678 
THE WILSON JOURNAL OF ORNITHOLOGY • Voi 124, No. 4. December 2012 
global detection function using the post-stratifica¬ 
tion approach, and variances and confidence 
intervals were derived by bootstrap methods in 
DISTANCE from 999 iterations (Thomas et al. 
2005). 
We used incidental Oma'o detections to 
identify suitable habitat to delineate the distribu¬ 
tion of “Oma'o in scrub alpine habitat. Formal 
passerine surveys have not been conducted in 
scrub alpine habitats on Mauna Loa (Camp el al. 
2009, Judge et al. 2011). Incidental “Oma’o 
detections were made during seabird nest search¬ 
ing and colony monitoring in the scrub alpine 
habitat on Mauna Loa between 2006 and 2010 
(Apr-Nov: SWJ, pers. obs.). We also compiled 
‘Oma'o detections since the 1990s recorded by 
HA VO, The Nature Conservancy, and USGS 
researchers who conducted botanical surveys in 
the Mauna Loa scrub alpine life zone (Ziegler 
2002). Detection rates were not calculated be¬ 
cause observations were not enumerated. 
RESULTS 
Oma'o were detected in high elevation wood¬ 
land habitat on the leeward 'dopes of Mauna Loa 
during point-transect surveys in 2010 after 
>30 yrs of apparent absence in the region. We 
detected 23 “Oma'o at 19 stations in the Papa 
(9 birds) and Northwest Kahuku (14 birds) tracts 
between 1,800 and 2,500 m elevation (Fig. I). We 
did not detect 'Oma’o in the Honomalino tract, 
which is predominantly wet mesic forest. Our 
detections occurred in the transition between 
woodland and alpine zones in habitat dominated 
by mamane. naio, and ’ohi'a trees. Canopy cover 
was classed as scattered (5-25% cover) in both 
tracts, and canopy height was mixed ranging 
between 5 and 10 m in the Papa tract and 5 to 
>10 m in the Northwest Kahuku tract. The 
understory' in both tracts was primarily bare 
ground due to ungulate grazing and lava flows, 
yet native shrubs and introduced grasses were 
present. Densities were relatively low at the two 
leeward tracts with an estimated (± SF) 0 16 ■+• 
0.05 and 0.50 ± 0.22 birds/ha in the Papa and 
Noithwest Kahuku tracts, respectively. 
Incidental sightings of ‘Oma'o in the Mauna 
Loa scrub alpine life zone ranged from 2.000 to 
3,200 m and only occurred on lava How substrates 
that were 800 to 5.000 years before present 
Oma“o were most often observed on puhoehoe, 
or smooth rope-like lava, as opposed to “a“a that 
consists of brittle and sharp chunks of lava. 
“Oma'o were also detected on infrared motion¬ 
sensor cameras trained on Hawaiian Petrel 
(Pterodroma simdwichensis) burrows at 2,850 m 
(Judge et al. in press). We identified 74.700 ha of 
potential habitat in the scrub alpine life zone on 
Mauna Loa. given the elevation, substrate age. 
and flow type where “Oma'o were observed 
(Fig. 1). There may be several thousand ‘Oma'o 
in the scrub alpine habitat based on our woodland 
estimates. 
DISCUSSION 
We detected ‘Oma'o in woodland habitat in 
relatively low densities during a 2010 bird survey 
of HA VO after the species was absent from 
leeward forests and woodlands on Hawai'i Island 
for >3 decades. A previous survey in 2005 on the 
fixed panel transects yielded no detections of 
“Oma'o (Tweed et al. 2007). nor had 'Oma'o been 
detected in any leeward forest or woodland during 
10 point-transect surveys dating to 1978 (Gorre- 
sen cl al. 2009). 'Oma'o detections in leeward 
woodlands of Hawai'i Island are significant 
because they comprise a major expansion in the 
species’ known range. The species is now 
distributed from windward wet, mid-elevation 
forests up to sparsely vegetated, high elevation 
scrub alpine and woodland habitats in both 
windward and leeward regions. 
It is likely the birds detected in the woodland 
habitat came from one of two local sources: 
colonizing individuals from leeward scrub alpine 
or from windward forest populations. Banding 
studies in mesic and wet forest habitats indicated 
Oma'o were highly sedentary (Ralph and Fancy 
1994, Wakclee 1996). Thus, it is possible the bird." 
we observed were not migratory or transitory but 
instead have dispersed into and re-colonized 
woodland habitats. It is most likely the leeward 
biids were from the nearby scrub alpine popula¬ 
tion that is now occupying the woodland-alpine 
transition zone. ‘Oma'o dispersing from the 
windward forests to leeward woodlands would 
have to either cross high elevation alpine habitats 
or lower elevation degraded forests (i.e., habitats 
without fruiting trees and shrubs). There are no 
data available to indicate if the birds migrated 
through alpine or degraded forests. A likely cause 
lor birds to disperse is that resources in the 
windward forests may be limiting because of 
drought (Timm and Diaz 2009) or habitat 
degradation (PMG, pers. obs.). It seems unlikely 
that 'Oma'o would have traveled 15-30 km across 
