Chang et al. • ALTITUDINAL CHANGES OF 8 15 N AND S 13 C IN MONTANE BIRDS 35 
TABLE 1. Longitude, latitude, altitude, and code for each feather collection site, Taroko National Park, Taiwan. 
Site 
Code 
Longitude 
Latitude 
Altitude (m) 
Bulouwan 
BL 
121° 57' 31" 
E 
24° 16' 95" N 
339 
Sibao 
SI 
121° 49' 00" 
E 
24° 20' 52" N 
956 
Loshao 
LO 
121° 45' 04" 
E 
24° 20' 78" N 
1,179 
Bilyu Sacred Tree 
BS 
121° 40' 03" 
E 
24° 18' 06" N 
2,231 
Guanyuan 
GY 
121° 33' 98" 
E 
24° 18' 76" N 
2,415 
Dayuling 
DY 
121° 31' 36" 
E 
24° 17' 96" N 
2,500 
Hehuan Farm 
HF 
121° 30' 71" 
E 
24° 17' 03" N 
2,740 
Siafongkou 
SF 
121° 29' 73" 
E 
24° 17' 20" N 
2,876 
significantly different stable isotope values in 
feathers collected from the same bird but grown in 
different years would indicate if the bird exhibited 
year-to-year molting site fidelity. Similarly, 
fluctuations in stable isotope values between 
feathers grown at different times within a single 
year would indicate altitudinal movements of the 
bird during the period of feather growth. Third, 
we explored if nitrogen and carbon isotopes can 
be used as tracers to study altitudinal migration of 
montane birds in Taiwan. If feather nitrogen and 
carbon isotopic profiles are distinguishable among 
sites along an altitudinal gradient, these two 
isotopes would provide an opportunity to sample 
individuals at other times of the year to identify 
their origins. 
METHODS 
Study Sites .—We mist-netted birds at eight 
banding sites at different altitudes (100-3,000+ m) 
in Taiwan’s Taroko National Park (Fig. 1, Table 1) 
from July 2007 to January 2008. Taroko is in the 
northern section of the Central Mountain Range of 
Taiwan, and includes high mountains and steep 
gorges. Climate and vegetation changes, along with 
altitude, in this area create vegetation zones that 
can be generally classified as broadleaved forests 
(< 1,500 m asl), mixed broadleaved and coniferous 
forests (between 1,500 and 3,000 m asl), and 
subalpine coniferous forests (>3,000 m asl) (Xu 
and Lin 1984). Most banding sites were in natural 
forests, but some areas within Sibao, Loshao, and 
Hehuan Farm have been used as vegetable 
plantations for many decades. The major vegeta¬ 
bles cultivated on these farmlands are cabbages, 
peas, spinach, and tomatoes. The overall latitudinal 
and longitudinal spreads of these eight banding 
sites are less than 1°. 
Feather Collections .—We collected feathers 
from each mist-netted bird for stable isotope 
analysis, but we also banded, classified as male or 
female (if possible), and recorded morphological 
and plumage characteristics for these birds prior 
to release. We collected the first primary feather 
from each wing of each bird mist-netted. Feathers 
which had a glossy color, no nicks in the outer 
webs, a visible terminal end of the rachis beyond 
the wing margin, and no abrasions were identified 
as new feathers grown in 2007. We assumed 
bleached and worn feathers were grown in 2006. 
We collected an old primary feather adjacent to 
the currently growing or newly grown feathers 
from each wing from those birds still in the 
process of molting, therefore having two gener¬ 
ations of feathers. Some birds were captured more 
than once, and the feathers replacing the first 
primary feathers we had pulled previously were 
again sampled. We selected primary feathers 
because they are molted shortly after breeding 
and re-grown prior to migration in most passer¬ 
ines (Pyle et al. 1997), making them the most 
likely plumage to represent the isotopic values of 
molting locations. 
All collected feathers were sealed in labeled, 
small paper envelopes and stored in a dry location 
prior to shipment to the United States for stable 
isotope analysis. All feathers were heated at 60° C 
for 30 min before shipping to meet the import 
requirements of the United States. 
Feather Cleaning and Preparation .—We ran¬ 
domly selected one of the two collected feathers, 
including old feathers grown in 2006 and new 
feathers grown in 2007, from each bird for 
analysis. Feathers were sonicated in distilled 
water for 30 min in preparation for isotopic 
analysis, followed by sonication in petroleum 
ether for an additional 30 min to remove 
contaminants from the feather surface. Feathers 
were then air-dried in a fume hood for 24 hrs. We 
cut sections weighing between 0.7 and 0.8 mg 
after feathers were cleaned and air-dried. We used 
most of the feather between the tip and the middle 
