Segura et al. • MICROHABITAT EFFECT ON NEST SURVIVAL 
507 
(Celtis tala), and secondarily in Coronillo (Scutia 
hiixifolia) and Molle (Schinus longifo/ius) (Segura 
and Arturi 2009). 
Our objective was lo evaluate the influence 
of surrounding vegetation cover on nest survival 
of Red-crested Cardinals in natural forests of 
central eastern Argentina. We hypothesized that 
increased nest cover reduces susceptibility to 
predation, and predicted higher microhabitat 
cover would have a significant positive influence 
on nest survival. 
METHODS 
Study Site.—We conducted the study at 'Es- 
tancia La Matilde' (35° 20' S, 57 G IT W) in 
central eastern Buenos Aires. Argentina. The 
study site was a flat area of 400 ha within the 
Biosphere Reserve Parque Costero del Sur (MAB- 
I'NESCO). It is semi-open grassland with several 
low chains of woodlands, mainly dominated by 
native tree species including Celtis chrcnhergiana 
(Tala, deciduous). Scutia hiixifolia (Coronillo. 
evergreen), and Schinus longifo/ius (Molle). 
Red-crested Cardinals are present in the study 
area during the reproductive and non-rcproductive 
seasons (Segura and Arturi 2012). Potential 
terrestrial nest predators in these forests arc 
white-eared opossum {Diddphis albiventris). 
lesser grison (Galictis cuja ). snakes (Philodryas 
'PP), and small rodents. Potential aerial nest 
predators are: Guira Cuckoo (Guira guira). 
Chi mango Careara (Milvago chi man go), and 
Narrow-billed Woodcreeper (Lepidocoktptes an- 
gustirostris). 
Nest Monitoring.—We collected data over three 
consecutive breeding seasons from 2005 to 2008. 
monitored 106 Red-crested Cardinal nests 
annually from October through February by 
extensively searching among suitable nesting 
habitat. We monitored the nests daily during the 
egg laying and hatching stages, and every 2 days 
tinting incubation and nestling stages. Nesting 
attempts that did not reach egg laying stage (i.e., 
nests in construction) and nests abandoned during 
egg laying or incubation were not considered in 
the analysis. Nests that failed due to Philornis 
ectoparasitism (Segura and Reboreda 2011) were 
not considered. We examined nest contents on 
each visit, by taking all eggs or chicks from the 
nest (Segura 2011 provides details of the null 
effect of nest monitoring on nest success). We 
checked nests until fledglings had left the nest or 
until depredated. We considered a nest successful 
if at least one young fledged. Nests with signs of 
predation or where chicks disappeared before the 
earliest possible Hedging date were considered 
depredated. The entire breeding cycle was 27 days 
(egg laying + incubation + nestling stages; Segura 
2011). 
We recorded physical characteristics of vege¬ 
tation cover in a 50-cm radius around the nest 
immediately after fledging. We measured the 
presence of leaves and branches at intervals of 
10 cm in a horizontal straight line in each of the 
four cardinal directions centering on the nest, and 
also 50 cm vertically above and below the nest. 
These measurements were taken twice at each 
nest, at the northern and southern side of the nest 
separately. We recorded the absence (0 = 0%), 
weak presence (1 = 1-50%), and abundant 
presence (2 = >50%) of leaves and branches 
covering the nest in each 10-cm interval. We 
calculated (he average for horizontal, above, and 
below measurements. We assumed vegetation 
cover surrounding the nest did not change 
throughout the breeding cycle (27 days). There 
was no significant association between our 
measurements of cover and date of the breeding 
season (Spearman's rank correlation; horizontal: 
p = - 0.04. P - 0.66; above: p = 0.01. P = 0.94; 
and below: p = —0.01, P — 0.95). 
We included date of the season and year as 
additional variables to control for intra- and inter¬ 
annual variation. Age of the nest was included as 
a variable to control for intra-nesting cycle 
variation- We also included physical characteris¬ 
tics of the nest site that may influence nest 
survival: (1) tree species where the nest was built 
(Tala and Coronillo trees). (2) nest height from 
the ground (m), and (3) nest location within the 
forest (‘center’ if the nest-tree was in the center of 
the continuous chains of forest parallel to the 
river, 'border' if it was in the border of the chains 
of forest, and ‘patches' if it was in small isolated 
forest patches more distant from (lie river). 
Data Analysis. —We estimated daily survival 
rates (DSR) using Program MARK (White and 
Burnham 1999. Dinsmore et al. 2002). Encounter 
histories were coded following Dinsmore et al. 
(2002). We calculated the number of days in each 
encounter history relative lo a date prior to the 
earliest initiated nest (I Oct = day 0). We used 
Akaike's Information Criterion adjusted for small 
sample sizes to compare models based on log- 
likelihood values (Burnham and Anderson 2002). 
We built all models without standardizing 
