444 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 3, September 2011 
Philippine Islands, (Proudfoot et al. 2007, Fuchs 
et al. 2008) (Table 1). We included published 
sequences from two representatives of New World 
'Otus' ( Megascops ): M. asio and M. hoyi for 
proximal outgroups. We rooted our trees using the 
more distantly related Bubo virginianus. 
Laboratory Procedures. —Samples used for 
DNA extraction were from muscle tissue. DNeasy 
Tissue Kit (Qiagen, Valencia, CA. USA) was used 
to isolate genomic DNA. We sequenced cyto¬ 
chrome b (cyt-b) and ND2 using primer pairs 
LI 4990/HI 5646. L15517/H16404 for cyt-b, and 
L52I6/H5766, L5758/H6313 for ND2 (Sorenson 
et al. 1999). A new primer was designed to 
sequence ND2 for several taxa: L52I7 (5'- 
CCCCATAATCTCAAAATCACATC-3') for O. 
m. nigrorum. O. m. megalotis, and O, longicornis. 
Reaction tubes contained 5.0 pi of 40 mM 
magnesium chloride, 2.0 pi of dNTP mix (2 niM 
for each nucleotide), 5.0 pi of 12 pM of primers, 
and 5.0 pi of Platinum PJx DNA polymerase 
enzyme in a 1:30 dilution (Invitrogen, Carlsbad. 
CA, USA) and 5.0 pi of template DNA. The 550- 
600 bp fragments were PCR-amplified ip 50 pi 
reaction capillary tubes using Rapidcycler© 
(Idaho Technologies. Salt Lake City, UT, USA) 
with the first cycle at 94° C for 15 sec, followed 
by 35 cycles at 0 sec at 94' C denaturation 
temperature, 55-58 c C annealing temperature at 
0 sec, and 50 sec at 70' C extension temperature. 
Amplified PCR-products were purified using 
Wizard PCR Preps Purification System (Promega 
Corporation, Madison. WI. USA). Sequencing 
was done using an Applied Biosystems 3730 
sequencer (Applied Biosystems, Foster City. CA. 
USA). Sequences were aligned and fragments 
assembled using GeneiousPro4.0.3 (Biomatters 
Ltd., Auckland, New Zealand). New sequences 
were submitted to GenBank with accession 
numbers JN131475 to JN131498. 
Phylogenetic Reconstruction.— Maximum Par¬ 
simony (MP) and Maximum Likelihood (ML) 
analyses used Program PAUP* 4.0b 10 (Swofford 
2003). A concatenated data set was generated and 
analyzed. All MP searches used equal weighting, 
heuristic search options with 1.000 replicates, 
tree-bisection-reconnection branch-swapping, and 
random addition of taxa. Clade support for MP 
and ML trees was obtained with bootstrap values 
?oQcf d J r ° m 1,000 re P |icati ons (Felsenstein 
1985). Program MODELTEST (Posada and 
Crandall 1998), using Akaike Information Crite- 
na, suggested General Time Reversible (GTR) 
plus Gamma model as the most appropriate site- 
specific model of evolution. We performed ML 
tree searches using the successive approximations 
method (Huelsenbeck 1998) in PAUP*4.0bl0 to 
obtain the best-fit trees and parameter estimates. 
Support for particular nodes was obtained using 
non-parametric bootstrap (Felsenstein 1985) as 
implemented in PAUP*4bl0 with 1,000 fast- 
addition bootstrap replicates in a likelihood 
framework. 
Bayesian analysis was performed in MRBAYES 
Version 3.1 using Markov Chain Monte Carlo 
(MCMC) tree searches (Huelsenbeck and Ronquist 
2001). Four independent searches were performed, 
each with a cold chain and three heated chains run 
for 400,000 generations with trees sampled ever} 
100 generations. 
Morphometric. Analysis .—Specimens (Table 2) 
from the Delaware Museum of Natural History 
(DMNH) and CMC were examined at the Houston 
Museum of Natural Science. Only adult speci¬ 
mens of known gender were included, as speci¬ 
mens of unknown age or gender could bias the 
results by over- or under-representing mean 
measurements. Twenty-one specimens (11 males 
and 10 females) of megalotis , 18 of everetti (10 
males, 8 females), and seven nigrorum (4 males, 3 
females) were included (Table 2), Culmen (cere 
to tip) and tarsus measurements were taken with 
Vernier calipers, and a ruler was used to measure 
natural wing chord and tail length (all measure¬ 
ments in mm) (Table 2). 
Significance of morphometric variation was 
assessed using ANOVA for differences between 
megalotis and everetti. whereas non-parametric 
Mann-Whitney LMests were used for comparisons 
of everetti and nigrorum , as only small samples 
were available for the latter form. Differences 
were examined separately for males and females 
in light of variation between male and female 
owls (Amadon 1959, Brooks and Arnold 2005). 
RESULTS 
DNA Sequence Analysis. —All samples were 
sequenced from frozen tissue. We obtained 1.033 
and 1.054 bp of ND2 and cyt-b, respectively. We 
truncated the first 15 bp at the 5' end of the 
downloaded sequences because of ambiguous 
alignment in that short segment. No other 
insertion or deletion was observed from that point 
upstream. Alignments were relatively straightfor¬ 
ward and were performed by the GeneiousAlign 
option within Geneious Pro 4.0.3 (Biomatters 
