Page 9 - Jakubas_et_al._2014_Storm_Petrels
P. 9
78 D. Jakubas et al.
We found a negative relationship between variation in body size, especially considering the
body mass and WS at the species and subspecies fact that the subspecies H. p. melitensis appears to
scales indicating that lighter birds were recorded be largely resident (Hashmi & Fliege 1994) and
in the areas characterized by strong wind. As evidence suggesting different wintering areas of
favourable foraging conditions for European populations breeding in the Irish Sea and western
Storm Petrels are defined by high wind speeds Irish coast (Fowler 2002).
leading to dramatic rapid changes in shear (Scott
et al. 2013), small body mass may be an advantage, Despite the fact that our analyses included
increasing manoeuvrability over waves in only individuals captured during the breeding
stronger wind conditions. season, inter-colony differences in body size
should be interpreted with caution as European
At the species scale, the longest wings were Storm Petrels are known from long-distance
recorded in birds breeding in the Mediterranean short-term movements. Overnight movements
(ME). This may be associated with diet and forag- from Shetland to Orkney or two- or three-day
ing technique differences. European Storm Petrels movements to Norway are not unusual (Fowler
in the Mediterranean exploit pelagic fish which 2002). Moreover, the high responsiveness of birds
are taken by diving, while individuals from the to tape luring using the “purr” call may result in
Atlantic populations feed mainly on mesozoo- catching birds from different colonies generating
plankton and ichthyoplankton by hovering above heterogeneity in the sample (Fowler 2002).
the water (D’Elbee & Hemery 1998, Albores‐
Barajas et al. 2011). In conclusion, our results indicate morpholog-
ical variation across the studied populations of
The small size of birds breeding in the Canaries European Storm Petrels, with a longitudinal
(compared to birds from other Atlantic and increase in wing length from west to east. More-
Mediterranean colonies; Castro et al. 2013, this over, the wing length of birds increased from south
study) may be an adaptation to the specific envi- to north with decreasing SST and AT, but only
ronment (strong winds, high temperature). Wing when the Atlantic subspecies was considered.
size is one of the most important traits differenti-
ating Southern and Northern Hemisphere Sexual size dimorphism
Hydrobatidae. Shorter wings in the Southern
Hemisphere subfamily Oceanitinae compared to Our results on SSD in the European Storm Petrel
the Northern Hemisphere Hydrobatinae are support the findings of previous works (Table 4)
explained as an adaptation to strong winds indicating female-biased SSD in tail and wing
(Carboneras 1992). It cannot be excluded that length (Fig. 6). It is believed that sexual selection
southernmost populations should be distin- appears to be the most important evolutionary
guished as a subspecies. force influencing SSD in seabirds. Selection
appears to favour the small male size in those
The forces driving patterns of body size varia- species that exhibit aerial display (Serrano-
tion remain speculative. Environmental factors on Meneses & Székely 2006). This might be true for
the breeding grounds seems to be important. the European Storm Petrel because males perform
However, it is also conceivable that wintering “display flights” at the beginning of the breeding
conditions may explain some of the observed period (Davies 1958). However, considering that
Table 4. Summary of studies investigating sex differences in the European Storm Petrel. Variables: BD — bill depth, BM — body
mass, HBL — head-bill length, RBL — rump band length, TL — tail length, TRS — tarsus length, WL — wing length. Location
abbreviations: see Table 1.
Location Reference Variables differing Variables in discriminant
(sample size) between sexes function (function correctness in %)
H. p. pelagicus
FAR this study (156) WL, TL, HBL, BM WL+HBL (75)
SHE1 Fowler et al. 1986 (56) lack of sex differences No function
WAL1 James 1983 (46) WL, TL TRS+WL (80), TRS+TL (76),
WL+TL (85),TL+TRS+WL (85)
POR Medeiros et al. 2012 (92) WL, BD, BM BD/BL+WL/TRS (63)
CAN Castro et al. 2013 (79) WL, TL 9 measurements (71),
WL+TL (no data)
H. p. melitensis Albores-Barajas et al. 2010 (46) WL, RBL
AEG WL+RBL (81)
