Biogeography of Luzon Island, Philippines

2014·7.6 MB·English

by  Benjamin Vallejo

B., Jr.: Biogeography of Luzon Island, Philippines
Vallejo,
Biogeography of Luzon Island, Philippines
Jr.
Benjamin Vallejo,
Institute of Environmental Science & Meteorology, College of Science, National Science Complex,
University of the Philippines, Diliman, Quezon City 1101, The Philippines; [email protected]
Abstract: Luzon Island is the largest and oldest of the oceanic islands of the Philippine archipelago. Previous bio¬
geographic research has determined centres of endemism within the island. While the Pleistocene Aggregated
Island Complex theory defines the island as one biogeographic region, recent phylogenetic studies suggest that
the island may be composed by distinct centres of endemism that correlate with tectonic features. Deep genetic
divergence between northern Luzon and southern Luzon clades support this hypothesis. Past researches have
downplayed the importance of vicariance in the modern biogeography of the island. However post vicariance dis¬
persal may have obscured historical and area relationships that are noted in coalesced islands. Nonetheless, the
molecular phylogenetic signature of pre Pleistocene vicariant events are there and further phylogenetic studies may
clarify these relationships.
Keywords: Pleistocene Aggregated Island Complex (PAIC), vicariance, dispersal, Luzon, Philippines, biogeography.
Introduction
The latitudinal extent of Luzon and the com¬
This essay builds upon the biogeographical plexity of its topography and the presence of inter-
hypotheses on the Philippines position in Wallacea montane valleys of varying sizes ensure a diversity
in Volume 1 of this book series. In my “The Phil¬ of climate not observed anywhere else in the Philip¬
ippines in Wallacea” I propose looking at the bio¬ pine archipelago (Dickerson et al. 1928). The alti¬
geography of the Philippines beyond the Pleisto¬ tude of the landforms and their exposure to the pre¬
cene Aggregation Island Coalescence (PAIC) theory vailing monsoon and trade winds generally defines
(Vallejo 2011). Here I focus on Luzon, the oldest the climate of the island.
of the oceanic Philippine islands. Luzon is the larg¬
est (104 688 km2) island of the Philippine archi¬
pelago. The roughly rectangular island is orientated Tectonics of Luzon Island
with its longest axis north to south from 18°32’N to
12°31’N (Fig. 1). The shape of the island gives rise The complex topography has tectonic origins.
to its name. The island resembles the traditional The island is hypothesized as a product of the accre¬
rice pestle or “lusong” of the Austronesian people. tion of at least four paleoislands (Hall 1996,1998).
The southern and south-eastern portion of The paleoislands correspond to the five montane
the island is composed of a series of peninsulas regions of Luzon (Devan-Song et al. 2012). The
trending southeast for about 150 km (Fig. 2). The island is part of the Philippine Mobile Belt (PMB)
northern end of the island are composed of sev¬ that defines the seismically active part of the ar¬
eral mountain blocs, most notable of which is the chipelago (Gervasio 1967). The island being part
Cordillera Central whose mountains attain an al¬ of the PMB is bounded by two subduction zones of
titude of more than 2000 meters. A large central opposite polarities. The eastern side is bounded by
plain defines the central portion of the island. This the west dipping Philippine Trench and the western
is bordered on the west by the Zambales mountain side is by the east dipping Manila Trench. New geo¬
ranges. The eastern side of the island is defined by physical evidence suggests that the Benham Rise
the Sierra Madre mountain range that trends from oceanic plateau east of Luzon has a thicker crust
north to south. At its northern end, a large valley, than what can be expected for oceanic crust (Lag-
the Cagayan Valley is situated between the Sierra may et al. 2009). This feature began colliding with
Madre to the east and the Cordillera Central to the northern Luzon starting in the Miocene thereby pro¬
west. foundly affecting Luzon’s tectonic evolution.
D. (ed.) 2014: Biodiversity, Biogeography and Nature Conservation in Wallacea and New Guinea, volume II
Telnov
116 118 120 122 124 126 128
a
CM
CO
to
CM
CO
(0
Figure 1. Luzon Island in the Philippine archipelago.
The Philippine archipelago and most especial¬ that absorb plate convergence such as the Philip¬
ly Luzon Island is defined by the left lateral strike pine Fault, Digdig Fault and the Northern Cordillera
slip Philippine fault system which longitudinally Fault. From a biogeographic standpoint, blocks may
cuts Luzon Island (Yumul et al. 2008) at Quezon correlate with patterns of endemism and distribu¬
and Nueva Ecija provinces. It originates from Davao tion of Philippine biota. This is the hypothesis first
Gulf in Mindanao, bisects Mindanao’s Agusan ba¬ proposed by Roy Dickerson and Elmer D Merrill in
sin, the passes through Leyte and Samar islands 1928 (Dickerson et al. 1928).
before terminating in Luzon’s north-western coast.
In Luzon the fault becomes braided (Fig. 3).
Thus with new geophysical data it is now pos¬ Dickerson’s hypothesis
sible to delineate the tectonic blocks that com¬
prise Luzon. These blocks are mobile, elastic and Roy Dickerson and associates of the Philippine
are related to the major fault features in Luzon Bureau of Science in “Distribution of Life in the
48
B., Jr.: Biogeography of Luzon Island, Philippines
Vallejo,
120 121 122 123 124
a>
CD
N
A
% its '
jr _ %
tW - North em Sierra
r
' m.
» a OO
' Tf
-v 4bJ£-m V ,
is
v
kv
Cordillera
Central Cagayan J.,1 Ty i
Mr-f-
Valley
4%
V' ^jr:
V y -kjfa Jift .4 r -
(£> CD
»>'v
Southern Sierra
Ifi
.L v Madre
V
.A f £
Zambales % -i—
)'(cid:9632)
• •,• -‘ •5>\_y\ w v _
Mountains v
-- n
V 3S
% \ v\ (cid:9632) (cid:9632)
%
PI *Ph
(cid:9632)v-
Bicol peninsula
n -- ?-
Southern Luzon volcanic belt (cid:9632)IML-
(cid:9632)
' «
Si. fli
TO
0 2040 80 120 160
Kilometers
^20* 121 122 123 124
Figure 2. Major physical geographic features of Luzon Island.
Philippines” (1928) proposed that the geological the Sulawesi - Eastern Mindanao route and the Tai¬
features of the Philippine islands could partly ex¬ wan - Batanes - Northern Luzon route.
plain the reason for the localization and endemism Luzon’s biota provides support and difficulties
of the Philippine biota. Dickerson’s hypothesis is for the dispersal hypothesis. Dickerson needed to
based on the fifth axiom of biogeography (Wallace explain the presence of the northern Luzon upland
1880) its emphasis on a general knowledge of geo¬ flora and fauna for which he had difficulties. While
logical history is necessary for understanding the the phenomenon of continental drift had been hy¬
evolution of the Philippine biota. pothesized by Alfred Wegener in 1924 (Wegener
Dickerson interpreted the distribution of floral 1966), the process by which this would happen had
and faunal elements in the Philippines using the not been proposed. Without the explanatory power
dispersalist paradigm. He proposed four coloniza¬ of plate tectonic theory, Dickerson had only the
tion routes by which dispersal happened. These are land bridge paradigm and the Wallace theory cor¬
the Palawan-Mindoro route, the Borneo-Sulu route, relating emergence of islands with geological and
49
Telnov D. (ed.) 2014: Biodiversity, Biogeography and Nature Conservation in Wallacea and New Guinea, volume II
Figure 3. Major tectonic features of Luzon Island (modified from the Philippine Institute of Volcanology
and Seismology, 2013).
consequently phylogenetic age of the biota. 1996, 1998) now known as the Pleistocene Aggre¬
The lack of a comprehensive theory to relate gated Island Complex (PAIC) theory which forms the
the presence of Philippine floristic provinces (sen- basis of the current idea explaining the presence
su Merrill) with tectonic features did not prevent of distinct island biotic regions in the Philippines
Dickerson to hypothesize on the physiognomy of (Heaney 1986, 1998, 1999, 2000).
geographical features of the Philippine islands in
geologic time. This radical speculation such as the
possibility that. Luzon and Mindanao were once sep¬ Biotic alliances in the Philippines and the distinc¬
arate archipelagic systems which he inferred from tiveness of Luzon (Fig. 4)
the distribution of the modern biota within each is¬
land, presaged the idea of island coalescence (Hall Elmer D. Merrill (Merill 1923) observed that
50
Vallejo, B., Jr.: Biogeography of Luzon Island, Philippines
Batanes-
Babuyan
Northern Sierra
Cordillera biogeographic boundary
Mad re
Zambales
Mountains Sierra Madre biogeographic boundary
Southern Sierra
* Mad re
Southern Luzon B Bicol Peninsula
Volcanic Belt
Aiimonao biogeographic
boundary
Figure 4. Luzon biotic regions (after Merrill 1928, modified with new distribution
and tectonics information).
when the Philippines is taken as a whole, its flo¬ the region. Merrill hypothesized a connection with
ral affinity with other regions in Southeast Asia is Taiwan but came to conclude that out of the 1100
definitely Indo-Malayan with a high proportion of species found in Taiwan and could occur in the Cor¬
endemic species. While floral regions do exist, its dillera given that the climate conditions are simi¬
distinctiveness is not sharp with the exception of lar less than 265 occur on Luzon and in Taiwan.
the Luzon Cordillera flora. It is on Luzon in the Cor¬ Thus puts a question on the land bridge connection
dillera where a flora and fauna with obvious Hima¬ of Luzon with Taiwan, which tectonic studies later
layan affinities are found. On other islands, it is on would conclude as unlikely. Merrill in conformity
high mountains where a distinct Sulawesian and with the land bridge paradigm suggests that only
Australian affinity can be found. The Cordillera flora the Cordilleras were once connected with Taiwan
with the exception of a few species is localized in and the rest of mainland Asia.
51
D. (ed.) 2014: Biodiversity, Biogeography and Nature Conservation in Wallacea and New Guinea, volume II
Telnov
A dispersal process cannot be ruled out for a geographers to examine whether biological disjunc¬
colonization route from Taiwan. Recent studies sug¬ tions observed in Luzon and the four microendemic
gest that at least for one species of skink, there is regions are congruent with tectonic features. In New
evidence of dispersal from Taiwan (Esselstyn et al. Caledonia, Michael Heads in describes biodiversity
2010). with respect to the West Caledonian fault system by
using molecular phylogenetic, tectonic and panbio-
geographic methods to describe biotic disjunction
Microendemism and mountain regions in Luzon with the fault system (Heads 2008). New Caledonia
is an ancient remnant of Gondwana and is rich in
Luzon has a high degree of microendemism archaic taxa. Heads hypothesizes that lateral strike
across taxa which has been extensively studied for slip fault systems may reveal more than biotic dis¬
the herpetofauna (Siler et al. 2010, 2011, Devan- junctions that any other geological feature. In other
Song et al. 2012) and mammals (Heaney et al. island systems in the Pacific basin, such disjunc¬
1998, Rickart et al. 2011). This microendemism tions have been also noted (Heads 1990, 2001,
is found mainly in the montane regions. For the 2008) in studies in cladogenesis.
herpetofauna, four to five regions with significant Two major schools of thought in biogeography
microendemism can be recognized. For the mam¬ examine the problem of disjunction and cladogene¬
mals, four regions of microendemism are recog¬ sis with respect to determining areas of endemism.
nized. These regions are: In cladistic biogeography (Platnick et al. 1978, Nel¬
son et al. 1981) especially in the parsimony analy¬
1) The Zambales Mountains, sis of endemism (PAE) approach, it is possible to
2) Cordillera Central Mountains, come up with correct inferences to historical rela¬
3) Northern Sierra Madre Mountains, and tionships among areas when by modelling a par¬
4) The Southern Sierra Madre Mountains and ticular combination of vicariance and non-response
the Southern Luzon and Bicol volcanic belt. to vicariance events (Brooks et al. 2003). In these
cases, vicariance is wholly responsible for species
Mt Pulag which lies in the Cordillera Central is distribution and species in each clade considered
a convergence zone of endemic flora with Gondwa- have a specific pattern of non-response to vicari¬
nan and Eurasian affinities (Buot Jr. et al. 1999). ance. These non-responses to vicariance generates
Of all the four regions of microendemism, the the correct area relationships. Another model that
Zambales and Cordillera mountains have received generates correct historical and area inferences is
more research interest in the last 20 years. Zam¬ when species distributions result from a particular
bales is relatively isolated from the rest of Luzon’s combination of extinction events especially for wide
mountains by the central Luzon plain (Brown et al. ranging species. Extinction events may split ranges
1996). This mountain range is essentially an ophi- analogous to geological vicariance events. Brooks
olite complex (Yumul et al. 2003, 2008). Recent et al describe three cases when PAE methods fail.
research indicates that the Zambales mountains In the third case, post vicariance dispersal may
as well as the nearby Cordilleras have substantial obscure historical and area relationships. I hypoth¬
patterns of endemism in small mammals (Balete esise that the third case is likely for the Philippine
et al. 2009, Rickart et al. 2011). The Zambales archipelago.
mountains have been also identified as an endemic
plant region (Merill 1923; Dickerson et al. 1928).
For example the umbrella plant genus Schefflera Cladogenesis in Luzon
J.R. Forst, G. Forst, 1775 has many endemics in
the Zambales and Cordillera Central ranges (Frodin The general distribution of endemics in Luzon
1986) but none outside these areas. can be roughly characterized by a northern group
and a southern group. The northern group clades
consists of Zambales and Cordillera representa¬
Can Luzon be a case by which biological disjunc¬ tives and the southern group clades which includes
tions can be examined as related to tectonic fea¬ species from the Bicol peninsula, have more affinity
tures? to representatives from the Visayas and Mindanao
islands. In the fantail birds Rhipidura Vigors, Hors-
The presence of one major fault system in the field, 1827 (Sanchez-Gonzalez et al. 2011), the
Philippine archipelago gives an opportunity for bio¬ older northern and southern clades have a deep
52
B., Jr.: Biogeography of Luzon Island, Philippines
Vallejo,
divergence suggesting an earlier colonization from However both species have evolved in higher mon¬
mainland Asia (Fig. 5). tane forests and have unlikely to have dispersed
A similar hypothesis can be proposed for the through the lowland forests of the southern Sierra
murines of Luzon (Jansa et al. 2006). The “old en¬ Madre. Nonetheless the distribution of endemic va¬
demics” which include Crateromys Thomas, 1895, ranids of Luzon and even of Panay in the Visayas
Phloeomys Waterhouse, 1839 cloud rats, Batomys appear to be delimited by their tectonic features
Thomas, 1895 and Carpomys Thomas, 1895 show with V. mabitang localized in the Panay Cordilleras
a deep divergence within the group, with the north¬ (Zamoras et al. 2008).
ern Luzon species more basal than the southern This distributional pattern is reflected in Phil¬
representatives. The Luzon endemic Phloeomys ippine Rafflesia R. Brown (Barcelona et al. 2009)
Waterhouse cloud rat genus is represented by a where R. maniliana Teschem is localized to the
northern Luzon and a southern Luzon species and Southern Luzon and Bicol Peninsula areas although
is more basal than Crateromys Thomas. Crateromys some individuals were collected in Cagayan, north¬
shows similar distributions in its representatives al¬ ern Luzon. Luzon endemics of Rafflesia R. Brown
though this genus has representatives in Mindoro, have been collected in the Bicol peninsula and in
Panay and Dinagat Islands. The “old endemics” the northern Sierra Madre ranges in Cagayan (Bar¬
colonized evolved at least 22 My, when northern celona et al. 2006, 2009; Madulid et al. 2006).
Luzon’s older tectonic features emerged from the Panay also has its own endemic Rafflesia found in
ocean. There is evidence that the “old endemics” the Panay Cordilleras (Barcelona et al. 2002).
have affinities to Australasian murine clades (Step-
pan et al. 2003) but are now extinct in the rest of
Asia (Jansa et al. 2006). More recent phylogenetic Revisiting PAIC theory
data from seven newly identified species of Apo-
mys Mearns, 1905 reinforce the hypothesis that The Pleistocene Aggregated Island Complex
northern Luzon has a distinct and endemic biota (PAIC) theory (Brown etal. 2002) is the paradigm for
(Heaney et al. 2011) and that central Luzon is a explaining the origin and dimensions of biodiversity
biogeographic break. in the Philippines (Heaney 1986; Heaney 1998).
In the biogeography of the Philippine varanids, The theory states that isolation and reconnection
the northern and southern Luzon disjunction can of islands that composed the palaeo “Greater Is¬
be noted (Welton et al. 2010). The northern Luzon lands” of the Philippine archipelago provided the
endemic Varan us bitatawa Welton, Siler, Bennett, vicariant mechanism for speciation. Also differen¬
Diesmos, Duya, Dugay, Rico, Van Weerd, Brown, tial dispersal abilities of the isolated species on the
2010 is morphologically distinct from the southern palaeoislands increased genetic isolation leading
Luzon and Bicol peninsular endemic V. olivaceus to speciation.
Hallowell, 1856. Both species are frugivorous. In PAIC theory can adequately explain speciation
total there are three frugivorous species of Vara- of Philippine taxa within the last 5 million years.
nus Merrem, 1820 in the world which includes V. Thus it is not surprising that the theory can explain
mabitangGaulke, Curio, 2001 from northern Panay phylogenetic relationships in taxa in the younger
(Gaulke et al. 2001). Philippine islands. PAIC theory predicts the follow¬
These frugivorous varanids are associated ing (Esselstyn et al. 2010):
with old growth rainforests which exists along the
eastern Philippines bioregion. V. bitatawa and V. 1) Populations in a given island should be
olivaceus are sister species, albeit with deep ge¬ more related from populations in other islands;
netic divergence and their ranges are separated by 2) Populations within an island should be ge¬
the lower elevations of the southern Sierra Madre netically more related than similar populations in
ranges (Fig. 6). The southern Sierra Madre is the other islands;
eastern Luzon terminus of the active Philippine 3) Monophyletic lineages should be found
Fault (Yumul et al. 2008). While the southern Sierra within one island and not across several islands.
Madre ranges which are very near to Manila have
been historically deforested, they were unlikely to Luzon being the oldest of the Philippine oce¬
have been prior to human settlement. Thus it would anic islands at 35 My and dating back to the Eo¬
have been possible for limited dispersal of the cene presents complications to the PAIC theory. In
northern species to southern Luzon. Similarly the contrast, predictions of PAIC is more easily verified
southern species could have dispersed to Luzon. for the younger PAIC islands like Negros and Panay
D. (ed.) 2014: Biodiversity, Biogeography and Nature Conservation in Wallacea and New Guinea, volume II
Telnov
C
5
0
(cid:9633) Y
^ D
R. cyaniceps {Cordillera, N, Sierra Mad re)
/ R . cyaniceps {S. Sierra Madre. Bicol)
<y
N R. saulii (Tabias)
<
atboventris (Negros-Panay) %
R
/?. samarensis (Leyte, Samar)
i
superciliaris (Mindanao)
j»jF ^ i
m *
-R. huichinsii(N. Mindanao monfane forests)
^ c>
/?. cinnamonamea (S. Mindanao monfane foresfsj
H
0
Figure 5. Phylogeny of endemic Philippine Rhipidura Vigors, Horsfield, 1827 (modified from Sanchez
and Moyle, 2011).
(8-5 My) (Diesmos et al. 2002) in studies of phy- al. 2009). Younger species originated from a more
logenies of reptiles and amphibians (Gaulke et al. recent colonization from Sundaland in the last 5
2007; Siler etal. 2007). My and they colonized Palawan and the oceanic
In Luzon, while there are taxa whose endemic islands of the central Visayas, where C. cebuensis
status can be explained by PAIC, other factors that (Steere, 1890) is an example. Copsych us can be di¬
explain endemism need to be considered. One is vided into two ecological groups with the more vag-
the habitat preference of a taxon. In Copsych us Wa- ile and coastal magpie-robins and the more inland
gler, 1827 Magpie Robins and Shamas, the older rainforest dwelling shamas. The diversification of
species the Luzon Shama C. luzoniensis (Kittlitz, the latter in the Philippines can be easily explained
1832) colonized Luzon and dispersed to Panay-Ne- by PAIC theory. In contrast the Philippine magpie-
gros PAIC and persisted there since in those island, robins even in historically isolated Sibuyan Island
there were more habitats that suited it (Sheldon et have their nearest relations from species in ocean-
B., Jr.: Biogeography of Luzon Island, Philippines
Vallejo,
p
Figure 6. Distribution of Luzon’s endemic varanids (after Siler et al. 2010).
ic Seychelles. Their affinities to presumed African islands of Luzon, Mindanao and Negros-Panay.
ancestors imply a very early colonization through However for taxa whose radiation occurred
the Gondwana landmasses of India. This also is during the Pleistocence, the predictions of PAIC
similar to the inferred phylogenetic history of the theory are well verified (Esselstyn et al. 2009) es¬
Philippine Eagle and its nearest extant related spe¬ pecially the isolation by distance model. However
cies, the Africal Bateleur eagle (Lerner et al. 2005; analysis of molecular variation (AMOVA) in these
Vallejo Jr 2011). However based on the molecular taxa shows the dominant role of intra-island iso¬
phylogenetic data, Copsychus Wagler diversifica¬ lation and breaks to gene flow. In fact this is the
tion in the Philippines can be dated to the early dominant proportion that accounts for genetic
Pleistocene and Pliocene at the latest and possibly variation in the largest and oldest oceanic islands
even earlier in the Miocene (Sheldon et al. 2009). of Luzon and Mindanao (Esselstyn et al. 2009). A
Despite island coalescence as a result of sea level possible factor that generated this variation is sym-
regressions, none of the Luzon and Negros- Panay patric speciation (Esselstyn et al. 2009) which may
PAIC species were able to colonize Mindanao. explain diversification in a coalesced island like Su¬
Similar patterns of diversification can be noted lawesi (Esselstyn et al. 2009).
in Philippine Rhipidura Vigors, Horsfield, 1827 (Fig.
5). Deep genetic divergences between and within
PAIC islands species and populations indicate a pre
Pleistocene colonization of the older PAIC oceanic
D. (ed.) 2014: Biodiversity, Biogeography and Nature Conservation in Wallacea and New Guinea, volume II
Telnov
Distribution anomalies in northern Luzon, other Integrating the paradigms, the Philippine islands
factors for endemism as “mini-Sulawesis”?
The current reconstructions of phylogenetic Recent studies on the distribution and phylog-
histories of amphibian and reptile taxa from north¬ eny of Luzon’s endemic flora and fauna suggest
ern Luzon show the affinities between the Zam- that the Pleistocene Aggregated Island Complex
bales, Cordillera and Northern Sierra Madre moun¬ (PAIC) theory to explain the evolution of biodiver¬
tains (Diesmos et al. 2004). One species of frog sity in the Philippines is simplistic and may apply
Platymantis pygmaeus Alcala, Brown, Diesmos, only for the Plio-Pleistocene epochs. However since
1998 from is found in northern Luzon and in the much of the species radiation occurred during this
oceanic island of Sibuyan in the Sibuyan Sea (Dies¬ time period, especially for the small mammals, this
mos et al. 2004). Sibuyan island was never con¬ pattern remains most observable. While phyloge¬
nected with any of the Pleistocene palaeoislands netic reconstructions of species diversification sup¬
(Heaney 1998). Despite intensive sampling in simi¬ port the PAIC theory, analysis of molecular genetic
lar habitats in southern Luzon and Bicol peninsula, variation suggests intra-island isolation possibly by
the species was never recorded. Since Platymantis sympatry.
Gunther, 1858 is rainforest associated does not The dynamic geological history of Southeast
disperse over the sea, this distribution remains Asia and most especially the Philippine archipelago
enigmatic. provides many opportunities for colonization and
For the northern Luzon murines, it is possible allopatric diversification. Allopatric diversification
that during the Pleistocene the extant Cordillera is likely determined by the physical geography of
species or their close relatives were found in lower the islands themselves. Luzon has the greatest es¬
elevations. Archaeological excavations in Callao timate of nucleotide diversity in shrews and much
Cave, Penablanca in Cagayan Province (85 m ele¬ of this occurred in the Holocene, as diversity has
vation) (17°42’11.74”N, 121°49’25.5”E) revealed more correlation with the age and size of the mod¬
the presence of Apomys Mearns and Batomys ern island than that of the paleaoisland (Esselstyn
Thomas fossils. Apomys is found from Luzon, Min¬ et al. 2009). This also implies the importance of
doro, Negros, Panay to Mindanao having dispersed orography and the likely influence of Holocene cli¬
to these areas during the Pleistocene sea level mate change which caused changes in vegetation
regressions (Steppan et al. 2003). The phyloge- and land cover. This may be a factor in the coloniza¬
netically older Batomys is found only in Luzon and tion of the Philippine islands by the bulbuls, where
Mindanao (Jansa et al. 2006) but now are found in certain clades effectively colonized the oceanic is¬
higher elevations. lands and others only the continental islands (Oli-
It is thus possible that the Pleistocene climate veros et al. 2010).
favoured the dispersal of the mountain murines of Another factor that makes it more difficult to
the Philippines across the Luzon central plains. If ascertain the direction of colonization which varies
this is possible then other taxa with similar habitat between taxa is the relative closeness of the Phil¬
requirements may have done so. However it is less ippine islands to each other (Jones et al. 2008).
likely that a close canopy lowland tropical rainforest Colonization came from various directions and
existed in the Luzon Pleistocene. It is more likely since the archipelago is close to the Sundaland is¬
that lowland rainforests contracted in the drier land of Borneo, much of the biotic affinity is broadly
and cooler climate with some areas serving as re- recognizable as Asian although with a significant
fugia and stepping stones to dispersal (Schneider Australo-Papuan component. For mammals, the
et al. 1999). This may also have been the factor Australo-Papuan component is demonstrated by
for vicariant speciation. Also it is possible that the the Chrotomys Thomas radiation (which includes
Pleistocene taxa were not as closely associated to Apomys Mearns) of “new endemic” rodents (Step-
rainforests as they are today but were more gener¬ pan et al. 2005; Jansa et al. 2006).
alist (Heaney et al. 2011). Thus they could have dis¬ If viewed in a deeper time scale, the affinities
persed through hypothesized montane forest refu- of the recent species radiation on Luzon may be
gia at lower elevations as has been demonstrated explained by pan biogeography and the correlation
in the archaeological record in Papua New Guinea of tectonic features with evolution. While the theory
(Pasveer et al. 2002). of island accretion is well described for Wallacea,
this investigatory angle is hampered by the current
lack of information on the timing of island accre-