Page 4 - Jakubas_et_al._2014_Storm_Petrels

Body size variation in the European Storm Petrel
patch indicating breeding. We ringed and meas-
ured all captured birds. We used a ruler for wing
and tail length (with 1 mm accuracy) and dial
calipers for all head measurements: head-bill
length (HBL), bill length (BL; along the dorsal
mid-line from the edge of the feathers at the base
of the culmen to the most distant part of the
hook), bill depth 1 (BD1; mandibles before the
opening of narines), bill depth 2 (BD2; mandibles
+ narines at the opening), bill depth 3 (BD3;
mandibles + narines at the base of the culmen),
bill width (BW; mandibles + narines at the base of
the bill)(Fig. 1), tarsus length (from distal end of Fig. 1. Morphometric measurements of the head: head-bill
tarsometatarsus to back of heel with leg bent) and
rump band length (measured following Albores-
Barajas et al. 2010). We weighed the birds with an
electronic scale (Ohaus, with 0.1 g accuracy). All
measurements were taken by the same person
(DJ). From each bird, we collected a feather or
blood sample (preserved in 96% ethanol) for
molecular sexing.
Laboratory analyses
We extracted DNA from the feathers and blood
(following evaporation of the alcohol) using the
Blood Mini kit (A&A Biotechnology, Gdynia,
Poland). We performed molecular sexing with the
primer pair F2550 and R2718 according to the pro-
tocol described by Griffiths et al. (1998), using a
50 °C annealing temperature for the PCR reaction.
The sex differences in the PCR products were
clearly visible in UV light when the fragments
were separated on a 2% agarose gel and stained
with Midori Green.
Data analyses
We performed all statistical calculations using
STATISTICA 8.0 (StatSoft, Inc., Tulsa, Oklahoma,
USA) and IBM SPSS 21.0 (IBM SPSS, Chicago,
Illinois, USA).
Geographical variation in body size. To inves-
tigate geographical variation in body size, we per-
formed analyses on unsexed individuals from 16
colonies using wing length and 13 colonies using
body mass, based on the published data (Table 1,
Fig. 2). For colonies for which data for males and
females were presented, we calculated wing
length and body mass for both sexes combined
using weighed mean with sample size as a weight.
We included in analyses only birds captured dur-
ing the breeding season.
Relationship between body size and environ-
mental variables. To investigate whether environ-
mental conditions affect the body size, we related
73
HBL
BL
BD3
BD1
BD2
BW
length (HBL), bill length (BL), bill depth 1 (BD1; before the
tube), bill depth 2 (BD2; at the tube), bill depth 3 (BD3; at the
base of the bill) and bill width (BW).
data on wing length and body mass with sea sur-
face temperature (SST), air temperature (AT) and
wind speed (WS) using Pearson correlation
coefficient. Data on sea surface temperature, with
a spatial resolution of 0.58 (latitude/longitude)
from the topmost possible water column (up to
5 m), were obtained from the Reyn_Smith OIv2
(Reynolds et al. 2002) database via the IRI/LDEO
Climate Data Library Reyn_SmithOIv2 monthly
sst: (http://iridl.ldeo.columbia.edu/SOURCES/
.NOAA/.NCEP/.EMC/.CMB/.GLOBAL/.Reyn_
SmithOIv2/.monthly/.sst). Considering possibly
long-distance foraging flights of the European
Storm Petrel, we selected the 1.5° box closest to
each colony (i.e. 167 km). Mean monthly SST val-
ues for June–September during the 10 years of
available data (1998–2007) were calculated for each
study site. AT and WS data were obtained from a
database on the website http://www. TuTiempo.net.
The meteorological stations closest to each breed-
ing site (up to 110 km) were chosen. We took data
from the same years as for SST (1998–2007). We
chose June–September because this period corre-
sponds to the breeding period of the European
Storm Petrel (Cramp 1998). We calculated mean
SST, AT and wind speed WS for periods including
breeding season (i.e., Jun–Sep, Jun–Aug, Jul–Aug
and Jul–Sep) for each site. We used various peri-
ods due to variation in the timing of breeding
(Cramp 1998). SST, AT and WS values for the stud-
ied periods differed significantly among the stud-
ied sites (Kruskal-Wallis test, all p < 0.001).
Based on the subspecies division and geo-
graphical localisation of colonies, we divided
all colonies studied into three subgroups