Palila Restoration

Developing Techniques for Palila Restoration Paul Banko, Steve Dougill, Dan Goltz, Elizabeth Gray, Luanne Johnson, Matt Kweskin, Pete Oboyski, John David Semones, and Michael Wiley USGS BRD Pacific Island Ecosystems Research Center The palila (Loxioides bailleui) is an endangered Hawaiian honeycreeper that has had the longest history of monitoring of any of Hawai'i's forest birds. Annual population estimates have fluctuated greatly with a mean population of 3,390 +/- 333 birds and a range of ~1,800-5,600 birds from 1980-1995. The palila is increasingly becoming restricted to the west slope of Mauna Kea, occupying ~5% of its historical range. Palila depend on mamane (Sophora chrysophylla) forest for the majority of their dietary intake as well as for nesting sites and shelter. The quantity and quality of this dry subalpine mamane woodland has declined over many decades due the grazing pressure of introduced herbivores but is beginning to recover in many areas where ungulate numbers have been reduced. However, despite improving habitat conditions, the palila has not increased in numbers. Restoration techniques are needed in order to speed the recovery of the palila population and return it to areas of its historic range, thus preventing extinction due to catastrophic events. These techniques are being developed through population monitoring, habitat management research, translocation, and breeding enhancement. Habitat management research has focused on the mamane forest, critical to the survival of the palila. Immature mamane pods are the most frequently eaten food of the palila; however, mamane flowers, leaf buds, and native Cydia sp. and geometrid caterpillars are also important components of the palila diet. To evaluate potential translocation sites we monitored the phenology of mamane trees since 1996, expanding an earlier study initiated in 1988 on the west slope of Mauna Kea. In addition, forest composition, structure, and recruitment was determined for six study sites in the Mauna Kea Forest Reserve and two sites in Pohakuloa Training Area using the point-centered quarter method. Results Pod availability was highly variable between breeding seasons. The west slope had higher pod densities compared to the north and south slopes and the two sites in Pohakaloa Training Area (Kipuka Alala and Pflats). We are currently analyzing relationships between mamane phenology and palila density variation, nest productivity, and roosting behavior. Future studies will include a more detailed analysis of forest composition, structure, and recruitment using the point-centered quarter method. We systematically searched along transects each breeding season to measure palila's annual productivity. Highly variable rates of seasonal nesting effort and success contribute to the variability in annual population estimates. 1996: 136 nests; 36 fledglings from 48 hatchlings (75%). 1997: 49 nests; 5 fledglings from 18 hatchlings (28%). 1998 (to mid-July): 29 nests; 8 fledglings from 12 hatchlings (66.6%). We are currently researching the relationships between predation, available food resources, disease, weather, and palila nesting effort and success. Techniques are also being developed for an "at the nest" supplemental feeding program (using high protein food like insects) in order to increase nesting productivity. Two translocations have been conducted using similar methods. The first translocation (T1), targeted second-year (SY) palila caught between Feb. and Mar. 1997. A total of 25 birds were mist-netted from the west slope core population, fitted with a radio transmitter and transported in individual cages by 4wd vehicles to the north slope of Mauna Kea within the Mauna Kea Forest Reserve. The north slope was chosen as it was within the palila?s historical range, contained suitable food resources, and had potential for further forest recovery and management. The second translocation (T2), targeting hatch-year (HY) birds, commenced in Oct. 1997, and ended in Mar. 1998. A total of 24 birds were transported to the north slope of Mauna Kea as in T1. To encourage breeding we moved after-second year (ASY) males and SY females in the later stages of the second translocation. After release in both T1 and T2 we located and observed birds daily using radio tracking receivers. Locations were recorded using a GPS unit to an accuracy of " 10m. In addition, daily roost sites were located for palila during T2. Native Cydia (Tortricidae) caterpillars are an important food resource for palila. These caterpillars feed on m~mane seeds and are consumed by palila feeding on mamane pods. We examined palila feces to determine the incidence of caterpillar consumption for nestling and adult birds. We also examined the frequency of parasitism on caterpillars by introduced wasps. We found caterpillar parts in 97% of nestling palila feces and 64% of adult palila feces. Caterpillar parasitism rates varied form 94% at low (6,200ft) elevation to 21% at high (9,000ft) elevation. We censused rats and mice (Mus musculus) with snap-traps placed along transects, ranging from tree-line to the forest reserve boundary. Traps were spaced at 150 m intervals and alternately placed on the ground or on a mamane limb. Feral cats and mongooses (Herpestes auropunctatus) were captured in live traps. We placed cat traps at the top, middle, and bottom of each transect. Additional cat traps were placed in areas where translocated palila or feral cat tracks were commonly seen. Acknowledgments: We would like to thank the U. S. Army Garrison, Hawai'i for funding the research. In addition, we would like to thank the United States Fish and Wildlife Service for their assistance, Department of Forestry and Wildlife for assistance and permission to work on Mauna Kea, The Peregrine Fund, and the many interns who aided with field work. Figure titles:

A second - year palila.



Mamane woodland on Mauna Kea.
Historic and current range of palila on the Big Island of Hawai'i.


Pod availability during the three breeding seasons (Mar. - May).


Results of translocations (T1 and T2) at the end of 8 weeks.
Feral cat capture rates on Mauna Kea.


Palila chicks depredated in a nest.