Background

Kapiti Island is home to a dwindling colony of Tītī/Sooty shearwater. With almost no chicks recorded fledgling for the past 4 seasons, the current population is getting dangerously close to collapse. On Kapiti Island, the main reason for this steep decrease is depredation by weka on tītī chicks. Both species are critically important for the island ecosystem and the situation is creating a conservation conundrum with one taonga species depredating on another one.
Evolution of known active tītī burrows on Kapiti. Source: archive data, Etienne Ossona de Mendez


Impacts on weka need to be accounted for when designing management alternatives for the Declining Tītī/Sooty shearwater. This expert elicitation aims to estimate impacts, or lack thereof, on the Kapiti weka population associated with 12 proposed Tītī/Sooty shearwater management alternatives.

General background information

  • Kapiti island is 1965 ha.
  • Kapiti island was declared free of mammalian predators in 1999.
  • In the summer of 2025, the total weka population on Kapiti Island is estimated at approximately 2,800 birds (range: 2,600–3,000), accounting for uncertainties in territory mapping.
  • The weka population is considered to be at carrying capacity on Kapiti.
  • The weka population has a low breeding rate with an average clutch of 1-2 eggs and 0.3 young per pair per annum in 1987 (A.J. Beauchamp, 1987).
  • On average, weka territories on Kapiti are estimated to be between 1.7 ha and 2 ha.
  • On Kapiti, 30% of the weka population are non-territorial individuals or subadults.

Questionnaire

Status quo

Actions

  • Monitor the tītī colony

Additional information

  • Under status quo there is no weka management on Kapiti to protect tītī
  • > 90% of tītī chicks are depredated by weka in the first 3 weeks following hatching

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2200.000 2600.000 3000.000 80
JZ 2400.000 3000.000 3500.000 76
QJ 2200.000 2800.000 3400.000 71
RM 2600.000 2800.000 3000.000 100
WL 2600.000 2800.000 3000.000 80
BO 2800.000 2800.000 2800.000 100
FR 2800.000 2800.000 2800.000 100
JA 2800.000 2800.000 2800.000 90
Aggregated 2250.426 2792.191 3391.482 95

Comments

Weka Numbers
  • JZ: Depends on climate change and drought impacts
  • DA: low value: depending on the dependence of weka on titi as a foodsource, if titi further decline and face (near) extirpation, the weka might reduce in numbers. But only minimally, as weka are reasonably territorial. highest value: the weka population is not dependent on titi as a foodsource. Most likely value: Weka are adaptable, and have some dependence of titi as a foodsource.
  • JA: No positive impact on titi numbers, continued decline.


Burrow modifications

Actions

  • Under the burrow modification management alternative 50 – 100 artificial tītī burrows would be installed at the tītī colony.
  • A weka exclusion flap would be placed on the entrance to each artificial burrow in early January and removed in early April.

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2200.000 2600.000 3000.000 80
JZ 2400.000 3000.000 3500.000 75
QJ 2200.000 2800.000 3400.000 71
RM 2600.000 2800.000 3000.000 100
WL 2600.000 2800.000 3000.000 80
BO 2800.000 2800.000 2800.000 100
FR 2800.000 2800.000 2800.000 100
JA 2800.000 2800.000 2800.000 90
Aggregated 2250.616 2794.605 3396.667 95

Comments

Weka Numbers
  • DA: introducing artificial burrows mimics the extirpation of titi for weka
  • JZ: The size of the colony is so small it will not make any difference to weka numbers
  • JA: I would expect little, if any, positive impact on titi chick / fledgling survival. This is because flaps are planned to be removed before titi chicks have fledged. Some weka kill chooks that are c. 2kg in weight, and young rabbits. Thus I would expect fledgling titi when exercising outside their burrows would be vulnerable to some weka. Thus I would expect the flaps to have little, if any, benefit to titi fledgling survival.

Supplementary feeding of weka

Actions

  • Under the supplementary feeding alternative supplemental food would be provided to weka at 3 locations in the tītī colony (area of c.2.8 ha) from January through April. This is hypothesised to reduce tītī depredation by satiating weka with other food (suggested food is a combination of salmon pellets and semihard cheese).

Additional information

  • Kapiti weka breeding typically extends from August to December.
  • Kapiti weka heavily rely on invertebrates for food.
  • Kapiti weka population shows high variability in breeding success over the years.
  • The colony is located in terminal forests mainly of five-finger (Pseudopanax arboreus), tawa (Beilschmiedia tawa), kanono (Coprosma grandifolia) and kohekohe (Didymocheton spectabilis).

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2600.000 2800.000 3200.00 70
FR 2800.000 2850.000 2900.00 90
JA 2800.000 2800.000 3000.00 90
JZ 2400.000 3000.000 3500.00 75
QJ 2200.000 2900.000 3500.00 65
RM 2600.000 2800.000 3000.00 100
WL 2600.000 2800.000 3000.00 80
BO 2800.000 2800.000 2800.00 100
Aggregated 2285.629 2851.695 3479.35 95

Comments

Weka Numbers
  • DA: I don’t think this will have a meaningful impact on the weka population. It may have a meaningful (negative) impact on the titi population.
  • JA: I expect this adlib supply of food would attract lots of extra non-territorial weka into the titi colony. What impact this would have on the survival of titi chicks and fledglings is hard to predict. I reckon it would be much like well-fed house cats; some still go out hunting and killing. Even if only a few of the weka attracted into the titi colony continued to hunt for titi, it would still be a disaster for the long-term viability of the titi population.

Weka fence

Actions

  • Under the weka fence alternative a fence would be placed around the tītī colony, excluding weka from an area of c.7.39 ha (the fence would be 1279 m long).
  • Upon creation, weka established within the fenced area are either killed or moved to a ferret-free area on the mainland within the rohe of Ngāti Toa Rangatira.
  • Upon creation, any overhanging vegetation within 3 m (either side of the fence) would need to be cleared.
  • Fenced area is monitored regularly to check for weka intrusion and overhanging vegetation.

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2200.00 2600.00 3000.000 80
FR 2750.00 2770.00 2790.000 95
JZ 2400.00 3000.00 3500.000 75
QJ 2200.00 2800.00 3400.000 71
RM 2600.00 2800.00 3000.000 100
WL 2600.00 2800.00 3000.000 80
BO 2800.00 2800.00 2800.000 100
JA 2800.00 2800.00 2800.000 100
Aggregated 2259.57 2791.73 3379.394 95

Comments

Weka Numbers
  • DA: This would also mimic the extirpation of titi, in addition to excluding a small amount of habitat for weka.
  • JA: Designing a weka proof fence is not difficult, Marlborough Sounds vegetable gardeners have been doing it for decades! A fence made of strong mesh that a weka cannot squeeze through, is about 1.5 m high, has a cap protruding to the outside to prevent a weka clambering over the top, and a foot on the outside to stop a weka burrowing underneath would be suitable. And, yes any vegetation above the fence would need to be removed to prevent it collapsing the fence or providing a ladder for weka to climb over it. Yes, it would be expensive of materials, transport and labour to erect, but I see this strategy as being the only viable one to keep weka out of the colony and so ensure the successful fledgling of the titi chicks.

Weka fence plus

Actions

  • Under the weka fence plus alternative a fence would be placed around the tītī colony, excluding weka from an area of c.7.39 ha (the fence would be 1279 m long).
  • Upon creation, weka established within the fenced area are either killed or moved to a ferret-free area on the mainland within the rohe of Ngāti Toa Rangatira.
  • Upon creation, any overhanging vegetation within 3 m (either side of the fence) would need to be cleared.
  • Fenced area is monitored regularly to check for weka intrusion and overhanging vegetation.
  • Speakers would also be used to play tītī calls at night to attract more tītī and boost recruitment at the colony.

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2100.000 2600.00 3000.000 80
FR 2750.000 2770.00 2790.000 95
JZ 2400.000 3000.00 3500.000 75
QJ 2200.000 2800.00 3400.000 71
RM 2600.000 2800.00 3000.000 100
WL 2600.000 2800.00 3000.000 80
BO 2800.000 2800.00 2800.000 100
JA 2800.000 2800.00 2800.000 100
Aggregated 2222.965 2781.11 3400.964 95

Comments

Weka Numbers
  • RM: I think the speaker playing tītī calls could also attract weka to the area (making it more important that the fence is 100% intact).
  • DA: no meaningful difference to weka from the ‘weka fence’ option
  • JA: As long as the fence is weka proof, and it must be given the expense of erecting and maintaining it, then using a sound system to attract extra titi into the colony site should have no impact on weka survival. Yes, it might attract more weka around the colony, but no problem because the fence will keep them out.

Burrow modifications + supplementary feeding of weka

Actions

  • Under the burrow modifications + supplementary feeding of weka alternative, 50–100 artificial tītī burrows would be installed at the tītī colony.
  • A weka exclusion flap would be placed on artificial burrow entrances in early January and removed in early April.
  • Supplemental food would be provided to weka at three locations in the tītī colony (approximately 2.8 ha) from January through April. This is hypothesised to reduce tītī depredation by satiating weka with other food (suggested food is a combination of salmon pellets and semihard cheese).

Additional information

  • Kapiti weka breeding typically extends from August to December.
  • Kapiti weka heavily rely on invertebrates for food.
  • Kapiti weka population shows high variability in breeding success over the years.
  • The colony is located in terminal forests mainly composed of five-finger (Pseudopanax arboreus), tawa (Beilschmiedia tawa), kanono (Coprosma grandifolia), and kohekohe (Didymocheton spectabilis).

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2200.0 2800.000 3200.000 70
FR 2800.0 2850.000 2900.000 90
JA 2800.0 2800.000 3000.000 90
JZ 2400.0 3000.000 3500.000 76
QJ 2200.0 2900.000 3500.000 65
RM 2600.0 2800.000 3000.000 100
WL 2600.0 2800.000 3000.000 80
BO 2800.0 2800.000 2800.000 100
Aggregated 2195.4 2829.527 3543.017 95

Comments

Weka Numbers
  • DA: as per supplementary feeding of weka, but with mimicking extirpation. This creates a broader range of potential increase and decrease
  • JA: I have no confidence that the combination of creating artificial titi burrows (even with flaps) and supplying supplementary food to weka in the colony will have any benefit for the titi population there, more than likely they will be detrimental to titi breeding success.

Weka deterrent

Actions

  • Under the weka deterrent alternative an olfactory based repellent such as peppermint essential oil would be dispensed during the day at 3 locations at the tītī colony (area of c. 2.8 ha) from January through April.
  • Audible repellents, such as karearea calls, would be played during the day at 3 locations at the tītī colony (area of c. 2.8 ha) from January through to April.
  • Concave mirrors would be put next to tītī burrows (n = 50 – 100) from January through to April.

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2200.000 2600.000 3000.000 80
JZ 2400.000 3000.000 3500.000 75
QJ 2200.000 2800.000 3400.000 71
RM 2600.000 2800.000 3000.000 100
WL 2600.000 2800.000 3000.000 80
BO 2800.000 2800.000 2800.000 100
FR 2800.000 2800.000 2800.000 100
JA 2800.000 2800.000 2800.000 100
Aggregated 2260.222 2793.273 3381.807 95

Comments

Weka Numbers
  • DA: These will likely have no effect on weka at all, seeing as weka are quick learners.
  • JA: I have no confidence that the smell of an essential oil combined with playing falcon calls will be a deterrent to weka moving into and about the titi colony. The weka is a highly adaptible species and so if there is food to be had when they are hungry then a smelly oil will not put them off. Also, while weka here in the Sounds are highly sensitive to falcon calls (alert and go to cover), if there are no consequences for weka (no falcons seen, no predations), then eventually the weka will ignore the calls and get on with foraging. Also, weka are active at night when titi are and falcons are not. I very much doubt that seeing their reflection in a mirror will have any repellent effect on weka being in the titi colony, that’s because there would be no consequences to individuals.

Weka deterrent + supplementary feeding

Actions

  • Under the weka deterrent + supplementary feeding alternative feeding would be supplied at 3 locations at the tītī colony (area of c. 2.8 ha) from January through April.
  • Olfactory based repellents such as peppermint essential oil would be dispensed during the day at 3 locations at the tītī colony (area of c. 2.8 ha) from January through to April.
  • Audible repellents, such as karearea calls, would be played during the day at 3 locations at the tītī colony (area of c. 2.8 ha) from January through to April.
  • Concave mirrors should be put next to tītī burrows (n = 50 – 100) from January through to April.

Additional information

  • Kapiti weka breeding typically extends from August to December.
  • Kapiti weka heavily rely on invertebrates for food.
  • Kapiti weka population shows high variability in breeding success over the years.
  • The colony is located in terminal forests mainly of five-finger (Pseudopanax arboreus), tawa (Beilschmiedia tawa), kanono (Coprosma grandifolia) and kohekohe (Didymocheton spectabilis).

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2600.00 2800.000 3200.00 70
FR 2800.00 2850.000 2900.00 85
JA 2800.00 2800.000 3000.00 100
JZ 2400.00 3000.000 3500.00 75
QJ 2200.00 2800.000 3400.00 71
RM 2600.00 2800.000 3000.00 100
WL 2600.00 2800.000 3000.00 80
BO 2800.00 2800.000 2800.00 100
Aggregated 2332.73 2840.822 3398.18 95

Comments

Weka Numbers
  • JA: I very much doubt that the combination of essential oil, mirrors and playback of falcon calls will have any deterrent effect on weka activity in the titi colony. As indicated above, I expect that the availability of supplementary food in the colony will attract more weka into it, and as a result there would continue to be very low breeding success of titi.

Burrow special

Actions

  • Under the burrow special management alternative 50 – 100 artificial tītī burrows would be installed at the tītī colony.
  • A weka exclusion flap would be placed on artificial burrow entries in early January and removed in early April.
  • From January through April, mana whenua would manually and daily block the entry of active tītī burrows (identified with chicks) pre-dawn and unblock pre dusk.
  • Olfactory based repellents such as peppermint essential oil would be dispensed during the day at 3 locations at the tītī colony (area of c. 2.8 ha) from January through to April.
  • Audible repellents, such as karearea calls, would be played during the day at 3 locations at the tītī colony (area of c.2.8 ha) from January through April.
  • Concave mirrors would be put next to tītī burrows (n = 50 – 100) from January through April.
  • Feeding would be supplied at 3 locations at the tītī colony (area of c.2.8 ha) from January through April. This is hypothesised to reduce tītī depredation by satiating weka with other food (suggested food is a combination of salmon pellets and semihard cheese).

Additional information

  • Kapiti weka breeding typically extends from August to December.
  • Kapiti weka heavily rely on invertebrates for food.
  • Kapiti weka population shows high variability in breeding success over the years.
  • The colony is located in terminal forests mainly of five-finger (Pseudopanax arboreus), tawa (Beilschmiedia tawa), kanono (Coprosma grandifolia) and kohekohe (Didymocheton spectabilis).

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2200.000 2800.000 3200.000 70
FR 2800.000 2850.000 2900.000 85
JZ 2400.000 3000.000 3500.000 75
QJ 2200.000 2800.000 3400.000 71
RM 2600.000 2800.000 3000.000 100
WL 2600.000 2800.000 3000.000 80
BO 2800.000 2800.000 2800.000 100
JA 2800.000 2800.000 2800.000 90
Aggregated 2236.731 2819.237 3468.169 95

Comments

Weka Numbers
  • DA: This amount of work does not sound feasible and will likely have little effect on the weka population
  • JA: The only additional activity in this strategy to hopefully enhance titi chick / fledgling survival is the blocking and unblocking of the active burrows. However, given that weka are active at night this is highly unlikely to result in improved titi chick and fledgling survival.

Select weka

Actions

  • Under the select weka alternative weka would be removed periodically (necessity based on observation) in early January and within a 250-meter radius of the colony. Selected weka are killed.

Additional information

  • Preliminary investigations show that weka that kill tītī are both territorial and non-territorial individuals established within a 250-meter radius of the colony.
  • Preliminary investigations (Stable isotope analysis and camera traps) show that weka that kill tītī are likely a group of 30 – 40 individuals.

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2000.000 2400.000 2800.000 70
JA 2500.000 2800.000 2800.000 90
JZ 2000.000 3000.000 3400.000 76
QJ 2200.000 2800.000 3400.000 71
RM 2600.000 2750.000 2950.000 100
WL 2550.000 2750.000 2950.000 80
BO 2800.000 2800.000 2800.000 100
FR 2800.000 2800.000 2800.000 100
Aggregated 2069.326 2716.713 3451.147 95

Comments

Weka Numbers
  • DA: the treatment area will behave like a sink of weka, slighly reducing their numbers
  • JZ: It depends on management ongoing. The site of the fence is very difficult terrain and it may need to be bigger.
  • JA: There may be a need to remove a few hundred weka each titi breeding season under this strategy. However given that the weka population is at carrying capacity, consists of nearly 3000 birds of which c. 30% are non-territorial individuals, I’m confident that the losses would be replaced by juveniles each year. However, I doubt that the process of selecting and removing weka will be quick enough to enhance titi chick and fledgling survival.

Select weka + burrow modification

Actions

  • Under the select weka alternative weka would be removed periodically (necessity based on observation) in early January and within a 250-meter radius of the colony. Selected weka are killed.
  • 50 – 100 artificial tītī burrows would be installed at the tītī colony.
  • A weka exclusion flap would be placed on artificial burrow entries in early January and removed in early April.

Additional information

  • Preliminary investigations show that weka that kill tītī are both territorial and non-territorial individuals established within a 250-meter radius of the colony.
  • Preliminary investigations (Stable isotope analysis and camera traps) show that weka that kill tītī are likely a group of 30 – 40 individuals.

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2000.000 2400.000 2800.000 70
JA 2500.000 2800.000 2800.000 90
JZ 2000.000 3000.000 3400.000 76
QJ 2200.000 2800.000 3400.000 71
RM 2600.000 2750.000 2950.000 100
WL 2550.000 2750.000 2950.000 80
BO 2800.000 2800.000 2800.000 100
FR 2800.000 2800.000 2800.000 100
Aggregated 2074.853 2715.675 3441.591 95

Comments

Weka Numbers
  • DA: as per ‘select’ weka, but also mimicking extirpation
  • JA: I expect that the killing of a relatively small number of weka in and nearby the titi colony will have no impact on the weka population from year to year. Also, I very much doubt that the actions suggested under this strategy would enhance the survival of titi chicks and fledglings.

Translocate weka

Actions

  • Under the translocate weka alternative there would be an annual translocation in early January of all weka within a 250-meter radius of the colony (30-40 individuals). These birds would be moved off Kapiti Island and released at a ferret-free site on the mainland within the rohe of Ngāti Toa Rangatira.

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2000.000 2400.000 2800.000 70
JA 2500.000 2800.000 2800.000 90
JZ 2400.000 3000.000 3500.000 76
QJ 2200.000 2800.000 3400.000 71
RM 2600.000 2750.000 2950.000 100
WL 2565.000 2765.000 2965.000 81
BO 2800.000 2800.000 2800.000 100
FR 2800.000 2800.000 2800.000 100
Aggregated 2139.771 2744.781 3424.197 95

Comments

Weka Numbers
  • JA: I expect that translocating weka out of the titi colony and about it will have much, if any, impact on the weka population on Kapiti from year to year. That’s because the weka population is so numerous, is at carrying capacity, and there are ample non-territorial weka to quickly replace territorial individuals. Also, I very much doubt that the actions suggested under this strategy would enhance the survival of titi chicks and fledglings.
  • JZ: I do not think this option is very plausable.

Tītī town

Actions

  • Under the tītī town alternative a weka exclusion fenced area with artificial burrows would be established at a new location on Kapiti Island. Tītī chicks from Tuteremoana and other locations would be translocated to this site annually.
  • Upon creation, weka established within the fenced area are either killed or moved to a ferret-free area on the mainland and within the rohe of Ngāti Toa Rangatira.
  • Fenced area is monitored regularly to check for weka intrusion.
  • Speakers would also be used to play tītī calls at night to attract more tītī and boost recruitment at the colony.

Additional information

  • Translocation site should be located in seral forest and cover an area of similar size to the current tītī colony (c.2.8 ha). By seral forest we mean a stage of earlier succession forest, following disturbance (clearance), characterized by transitional plant communities such as kanuka scrub, flax tussocklands, five fingers, heketara, lancewood, mahoe and mingimingi shrublands. Chosen site would be in a more accessible location than the current colony.

How many adult weka do you think will be on Kapiti by 2050 under this management strategy?


Expert Minimum Plausible Most Likely Maximum Plausible Confidence
DA 2200.000 2600.000 3000.000 80
FR 2780.000 2790.000 2800.000 95
JZ 2400.000 3000.000 3500.000 75
QJ 2200.000 2800.000 3400.000 71
RM 2600.000 2750.000 2950.000 100
WL 2600.000 2800.000 3000.000 80
BO 2800.000 2800.000 2800.000 100
JA 2800.000 2800.000 2800.000 100
Aggregated 2242.271 2788.887 3394.123 95

Comments

Weka Numbers
  • JA: I expect this strategy would have no lasting impact on the weka population. I expect that the cost of building and maintaining a weka-proof fence at a lower altitude site in seral forest would be much less than at the top of the island. As long as the translocated titi fledglings are willing to return to the new site once they reach breeding age, then the titi population would be expected to gradually increase given that very few (those that land outside the fenced area), if any, would be killed by weka. If this strategy is ultimately given the go-ahead, detailed discussions about it should be undertaken with seabird / titi experts to determine whether they think a low altitude site would be suitable for titi, and also when selecting the actual site (for example, I suspect it would need to include suitable natural or artificial take-off sites, and be free draining so that burrows don’t get flooded).
  • RM: This answer would be affected by the size of the new tītī town, but since it is not mentioned I kept the numbers as the guess as the previous answers involving killing/translocating weka under the assumption that the new area would be of a similar size.