Baubock et al. • CLIMATE CHANGE DOES NOT AFFECT PROTANDRY 
209 
bird arrival dates at the Center, which is on the 
western side of Cape Cod Bay. Plymouth County. 
Massachusetts. USA (4U 50' N. 70 30' W). This 
coastal migratory route is characterized by brushy, 
second growth deciduous woodland 7-10 m in 
height, dominated by oak ( Quercus ) tree species, 
encaceous shrubs, and thorny greenbrier (Smilax 
spp.) vines. It is bordered on the cast and south by 
a steep, eroding coastal bluff and on the west and 
north by brushy wetlands. Average annual tem¬ 
perature during the period of study was 9.7 C. 
and annual precipitation was 129.5 cm. 
The Center operated between 45 and 50 nylon 
mist nets (12 X 2.6 m, 4 panels. 36 mm extended 
mesh) annually from 1970 to 2002, inclusive. 
Less complete coverage and imprecise records 
regarding capture effort expended were available 
during the first 4 years of the observatory's 
existence (1966-1969) and we excluded these 
years from analysis. Each bird captured was 
marked with a unique identification band, which 
prevented double counting of individuals. Nets 
were kept at fixed locations throughout the study. 
Nets were generally operated from 0.5 hr prior to 
sunrise to 0,5 hr after sunset, except for closures 
during adverse weather conditions. Thus, 50 nets 
open for 12 hrs equals 600 net hrs per day. 
Sampling was conducted from 5 to 7 days per 
week during the spring (15 Apr-15 Jun) migration 
period. 
The staff banded 205,454 individuals of 159 
species during the study years. We considered 
only passerines for which males and females 
could be reliably distinguished (n = 7 species) 
(Table I). We considered the first capture of a 
bird as that individual's arrival date—i.e., we did 
not consider repeat captures for individual birds. 
We did not consider locally breeding birds, 
which we identified on the basis of well- 
developed brood patches or cloacal protuberanc¬ 
es; we did not consider spring captures of 
hatching-year individuals. 
The data we analyzed represented birds cap¬ 
tured and banded during spring migration. Indi¬ 
viduals from each species did not necessarily 
belong to a single breeding population. Thus, w-e 
reler to our data as representing migration 
cohorts—i.e.. groups of birds that passed through 
Manomet during spring migration. The banding 
data are a subset of those used in a previous study 
of migratory responses to climate change (Miller- 
Rushing et al. 2008). 
Climate Data .—We calculated mean monthly 
temperatures for Manomet. Massachusetts by 
averaging data from five weather stations within 
a 20-km radius. We avoided possible microcli¬ 
matic anomalies that may occur at a single station 
by averaging temperatures from several weather 
stations (Pielke et al. 2002). Mean temperatures in 
March. April, and May were significantly corre¬ 
lated from Wilmington. North Carolina north to 
Manomet (P < 0.05 for all correlations). We 
calculated the mean arrival date for each bird 
species and averaged the previous 2 months’ 
temperatures to obtain the relevant mean temper¬ 
ature index for each bird. Wc obtained the 
monthly minimum and maximum temperatures 
from the Plymouth-Kingston weather station in 
Plymouth, Massachusetts. 
We obtained annual indices for North Atlantic 
Oscillation (NAO) and the El Nino Southern 
Oscillation (ENSO) from the National Center for 
Atmospheric Research (Boulder. CO. USA; www. 
cgd.ucar.edu/). These indices integrate many 
climate variables over large regional areas. 
Positive NAO index values correlate with south¬ 
erly winds in the eastern United States (Hurrell 
1995). while negative values correlate with 
increased precipitation in (he eastern United 
States (Mauget 2006). Positive ENSO index 
values correlate with relatively warm tempera¬ 
tures during winter and spring in the southeastern 
United States, and with relatively cool tempera¬ 
tures in Central America, southern Caribbean, and 
much of South America (Trenberth and Caron 
2000). Positive ENSO index values also correlate 
with relatively dry years in the southeastern 
United States and with relatively wet years in 
Central America, southern Caribbean, and north¬ 
ern South America (Trenberth and Caron 2000). 
March-May temperatures at Manomet were not 
correlated with NAO (/■ = 0.03, P = 0.89) nor 
ENSO (/- = 0.03, P = 0.86). 
Data Analysis.—Wc converted all arrival dates 
from calendar dates to days after the vernal 
equinox. This transformation avoided biases in 
Julian or calendar dates caused by changes in 
timing of the vernal equinox. We calculated the 
mean arrival date separately for males and 
females of each species in each year, differences 
between male and female arrival, number of 
males and females banded in each year, and the 
ratio of males to females. A small number of birds 
ol the earliest or latest arriving species may have 
arrived prior to 15 April or after 15 June in 
