CHRISTMAS  ISLAND

EXPEDITION

The Christmas Island Goshawk Accipiter natalis

Introduction 

The Christmas Island Goshawk Accipiter natalis is currently listed as Endangered under the Environment Protection and Biodiversity Conservation Act 1999 (EPBC Act).  It is currently thought to be a subspecies of the Brown Goshawk Accipiter fasciatus, but its taxonomy needs resolution.  The Christmas Island Goshawk is considered to be the rarest endemic bird on the Island and prior to the expedition to Christmas Island there had never been any detailed studies on this species.  In 2000 the Christmas Island Goshawk met the criteria as being Critically Endangered. 

Mature Male Christmas Island Goshawk

Photo by Michelle Manhal  

The Australasian Raptor Association (ARA) was approached by Parks Australia North-Christmas Island to study the Critically Endangered Christmas Island Goshawk, of which six members of the ARA, which included a Zookeeper from Healesville Sanctuary, conducted an expedition to trap and colour band a proportion of the Christmas Island Goshawk population during the 3^rd August through to the 2^nd September 2004. 

Method 

To determine the size of the Island’s Goshawk population and the ecological requirements for this species, a mark-recapture (banding) was seen as the first stage of this expedition.  During the month long expedition 52 Christmas Island Goshawks where trapped and colour banded.

Two active nests were discovered during this study.  Both were in Macaranga trees near the edge of made (single lane roads) in primary rainforest.  The last time there was a discovery of a nest site or its description for this species was in 1940. 

The information we received from Parks Australia North – Christmas Island when arriving on the Island, was that the CI (Christmas Island) Goshawk was easy to find.

Juvenile male CI Goshawk would regularly come to the Research station for a visit.  Photo by Michelle Manhal. 

It was apparently a simple matter of driving around, stopping from time to time and opening your car door and waiting.  The CI Goshawk would apparently just come and land on the door of your car, as they are a curious species, not like the Brown Goshawk accipiter fasciatus.   Despite these assurances we found this species a little less curious than first described. 

The expedition team had 2 vehicles at their disposal to which the survey effort was divided between 2 survey teams for 10 hours per day.  This allowed 30 survey days per team and 2 teams operating simultaneously for the duration of the expedition gives c. 600 search hours. 

Trapping Method

Two trapping methods where used during our time on Christmas Island.  The first was a stationary trap, six 600 x 600 x 900mm traps where assembled, a modified version of, the ‘falling lid’ traps of Kenward and Marcström 1983 (and here referred to as ‘Strine Goshawk trap’) were stationed at strategic locations. These traps were initially checked up to six times per day and then a minimum of three times per day after the first week.  Reassurances had been made that the Strine traps worked exceptionally well in Tasmania, but only 6 goshawks (3 being re-traps) where ever trapped using the Strine Goshawk Traps.  The traps were placed on steel frames with four steel legs 500mm tall to prevent Robber Crab Birgus latro and Red Crab Gecarcoidea natalis from triggering traps when attempting to make entry. 

Fig 1: Robber Crabs still able to access trap.         Fig 2: Free feeds given to stop crab access to trap. 

  Red Crab Gecarcoidea natalis                  Robber Crab Birgus latro

                                                  Photos by Michelle Manhal

The traps were placed in 12 locations on the edge of or close to tracks in primary rainforest or nearby clearings. 

It was suggested (Michelle Manhal) after about 2-3 days that a Bal Chatri be made and trialled, to which a 2^nd Bal Chatri (Berger and Mueller 1959 and Thorstrom 1996) was made (by Michelle Manhal) on Christmas Island and these were transported (one in each vehicle) at all times.  Once an un-banded Goshawk had been sighted, a trap was placed on the ground or at times on the back of the stationary vehicle in clear view of the Goshawk.  62% of the Goshawks trapped using a Bal Chatri were first detected during active surveys in vehicles or by walking either along tracks or through rainforest areas using GPS.

The active surveys involved teams of two people (although Michelle Manhal tended to work or survey majority of the time alone) in a vehicle driving at 10-20km/hr along made tracks or roads.  Effective survey distances from the vehicles varied considerably depending upon the terrain, density and proximity of vegetation from the vehicle.  Usually the Goshawk would be flushed up from the ground in response to the approaching vehicle.  Once a Goshawk was located, attempts were immediately made to identify if it was wearing bands (this was done through the use of binoculars).  If no bands were seen a Bal Chatri was then placed as close as practical and in clear view of the Goshawk without flushing it further.  The Bal Chatri would be placed in such a way as to allow the Goshawk an unobstructed view of the trap and with a clear flight path to it.  

On several occasions the trap would be moved if the Goshawk had made no attempt to land on the trap within a short space of time, this was dependant on the Goshawks behaviour.  

2^nd year female CI Goshawk on a Bal Chatris, prior to capture. 

Photo by Michelle Manhal

This “reading” and or “understanding” of the bird’s behaviour and responding by moving the trap to a position where it may be more attractive to incite an attack by the Goshawk was vital to much of the success in using these traps.  Even an adult female that was first observed carrying and then eating a CI White-eye Zosterops natalis was successfully trapped within an hour by patiently and repeatedly re-positioning the trap in response to the Goshawk’s behaviour.  During the trapping process it was vital that all other persons move from the area as more than one person in such close proximity to the Goshawk did unsettle the bird.  It was, at times, difficult to assess whether the Goshawk had been snared (see photo above), so caution was paramount when approaching the trap.  This trapping method was conveyed (taught) to all expedition volunteers by Michelle Manhal.

Morphometrics 

Once captured, Goshawks were held in dark blue cotton weighing bags with a draw string around the top for ease of opening and closing, a zipper sewn down one side of the bag gave easy access to the bird during measuring.  Up to 28 linear measurements were taken for each individual and recorded on a data sheet.  Each banded individual is assigned a unique 3 character identity code (such as F03).  The first character indicates the sex of the Goshawk (in this case F for female) and the two numeric characters that follow “03” indicate that this is the third female banded.   Additional measurements were taken to aid in determining the most appropriate band size for each sex of this population of Goshawks.  The Australian Bird and Bat Banding Scheme currently has no recommended band size for this subspecies and part of the conditions of conducting this study was to take measurements and recommend appropriate band sizes.  One metal band was placed on each leg of each Goshawk trapped.  A stainless steel band (size 9) supplied by the Australian Bird and Bat Banding Scheme (ABBBS), Commonwealth of Australia was placed on one leg.  Each of these was powder coated one of four colours.  For the males a second un-numbered, powder coated stainless steel band would be placed on the opposite leg.  All females had a similar ABBBS but appropriately larger (size 11) powder coated band placed on the left leg.  An aluminium band anodised with two colours (one above the other) is placed on the right leg of each female.  Each band has a readable double character code repeated one character above the other, four times around each band.  The characters are alpha or numeric.  These bands are attached with two pop rivets. 

Data collection.                    Photos by Michelle Manhal             Band sizing. 

In addition to taking morphometric measurements, DNA material was collected (16 blood samples and 25 feather samples) which is due to be deposited with the Museum of Victoria for genetic analysis.  A broader systematic genetic analysis of all Accipiters in the Australian Papuan and Austro-Oriental regions will prove very illuminating in placing the true specific status of this taxon. 

Moult Data 

Data on moult and plumage condition was collected in two ways.  A moult score was recorded for each Goshawk and digital images were taken of most birds following banding and prior to release.  A three point scoring system was used for each of the following wing feathers: primaries 1-10, secondaries, 1-10, tertiaries 1-4 and tail feathers 1-6.  Each feather was scored on one side of the body only and that was the right side; M for feather missing; N for a new feather or growing feather; and O for an old or worn feather.  Head and body feather tracts were scored simply by presence or absence of growing pin feathers.  A score of ‘Y’ for ‘yes in moult’ or ‘N’ for ‘not in moult’ was recorded for crown moult and separately for body moult.  

Normally one photo of the under wing opened and body, one of back with wing opened and a head photo either front-on or from the side.  Photographs were taken to illustrate, patterns of moult. Age specific plumage and soft part colouration and iris colour.   

Juvenile CI Goshawk         2^nd year CI Goshawk                Adult CI Goshawk

2^nd year CI Goshawk                                     Juvenile CI Goshawk 

Photos by Michelle Manhal

The active survey and trapping effort was evenly divided between morning and afternoon on most days.  With approximately 5 hours being spent in each half of each day by all survey teams we note that 66% of males were trapped in the afternoon contrasting with 68% of females being trapped in the morning.

Habitat 

Habitat has decreased by 25% and Goshawks are competing with another diurnal raptor, the Australian Kestrel Falco cenchroides which benefits from past rainforest clearing.  Goshawks and Kestrels will be competing for those small vertebrate prey species remaining.  Especially those restricted small mammal and reptile species venturing out into more open areas.  The impacts of the Yellow Crazy Ant are also locally reducing prey abundance and diversity particularly at the sight of the multi queened super colonies.  Even several years after eradication of the Yellow Crazy Ants from these areas, large areas of rainforest continue to be greatly affected with vegetation diversity and structural changes and a persistent reduction in the vertebrate and invertebrate biodiversity in these areas.  Surveys in some of theses areas failed to detect a single Goshawk. 

Before Yellow Crazy Ants                     and              After Yellow Crazy Ants

Photos by Michelle Manhal 

 Habitats of banding sites.                                                    Christmas Island  

                                                                               Photo by Michelle Manhal

Resightings

The resighting data in the below table includes all 55 Goshawks banded and covers the period from August until the end of November 2004.

 Diet

Eight regurgitated pellets, containing the hard remains of dietary items, were obtained from Goshawks while they were being banded.  These pellets were analysed for prey remains by the BMP team, using the Christmas Island Arthropod Collection and specimens of vertebrates housed at PANCI’s lab. 

 Recommendations

A similarly designed and conducted study a year later will consolidate our understanding of the ecological requirements of this endangered species.  Especially in these key areas:  Cohort survivorship; population age structure; population size; pair, site and territory fidelity; and gain a more comprehensive data set on the diet especially as it relates to age and sex of Goshawks.  Most importantly of all an estimate of the population size should be possible following a successful second survey.  An education campaign to encourage both reports of sightings of Goshawks and to foster an appreciation of this endemic species may reduce some of these threats.

All photographs and information presented on this site are covered by © copyright 2009.  Written authorisation of all materials must be sort prior to use.