730 
THE WILSON JOURNAL OF ORNITHOLOGY . Vol 124. No. 4. December 2012 
*“» *— a, Woodbun 
Day 
0 
Nestling uppcarance 
Dark pink mass, head down 
£S=S«“~s^sr.-— 
Pin feathers he,„£ sarfaee *"» «" —* 
’XSiZ7 d ° W " y Pi " “ «*•• » "neign down midline of back 
alUivcr telly" 0 ° f baCk and nCCk - SidCS ° f brCaSt ' chin ’ ***** of ^ 
- Set ::t r ,h shc r s ^ rr co,cm - ~ •* <-» 
uTmS ' 5 £*-2 ~ “rs 2:,: 
Two pronounced lulls of down on head' head f„|’| v V' rl' a,,d other WI "S lathers still in feather sheaths 
breas, fully feathered, T* ***" ^ ° f ^ ‘ >U1 offc-hr sheaths: 
devcioping tha, nre short end s.uhhvf f,=^ *“■ ^ °” -* « *=■*■" 
Mayfield’s ,1975) approach, bu, we included il 
or comparison purposes with other studies of 
Swatnsons Warbkrs. The logistic exposure 
method, a binomial distribution, requires each 
exposure interval to be assigned either a 0 or a 1 
°r*~ intervals, respectively 
(Shaffer -OWi. Exposure intervals of known and 
unknown lale were truncated as suggested hy 
Manohs et al. ,2000). We assigned each day of the 
breeding season a number starting with 28 April 
the earliest nest initiation we observed Nest 
observations were frequently >] day and we 
eraged the exposure period to assign a date to 
each interval. Observa.ions from the budding 
stage were excluded from the analysis; thus nest 
age began with the hatching day A wa 
Zffo £ Tr* if 11 SL one 
CBurhiins “ l ”"* "* pe ™ d 
(BumVamanTA T Candid “'= models 
urnnam and Anderson 2002), based on o..r n,mi 
5&3F*3s& 
2) day oi ,he - vear ^rant et al ?()05i Ju 
or (3) nest age or stage (Gran, e, al. 2005) We dfd 
not assume constant survival during the nesting 
cycle and modeled linear, quadratic, and cubic 
ellects of nest age as well as linear and quadratic 
effects of date following Grant et al. (2005). The 
biological models focused on Brown-headed 
Cowbird parasitism. The features examined for 
habitat and landscape models focused on nest 
height, cane density, and distance of nests to edges 
including roads, sloughs, and swamps. 
We used logistic-exposure models in PROC 
GHNMOD (SAS Institute 2004. Shaffer 2004) and 
calculated Pearson correlation coefficients for the 
oZ ble , S , USing PR0C CORR (SAS Institute 
-004). Highly correlated variables (r > 0.70) 
vvere not included in the same model. We used 
Hie I losmer and Lemeshow (2000) goodness of fit 
,eM lo S au S e 'he fit of the global model. Candidate 
models were ranked according to Akaike's 
Information Criterion for small sample sizes 
(AIC ) (Bumham and Anderson 2002). Models 
with lower AAIC r values and higher AIC, weights 
were considered to he the best supported. ' 
\Ve used null hypothesis testing, setting P < 
' fo f significance, to compare the effect of 
roun-headed Cowbird parasitism on hatching 
an Hedging rates of Swainson’s Warblers. All 
estimates relating to the length of nesting periods 
M C l ," 1 Ch SIZe are ex P re ssed mean ±95% Cl. 
,;! y IC dai, y surv ival rates and microclimate 
dta are expressed as mean ± SE. 
