Miranda el al. • TAXONOMY OF PHILIPPINE SCOPS OWLS 
443 
cycles. The contraction and expansion of biomes 
during these cycles led to isolation, or aided in 
colonization as adjacent islands merged. The 
montane region expanded down along mountain 
slopes, while lowland dipterocarp forest retracted 
as average temperature decreased over time. This 
model predicts the lowland dipterocarp forest, 
while retracting at the upper elevalionaJ boundary 
of the mountains, would have expanded as more 
lowland areas were exposed with descending sea 
level. Concomitantly, adjacent islands would be 
closer together, if not coalesced. This process w-as 
hypothesized to have acted as 'species pumps' in 
topographically diverse islands such as the 
Philippines, that accelerated speciation compared 
to less topographically diverse regions (Steppan et 
al. 2003). The biome elevationa! shift hypothesis 
predicts that species in montane regions of 
adjacent islands are sister taxa, and lowland 
species between adjacent islands are sister taxa. 
There are two recognized strigid genera within 
the Philippines, O/us and Mimizuku. There arc 
seven Philippine species of Otus, five of which are 
small island endemics: Luzon Scops Owl (0. 
longicomis), Mindoro Scops Owl (O. mindoren- 
sis), Mindanao Scops Owl ( 0. minis). Philippine 
Scops Owl (O. megalotis), and Palawan Scops 
Owl (O. fuliginosus). The remaining two O/us are 
not restricted to the Philippines: O. mantmanensis 
on small islands off Borneo and the Sulu 
Archipelago, and O. elegans in Batanes Islands 
north of Luzon (Kennedy et al. 2000). 
Early classification based on morphology 
placed Mimizuku gurneyi as a relict of an ancient 
lineage, and clustered nigrorum and longicomis 
as races ol the widespread O. spilocephalus 
(Burton 1973). Marshall's (1978) classification 
led to species status for O. longicomis, O. 
mindorensis, and O. minis. Research based on 
vocalizations led to more major taxonomic 
rearrangements with an increase in the number 
of Asian scops owl species (Roberts and King 
1986. Marshall and King 1988. Becking I994~ 
Lambert and Rasmussen 1998). Earlier systemat¬ 
ic 4 ' w ork on Philippine scops owls using different 
mtDNA genes suggested divergence of the 
lowland/highland eludes and noted the relatively 
high genetic distances among endemic Philippine 
scops owls (Miranda et al. 1998). Merger of 
Mimizuku with Otus was also suggested in that 
paper and elsewhere (Mindell et al. 1997; Miranda 
et al. 1997, 1998). Kdnig and Weick (2008) cited 
this previous phylogenetic evidence and men¬ 
tioned the differences in vocalizations among the 
O. megalotis subspecies, but did not designate the 
three O. megalotis subspecies as distinct species. 
Wink ct al. (2008. 2009) also supported the 
invalidity of the genus Mimizuku. 
An assessment of genetic structure among the 
three megalotis subspecies based on larger 
sampling should clarify their biogeography and 
taxonomy. For example, the three island blocks of 
Greater Luzon, Greater Mindanao, and Greater 
Panay-Negros were separated from each other by 
narrow straits of ~25 km in width (Steppan et al. 
2003) . There should be little or no genetic 
differentiation among subspecies, assuming these 
straits did not pose a barrier to dispersal. 
However, if the distinct phenotypic variation 
observed among the subspecies is congruent with 
the mtDNA substructure, one can invoke the 
notion of reproductive isolation due to a geo¬ 
graphic barrier (i.e., the open sea). 
We used mitochondrial DNA sequences of two 
genes. NADH dehydrogenase 2 (ND2) and 
cytochrome b (cyt -b). to address two questions. 
(I) Does the presence of multiple scops owl 
species on islands reflect speciation events or do 
species assemblages result from multiple coloni¬ 
zation events? (2) Do the patterns of mtDNA 
divergences as shown by gene trees and morpho¬ 
logical divergences reflect the current taxonomy 
for the Philippine Scops Owl (O. megalotis )? 
Resolution of taxonomy to reflect evolutionary 
divergences has far-reaching implications for 
studies of conservation and biodiversity (Zink 
2004) . especially in a biodiversity and conserva¬ 
tion hotspot such as the Philippines (Mittermeier 
et al. 1999, Myers et a). 2000, Sodhi et al. 2004, 
Peterson 2006). 
METHODS 
Taxonomic Sampling .—Tissue samples were 
obtained during the 1991-1993 Philippine Biodi¬ 
versity Inventory conducted by the Cincinnati 
Museum Center (CMC) and the Philippine 
National Museum. We used 12 individuals 
representing four species of Philippine scops 
owls: M. gurneyi, O. longicomis, O. minus, and 
O. megalotis, which is represented by three 
subspecies: O. m. megalotis (lowland Luzon), 
°. m. everetti (lowland Mindanao), and O. m. 
nigrorum (Panay). 
We used sequences of several Old World Otus 
taxa from neighboring islands to increase taxon 
sampling and orient colonization patterns of the 
