National recovery plan for the Lord Howe Woodhen (Gallirallus sylvestris)

NSW National Parks and Wildlife Service, 2002

4. Ecology

4.1 Feeding Ecology

The Lord Howe Woodhen forages from the ground during the day and occasionally at night (Marchant and Higgins 1993). It moves slowly across the forest floor, using its bill to sift among fallen leaves and rotting timber for arthropods and for digging in the underlying soil for earthworms (Marchant and Higgins 1993). It has not been recorded using its feet to scratch (Miller and Mullette 1985).

Miller and Mullette (1985) investigated the foraging behaviour and diet of the Woodhen. Some 52.2% of foraging actions involved digging with the bill in the soil, which usually resulted in the capture of earthworms from depths as great as 10 cm. When feeding like this, individual Woodhens often worked a small area (0.5 m2) for up to 30 minutes. Some 30.7% of foraging actions involved sweeping the half open bill through the leaf litter and catching arthropods. When feeding, the Woodhen regularly moved quite large objects (eg. palm fronds) to expose earth or litter for foraging.

The vast majority of the Woodhen's diet (>80% by both frequency and weight) comprised earthworms (Miller and Mullette 1985). The same study found that white grubs (coleopteran and hemipteran larvae) were the second most consumed diet item. Informal observations have also been recorded of Woodhens taking flying cicadas, the blossoms of the green plum (Randia stipulosa), and occasionally lichen and fungi (Miller and Mullette 1985), as well as molluscs, spiders, millipedes, amphipods and isopods (Fullagar 1985, Marchant and Higgins 1993).

Woodhens readily scavenge food from walkers and residents (eg. butter, porridge, stew, biscuits, meat, chocolate and bread, Marchant and Higgins 1993). Their scavenging behaviour has included eating the flesh and insect larvae in dead Providence Petrels. Bester et al. (in prep) found that the greatest cause of Providence Petrel mortality was attributed to Woodhen predation. The Woodhen predation rate on Providence Petrels during 2000 was 14.7% (ie. 46 losses / 312 burrows with eggs). Woodhens took substantial numbers of petrel chicks, not only entering burrows to take them, but also by excavating small holes in the roof of the nest chamber through which to extract the chick below. However, the predation rate on eggs was much lower as the Woodhens would refrain from entering the burrows whilst the adult was in attendance. Woodhens took mostly those eggs that were abandoned and later pushed out of the burrow by rising floodwaters or by prospecting non-breeding birds.

It is likely that the Lord Howe Island Woodhen is not posing a threat to the long-term survival of this species at this stage, and that the Woodhen may depend upon Providence Petrel chicks when other food sources are in short supply.

Additionally the Woodhen has been observed to kill and eat introduced Rats and Mice (R. Harden, NPWS, pers. comm.).

In the settlement area, Woodhens often eat food provided for domestic poultry and are regularly observed breaking and eating chicken eggs. The Woodhen drinks fresh water from streams, pools and droplets on moss.

It is significant that earthworms, the predominant item in the Woodhen's diet, were also found to be a major item in the diet of feral Pigs located on the island. Other potential food competitors in Woodhen habitat include the Buff-banded Rail (Gallirallus philippensis) and Purple Swamphen (Porphyrio porphyrio), numbers of which appear to have increased in the settlement area in the last five years (D. Hiscox, LHIB, pers. comm.).

The self-introduced Blackbird (Turdus merula) and Song Thrush (T. philomelos) forage in similar habitat and in the same way as Woodhens and may compete with them in the settlement area, the only part of the island where these species co-occur regularly.

Studies were undertaken before the release to the wild of captive bred Woodhens to locate sites supporting good populations of soil invertebrates. The studies showed that densities of soil invertebrates were significantly higher in lowland forest than in the Woodhen's habitat on Mount Gower (Miller and Kingston 1980). Work by Miller and Mullette (1985) showed that lower altitude Woodhen territories (established after the rehabilitation of the species) showed a higher proportion of successfully breeding pairs and higher numbers of chicks fledged per pair. This may relate to the greater food resources available in lowland territories but also to the unoccupied space available for population expansion at that early stage in the species' rehabilitation.

4.2 Social Organisation

The Woodhen has been reported to be monogamous, usually pairing for life (Miller and Kingston 1980, Lourie-Fraser 1985). Observations of Woodhen pairs by Island residents indicate that birds will change partners (D. Hiscox, pers comm). If pairs split then one of them moves to fill a vacancy in an adjacent territory; seldom do they move any further (Miller and Mullette 1985). On Mount Gower in the late 1970s the population reached its lowest level of only three to four breeding pairs (Miller and Mullette 1985). Harden and Robertshaw (1988, 1989) analysed post rehabilitation banding records and found pairs were more stable in the southern mountain area of the island than in the settlement area. For example, between 1986 and 1988, in the former area, four of an original eight known pairs were still together, while in the latter area, only three of an original 16 pairs were still together. The reasons for pairs splitting are unknown.

Pairs defend an exclusive territory in which they forage and breed and, once established, rarely move further than 500 m (Harden 1986). Between 1978 and 1980, territories on Mount Gower averaged 2 to 3 ha (Lourie-Fraser 1985). Since rehabilitation, lowland territories are of the order of 1 to 4 ha (R. Harden, NPWS, pers. comm.).

Most commonly, Woodhens reproduce from August to January and continue raising young until April. However, the start and finish dates of breeding can vary between years and there are breeding records for much of the year (Miller and Mullette 1985).

Woodhens construct a nest in a shallow depression on the ground, lined with dry grass and leaves, between 10 and 25 cm across and about 2.5 cm deep. It is located on the ground under dense ferny vegetation, or in the unused burrow of a Providence Petrel (Miller and Kingston 1980). One to four eggs are laid 24 to 36 hours apart and incubated by both parents for 20 - 23 days. The young are precocial and move from the nest within 2 days of hatching. Within the territory, the pair builds 3 or 4 nursery nests in which to brood young chicks at night. Both parents brood and feed the chicks, helped by young of the previous brood. Chicks fledge at 28 days of age and are approaching adult size after about 80 days. However, they are not fully grown until about 12 months old, after which time they can be sexed based on measurements.

The adults expel juvenile Woodhens from the parental territory in the winter months (Lourie-Fraser 1985) and at about four months of age (Marchant and Higgins 1993). The juveniles then move about until they either find a mate in an existing territory or establish a new territory with a mate (Harden and Robertshaw 1988).

They can breed at nine months of age (Marchant and Higgins 1993). Juveniles that do not establish in a territory by the next breeding season generally do not survive (Harden and Robertshaw 1988, 1989). Earlier hatched juveniles have a higher survival rate than later hatched birds and it has been postulated that the former may have a time and size advantage in establishing a territory (Harden and Robertshaw 1989).

Breeding success is greater in the lowlands than on Mount Gower and greater in the settlement area than in the southern mountains (Marchant and Higgins 1993, Harden and Robertshaw 1988, 1989). For example, between 1986 and 1990, lowland pairs produced between 2.9 and 3.6 young per year, while over the same period in the southern mountains (both lowland and mountain sites), pairs produced between 1.2 and 1.7 young per year.

Conversely, between 1986 and 1995 the mean juvenile mortality in the first year was much lower on Mount Gower (41%) than southern mountains lowland sites (54%) and the settlement area (71%) (R. Harden, NPWS, pers. comm.). The reasons for these differences are not known. This relatively high avian rate of reproduction more than compensates for adult mortality but leads to a proportionally high rate of juvenile mortality in the first year (c. 60%) (Harden and Robertshaw, 1989). The high juvenile mortality rate is likely to be a consequence of their social organisation and possibly limited high quality habitat.

Some settlement residents regularly feed some Woodhen pairs residing in the vicinity of their houses. Anecdotal evidence (D. Hiscox, LHIB, pers. comm.) suggests that it is these pairs that produce the most young in the Settlement area, even when they live in apparently sub-optimal habitat (eg. garden vegetation on calcerenite). Feed provided for domestic poultry may also be an important source of food for the Woodhen within the Settlement area.

An unusual example of social organisation arose in 1981 - 82 as a result of a captive released female pairing with a wild male at the King's house near Salmon Beach (Miller and Mullette 1985). With supplementary feeding by the King's, the pair produced 16 chicks in 18 months and, within four years, had produced 36 chicks. Young from one brood assisted in the defence and feeding of chicks from subsequent broods and in the defence of the territory. This behaviour was never observed in the remnant population on Mount Gower where all available territories were occupied. The young from this exceptional pair almost certainly occupied the Far Flats and the settlement area. However, the original breeding pair died in 1995, which may have contributed to fewer Woodhens being recorded in the settlement.