With assistance from three trained observers, daily censuses began when adult BCNH arrived at the park (2 April 2010 and 30 March 2011) and continued until all young had dispersed (20 September 2010 and 12 August 2011).
Movement of BCNH during censuses was minimal and consisted primarily of nest building and courtship.
Peak abundance of active nests can be reliably determined from a single count conducted during peak nesting season because of the conspicuous nature of active BCNH nests (Kelly et al., 2007).
The ratio of young to active nests, or young to nesting pairs, is frequently used to quantify reproductive success in colonial birds including BCNH (e.g., Tremblay and Ellison, 1979; Hoefler, 1980; Crouch et al., 2002; Hothem and Hatch, 2004; Levengood et al., 2005; Kelly et al., 2007).
Young were unlikely to have dispersed before the peak abundance censuses on 14 June 2010 and 8 June 2011 because BCNH disperse at approximately 58 d of age (Levengood et al., 2005) and peak abundances of young occurred less than 58 d after the first young hatched.
Because of the endangered status of BCNH in Illinois and under advisement of the Illinois Department of Natural Resources observation methods involving touching BCNH (e.g., banding) were avoided (Maggie Cole, pers.
The close proximity of two habitat patches (grove and avenue) and my ability to census the entire BCNH colony daily presented an opportunity to investigate whether the colony exhibited behavioral flexibility in terms of fine-scale habitat patch selection in response to fluctuations in colony size.
To determine the effect of colony size on habitat patch selection, I used logistic regression wherein the habitat patch selection of an individual BCNH was the binary dependent variable of interest.
The first predictor was colony size ([n.sub.total]), defined as the number of adult BCNH in the colony determined via census.
BCNH nested in very high density in Lincoln Park, with 217 nests/ha in 2010 and 315 nests/ha in 2011.
In 2010 although many BCNH pairs initially began nesting on the grove, only one pair remained there for the entire breeding season and they did not produce young.
where [P.sub.avenue] was the probability of selecting the avenue habitat patch, [n.sub.colony] was number of BCNH in the colony and y was a categorical variable for year (2010 or 2011).