590.5 Fl N.S. no. 112(2006) oology NO. 112 IANA The Mammals of Mt. Kitanglad Nature Park, Mindanao Island, Philippines Lawrence R. Heaney Bias R. Tabaranza, Jr. Eric A. Rickart Danilo S. Balete Nina R. Ingle December 13,2006 Publication 1544 IRARY HAU PUBLISHEDBY FIELD MUSEUM OF NATURAL HISTORY FIELDIANA Zoology NEW SERIES, NO. 112 The Mammals of Mt. Kitanglad Nature Park, Mindanao, Philippines Lawrence R. Heaney Department of Zoology Field Museum of Natural History 1400 S Lake Shore Drive Chicago, Illinois 60605 USA Bias R. Tabaranza, Jr. Department of Biology Iligan Institute of Technology, Mindanao State University, Iligan City, Lanao del Norte, Philippines Current address: Haribon Foundation 2F Santos & Sons Bldg, 973 Aurora Blvd. Cuhao, Quezon City, Philippines Eric A. Rickart Utah Museum of Natural History University of Utah 1390 East Presidents Circle Salt Lake City, Utah 84112 USA Danilo S. Balete Department of Zoology Field Museum of Natural History Chicago, Illinois 60605 USA Current address: Laksambuhay Conservation, Inc. 10241 Mt. Bulusan St., U-2 Los Banos, Laguna, Philippines Nina R. Ingle Department of Natural Resources Cornell University Ithaca, NY 14853 USA Current address: 309 6th A St. Ecoland 8021 Davao City, Philippines Accepted July 28, 2006 Published December 13, 2006 Publication 1544 PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY BIOLOGY LIBRARY 101 BURRILL HALL FFR n 7 ?nn7 2006 Field Museum of Natural History ISSN 0015-0754 PRINTED IN THE UNITED STATES OF AMERICA pi Table of Contents Abstract 1 Introduction 1 Study Area and Methods 3 Geology of the Kitanglad Range 3 Climate of the Kitanglad Range 4 Methods 5 Study Sites 7 Field Work by Previous Investigators 13 Accounts of Species 14 Order Erinaceomorpha 14 Order Soricomorpha 17 Order Scandentia 20 Order Dermoptera 20 Order Chiroptera 21 Order Primates 38 Order Rodentia 38 Order Carnivora 49 Order Artiodactyla 50 Analysis and Discussion 51 Adequacy of Sampling 51 Patterns of Species Richness of Small Fruit Bats 52 Patterns of Species Richness of Non-volant Small Mammals 53 Mammalian Biogeography on Mindanao 54 Distribution and Ecology of Non-Native Species 55 Conservation and Management of the Kitanglad Mammal Fauna 55 Conclusions 58 Acknowledgments 58 Literature Cited 59 List of Figures 1 . Photograph of the Kitanglad Range taken from above Silipon, below Sites 5 and 6 3 2. Map of Mindanao Island and adjacent areas 4 3. Map of topographic features and collecting sites on the Kitanglad Range 6 4. Photograph of lowland rain forest at 1 100 m (Site 2) 9 5. Photograph of lower montane rain forest at 1600 m (Site 3) 10 6. Photograph of lower mossy forest at 2250 m (Site 6) 11 7. Photograph of upper mossy forest at 2800 m (Site 9) 12 8. Elevational ranges of small mammals from Mindanao 15 9. Elevational ranges of fruit bats from Mindanao 23 10. Species accumulation curves for small fruit bats 51 1 1. Species accumulation curves from small-mammal surveys 52 12. Number of species of small fruit bats along four elevational transects 53 13. Relative abundance of small fruit bats along four elevational transects 53 14. Number of species of small mammals along two elevational transects 54 15. Relative abundance of small mammals along two elevational transects 54 16. Change in forest cover on Mindanao Island. 1880-1988 56 List of Tables 1. Temperatures recorded at primary sites in 1992-1993 5 2. Numbers of small non-volant mammals captured at each site 16 3. Diel activity patterns of small non-volant mammals 17 4. Bait attractiveness to small non-volant mammals 17 5. Stomach contents of small non-volant mammals 18 6. External and cranial measurements of Erinaceidae and Tupaiidae 19 7. External and cranial measurements of Soricidae 19 8. Number of small fruit bats captured at each site 22 9. External and cranial measurements of fruit bats (Pteropodidae) 24 10. Number of insectivorous bats captured at each site 30 11. External and cranial measurements of Rhinolophidae 32 12. External and cranial measurements of Vespertilionidae and Molossidae 35 13. External and cranial measurements of rodents (Sciuridae and Muridae, Apomys through Crunomys) 39 14. External and cranial measurements of rodents (Muridae. Limnomys through Tarsomys) 45 The Mammals of Mt. Kitanglad Nature Park, Mindanao, Philippines Lawrence R. Heaney, Bias R. Tabaranza, Jr., Eric A. Rickart, Danilo S. Balete, and Nina R. Ingle Abstract Field surveys within and adjacent to the Mt. Kitanglad Nature Park in the Kitanglad Range of Bukidnon Province, north-central Mindanao, from 1992 to 1999, along with examination of previously existing specimens, have allowed us to document the local presence of 58 species of mammals, 53 native and five non-native. These include one gymnure (Erinaceidae), two shrews (Soricidae), one tree shrew (Tupaiidae), one flying lemur (Cynocephalidae), 14 fruit bats (Pteropodidae), eight roundleaf and horseshoe bats (Rhinolophidae), nine evening bats (Vespertilionidae), one mastiff bat (Molossidae), two primates (Tarsiidae and Cercopithecidae), three squirrels (Sciuridae), 14 mice and rats (Muridae), two civets (Viverridae), one pig (Suidae), and one deer (Cervidae). Mt. Kitanglad Nature Park has one of the most diverse mammal faunas in the Philippines, exceeding that of the more widely known Mt. Apo. Three species, a bat (Alionycteris paucidentata) and two native mice {Crunomys suncoides and Limnomys bryophdus), are currently known only from high elevations in the Kitanglad Range. Species richness of bats declined with increasing elevation, but richness of non-volant small mammals increased five-fold from lowlands to a peak at ca. 2250 m, and then declined with further increases in elevation. We found distinctive mammal communities in lowland rainforest (up to about 1200 m elevation), montane rainforest (ca. 1200 m to 1900 m), and mossy rainforest (2000 m to the peak at 2950 m). We conclude that all three rainforest types, at all elevations, are important to the success of the park as a biological reserve. Over-hunting of large mammals and illegal logging both pose serious problems. Lowland rainforest has been removed on much of Mindanao, including the vicinity of the park, and thus is the habitat type that is currently most threatened. Habitat destruction, especially of lowland rainforest, threatens the mammals in the Kitanglad Range as well as the economic and social stability of the human population of northern Mindanao. Introduction With at least 180 species of native mammals, of which minimally 115 are endemic, the Philippine Islands have one of the highest levels of endemism of any country (Balete et al., in press; Heaney et al., 1998; Rickart et al, 2002, 2003, 2005). Those countries with more endemic species are much larger; based on relative geo- graphic area, the Philippines has one of the greatest concentrations of unique species, ex- ceeding Brazil, Madagascar, and the nations of East Africa (Heaney & Regalado, 1998; Mitter- meier et al., 1997, 2004; Ong et al., 2002). This great diversity of mammals is associated strongly with both the tropical climate and the complex geological history of the Philippines. Most of its islands are oceanic in origin, and have never been connected either to the Asian mainland or to each other (Hall, 1998, 2002; Heaney, 1985, 1986, 2004; Steppan et al., 2003), and form a mosaic of unique centers of biodiversity. Thus, FIELDIANA: ZOOLOGY, N.S., NO. 112, DECEMBER 13, 2006, PP. 1-63 on each of the larger islands, at least 50% of the non-flying mammals occur nowhere else. Mind- anao, the subject of this monograph, is the second largest island in the Philippines at 94,875 km 2 . During periods of low sea level during the Pleistocene, Mindanao was connected to Leyte, Samar, Bohol, and nearby smaller islands, and thus was even larger than today, but this "ice-age island" of Greater Mindanao was never connected to any larger land mass. Of the 27 currently known species of non-flying mam- mals that occur naturally on Greater Mindanao, 22, or 81.5%, live nowhere else (Heaney, 1986, 2000, 2004; Heaney & Regalado, 1998; Ong et al., 2002). Within Greater Mindanao, many mammal species are widespread, especially those that live in lowland forest, but several subcenters of endemism can also be identified. At least one species of mammal (an as-yet undescribed species of forest mouse, Apomys; Steppan et al., 2003) occurs only on Bohol, Leyte, and probably Samar and associated smaller islands. Three species of small mammals are restricted to Dinagat and adjacent islands {Podogymnura aureospinula, Batomys russatus, and Crateromys australis), and it appears that one species is restricted to the Zamboanga Peninsula (Croci- dura grandis; Heaney, 2004, Heaney et al., 2006), though both this species and Zamboanga are poorly known. Additionally, there are many species that occur only in montane and/or mossy forest at high elevations on Mindanao; at least eight species are currently known only from such places, establishing this habitat as another sub- center of endemism. Finally, as documented here, three species {Alionycteris paucidentata, Crunomys suncoides, and Limnomys bryophilus) are currently known only from the Kitanglad Range, though we suspect that all three occur more widely on the many poorly known moun- tains of central Mindanao. Although these basic facts of mammalian distribution have been demonstrated, much of this information has come to light only since 1992 when we began our field studies on Mind- anao (which resulted in the discovery of two of the three species apparently endemic to the Kitanglad Range). Furthermore, although earli- er research on Mindanao has provided a great deal of information on the distribution and ecology of some species (e.g., Hollister, 1913; Hoogstraal, 1951; Musser, 1977, 1982a,b; Rabor, 1986; Sanborn, 1952; Taylor, 1934), many species were poorly known in nearly all respects, especially those from middle and high elevations. Our field and museum-based studies have resulted in a substantial amount of new in- formation on elevational distributions (summa- rized in Heaney, 2001), systematics (Rickart et al., 1998, 2003), and ecology (Ingle, 2001, 2003). We chose to focus our field studies in the Mt. Kitanglad Nature Park, one of the newer (1990) and largest (31,297 ha) of the national parks in the Philippines (Mallari et al, 2001), in Bukid- non Province of north-central Mindanao (Fig. 1). Little published information on the mammals of the park existed at the time we began our studies, though we learned that several collections had been made earlier; these are summarized here. This publication summarizes all currently available information on systemat- ics, distribution, ecology, and conservation status of the mammals of the Kitanglad Range, in part to serve as a baseline for research in other parts of Mindanao, where mammals remain less well known. We chose to work in the Kitanglad Range for two primary reasons. First, the Kitanglad Range includes the second-highest peak on Mindanao (2938 m) and has an extensive area at high elevation (Fig. 2). This led us to expect that we would be able to conduct a thorough assessment of the pattern of distribution and variation in mammalian assemblages along most of the elevational gradient that exists on Mind- anao. Second, the Kitanglad Range supports one of the larger remaining tracts of old-growth forest on Mindanao (Environmental Science for Social Change, 2000), which we hoped would allow us to sample a region with a mammal fauna as little disturbed by humans as possible. We began our sampling at the northwest corner of the park near Barangay San Vicente, Baun- gon Municipality, where old-growth forest occurred in patches as low as 700 m elevation and continuous old-growth forest reached as low as 900 m elevation and perhaps lower, and proceeded to near the highest elevations in the range. We note that photographs of live or fresh specimens and skulls of nearly all mammal species from the Kitanglad Range are avail- able through a web site (http://www.fmnh.org/ philippine_mammals/); links to pdf versions of simple, two-page field guides to non-volant mammals and to bats are also available (http:// fm2.fieldmuseum.org/animalguides/). FIELDIANA: ZOOLOGY Fig. 1. Photograph of the Kitanglad Range taken from above Silipon, below Sites 5 and 6, on 04 March 1993; Site 5 is located on the ridge in the center of the photograph. Study Area and Methods Geology of the Kitanglad Range The Kitanglad Range (Fig. 2) is a part of the Central Mindanao volcanic sector, which is the largest single volcanic field in the Philippines. The sector extends from Mt. Apo in the south to Camiguin Island in the north; all of the mountains in this region have originated during the last 3 million yr, a very brief period for such extensive mountain-building (Sajona et al., 1997). The Central Mindanao volcanic sector is bounded on the east by the Philippine Fault, which runs north-south, largely beneath the Agusan River, and the Eastern Mindanao sector, which extends from Mt. Hilong-hilong in the north to Mt. Diwata in the center and the Mt. Kampalili/Mayo complex in the south. This area is geologically older, with volcanics dated to ca. 47, 46, 29, 19, 12, 11, and 3 million yr ago (Ma), as well as some that are quite recent (less than 0.25 Ma). The Central Mindanao sector is bounded on the southwest by the Cotabato Fault, which largely lies beneath the Mindanao and Alah rivers, and the Daguma/Sarangani sector, which includes the area from Mt. Talayan to Mt. Daguma and Mt. Busa, as well as the portion of the Zamboanga Peninsula to the west of Mt. Dapiak (see fig. 1 in Sajona et al., 1997). This sector also has older volcanics than Central Mindanao, with dates ranging from 32 to 6 Ma, plus Quaternary flows. Within the Central Mindanao sector are numerous Plio-Quaternary lava flows. Beneath the lava flows are late Oligocene to early Miocene limestones overlying much-older peri- dolites and gabbroic/plagiogranitic formations (which probably formed deep beneath the surface). The base of the volcanic field is made up of lavas ranging from about 2.5 to 0.6 Ma. The oldest known surface flow is from Marawi (ca. 40 km west of the Kitanglad Range), dated at 2.31 0.11 Ma. The Kitanglad Range and Mt. Kalatungan, immediately to the south of Kitanglad, "consist of several E-W oriented basaltic edifices ... and rare andesitic/dacitic adventive domes" of Quaternary age (Sajona et al., 1997, p. 130). There are no dated flows from Kitanglad, but Mt. Kalatungan has flows dated from 0.4 0.05 Ma and 0.25 0.07 Ma, and volcanic material dated to 1.15 0.27 Ma is known from Quezon, just southeast of Mt. Kalatungan. Other volcanics in eastern Central Mindanao range from 0.8 to 0.25 Ma (Sajona et al., 1997). HEANEY ET AL.: THE MAMMALS OF MT. KITANGLAD NEGROS BOHOL T~^" y CAMIGUIN SIARGAO BUCAS GRANDE Mount Hilong-hilong SIQUIJOR SULU SEA Mount Dapiak Mount Malindang Mount Tago . _ Mount \<% 'Balatukan \% Mount i $ Lumut/ \j^ Kimangkil Diwata Range / J Mount Sugarloaf S Vv j> Mount ' 1 MOROGULF Pia ^^ an i _^-^i_^^ Mount f , C. /- 5 Talayan \_^>-BASILIAN ^iv?/ V^ o Mount ^^ Qj Daguma $V/- Mount' Busa Mount Matutum ~\ Sea level - 1000 meters 1 1000 -2000 meters 50 I Above 2000 meters n I | I I Kilometers Kitanglad Range Kalatungan Range MINDANAO Mount Apo Mount Latian Complex o* Fig. 2. Map of Mindanao Island and adjacent areas, showing major topographic features and the locations of mountains referred to in the text. These data indicate that both the East Mind- anao sector and the Daguma/Sarangani sector are substantially older than the Central Mind- anao sector, and probably formed as separate islands that coalesced into a single dry-land island with the other portions of Mindanao only within the last 3 to 5 Ma as a result of volcanic activity in Central Mindanao (Hall, 2002; Hamilton, 1979; Mitchell et al., 1986; Sajona et al, 1997). Because the Kitanglad Range did not exist prior to about 400,000 yr ago, the mammals that now live there have colonized the mountain from elsewhere on Mindanao at some point(s) since then. Climate of the Kitanglad Range Limited data on climate are available for the Kitanglad Range, but patterns are clear. High and low temperatures taken during 2-wk periods at primary survey sites during 1992 and 1993 (Table 1) show mean highs of 25.5C at 1100 m elevation declining to 18.1C at 2250 m eleva- tion, for an average lapse rate of 0.64 C per 100- m increase in elevation. Low temperatures at the same sites were 18.6 C and 11.5C, for an average lapse rate of 0.62C per 100-m change in elevation. These lapse rates are identical, or nearly identical, to the average lapse rate of FIELDIANA: ZOOLOGY Table 1. Mean high and low temperatures recorded on the Kitanglad Range at primary survey sites during 1992 and 1993 (see Methods), n = number of days of recording. Temperatures are given as mean and range. Site Elevation Dates Daily highs Daily lows Site 2 1100 m 17-30 April 1992 14 Site 3 1600 m 02-21 May 1992 16 Site 5 1900 m 10-24 March 1993 13 Site 6 2250 m 18 Mar-12 Apr 1993 14 25.5C, 23.0-27.0C 21.6C, 19.0-24.0C 19.5C, 17.5-23.0C 18.1C, 16-22.5C 18.6C, 17.5-20.0C 16.1C, 14.5-17.5C 12.3C, 11.0-13.0C U.5C, 10.5-13.0C 0.64C per 100 m for moist air worldwide (Lomolino et al., 2006). With a peak elevation of 2938 m, we anticipate a temperature differen- tial from the seashore to the peak of 18.8C at any given point in time. Thus, on days when Cagayan de Oro experiences its mean annual daily high temperature of 27C, the peak is likely to have high temperatures of about 8.3C. Because the ability of air to retain water declines as it cools, the cooling of air as it rises over the Kitanglad Range produces heavy rainfall. Average annual rainfall for Impalutao, Impasugong Municipality (600 m elevation), was 275 cm/yr (n = 1 1 yr); at Sumilao (700 m), 291 cm/yr (n = 9 yr); and at Chinchona (1250 m), 380 cm/yr (n = 8 yr; Manalo, 1956). Rainfall at Site 15 (1450 m) during a single year was estimated as 380 cm (Ingle, 2001). These data yield an average rate of increase in rainfall of 12.4 cm per 100 m elevation. From this rate of change, we project that rainfall should average about 480 cm/yr at 2250 m (Site 6). If the increase in rainfall were to continue to the peak at 2938 m, rainfall would be about 564 cm/yr. Similarly, rainfall would decline to about 200 cm/yr near the coast. Thus, annual rainfall on the higher portions of the mountain is very likely to be more than twice, and possibly nearly three times, that in lowland areas. Rainfall in the Kitanglad Range is generally seasonal, with a moderately dry season generally extending from mid-November to early May, and a wet season during the other months. In Malaybalay, the driest months average more than 10 cm of rain, and the wettest average over 30 cm (Ingle, 2001). However, these averages conceal high variability from year to year, perhaps associated with the El Nino phenome- non, when rains decline dramatically, and La Nina, when rains are especially heavy. Given the combination of declining tempera- ture and increasing rainfall with increasing elevation, it follows that humidity is often very high on the mountain. This is consistent with our observation of increasingly abundant and fre- quent fog at our sites from 1600 to 2600 m elevation, and increasingly abundant epiphytes, especially moss, over this elevational range, as described below. Methods To make comparisons possible among sites along this transect of the Kitanglad Range and between transects on different islands, our methods followed the same procedures as used previously (Heaney et al., 1989, 1991, 1999; Rickart et al., 1993). Sampling was conducted in the lowest area of old-growth rainforest that we could locate and reach, and was then extended at sites along the elevational gradient to near the highest peaks (Fig. 3). This included sites in lowland dipterocarp forest (1100 m), montane rainforest (1600 m), transitional montane/mossy forest (1800 m), lower mossy forest (1900 m), and mossy rainforest (2250 m, 2375 m, 2600 m, and 2800 m); detailed descriptions of the sites follow. Additionally, we briefly sampled an area of partially logged lowland forest at ca. 875 m, and Ingle (2001, 2003) conducted extended studies of bats at 1450 m in regenerating montane forest. At our 10 primary sites (Sites 1-9 and 15, described below), we sampled within a ca. 50-m elevational range of the mid-point cited in the site descriptions, and within about 300 m linear distance of the central point (which typically was our camp site). At most sites, we used observation, netting, and trapping to sample the fauna, and we stayed at a given site for at least 7 days, or until we added no new species to our lists for the site. At Sites 1-6, surveyed in 1992-1993, bats were netted in mist nets with four shelves; each net was 12 m in length. Nets were placed in locations chosen by experienced individuals as being likely to receive bat traffic, i.e., especially along ridges, but also across trails and streams, with the bottom of the net usually ranging from 0.5 m to HEANEY ET AL.: THE MAMMALS OF MT. KITANGLAD -8*15 1 124*45' 1 124*50' 1 124*55' 1 125*00' Sumilao 100 km -8*10 125*00' I Fig. 3. Map of topographic features and collecting sites on the Kitanglad Range. Elevation is in meters. 2 m above the ground level, but at each site we set some nets 4-8 m above ground. Nets were tended continuously from early dusk until about 2100 h. Nets were left open thereafter, and bats were removed at dawn. Most nets were left in place for 3 days, then moved to a different location. At Site 15, bats were captured in three to five net systems consisting of two to six 6-m-long nets stacked one on top of each other, with the top net up to 12 m above the ground. Nets were raised and lowered with the use of pulleys (Ingle. 1992. 1993). Between August 1998 and Novem- ber 1999. netting was conducted monthly, usually for three to four nights. Nets were tended every 30 min to 2200 h. and every 1-2 h thereafter until dawn (Ingle, 2001, 2003). At Sites 1-9, non-volant small mammals were captured in Victor rat traps (ca. 90% of traps used) and National live traps (ca. 10%). Traps were checked each morning soon after dawn (ca. 0600-0700 h). and late each afternoon (ca. 1600- 1700 h). Traps were placed 3-5 m apart, with 95% set on the ground and 5% set on vines or horizontal branches, and left in place for 3 days. About 80% of the traps were set by experienced trappers. At Sites 2-6, we recorded the time when we recovered trapped animals, and used Chi-squared tests to determine diel activity patterns by calculating expected capture frequen- cies based on a 14-h nocturnal and/or crepuscu- lar period, and a 10-h period of daylight. Traps were baited each morning and afternoon with either fresh, lightly roasted coconut coated with peanut butter, or with live earthworms. Only coconut bait was used in 1992 (3,047 trap- nights), but both baits were used in 1993, for 4,077 trap-nights (79%) with coconut and 1.067 trap-nights (21%) with earthworms. Bait type was rotated in a nearly random pattern among trap lines and among days within trap lines. In 1993. we recorded which bait had been used to FIELDIANA: ZOOLOGY trap each individual, and used Chi-squared tests to determine bait preference of different species by calculating expected frequencies based on the relative effort with each type of bait. Most small non-volant mammals collected from Site 15 and all obtained from the nearby Sites 13, 14, 16, and 17 were trapped by local people in unbaited native traps. There were four types of native traps, all made of string and either bamboo or wood. The giman and latugpi function as snap traps. The balod is a snare trap, and the bayobo is a balod with a low fence of stakes to funnel the animal into the trap. All trappers were Binukid-speaking residents of Barangay Lupiagan; most are members of the Higaonon ethnic group, but some were Taalan- dig and Bukidnon. They were asked for in- formation on the name and habits of the animals trapped. A few specimens were taken in buildings at Site 15 with snap traps. Voucher specimens were preserved for all species that were trapped or netted; all were cataloged and studied at the Field Museum of Natural History (FMNH), and half have been returned to the National Museum of the Philippines. Some voucher specimens from Sites 10-15 were deposited with the Park Superinten- dent (PASU) of Mt. Kitanglad Range Nature Park. Most specimens (ca. 95%) were preserved in formalin in the field; these were injected with an unbuffered, saturated solution of formalin, then placed in a 10% solution to which a small amount of soap had been added to allow formalin to penetrate to the skin. After several weeks, they were transferred to ethanol for permanent storage. Many of these had skulls removed and cleaned for study at FMNH. The rest were prepared as skeletons, stored in 70% ethanol in the field, then cleaned for study at FMNH. Tissue samples for genetic studies were taken from fresh specimens and frozen in liquid nitrogen. Specimens prepared as skeletons were exam- ined for reproductive status in the field; those preserved in formalin were examined in the museum. Embryo size was measured as crown- to-rump length (CRL). Uterine scars were detected as dark areas of pigment; in some instances, these could be assessed as being either recent or old. Nipples were recorded as enlarged (having nursed young) or small (not having nursed); in our experience with the species reported here, nipples do not fully regress after nursing. Multiparous females were defined as those showing evidence of enlarged nipples and uterine scars; females that were pregnant but had small nipples were considered primiparous. Nulliparous females were small, non-breeding individuals with small nipples and lacking uterine scars. For male specimens, we recorded testis position (abdominal or scrotal) and size (length and width). Stomach contents were examined at FMNH from specimens in fluid. Stomachs were re- moved, and the contents emptied into shallow glass plates for examination under a binocular microscope. Standard external measurements (total length, tail vertebrae, hind foot including claw, and ear) and weight (in grams) were taken in the field on fresh specimens. Cranial measurements were taken at FMNH by L. R. Heaney, using digital calipers graduated to 0.01 mm. Values in the text and tables are given as mean one standard deviation. Temperatures were taken daily at 0800 h from a maximum-minimum thermometer that was attached to the side of a tree in a shaded spot; readings were rounded to the nearest half-degree centigrade. The nomenclature of mammals used here follows Heaney et al. (1998), except that ordi- nal-level names follow Wilson and Reeder (2005). Study Sites With a maximum elevation of 2938 m, Mt. Kitanglad is the highest peak in the Kitanglad Range, and the second highest on Mindanao, following Mt. Apo (2954 m). The Kitanglad Range is generally steep and rugged, rising rapidly from a broad and relatively flat volcanic plain at ca. 900 m to several peaks in excess of 2400 m (Fig. 3). The Kitanglad Range is con- nected by land over 1000 m to Mt. Kalatungan and Mt. Piagayungan to the south, together comprising one of the largest upland areas on Mindanao (Fig. 2). The entire range is some- times referred to as "Mt. Kitanglad," and specimens in museums often bear that name when they actually come from one of the associated peaks. We use the term "Mt. Kitan- glad" to refer only to the peak itself (Fig. 3), and "Kitanglad" to refer to the range (including Mts. Dulang-dulang, Nangkabulos, Imbayao, and others as shown in Fig. 3). HEANEY ET AL.: THE MAMMALS OF MT. KITANGLAD The vegetation on the Kitanglad Range originally consisted of three primary associa- tions. Lowland rainforest, dominated by trees of the family Dipterocarpaceae and with figs (Ficus spp.) commonly included, extended from the base of the range up to about 1200 m. Although some old-growth forest remains, especially at the upper edge, much has been removed and the land converted to agriculture or used for pasture, and much is covered by unproductive saw grass (Imperata cylindrical locally called cogon). Patches of regenerating secondary lowland forest are fairly common, especially along steep hill- sides near streams. In lowland forest, canopy height is usually great (often exceeding 25 m). trees have large leaves (8-15 cm length), the ground has only a thin layer of dead leaves and little or no humus, and moss is scarce. In such habitats, ants are abundant, and earthworms are usually scarce. Beginning at about 1200 m. lowland forest is replaced over a short transition zone by montane forest, which extends to about 1900 m. This forest has very few dipterocarps or figs, but rather is dominated by oaks (often Lithocarpus), laurels (often Cinnamomum), wild cherry (Pru- nus), or oil-fruit trees (Elaeocarpus: Ingle. 2001). In montane forest, canopy height is usually less than 25 m, often 20 m or less, and leaves are smaller (often 5-8 cm). Climbing pandans (Frey- cinetia spp.) and climbing bamboos (Schyzosta- chium spp.) are often abundant. The ground has a continuous layer of dead leaves and humus (though these often are only a few centimeters thick), and moss grows commonly on tree trunks and logs with scattered thin patches on the ground. Ants are uncommon, and earthworms fairly common. Beginning at about 1900 m and extending up to the peaks, the predominant vegetation is mossy forest. Here, canopy height rarely exceeds 20 m. and more often is less than 12 m. Oaks (Lithocarpus) and laurels (e.g., Litsea) are common, and conifers (Podocarpaceae). includ- ing Dacrydium, Phyllocladus, and Podocarpus. often are common and conspicuous because of their large girth (sometimes >1 m) and height (up to ca. 20 m). Leaves are usually small (2- 4 cm). Tree ferns (Cyathea spp.). saxifrage (Polysoma), and oil-fruit trees (Elaeocarpus) often are abundant (Pipoly & Madulid. 1997). The ground has a thick layer of dead leaves and humus that often exceeds 1 m in depth, and moss grows on most surfaces, often forming a thick layer. Rhododendron, Melastoma, and other shrubs are often common. Ants are absent, and earthworms usually are abundant. In all of these habitats, landslides are common because of the steep slopes and high rainfall, and natural fires occasionally occur during the dry season. As a result, there is a mosaic of older and younger successional stages, though the older ones predominate. We refer to such forest as "old-growth." rather than "primary.'* because the former term places less emphasis on the complete absence of any disturbance. Tradition- ally, nearly all of the Kitanglad Range has been used as a source of meat and medicinal plants, with trails reaching nearly every part. As detailed below, we conducted our survey of mammals on four of the mountain peaks comprising the Kitanglad Range: Mounts Du- lang-dulang. Imbayao. Kitanglad. and Nangka- bulos (Fig. 3). The survey included 17 sites along elevational gradients and vegetational types from disturbed lowland forest at 825 m to old-growth mossy forest at 2800 m. The main component of our study was conducted from 1992 to 1993. Additional data were gathered in conjunction with the dissertation research of N. R. Ingle in 1996-1999. We have also included data from earlier surveys of the same area, based primarily on specimens that were deposited at the Field Museum: The Danish Philippine Expedition of 1951-1952 (Sanborn, 1953) and the Philippine Zoological Expedition of 1960 (Ripley & Rabor. 1961). During the first year of our survey, from April to May 1992, we focused our efforts on Mt. Imbayao. at the northwestern side of the Kitanglad Range. We established four study sites at the following elevations and vegetation types: disturbed lowland forest at 825 m (Site 1), old-growth lowland forest at 1100m (Site 2), old-growth montane forest at 1600 m (Site 3), and transitional montane/mossy forest at 1800 m (Site 4). The following year. March to April 1993, we surveyed Mt. Nangkabulos. on the north-central portion of the range, and Mt. Dulang-dulang on the eastern front. On Mt. Nangkabulos we established two survey sites: in transitional montane/mossy forest at 1900 m (Site 5) and old-growth mossy forest at 2250 m (Site 6). On Mt. Dulang-dulang. one of us (B.R.T.) contin- ued the survey from May to June 1993. Three sites were surveyed during this period: old- FIELDIANA: ZOOLOGY growth mossy forest at 2300 m (Site 7), 2600 m (Site 8), and 2800 m (Site 9). Finally, from July 1997 to November 1999, one of us (N.R.I.) conducted an extended study of seed dispersal in residual montane forest at 1450 m, on the northern flank of Mt. Kitanglad (Site 15), where she netted bats (Ingle, 2001, 2003). At the same time, she coordinated small- mammal trapping with local hunters and ob- tained specimens of small mammals from nearby areas, including the village of Lupiagan at 1200 m (Site 13) on the edge of residual montane forest, and three sites in similar vegetation at 1350 m to 1500 m elevation (Sites 14, 16, and 17). Several native traps were used to catch rodents (see Methods). Further, in 1996 N. R. Ingle and J. L. Sedlock made observations on a Pteropus vampyrus roost (Site 10), and found a skull in forest and abandoned lower mandibles from these bats in a hunter's hut at 1000 m and 1100 m, respectively (Sites 11 and 12). The following are brief site descriptions and inclusive dates of survey: Site 1 Mt. Imbayao, 15 km S, 7 km E Baungon, San Vicente Municipality, 811.5'N, 12444.5'E (25 April 1992). This site was in heavily disturbed lowland forest at 875 m elevation, with a mixture of slash-and-burn gardens (kaingin) and partially logged forest, along the Tumalaong River, one of the two headwaters of the Sumaluan River. Slopes were variable, ranging from 0-1 5, most of which were in the higher range. The standing trees probably had been as tall as those at Site 2, but the bigger ones were all gone, with only the stumps, up to 1 m in diameter, remaining. Moss cover on trees was less common, but climbing pandans were just as common as at Site 2. Piper and Ficus spp. were more common here than at Site 2, as were domestic bananas (Musa sp.) in the clearings. We conducted limited netting of bats but no trapping at this site. Site 2 Mt. Imbayao, 15 km S, 6 km E Baungon, San Vicente Municipality, 811'N, 12444.5'E (17-30 April 1992). This site was in old-growth lowland forest at 1100m elevation (Fig. 4), on a narrow ridge along the northeast fork and headwaters of the Sumaluan River. Slopes on the sides of the ridge were 30-40, but sometimes 70. Canopy height was 20-25 m, with emergent trees reaching ca. 30 m; diameter at breast height (DBH) ranged from 25 cm to 60 cm. Canopy leaf size averaged 8-15 cm; no leaves were emarginate. Canopy vines, including Fig. 4. Photograph of habitat at 1 100 m (Site 2) in April 1992. climbing bamboo and pandans, and epiphytes, including orchids, ferns, and moss, were present but not common. The understory included spiny palms, rattan, saplings, and tree ferns. A thin layer of leaf litter covered most of the ground. Moss was present only on fallen logs and the bases of large trees. The soil was weathered volcanic ash, with some small stones; a thin layer of humus covered most of the ground, reaching a maximum depth of ca. 5 cm. Fallen logs were common, and large exposed rocks were present along the ridge and streams. Site 3 Mt. Imbayao, 15 km S, 7 km E Baungon, San Vicente Municipality, 1600 m elev., 810'N, 12445'E (02-21 May 1992). This site was along a high ridge in old-growth montane forest (Fig. 5) with ca. 5-10 slopes on top of the ridge and about 30-60 slopes on the sides. Emergent trees, including gymnos- perms (Agathis), had an average height of 20 m with DBH (above buttresses) of ca. 30-60 cm. The canopy trees, including some oaks (Litho- HEANEY ET AL.: THE MAMMALS OF MT. KITANGLAD Fig. 5. Photograph of lower montane rain forest at 1600 m (Site 3) in May 1992. carpus) and laurels (Cinnamomum) were ca. 15 m high, with an average DBH (above buttresses) of 30 cm. Leaves averaged 5-8 cm; about half had emarginate edges. Some common epiphytes were ferns (including bird's nest ferns), orchids, and hanging moss. Canopy vines, including pandans (Freycinetia spp.) and rattans {Calamus spp.) were moderately common. Understory and ground-cover plants were dominated by ferns, saplings, a few tree ferns, erect screw-pine {Pandanus), and some Melastoma shrubs. Occa- sional patches of tall '"saw grass** were observed. Ficus density was low but density of other fruiting trees was fairly high in the canopy and the understory. Density of moss was low to moderate, found mainly on tree trunks, logs, and old trees; a small amount of moss grew on the ground. The ground surface was covered by about 2-6 cm of leaf litter; this overlaid a humus layer that typically was 4-8 cm, but ranged from 2 cm to 20 cm. The humus layer was generally wet and overlaid weathered red ash. On-site disturbance consisted of a few scattered tree falls. Site 4 Mt. Imbayao, 15 km S, 7 km E Baungon, San Vicente, Municipality, 1800 m elev., 89'N, 12445'E (10-21 May 1992). The forest at this site was old-growth transitional montane/mossy forest, near the top of a small peak. The moderately open canopy included trees that were short, ca. 15 m, on the ridge-sides, and ca. 10 m on the ridge-tops. Undergrowth was heavy, and fallen logs were common. Moss covered tree trunks, limbs, and twigs from the ground to the canopy, growing thickly on the trunks and ground but thinner elsewhere. Ferns, orchids, and other plants grew abundantly as epiphytes, and climbing pandans were abundant. Leaf litter was abundant, and the humus layer was thick (more than 1 m in most places) and spongy. A thick root mat on the surface gave the ground a springy resilience. Long strings of "Spanish moss" hung from many trees. The forest was frequently shrouded by fog. Site 5 Mt. Nangkabulos, 16.5 km S, 4 km E Camp Phillips, 1900 m elev., 8 : 10.5'N, 124=51'E (10-24 March 1993). This site lay in old-growth transitional montane/mossy forest, in an area characterized by steep slopes and fairly narrow ridge-tops. Canopy height was typically 15- 20 m. but emergents (primarily a conifer with emarginate leaves, probably Phyllocladus sp.) reached 20-25 m; none of the trees had but- tresses. DBH averaged 20-30 cm, and leaf size averaged about 20 mm in the canopy, with some up to 60 mm near the ground. Oaks and laurels were common, and a few strangler fig trees were present: dipterocarps and Musa appeared to be absent. Ferns and moss were the common epiphytes: "Spanish moss" and orchids were less common. Canopy vines, especially climbing pandans, were common. Fallen logs, often quite rotten, were also common. Ground cover con- sisted of ferns, climbing ferns, small Melastoma shrubs with ripe red berries, raspberry (Rubus sp.). and moss. Leaf litter covered virtually the entire ground surface, usually 1-2 cm deep, and was underlaid by a layer of moist humus, 10- 30 cm deep, on top of weathered volcanic ash. Site 6 Mt. Nangkabulos. 15.5 km S, 4 km E Camp Phillips, 2250 m elev., 89.5'N, 124=51 'E (18 March- 12 April 1993). This site was located in old-growth mossy forest (Fig. 6) characterized by steep slopes, typically between 20 : and 45 : . The canopy was relatively open, and averaged 7- 10 m in height, with emergents reaching 12-15 m on the ridge-tops and 14-18 m on the slopes. None of the trees had buttresses, and DBH 10 FIELDIANA: ZOOLOGY .' - -4. t\ Fig. 6. Photograph of lower mossy forest at 2250 m (Site 6) on 04 April 1993. A tarp above the work table in camp is visible. averaged 20-50 cm, but a few reached a DBH of 110 cm. The largest trees with the widest di- ameter were gymnosperms of at least two genera (probably Dacrydium and Phyllocladus). Leaf size in the canopy was typically 1>-A cm, with most emergent leaves only 2-3 cm; very few were emarginate. Moss, "Spanish moss," ferns, and orchids were the common epiphytes; canopy vines were nearly absent. Fallen logs were common; many were quite rotten. Understory and ground cover plants were abundant, in- cluding Rhododendron, fruit-bearing shrubs, sap- lings, ferns, and climbing ferns. Moss was common on and near the ground, 1-3 cm thick on trunks and fallen logs, but was scarce more than 2 m above the ground. Leaf litter covered virtually the entire ground to a depth of 5- 20 mm, lying on top of a layer of humus 10- 50 cm in thickness that overlaid weathered volcanic ash. Scattered large rocks (up to 4 m in diameter) protruded through the humus and leaves. Site 7 Mt. Dulang-dulang, 15 km S, 11 km W Dalwangan, Malaybalay City, 2375 m elev., 87.5'N, 12456'E (26 May-04 June 1993). This site was located in old-growth mossy forest above a river and a falls bounded by very steep cliffs and vegetated slopes of about 45-70. Moss was present, but not very common, on trunks of standing trees, fallen logs, and root tangles. A few large conifers (probably Dacrydium and Phyllocladus) were present (DBH = 100 cm) as well as many medium-sized trees (DBH = 30- 60 cm). Lianas were rather common, and epi- phytes were predominantly moss, orchids, and ferns. Understory plants included ferns, tree ferns, saplings, and a common plant (Drimys piperita) called "ali," which local people use as medicine. The ground had a moderate cover of dried leaves that lay over thick humus. A mountaineering trail traversed the area. A few scattered tree falls were noted. Site 8 Mt. Dulang-dulang, 15 km S, 1 1.5 km W Dalwangan, Malaybalay City, 2600 m elev., 87.5'N, 12456'E (29 May-04 June 1993). This site was in mossy rainforest about a kilometer from Site 7. The trees on average were about 20- 25 m, with trunks more profusely covered with moss than the trees of Site 7. The understory was quite dense, consisting of saplings, tree ferns, the medicinal plant Drimys piperita, a shrub (Mela- stoma sp.), and Clethra canescens (small trees with wind-dispersed seeds often common in open areas). The ridge-top was relatively broad (5- 10 m) and flat, and was covered with a moderate amount of leaf litter that lay over a thick layer of humus. The mountaineering trail that traversed Site 7 also traversed this area. Some tree-falls had created openings in the canopy. Site 9 Mt. Dulang-dulang, 15 km S,12.5 km W Dalwangan, Malaybalay City, 2800 m elev., 87.5'N, 12456'E (29 May-04 June 1993). The vegetation at this site was typical mossy forest (Fig. 7) except at the peak itself, which was elfin forest. The average height of emergent trees was about 10 m; DBH was about 20 cm but often looked larger because of the profusion of moss on trunks, branches, and twigs. Canopy leaf size ranged from 1 cm to 5 cm. Understory plants were uncommon and the ground was abundantly covered with leaf litter. The peak was character- ized by sturdy but stunted trees with gnarled and twisted trunks and thickly cuticled leaves. Some grasses and a few herbaceous plants were also present. On the peak and a few meters below, mountaineers had cleared some areas in order HEANEY ET AL.: THE MAMMALS OF MT. KITANGLAD 11 Fig. 7. Photograph of upper mossy forest at 2800 m (Site 9) on Mt. Dulang-dulang taken in July 1993. to pitch tents, with minor disturbance to the vegetation. Site 10 Mt. Kitanglad, Manolo Fortich Municipality, 1000 m elev., (April-May 1996). This site consisted of some remnant lowland forest on a steep bank along the Mangima Creek, surrounded by grassland and agricultural areas. It is located ca. 7 km southwest of Site 15. A roost of Pteropus vampyrus was observed here in April and May 1996. Site 11 Mt. Kitanglad, 9.6 km S, 3.7 km W Sumilao Poblacion, 1000 m elev., 813'50"N, 12453'10"E (1996). A Pteropus vampyrus skull was found in the forest at this site. It was not known if the individual had died at the site. Site 12 Mt. Kitanglad, Sitio Bagalangit, Barangay Kalugmanan, 9.8 km S, 3.7 km W Sumilao Poblacion, 1100 m elev., 8 13'20"N, 124"53T0"E (1996). Lower mandibles of mam- mals were recovered from a hunter's house located in grassland at this site. Disturbed forest remained along watercourses. It was not known where the animals had been hunted. Site 13 Mt. Kitanglad, Barangay Lupiagan, Sumilao Poblacion, 1200 m elev., 812'N, 124 55'45"E (02 March 1999, 14 June 1999). Lupiagan was a community of about 20-30 mostly wooden houses at the end of an unpaved road, surrounded by vegetable and corn farms, with remnant forest remaining along steep slopes leading down to rivers. A few specimens of small mammals were obtained from residents in this village. Site 14 Mt. Kitanglad, Barangay Lupiagan, 10.4 km S 3 km W Sumilao Poblacion, 1350 m elev., 8 11'50"N, 124 55'45"E (17 March 1999; 04 May 1999; 04, 21 and 23 September 1999; 15 and 16 October 1999). This site consisted of a small human settlement of at most 10 houses along Tinag-i Creek. Vegetation in the area was a mix of remnant montane forest, grassland/ fernland, and subsistence agricultural plots, not unlike Site 15. Several residents supplied some rodent specimens they trapped in remnant montane forest near Tinag-i Creek. Site 15 Mt. Kitanglad, 10.6 km S, 2.8 km W Sumilao Poblacion, 1450 m elev., 811'20"N, 124 55'20"E, (July 1997 to November 1999). This was the site for N. R. Ingle's research on seed dispersal (Ingle, 2001, 2003), with bat netting conducted as part of the study; specimens of small non-volant mammals were taken or 12 FIELDIANA: ZOOLOGY procured opportunistically. The area was a mo- saic of montane forest and successional vegeta- tion, bordered by grassland. It was selected because of its relatively level terrain, but steep forested slopes led down to creeks. The montane forest canopy was 20-25 m high and was relatively open. Maximum DBH recorded was 65 cm. Average stem density was 650 stems/ha for stems >10 cm. Canopy trees in the genera Lithocarpus (oaks) and Elaeocarpus (oil-fruit trees) dominated the forest, comprising 32% and 22% of basal area, respectively. Other canopy trees were Castanopsis cf. javanica, Michelia philippinensis, Gordonia luzonica, and the conifer Phyllocladus hypophyllus. Below the canopy, Cinnamomum sp., Actinodaphne diver si- folia, Prunus grisea, Alstonia macrophylla, and Helicia cf. robusta were common, occurring at densities of ca. 15 stems/ha. Erect pandans were also common. Forest floor vegetation included forest seedlings and saplings, shrubs such as Lasianthus spp., and, in some areas, ground ferns. Ficus were present but uncommon. Lianas were present but not abundant; climbing pan- dans were relatively abundant. An old logging road cut a 6-m wide swath through the forest. There were several areas of successional vegeta- tion, which were inferred to be of different ages from the stature of the vegetation and its composition. Alphitonia philippinensis, Mus- saenda philippinensis, Trema sp., and Dodonaea angustifolia were common in the older succes- sional vegetation, which probably was estab- lished 20-30 yr previously. In a 1-ha abandoned slash-and-burn farm surrounded by forest and cleared about 5 yr previous to the study, common species were Aralia bipinnata, Cypholo- phus moluccanus, Melastoma malabathricum, and Saurauia spp. Plants were up to 5 cm diameter and up to 5 m tall. The Department of Agricul- ture maintained a research and demonstration farm nearby, and small-diameter timber had been collected for house construction and for firewood both in the research farm and the village, but human disturbance was at moder- ately low levels. Site 16 Mt. Kitanglad, Barangay Lupiagan, 10.7 km S, 2.9 km W of Sumilao Poblacion, 1450 m elev., 811'10"N, 12455'10"E (26-27 November 1999). This site was along the west side of the Miaray Creek, across from Site 15, where the vegetation was montane forest similar in composition to that of Site 15, except that there had been less disturbance because of the steep slopes. Site 17 Mt. Kitanglad, Barangay Lupiagan, 11.5 km S, 2.2 km W of Sumilao Poblacion, 1500 m elev., 811'0"N, 12455'35"E. (02 May 1999, 28 and 30 September 1999). This site was called Tawasan by the local people. It was situated in secondary forest adjacent to residual montane forest and grasslands and a vegetable farm. Field Work by Previous Investigators In 1960, the Peabody Museum of Yale University and Silliman University sponsored a wildlife inventory of Mt. Kitanglad headed by Prof. Rudolfo B. Gonzales (Ripley & Rabor, 1961; Rabor, 1966). The results of their mammal survey were unpublished, and we were unsuccess- ful in locating any field notes; some, perhaps all, of the mammal specimens were deposited at FMNH. One source of data for their sites is an article on birds (Ripley & Rabor, 1961). In that publication, however, the collection sites, their corresponding elevations, and dates of collection were not enumerated, but were mentioned in the species accounts of selected birds. The team visited the area twice that year, from April to May and during the last week of December, according to Ripley and Rabor (1961). However, dates on the FMNH specimen labels of several small mammals indicated that the team started field work during the last week of March 1960. Information about the sites enumerated below (Sites 18-20, not mapped in Fig. 3), including their elevations and the dates of collection, were based solely on data from specimen labels at FMNH. It appears that the collection of small mammals was opportunistic. Based on elevation, we surmise that the vegetation at their Site 20 approximated that of ours at Site 4, and that Site 19 was probably in old-growth rather than residual forest, perhaps similar to our Site 15. Site 18 Mt. Kitanglad, Malaybalay City, 4200-4300 feet elev. (ca. 1300 m), (24-29 March 1960, 30 April-09 May 1960). Site 19 Mt. Kitanglad, Malaybalay City, 5000 feet elev. (ca. 1500 m), (03-07 April 1960). Site 20 Mt. Kitanglad, Malaybalay City, 5800-6200 feet elev. (ca. 1800 m). In 1951-1952, an extensive survey of birds, mammals, and other wildlife was conducted on several islands by the Danish Philippine Expedi- tion (DPE) headed by F. Salomonsen. The group HEANEY ET AL.: THE MAMMALS OF MT. KITANGLAD 13 visited several areas on Mindanao, including Mt. Kitanglad, between August 1951 and March 1952. Based on the dates in published species accounts of birds and mammals (Salomonsen, 1953; Sanborn, 1953), Salomonsen and party spent at least 3 mo in Bukidnon, surveying Mt. Kitanglad and Mt. Kaatoan from October to December 1951. The party then proceeded to Agusan the following month, most likely through Cabanglasan, Bukidnon Province, where they collected a tarsier, Tarsius syrichta, during the first week of December. Almost all of the specimens collected during the DPE were deposited at the Zoological Museum in Copenhagen. A few mammals, however, were deposited in the FMNH (see Species Accounts). The results of their Kitanglad sojourn, particularly regarding birds and mam- mals, were promptly published (Salomonsen, 1953; Sanborn, 1953). In these publications, however, the collection sites and their corre- sponding elevations and dates of collection were not routinely specified. Our attempt to recon- struct the Salomonsen party's collection sites and dates on the Kitanglad Range was unsuccessful because no further data were available from the Zoological Museum in Copenhagen. As with the Gonzales collection a decade later, we thus lost an opportunity to make a detailed comparison of the ensuing changes in small-mammal commu- nities during the last four decades, a period of massive habitat destruction. In the absence of further data, the following sites, dates, and elevations were reconstructed from available data on the labels of specimens from the DPE that are in the FMNH, as well as from relevant published accounts of birds and mammals (Salomonsen, 1953; Sanborn, 1953). Site 21 Cabanglasan, no elev. data (Decem- ber 1951). This site lies east of Malaybalay, at the border of Bukidnon and Agusan del Sur. Site 22 Mt. Kaatoan, Cinchona, 1250 m elev., (12-23 November 1951). Site 23 Mt. Kitanglad, 1600 m elev., (16-18 December 1951). Site 24 Dilirig Caves, Manolo Fortich Mu- nicipality, Bukidnon (January-February 1933). Sometimes also spelled "Dalirig," this barangay is located along the boundary of the municipal- ities of Manolo Fortich and Sumilao, beside the primary road (formerly Highway 315, now National Road 3), about 4.5 km E Manolo Fortich poblacion at ca. 450 m. These caves are currently a tourist attraction, and were formerly heavily minded for guano. In 1933, L. H. Phillips captured Eonycteris robusta, Eonycteris spelaea, and Hipposideros diadema. Several specimens were deposited in the FMNH but most are in the Museum of Comparative Zoology (MCZ), Harvard University. No field notes are associat- ed with the L. H. Phillips collection (J. Chu- pasko, pers. comm.). Also deposited at the Field Museum is a specimen of Hipposideros diadema that Lim Boo Liat collected in 1960 from an unknown cave in Bukidnon, designated as Site 25. Un- fortunately, the collectors' field notes and catalogs for both collections cannot be traced. Site 25 Unknown cave, unknown locality in Bukidnon, unknown elevation (26 January 1960). Lim Boo Liat collected an H. diadema in Bukidnon, but details are absent. Lastly, a large series of small mammals was collected in September-December 1965 by L. Bregulla and deposited at the Senckenberg Museum, Frankfurt (SMF). This collection pro- vided the type specimen of Alionycteris pauci- dentata (Kock, 1969b). Bregulla apparently collected in several other parts of the Philippines in 1964-1966, including the Central Cordillera of Luzon, where he obtained the type specimen of Otopteropus cartilagonodus (Kock, 1969c). Aside from papers on Dyacopterus spadieeus (Kock, 1969a), and records of Tarsomys echinatus and Limnomys sibuanus (Musser, 1994), no other details on Bregulla's collecting activities on Mindanao or elsewhere in the Philippines are available in the literature. Heaney examined some specimens from this collection in June 1989. and cranial measurements of some species are included here. Accounts of Species Order Erinaceomorpha Family Erinaceidae Hedgehogs and Gymnures Podogymnura truei Mearns, 1905 The Mindanao gymnure is endemic to Mind- anao Island, where it is widespread in montane and mossy forest at 1 300 m and higher; there are no historical records from lowland forest (Fig. 8; Heaney et al., 1998). The only congeneric species, P. aureospinula, occurs in secondary and old- growth lowland forest on Dinagat (Heaney & Rabor, 1982) and on Bucas Grande (Tabaranza, unpubl. data). During 1992 and 1993, we 14 FIELDIANA: ZOOLOGY 3000 2500 C 2000 o 1500 Q) s 1000 500 t peak of Mt. Kitanglad, Mindanao t * -d -S fl -3 3 Fig. 8. Elevational ranges of small mammals from Mindanao. Records from the 1960s are indicated with open squares, and those from the 1990s by solid squares. The stars indicate documented records from the 1990s without voucher specimens (see text). captured 94 individuals in transitional montane/ mossy and mossy forest from 1800 m (Site 4) to 2800 m (Site 9); overall, it was the most common small mammal at mossy forest sites (Table 2; Fig. 8). We captured them in traps set on the ground, usually along runways, in front of holes, among mossy root tangles, under rotting fallen trees, or under the bases of standing live or dead trees. In 1999 we obtained 10 specimens trapped by local people in residual montane forest at 1450-1500 m (Site 15), using unbaited native traps called giman and balod. In December 1951, the DPE collected four individuals at 1600 m (Site 23), of which one is deposited at the FMNH (Sanborn, 1953). In March, April, and June 1993, we captured nine pregnant females with a mean weight of 72.6 9.4 g (range = 60-84 g). All but one (which had twins) had single embryos with mean CRL of 17.2 14.0 mm (range = 4-40 mm, n = 9). During the same period, 18 adult females (64.6 7.0 g, range = 55-78 g) had large mammae but were not pregnant and four others were nulliparous (59.5 7.3 g, range = 52-68 g). None of the females captured in May 1992 (n = 13) and November 1999 (n = 2) was pregnant, but several had swollen uteri. Adult males weighed an average of 67.8 6.3 g (range = 52-82 g, n = 33); young adults were slightly lighter, averaging 62.8 8.5 g (range = 52-72 g, n = 5). Testis size was measured in two young adults and seven adults; size ranged from 9x6 mm to 14 X 7 mm. Juveniles and subadults of both sexes were recorded from March through May. There were only few of them, comprising less than 6% of the total captures in 1992, 1993, and 1999. Mean weight of the two juveniles and four subadults was 53.0 6.3 g (range = 43-62 g). Of 72 standard- ized captures, all were recorded as nocturnal/ crepuscular, and none occurred in the daytime (Table 3), indicating strongly nocturnal activity. Earthworms were significantly more attractive as bait to P. truei than was toasted coconut; the only other small mammal that significantly preferred earthworms was Tarsomys apoensis (Table 4). Stomach contents of 55% of P. truei (n = 31) consisted chiefly of earthworms, crudely chewed into ca. 3-mm to 20-mm sections. In several instances, the sections were not com- pletely severed, and portions of more than 20 mm length were found in the stomachs; one contained a ca. 50-mm earthworm that was HEANEY ET AL.: THE MAMMALS OF MT. KITANGLAD 15 Table 2. The number of small non-volant mammals trapped at principal sites in the Kitanglad Range in 1992 and 1993. The number of standardized captures per 100 trap-nights are given in parentheses. Values in brackets include non-standardized captures; see Methods. Site 2 Site 3 Site 4 Site 5 Site 6 Site 7 Site 8 Site 9 1100 m 1600 m 1800 m 1900 m 2250 m 2375 m 2600 m 2800 m Crocidura beat us 1* 0** 1 (0.08) 1 (0.06) 0** 0** 2 (0.43) Podogymnura truei 13 29 30 6 9 7 (1.07) (1.78) (1.49) (0.82) (1.94) (2.26) Urogale everetti 2 (0.16) 1 (0.06) 1 (0.05) Apomys hylocoetes 8 58 [62] 9 2 4 (0-43) (2.83) (1.23) (0.43) (1.29) Apomys insignis 4 (0.35) 29 (2.39) 24 (1-54) Batomys salomonseni 4 (0.33) 14 (0.86) 15 [16] (0.75) 3 (0.41) Crunomys melanius 2 (0.17) Crunomys suncoides 1 (0.05) Limnomys bryophilus 16 (0.80) 1 (0.14) 4 (0.86) 5 (1.61) Limnomys sibuanus 4 (0.25) 5 [6] (0.25) Rattus everetti 2 (0.37) 4 (0.35) 1 (0.08) Rattus exulans 1 (0.06) Rattus tanezumi 1 (0.06) Tarsomys apoensis 4 (0.25) 15 [17] (0.75) 3 (0.41) 0** 4 (1.29) Total captures 2 [+2] 10 50 [+4] 87 140 [+8] 22 17 20 Total trap-nights 540 [+137] 1154 1216 [+172] 1626 201 1 [+54] 732 465 310 Total mammals 0.37 [0.29] 0.87 4.11 [3.89] 5.35 6.96 [7.17] 3.0 3.66 6.45 per 100 trap-nights Number of species 1 + 1* 3 + 1** 6 10 8 + 1** 5+ 1** 4+ 1** 4 * caught by hand; ** presence inferred largely intact. In these individuals, finely chewed arthropod parts were a minor element of the stomach contents. The remaining stomachs contained finely chewed arthropod parts in- cluding hymenopteran wings and coleopteran elytra. No discernable plant materials were evident in any stomach (Table 5). These data suggest that this species feeds on invertebrates, with strong tendency toward vermivory. These feeding habits are similar to those of Chrotomys and Rhynchomys from Luzon (Balete & Heaney, 1997; Heaney et al., 1998; Rickart et al., 1991). Females are slightly larger than males in some external measurements but are very similar in most cranial measurements (Table 6). Based on a small series from Kitanglad, Sanborn (1953) described a subspecies, Podogymnura truei min- ima, that differed from the nominate form of the species from Mt. Apo in being of smaller cranial size; he used only four specimens from Mt. Apo for comparison. Direct comparison of our large series from Kitanglad with a large sample from Mt. Apo (both in FMNH) shows only a slight difference in size (Table 6); we thus agree with Corbet and Hill (1992) that recognizing two subspecies may not be justified. Specimens from Kitanglad have a standard karyotype of 2N = 40, FN = 76 (Rickart, 2003). SPECIMENS EXAMINED Total 109. Site 4 (13 FMNH); Site 5 (29 FMNH); Site 6 (30 FMNH); Site 7 (6 FMNH); Site 8 (9 FMNH); Site 9 (7 FMNH); Site 15 (3 FMNH); Site 16 (5 FMNH); Site 17 (1 FMNH); Site 19 (3 FMNH); Site 20 (2 FMNH); Site 23 (1 FMNH). 16 FIELDIANA: ZOOLOGY Table 3. Diel activity patterns of small mammals on the Kitanglad Range taken in 1992 and 1993 at Sites 2-8. Diurnal activity period was 10 h, noctural/crepuscular was 14 h. Chi-squared test for expectation of activity level equal to length of diel period. Nocturnal/crepuscular Diurnal Species Observed Expected Observed Expected Total Podogvmnura truei 72** 41.76 30.24 72 Crocidura beatus 2 1.74 1 1.26 3 Urogale everetti 3 2.32 1 1.68 4 Apomys insignis 58** 33.64 24.36 58 Apomys hylocoetes 66** 39.44 2 28.56 68 Batomys salomonseni 33** 19.72 1 14.28 34 Crunomvs melanius 2 1.16 0.84 2 Crunomys suncoides 1 0.58 0.42 1 Limnomys bryophtfus 16** 9.28 6.72 16 Limnomys sibuanus 10** 5.80 4.20 10 Rattus everetti 6* 4.06 1 2.44 7 Rattus exulans 1 0.58 0.42 1 Rattus tanezumi 1 0.58 0.42 1 Tarsomys apoensis 13 12.18 8 8.82 21 Total 283** 172.84 15 125.16 298 * p < 0.05. ** p < 0.01. Order Soricomorpha Family Soricidae Shrews Crocidura beatus Miller, 1910 The Mindanao shrew is endemic to Cami- guin Island and the Mindanao Faunal Region, including at least Biliran, Leyte, Maripipi, and Mindanao; it is common in primary forest particularly at higher elevations, uncommon in secondary forest, and absent outside of forest (Heaney et al., 1998, 2006). There are histor- ical records from near sea level on Mindanao, indicating that this species originally had (and may still have) the greatest elevational range of any native mammal species on Mindanao (Fig. 8). During 1992 and 1993, we captured Table 4. Bait attractiveness of roasted coconut with peanut butter vs. live earthworms at Sites 5-9 in 1993. Chi-squared test for expectation of equal trap success with the two baits, using traps set in a standardized manner, based on number of standardized trap-nights at the site(s) where the given species occurred (see Methods). Coconut Earthworm Species Observed Expected Observed Expected Total 33 63.99 48** 17.01 81 1 2.37 2 0.63 3 58 62.41 21 16.59 79 21 20.75 4 4.25 25 30* 25.28 2 6.72 32 18 20.54 8 5.46 26 8 7.29 1 1.71 9 15 20.54 11** 5.46 26 184 221.99 97** 59.01 281 1375 (0.83) 275(0.17) 1626 1589(0.79) 422 (0.21) 2011 1137(0.74) 370 (0.26) 1507 4077 (0.79) 1067 (0.21) 5144 Podogymnura truei Crocidura beatus Apomys hylocoetes Apomys insignis 1 Batomys salomonseni Limnomys bryophilus Limnomys sibuanus 2 Tarsomys apoensis All species (all sites) 1900 m, standardized trap-nights 2250 m, standardized trap-nights 2375-2800 m, standardized trap-nights Total standardized trap-nights * p < 0.05. ** p < 0.01. 1 1900 m (Site 5) only. 2 1900 m and 2250 m (Sites 5 and 6) only. HEANEY ET AL.: THE MAMMALS OF MT. KITANGLAD 17 Table 5. Stomach contents of native non-volant small mammals from the Kitanglad Range. Numerical values indicate the percent of stomachs that contained some identifiable volume of each type of material. Vegetable matter Species n (seeds, etc.) Arthropod exoskeleton Earthworm Podogymnura truei 31 100 55 Crocidura beatus 3 100 Urogale everetti 8 12 100 50 Exilisciurns concinnus 2 100 100 Apomys hylocoetes 40 95 100 Apotnys in sign is 35 91 97 Batomys salomonseni 19 100 Bullinnts bagobus 3 100 100 (trace) Cnmomys melanius 2 100 100 Crunomys suncoides 1 100 Limnomys bryophilus 16 100 Limnomys sibuanus 4 100 Rattus everetti 5 100 80 (trace) Tarsomys apoensis 10 100 50 (trace) five individuals at sites in lowland forest at 1100m (Site 2), transitional montane/mossy forest at 1800 m (Site 4), and mossy forest at 1900 m and 2600-2800 m (Sites 5 and 8, res- pectively; Table 2; Fig. 8). One of the trap- ped individuals was captured during the day (Table 3). Two were caught in traps baited with coconut coated with peanut butter, two in traps baited with live earthworms (Table 4), and one was hand-caught during daylight by a particularly alert field assistant. Additional- ly, in 1999 we found a dead shrew on the ground in residual montane forest at 1450 m (Site 15). Stomach contents of three individuals con- sisted solely of finely chewed arthropod exo- skeleton, including wings. These are consistent with insectivorous feeding habits. We suspect that the two individuals caught in traps baited with coconut were accidental captures, because their position in the trap suggested that they bumped the bait pedal. Cranial and external measurements were similar in both sexes, and show only slight variation from those on Leyte, Biliran, and Maripipi (Table 7; Heaney & Ruedi, 1994; Rickart et al., 1993). Direct comparison of three adult crania from Kitanglad with specimens from Duminagat, Mt. Malindang (FMNH 87391), Dabiak, Zamboanga (FMNH 80360), and Camiguin Island (FMNH 167855) showed no evident differences. Based on a single speci- men from Leyte, this species has a karyotype of 2N = 38 (Rickart, 2003). SPECIMENS EXAMINED Total 6. Site 2 (1 FMNH); Site 4 (1 FMNH); Site 5 (I FMNH); Site 8 (2 FMNH); Site 15(1 FMNH). Suncus minimis (Linnaeus, 1766) The Asian house shrew is a widespread commensal species recorded from Asia to Indo- Australia; it occurs throughout the Philippines, but it is not a native species. Usually it lives in urban and agricultural areas, and occasionally in disturbed forest. However, on islands with few native small mammal species, such as Camiguin and Negros, S. murinus is common in old-growth forest, particularly where there is natural distur- bance from landslides at upper elevations (Hea- ney et al., 1989, 1998, 2006). In the Kitanglad Range, we found them only in association with a residential area and heavily disturbed forest, and not in primary forest (Table 1). Similar occurrence patterns were observed on Mt. Isarog, Luzon (Heaney et al., 1999) and on Biliran. Leyte, and Maripipi (Rickart et al., 1993). In May and June 1999, we captured two males; one was caught by hand in a coffee garden close to houses in a village at 1200 m (Site 13), and the other in an unbaited native snap trap (giman) in residual montane forest, near the Tinag-i Creek at 1350 m (Site 14). An adult male caught on 04 May 1999 weighed 40.6 g and had scrotal testes measuring 5x7 mm, whereas a subadult male caught on 14 June weighed 27 g and had abdominal testes measuring 3x6 mm. A Lupiagan village leader aged more than 70 yr stated that S. murinus had not been present FIELDIANA: ZOOLOGY g 52 V, Q CU TT & u E M E o - ~ u - <*- O CA J5 ca -E C b W) E a J 3 >, 3 u _C M U e T3 u 'E ; J u E u u Q ca E go -j a E E 7Z & < CO U "O "oo 9) u E E O ca u n .a ^ '^ 4J <*- n "^ c M <- tso c g a * 5 oo ca -o M so c S ' 3 .9 5 n - > w ftca w 3 fi 25 M < 60 ra Hcn - 3 If +1 d +1 d d +1 o +1 in + ! in in rT _2 se r- r-j = s o\ SS o\ d 2 "* 'r, "-. 1- = -= a +l no 3 +1 d NO +1 I/O n d +i co d r i (*1 od d +i o CN M od X no d C*i On oi r> >n t~ no +1 On +| On vq o\ 4 od od od od +t r +l - co On in >n W1 ON ON ON 9C 1 0^ + 1 +1 ON +1 On 7 +l ) **1 r~- NO Br NO CO VI -^ -*cf **J rv On O^ od ON od Q\ On NO NO NO o M (N M (N +1 On +1 CO d I +1 +1 NO in + 1 in NO n r~ csj On in yz t- od IO d xj 9N 90 no in in Cl m co co +1 oo +| r~ (N Tl- CO r- NO NO * +i S o <1N o 2 S ^ 5 m .o o c S (N ^ N_ O T3 ^ JS N 60'^ S i> ^E E ^ c o I .CO - g> .3 :0 15 ^ ~- c O "O P .S e3 T3 Q JS rt 60 o o. S l g 2 E 4- O a o u. c U -j E _ J2 3 - <-. 60 3 3 a 2 n*- _c r-1 ~ y: -J 3 _ E 3 1) 1/3 -3 CO 73 H c 3 - c/> 3 o,u 3 c u ctf -S g b "o Z u == c .o E s U r 3.3 CO (V> d e^ Z. 3 60 - ~ c , c 3 T3 -3 r^ vo m r- VO r- VO x oo oo X -= o o i-- 5 = + 1 +1 in +1 VO +1 O B ^r A iC rs _ X o r- = * * *r * VO o r~ +1 00 +1 00 m - rr fNj 06 06 06 06 00 tN _ _ NO + 1 On +1 00 +1 00 On OO (N r^ c~, t~- v~> 00 00 r-- r- r-~ r- a ON On + | On +| On +| r~- O O O ^! O ^. O O +1 +1 + 1 +1 m vo * O ON O On 3 Ov O O - o *. o ^ ^ + 1 +1 - On O r- ^t r- r-- (N r-- ! + 1 +1 + 1 . + 1 - +1 3 U. S tu rvi + 1 vo +| NO r-- r^ r~ fN On O ~ O NO o + 1 NO +| t-~ vO Ov t r- nC OO cni rj =>' on d + 1 7 +1 vn vo fN NO r*1 +1 7 +1 7 + 1 o 00 r^ NO m fN r*- t CN O O * * * 1*1 fN t ^f t Tf fN _; _; >-^ ^ ^ r~ r^ o> n n X O On - - NO ~ t_ ? ON X O c> fN fN -' + 1 >T) + 1 fN NO + 1 On X + 1 ON + 1 7 + 1 7- +1 \r V. n Tf X ^C ^c X <* -T * * >n t m r^ ^ 5 00 r. ~ * On m J fN~ d d t^ d d +1 -r + 1 -T + 1 f*^ fN + * n +1 +1 r^ +1 12 T -r r*^ fN -r m fN fn fN C in -r co fN fN fN r^ ~ X O m _ iC cr m -r d d d * d d +1 -1- +1 -f + 1 rN + 1 fN +1 +1 +1 m n fN rj fN fn fN fN _-. f*N ri f^ fN rfv. fN fN fN fN fN + 1 rp +| r- ^ -> ^t 6 r< w r, rs +1 ? +1 2 + 1 r- o r- 00 vo ^ M ^ p,J + 1 OO r~- _, ^cr-r- ^ ^" ^ ^ r-\D^cvo r--r- 00 o fN Ci ^ r- ^ o +1 ^0 +1 r-7 n 00 >o sO vO v \0 vo fN + 1 fN +| fN ro OC O fN -^ a< o m - r~- +1 3 +1 - P >"i fN + 1^+1* no ^4- no m r- t-~ r- r- 8 2 fN 2 + 1 fN +1 0! Ov 4-1 .P U 3 U 3 c s 2 g S is "2 c 2 C i ^ S W3 5 -- " cd a c u U OJ *- 60 8 . w a - 2-S t) JO -4- s t- KJ <$ _. V3 en 3; M U cs) . G ** " a +1 S a - *= ^ &- c ^ d > ^ c '5b cd o> T3 ' 5 . S S 1 5 jS ~i = S e U ^ oi -= ;i 0) ? g a to - ji 8 . 5 = if .= M 8 jg H _ rs ti C = - _^ = * u S. i 'I a, t: ^r m on no Oi tl tl ON On o\ On 4 ^ - OC OO l^ s s +1 VO +1 no NO W) U-> 4 NO NO NO NO o __ o +1 * +1 qi OO 00 OO 00 ON O Cn] o ~ o ~ . +1^+1^ O On 0O NO O O O NO <= _ +1 rji +| M ; ON O 0O cnI ' (N Cn| r*~i _.+!_. +1 On + 1 +1 4 On >ri ON OO ON I^ o o o o ON 00 -yi On 11.7 11.5- 1.6 O on + 1 00 OO tl - +1 - \D O ^O *? +l 7 +l 7 ^ ^t in n fN (N (N 3 ^ en a eg 35 > S ^^ en en dj S3 CD -o c en - a 8.1 1> r- .E <*- -o 'E J.S ox) g C - S3 frt 3 C u S3 r= OCL, QS cu C >- 00 S3 S3 ll as| d> go Ho* g "~ T3 .H 2| CO 60 (U - ss r - E ^t en H E E * a "o S a ~ s. _2 O U M c 1 a B 2 JI "S o W ^ T. cu > .= IE J| o 9 "it c y E A 1 Z! Ft 5 N Jl o CU >> > "O "B e .- Tt o < B .5 _i; ^_ 9 3 * M 1 3 V c A w = .= *B = E M = _o M H "5 = H _^ c X r 8 V I =. X ce r- co +1 ON i o + 1 4 -r 4 m ro ro r*" ( CO fl <^> n O _ O )0 +1 O in 1 _ + 1 3 _ iri Tf -T -1- Tt o <*> ro r^i r*^ r*-j +1 f"J> ^t w * r*\ ro f*S r*-j r*"i m oo oo oo r^t^-oo t^r-~r~vo r*- i^ vo ^o ^* oo Tj- Tt Tj" f*% f*S rn o -r ^, +1 o +1 00 ~ o c o ON o o 00 X o OO jC X o ^ d ^ ~ m ON ON +l - +i 7 7 ON ON ON Wl ON C i Ojv m m m on on on rr ^ -a- oo as r- oo * * on m i^i r- Tt tt m r<~, Tt ft on m ^o ^o o ^. n in r o^irof^ mmro o^mm +1 - ON ON on on m oo oo r*- oo o +l 7 o o soon >/^ ^o in >o so v~) so mm m r*imr^j r^ m n +1 m ON SO ^t r*- t on r^i + 1 o> oo oo oo ^- m, ui oo m t o oo t 000 co co co cc r~ oo fc 2^ u. 2 S- s ^ 5 5 5 5 t I HEANEY ET AL.: THE MAMMALS OF MT. KITANGLAD 35 Rickart et al.. 1993). Sedlock (2001) described the echolocation call on Luzon. SPECIMENS EXAMINED Total 1. Site 2 (1 FMNH). Murina cyclotis Dobson, 1872 The round-eared tube-nosed bat ranges from Sri Lanka to Hainan, Borneo, and the Philip- pines. Philippine records are from within the Luzon, Mindanao, and Sibuyan faunal regions, where it has been documented in primary and lightly to heavily disturbed lowland and montane forest from 250 m to 1500 m (Heaney et al., 1998; Sedlock, 2001). On Kitanglad in April 1992. we obtained two specimens in old-growth lowland forest at 1100 m elevation (Site 2; Table 10). An adult female (8 g) was pregnant with a single embryo (CRL = 6 mm). The adult male had scrotal testes. Pregnant and lactating females were recorded in April and May on Mt. Isarog, Luzon (Heaney et al., 1999). External and cranial measurements (Table 12) of both specimens are consistently smaller than those of specimens from the islands of Biliran and Catanduanes. and even smaller than those of specimens from Mt. Isarog, Luzon (Heaney et al., 1991, 1999; Rickart et al., 1993). Sedlock (2001) has described the echolocation call on Luzon. Specimens from Luzon had a karyotype of 2N = 44, FN = 50 (Rickart et al., 1999). SPECIMENS EXAMINED Total 2. Site 2 (2 FMNH). Myotis sp. On Mt. Kitanglad in April 1992, we netted a single small, dark brown Myotis (adult female, 6 g; FMNH 147067) in old-growth lowland forest at 1100 m (Site 2; Table 10). It was pregnant with a single embryo (CRL = 21 mm). The braincase is domed, rising sharply from the rostrum, which is relatively long and narrow. The canines are long, and the upper unicuspid (PM - ) is present but small and only slightly out of line in the toothrow. The base of the thumb is not broad and fleshy, but rather is similar to those of Myotis muricola bronni. In these features, this specimen differs from Myotis rossetti and Myotis ridleyi of Southeast Asia (Corbet & Hill, 1992), but is similar to those of Myotis ater and Myotis muricola, from which it differs in size (Table 12). Myotis ater is known from Vietnam and peninsular Malaysia to Sulawesi, the Mollucas. and New Guinea (Simmons, 2005). We re- examined specimens previously referred to M. muricola from the Philippines, and are readily able to refer series from several localities to this larger species. Specimens from Culion Island, Palawan Province (FMNH 63653-63654, 63679- 63681) originally referred to Myotis nugax by Sanborn (1952), and two specimens from Mt. Isarog, southern Luzon originally referred to M. muricola (USNM 573778-573779; Heaney et al., 1999) have measurements nearly identical to those given by Hill (1983) and Corbet and Hill (1992) for M. ater, and share the broader rostrum and zygomatic arches, more inflated braincase, and proportionately smaller PM 3 relative to M. muricola. Other series we examined are typical M. muricola browni; these include specimens from Negros Island (FMNH 145546), from Cagayan (176551), Kalinga (FMNH 167239), and Laguna (FMNH 177469) provinces on Luzon Island, plus previously examined specimens with pub- lished measurements from Biliran, Leyte, and Maripipi islands (Rickart et al., 1993) and the holotype of M. muricola browni from Saub. Cotabato Province, Mindanao (Taylor, 1934). These data demonstrate that the two species are broadly sympatric over at least part of the Philippines. Our single specimen from Kitanglad differs from both M. ater and M. muricola, and does not fit within any other species of Myotis from Southeast Asia (Corbet & Hill, 1992) or Walla- cea (Flannery, 1995). For this reason, we do not assign it to any species at this time. SPECIMENS EXAMINED Total 1. Site 2 (1 FMNH). Philetor brachypterus (Temminck, 1840) The short-winged pipistrelle is distributed from Nepal to New Guinea, and occurs through- out most of the Philippines where it has been documented in primary and disturbed lowland forest from 475 m to 900 m (Heaney et al., 1998: Sedlock, 2001). On Kitanglad, we netted an adult male in disturbed lowland forest at 800 m (Site 1) and three others in residual montane forest at 1450 m (Site 15; Table 10). An adult female (14 g) netted in April was pregnant with a single near-term embryo (CRL = 25 mm) and an adult male (15 g) netted in November had scrotal testes. In most external and cranial measurements (Table 12), the Kitanglad specimens are similar to those from Leyte and Mt. Isarog, Luzon 36 FIELDIANA: ZOOLOGY (Heaney et al., 1999; Rickart et al., 1993), and to specimens from Catanduanes (FMNH 140603) and Negros (FMNH 145548). Sedlock (2001) has described the echolocation call from Mt. Makil- ing, Luzon. SPECIMENS EXAMINED Total 4. Site 1 (1 FMNH); Site 11 (3 FMNH). Pipistrellus javanicus group Although recent publications on Philippine pipistrelles have considered only a single species from the P. javanicus group to be present (e.g., Heaney et al., 1998, 1999), a study on Mt. Makiling, Laguna Province, Luzon, indicates that specimens previously referred to P. javanicus represent two species, which differ in echoloca- tion calls and also have slight morphological differences in pelage, body size, and skull morphology (Sedlock, 2001). Examination of 30 specimens from Mt. Kitanglad revealed two and perhaps three species. We report on these briefly here, but note the importance of more extensive study. The majority of the specimens from Kitanglad are small, with condylo-incisive length of adults averaging 12.52 mm for males and 1 1.98 mm for females, forearm of 34.0 mm (males) and 34 mm (females), and total length of 80.4 mm for males and 84 mm for females (Table 12). This small morph includes 24 individuals from Sites 2, 3, 5, 6, and 15 (Table 10). These are quite similar to specimens of the smaller morphotype from Mt. Makiling, Laguna (FMNH 142035, 142041- 142045), as well as specimens from Kalinga Province, Luzon (FMNH 167237-167238). Al- though these represent the smallest Pipistrellus from Kitanglad, they are substantially larger than Pipistrellus tenuis (Ingle & Heaney, 1992). Five additional adult specimens from Kitan- glad (FMNH 147069-147070, 147072, 147076, 147854) are larger, with condylo-incisive length averaging 13.21 for four males and 13.18 mm for one female, forearm of 35.5 mm and 37 mm, and total length of 88.2 mm and 84 mm for males and females, respectively (Table 12). They are similar in size to a large morphotype from Mt. Makiling, Luzon (FMNH 142036, 166446- 166449), but some external features appear to differ, and the two large morphotypes may represent different species (Sedlock and Heaney, unpubl. data). These larger bats were captured at Sites 2 (1100 m; 3 of 5), 4 (one from 1800 m), and 5 (1900 m; 1 of 2), indicating a lower elevational range than the small morphotype. There is no size overlap between the "small" and "large" morphotypes from Kitanglad; those individuals nearest in size are separated by more than 0.5 mm in condylo-incisive length. In addition to differences in cranial size, these two morphotypes differ in the attachment of the plagiopatagium on the lateral side of the ankle (Sedlock and Heaney, unpubl. data), in the degree of swelling of the soft tissue on the rostrum, and in the color of the ventral pelage. In addition to these two well-defined morpho- types on Kitanglad, a single subadult taken at 1450 m (Site 15; FMNH 166473) is large, with condylo-incisive length (13.26 mm), forearm (37.7 mm), and total length (84 mm) and most other measurements similar to the "large morpho- type" but with a wider interorbital region; this specimen may simply represent a young individual of the large morphotype, but more study is needed. In both morphotypes, males outnumbered females by at least four to one. An adult female of the small morphotype captured in early May had a swollen uterus and was lactating. Specimens of what appears to be a large morphotype from Mt. Isarog, Luzon, had a karyotype of 2N = 38, FN = 48, which differs from that of Malaysian specimens referred to P. javanicus (Rickart et al., 1999). SPECIMENS EXAMINED Total 30. Site 2 (5 FMNH); Site 3 (2 FMNH); Site 4 (1 FMNH); Site 5 (2 FMNH); Site 6 (15 FMNH); Site 15 (5 FMNH). Family Molossidae Otomops sp. Mastiff bats of the genus Otomops occur from Africa to New Guinea. A specimen from Balbalasang, Kalinga Province, Luzon, repre- sented the first published Philippine record (Heaney et al., 2005a), but the first captured was one from Baguio, Benguet Province, Luzon, taken on 15 July 1991 by R. I. Crombie (FMNH 142615). In August 1999, R. Baylomo and A. Pepay captured an adult male (19 g) with scrotal testes (FMNH 167382) in residual montane forest at 1450 m (Site 15; Table 10). In most external and cranial measurements (Table 12) the Kitanglad specimen is notably smaller than those from Luzon, and there are cranial shape differences; the Philippine specimens of Otomops are under study by J. Eger (pers. comm.). SPECIMENS EXAMINED Total 1. Site 15 (1 FMNH). HEANEY ET AL.: THE MAMMALS OF MT. KITANGLAD 37 Order Primates Family Tarsiidae Tarsius syrichta (Linnaeus, 1 758) The Philippine tarsier is endemic to the Mindanao Faunal Region, with records from the islands of Basilan, Biliran, Bohol. Dinagat, Leyte, Maripipi, Mindanao, Samar, and Siargao. It is common in primary and secondary lowland forest and shrubby agricultural areas from sea level to 700 m (Dagosto & Gebo, 1997; Neri- Arboleda & Arboleda, 2002; Rickart et al, 1993). Although we obtained no specimens on Ki- tanglad, we observed tarsiers in disturbed low- land forest at 825 m (Site 1). In 1951, the DPE collected a single female at 450 m in Cabangla- san, Bukidnon, close to the boundary with Agusan (Sanborn, 1953). A radio-tracking study in the wild on Leyte indicated a small home range in males, ca. 0.6-2.0 ha, typically with three to four sleeping sites within a home range (Dagosto et al., 2001). The same study showed that foraging and traveling are done close to the ground, mainly at 1-2 m above ground. Specimens from Mindanao and Leyte had karyotypes of 2N = 80 (Rickart, 2003). SPECIMENS EXAMINED. None. Family Cercopithecidae Macaca fascicularis (Raffles, 1821) The long-tailed macaque occurs from Burma to Timor and throughout the Philippines, where it occurs in primary and secondary lowland forest and primary montane forest, and in adjacent agricultural areas from sea level to 1800 m (Fooden, 1995; Heaney et al., 1998). On Kitanglad, we sighted the species in disturbed lowland forest at 875 m (Site 1), old-growth lowland forest at 1100m (Site 2), residual montane forest at 1450 m (Site 15), old-growth montane forest at 1600 m (Site 3), transitional montane/mossy forest at 1800 m (Site 4) and 1900 m (Site 5), and old-growth mossy forest at 2250 m (Site 6). According to the local farmers and hunters near Site 15 in 1997-1999, monkeys were formerly more common, often raiding gardens planted with corn on the lower slopes. We were in the forest at this site on almost a daily basis for many months, but the long-tailed macaque was rarely heard or seen. When seen, no more than 10 individuals were observed. Local people hunted the monkeys for sale as pets to local buyers, to agents supplying monkey farms near Manila, and for food. A lower mandible was found in a hunter's hut at 1 100 m (Site 12), along with rodent and flying fox mandibles. SPECIMENS EXAMINED Total 1. Site 12 (1 FMNH). Order Rodentia Family Sciuridae Exilisciurus concinnus (Thomas, 1888) The Philippine pygmy squirrel is endemic to and widespread within the Mindanao Faunal Region, with records from Basilan, Biliran, Bohol. Dinagat, Leyte, Mindanao, Samar, and Siargao. It occurs in primary and secondary lowland and montane forest from sea level to 2000 m (Heaney, 1985; Heaney et al., 1998). On the Kitanglad Range, we caught single individ- uals in old-growth montane forest at 1350 m (Site 14), in residual montane forest at 1450 m (Site 15), and in transitional montane/mossy forest at 1800 m (Site 4; Fig. 8). The specimens collected at Sites 13 and 14 were caught by local residents using unbaited native traps called giman. In addition, we observed several individ- uals running quickly along fallen logs on the ground at the edge of kaingin and partially logged lowland forest at 825 m (Site 1); another was seen scampering up a tree trunk in old- growth montane forest at 1600 m (Site 3). They produced a sharp chirp similar to that of Holarctic chipmunks ( Tamias). Two adult males and an adult female were collected in April 1960 in transitional montane/mossy forest at 1800 m (Site 20). Stomach contents of two females contained finely ground pale and reddish-brown plant materials, and both also contained roughly chewed arthropod larvae. Three adult females weighed an average of 31 g (range = 25-35 g), one of which, caught in May at Site 4, was pregnant with a single embryo (CRL = 12 mm) and another, caught in April at Site 15, had a swollen uterus. The pregnant female was multiparous and had three placental scars. A female (25 g) captured in September at Site 14 had large mammae but no visible placental scars. In most external and cranial measurements (Table 13), the Kitanglad specimens are slightly larger than those from Biliran, Dinagat, Leyte, and Siargao, but indistinguishable from speci- mens taken elsewhere on Mindanao (Heaney, 1985; Heaney & Rabor. 1982; Rickart et al., 38 FIELDIANA: ZOOLOGY -2 o I* 18 J* *-> -> S 4> 2 ITS "O - ^ 4> eg 3 a ? u ^3 re e go..! gi ^-o ^ 8 c fa u 5= c a 3 > ,- i) x> !-> u u g 3 ;rt Q> u *~ re 3 . g C =3 05 S S re 3 d 3 h La c '53 M J3 N -0 4) ^ 'J 'i O." > 5* . > m Cm u c C -.1 U , = c 5 .5 re 3 > Cl a 0) c/3 -0 +1 S" - re U (/I "O c 2 CS ! sd r^ ^d r^sor-^so 1 1 .2 j< I aa oc in vd +1 so +| vd +| so no no r~ o no +1 00 +| 00 3 CI ON o\ 00 n <* tj- ci ci ci o "l O "i no in no m r~ r~ r~ r~ m 00 (N + i y + i y m tJ- tt in (N 00 NO 00 r- ON On fS rf * *' ** on c 00 in on m r^i en ro rn on ^t in . ' n^ - ^ tl N tl . 00 ON 00 in ^- 00 in r~ rn <^' no in fN r<^ fN fN + 1 m +1 CI +1 1". +1 Cl 4 CN. +1 r4 NO + 1 n ON + 1 r 1 +1 r ] 4 +1 ON r i m +1 in Cl ON i~- O ON N r 1 3 r- y. 00 00 CN, no cn| rNi Nt (ni Nt r, fS (N f) d -J & n r. as en a n | 2 C_ 2 E 3 tn | E c ^ a u 5 g 1 ^> B u 3 3 o -g *N en ^ 03 2 C tU > a _c u -. u 3 d Q 60c l> S 1 E 2 -5 ~ in < 03 "3 en Q2 tu E 6 OO tn C + 1 N u a * 03 _tu J ,2 "5/5 S E^ . C/3 03 tu c 01) J-S < a HS f tn 1 ft lit S I 2 01 ft O On - p - 4) o M m cc I~- t~- Tt OO fN v> O * O ft ' >n O 3 t ft +1 +i +1 w> r- r- CO r I _- 90 00 4 * __ \fi vn VN *r Tf t ft ft o 9i > O >n n I r~;Ti;-> On OnN '''' OnOn On >o m (N . o r^ r-' r~ r~ no '^o vo n p~ t r i On oo i On On On "' 4 "i 't ' ^- tt *> ON rn r-i m f*^ m r~i <^-j *