The amphibians and reptiles of Luzon Island, Philippines, VIII: the herpetofauna of Cagayan and Isabela Provinces, northern Sierra Madre Mountain Range

Abstract We provide the first report on the herpetological biodiversity (amphibians and reptiles) of the northern Sierra Madre Mountain Range (Cagayan and Isabela provinces), northeast Luzon Island, Philippines. New data from extensive previously unpublished surveys in the Municipalities of Gonzaga, Gattaran, Lasam, Santa Ana, and Baggao (Cagayan Province), as well as fieldwork in the Municipalities of Cabagan, San Mariano, and Palanan (Isabela Province), combined with all available historical museum records, suggest this region is quite diverse. Our new data indicate that at least 101 species are present (29 amphibians, 30 lizards, 35 snakes, two freshwater turtles, three marine turtles, and two crocodilians) and now represented with well-documented records and/or voucher specimens, confirmed in institutional biodiversity repositories. A high percentage of Philippine endemic species constitute the local fauna (approximately 70%). The results of this and other recent studies signify that the herpetological diversity of the northern Philippines is far more diverse than previously imagined. Thirty-eight percent of our recorded species are associated with unresolved taxonomic issues (suspected new species or species complexes in need of taxonomic partitioning). This suggests that despite past and present efforts to comprehensively characterize the fauna, the herpetological biodiversity of the northern Philippines is still substantially underestimated and warranting of further study.


Introduction
The highly distinctive terrestrial vertebrate fauna of the northeastern Philippines has been the subject of intense interest, speculation, and debate since the first historical explorations of the northern extremes of the archipelago (Wallace 1860(Wallace , 1876Everett 1889;Boulenger 1894;Stejneger 1907;Hoogstral 1951;Allen et al. 2004Allen et al. , 2006. Although many past and recent explorations of this unique part of southeast Asia highlighted spectacular endemic species (Stejneger 1907;Ota and Ross 1994;Oliveros et al. 2011), the dominant view of the Philippines by the beginning of the 20 th century was the biogeographer's concept of a "fringing" archipelago (Dickerson 1928;Kloss 1929;Darlington 1957;Myers 1960Myers 1962Brown and Alcala 1970a;). According to this perception, archipelagos near a continental source for invasion by vertebrate colonists should show distribution patterns consistent with the classic "immigrant pattern" of faunal distributions (Myers 1962;Brown and Alcala 1970a;Lomolino et al. 2010). Thus, early biogeographers expected species to be distributed along possible migration corridors, with various groups extending no further in distance from the continental source, than their relative dispersal abilities would allow (Taylor 1928;Inger 1954;Darlington 1957;Myers 1962;Carlquist 1965;Brown and Alcala 1970a). With respect to the northern Philippines, the most often-cited dispersal corridors included the western island arc (Borneo-Palawan-Mindoro) and the eastern Island chain (Sulu Archipelago-Mindanao-Leyte-Samar; Myers 1962;Esselstyn et al. 2004;Brown and Guttman 2002;Jones and Kennedy 2008;, with more limited evidence in support of southward colonization from Taiwan (Taylor 1928;Kennedy et al. 2000;Esselstyn and Oliveros 2010).
In the context of this biogeographical world view, islands like Luzon, at the tail ends of island chains and possible dispersal routes from the continental source (Diamond and Gilpin 1983;Brown and Guttman 2002;Jones and Kennedy 2008;but see Taylor 1928;Kennedy et al. 2000;Esselstyn and Oliveros 2010) were viewed as the extreme end points of faunal dispersal and dispersion (Huxley 1868;Darlington 1928;Myers 1962;Esselstyn et al. 2004). As a consequence, numerous classic works consider the biodiversity of such islands as "depauperate" in the sense that they contained a reduced set of species shared with a continental mainland source (Dickerson 1928;Taylor 1928;Inger 1954;Brown and Alcala 1970a;Dickinson et al. 1991;de Jong 1996;Lomolino et al. 2010). The view of a depauperate Luzon fauna has persisted throughout the last half century in discussions of its herpetofauna (Inger 1954;Leviton 1963;Brown and Alcala 1970a, 1978. Recently however, a renewed interest in faunistic studies of the northern Philippines (Brown et al. 1996(Brown et al. , 2000aDiesmos et al. 2005;Siler et al. 2011a) has produced a series of notable discoveries R. Brown et al. 1999R. Brown et al. , 2000bR. Brown et al. , 2008R. Brown et al. , 2009W. Brown et al. 1997aW. Brown et al. ,b,c, 1999aSiler et al. 2009Siler et al. , 2010aWelton et al. 2010;Fuiten et al. 2011), drawing attention to high levels of species diversity, preponderance of inferred autochthonous speciation, and substantial endemism in the northern reaches of the archipelago (Diesmos et al. 2005;. Together these studies have suggested that the northern portions of the archipelago may, in fact, be substantially more biologically diverse than currently appreciated. Thus, it is conceivable that, despite past expectations, species richness at a given northern Luzon site may be potentially as high as that demonstrated for the southern portion of the archipelago, adjacent to the Sunda Shelf (Nuñeza et al. 2010;Diesmos et al. 2005;Diesmos and Brown 2011;.
Recent works suggest that the northern end of Luzon Island (Fig. 1) and the islands between Luzon and Taiwan (Oliveros et al. 2010) represent the very last extent of conceivable dispersion of faunal elements of Sundaic origin (Dickerson 1928;Inger 1954Inger , 1999Inger and Voris, 2001). Recent studies have considered the diversity of herpetofaunas of the islands north of Luzon (Oliveros et al. 2010) and the northern end of the Cordillera Mountains of northwest Luzon (Diesmos et al. 2005;. In this paper we take the first step towards gaining a better understanding of the faunal communities of the northeastern-most extreme of Luzon by considering the amphibians and reptiles of the northern Sierra Madre Mountains (Fig. 1). We provide the first attempt to synthesize the known herpetological diversity of Cagayan and Isabela provinces (Fig. 1), northern Luzon Island (see also van Beijnen 2007, Diesmos 2008. We present data from our own survey work, as well as those from historical museum collections derived from the northern Sierra Madre Mountains. Because very little has been published previously on the herpetological communities of the area, all of these records constitute major range extensions and substantial expansion of our knowledge of resident biodiversity. This work contributes to a growing body of recent literature demonstrating that the herpetological communities of Luzon Island are species rich, composed of high percentages of endemic taxa, and are regionally unique in comparisons to the other zoogeographical regions of Luzon (Auffenberg 1988;Brown et al. 1996Brown et al. , 2000aDiesmos et al. 2005;Welton et al. 2010;Siler et al. 2011a;McLeod et al. 2011;Devan-Song and Brown 2012).
Cagayan and Isabela Provinces: Geography and Landscape. Cagayan and Isabela provinces lie at the extreme northeastern portion of Luzon (Fig. 1), with land areas totaling more than 9,000 and 10,664 square kilometers, respectively. Cagayan contains 28 municipalities and 825 barangays (villages) while Isabela contains 35 municipalities and 1018 barangays. Their capital cities are Tuguegarao and Ilagan, respectively. Inhabited by six major enthnolinguistic groups (Ilocanos, Ibanags, Malauegs, Itawis, Gaddangs, and Aetas), together they are home to more than 2.6 million human residents (NSO, 2010).
Both Provinces are dominated by three strikingly distinct geographical and topographical features: the wide alluvial plains surrounding the Cagayan River valley, the northern extremes of a strikingly elongate north-south mountain range (The Sierra Madre;, and the narrow strip of coastal forest along the north (Figs 2-3) and the east coasts of northern Luzon and the Philippine Sea (Fig. 1). Portions of the southwestern corner of Cagayan (bordering the province of Kalinga) and all of the western portions of Isabela (bordering Kalinga, Mountain, and Ifugao provinces) abut the foothills of the central Cordillera Mountains of Luzon (Fig. 1). Roughly a third of the land area of these provinces is near sea level; the majority of the remaining area constitutes the mountainous terrain of the northern Sierra Madre Mountain Range and the sprawling foothills to the west and east of this elongate mountain massif.
The Babuyan Island Group across the Balintang channel to the north of Luzon ( Fig. 1) is included administratively in Cagayan Province; this biogeographically distinct region has recently been reviewed for its herpetofauna (Oliveros et al. 2010) and will not be treated in detail here.

Materials and methods
We surveyed amphibian and reptile diversity at numerous sites throughout Cagayan and Isabela provinces (Table 2) using standardized sampling techniques (Heyer et al. 1994) and specimen collection and preservation methodology (Simmons 2002;ASIH 2004). Our most recent surveys (July-August, 2011) involved intensive elevational transects at the extreme northern end of the Sierra Madre Mountain Range in the Mt. Cagua area (Municipality of Gonzaga, Barangays Magrafil and and Santa Clara; Fig. 1). Surveys were conducted in early mornings, mid-day, afternoons, and evenings by experienced teams of four to eight individuals, sampling a wide variety of habitat types within each general study location. Habitats included dry forest on ridges, moist ravines, forest trails at all elevations, dry intermittent streambeds, small streams, large rivers, forest gaps and edges, and grassy open areas . Investigators at each sampling location made extensive surveys of each area (on foot) to ascertain habitat types and then visited each at varying times of the day. Nocturnal searches (1800-2400 hr) were conducted at each habitat type, within each sampling site, on dry and rainy nights. By concentrating field survey efforts to span the end of the dry season and the beginning of the rainy season (June-August) we were able to assure that each habitat type at each location was sampled under differing atmospheric conditions. Sampling Locations. Data presented here include results of our own surveys (Table 1) and a variety of collections, both intensive and incidental, from major U.S. Museum collections (see acknowledgements). In addition, an extensive series of collections housed at the USNM (field work of R. I. Crombie), KU and PNM (field work of ACD and surveys of MVW, EJ, and DR) targeted several localities to the south, in central Cagayan Province and Isabela Province. To be as comprehensive as possible in our treatment of Cagayan and Isabela, we include all of these records here, with the caveat that methods of surveying herpetological communities most likely differed among collection efforts and locations.

Results
We document 101 species of amphibians and reptiles from Cagayan and Isabela provinces, including 29 frog species, 30 lizards, 35 snakes, two freshwater turtles, three marine turtles, and two crocodilian taxa (Table 2). Taken together this diversity represents approximately 35% percent of the total Philippine herpetofauna (approximately 350 species ;Brown 2007;Diesmos and Brown 2011;Brown and Stuart 2012) and 70% of the taxa recorded are Philippine endemics. Below we provide accounts for each species, provide notes on their natural history and habitat, and highlight many unresolved taxonomic problems (involving 38% of the species included) that are relevant to particular taxa. We also comment on the conservation status of individual species when data presented here suggest that existing conservation status assessments (IUCN 2010 are out of date McLeod et al. 2011; or will soon require revision.          (Brown & Alcala 1980) N N Parvoscincus decipiens (Boulenger, 1895) (Stejneger, 1908) N N Parvoscincus tagapayao (Brown, McGuire, Ferner & Alcala 1999)

Platymantis corrugatus (Duméril, 1853)
Platymantis corrugatus (Fig. 20), as presently recognized, is a widespread endemic species found throughout the archipelago. There is considerable color pattern variation, but the species can be generally diagnosed by its medium body size, some form of a dark (gray, brown, or black) facial mask, and elongate tubercular ridges running along the dorsal surface. We observed this species at many locations calling most intensively at sunset (1800-1900 hr) after which it only called intermittently. The species commonly calls from beneath some kind of ground cover (leaf of other debris) on the forest floor. Its call sounds to the human ear like a raspy "whaaah…whaaah." Cagayan Province-

Platymantis cornutus (Taylor 1922)
Originally described on the basis of a single specimen from Balbalan, Kalinga, in the northern Cordillera Mountain Range (holotype CAS 231501; Taylor 1920Taylor , 1922a, this species ( Fig. 21) is widespread, commonly encountered, and locally abundant (given sufficient precipitation) at mid-to high-elevation sites in the Sierra Madre Range (Brown et al. 2000a;Diesmos et al. 2005;Siler et al. 2011a). We have no reliable records of any other member of the P. guentheri Group (Brown et al. 1997b(Brown et al. , 1997c of frogs at the same localities where P. cornutus has been recorded in the mountains of extreme northern Luzon, rendering our confidence in this identification very high. Platymantis cornutus calls from understory vegetation immediately following rain and is most frequently encountered on axils and along fronds of aerial ferns. This species deposits direct-developing embryos in small clutches (6-8 eggs) on fern axils ). It has one of the most rapid advertisement calls of any Philippine Platymantis, sounding to the human ear like "Tuk-tuk-tuk-tuk…" with 10-20 rapidly-delivered individual pulses. Geographic records reported here contribute to the continued expansion of this species' range throughout much of northern Luzon, supporting Brown et al.'s (2012) action of downgrading this species from "Vulnerable" (VU) to "Near Threatened" .

Platymantis polillensis (Taylor 1922)
Platymantis polillensis (Fig. 22) is a small, herbaceous-layer specializing arboreal species encountered most often in ferns and shrubs colonizing disturbed forest edges, secondary growth forest, forest gaps, and tree falls. Previously considered "Critically Endangered," or "Endangered" ) and endemic to the island of Polillo (Quezon Province, off the coast of SE Luzon; holotype CAS 62250), this species is now known to be widespread, commonly encountered (given occurrence of precipitation and preferred habitat type), and often locally abundant (Brown et al. 2000aSiler et al. 2011a;McLeod et al. 2011). The major range extension reported here supports Brown et al.'s (2012) downgrading of this species conservation status to  "Near Threatened" (IUCN 2010) based on its additional presence in Aurora Province, southern Luzon. This species calls with a slow series of amplitude-modulated high frequency "chirps" following sufficient precipitation.

Platymantis pygmaeus Alcala, Brown & Diesmos, 1998
Platymantis pygmaeus (Fig. 23) was originally described from Palanan, Isabela Province, and is now known to be widespread and abundant in Bulacan, Quezon, Aurora, Kalinga, Isabela, Cagayan, and Ilocos Norte provinces Brown 2000b;McLeod et al. 2011). The substantial distributional record reported here, while not surprising, constitutes additional evidence in support of Brown et al.'s (2012) downgrading of this species conservation status from "Vulnerable" (IUCN 2011) to "Near Threatened" (IUCN 2010). This is the smallest species of Platymantis in the Philippines (male SVL 12-15 mm) and it can be recognized in life by its high frequency "click-click-click…" advertisement call and its preference for calling from low (0.3-1.0 m), shrub layer vegetation.

Platymantis sierramadrensis Brown, Alcaka, Ong & Diesmos, 1999
Platymantis sierramadrensis was described on the basis of specimens from Barangay Umiray, Municipality of General Nakar, Quezon Province (holotype PNM 6465), from Aurora Province (paratypes 204738, 204742-45), and other, non-type material from Palanan, Isabela Province (CAS 204739, 204740, CAS 204741). Subsequent confusion in identification of Platymantis sierramadrensis has involved a suspicion that two separate taxa may have been attributed to this species, a confusion that may have undermined the type description Brown et al. 2000a, Siler et al. 2011a). Since the realization of this potential problem, we have twice noted (Brown et al. 2000a;Siler et al. 2011a) the presence of two sympatric small bodied P. hazelae Group (Brown et al. 1997b) species, one of which appears to be most abundant at lower elevations (approximately 400-700 m) in disturbed habitats and another that is often encountered at the upper end of this elevational range, but is most abundant at elevations above 900 m. We consider the lower elevation species, with a "chirp" mating call, to be the widespread, common species P. polillensis, and the slightly larger bodied, high elevation species, tentatively assigned to P. sierramadrensis. The latter calls with a pure, constant frequency call, sounding to the human ear like the ringing of a small bell (thus differing from the "chirp" call of P. polillensis). Current IUCN conservation classification for this species is "Vulnerable (B1ab(iii))," based on our assessment from 2004 (IUCN 2011). Considering the taxonomic confusion still surrounding this species, the lack of reliable past records, and the absence of any convincing evidence  of population, area of occurrence, or habitat decline, we now consider this species to be "Data Deficient (DD;IUCN 2010. Once the taxonomy of this species is clarified with a return to the type locality in General Nakar to determine which call type occurs there, direct, field-based data gathered from natural populations (and not inferred from forest cover) will be necessary to reconsider a higher possible conservation threat level

Platymantis taylori Brown, Alcala, Diesmos, 1999
Since the time of its discovery , this species (Fig. 24) has been documented only at the Municipality of Palanan (Barangay Didian). This taxon was diagnosed primarily on the basis of its relatively large body size and distinctive advertisement call, sounding to the human ear like the buzz produced by a Geiger counter. This species previously has been classified by IUCN as "Endangered" (EN; B1ab(iii); IUCN 2011, 2011), on the basis of its purported limited range and anticipated decline in habitat due to the presence of logging at low elevations along Luzon's east coast near Palanan.
Long overdue for a conservation status revision, we categorize this species as "Data Deficient" (DD) because (1) it has been recorded only once and no repeat surveys to the immediate or surrounding areas have been undertaken to determine the extent of its range, and (2) there is no evidence that this taxon requires intact, low-elevation forest and no evidence to suggest that it is range-restricted. Thus, there is no way to determine whether continued degradation of lowland coastal forests in Palanan will adversely affect this species. Originally characterized as "common and widespread" at the original collection site IUCN 2004), its range presumably includes an extremely large protected area, supporting our conviction that this species must be downgraded to a low conservation threat category (e.g., "Near Threatened," NT) or, more appropriately, considered "Data Deficient" until some attempt is made to study it in the field and more surveys in surrounding areas are conducted. Platymantis taylori is another example of a case in which negative data have been used inappropriately for conservation status assessment ), resulting in a higher level of threat category when, in reality, virtually nothing is known of its biology, natural history, habitat requirements, and actual conservation status.

Platymantis sp. 1 "Yokyok"
This distinctive form (Fig. 25) is now known from two sites in Cagayan and Isabela provinces (both between 400 and 500 m in disturbed forested habitats). We suspect that this possible undescribed species is much more widespread and will be frequently encountered if surveys can be conducted in intervening localities. This terrestrial spe-  cies is slightly smaller than the morphologically similar P. cagayanensis and P. sp. 3 "seeyok," and calls with a long pulse train, sounding to the human ear like "Yok-yokyok-yok…." Cagayan Province-Location 1b: KU 330628-35. Isabela Province-Location 3: KU 307608-09, 327587; Location 30: no specimens (MVW photo voucher).

Platymantis sp. 2 "Cheep-cheep"
We encountered another potentially distinct species (Fig. 26) of Platymantis at both high and low elevation sites on Mt. Cagua, Municipality of Gonzaga. The suspected new species appears phenotypically most similar to P. lawtoni from Sibuyan Island (Brown and Alcala 1974;Brown 1998, 1999), but is distinguished from other Luzon taxa by its distinct coloration, smooth dorsum, semi-aquatic microhabitat preference, and distinctive "cheep-cheep-cheep…" vocalizations.

Platymantis sp. 3 "See-yok"
This suspected new species (Figs 27,28) was first observed in Old Balbalan Town (Kalinga Province; RMB and ACD, personal observations) and has since been recorded at many sites throughout central and northern Luzon. Morphologically most similar to P. cagayanensis, this species can reliably be identified by its silvery iris (versus the bright yellow-orange iris in P. cagayanensis) and by distinctive advertisement call, sounding to the human ear like "seee-yok…seee-yok" (Brown et al. unpublished data).

Platymantis sp.
Without genetic data, information on mating calls, and/or photographs in life, numerous museum specimens of ground-dwelling, medium sized, dorsally tuberculate members of the genus Platymantis cannot confidently be identified to species. Many have previously been identified by field collectors as P. dorsalis on the basis of generalized morphological similarity to that southern Luzon (type locality: Laguna Bay) species (Brown et al. 1997c;Brown 1998, 1999). They are clearly morphologically distinguishable from the terrestrial species P. corrugatus (color pattern differences and presence of dorsolateral dermal tubercular ridges in P. corrugatus), P. sp. 2 "cheep-cheep" (color pattern differences, and absence of any dorsal tubercles in P. sp. 2 "cheep-cheep), P. pygmaeus (much larger body size), and the arboreal species P. cornutus, P. polillensis, and P. sierramadrensis (all of which have expanded finger and toe pads). With on-going taxonomic work, these specimens may be identifiable to P. cagayanensis, P. sp. 1 "Yokyok," P. sp. 3 "See-yok" or they may eventually prove to be new, undescribed species.

Family Dicroglossidae
Fejervarya moodiei (Taylor 1920) Fejervarya moodiei (Fig. 29) is a widespread, endemic estuarine specialist that can be found in a variety of coastal areas including brackish water swamps. Previously considered conspecific with the widespread Southeast Asian species F. cancrivora, recent genetic evidence suggests that the Philippine populations are genetically distinct; the available name for the Philippine population is Fejervarya moodiei (Kurniawan et al. 2010(Kurniawan et al. , 2011. Widespread and common at most coastal areas throughout the Philippines, this species is clearly most appropriately considered "Least Concern" (LC; IUCN 2011).

Fejervarya vittigera (Wiegmann, 1834)
Fejervarya vittigera is a widespread, low elevation species typically observed in highly disturbed areas with standing water (rice fields, ponds and lakes) or along small, denuded streams near coastal areas or canals bordering agricultural areas. Our specimens were found along muddy stream banks in disturbed forests at the edge of agricultural plantations. Although until now this species has always been considered "Least Concern" (IUCN 2011), and not threatened, recent evidence suggests that populations of this endemic low elevation taxon may be in decline (ACD and M. L. Diesmos, personal observation) due to the spread of exceptionally high density populations of the introduced (Diesmos et al. 2006) Asian Bullfrog, Hoplobatrachus rugulosus (see below), which appears to displace, out-compete, or otherwise competitively exclude F. vittigera in some areas

Hoplobatrachus rugulosus (Wiegmann, 1834)
This introduced species (Fig. 30) was first detected in Laguna province in 1996 (Diesmos et al. 2006), but has since been encountered throughout low-lying valley systems bisecting most major islands in the Philippines. Hoplobatrachus rugulosus achieves remarkable population densities in large areas of rice cultivation and we have witnessed thousands of individuals in a single day's hike, actively foraging during day light hours,  voraciously hunting any potential prey item (including juveniles of their same species and sympatric congeners; ACD and RMB, personal observations). In 2001 RMB and ACD drove the length of the Cagayan Valley, stopping frequently to interview farmers about the densities of frogs in their fields. All reported that these distinctively larger frogs were now the dominant species in the area (and the smaller, previously more common species [presumably Fejervarya vittigera] was now far less common). Additionally, in recent trips to Ilocos Norte , ACD and party found exceptionally high densities of H. rugulosus in agricultural areas and along riverbanks and very few native F. vittigera in the presence of this invasive species.

Limnonectes macrocephalus (Inger 1954)
The Luzon fanged frog Limnonectes macrocephalus ( Fig. 31) inhabits rivers and streams from sea level up to high elevation forests. Although targeted by humans for food and potentially at risk from predation and competition from invasive species (Diesmos et al. 2006), the Luzon fanged frog has always been characterized as common in mid-to high-elevation forests (Brown et al. 1996(Brown et al. , 2000aDiesmos et al. 2005;Siler et al. 2011a;McLeod et al. 2011). Although this is Luzon's largest species, low elevation populations, subject to predation by humans and introduced frog species, consistently have a smaller average body size than do high-elevation populations inhabiting inaccessible montane areas (RMB and ACD, personal observations). The largest individuals have been documented from small, high-elevation mountain streams that lack above ground connections to large rivers at lower elevations (Brown et al. 2000a;Diesmos et al. 2005). Thus, we assume a lack of connectedness has impeded subsistence harvesting in these areas and L. macrocephalus' indeterminate growth pattern had allowed these populations to achieve high average body sizes (of up to 350-400 g in males) in the absence of human predation.

Limnonectes woodworthi (Taylor 1923)
Limnonectes woodworthi (Fig. 32) is a commonly encountered stream frog in the mountains of southern Luzon and throughout the Bicol Peninsula Brown et al. 1996Brown et al. , 2000a; more recent studies have determined that this species may also occur farther north, along the foothills of the Sierra Madre Mountains in Aurora Province (Siler et al. 2011a), Isabela Province (ACD, unpublished data), and as far north as Cagayan Province, the Babuyan Islands, and Ilocos Norte Province (Oliveros et al. 2010;). However, the northern  populations have a somewhat distinctive color pattern, suggesting they may be taxonomically differentiated. Future studies involving morphometrics, advertisement calls, and genetic data will be necessary to test for the presence of possible species boundaries within L. woodworthi.

Occidozyga laevis (Günther, 1859)
Occidozyga laevis (Fig. 33) is a common, widespread species known throughout many of the islands and neighboring continental landmasses of Southeast Asia (Inger 1954(Inger , 1999Inger and Voris 2001). Although we have noted body size and call variation at a few sites in the Philippines (ACD and RMB, unpublished data), no taxonomic studies have as of yet targeted this variable taxon. Our specimens were collected along banks of rivers and streams (in quiet side-pools and adjacent puddles), or in puddles on basins on the forest floor, adjacent to flowing water. In the Philippines, individuals aggregate to form breeding groups, with males emitting clicking pulses, sounding to the human ear like the tapping together of small stones.

Kaloula kalingensis Taylor 1922
Kaloula kalingensis (Fig. 34) originally was described from Balbalan, Kalinga Province (Taylor 1922a). However, as currently understood, this taxon is now considered common and widespread throughout much of northern Luzon (Brown et al. 1996(Brown et al. , 2000aDiesmos et al. 2005;Siler et al. 2011a;McLeod et al. 2011;Blackburn et al. in review). Typically encountered in water-filled holes in trees (30-100 cm trunk diameter; holes 1-4 m above the ground) in low to mid-elevation forested areas, this species tolerates high levels of disturbance and is often even found in thick invasive stands of introduced species of bamboo, provided that water-filled cavities provide its favored calling, courtship, breeding, and egg deposition microhabitat (Brown and Alcala 1982;personal observation). These observations recently prompted  to downgrade this species IUCN conservation status from "Vulnerable" (IUCN 2011) to "Near Threatened" (IUCN 2010) and our data support this action. However, recent molecular studies by Blackburn et al. (in review) suggest that K. ka-  lingensis may be a complex of three or four taxonomically distinct entities, which may result in one or more of these putative species (or, at least significant evolutionary units [ESUs] for conservation) to exhibit a more restricted geographical range. If so, the conservation status of these individual putative species (or ESUs) will need to be individually assessed for conservation threats using field-based data of the actual population abundances and distribution (i.e., not inferences from forest cover).

Kaloula picta (Duméril and Bibron, 1841)
Kaloula picta (Fig. 35) is a widespread Philippine endemic, distributed widely in low elevation agricultural areas, along riparian habitats in the foothills of mountain systems, and along low-elevation river valleys and coastal areas (Inger 1954;Brown and Alcala 1970a;. Nearly genetically identical throughout the archipelago (Blackburn et al. in review), Kaloula picta may be another species that has recently undergone rapid range expansion as a result of population transplantation in agricultural shipments, coupled with the conversion of most low-elevation coastal floodplains into its preferred habitat (i.e., flooded rice fields; Brown et al. 2010a

Kaloula pulchra Gray, 1825
Kaloula pulchra is an invasive species (Diesmos et al. 2006) only detected in the country in the last decade and suspected of being introduced through the pet trade. This species has become widely distributed on Luzon and several other islands. Recent observations suggest that, in disturbed habitats, K. pulchra's impact on native species may be increasing ). We encountered this species in agricultural areas and heavily disturbed riparian habitats (polluted streams near residential areas) along the Cagayan Valley and wide floodplains surrounding the Cagayan River. It is considered "Least Concern" (LC; IUCN 2011).

Kaloula rigida Taylor 1922
Kaloula rigida (Fig. 36) was described from Baguio City, Benguet Province (Taylor 1922a) and is now known to be widespread in Kalinga, Apayao, Ifugao, and Benguet Provinces of the Cordillera Mountain Range and Isabela and Cagayan Provinces of the northern Sierra Madre (Taylor 1922a;Inger 1954;Diesmos et al. 2005;). This species is a fossorial, ephemeral, pool-breeding specialist that emerges immediately following heavy rains at the onset of the rainy season  (June-August) but may be otherwise undetectable if field surveys are conducted in dry months ). Our new records, constituting a major range extension and confirmation of this species continued existence in heavily disturbed forest, further support Brown et al.'s (2012) downgrading of the IUCN (2011) conservation status for this species from "Vulnerable" to "Near Threatened" (IUCN 2010). This species calls in large choruses in temporary pools following heavy rains. Individuals call with repeated pulses, sounding to the human ear like the striking together of two pieces of wood; in large choruses, the collective sound of many individuals calling sounds like a single-stroke engine or small generator.

Hylarana similis (Günther, 1873)
Hylarana similis (Fig. 37) is ubiquitous throughout Luzon and associated landbridge islands Diesmos 2002, 2009) where it is locally abundant in all riparian habitats sampled (Brown and Guttman 2002). This species ranges from coastal plains near sea level, to the foothills of all of Luzon's major mountain ranges, where it is particularly abundant, to mid-and high-elevation forested regions. Without any evidence of population declines and considering its wide distribution,  argued for the downgrading of this species from "Vulnerable (IUCN 2010) to "Near Threatened" (IUCN 2010). This latter designation was considered a compromise because although no declines have been noted (and given current IUCN criteria for assessing conservation threat, this species is most appropriately classified as "Least Concern"), recent studies have determined that this species exhibits high levels of chytrid fungus infection at one low elevation site in southwest Luzon (Swei et al. 2011

Sanguirana luzonensis (Boulenger, 1896)
This widespread, Luzon faunal-region Diesmos 2002, 2009) endemic ( Fig. 38) is morphologically variable and exhibits a particularly broad set of habitat tolerances, from coastal waterways, to disturbed lowland riparian habitats, and rivers and streams at the foothills of all Luzon mountain ranges. This species is particularly abundant from low-(200-300 m) to high (up to 1700-1800 m) elevation forested  areas and appears quite tolerant of anthropogenic disturbances. It is found on rocks, exposed gravel beds, muddy banks, and low, shrub layer vegetation along nearly all of Luzon's waterways. The major range extensions and wide variety of habitat types reported here support Brown et al.'s (2012) downgrading of the conservation status (IUCN 2011) of this ubiquitous, disturbance-tolerant species, from "Near Threatened" to "Least Concern" (IUCN 2010). Sanguirana luzonensis calls in quiet side pools or when water levels are low and ambient noise is reduced; thus it appears to breed in the late dry season (March-May) and calls with a soft series of descending-frequency "peeps" and "whistles" (Brown et al. 2000a(Brown et al. , 2000bFuiten et al. 2011

Sanguirana tipanan (Brown, McGuire, and Diesmos, 2000)
The presence of this species (Fig. 39), originally described from Aurora Province (Brown et al. 2000a,b), in Palanan has been confirmed by ACD; to date no specimens are available in museum collections from Cagayan or Isabela Provinces. Although no follow-up surveys have been performed in Palanan, additional surveys in Aurora Province  have documented this taxon at four new sites, suggesting that it probably no longer qualifies for "Vulnerable" (VU; IUCN 2010, 2011) status. This species was not documented in our extensive montane surveys at the northern tip of Luzon (Mt. Cagua, Municipality of Gonzaga). However, until more fieldwork is conducted in the intervening forested mountains of Isabela and Cagayan Provinces to determine the extent of this species range, little can be interpreted from the apparent northern extent of S. tipanan's occurrence at Palanan.

Philautus surdus (Peters, 1863)
A single specimen of this widespread Luzon-region rhacophorid frog has been collected in Palanan. The loud "crunch…crunch" vocalizations of this species have been heard by the authors at Barangay Nassiping, Municipality of Gattaran.

Polypedates leucomystax (Gravenhorst, 1829)
Philippine Polypedates leucomystax (Fig. 40) is a genetically distinct variant of a widespread species complex ranging throughout much of Southeast Asia (Inger 1954,   (Brown et al. 2000a,b). This species has been ovserved at Location 30 by ACD. Photo: J. McGuire. 1999; Brown et al. 2010a;Kuriashi et al. 2012). Within the archipelago, this species is genetically identical throughout most of its range, but with two genetic types occurring in the Mindanao faunal region Diesmos, 2002, 2009), one of which is shared with northern Borneo and southern Peninsular Malaysia, suggesting two invasions of the Philippines . The existence of a widespread single haplotype throughout the Philippines suggested to Brown et al. (2010a) that this distribution may have arisen from demographic range expansion following the last several centuries of habitat conversion and human mediated dispersal throughout the country. This species is known from dry, coastal areas near agriculture, to 1000+ m high in the Northern Cordillera where it has been found in pristine forests at high elevation (Diesmos et al. 2005(Diesmos et al. , 2006. Polypedates leucomystax constructs foam nests above water (Brown and Alcala 1982) and calls with loud, single "Craaaak!" or "Plehht!" vocalizations.

Rhacophorus pardalis Günther, 1859
This species of Southeast Asian "flying frog" (Fig. 41) is also known from the islands of Indonesia and Malaysia (Brown and Alcala 1994;Inger 1999). In the Philippines it breeds in vegetation above stagnant water in side pools along rivers, water buffalo wallows, or temporary pools in forests. Rhacophorus pardalis constructs foam nests above water (Brown and Alcala 1982) and calls with soft "rattle" or "buzz" (personal observations).

Rhacophorus appendiculatus (Günther, 1858)
Rhacophorus appendiculatus, although considered widely distributed (Inger 1954(Inger , 1999Brown and Alcala 1994;Alcal and Brown 1998) on numerous Philippine islands, is patchily distributed on Luzon (Brown and Alcala 1994;McLeod et al. 2011). We have most often encountered this species following heavy rains, in dense choruses surrounding large temporary swamps or pools in forests of varying degrees of disturbance, and at low-to mid-elevations (300-700 m; see Siler et al. 2011a;McLeod et al. 2011). Two recently collected small specimens (Fig. 42) from high-elevation forests on Mt. Cagua appear to fit this species diagnosis (Brown  and Alcala 1994) with the caveat that their small body size and reduced tarsal dermal fringes suggest to us at least the possibility that some morphological variation in this group, and possible taxonomic significance if bolstered by future studies of ecology, morphology, habitat, genetic, and call variation.

Bronchocela marmorata Gray, 1845
We collected individuals of this widespread northern Luzon species (Fig. 43) 4-8 m above the ground in secondary growth trees and agricultural hedgerows. Our specimens are clearly diagnosable, in accordance with Taylor's (1922b) definition, as B. marmorata. However at numerous sites throughout the southern portions of Luzon (Brown et al. 2000aSiler et al. 2011a;McLeod et al, 2011), specimens appear to match the definition of B. cristatella (Kuhl 1820;Hallermann 2005), and yet are genetically identical to specimens that key out to B. marmorata (Hallermann 2005;Brown, Welton, Rock, Siler, and Diesmos, unpublished data). These observations sug-  gest the strong possibility that the characters utilized to define these two nominal species' on Luzon vary clinally and/or ontogenetically. Clearly, further study is warranted; we note that if, as we suspect, only a single species in this group exists on Luzon (as unpublished genetic data would suggest), the correct name for that species would be B. marmorata (Taylor 1922b

Draco spilopterus (Wiegmann, 1834)
This widely distributed Luzon and Visayan faunal-region Diesmos 2002, 2009) endemic achieves particularly high densities in coastal coconut palm plantations, but it is also found at lower densities in disturbed and primary forests throughout the northern Philippines (McGuire and Alcala 2000). Our specimens (Fig.  44) were collected in patchy, disturbed (selectively logged) forests at low elevations, adjacent to clearings caused by shifting, slash-and-burn agriculture.

Cyrtodactylus philippinicus (Steindacher, 1867)
One of the most common squamates in the northern Philippines, Cyrtodactylus philippinicus (Fig. 45) is common from low-to mid-elevation forests, at elevations of 800 or 900 m (Brown et al. 1996(Brown et al. , 2000aDiesmos et al. 2005;Siler et al. 2011a). This species is typically found along riparian habitats, and is active at night on rocks and boulders, over-hanging stumps and logs, or on root balls of large trees, exposed by flowing water.

Gehyra mutilata (Wiegmann, 1834)
Gehyra mutilata (Fig. 46) is a common, widespread "house" gecko that differs from the other most common human commensals (Hemidactylus platyurus and H. frenatus) in that it prefers dark perches, away from overhead lights, probably as a result of competitive interactions with these latter species (Ota 1989). Our specimens were found on darkened walls of houses and on trunks of trees in residential areas.

Gekko gecko (Linnaeus, 1758)
Gekko gecko (Fig. 47) is a widespread Southeast Asian species. In the northwrn Sierra Madre, specimens have been observed on a variety of man-made structures in and around human habitation (but this species is less frequently encountered in forests). Although we heard its distinctive vocalizations at many of the sites we visited in Cagayan and Isabela Provinces 19,(20)(21)(22)(28)(29)(30), we seldom endeavored to collect this identifiable and well-known species. One specimen, captured on a house in a small village, was brought to us by a resident of Barangay Magrafil.

Gekko kikuchii (Oshima, 1912)
Gekko kikuchii (Fig. 48) is the name available for the genetically distinct northern Luzon and Lanyu Island (Taiwan) lineage ) of the Gekko mindorensis complex (Bauer 1994;Roesler et al. 2011;). This species is found on large boulders in riparian habitats and on concrete structures bordering water (culverts, bridge pylons, walls, and cinderblock buildings). Clearly, the widespread Philippine Gekko mindorensis Complex (Taylor 1922;Brown and Alcala 1978;) will require extensive taxonomic revision in the near future (Siler et al. unpublished data).

Hemidactylus frenatus Duméril & Bibron, 1836
One of the most common "house" geckos ( Fig. 49) in the Philippines, this species is frequently encountered under exterior lights of buildings, preying on insects attracted to artificial illumination. It can be easily diagnosed from H. platyurus by its round tail and smooth, non-frilled flanks.

Hemidactylus stejnegeri Ota & Hikida, 1989
Diagnosed as a triploid species (Ota and Hikida 1989a,b), this taxon has only been encountered a few times in recent years (once in downtown Tuguegarao City, ACD personal observation). We suspect that it may be widespread and fairly common (and possibly common in existing collections) but that it escapes recognition by herpetologists who may, at a glance, misidentify this species as H. platyurus or H. frenatus. Outside of the Philippines, it has been documented in Taiwan and Vietnam (Ota et al. 1993).

Lepidodactylus cf. lugubris (Duméril & Bibron, 1836)
Although Brown and Alcala (1978) noted no populations of Lepidodactylus on Luzon Island, we have consistently (over the past decade) captured small numbers of specimens (Fig. 50) from a variety of habitats on the Bicol Peninsula, Bulucan, Aurora, Kalinga, Ilocos, and now Cagayan and Isabela provinces (Brown et al. 2000aDiesmos et al. 2005;Siler et al. 2011a;McLeod et al. 2011). We continue to identify these as L. cf. lugubris, albeit with the same reservations as articulated by Brown and   Alcala (1978): these populations are quite variable, may represent one or more undescribed species (or triploid clone of L. lugubris; R. Fisher, personal communication), often resemble L. planicaudus (i.e., possessing lateral tail tuberculation ;Stejneger 1905 [known from Polillo and Mindoro islands, immediately adjacent to Luzon; Brown and Alcala 1978]), or L. balioburius (Ota and Crombie 1989 [known only from Batan Island, north of Luzon; Oliveros et al. 2011]), and/or may be referrable to other possible taxa currently residing in the synonymy of L. planicaudus (e.g., P. naujanensis, Taylor 1919). Once detailed studies (in particular, with genetic data) become available, the taxonomic status of these Luzon populations will require careful consideration.

Luperosaurus cf. kubli Brown, Diesmos & Duya, 2007
A single specimen provisionally assigned to the rare species Luperosaurus kubli ; Fig. 51) has been observed in ultrabasic forests at Dyadyadin but to date, no specimens have been secured. Described from Nagtipunan (Qurino Province, just south of Isabela Province), L. kubli has been observed and collected only once (holotype = PNM 9156) and is considered a member of the large, robust-bodied Philippine endemic clade of Luperosaurus (Brown et al. 2000c). Like most species of Luperosaurus, the microhabitat preference of this species is unknown (Brown and Diesmos 2000).

Pseudogekko compressicorpus (Taylor 1915)
As currently defined, this species is widely distributed throughout the Philippines (Fig. 52), from extreme southwestern Mindanao, throughout the eastern island arc (Leyte-Samar) and the Bicol faunal region (including Polillo and Catanduanes islands), and widely throughout the rest of Luzon (Brown and Alcala 1978;RMB and CDS, unpublished data). This species is typically encountered on large leaves in shrub-and understory layer vegetation, at low-to mid-montane forested sites. Interestingly, it is often encountered on leaves at night following heavy rains (RMB, personal observation).

Brachymeles bicolor (Gray, 1845)
Brachymeles bicolor (Fig. 53) has remained one of the Philippines most distinctive and enigmatic skinks since the time of its original discovery (Gray 1845). Detected only a few times since its original description (Brown and Alcala 1980;Brown et al. 2000a;Diesmos et al. 2005), the species was recently redescribed on the basis of material from Aurora, Isabela, and Cagayan Provinces (Siler et al. 2011c). Brachymeles  bicolor is a forest species that is now predictably found in mid-to high-elevation forests (400-1200 m) and can be located by digging around rotting logs, stumps, and tree buttresses in forests of varying levels of disturbance. Specimens of Brachymeles bicolor were caught in both contiguous forests and natural forest remnants at elevations between 150 and 400 m within boundary of San Mariano town, Isabela Prov.

Brachymeles bonitae Duméril & Bibron, 1839
As currently defined, Brachymeles bonitae (Fig. 54) is a common, widespread species, endemic to the Luzon faunal region Diesmos 2002, 2009), Masbate, Mindoro, Lubang, and Camiguin Norte islands (Brown and Alcala 1980;Oliveros et al. 2010;Siler et al. 2011b). Commonly found within, around, and under rotting logs in loose soil, this slender burrowing skink appears to tolerate varying degrees of forest disturbance, but is usually absent in low-elevation plantations and agricultural areas, where it appears to be replaced by larger, surface dwelling species of Brachymeles (CDS, personal observation). The low-elevation (150-300 m) forests and coconut palm plantations in the Municipality of Gonzaga (Locations 1b, 2) appear to be an exception in that we found this species commonly in disturbed areas like coconut groves, but in the absence of larger pentadactyl species (e.g., B. kadwa; Siler and Brown 2010). Recent genetic evidence suggests B. bonitae is paraphyletic with respect to B. tridactylus of the Visayan island group (Brown and Alcala 1980;Diesmos 2002, 2009) suggesting that this may be yet another complex of multiple evolutionary lineages, deserving of taxonomic partitioning Brown 2010, 2011;Siler et al. 2011b;in press

Brachymeles kadwa Siler & Brown 2010
This recently described species (Fig. 55) is known from numerous localities across Luzon (Bicol Peninsula, Aurora, Isabela, Laguna, and Cagayan provinces) and also Calayan and Camiguin Norte islands, north of Luzon (Oliveros et al. 2010;Siler and Brown 2010). A large-bodied, pentadactyl species, it is active on the surface in a variety of forest types (with varying levels of disturbance) and, as a result of this activity pattern, it is frequently collected in pitfall traps. Often, it is exceedingly common in low-elevation coconut palm plantations Oliveros et al. 2010

Brachymeles muntingkamay Siler, Rico, Duya & Brown 2009
Recently discovered and described ; Fig. 56), this species previously was known only from Mt. Palali in Nueva Viscaya Province, south-central Luzon (Caraballo Mountain Range). We were quite surprised to find a large population of this distinctive species so far north in the northern Sierra Madre; genetic analysis confirms that the northern population is closely related to the southern, Mt. Palali lineage, suggesting that this species most likely occurs at intervening localities in mid-to highelevation, reasonably intact forests (700-1000 m) in four or five provinces (Siler et al. in press a). Specimens at these known locations were found inside rotten logs or in soil beneath rotting logs; none were collected in pitfall traps, suggesting that their largely fossorial lifestyle may have prevented discovery until targeted survey efforts associated with CDS's work ) resulted in their detection. Recent findings suggest that B. muntingkamay was also observed in a patch of lowland forest in Nassiping, San Mariano (Isabela Prov.), at 60 masl, midway between Mt. Cagua (Cagayan Prov.) and Mt. Palali (Quirino Prov.).

Eutropis multicarinata borealis (Brown & Alcala 1980)
Known from the northern Philippines and Lanyu Island near Taiwan, this species is considered widespread throughout Luzon and associated smaller island groups (Brown and Alcala 1980;Brown et al. 1996Brown et al. , 2000aOliveros et al. 2010;Diesmos et al. 2005;. Eutropis m. borealis (Fig. 58) exhibits inordinate amounts of geographically-based body size, scalation, and color pattern variation (Brown and Alcala 1980), suggesting to us that it may be composed of multiple independent evolutionary lineages worthy of taxonomic recognition. Future genetic studies with dense   geographical sampling will be necessary to test the hypothesis of a single species, composed of only two subspecies (Brown and Alcala 1980

Eutropis multifasciata (Kuhl, 1820)
This widespread Southeast Asian species (Brown and Alcala 1970a;Manthey and Grossman 1997;Fig. 59) is commonly encountered in the Philippines at lower elevations (coastal areas to several hundred meters in elevation) along edges of agricultural land surrounding disturbed forest patches. Eutropis multifasciata is active in open sun at midday and can be observed actively forging in the open and retreating into nearby shrubs when disturbed. This species is notable for its striking color polymorphism on lateral surfaces (bright green, orange, or yellow display surfaces), often with multiple color patterns exhibited within the same population.

Lamprolepis smaragdina philippinica Mertens, 1829
Lamprolepis smaragdina philippinica (Fig. 60) is one of the most locally abundant lizards in coastal areas throughout the archipelago. Also encountered in agricultural plantations (coconut palm plantations, avocado, cacao, and mango plantations) and in regenerating forest nurseries and riparian corridors, L. s. philippinica exhibits geographically based color variation, with fully green individuals at some localities, brown patches on the head and dorsal surfaces or forelimbs at other sites, and all gray-brown individuals at two known areas (Siler and Linkem 2011). These observations suggest to us that taxonomic partitioning of this species will most likely be necessary with future study

Lipinia cf. vulcania Girard 1857
Girard's (1857) single specimen of this distinctive species reportedly originated at1700 m asl on Dapitan Peak, Zamboanga Peninsula, of Mindanao Island; this unique specimen is now presumed lost (Brown and Alcala 1980). With some hesitation, Brown and Alcala (1980) referred an additional specimen from Luzon (specific locality data unknown) to L. vulcania. We suspect the two available specimens from Luzon (CAS 16472 and ACD [PNM] 2036) will eventually be recognized as a new species if researchers can visit the type locality on Mindanao and secure additional comparative material that would allow for a thorough taxonomic study. The Luzon population we refer to L. cf. vulcania (Fig. 61) is most likely an undescribed, but related species.

Otosaurus cumingi Gray 1845
This is the largest Philippine species in the Sphenomorphus Group (Greer and Parker 1967) and was formerly referred to the genus Otosaurus (Taylor 1922b), but later transferred to Sphenomorphus (Brown and Alcala 1980). New phylogenetic studies by Linkem et al. (2011) have resulted in the resurrection of a monotypic Otosaurus to accommodate this highly distinctive Philippine "giant" Sphenomorphus Group species. Otosaurus cumingi (Figs 62, 63) is most frequently encountered active at midday in mid-elevation forests (200-500 m) of varying levels of disturbance. It is a wide-ranging species, and can be encountered on the forest floor, not necessarily confined to riparian corridors or woody debris microhabitats.

Pinoyscincus abdictus aquilonius (Brown & Alcala 1980)
Recently transferred from the paraphyletic genus Sphenomorphus to the newly recognized genus Pinoyscincus ) on the basis of a multilocus phylogenetic analysis and a survey of new morphological characters, P. abdictus aquilonius (Fig. 64) is a medium-bodied forest species with a preference for intact, low-to mid-elevation habitats (100-600 m) with minimal disturbance (Taylor 1922b;Brown and Alcala 1980;Brown et al. 2000aMcLeod et al. 2011

Parvoscincus decipiens (Boulenger, 1895)
Recently transferred from the paraphyletic genus Sphenomorphus to an expanded Parvoscincus (Ferner et al. 1997) on the basis of a multilocus phylogenetic analysis and a survey of new morphological characters , P. decipiens (Fig. 61) is a small-bodied forest species with a preference for intact, mid-to high-elevation habi- tats (400-1200 m) with minimal disturbance (Taylor 1922b;Brown and Alcala 1980;Brown et al. 2000aMcLeod et al. 2011

Parvoscincus cf. decipiens
A second small leaf-litter skink species, related to P. decipiens (Linken and Brown in review; Fig. 66) is present at two localities in Cagayan and Isabela Provinces. With near identical ecological habits and microhabitat preference, this undescribed new species is only distinguishable on the basis of coloration, a few differences in scalation, and pronounced genetic variation (Linkem and Brown in review

Parvoscincus leucospilos (Peters, 1872)
Yet another species recently transferred from the paraphyletic genus Sphenomorphus on the basis of an extensive phylogenetic analysis and a survey of morphological characters , P. leucospilos (Fig. 67) is a semiaquatic species that has escaped detection in the Philippines since its original description (review : Taylor 1922b;Alcala 1980) until Brown et al. (2000a) discovered a single specimen in Aurora Province in 1997. Since that time we have found numerous populations of this enigmatic species by focusing on riparian streams in intact forest between 200 and 800 m (Brown et al. 2000aMcLeod et al. 2011). Our sighting of this species at 600 m on Mt. Cagua resulted from disturbing the nocturnal resting place of a specimen (under stream side wet leaf litter); following disturbance, the species dove into the nearby running water and escaped capture, a strategy typical for this ecologically unique taxon (RMB, ACD, CDS, personal observation).

Parvoscincus steerei (Stejneger, 1908)
Considered a widespread, variable Philippine endemic, Parvoscincus steerei, like its congener, P. leucospilos, has been transferred from the paraphyletic genus Sphenomorphus on the basis of a robust phylogenetic analysis and a survey of new morphological characters ; an ongoing phylogeographic study, coupled with a review of its conservative morphology may result in taxonomic partitioning once sufficient data have been accumulated. Past studies have noted considerable color pattern and body size variation in this small species of Philippine skink (Taylor 1922b;Brown et al. 1996Brown et al. , 2000aBrown and Alcala 1980;). This species is common across a wide range of habitats and elevational gradients but is most frequently encountered under leaf litter and woody debris in riparian habitats between 300 and 900 m above sea level.

Parvoscincus tagapayo (Brown, McGuire, Ferner & Alcala 1999)
Discovered only in 1999 (Brown et al. , 2000a and originally considered rare and limited to mid-elevation primary forest, this species is now known to occur in a variety of habitat types in Kalinga, Apayao, Isabela, Nueva Viscaya, Aurora, Ilocos Norte , and now Cagayan Provinces. This mid-to high-elevation forest obligate qualifies as "Vulnerable" (VU: B1ab(iii)) because it is known only from a limited (probably less than 20,000 km 2 , but more likely more than 5,000 km 2 ) extent of occurrence and fragmented habitat type, with habitats (low-to mid-montane dry forests) likely to decline in quality and possibly in extent. Future re-evaluations based on field surveys of actual populations will be necessary to confidently establish the conservation status of this species.

Varanus marmoratus (Wiegmann, 1834)
This Luzon faunal region Diesmos 2002, 2009) monitor lizard (Fig.  68) is ubiquitously present in low elevation habitats, including completely denuded coastal areas, agricultural plantations, scrubby vegetation, matrices of secondary growth and primary forest patches, and along forest edges from low-to mid-elevations (Gaulke 1991a(Gaulke , 1991b(Gaulke , 1992a(Gaulke , 1992b. Frequently observed scavenging around human habitats, and heavily disturbed riparian habitats, this species appears to have benefitted from the activities of humans in the northern Sierra Madre. That said, exploitation of this species (for food, leather and pets) is pronounced (Gaulke et al. 1992b(Gaulke et al. , 1998Welton et al. 2012

Varanus bitatawa Welton, Siler, Benett, Diesmos, Duya, Dugay, Rico, van Weerd & Brown 2010
This distinctive species of arboreal, frugivorous, large-bodied monitor lizard (Figs 69,70) was discovered by scientists during the past decade but not described until 2010 (holotype PNM 9719 from Aurora Province); the species was previously well known to Agta tribes peoples (Estioko- Griffin and Griffin, 1975;Griffin and Estioko-Griffin 1985) who consider it a choice delicacy and recognize it with a distinctive local name ("Bitatawa;" Welton et al. 2010). Varanus bitatawa appears to be widespread and common in forested regions of the northern Sierra Madre , extending as far south as the Lingayen-Dingalan geologic fault (Defant et al. 1989;Yumul et al. 2003) and three low-lying, arid river valleys constituting the Mid-Sierra Filter Zone ). This hypothesized barrier divides northern Aurora Province from southern Isabela Province (Fig. 71) and may have served as an ecological or physical barrier to dispersal, possibly promoting divergence between V. bitatawa and its closest relative, V. olivaceus, (Auffenberg 1976(Auffenberg , 1979(Auffenberg , 1988 from Bulacan and Quezon Provinces, Polillo, and Catanduañes islands, and the Bicol faunal region. Our new records from barangays Magrafil and Santa Clara (Figs 69, 70), Municipality of Gonzaga, are the northernmost records for this spe-    cies . Residents and wildlife managers in the vicinity of Gonzaga report that V. bitatawa is a prized target for local consumption in bush meat trade (Fig. 70), and is targeted, in particular, by Agta tribal groups who heavily hunt this species for its meat, preferring it to the more common V. marmoratus . Although a recent survey found conspicuous signs of arboreal monitors in the northern Cordillera Mountains , to date V. bitatawa appears to be restricted to the northern Sierra Madre where it is abundant, frequently encountered by hunters, and heavily hunted for bush meat.

Broghammerus reticulatus (Schneider 1801)
Reticulated pythons (Fig. 72) are common in a wide variety of low-to mid-elevation habitats, including residential areas, agricultural plantations, and the slashand-burn shifting disturbed forests typical of the foothills of major Sierra Madre mountain slopes. Also hunted for meat and leather (Gaulke 1998), this species may require additional measures of protection if commercial exploitation becomes prominent in Isabela and/or Cagayan Provinces. Our specimens were collected in riparian habitats where they were actively hunting on the ground and in low vegetation strata at night.

Ahaetulla prasina preocularis (Taylor 1922)
Widely distributed throughout the Philippines (Leviton 1963a(Leviton , 1967, this species (Figs 73, 74) is most often encountered sleeping on branches of bushes and saplings in secondary growth and selectively logged primary growth forest, on the edges of agricultural plantations, and in shrubs surrounding residential areas. Cagayan

Boiga cynodon (Boie, 1827)
Widely distributed in Southeast Asia (Leviton 1970), B. cynodon (Fig. 75) is highly variable in color pattern, ranging from blond to tan and patternless, to gray with brown and black irregular transverse saddles. We collected two specimens away from water, actively foraging in understory vegetation in a mixture of secondary growth and selectively logged primary growth forest (on the lower slopes of Mt. Cagua); several specimens were captured in low-elevation forest patches at Gattaran in stream-side understory vegetation.

Boiga dendrophila divergens Taylor 1922
One specimen of this widespread endemic Luzon subspecies (Fig. 76) has been collected at the Municipality of Santa Ana (circumstances of collection unknown). This species is undoubtedly widespread throughout low elevation and coastal habitats of northeastern Luzon, and its presence has been confirmed in the Babuyan Islands off the northeast tip of Luzon as well .

Boiga philippina (Peters, 1867)
A single specimen was collected in the Nassiping Forest Reserve where it was active at night in lower branches of streamside understory vegetation. Also now known from the Babuyan Islands (Oliveros et al. 2010), this species (Fig. 77) appears to be a northern Philippine taxon, in accordance with its reported type locality ("northwestern Luzon;" Peters 1867).

Calamaria bitorques Peters 1872
Previously collected in the southern Sierra Madre (Aurora Province; Brown et al. 2000a;, this species (Fig. 78) is infrequently encountered but can be distinguished from C. gervaisi on the basis of color pattern and larger maximum body size (Inger and Marx 1965).

Coelognathus erythrurus manillensis (Jan, 1863)
We encountered this Luzon endemic (Leviton 1979;Fig. 80) actively foraging (in the late morning) on the ground in dry forest at low elevations in Cagayan Province in the Nassiping Forest Reserve.

Cyclocorus lineatus lineatus (Reinhardt, 1843)
One of only four snake genera endemic to the Philippines Leviton 1963a), Luzon populations of Cyclocorus lineatus (Leviton 1965a ; Fig. 81) have commonly been encountered by us under the cover of rocks, loose soil, logs, and other debris along banks of streams and rivers (personal observations). An additional specimen was collected from within soft, dry rot decaying wood matter in a large stump on the bank of a stream at 600 m on Mt. Cagua. Cagayan

Dendrelaphis luzonensis Leviton 1961
Luzon populations of Dendrelaphis luzonensis (Leviton 1961(Leviton , 1968van Rooijen and Vogel 2012;Fig. 82) are most often encountered asleep in shrubbery and understory vegetation, ferns and palms, surrounding the banks of streams and rivers, especially at low elevations in and around agricultural areas and on forest edges. We collected a specimen on the lower slopes of Mt. Cagua sleeping in vines clinging to a large tree trunk in selectively logged forest away from water.

Dendrelaphis marenae Vogel & van Rooijen, 2008
Common in residential and agricultural areas where they are most often seen active during the day on the ground or sleeping in bushes at night, this common species is widely distributed throughout the northern Philippines (Leviton 1968). Dendrelaphis marenae (Fig. 83) has recently been morphologically distinguished and diagnosed as a species distinct from the Indochinese D. pictus, and the eastern Indonesian D. grismeri (Vogel and van Rooijen 2008).

Dryophiops philippina Boulenger, 1896
This widespread Philippine endemic (Fig. 84) has been collected in recent years with increasing frequency as workers target the remaining low-elevation and coastal forests of the Philippines. We observed one specimen in ultrabasic forests near the municipality of Palanan.

Gonyosoma oxycephalum (Boie, 1827)
This widespread, non-endemic, Southeast Asian rat snake (Fig. 85) has been documented throughout the Philippines in a wide variety of habitats. Our records originated in forested areas at low elevations in Binatug and Palanan.

Hologerrhum philippinum Günther, 1858
Seldom encountered, the genus Hologerrhum is one of only four snake genera endemic to the Philippines Leviton 1963aLeviton , 1983. Collected only a few times in the past two decades (Brown et al. 1996McLeod et al. 2011;Phenix et al. 2011), this species has been encountered in recent years in dry forest among bamboo stands (ACD, personal observation) and multiple times in dry streambeds under rocks. Our specimen from Mt. Cagua (Figs 86, 87) was collected in a dry ravine, under a small rock, at 600 m in selective logged primary forest. In contrast to other recent specimens with salmon-red or pinkish red ventral surfaces, our Mt. Cagua specimen had a bright yellow venter. The only other species in the genus (H. dermali from Panay Island; Ferner et al. 2001;Gaulke 2001) also has a bright yellow venter but differs from H. philippinum by the presence of a mid-ventral black stripe .

Lycodon capucinus (Boie, 1827)
Widespread and common throughout the Philippines and Southeast Asia (Leviton 1963a(Leviton , 1965bManthey and Grossman 1997;Inger and Voris 2001), L. capucinus (Fig.  88) is a frequently encountered snake in residential and agricultural areas at low elevations. We collected one road kill specimen on the road to Barangay Magrafil along the north coast Luzon highway in the Municipality of Gonzaga. The recent phylogenetic study of Siler et al. (2013) revealed moderate levels of genetic diversity among sampled populations of L. capucinus across its recognized range in Southeast Asia. The species relationship with the morphologically similar species L. aulicus (Linnaeus 1758) has long been controversial (review: Siler et al. in press b), and future studies focused on the L. aulicus and L. capucinus will be needed to fully resolve species boundaries within this widespread species complex.

Lycodon muelleri Duméril, Bibron & Duméril, 1854
Lycodon muelleri (Fig 89) has been recorded throughout Luzon (Leviton 1965b;Brown et al. 2000a;) and is frequently collected in low vegetation along streams and rivers at low elevations (<500 m). Our Isabela Province specimen was collected in the Municipality of San Mariano, at night, on the buttress of a large tree adjacent to a mountain stream. In a recent phylogenetic study of Southeast Asian wolf snakes, a deep genetic divergence was observed between populations of L. muelleri from northern and central Luzon and populations sampled on the Bicol Peninsula in southeast Luzon (Siler et al. 2013).

Lycodon solivagus Ota and Ross, 1984
A single specimen of this species was collected close to the type locality (Ota and Ross 1994) where it was found dead on the road between Barrio Battalan and Lasam Centro (Municipality of Lasam); this damaged specimen was identified by its distinctive dentition. The only other specimen ever collected is a single individual from Barangay Paitan, Municipality of Quezon, Nueva Vizcaya Province (KU 325974). Unfortunately, tissue samples of this species presently are not available, and how this species is related to the many other Philippine endemic wolf snakes remains undetermined (Siler et al. 2013).
Oligodon ancorus (Girard, 1858) Leviton (1962) listed sites for this species throughout much of Luzon. The one specimen recently collected in Cagayan province was encountered as it rested on top of a rock by a small stream.

Psammodynastes pulverulentus (Boie, 1827)
This common, widespread species (Manthey and Grossman 1997;Inger and Voris 2001;Fig. 90) has been documented throughout Luzon (Leviton 1983;Brown et al. 2000aMcLeod et al. 2011); one of our specimens was collected on the ground on a stream bank at night and the other was encountered asleep among the lower branches of a small bush on a river bank.
Pseudorhabdion cf. mcnamarae (Taylor 1917) Pseudorhabdion mcnamarae was originally described from northern Negros (Visayan island group; Taylor 1917Taylor , 1922aLeviton and Brown 1959), but at least one specimen from Luzon has ostensibly been referred to this species (CAS 61544, from Balbalan, Kalinga Province). Although we suspect that Pseudorhabdion mcnamarae is actually restricted  to the Visayan faunal region Gaulke 2001;Ferner et al. 2001) and that the Luzon population (Fig 91) represents another, possibly undescribed, species, we do not recommend taxonomic action until a detailed study of this group can be undertaken.

Pseudorhabdion cf. talonuran Brown, Leviton & Sison, 1999
Another specimen of a distinctive species of Pseudorhabion has been collected in the Dibanti River basin, Barangay Dibuluan (Municipality of San Mariano). This specimen most closely resembles Pseudorhabdion talonuran, a high-elevation species from Panay Island Ferner et al. 2001), suggesting a biogeographically improbable, disjunct distribution and the possibility that this single specimen may constitute the first record of a new species from the northern Sierra Madre.

Ptyas luzonensis (Günther, 1873)
Now considered common and widespread throughout the Luzon and Visayan faunal regions (Leviton 1983;Ross et al. 1987), this species (Fig. 92) has been documented throughout Luzon at a variety of forested sites McLeod et al. 2011). Whereas most recent specimens have been encountered at night, asleep in branches of understory trees, along the banks of streams in selectively logged primary and secondary growth forests McLeod et al. 2011), the one Cagayan Province record was collected on the ground where it was actively hunting at night along a dry ridge.

Rhabdophis spilogaster (Boie, 1827)
Rhabdophis spilogaster (Fig. 93) is diurnally active in riparian habitats at low-to mid-elevations throughout Luzon. Specimens were collected in artificial fish ponds in residential areas, swimming in side pools of small streams, along the borders of flooded rice fields, and among rocks on stream banks in regenerating secondary growth forest.

Tropidonophis dendrophiops (Günther, 1883)
This moderately common natricine snake (Fig. 94) is an ecological generalist that is frequently encountered in riparian habitats (Leviton 1963a;Malnate and Underwood 1988). Our specimens were collected during mid-to late morning when they were active in nearly dry streambeds; specimens were first observed actively crawling among rocks and other debris.

Hemibungarus calligaster calligaster (Wiegmann, 1835)
A single specimen of this widespread coral snake (Leviton 1963b;Brown 2006;Siler and Welton 2010;Fig. 95) was collected in a forest fragment in Santa Ana. It is widespread throughout the eastern seaboard of Luzon but has not yet been documented in the Cordillera (Diesmos et al. 2005;.

Naja philippinensis Taylor 1922
A single specimen of the distinctive Philippine cobra (Leviton et al. 1964b) was collected at Barangay San Pedro, Municipality of Lasam. Recent encounters with this species in Aurora Province  plus historical records (Leviton et al. 1964b) suggest that it is widespread and common throughout the Sierra Madre. Persecution and exploitation of this species have been identified as potential conservation threats (Gaulke 1998. Circumstances of capture of the single documented Cagayan Province specimen are unclear; the animal was most likely captured by residents in agricultural areas surrounding human settlements (personal observations).

Ophiophagus hannah (Cantor 1936)
Residents of San Mariano and Palanan related to us numerous instances of sightings and resident killings of very large, light tan-colored cobras in the vicinity of settlements and agricultural areas. This species has been reported widely on Luzon (Leviton 1964b;McLeod et al. 2011;Devan-Song and Brown 2012) and is known to the residents of Isabela as well. One of us (ACD) sighted an additional king cobra in a forest fragment at Apaya (Location 23). Thus, our own records plus resident reports provide, in our opinion, sufficient credibility to include O. hannah in this report, although we consider these records unconfirmed until voucher specimens become available.

Family Homalopsidae
Cerberus schneideri (Schlegel, 1837) Dog-faced water snakes (Fig. 96) are distributed in coastal areas throughout much of Southeast Asia (Gyi 1970). In a recent comprehensive systematic review, Murphy et al. (2012) identified species-level diagnostic differences corresponding to previously identified phylogenetic breaks (Alfaro et al. 2004(Alfaro et al. , 2008, necessitating the elevation of a formerly synonymized name (C. schneideri) to accommodatethe distinctive lineage distributed throughout the coasts of Malaysia, Indonesia and the Philippines. The taxon C. rynchops is now restricted to the coasts of Thailand, Myannmar, the Indian subcontinent, and Sri Lanka. Cerberus schneideri has been documented on most major islands of the Philippines and our specimen originated in the Municipality of Santa Ana, along the northeast coast of Luzon.

Oxyrhabdium leporinum leporinum (Günther, 1858)
Adults of this common Luzon faunal region endemic (Fig. 97) are frequently encountered actively foraging along stream banks in forests of varying levels of disturbance (Leviton 1964c;Brown et al. 2000aDiesmos et al. 2005;; juveniles are most frequently found at night, sleeping perched in herbaceous layer vegetation, ferns, and small shrubs in riparian habitats (McLeod et al. 2011). Our specimens were found coiled on axils of ferns along a stream bank in selectively logged forest at 600+ masl.

Ramphotyphlops braminus (Daudin, 1803)
Specimens of this common, parthenogenetic, presumably introduced species were collected under rocks, palm fronds, and other debris on the edge of forests, and within selectively logged forests at low-to mid-elevations.

Typhlops ruficaudus (Gray, 1845)
A single specimen of this species was collected in a fern axil in a small primary growth forest fragment. As the taxonomy of Philippine typhlopids has improved in the past decade (McDowell 1974;Wynn and Leviton 1993;McDiarmid et al. 1999) this species has emerged as a moderately common component of Luzon's herpetofauna (McLeod et al. 2011).

Typhlops sp. 1
This distinctive, probable new species appears related to the T. ruficaudus Group (Wynn and Leviton 1993), but differs from other members of that group on the basis of several characters of scalation and body size.

Typhlops sp. 2
This distinctive, probable new species appears phenotypically most similar to T. luzonensis, but differs from that species on the basis of several characters of scalation and body size.

Family Viperidae
Trimeresurus flavomaculatus (Gray, 1842) Exceedingly common on the lower slopes of Mt. Cagua, T. flavomaculatus (Fig. 98) was observed in high densities surrounding ephemeral pools during the start of the rainy season at mid-elevations (400-700 m). We encountered multiple individuals per night at the same temporary pond as they actively hunted frogs (Occidozyga laevis, Polypedates leucomystax, Rhacophorus pardalis, Kaloula rigida and K. picta) in the lower strata (30-100 cm above the forest floor) of shrub layer vegetation or in temporary pools along muddy paths in secondary forest. Trimereserus flavomaculatus is widespread and common throughout the Luzon faunal region (Leviton 1964a;Brown et al. 1996Brown et al. , 2000aDiesmos et al. 2005;McLeod et al. 2011;Devan-Song and Brown 2012

Tropidolaemus subannulatus (Gray, 1842)
This common and widespread Luzon and Visayan region pit viper is frequently encountered in forested areas from sea level to mid-montane elevations. Although no specimens are available, it has been photographed in the Dibanti River area, Municipality of San Mariano.

Family Trionychidae
Pelochelys cantorii Gray, 1864 Diesmos et al. (2008) documented the presence of this species (Figs 100, 101) at several sites in Cagayan and Isabela Province. The taxonomic status of Philippine populations of Pelochelys cantorii requires re-evaluation. As currently understood, Pelochelys cantorii occurs from southern India through Bangladesh, southern China, Myanmar, Vietnam, Cambodia, Laos, Thailand, Malaysia, and southern Borneo, Indonesia (Lim and Das 1999;Ernst et al. 2000;Webb 1995Webb , 2002Stuart and Platt 2004;Fritz and Havas 2007). We consider this widespread distribution unlikely for a single species. Morphological variation in skull and carapace morphology (neural bone counts) has been reported between Philippine and mainland Asian populations of P. cantorii (Baur, 1891;Taylor 1920bTaylor , 1921, suggesting that Philippine populations may be distinct. If so, Gray's (1864) epithet (Pelochelys cumingii) would be the appropriate name for this possible Philippine endemic.

Caretta caretta (Linnaaeus, 1758)
Loggerhead Turtles have been observed in coastal waters of Isabela Province (van Lavieren et al. 2009) but studies of their habitats and natural history along the east coast of Luzon are lacking.

Chelonia mydas (Linnaaeus, 1758)
Green Turtles have been observed nesting on the beaches of the coast of Isabela Province (MVW, personal observations); we expect the vast, undeveloped coastline of Cagayan Province also supports nesting populations to the north.

Eretmochelys imbricata (Linnaeus, 1766)
Hawksbill Turtles nest on the beaches of the coast of Isabela Province (MVW, personal observations) and we expect this species also nests along the beaches of Cagayan Province to the north.

Crocodylus mindorensis Schmidt, 1935
The target of an extensive local conservation program (van Weerd and van der Ploeg 2004a,b; van der Ploeg and van Weerd 2004;van Weerd 2010;van der Ploeg et al. 2011avan der Ploeg et al. , 2011bvan der Ploeg et al. , 2011c, Crocodylus mindorensis (Figs 102, 103) is now known from the municipalities of Maconacon, Divilacan, Palanan, San Mariano, Benito Soliven and San Guillermo in Isabela Province. Tolerant of substantial habitat disturbance and capable of living in close proximity with humans, remaining populations of this species are most severely threatened by continued habitat loss and degradation, and persecution by people. The USNM specimens were purchased from a dealer and reportedly came from the northern part of the Cagayan Valley (exact locality unknown). Recent records include field observations of live Crocodylus mindorensis animals in riverine and coastal habitats, including under saline conditions, in the municipalities of Maconacon, Divilacan and Palanan along the Pacific coast of Isabela and in inland freshwater rivers, creeks and small ponds in San Mariano, Benito Soliven and San Guillermo in Isabela Province. There are, as yet, unconfirmed reports of C. mindorensis in the municipality of Baggao in Cagayan, and the species occurred with certainty until recently in small lakes and rivers in Peñablanca, also in Cagayan Province. There are, furthermore, unconfirmed reports of C. mindorensis from the tributaries to the Cagayan River that originate in the Cordillera Mountains along the western edge of Cagayan Valley, in both Isabela and Cagayan Provinces.

Crocodylus porosus Schneider, 1801
The large-bodied saltwater crocodile Crocodylus porosus is distributed in estuarine and coastal areas, swamps and rivers, from India, throughout Southeast Asia, eastward to Australia (Iskandar 2000). Recent records include documented field observations of Crocodylus porosus in coastal areas in the Municipalities of Maconacon and Palanan, Isabela Province (Fig. 104)

Discussion
Recent studies focusing on the major subcenters of herpetological endemism of the Luzon faunal region Diesmos 2002 2009) have documented between 52 to 85 species for a given area (Ross and Gonzales 1992;Brown et al. 1996Brown et al. , 2000aunpublished data;Diesmos et al. 2005;Oliveros et al. 2010;Siler et al. 2011a;McLeod et al. 2011;Devan-Song and Brown 2012). As reviewed by Auffenberg (1988; see also Hall, 1998Hall, , 2001Hall, , 2002Yumul et al. 2003Yumul et al. , 2009, the occurrence of multiple geological components of central and southern Luzon correspond in some cases to approximate paleoisland precursors that accreted in recent geological history (mid-to late Cenozoic) to form today's geographically complex, tectonically active southern portion of Luzon (Hall 2002;Yumul et al. 2003). In contrast, northern Luzon, at the northern end of the Philippine Sea Plate and contained within east-dipping Manila Trench and the west-dipping proto-East Luzon Trough (Florendo 1994;Hall 2002;Yumul et al. 2003) is characterized by two main montane components: the Cordillera and the Sierra Madre. Because of the intriguing possibility of a history of isolation on past paleoisland microterranes or ancient montane formations, biogeographers have often noted the occurrence of regional faunas that appear to coincide with the major geological components of the island (Auffenberg 1998;Brown et al. 1996;Diesmos et al. 2005;Siler et al. 2011a;Welton et al. 2010Welton et al. , 2012. Thus, our current expectation is that some degree of local faunal endemism should be discernable and uniquely associated with the Cordilleras, the Sierra Madre, the Zambales, the volcanoes of southern Luzon, and the Bicol Peninsula (and associated islands of Polillo and Catanduanes; Brown et al. 1996Brown et al. , 2000aRoss and Gonzales 1992;Diesmos et al. 2005;Devan-Song and Brown 2012).
This study further supports our general expectation of within-Luzon biogeographic provincialism in the sense that the northern Sierra Madre, like the other major components of Luzon, contains a high percentage of locally endemic amphibians and reptiles. At the same time, this study confirms recent and unexpected findings of a considerable degree of overlap in faunal elements between the northern Cordilleras and the northern Sierra Madre (Diesmos et al. 2005;. In this study we have made a preliminary, first-pass, enumeration of the amphibians and reptiles of the northern Sierra Madre and documented species diversity of this region at more than 100 species. This level of diversity surpasses all but one faunistic study for Luzon (Siler et al. 2011a) and rivals other regional high-diversity areas throughout the archipelago. Only the cumulative results of the central Mindanao surveys of Taylor (based on several years of work in Bunawan, Agusan del Norte, Mindanao Island; Taylor 1920aTaylor , b, 1921Taylor , 1922aTaylor , b, c, d, 1923, the eastern Mindanao surveys of Brown and Alcala (unpublished data, available at California Academy of Sciences Herpetological database web portal), and new, repeated western Mindanao surveys of Brown et al. (unpublished data, available at the University of Kansas Specify database web portal) include species diversity estimates upwards of 80-90 species for a given site.
In a 10-day preliminary survey at Aurora National Park, Brown et al. (2000a), used species accumulation curves to demonstrate that even substantial sampling efforts (involving well trained field teams with 4-6 workers) will underestimate diversity-as measured by species accumulation curves that failed to asymptote during their intensive, but brief, field survey effort. In a more recent study a decade later, Siler et al. (2011a) increased Aurora diversity to 85 species by visiting different drainages in the same park, focusing on different habitat types, sampling a wider range of elevations, and concentrating survey efforts across a broader range of atmospheric conditions. The lessons from the combinations of these two studies, as well as two studies from the northern Cordilleras (Diesmos et al. 2005;) are very clear-and provide important "best practice" guidelines for future biodiversity inventories and conservation efforts in this and other similar archipelagos.
We concentrate on these two pairs of studies because they are the only two of their kind for Luzon: the northern Cordillera surveys (Diesmos et al. 2005;) and the repeat assessments at Aurora National Park (Brown et al. 2000a;Siler et al. 2011a). Together, these studies indicate that no site can be reasonably characterized for resident biodiversity with a single visit, even if a considerable effort is exerted over a multi-week field expedition (typical of the last several decades of field surveys throughout the Philippines; personal communications with A. C. Alcala, D. S. Rabor, L. A. Heaney, R. S. Kennedy, P. C. Gonzales, and colleagues). Because of annual herpetological species activity patterns associated with seasonality and the reproductive effort (Taylor 1917;Brown and Alcala 1963, 1981, 1986Auffenberg 1988, 1989;review: Brown et al. 2002;Gaulke 2011), it is imperative that expeditionary inventory fieldwork focuses on a given area for minimum of one survey in the dry season and a follow-up effort in the rainy ("monsoon") season. Our recent efforts have attempted to maximize atmospheric variation (principally, variation in occurrence and severity of precipitation events) by concentrating survey work at the beginning of the rainy season (June-August) in hopes of sampling the same sites when dry and, subsequently, following the first heavy rains. However, even these sequential efforts targeting major shifts in habitat and climactic variables fail to capture annual variation in the reproductive effort (and thus, detectability of a given species) in the majority of species present (Alcala, 1967;Brown 1967, 1982;Alcala 1970b, 1982;Alcala and Alcala 1980;Auffenberg, 1988;Auffenberg 1988, 1989;Gaulke 2011).
A second major lesson involves the naturally patchy distribution of many amphibian and reptile species (Brown and Alcala 1963, 1981, 1986. The experience drawn from the Siler et al. (2011a) follow up to the Brown et al. (2000a) Aurora surveys indicate the need for concentrated survey efforts across many sub-sites (habitat types) within a general area. The Siler et al. (2011a) effort nearly doubled the known herpetological diversity of Aurora National Park by focusing on south versus north facing slopes, different streams and river drainages, geological variables (karst versus volcanic soils), variation in elevational gradients, and different forest types/plant communities. Most of Siler et al.'s (2011a) sampling locations were within kilometers of Brown et al.'s (2000a) work; the critical difference, in our opinion, was the simple fact that Siler et al. detected additional species by sampling a broader range of habitat heterogeneity and climactic variability. Because of our collective experience with the last 15 years of survey work on Luzon, we are now compelled to be extremely cautious about generalizing conclusions regarding patterns of biodiversity, abundance, biogeography, endemism, and especially conservation status from any single-visit faunal inventories. As  emphasized in a recent study of Ilocos Norte Province (northwest Luzon), arid conditions associated with the dry season render conclusions about amphibian communities derived from surveys during these times moot for the fundamental reason that negative data are uninformative for the purposes of determining species presences/absences. The same is clearly true for any single-visit survey to a given area: the apparent absence of a species (non-detection or negative data) during a survey actually tells us nothing about the abundance of that taxon, its distribution or, most significantly, its conservation status .
With these caveats in mind, diversity patterns for the reasonably well surveyed areas of Luzon include 52-55 species for the Northern Cordillera (Diesmos et al. 2005;, 63 species for Bulacan Province, southern Sierra Madre (McLeod et al. 2011), 52-60 species from the Zambales Mountains (Brown et al. 1996;Devan-Song and Brown 2012), 52 species from the Babuyan Island Group, north of Luzon , 58 species for Catanduanes Island (Ross and Gonzales, 1992), 52-56 species from the coastal forests of Subic Bay (Devan-Song and Brown 2012), and 85 species from the central Sierra Madre (Brown et al. 2000a;Siler et al. 2011a). A number of species are conspicuously absent from our total estimates; many of these being common throughout coastal areas of Luzon (Inger 1954;Leviton 1962Leviton , 1963aLeviton , b, 1964aLeviton , b, 1965aLeviton , b, 1967Leviton , 1970aBrown and Alcala 1974;), but we predict that these species will be recorded during future surveys in the northern Sierra Madre. These species include widespread endemic amphibians like Rhacophorus bimaculatus, the introduced species Hylarana erythraea (Diesmos et al. 2006), and several undescribed Platymantis species now known from Aurora Province to the south of Isabela (Brown et al. 2000a;Siler et al. 2011a). Lizard species that we expect will eventually be recorded from the northern Sierra Madre include Gonocephalus sophiae, Hydrosaurus pustulatus, Luperosaurus angliit , Brachymeles boulengeri, B. elerae , Dasia grisea, Emoia atrocostata, Eutropis bontocensis, Parvoscincus luzonensis, P. lawtoni, and P. igorotorum . Snakes that we expect to be residents in the northern Sierra Madre include Boiga angulata, Chrysopelea paradisi, Pseudorhabdium oxycephalum, Rhabdophis baurbori, Myersophis alpestris Leviton 1983), Acutotyphlops banaorum (Wallach et al. 2007), Typhlops luzonensis, T. ruber, and T. cumingi (McDowell 1974;McDiarmid et al. 1999). Additionally, we anticipate that the introduced freshwater turtles Pelodiscus sinensis and Chrysemys picta ) will eventually be encountered in the northern Sierra Madre. Coastal areas (with their potential habitats for sea snakes, marine turtles, rare forest geckos, and selected scincid lizards; Alcala, 1986;W. Brown et al. 1978W. Brown et al. , 1980R. Brown et al. 2007R. Brown et al. , 2011 are likely to support additional species diversity. At present the Luzon faunal region's herpetological diversity stands at more than 150 species. A total of 49 total amphibian species have been documented, 44 of which are native (5 introduced; Diesmos et al. 2006), and 32 of which are endemic Alcala 1970a, 1994;Brown 2007;Diesmos and Brown 2011). Luzon supports at least 106 native reptiles, 76 of which are endemic to this faunal region (Leviton 1963a;Alcala 1978, 1980;. If the percentage of species associated with unresolved taxonomic problems identified here (~38%) can be extrapolated to the total fauna, as many as 20 of Luzon's current amphibian species and as many as 40 current reptile taxa may be associated with future taxonomic changes. The majority of these will most likely involve partitioning of species complexes into two or more distinct evolutionary units (Ron and Brown 2008;Bain et al. 2008;Brown and Stuart 2012).
In addition to this expected increase in biodiversity associated with refined taxonomic partitioning (e.g., splitting) of species groups, the northern Philippines has been the focus of the majority of de novo new species discovery in recent decades (W. Brown et al. 1997a, b, c;1999a, b;Brown 1998, 1999;R. Brown et al. 1999R. Brown et al. , 2000bR. Brown et al. , 2007R. Brown et al. , 2011Siler and Brown 2010;Welton et al. 2010;Diesmos and Brown 2011). Our estimates place the remaining known new taxa (i.e., already in collections, represented by specimens clearly identified as new species) awaiting description in the Luzon faunal region at approximately 25 amphibian and 15 reptile species (RMB, A. C. Alcala, ACD, and CDS, unpublished data). When the two sources of undescribed herpetological diversity for Luzon are combined, a striking potentiality for unknown diversity emerges: we anticipate that the diversity of the island may grow to as many as 90-100 (70-80% endemic) amphibian species and as many as 150-160 reptile species with ongoing biodiversity studies in the near future. Clearly the herpetological biodiversity of the northern Philippines is substantially underestimated.
The results of this and related studies from the northern Philippines contribute to an ongoing revision of the biogeographic characterization of the Philippines as a "fringing archipelago" with a depauperate fauna in its northern regions (Dickerson 1928;Taylor 1928;Darlington 1957;Myers 1960Myers , 1962Carlquist 1965;Leviton 1963a;Brown and Alcala 1970a). As more recent studies have focused on in situ diversification within the archipelago (Heaney 2000;Brown and Guttman 2002;Evans et al. 2003;Heaney et al. 2005;Jansa et al. 2006;Siler et al. 2010cSiler et al. , 2011bSiler and Brown 2011;), the northern Philippines has emerged as a major regional hotspot for the autochthonous production of vertebrate biodiversity via a variety of evolutionary processes of diversification .
Conservation of Luzon's vertebrate biodiversity-in particular the more spectacular Philippine evolutionary radiations and complex ecological communities supported by the remaining forested areas of Luzon-remains an on-going effort, challenged by rapid development, large-scale extractive logging and mining industries and conversion of natural habitats into agricultural lands driven by a burgeoning human population (Liu et al. 1993;Uitamo 1999;van der Ploeg et al. 2011d). A suite of recent studies has shown that some forested regions closest to the country's large major metropolitan areas remain among the least studied of Luzon's forests Diesmos and Brown 2011;McLeod et al. 2011;Devan-Song and Brown 2012); these areas are immediate priorities for comprehensive faunal surveys of the type presented here. In contrast, forested areas that have been properly surveyed for herpetological diversity rank among the areas supporting the country's most diverse herpetological communities (Brown et al. 2000aSiler et al. 2011a). Before reasonably well-informed, biologically meaningful conservation measures are to be effective, a basic understanding of distribution patterns and cross-taxon congruence of Luzon's vertebrate biodiversity will be necessary van Weerd and Udo de Haes 2010;Diesmos and Brown 2011). In the absence of actual, field-based, empirical, survey data, conservation status assessments (IUCN 2010) and priority setting exercises Diesmos and Brown 2011) will remain incomplete, uninformed, and overly reliant on secondary sources, extrapolation, and "expert" opinion (Diesmos et al. 2005;). In the northern Sierra Madre where, despite the fact that large areas are protected on paper, threats to the remaining large tracts of forested areas have been clearly identified (NORDECO 1988;Hicks 2000;van Weerd and Udo de Haes 2010;van der Ploeg et al. 2011d;Minter et al. 2012), and a major challenge will be to monitor herpetological communities through time in order to assess communities' responses to land use changes, climate change, resource extraction, introduced species, emerging infectious disease, and habitat degradation. With the initial baseline information provided here, tremendous opportunities exist for future studies in taxonomy, biogeography, ecology and conservation of northern Luzon's amphibians and reptiles.