1 INTRODUCTION

On 26th June 2003, 38 Pateke were released onto Tangiaro Stream at Port Charles, in accordance with the Pateke/Brown Teal Recovery Plan: "To initiate, by a variety of means, the establishment of new Pateke populations at a minimum of five locations on the New Zealand mainland".

2 PREDATOR/PEST CONTROL

The area trapped was in the Port Charles catchment, and along the main road leading into Port Charles. The majority of traps were set on private land, as the area surrounding the release site is privately owned. The traps were maintained by independent contractors (Lettecia Williams and Tony Horton), and were serviced according to type, with live capture traps daily, and kill fortnightly or monthly. Live capture traps were used predominantly, as opposed to kill traps, due to the number of domestic cats in the area. The death of a domestic cat so early in this project could seriously damage community relations. To help with identification of domestic cats, as well as give native birds a better chance, collars with bells were given to all known local cat owners. It was then up to the local if they wanted to put the collar on the cat.

In mid December, a meeting to revise the predator control effort was held with members of the Recovery Group, Hauraki Area staff, and locals involved in predator control. It was recommended that a three tiered trapping systems be implemented:

• Zone 1: This concentrated on intensive trapping around known nest, location and flock sites
  
• Zone 2: This concentrated on habitat trapping, looking at access ways into sites as well as bush margins and roads.
  
• Zone 3: This concentrated on landscape trapping, placing traps on ridges and gullies around the Port Charles catchment. This landscape trapping would also tie in with the predator control of the Moehau Kiwi Sanctuary.

Zones 1 & 2 have been implemented, and Zone 3 is currently being set up, after funding was provided by the Recovery Group. Traps were also laid response to sightings of cats, and Pateke deaths, and had considerable success. Table 1 summarises the types of traps used, and Table 2 which species have been caught. Figure 1 shows the capture sites of stoats, cats and weasels.

2.1 Cats

Cats were controlled using Holden cage traps, 3 x Victor leg-hold traps in submarine boxes, and Steve Allen (SA) Coni-Bear traps. This provided 10144 corrected traps nights for a catch of 29 cats ( 0.286 cats per 100 CTN). Initially bait was tinned fish-based cat-food, but was changed to salted rabbit pieces as the tinned cat-food went off quickly. Additional traps were put out in response to sightings of cats or known deaths of Pateke. Cage traps and leg-holds were checked daily.

2.2 Mustelids

Mustelids were controlled using Mk 6 Fenn traps in a single entrance wooden box. Traps were initially baited using hens eggs, but were changed to salted rabbit meat when this became readily available. Each trap was checked fortnightly during peak stoat breeding season and monthly during non-peak seasons, using the same regime as Moehau Kiwi Sanctuary.

2.3 Rats

Victor Professional snap traps in covers were set out around feeders, and various cat traps, to control rats. Although not a predator of Pateke, rats were recorded as being a nuisance by stealing bait from traps and attacking mallard ducklings at feeders. Traps were baited using peanut butter, and were checked weekly, or daily if near a live capture trap.

2.4 Mallard Ducks

1 attempt at mallard control was undertaken to breakup a female mallard/male Pateke pair using subsonic .22 ammunition. Although the aim was to kill the female mallard, the ammunition used was not powerful enough to kill the bird. However, the pair separated after the attempt. This was done to reduce the possibility of Mallard/Pateke hybrids being produced.

2.5 Mynas and Sparrows

One attempt at myna control was carried out using alphachlorolase paste spread on stale bread. Although remains of mynas and sparrows were found it is unsure how many birds were actually killed during this operation. This was done on private land away from the release site to reduce the chances of accidentally poisoning any native birds.

2.6 Non-target species

Hedgehogs, possums, rabbits, mice, blackbirds, kingfisher and myna were all caught as non-target species.

TABLE 1: Trap types

TABLE 2: Species caught

• Includes 3 cats that were captured and returned to owners

Figure 1: Capture sites of stoats, cats and weasels

3 POST RELEASE MANAGEMENT

3.1 Feeding

10 feeders containing Pateke feed were put out post release. The feeders were spaced approximately 20 metres apart along the true left of Tangiaro Stream (release site), and were either attached to battens driven into the ground, or nailed to trees. Feeders were serviced with a mixture of kibbled maize and Pateke pellets on the following regime:

1. Every working day for the first month: 4 feeders were then removed, as they were not being used, and two were shifted to the sandbank at the river mouth where the highest amount of usage occurred.
2. Three times a week for a further week
3. Two times a week for a further week
4. Once a week for a further week.

During this time the amount of feed being put into the feeder was reduced. Observations of feeder usage by all birds, Pateke numbers and identification occurred.

After this, feeding was discontinued. No Mallard, Myna or sparrow control occurred during this time, although the Ranger and locals both noted that the population of these birds increased while the feeding regime occurred and decreased once it had finished. An electric fence was put out to keep cattle away from feeders and feeding sites.

3.2 Telemetry Monitoring

Monitoring was carried out every working day for the first month post release. This then dropped to three checks a week where possible. During this monitoring period each bird was monitored for mortality signal and distribution. Where possible, each bird was tracked down and its location logged by GPS and assigned a number. Sightings and numbers of both released and wild birds were recorded (see section 5.0 for results). It was discovered that the TR4 receivers being used by the project had been configured to read one channel below the actual transmitter channel. This was not a big issue, but sometimes made remembering data more difficult, and has been rectified by having the receivers re-configured. Throughout the breeding season females were monitored for nesting. Because most of the land the Pateke use is privately owned, by a series of landowners, access issues made monitoring, capturing and retrieving dead Pateke more difficult.

Table 3: Progress against targets

3.3 Predator/Pest Control

See section 3.0

4 MONITORING PATEKE AT PORT CHARLES

4.1 Flock counts

Objective: To understand the relationship between flock counts and total Pateke population (adults and juveniles) in the source area and, therefore, be able to use flock counts as a means of determining trends in Pateke abundance under specific research-by-management scenarios.

Monitoring Target: Locate any historical and new flock sites in the Coromandel region.

Counts were undertaken at historical and 'new' flock sites twice during the annual flock count period (late Feb to early March). Although a further 2 counts were planned these were not achieved due to inclement weather. Historical flock sites further afield (Kennedy Bay, Opito Bay, Otama Bay and Colville Bay) were visited outside this time and searched but no Pateke were seen. Table 3 shows numbers of Pateke seen during the flock sites visits.

Stony Stream, Denize and Karuna ponds, and Big Bay were also visited on a number of different occasions in response to reported sightings. One wild Pateke was recorded on the Stony River, up to 6 at Sandy Bay, up to 4 birds seen on both the Denize and Karuna Ponds, and up to 13 birds at the Big Bay flock site. Four wild Pateke were also seen on the Waikawau River, although it is unsure if these are the same birds that have been seen at both Karuna and Denize ponds. Table 4 and Figure 2 summarise the Flock Sites and Counts

TABLE 4: Flock site info

Note: First number indicates total birds seen; second number indicates numbers of birds with transmitters seen.

FIGURE 2: Flock sites

4.2 Survival of Released Pateke

Objective: To test the hypothesis that maximum practicable management results in a high (management objective of 50%+) adult survival at Port Charles in the first 12 months post release.

Performance Measure: Annual survivorship determined for 18 males and 20 females over 10 months from June 2003 to April 2004.

Monitoring Target: N = 38 adults monitored.

From the 38 Pateke monitored, there have been 16 deaths to date (see section 5.5 for additional information). One female has a failed transmitter, but has been seen and identified occasionally, although it cannot be confirmed that she is still alive at the date of this report. There are 4 long term missing birds that have been missing for more than 2 months (2 male, 2 female), and 6 birds that have had transmitters removed (it was hoped to have removed the transmitters from all the 03-04 release birds, but this was not possible due to catching difficulties at the flock site), 5 female and 1 male. It must be noted however that 2 (1 male, 1 female) of these birds have been included in the deaths as they were killed after their transmitters were removed.

An additional 3 wild birds were caught and had metal bands attached, but not colour bands or transmitters, so these birds were not monitored. There were no harness failures. All other birds are still being monitored. The released Pateke survival calculation is based on all 38 birds in the sample at the start of the period (26/06/03), for which the outcome is known for at the end of the period (13/05/04).

• Released Pateke survival = 44.8%
  
• Male survival = 30.8%
  
• Female survival = 56.5%
  

Aerial tracking for long term missing birds occurred on three occasions. This was done when three or more birds were missing, or when Moehau Kiwi Sanctuary was doing aerial tracking of kiwi, to make the operation more cost-effective:

• A flight on 20/10/03 picked up a Pateke in the Oneura Stream behind Little Sandy Bay, when looked for on the ground it was unable to be found. This bird was later found in a drainage ditch at Waikawau Bay.
  
• The heli-flight on 8/02/04 picked up 2 mortality signals. Both signals came from a hill on the western side of Port Charles. These signals were located at 317m and at 115m altitude, but when tracked down the signal coming from high point 317 could not be found and the signal coming from high point 115 was found to be alive and living on the Tangiaro River.

4.3 Hatching Success and Duckling Survival

Objective: To determine hatching success and duckling survival of Pateke.

Performance Measure: Survival to fledging of any young from any released birds

Monitoring Target: Any ducklings produced

Two known nests were monitored during the nesting season, but were not approached to minimize risk of abandonment:

• It is possible that the first nest was preyed upon, but that cannot be confirmed as the females' transmitter failed after leaving the nest. One whole egg (empty), and the remains of 3 others, were found in the nest.
  
• The second known nest contained 4 eggs, checked when the female was away. From this 3 ducklings hatched (75% success). An attempt was made to put transmitters on the 3 ducklings from this nest but the adults were both killed and the ducklings fled before this was achieved

Another nest was found when dogs disturbed a nesting bird. This may have been a released female, but this cannot be confirmed). Six eggs from this nest were placed under a bantam hen, to allow time to decide what to do with them. Due to miscalculation of their hatch window, one hatched during this time (16.7% success), and the duckling was sent to Otorohanga Zoological Society to be raised, and was released with 42 others on 13 May 2004. The unhatched eggs were sent to IVABS at Massey University for autopsy (see COD for results). Figure 3 shows the nest sites at Port Charles:

FIGURE 3: Nest Sites

4.4 Juvenile Survival

Objective: To determine survival of juveniles from fledging to recruitment into the population (survival to one year and survival to breeding)
Performance Measure: Fledge to recruitment of 2 juveniles per brood from released birds

Monitoring Target: 2 juveniles per brood from released birds

Due to the fact that both monitored nests failed, this was not be able to be determined. No wild nests or juveniles were monitored.

4.5 Cause of Death

Objective: To determine cause of death (COD) of deceased radio-tagged juveniles and adults.

Performance Measure: COD determined for at least 50% of deceased radio-tagged birds.

Any bird carrying a transmitter that had switched to mortality was tracked down and, where possible, the carcass recovered. Carcasses were sent to IVABS at Massey University for autopsy, apart from one that had been hit by a car that was kept for aversion training purposes. At each scene, digital photos were taken, and the area was searched for sign of cause of death.

To date there have been 16 deaths (9 female and 7 male). Table 6 summarises the deaths of all Pateke during the monitoring season. Autopsies were carried out on 5 eggs taken from an abandoned wild nest. 80% of the eggs were early embryonic death, 20% were late embryonic death.

Figures 4 and 5 show the time of day, and month the deaths occurred in, and Figure 6 shows the distribution of death sites.

TABLE 6: Cause of Deaths

* Cat/ferret deaths have been included in the cat column as there are no ferrets in Northern Coromandel.

Figure 4: Time of death

Figure 5: Month of death

Figure 6: Site and cause of death

Port Charles

Waikawau Bay

4.6 Dispersal

Objective: To determine seasonal dispersal patterns of juveniles and adults and their subsequent geographic range.

Performance Measure: 70% of dispersed/missing radio-tagged birds relocated and further monitored.

Released birds were monitored every working day for the first month, then 3 times a week where possible for the rest of the financial year. Each bird was tracked down and its location was logged by GPS. Birds that had been missing for more than a fortnight were searched for at likely dispersal or flock sites. High vantage points, aircraft and vehicles were also used to search for missing birds. Aerial tracking occurred when it became cost effective. Of the 13 surviving birds (excluding missing, TX failures and harness removals) all were monitored to the fullest possibility. Figure 7 shows the locations and distribution of all released Pateke.

Figure7: Released Pateke locations.

Port Charles

Waikawau Bay

5 ADVOCACY

Various activities were undertaken to inform and educate the public, land owners and DOC about Pateke progress and issues, including:

• Roadside stops to talk to locals and visitors about Pateke.
  
• Moehau Environment Group (MEG) summer program talk and tour.
  
• Regular progress reports in The Panui, a local newsletter.
  
• Letter drops to locals on Pateke progress and issues.
  
• Pateke aversion training day for dogs.
  
• Several articles in the Hauraki Herald on various Pateke topics.
  
• Release coverage on TV3 national news.
  
• Contributions to Rarebits.
  
• Collars with bells for cat owners.
  
• Talk at Colville School on monitoring and Pateke.
  
• Road signs warning people of Pateke on roads at night.
  
• Letter drop for duck shooters before duck shooting season.

6 DISCUSSION/RECOMMENDATIONS

• The use of feeders post-release was an important factor in allowing the released Pateke the best possible start. It was not necessary to continue the feeding for as long as was done, and it is recommended that the feeding regime be reduce to a month for the second release. With the increased numbers of Mallards, Myna, and Pukeko, it is also recommended that some pest control be undertaken.
  
• The monitoring of birds was made more difficult with the difference in TX and TR4 channels, causing some confusion when people other that core staff were doing monitoring. This issue has been solved by having the TR4 receivers re-configured
  
• Land access issues were a barrier to the effective monitoring of the Pateke, although able to be worked around through using the trapping contractor to monitor any birds on land where DOC staff could not access. Sorting out land access in the future would greatly improve the running of the Pateke project.
  
• While the predator control is not fully implemented yet, it has still been effective. Cats and rats are being caught in high numbers, 21.4% and 54.2% of total captures respectively. Stoat captures are low but this reflects of what is happening in the whole of the Moehau Kiwi Sanctuary.
  
• A much higher number of Pateke were counted during the annual flock counts than expected, with Big Bay and Port Charles having the largest numbers. For Port Charles, this is not surprising given that Tangiaro Stream was the release site. What is surprising is that only 6 released Pateke were seen at the release site, but significantly more are known to use the release site as habitat. The remainder seen were wild birds. It is possible that these 2 sites have the highest number of Pateke due to the fact that they are both tidal, therefore having a constant supply of water, as well as large tidal flats and wet pasture areas for feeding (NB: to date there have been no observations of Pateke feeding on the mudflats). Although the other sites visited during the annual flock count had few or no Pateke at them, it is known that wild Pateke do frequent these sites.
  
• Equipment failure has not been a major problem. The primary TR 4 reciever failed once, and was repaired by Sirtrack, and the back-up receiver was used in the interim. Transmitter failure has only occurred once, and the Pateke carrying it has been identified a number of times since the failure, but not on a regular basis. This failure did reduce the chances of gaining accurate data on nesting, and perhaps duckling survival.
  
• The issue of missing birds continues to be a problem, with ground searching and aerial tracking bringing limited success. The usefulness will increase with more experience using the method, but the topography of the area makes aerial telemetry more difficult, and the frequency of its use needs to be weighed up against cost.
  
• The season produced one 'natural' nest, which hatched three ducklings. However, these birds disappeared soon after both parents were killed by a cat/dog. One other nest (assisted by a bantam hen) produced one fledgling, but this is not counted in the nesting success figures. All that can be said from this is that there was no real breeding success in 03-04
  
• Figure 6 shows that most cat kills occurred in 2 areas.
  
• Site 1: on the western side of the Tangiaro Stream. It is thought that the cat kills at this site were made by the same cat, as two Pateke were found in the same place, although not at the same time. There were also 2 more Pateke killed within 30 m of this site. A 6 kg female cat was caught a short time after the fourth bird died, and since then there have been no more Pateke deaths in that area.
  
• Site 2: at Parakete, along Harriet King's Road. This site also has indications that the kills were made by the same animal, as two of the Pateke bodies were found together. Wild cats have been caught in this area, and domestic cats and dogs are known to live near by. The cause of death could not be determined beyond "cat/dog", as the bodies had been dismembered to the point of being barely recognisible.
  
• Although road signs have been put up in response to the first Pateke death by vehicle, two more deaths by vehicle have since occurred. Further advocacy and education may be needed to reduce this as a cause of Pateke deaths.
  
• The dispersal of birds can be seen in Figure 7, with the majority still at or near the release site. Released Pateke have moved further afield to Sandy Bay and Waikawau Bay, where there are small populations of wild Pateke. The released bird at Waikawau Bay has not been seen in the company of wild Pateke.