48 
PACIFIC SCIENCE, January, 1947 
100 birds an acre at times at Kipuka Puaulu 
(4,000 feet). Guava thickets at elevations 
of 1,000 to 2,000 feet are also tenanted in 
good numbers. However, virgin tree fern 
forests favorable to the native Hawaiian 
birds are in some areas without Leiothrix. A 
forest of this type at the Twin Craters (3,800 
feet) in Hawaii National Park usually had 
not more than one individual per acre. Simi¬ 
lar forests of the Upper Olaa Forest Reserve 
near a farming district exhibited a somewhat 
greater density. 
On Maui in the Kipahulu Valley the vir¬ 
gin, undisturbed ohia lehua forests (eleva¬ 
tion 4,000 to 6,000 feet and rainfall 250 to 
350 inches a year) had very few birds of this 
species. In koa forests in this same valley, 
but between 2,000 and 4,000 feet elevation 
and with a rainfall of 150 to 250 inches a 
year, the population density of this species 
was slightly greater, an indication that the 
bird was generally present, but uncommon. 
The birds are seldom seen more than 15 
feet above the ground and then only when 
the ground cover is very thick, or when the 
birds are in heavy tree foliage. Only occa¬ 
sionally can they be observed in the higher 
strata of the ohia and koa trees. Most fre¬ 
quently the birds are heard in the low thickets 
and not seen. If one is quiet it is possible to 
observe the bird flitting from shrub to shrub 
next to the ground. These movements com¬ 
bine short fluttering flights and hops of a few 
inches. 
FLOCKING 
We have seldom observed a flight of more 
than approximately 50 feet, but others report 
that flights of 200 feet are not uncommon in 
the more open parts of the range of the 
species. Longer flights are more frequently 
found during the winter, when the birds are 
banded together into traveling groups of 20 
or even 100 birds. Some of the flocks in 
favorable places seem to confine their move¬ 
ments to a limited area, as at Kipuka Puaulu 
on Hawaii. However, some observers think 
the flocks represent groups migrating into 
the lowlands for the winter. There is a 
movement at this time, but it is more of a 
general diffusion into peripheral areas not 
heavily occupied during the breeding season, 
and it is by no means limited to downward 
movements, for, as mentioned previously, in 
winter the birds appear conspicuously in the 
ohia lehua forest of Mauna Loa at high ele¬ 
vations. 
By May at Kipuka Puaulu the bands have 
broken up, and most of the birds are segre¬ 
gated by pairs. By late summer (July and 
August) small bands of 4 to 12 birds are 
common. Consequently, it appears that flocks 
are re-formed almost immediately after pa¬ 
rental duties at the nest are completed. How¬ 
ever, some of the birds remain in flocks 
throughout most of the breeding season 
(March through June). 
NESTING 
Perhaps nesting occurs at all elevations 
found in the spring and summer range of 
Leiothrix, but on Oahu nests have been 
found only between 500 and 2,500 feet ele¬ 
vation. On Hawaii the highest elevation at 
which a nest was found was 6,100 feet at 
Kipuka Kulalio. Without exception on Oahu 
the nests have been less than 10 feet from 
the ground and located in the densest parts 
of the understory; one nest was only 18 
inches above the ground. On Hawaii nests 
have been found from 3 to 7 feet from the 
ground. Choice of the nest site seems to 
depend not so much on the particular kind 
of vegetation as on its density, for nests on 
Oahu have been found in staghorn fern 
(Dicranopteris linearis ), oi, palm foxtail, 
and guava, on the finer branches of low- 
growing hau trees, and attached to the stems 
of ti leaves. On Hawaii, sites selected for 
nests included dense shrubs (aalii, Dodonaea 
viscosa; pukeawe, Styphelia Tameiameiae ), 
the branches of trees (mamani, Sophora 
