Ocampo et al. • BREEDING BIOLOGY OF HYPO PYRRHUS 
539 
FIG. 1. (A) The Alto San Miguel in the nonhem central 
Andes. Depanment of Antioquia. Colombia (06 02' 05" N. 
75 36' 58" W). (B) Aerial photograph depicting the end¬ 
points of our 2-km transect along the upper Medellin River 
at Alto San Miguel where seven groups of Red-bellied 
Grackles (Hypopyrrhus pyrohypogaster) were nesting co¬ 
operatively for at least 4 consecutive years (2006-2009). 
watershed of the Medellin River in the northern 
central cordillera of the Andes, Department of 
Antioquia. Colombia (Fig. 1). The region is in the 
lower montane wet forest’ life zone at an 
elevation of 1,800 to 2,100 m: the mean annual 
temperature and precipitation is 16 C and 
TOGO mm, respectively (Espinal 1992). The upper 
watershed of the Medellin River has a long history 
°l human disturbance due to the influence of 
'mall urban settlements and conversion of forest 
pastures for cattle ranching and for forest 
plantations. The current landscape is dominated 
hv second growth, pastures with isolated trees, 
semi-open areas at different stages of succession, 
and Eucalyptus and pine (Pinas panda) planta¬ 
ins. Patches of native montane forest remain in 
'he upper hills and the Alto San Miguel Nature 
Reserve. Red-bellied Grackles have been ob¬ 
served nesting at this location since at least 
'999 (AMC, pers. obs.). We describe the nests 
and behaviors of seven groups observed along the 
river. 
We followed individuals along a 2-km transect 
tr °m January through July 2006—2009 along the 
Medellin River that exhibited nesting or parental 
behavior and searched for nests and nesting 
groups following Martin and Geupel (1993). We 
recorded nest stage (under construction, incuba- 
hon. nestling, or inactive), the species of tree or 
s hrub, and height for each nest. We measured 
nests whenever possible (inner and external 
diameters, depth, and height) after nestlings 
Hedged or nests were depredated, and eggs 
(length, width, fresh mass at day 0); we weighed 
eggs every 2 days for one nest to estimate mass 
loss rate during incubation. We measured nest¬ 
lings' body mass every day until they fledged, and 
estimated nestling growth rate following Martin 
et al. (2011) using the logistic growth curve based 
on the equation: W (t) = A/( 1 + e 1 K< " w'here 
W(i) denotes body mass of a nestling at time /. A is 
the asymptotic mass that nestlings approach, f, is 
the inflection point of the curve, and K is a 
constant scaling rale of growth (Ricklefs 1967. 
Ricklefs 1968. Rentes and Martin 2002). All 
measurements were taken on site w ith an accuracy 
of 0.05 g (FlipScale F2 ) for mass and 0.1 mm for 
external measurements with a caliper. 
We monitored nests every 1 to 3 days during 
incubation, or until the last egg in the clutch 
hatched, and during the nestling period from 
clutch hatching until nestlings fledged. Every 
monitored nest was observed for 1 hr in the 
mornings within the 0900-1100 hrs period, and 
from distances > 15 m. We recorded brooding 
behavior, nest attentiveness (proportion of time 
adults were on the nest over total time of 
observation), provisioning rate (food deliveries/ 
hr), and type of food items (c.g., arthropods, fruits 
and flower parts, and small vertebrates) given to 
nestlings by direct observation. We estimated the 
proportion of eggs that hatched (hatching success) 
and the proportion of young that fledged (fledg¬ 
ling success) based on observations of 21 nests. 
We calculated daily survival rate and nesting 
success using the Mayfield estimator (Mayfield 
1961), but used exposure days in incubation and 
nestling periods combined because we did not 
have information on the exact hatching day for a 
number of clutches. We assumed nestlings 
Hedged successfully when nests were empty and 
Hedglings remained in the group territory'. We 
observed the first and subsequent hatchings of the 
clutch of two nests allowing us to ascertain if all 
eggs of the clutch hatched synchronously (within 
a 24-hr period). 
Size. age. and gender (M/Ft composition of 
each breeding group was based on counts of the 
number of adult and immature individuals. Adults 
were recognized by their scarlet bellies and shiny 
black plumage, and their bicolored irides (pale 
yellow with a bright red outer margin). First-year 
birds were recognized by their pale red belly, dull 
