442 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 124, No. 3. September 2012 
and. likely, abundance can vary over relatively 
short distances, perhaps in relation to abrupt 
changes in habitat that occur on the plots (Sheth 
et al. 2009). Observations also indicated consid¬ 
erable small-scale variation in abundance with 
few or no individuals recorded in some areas ol 
each plot (Fig. 5). Clumped distributions could, as 
with nets, reflect differences in habitat suitability. 
For example, Clyphorynehus was rare or absent in 
some of the wetter habitats, accounting for some 
of the spatial variation in abundance. 
Observed spatial variation in abundance might 
be partially explained by sampling methods, 
particularly for observational data. Observations 
cover a greater proportion of each plot than do 
mist nets but Clyphorynehus vocalizations are 
relatively weak and individuals are not likely to 
be equally detectable at all distances from transect 
center lines (i.e.. probabilities for detecting 
individuals may decline away from transects). 
Distance effects may account for some of the 
autocorrelation and clustering patterns in the 
count data. 
It is clear, based on recaptures of marked 
individuals, that many individuals confine most of 
their activity to relatively small areas. This is 
evident based both on recaptures within a year 
(e.g., from Jan to Mar) and between years; many 
recaptures are from the same or adjacent nets (i.e.. 
50 m from the original location) (Fig. 3). 
Differences in patterns of abundance between 
plots could arise for several reasons. Puma, for 
example, has more areas with apparently less 
suitable habitat (e.g.. dense vine tangles, swamps), 
which might account for the fact there were more 
areas on Puma with few or no observations 
(Fig. 5). Harpia tends to be more uniformly terra 
firtne although there still were areas without 
records. Differences in availability of suitable 
habitat might account for the greater within-year 
recapture distances on Puma relative to Harpia 
(Fig. 3). Individuals might move over longer 
distances in search of suitable sites or might have 
larger home ranges; either could lead to longer 
recapture distances. 
Relatively long movements associated with 
large home ranges or territories have been 
suggested to account for the frequent number of 
captures ol Clyphorynehus in mist nets (e.g., 
Marra and Remsen 1997). Some longer-distance 
movements (i.e.. >500 m) occurred at our site 
(Fig. 3) (as also reported by English [ 19981 at a 
nearby site) and may represent exploratory 
movements, as individuals typically moved back 
to original capture locations. However, move¬ 
ments of recaptured individuals in our study 
generally were relatively short (Fig. 2), indicating 
relatively small home ranges. It we assume that 
average recapture distance (Fig. 3) indicates the 
diameter of a circular home range, home ranges 
would be 0.7 ha in Harpia and 1.2 ha in Puma. If 
wc use the recapture distance as the radius ol a 
circular home range, these values would be 2.8 
and 4.7 ha, respectively, for the two plots. Home 
ranges, in either case, would be relatively small 
and similar to or smaller than reported for many 
other species in Amazonian forests (e.g.. Terborgh 
et al. 1990). Johnson et al. (2011). for example, 
reported a territory size of 5.2 ha for Glyplw- 
rynehus in terra firtne forest in Brazil, 
Approximately half of all individuals were not 
recaptured (Fig. I) and may represent transients 
or young of the year, birds that are not likely to be 
recaptured. An analysis of apparent survival rates 
(Blake and Loiscllc 2008) indicated transients 
likely were an important component of Gfypho- 
rynchus population dynamics. In other cases, 
individuals may be captured when on the edge 
of their home range with most of the range beyond 
the normal area covered by nets. Thus, with nets 
opened at a particular location for only two 
mornings per year, it is perhaps not surprising that 
some individuals are not recaptured. It is not 
uncommon for individuals of many species to not 
be recaptured for many years (unpubl. data). 
Glyphorynehus abundance also varies substan¬ 
tially at larger scales of comparison, whether 
abundance is represented by captures or observa¬ 
tions. Capture rates at our study site are the 
highest of which we arc aw'are. even when 
compared to sites where Glyphorynehus is among 
the most frequently encountered species. Beja 
et al. (2010) reported Glyphorynehus to be most 
frequently captured in terra ftrme forest in central 
Amazonia and, although capture rates were not 
given, the data provided suggest a capture rate of 
~0.7 birds/100 nin/hr. Similarly low rates (~0.6- 
1.15/100 mn/hr) have been reported from other 
South American sites such as the forest fragments 
project near Manaus (Bierregaard 1990. Stratford 
and Stoulfer 2001). Slightly higher capture rates 
of Glyphorynehus were reported in control and 
logged forest plots in Brazil (—1.1 to 2.3 birds/ 
100 mn/hr; Wunderle et al. 2006) and primary 
(2.4 birds/100 mn/hr) and logged forest (1.3-1-9 
birds/100 mn/hr) in Venezuela (Mason 1996). 
