Blake and Loiselle • VARIATION IN GL YPHOR YNCHUS ABUNDANCE 
437 
and recaptures provides the data needed for such 
estimates (Jullien and Clobert 2000, Blake and 
Loiselle 2008). Glyphorynchus has figured prom¬ 
inently in studies that used molecular approaches 
to examine patterns of diversification (e.g.. Bates 
2000, 2002; Mila et al. 2009; Thomassen et aJ. 
2009). 
The major goal of our study was to provide 
information on temporal and spatial patterns of 
Glyphorynchus abundance from a lowland forest 
site in Ecuador. We used netting and observation 
data to examine spatial variation in abundance (we 
use the term abundance to refer to numbers of 
birds, whether based on captures or observations) 
at both within and between-plot (100-ha plot) 
scales, movements (based on recapture informa¬ 
tion). patterns of habitat use (based on captures and 
observations), and temporal changes in these 
metrics. Specifically, we examined whether Gly- 
phorynchus was randomly distributed across 
two 100-ha study plots or whether it displayed a 
clumped distribution pattern indicative of responses 
to local habitat differences; whether abundance is 
relatively stable across years; and whether patterns 
of abundance (both spatial and temporal) are similar 
on two 100-ha plots in the same area of lowland 
forest. Our results provide a unique perspective on 
the distribution and abundance patterns of an 
important neotropical species. 
METHODS 
Study Site.—We conducted our research at 
Tiputini Biodiversity Station (TBS), Orellana 
Province, Ecuador (—00 37' S, 76 10' W; 
190-270 m elevation). TBS was founded in 1994 
by Universidad San Francisco de Quito (USFQ) 
on a tract of undisturbed lowland rainforest within 
the — 1.7-million ha Yasum Biosphere Reserve, 
one of the most biologically diverse regions of the 
world (Bass et al. 2010). The station and nearby 
areas support a variety of habitats including terra 
lirine and vdrzea forest, palm (e.g.. Manritia 
flexuosa) swamps and other wetlands, as well as 
areas of natural succession that follow treefalls, 
windthrows. or other natural disturbances. The 
mean annual precipitation at Yasum' Research 
Station. —30 km WSW of TBS, is —3,100 mm 
(http://w wav. biologia.pucc.edu.ec/natura.plip ?c=337). 
Wet months are from April through June: January 
and August are relatively dry with January often 
very dry (pers. obs. l. 
Two —100-ha plots (Harpia and Puma; -I X 
1 km each) were established in terra firme forest 
during 2001. Both plots were gridded (100 X 
200 m grid lines) and marked W'ith tagged, 1.5-m 
tall PVC tubes every 50 m along each grid line. 
The Harpia plot ranges from 201 to 233 m in 
elevation and is characterized by dissected upland 
forest. The Puma plot has less topographic relief 
overall, although the range in elevation is similar, 
from 209 to 235 m. Both areas experience partial, 
temporary inundation ( — 5 to 10 ha. depending on 
the height of the flood) when small streams fill 
and overflow their banks as the Tiputini River 
rises; Puma has more areas that fill with standing 
water after heavy rains. 
Points at 50-m intervals within each plot were 
classified into one of six habitat categories 
(mature upland forest, mixed upland forest, 
palm-hardwood swamp, palm swamp, second 
growth, gap) based on a subjective evaluation of 
the dominant habitat type within a 25-m radius of 
that point in 2005. Most of the habitat types have 
remained the same throughout the study despite 
some changes in habitats around points (e.g., as a 
result of treefalls). 
Sampling Methods: Captures .—Birds were 
captured with mist nets (12 X 2.6 m. 36-mm 
mesh) set at ground level. Nets were arranged in a 
series of eight sets of 12 nets on each plot (96 nets 
per plot). Each set of 12 nets formed a rectangle 
(100 X 200 in) with nets set — 50 m apart; max¬ 
imum distance between nets on a plot was 
—900 m. Nets on the two plots were —1.7 km 
apart at their closest point. Barlow and Peres 
(2004) concluded, based on recaptures of marked 
individuals, that plots 500 m apart in the Brazilian 
Amazon were spatially independent; <1% of 
birds were recaptured between plots in their study. 
We recorded six recaptures (all species combined) 
between plots (<0.1% of total captures; >13,000 
captures) and have reobserved only a few' color- 
banded individuals between plots. Each set of nets 
was operated for 1 day (—0600 to 1230 hrs) in 
January (peak of breeding for many species) and 
March (late breeding season for many species), 
starling in March 2001 and ending in March 2010. 
All captured birds were banded with numbered, 
aluminum leg bands. Mist nets, as with any 
sampling method, are subject to biases, but 
because both plots in this study were sampled 
the same way. any biases associated with mist 
nets are similar on both plots. 
Sampling Methods: Observations .—We used 
direct observations to sample birds during Februarv 
(2002-2006. 2009. 2010) and April (2002-2005); 
