MP, was drawn up in 1989 once the intense monitoring program was completed. The reduced program involved monitoring 63 of the original core sites. Again the program had two components:
(i) The core sample was monitored every three weeks (or seven times) from May to September.
(ii) The 63 study blocks around the core sites were surveyed four times per year; twice between May and September, once in November/December and once in February/March.
This monitoring program required one full time staff plus additional staff to help survey study blocks. This program ran until 1992.
After 1992 the monitoring program underwent a further reduction. Changes to staff workloads and responsibilities meant that much less time could be devoted to monitoring. In 1992 and 1993, only 19 study blocks were surveyed on a single round. Monitoring is not currently continuing due to other demands on PAN staff, although there are plans to develop an on-going monitoring program (P. Meek, PAN, personal communication).
Objectives i, iii, and iv are currently relevant for assessing the status of the population, although the reafforestation is probably not likely to have an effect on the breeding success near clearings for a number of years or even decades.
The current rate of monitoring is not sufficient to detect movements of pairs within study blocks and would not produce breeding success figures that could be compared with previous years. Therefore, it cannot address adequately objectives i and iii. Furthermore, the current monitoring program is not surveying areas near clearings and therefore is not addressing objective iv. Moreover, the differences in breeding success between wind affected and unaffected areas which was detected by the original monitoring program (Reville et al. 1990a & b) were significant, but the magnitude of the difference was such that it would require more accurate estimates of breeding success than can be determined from the current methods of monitoring. The results and conclusions of the ABMP were recently reviewed by an expert panel (Meek, in prep.).
Attempts to rehabilitate the forest after mining were made from 1975 - 1985 but these attempts were only partially successful. In 1987, rehabilitation methods were reviewed and recommendations made for the future (Carew-Reid 1987). The Christmas Island Rainforest Rehabilitation Program (CIRRP) commenced in 1989 following the recommendations of Carew-Reid (1987). The aim of this program was to revegetate clearings in areas adjacent Abbott's Booby nesting habitat in an attempt to reduce the wind turbulence created by the clearings. This was accomplished by building planting beds in rows along the contours of the main slope with the resulting trees intercepting the prevailing winds and acting as a wind break. The planting program used both fast-growing pioneer species and slower-growing forest species. The CIRRP has recently been reviewed by an independent body (CSIRO 1996). They recommended a number of major changes to the CIRRP which are currently being considered. CSIRO (xxxx) criticised the absence of a monitoring to examine any impact of the rehabilitation plantings on the breeding success of Abbott's Booby.
The mining operations have resulted in approximately 70 clearings across the island, covering about 32 square kilometres or 24% of the island (Carew-Reid 1987). Carew-Reid suggests that 26% of the area cleared on Christmas Island for mining needs to be rehabilitated as a priority; this comes to approximately 8.32 sq. km. or 832 ha. Clearings occurring within the distribution of Abbott's Booby have been ranked in order of importance for rehabilitation. Nine areas have been classified as priority 1 and seven as priority 2 (Carew-Reid 1987). Priority 1 areas cover about 365 ha and priority 2 areas cover approximately 240 ha. Each year the rehabilitation program has planted out on average about 10 ha, but the area covered has depended on the funding available for earthworks (R. Hart, PAN, pers. comm.). To date about 30% of priority areas have been planted out. Rehabilitation probably begins to reduce wind velocity in the clearings once the plants form an almost continuous canopy over the clearing, and this occurs after about 18 months to two years after planting (R. Hart pers. comm.), but it would probably may take at least ten years before rehabilitation has an effect on the wind turbulence in the canopy (see Brett 1989). Thus it is not known how many years it will take before rehabilitation has an effect on the breeding success of Abbott's Boobies in the adjacent areas.
The Christmas Island phosphate mining company currently pays a royalty to the Commonwealth Government, a percentage of which is returned to PAN on Christmas Island to fund the nursery operations and rainforest rehabilitation program. In 1995/96 the amount of royalty received by the rehabilitation program was approximately $719,805; in 1996/96 it was approximately $756,562.