OCCURRENCE AND IDENTIFICATION OF THE LEACH'S STORM-PETREL COMPLEX OFF SOUTHERN CALIFORNIA 
breeding season (November-April). Although 
measurements of Ainleys and Chapman’s 
overlap (Table 1), differences in plumage as- 
pect (such as rump pattern, which is difficult 
to qualify; see Figures 6 and 16), plumage 
wear, and molt timing allow many specimens 
to be identified, particularly when compared 
with series of specimens collected at the 
breeding islands. 
After breeding, Ainley’s Storm-Petrels ap- 
pear to withdraw southward, based on limit- 
ed specimen data. An adult female in worn 
plumage (SDNHM #29925) collected 360 
kilometers southwest of Guadalupe Island on 
21 May, and a juvenile in fresh plumage 
(SDNHM #29924) collected 550 kilometers 
west of Clarion Island, Mexico on 24 May (at 
about 18° S), both suggest a southward post- 
breeding dispersal, as does an adult in wing 
molt (with p4 shed) collected on 1 1 June at 4° 
S, 93° W (CAS #484), a location about 400 
kilometers southwest of the Galapagos Is- 
lands. A specimen completing wing molt 
(with p9-pl0 growing) on 9 October (CAS 
#471), taken about 500 kilometers south- 
southeast of Mexico’s Revillagigedo Islands, 
may have been a northbound migrant. 
Potential identification criteria 
With the newly described complexity in the 
Band-rumped Storm-Petrel complex (Smith 
and Friesen 2007, Smith et al. 2007), criteria 
for identifying some taxa away from the 
breeding islands have been proposed (Robb et 
al. 2008). Members of the Leach’s Storm-Pe- 
trel complex have not previously had such 
treatment in the literature, certainly not from 
the perspective of at-sea identification. For 
the purposes of this article, we assume ob- 
servers are familiar with nominate Leach’s and 
do not cover identification of that taxon rela- 
tive to other species of storm-petrels. The ac- 
counts below summarize our thoughts on po- 
tential identification criteria, based on study 
of specimens and encounters with birds at sea 
off southern California and northwestern 
Mexico, As in most birds, wing molt in storm- 
petrels starts with the innermost primary (pi) 
and progresses sequentially outward to the 
outermost (plO). 
All measurements are given in millimeters 
(Table 1), but we augment more general “field 
guide” measurements, such as length and 
wingspan, with measurements in inches, to 
which most North American birders are more 
accustomed. For consistency, all measure- 
ments were made by Howell. Wingspan meas- 
urements come from 30 fresh specimens of 
Leach’s collected by the late Larry Spear, and 
by a formula derived from these. In nominate 
Leach’s, wing chord averages 33% (32-34%) 
Figure 5. Variation in rump pattern of Leadi's Storm-Petrels collected in summer (early May to early July) off Oregon, showing the 
typical range oUeucorhoa rump patterns, from all-white (score 1) to having a broad dusky median patch (score 3). Relative to 
Townsend's, note the relatively long and deeply forked tails of these birds. SDNHM specimens. Photograph by Steve N. 6. Howell. 
Figure 6. Variation in rump pattern of Chapman's Storm-Petrels collected in summer at colonies off Baja California, showing a typical 
range from some white (score 2.5) to all-dark (score 5). SDNHM specimens. Photograph by Steve N. G. Howell. 
of wingspan, as is true of Fork-tailed Storm- 
Petrel (0. furcata) and Ashy Storm-petrel (0. 
bomocbiva), among other species of Ocean- 
odwma. We have made the assumption that 
this ratio also holds for Chapman’s, 
Townsend’s, and Ainley’s, which allows hypo- 
thetical wingspans to be calculated from wing 
chords. Most conventional measurements 
overlap among all taxa in the Leach’s Storm- 
Petrel complex (Table 1), leading some recent 
authors (e.g., Unitt 2004) to suggest that even 
maintaining nominate leiicorboa and socor- 
rocnsis as distinct subspecies might not be 
tenable. Considered in combination, howev- 
er, distinctions among vocalizations, meas- 
urements, and plumage support the recogni- 
tion of at least four taxa. 
The Leach’s Storm-Petrel complex is well 
known for marked variation in “rump” pat- 
terns (technically, the uppcrtail coverts). 
VOLUME 63 (2009) • NUMBER 4 
543 
