�

BANWANATURAL SCIENCE

Volume 4, Number �, 2007Special Issue: 2007 Proceedings of the

Wildlife Conservation Society of the Philippines

Editor-in-ChiefEufemio T. Rasco, Jr.

Banwa Editorial BoardReynaldo G. Abad

Peter J. BattSylvia B. ConcepcionRicardo M. de Ungria

Larry M. DigalAntonio G. Moran

Virginia P. ObsiomaEdmundo B. PrantillaEufemio T. Rasco, Jr.

Gilda C. Rivero

WCSP Board of Reviewers

Banwa is a biannual refereed journal publishedby the University of the Philippines Mindanao.

Ruth Grace B. Rosell-AmbalRowena Reyes-Boquiren

Rafe M. BrownArvin C. Diesmos

Jake EsselstynJames W. Grier

Lawrence R. HeaneyTodd E. Katzner

Myrissa V. Lepiten-TabaoIreneo L. Lit, Jr.

Rolando L. MetinPerry S. Ong

Mary Rose C. PosaEric A. RickartJodi L. SedlockWoody TurnerDoug Watkins

Indira Dayang L. WidmannPeter Widmann

2 �

Copyright © 2007 by WCSP and the individual authorsand the University of the Philippines MindanaoAll rights reserved. Reproduction without permissionfrom the publisher is strictly prohibited.

ISSN �656-�7�9

Banwa is published twice a year underthe auspices of the Office of Research

Contributions may be submitted to:

The Editorial BoardBanwaOffice of ResearchUniversity of the Philippines MindanaoMintal, Tugbok District8022 Davao City, Philippines

For inquiries, call or fax (6�82) 29�-�8�9, orVisit our website: http://www.upmin.edu.ph/ojs/index.php/banwa/Inquiries and submissions may also be addressed to: banwa@upmin.edu.ph, or research@upmin.edu.ph

Editorial staff: Abraham A. Garcia, Jr., Production Editor & Publication Designer Cheryl T. Salili, Editorial Assistant Mitchiko A. Lopez, Editorial Assistant

Design and photo credits:Front cover: Kit design of the �6th Biodiversity Symposium–Wildlife

Conservation Society of the Philippines); Polillo Tarictic Hornbill (Penelopides manillae subnigra)–Juan Carlos Gonzalez, Polillo Islands Biodiversity Conservation Foundation Inc.; Guenther’s forest frog (Platymantis guentheri) taken from Tinagong Dagat in Mt. Hamiguitan, Davao Oriental.–Elsa May Delima; Madagascar fish eagle (Haliaeetus vociferoids)–Dr. Lily Arison Rene de Roland, The Peregrine Fund

Back cover: A Sakalava fisherman in Madagascar prepares his catch for transport to market by drying and smoking the fish in front of open, wood-burning fires.–Dr. Rick Watson, The Peregrine Fund; Apomys sp. at Mt. Cetaceo, Penablanca Protected Landscape and Seascap–Mariano Roy Duya, Conservation International Philippines

Internal Review of Articles, Solicitation of Peer Reviews, Guidance for Revision of Manuscipts, and Proofreading by the

WCSP Editorial CommitteeWCSP Editorial Committee

WCSP Proceedings EditorCarlo Custodio

Board of the Wildlife Conservation Society of the Philppines(April 2006-April 2007)

Perry S. Ong : President Carlo C. Custodio : Vice President Ma. Nancy P. Ibuna : Secretary Grace R. Ambal : Treasurer Indira Dayang L. Widmann : Auditor Jayson C. Ibañez : Member Tom Brooks : Member Arvin C. Diesmos : Member Mariano Roy M. Duya : Member Marisol D.G. Pedregosa : Member Ely L. Alcala : Member Myrissa Lepiten-Tabao : Member Lawrence R. Heaney : Member Angel C. Alcala : Member Emeritus Blas R. Tabaranza, Jr. : Member Emeritus

(April 2007-April 2008) Perry S. Ong : President Carlo C. Custodio : Vice President Ma. Nancy P. Ibuna : Secretary Grace R. Ambal : Treasurer Jayson C. Ibañez : Auditor Mariano Roy M. Duya : Member Indira Dayang L. Widmann : Member Marisol Pedregosa-Hospodarsky : Member Tom Brooks : Member Ely L. Alcala : Member Arvin C. Diesmos : Member Sabine Schoppe : Member Juan Carlos T. Gonzales : Member Angel C. Alcala : Member Emeritus Blas R. Tabaranza, Jr. : Member Emeritus Lawrence R. Heaney : Member Emeritus

The Wildlife Conservation Society of the Philippines (wcsp.8k.com) is a professional society of about 200 members with the goal of advancing wildlife research and conservation in the Philippines. The WCSP meets annually in April. Proceedings are published as special issues in national journals.

4 5

Message

Since its establishment in �99�, the Wildlife Conservation Society of the Philippines (WCSP) focuses its efforts in advancing wildlife research and conservation in the Philippines. Promoting collaborative research, providing technical assistance and training and increasing public awareness has since become central elements of the WCSP members’ DNA. WCSP capitalizes on its core competencies in the field of wildlife biology to create new value and provide significant technical and scientific inputs for the furtherance of wildlife conservation in the Philippines.

Wildlife conservation is now in an era of radical change. WCSP measures up to the challenge by way of holding annual symposium wherein current research findings and trends in wildlife conservation are presented.

The publication of the proceedings of the �6th Annual Philippine Biodiversity Symposium held on �6-�8 April 2007 with the theme “Renewing ties: Scientists and Grass-roots Practitioners for Biodiversity Conservation” is one concrete step towards popularizing and gathering support for wildlife conservation in the country. We recognize that wildlife biologists and scientists cannot do it alone. Other members of the society, therefore, must contribute if we are to respond appropriately to the increasingly diverse requirements of wildlife and biodiversity conservation.

Let us, and I am sure we will, continue to evolve, to maintain our dynamism and to look at our work from the wildlife’s point of view.

PERRY S. ONG PresidentWildlife Conservation Society of the Philippines

Message

The University of the Philippines Mindanao’s strives towards becoming a Center of Excellence in the field of research. The University continues to enhance its capabilities to transmit and to generate exchanges of knowledge in an evolving culture of research.

The Banwa Journal, an academic journal of the University, was established primarily as a venue for researchers and scholars to publish their research output on: culture and the arts; biosystems and the environment; and strategic policy and sustainable management. It published its first issue in April 2004. It was initiated by the College of Humanities and Social Sciences. Banwa means “community” in the vernacular.

For these issues, the Banwa is privileged to publish the conference proceedings of the Wildlife Conservation Society of the Philippines. We take pride in supporting the conservation efforts of the Society. We have alloted three issues for their 2006 and 2007 Annual Philippine Biodiversity Symposiums. In this humble way, we contribute to the increasing awareness of the conservation efforts of a dedicated community of wildlife biologists and scientists.

Madayaw sa tanan!

GILDA C. RIVEROChancellorUniversity of the Philippines Mindanao

6 7

Article Digest 7

Wildlife Trade in Southern Palawan, Philippines 12 Rommel M. Cruz, Deborah Villafuerte-van den Beukel, Indira Lacerna-Widmann, Sabine Schoppe, and Peter Widmann

The Herpetological Importance of Mt. Hamiguitan 27Range, Mindanao Island, Philippines Elsa May M. Delima, Arvin C. Diesmos, and Jayson C. Ibañez

Report on a Survey of Mammals of the Sierra Madre 41Range, Luzon Island, Philippines Mariano Roy M. Duya, Philip A. Alviola, Melizar V. Duya, Danilo S. Balete, and Lawrence R. Heaney

Density and Feeding Preference of the Polillo Tarictic 69Hornbill Penelopides manillae subnigra in Fragmented Forests of Polillo Island Ana Katrina Mamangun and Juan Carlos Gonzalez

Community-Based Wetland Conservation Protects 83Endangered Species in Madagascar: Lessons from Science and Conservation Richard T. Watson, Lily Arison René de Roland, Jeanneney Rabearivony, and Russell Thorstrom

Banwa Instructions for Authors 98

Contents

2007 Proceedings of the Wildlife Conservation Society of the Philippines16th Annual Philippine Biodiversity Symposium

Wildlife Trade in Southern Palawan, Philippines Cruz, R.M., D. Villafuerte-van den Beukel, I. Lacerna-Widmann, S. Schoppe, and P. Widmann

Southern Palawan is one of most important source areas for animals entering the illegal wildlife trade in the Philippines and Barangay Culasian in Rizal is a poaching “hotspot.” This study aims to contribute to the understanding of the illegal wildlife trade business. Informal taped interviews were conducted in Rizal, the largest municipality in Palawan, with former wildlife traders, motorcycle drivers who are hired to transport the wildlife, middlemen and concerned government personnel. Information was gathered on the species traded, the terms of transport, origins, destinations and prices. Information obtained indicates that wildlife trade extends to localities in the municipalities of Rizal, Quezon, Brooke’s Point, Bataraza, and Balabac. Confiscation records show that 25 species were illegally traded in the years 2000 to 2006. Of these, 22 were sold for the pet trade, two for bush meat, and one for hobby or traditional medicine industry. Of all traded wildlife species, �9 are listed as globally threatened or near-threatened according to IUCN. Thirteen (52%) of the confiscated species are endemic to the Palawan faunal region. Beetles were the most traded wildlife in terms of number of individuals. Among birds, the Hill Myna was heavily traded for the pet trade followed by the Blue-naped Parrot. The Palawan hornbill was poached from its nest for the pet trade or for food. Of all traded wildlife recorded in this study, the Philippine cockatoo is the only critically endangered species. Freshwater turtles are also commonly traded. The 7-year confiscation record shows that 2,174 mynas and parrots were confiscated representing a mere 6.5% of the estimated traded number of ��,250 in seven years (4,750 individuals/year). Most shipments go undetected, and even if traders are apprehended law enforcement is weak. A major problem is that wildlife smuggling is highly organized, involving powerful and influential circles. We conclude that the confiscated animals represent only a very small percentage of the actual trade. A thorough understanding of the dynamics and complexity of trade is necessary to develop the expertise necessary to implement the anti-illegal trade law.

Article Digest

8 9

The creation of the Culasian Managed Resource Protected Area in Rizal under the Southern Palawan Anti-Poaching Initiative (SPAPI) has led to a decline in poaching activity inside the �,954-ha forest in the last two years. Alternative livelihood programs, conservation education and advocacy intensification have also proved to be a powerful tool to change perception of poachers towards conservation.

The Herpetological Importance of Mt. Hamiguitan Range, Mindanao Island, PhilippinesDelima, E.M.M., A.C. Diesmos, and J.C. Ibañez

This paper presents the results of biodiversity surveys in areas within Mt. Hamiguitan on the eastern part of Mindanao Island, Philippines focusing on frogs and reptiles. The data is the first documentation of these vertebrate groups for this mountain range. Three habitat types were visited corresponding to the following sites: dipterocarp, transitional dipterocarp -montane and mossy-pygmy. In each of these sites, frogs, lizards and snakes were sought using standard sampling methods for frogs and reptiles. A total of �4 species were recorded (�6 frogs, �4 lizards and 5 snakes). Majority of the species (25 species: 74%) are found solely in the Philippines while �� of the 25 species are currently known in Mindanao and neighboring islands. The study also accounted for the presence of species which have limited preserved samples, formerly known to occur in other localities, or those referred to as site specific endemic species. The list includes Limnonectes cf. diautus, Philautus poecilus, Brachymeles gracilis hilong, Sphenormorphus cf. diwata and Tropidophorus davaoensis. In terms of elevational distribution, the bulk of the species encountered were found in lower elevations, in the dipterocarp forest site which appears to be outside the boundaries of the proclaimed protected area. Because of this and due to the high occurrence of species restricted to the Philippines, recommendations on reassessing current conservation efforts in Mt. Hamiguitan appear vital and timely.

Report on a Survey of Mammals of the Sierra Madre Range, Luzon Island, PhilippinesDuya, M.R.M., P.A. Alviola, M.V. Duya, D.S. Balete, and L.R. Heaney

To prepare an inventory of mammals, �� sites were surveyed along the Sierra Madre mountain in 2002 to 2005. The survey covered habitats from lowland to mossy forest, at elevations �00 to �500 m a.s.l. A total

Article Digest

of �8 species of mammals was recorded. This includes �0 fruit bats, �5 insectivorous bats, nine rodents, two ungulates, one primate and one carnivore. Eight are new records for the Sierra Madre. This includes two potentially new species of rodents in the genera Apomys and Chrotomys: Kerivoula cf. papillosa, not previously known from the Philippines; and Coelops hirsutus, previously known only from Mindanao and Mindoro Islands. Rhinolophus inops, R. philippinensis, Harpiocephalus harpia and Murina cyclotis have not been reported previously, and are generally poorly known overall. Initial examination of Pipistrellus javanicus, Rhinolophus arcuatus and Rhinolophus virgo suggests some differences from the known species, and further taxonomic study is needed. We captured Archboldomys musseri on Mt. Cetaceo and nowhere else, supporting previous evidence that it is endemic to that mountain alone. The native large mammals were scarce, due to both habitat destruction and heavy hunting. A modified mist netting technique for insectivorous bats (V-net) was employed during the survey. We used two mist-nets and arranged it in a V configuration wherein one end of each net is attached to a common secured pole. One arm of the V-net is fixedly positioned and serves as a wall whereas the other arm is mobile. Two persons were assigned to guard the nets, one positioned at the mobile arm of the V-net and the other at its side along the wall. When a bat enters, the mobile arm is shut quickly, trapping and entangling them along the interior of the wall. This technique contributed to the success of capturing most of the insectivorous bats during the survey. Our data clearly show that the mammalian fauna of the Sierra Madre has been poorly known, and is more diverse than previously documented, with the presence of endemic species being especially noteworthy. More surveys are needed in areas where few or no data are currently available; particularly in karst, ultrabasic and mossy forest, to fully document and thereby permit protection of its diversity.

Density and Feeding Preference of the Polillo Tarictic Hornbill Penelopides manillae subnigra in Fragmented Forests of Polillo IslandMamangun, A.K. and J.C. Gonzalez

The density and feeding preference of the Polillo tarictic hornbill (Penelopides manillae subnigra) were compared between a heavily disturbed secondary forest and residual lowland forest. Using transect analysis, individuals were counted and the density of tarictics in each forest site was computed. The disturbed forest site exhibited higher tarictic density (4.75 individuals per km²) than the residual forest site

Article Digest

�0 ��

Article Digest

(1.25 individuals per km²). The difference in tarictic density between the two sites was affected by several factors such as the abundance of fruiting trees in each site, the presence of nesting trees and also the degree of anthropogenic disturbance. The disturbed forest site was more fragmented (�9.96% forested area) than the residual forest site (26.�9% forested area). It was concluded the both forest sites were suitable in sustaining tarictic populations. The disturbed forest site contained more fruiting trees foraged and dispersed by tarictics thus a higher density was observed. In the residual forest, large trees were observed that are essential for nesting during breeding season. Male tarictics are territorial especially during this time thus they drive away other tarictics to protect their nests and as a result, lower density was observed. Historically, forests of Polillo have undergone fragmentation. Being endemic, tarictics are subjected to limited habitat thus they adapt and tolerate anthropogenic disturbances to forest habitat available.

Community-Based Wetland Conservation Protects Endangered Species in Madagascar: Lessons from Science and ConservationWatson, R.T., L.A.R. de Roland, J. Rabearivony, and R. Thorstrom

The Peregrine Fund’s research in the early �990s showed that survival of the Madagascar fish eagle (Haliaeetus vociferoids) is threatened by human persecution and habitat loss. Of a global fish eagle population estimated at �00-�20 breeding pairs, the single highest concentration of �0 pairs breed on three adjacent lakes in western Madagascar. Fishing on the lakes is the main livelihood of local Sakalava people who have harmoniously shared these important wetlands with fish eagles for centuries. From 1991 through 1995 we documented a massive influx of migrant fishermen who abused local traditional resource extraction rules and threatened the livelihood of local inhabitants as well as survival and production of one of the world’s most endangered eagles. The economic incentive to endure the hardship of migration to the lakes and camping on the lakeshore for the season was strong. In �995 per-capita income from fishing was about $1500 for the six month season, about 7.5-times the national annual average. Fish stocks were rapidly diminished through the fishing season as catches diminished to the point where fishermen gave up fishing before the end of the season. Fish stocks were lowest when Madagascar fish eagle nestlings fledged, affecting annual productivity. The most serious impact of fishermen may be on the lake-side forest, which was used as a source of dugout canoes and wood to fuel fish-drying fires. To conserve this important breeding site we worked collaboratively with the local

community to enhance and enforce traditional resource utilization rules that helped prevent loss of fish eagle breeding habitat, reduce nest site disturbance, and sustain prey availability. We used a �996 law to empower rural communities to control natural resource use by creating two community associations with authority to enforce local rules. We helped the associations become effective through training, advice, logistical, and scientific support. The associations passed their three-year probationary period, have started a �0-year assessment period, have been awarded WWF’s “Gift to the Earth” for their exemplary role in conservation, and our model is being duplicated by others throughout Madagascar. Ten pairs of Madagascar fish eagles continue to breed on the lakes, and annual productivity has recovered to normal levels. Our goal is to spread this model to other valuable conservation sites in Madagascar: Tambohorano, for Madagascar fish eagles, and Bealanana, for Madagascar harriers and the Madagascar pochard, a species we rediscovered in late 2006 after the species was declared probably extinct.

Article Digest

�2 ��

Introduction

In recognition of its importance to global biodiversity, the entire island of Palawan was declared a Biosphere Reserve with two world heritage sites, an Endemic Bird Area, and a Philippine Priority Area for Biodiversity Conservation (Diesmos and Palomar, 2004). Yet, illegal trade in flora and fauna is a major concern. Despite lack of properly documented information, apprehension data alone provide enough evidence of rampant illegal trade. Data from �999-2002 indicate that a large number of hunted wildlife are birds, particularly talking mynas, wild quail and blue-naped parrots (Lasmarias, 2004). Several endangered and endemic species are also commonly traded, e.g. Palawan hornbill, leopard cat, Palawan peacock pheasant, and red jungle fowl. Excessive hunting is putting many of the rare and Palawan endemic species at the brink of extinction (Werner and Allen, 2000; Widmann, �998). The southern part of Palawan is the center of illegal trade in the province and one of the hottest hotspots for illegal wildlife trade in the Philippines (van den Beukel et al., 2006; Widmann, 2006). Levels of poaching and hunting are rampant in Bataraza, and Rio Tuba was identified as a local center for illegal wildlife trade (Widmann and Diaz, 2004). Barangay Culasian in Rizal was identified as a poaching “hotspot,” thus it was made as the project site for Southern Palawan Anti-Poaching Initiative (SPAPI) (Widmann, 2006). It is difficult to ascertain the impact of hunting on wildlife populations in Palawan. First, confiscation data do not indicate where these species were collected and second, there is little information on the population of the hunted wildlife species (Lasmarias, 2004). However, assuming that the volume confiscated only represents a small proportion of the total number of illegal wildlife, we can assume that populations in the wild are at risk if hunting cannot be controlled. But to be more conclusive, there is a need to explore the dynamics, extent, and impacts of illegal wildlife trade in the Palawan Corridor and this should be among the priority research needs in Palawan (Diesmos and Palomar, 2004). The present study aims to contribute to the understanding of the illegal wildlife trade in Southern Palawan. It also aims to assess the origin and destination of wildlife, uncover operation modes of wildlife traders, transportation mode and trade routes.

Cruz et al. 2007. Banwa 4(�):�2-26.

Wildlife Trade in Southern Palawan, Philippines

Rommel M. Cruz1, Deborah Villafuerte-van den Beukel, Indira Lacerna-Widmann, Sabine Schoppe, and Peter Widmann� Correspondening author. KATALA Foundation Inc. (KFI), P.O. Box �90, National

Highway, Barangay San Jose, Puerto Princesa City 5�00, Palawan, Philippines. rumil_0�@yahoo.com.

This research study was conducted in collaboration with Critical Ecosystem Partnership Fund (CEPF), Loro Parque Fundacion (LPF) and funding partners.

Abstract

Southern Palawan is one of the hottest hotspots for illegal trade of wildlife in the Philippines. Large numbers of wildlife are transported either by fishing vessels or private chartered planes from the south of Palawan to Zamboanga, Cebu, Manila, Batangas and even to Malaysia. Parrots and mynas are among the species of birds most traded due to their huge demand in the market. Other birds that are also under considerable pressure of poaching are Palawan hornbill and White-bellied sea eagle. Apart from birds, other Palawan wildlife included in the illegal shipments are Palawan pangolin, Balabac mouse deer, Palawan bearcat, Palawan bearded pig, Southern Palawan tree squirrel, freshwater turtles and beetles. The present study identified species of conservation priority involved in trade. The study also presents data on traded wildlife species in Palawan including their market value, modes of transport, operation of wildlife traders in Palawan, and trade routes.

Keywords: Palawan, Philippines, wildlife trade

Banwa. 2007. 4(�):�2-26.

�4 �5

Materials and Methods

Interviews were conducted in the Municipality of Rizal located 20� km south of the provincial capital, Puerto Princesa City (Figure �). Rizal is the largest municipality in Palawan with a total land area of �25,9�5.45 ha. Rizal has a net in-migration rate of 80% (Anda and Tabangay-Baldera, 2004). Informal taped interviews were conducted with former wildlife traders, hired motorcycle drivers, middlemen, Municipal Environment and Natural Resources Officers (MENRO), Philippine National Police (PNP) officers, Palawan Council for Sustainable Development Staff (PCSDS), and Department of Environment and Natural Resources (DENR) personnel who are aware of the wildlife trade and were involved in wildlife conservation. Information was gathered on the species traded, confiscations, means and ways of transporting wildlife, origins and destinations, price paid to collectors and selling prices.

Cruz et al. 2007. Banwa 4(�):�2-26. Cruz et al. 2007. Banwa 4(�):�2-26.

For the purpose of this study and in accordance with Republic Act 9147, wildlife trade is defined here as an act of engaging in the exchange, export or import, purchase or selling wildlife, their derivatives or by-products, locally or internationally. Wildlife traders or shippers are defined as persons or group of persons doing wildlife trafficking. A trader is identified as the person financially sustaining the middlemen and poachers to collect, transport and deliver the wildlife from its origin to the trader. The function of the trader is to gather wildlife and supply the demand in and outside Palawan and the Philippines. Middlemen are persons whose primary duties are to buy wildlife from various poachers, heap and deliver them to traders.

Results and Discussion

Confiscation records from DENR and PCSDS from 2000-2006 that were compiled by Katala Foundation (KFI, 2007) show that 25 wildlife species or species groups were illegally traded in Southern Palawan (Table 1). Of these, 22 species are specifically recorded as traded wildlife, two are bush meat, and one is for the hobby or traditional medicine industry. Of all traded wildlife species, �9 are listed as “globally threatened” or “near-threatened” by IUCN (2007). Thirteen (52%) of the confiscated species are endemic to the Palawan faunal region.

Beetle trade Beetles are the most traded wildlife in terms of number of individuals. The major source areas are Narra, Aborlan, Brooke’s Point, Rizal, Quezon and Bataraza based on the interviews conducted with traders. Unlike other species, beetles are small, easy to collect, and transport, and one of the most expensive and demanded wildlife. The demand for beetles, particularly of the genera Dorcus and Odontolabis is documented through recent confiscations in Palawan. The beetles are offered over the internet and sold in vending machines in Japan. Most beetles are collected and traded for their supposedly aphrodisiac qualities. Female beetles for instance, are soaked in wine beverages and other liquors in some bars and restaurants in Metro Manila. A serving of beetle in a glass of tequila would cost as much as a few thousand pesos. The taxa Odontolabis and Dorcus titanus palawanicus are sold in pairs to insect collectors in the Philippines

Figure 1. Map of Palawan, Philippines where the Municipality of Rizal is highlighted

Busuanga •• Coron

El Nido •Sandoval

• TaytaySan Vicente •

• Roxas

• Puerto Princesa

• Aborlan• NarraQuezon •

RIZAL •

• Brooke’s PointBataraza •

• Balabac

Cagayancillo

Cuyo •

�6 �7

carrying the insect-filled baggage is the primary reason. Beetles are restrained using thick plastic and tied with a cotton string. They are packed in bags that could accommodate as many as �00-�00 individuals. Feeding is stopped once the animals are packed, hence a delay in shipment would eventually kill the insects. Beetles are hand carried and delivered to restaurants and to insect collectors in Manila. According to World Bank (2005) some beetles are available through international dealers on the internet at USD200.00 or more per pair, presenting a considerable profit to urban traders. The trade of beetles in Palawan had been ongoing for almost a decade until it was discovered recently in Palawan. The high demand for beetles was noted in the large volumes confiscated in seaport and airport terminals in Puerto Princesa City in 2006 (KFI, 2007). In these confiscations it was known that Europeans were involved in the trade.

Bird trade Among birds, the Hill Myna is most sought after for the pet trade. It is bought from collectors at Php400.00-900.00/individual depending on the age (Table 2). This species is mainly found in Rizal where nesting trees (Manggis; Koompassia excelsa) are abundant. The high demand is due to its unique characteristic, particularly its ability to mimic human voices and sounds (van den Beukel et al., 2006). Hill myna and other bird species are mostly poached by members of the Palaw’an and Tagbanua ethnic groups.

Cruz et al. 2007. Banwa 4(�):�2-26.

for breeding purposes. The prices of beetles vary depending on the size; the larger the beetle, the higher the price. The price of beetles in source areas showed a big difference from the market price in Manila (Figure 2). The insect trade in East and Southeast Asia began in the mid-�990s when economic returns from other wildlife were declining, and it peaked in the late �990s (World Bank, 2005). In Palawan, beetles are usually transported weekly via commercial shipping lines, or daily through passenger planes. The risk of detection is high but very few confiscations are made. According to informants, connivance of port officials, x-ray machine operators, and ship crew and even porters

Cruz et al. 2007. Banwa 4(�):�2-26.

Table 1. Traded wildlife from Southern Palawan confiscated between 2000-2006 (Source: KFI, 2007)

Figure 2. Beetles price per individual (Source: Palawan beetle traders)

�00

90

80

70

60

50

40

�0

20

�0

0 �� �0.9 �0.8 �0.5 �0 9.5 9 8 7

Am

ount

x �

0� (Ph

P)

Size (cm)

Scientific Name Species No. IUCN Endemic in of status Palawan individuals (2007) (√ = yes)

Dorcus spp., Beetles 3,926 √ Odontolabis spp.Gracula religiosa Hill myna 1,522 LR/LC subspecies palawanensisTanygnathus lucionensis Blue-naped parrot 652 LR/NT Freshwater turtles 2�� depending on speciesAnthracoceros marchei Palawan hornbill 38 VU √Gallus gallus Red jungle-fowl �5 Macaca fascicularis Long-tailed macaque 27 LR/NT Haliaetus leucogaster White-bellied sea-eagle 2� Polyplectron emphanum Palawan peacock-pheasant 22 VU √ (napoleonis)Prionailurus bengalensis Leopard cat 21 LR/ LC Manis culionensis Palawan pangolin 18 LR/NT √Aonyx (Amblonyx) cinerea Small-clawed otter 15 NT Tragulus nigricans Balabac mouse deer 15 √Arctictis binturong whitei Palawan bearcat 14 LR/NT √Python reticulatus Reticulated python �4 Cacatua haematuropygia Philippine cockatoo �� CR Sundasciurus steerii Southern Palawan 13 LR/NT √ tree squirrelSpizaetus cirrhatus Changeable hawk-eagle 10 LR/NT Hylopetes nigpripes Palawan flying squirrel 10 LR/NT √Hystrix pumila Palawan porcupine 8 LR/ LC √Spilornis cheela Crested serpent-eagle 5 LR/ LC Tupaia palawanensis Palawan tree-shrew 5 VU √Chalcophaps indica Common emerald dove � LC Accipiter trivirgatus Crested goshawk 2 LC Sus barbatus Palawan bearded pig 1 VU √ ahoenobarbus

Legend: LR: low risk, LC: least concern, NT: near threatened, VU: vulnerable, CR: critically endangered.

�8 �9

Cruz et al. 2007. Banwa 4(�):�2-26.

The blue-naped parrot ranks second among the commonly traded bird species. This species is however sold at a very low price (Php30.00-35.00/individual) due to its abundance in Palawan (Table 2). Its attractive plumage and ability to mimic sounds contributes to its high demand as a pet. The Palawan hornbill is poached for the pet trade or as food. Interviews with poachers revealed that every breeding season (January-June), a minimum of two hornbills – hatchlings and rearing parent bird - are poached per nest. Evidence was found in 2005 when KFI led an apprehension in Sitio Pinatitig in Barangay Culasian and three hornbills – two hatchlings and one parent – were confiscated and turned over to PWRCC. The red jungle fowl is usually caught in snares and is one of the most important sources of bush meat for ethnic groups in Palawan. However, it is also sold alive for interbreeding with fighting cocks (Lacerna and Widmann, �999). Among all recorded traded wildlife in this study, the Philippine cockatoo is the only “critically endangered” species (IUCN, 2007). It was not among the top ten traded wildlife in Southern Palawan from 2000 – 2006 (KFI, 2007), however, the total number of 13 confiscated individuals from 2000 to 2006 is high in relation to its remaining wild world population estimated at �000 individuals (Widmann, 2002). The drastic decline of the cockatoo population in the �990’s was due to hunting, poaching and habitat destruction (Widmann et al., 2005). The poached cockatoos originated mainly from the far south of Palawan particularly from the Balabac Group of Islands and Bataraza where small remnant populations still occur. The hunting is triggered by high market prices starting in Palawan at Php�,000.00-�,000.00 per individual, in Manila it is sold at PhP2,500.00-8,500.00 per individual (van den Beukel et al., 2006) (Table 2).

Due to high market value of the Philippine cockatoo, poachers have developed territoriality over nesting trees. Every poacher respects the “ownership” of nesting trees of other poachers. Robbing each others nesting trees is not traditionally practiced. A poacher will poach all hatchlings of “his” trees every breeding season and will only leave the breeding pair to sustain his income for the next year. For rare species like the Philippine cockatoo that have only a very limited number of breeding pairs in the study area, this easily leads to local extinction once the existing breeding pairs are no longer fertile.

Reptile trade Freshwater turtles are also commonly traded as indicated in the confiscation records of the DENR (KFI, 2007). Earlier studies on the freshwater turtle trade in Southern Palawan had found three species (Cuora amboinensis, Dogania subplana and Cyclemys dentata) that are traded (Gavino and Schoppe, 2004; Regodos and Schoppe, 2005). These studies confirm the report of Widmann (1998) who wrote that populations of C. amboinensis were decreasing in densely settled areas because of local consumption. Cuora amboinensis and C. dentata are sold within Bataraza for local consumption for Php50.00-60.00/individual and D. subplana for Php20.00-50.00/individual (Gavino and Schoppe, 2004). Prices are similar to those reported by informants of this study (Table 2).

Tracking the wildlife trade Information obtained through interviews in Rizal indicate that wildlife trade in Southern Palawan extends to localities in the municipalities of Rizal, Quezon, Brooke’s Point, Bataraza, and Balabac. Poachers are usually members of Palaw’an and Tagbanua ethnic groups who depend mainly on hunting and poaching as source of income. Middlemen buy and deliver the stocks directly to wildlife traders who transport these to Manila or other parts of the Philippines. A former practice was to peddle wildlife along the roads of major cities, or to sell to pet shop or restaurant owners who serve “exotic” food. Nowadays, wildlife goes directly to collectors and pet shop owners, on a cash-on-delivery basis. The income generated from wildlife trade is high and therefore attracts many people in Palawan to participate in the trade. Wildlife is obtained from barangays in the west coast of Palawan like Sowangan, Tagusao, Quinlogan and Sawmill of the Municipality of Quezon and Iraan, Punta Baja, Campong-ulay, Ransang, Upper Culasian, Panalingaan, Taburi, Latud and Canipaan of the Municipality of Rizal (Figure �).

Cruz et al. 2007. Banwa 4(�):�2-26.

Table 2. Prices of some wildlife species in source areas in 2006

Species Local price (PhP)/individual

Hill myna Gracula religiosa 400- 900Blue-naped parrot Tanygnathus lucionensis �0-�5Philippine cockatoo Cacatua haematuropygia �,000-�,000Balabac mouse-deer Tragulus nigricans 600-�200Palawan pangolin Manis culionensis �00Freshwater turtles 20-50

20 2�

Cruz et al. 2007. Banwa 4(�):�2-26. Cruz et al. 2007. Banwa 4(�):�2-26.

are fully fed and dosed with intoxicants in order to prevent them from creating noise. They are restrained by wrapping a damp cloth around their bodies and by putting them inside a bag with cell-like compartments. In an earlier study, de Leon (2005) described that wildlife is transported via motorcycles or shuttle vans to loading areas where hired fishing boats are waiting to transport the cargo to Manila via Batangas or Bulacan. The rent for these boats can be as high as PhP60,000 in addition to the docking fee required by certain resorts. Some years ago birds were still frequently transported onboard passenger boats to Manila or Cebu where the bird cage/boxes were kept in the cabins (de Leon, 2005). Whether on private boats or on passenger boats, loading time is preferably in the evening to minimize detection and stress. Similar observations were made by a trade study of Haribon (2004). The present study showed that several surveillances are usually conducted by traders to ensure that there is no apprehension and docking is usually done in the evening to minimize detection risks. The identified loading areas include eight

From the source areas, wildlife is collected and transported via motorcycles, boat and or jeepneys to stock houses where other wildlife is accumulated until the required volume is reached. A total of �8 stock houses were identified with the help of the informants, nine in Rizal, four in Quezon, two in Brooke’s Point, two in Puerto Princesa City and one in Bataraza (Figure 4). The stock houses are not permanent, and from time to time, wildlife stocks are transferred to other houses within the same areas to prevent detection. A similar pattern was also observed in a World Bank (2005) financed study on illegal wildlife trade in East and Southeast Asia. The present study revealed that most traders operate in Rizal and Bataraza, some having more than one stock house. A total of 17 wildlife traders were identified. A few reportedly have connections with the armed forces, which might explain why they remain relatively untouched by law enforcers. Birds are fed with a diet of dog food, banana and papaya while beetles are kept in tiny labeled, transparent plastic jars and fed with chopped sugarcane or old newspapers. Before transport, birds

Figure 3. Wildlife source areas in Palawan Figure 4. Location of wildlife stock houses in Palawan

22 2�

Cruz et al. 2007. Banwa 4(�):�2-26.

in Southern Palawan (Balabac, Bataraza, Pulot Shore at the boundary of Brooke’s Point and Espanola, and Candawaga coast, Ransang coast, Malakibay coast, Salungsong coast, Bansi coast and Bunog Shore in Rizal), one in Puerto Princesa City (Sitio Tacduan) and four in Northern Palawan, particularly in El Nido, Taytay, Roxas and San Vicente. Some wildlife traders join shipments of mangrove tanbark from Rizal to Zamboanga City for pet shops or street peddling. Sometimes, these reach Sabah, Malaysia in connection with the tanbark and timber trade. The final destinations in the Philippines include Manila, Cebu, Zamboanga City and General Santos City (Figure 5). Manila is the major destination of most birds from the provinces. In Manila, the cargo is picked up directly by pet shop owners in locations preferred by the trader. No direct deliveries are done by traders except for beetles.

Volume of trade Parrots are usually shipped �-� times a month during the breeding season (January-July) and once a month outside of the breeding season. This results in an average of �9 shipments per year. A minimum of 250-�00 heads of birds and numerous other wildlife comprise one shipment, in order to make the trip economically feasible. Hence, a conservative estimate of 4,750 parrots are illegally traded every year. Based on the confiscation records, shipments are usually composed of 70% mynas and �0% parrots. For this number of birds, the collectors earn PhP1,705,250.00 (=USD28,420.00) (i.e., �,�25 mynas at PhP500.00 each and �,425 parrots at PhP�0.00 each). The total number of poachers involved in the business is unknown, but it is apparent that the wildlife trade is a lucrative business, especially at the higher levels of the trade hierarchy (middlemen and traders). Confiscation records from 2000-2006 showed a total of 2,174 confiscated mynas and parrots in seven years (KFI, 2007). This represents a mere 6.5% of the estimated traded number of ��,250 in seven years (4,750 individuals/year) as most shipments go undetected as law enforcement is weak. Moreover, most apprehended traders remain unpenalized. Only a few court cases against traders are pending at present. A World Bank (2005) study of illegal wildlife trade in East and Southeast Asia indicated that even when caught, fines and other penalties are generally much less than the gain from trade. De Leon (2005) stated that prosecution of cases involving wildlife law violation has been likewise difficult, as there is no special court in the Philippines specifically handling cases of environmental law violations. Enforcing laws protecting wildlife is often difficult because there are few and inadequately trained enforcement personnel, and enforcement policies and strategies are ineffective. A major problem is that wildlife smuggling is highly organized, with powerful and influential circles involved.

Conclusion

We conclude that the volume of illegally traded wildlife revealed by this study represents only a very small percentage of the actual volume traded. Understanding the dynamics and complexity of trade as well as enhancing the expertise of grass-roots law enforcers is necessary to curb the illegal trade. In addition to lack of manpower, indifference and corruption, law enforcers have insufficient knowledge on species

Cruz et al. 2007. Banwa 4(�):�2-26.

Figure 5. Wildlife source areas in Palawan

24 25

Cruz et al. 2007. Banwa 4(�):�2-26.

and legal status identification. Therefore, there is a need for training on species and legal status identification for law enforcers. To ensure a more efficient curb on wildlife trade, it is essential that laws are strictly implemented and penalties imposed on violators. Patrols of smaller sea ports and offshore areas need to be increased to detect illegal shipments. Personnel of Anti-Wildlife Poaching Units need to monitor checkpoints, especially during the breeding season. To determine the impact of trade on wild populations, studies on the population/abundance of the main species involved in illegal trade should be conducted. Beissinger and Bucher (�992) recognized the need for novel concepts for co-management of species and ecosystems to achieve sustained natural resource utilization in the face of population growth and poverty. These may include the application of managed resource protected areas in the most vulnerable lowland ecosystems, as well as the application of sustainable use concepts for parrots and mynas in the future, as has been proposed for neotropical parrots for quite some time. This could provide a small, but sustainable and legal flow of live birds for the pet market. In the study area, the establishment of the Culasian Managed Resource Protected Area in Rizal under the SPAPI resulted in a decline in poaching activity inside the �,954 ha of forests in the past two years. The value of an alternative livelihood program for poachers was recognized. Conservation education has also proved to be a powerful tool for changing the perception of poachers towards conservation. Advocacy is also needed to change the attitude of pet lovers and collectors towards wildlife, its products and derivatives. The improvements achieved by the implementation of SPAPI in Culasian should inspire efforts to extend this strategy to other barangays and municipalities in Southern Palawan. Sustainable use schemes for wildlife in the Philippines require an effective control of the illegal and unsustainable trade in wildlife.

Acknowledgment

The authors are indebted to the CEPF who funded the SPAPI. We are grateful to LPF and funding partners. We thank Janice Tupas, KFI volunteers and especially the Wildlife Enforcement Officers of Culasian. Thanks to CI–Palawan for the maps. To Erwin van den Beukel for his untiring support. To Joie Matillano of WPU, DENR-PENRO, Rainier Manalo and Dong Guion of PWRCC, Marge Babon

Cruz et al. 2007. Banwa 4(�):�2-26.

and to the Centre for Environmental Awareness and Education, Philippine Star and PCSDS, our thanks. Thanks to the people of Rizal especially the Palaw’an tribe and Chieftain Adok Desig.

References

Anda, R.D. and J. G Tabangay-Baldera. 2004. Surublien: Strategies to conserve Palawan’s biodiversity. Provincial Government of Palawan, Palawan Council for Sustainable Development Staff, Department of Environment and Natural Resources-MIMAROPA Region IV, Palawan NGO Network, Inc. and Conservation International Philippines, Puerto Princesa City, Philippines. �24 p.

Beissinger, S.R. and E.H. Bucher. �992. Sustainable harvesting of parrots for conservation. In: Beissinger S.R. & N.F.R. Snyder (eds.). New World parrots in crisis – solutions from conservation biology. Smithsonian Institution Press, Washington, USA and London, UK. pp.7�-��5.

De Leon, J. 2005. Development of collaborate mechanism among the various stakeholders to abate illegal wildlife trade in Manila. Unpublished thesis, Master in Public Management, The Faculty of the Master in Public Management Program, Institute of Public Management, Graduate School of Public and Development Management, Development Academy of the Philippines. �5� p.

Diesmos A.C. and N.A. Palomar. 2004. The status of biological diversity in the Palawan corridor. pp �-7. In: Anda, R.D. and J. G Tabangay-Baldera (eds.). Surublien: Strategies to conserve Palawan’s Biodiversity. Provincial Government of Palawan, Palawan Council for Sustainable Development Staff, Department of Environment and Natural Resources-MIMAROPA Region IV, Palawan NGO Network, Inc. and Conservation International Philippines, Puerto Princesa City, Philippines. �24 p.

Gavino, C.M. and S. Schoppe. 2004. First information on the trade of freshwater turtles in Palawan. Agham Mindanaw 2:55-62.

Haribon, 2004. The Haribon campaign on illegal wildlife trade. �0 January 2007. <www.haribon.org.>.

IUCN. 2007. IUCN Red list of threatened species. �� September 2007. <www.redlist.org>.

KFI. 2007. Wildlife confiscation records from DENR and PSCSD for the years of 2000 – 2006. Katala Foundation Incorporated (compiler), Santiago Compound, National Highway, San Jose, Puerto Princesa City, 5�00 Palawan, Philippines.

26 27

Lacerna I.D. and P. Widmann. �999. Biodiversity utilization in a Tagbanua community in southern Palawan, Philippines. In: Goeltenboth, F., P.P. Milan and V. Asio (eds.). Aspects of ecosystems management in tropical Asia. International Conference on Applied Tropical Ecology, ViSCA, Baybay, Leyte, Philippines. Pp. 52-64.

Lasmarias, N. 2004. Profile of threats to biodiversity. In: Anda, R.D. and J.G. Tabangay-Baldera (eds.) Surublien: Strategies to conserve Palawan’s Biodiversity. Provincial Government of Palawan, Palawan Council for Sustainable Development Staff, Department of Environment and Natural Resources-MIMAROPA Region IV, Palawan NGO Network, Inc. and Conservation International Philippines, Puerto Princesa City, Philippines. �24 p.

van den Beukel, D.V., R. Cruz and P. Widmann. 2006. Trapping and hunting of wildlife in Rizal, Palawan, Philippines. Paper presented during the 5th WCSP Annual Philippine Biodiversity Symposium, Legend Hotel Palawan, Puerto Princesa City, April 4-8, 2006.

Widmann, P. 2006. Initial Protected Areas Plan for Culasian Managed Resources Protected Area (CMRPA) and Candawaga Wetland Reserve (CWR). Unpubl. working paper submitted to SPAPI project partners. Puerto Princesa City, Philippines.

Widmann P. and S. Diaz. 2004. Rapid bird survey in mangroves of Northern Bataraza, Palawan, Philippines. A Final Report. Assessment of mangroves and associated fauna in Bataraza and Balabac, Palawan. Proponent: Conservation International Philippines. Pp. �5-50.

World Bank. 2005. Going, going, gone: The illegal trade in wildlife in East and Southeast Asia. World Bank, Environmental and Social Development Department, East Asia and Pacific Region. Pp. �-�0.

Cruz et al. 2007. Banwa 4(�):�2-26.

The Herpetological Importance of Mt. Hamiguitan Range, Mindanao Island, Philippines

Elsa May M. Delima�, Arvin C. Diesmos2,� and Jayson C. Ibañez4

� Corresponding author. Philippine Eagle Foundation, VAL Learning Village, Ruby Street Marfori Heights, Davao City, Philippines. delimaelsa@yahoo.com.

2 Herpetology Section, Zoology Division, National Museum of the Philippines,Padre Burgos Avenue, Ermita �000, Manila, Philippines. arvin.diesmos@gmail.com.

� Conservation Ecology Laboratory, Department of Biological Sciences, National University of Singapore, Block S� �4 Science Drive 4, Singapore ��754�. arvin.diesmos@gmail.com.

4 Philippine Eagle Foundation, VAL Learning Village, Ruby Street Marfori Heights, Davao City, Philippines. falcon2car@yahoo.com.

The Study is part of the “Eastern Mindanao Corridor Biodiversity Archiving and Assessment Project” of the Philippine Eagle Foundation funded by the Critical Ecosystem Partnership Fund.

Abstract

We provide the first accounts of the amphibians and reptiles of Mt. Hamiguitan Range in south eastern Mindanao. Three sites were visited: dipterocarp, transitional dipterocarp-montane and mossy-pygmy forests. The combination of transect sampling, pitfall trapping and microhabitat searches produced records of �4 species (�5 frogs, �4 lizards and five snakes). We provide information on the herpetofaunal assemblage of Mt. Hamiguitan including data on species richness, elevational distribution and microhabitat preferences. High levels of species richness and endemism were observed especially in the dipterocarp forest site located outside the boundaries of the protected area. Our data suggest that Mt. Hamiguitan range should be considered an important subcenter of herpetological diversity. Future conservation efforts should focus particularly on lowland forests.

Keywords: amphibians, conservation, Mindanao Island, reptiles, species richness

Banwa. 2007. 4(�):27-40.

28 29

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

Introduction The herpetology of Mindanao Island in southern Philippines is moderately understood. Herpetological studies on the island have been few and far in between. Among these are the studies conducted by Taylor (�922), the Philippine Zoological Expedition of the Chicago Natural History Museum of �946–�947 (Inger, �954), Brown and Alcala (�986), National Museum of the Philippines-Cincinnati Museum of Natural History “Philippine Biodiversity Inventory” (Brown et al., 2000; Kennedy et al., �997) and lizard and snake accounts in select areas of eastern Mindanao (Smith, �99�a and �99�b). There are subsequent studies done in the area but these remain unpublished. Therefore, there exists a large gap of knowledge of the Mindanao herpetofauna. Mindanao island forms the majority of the Mindanao Pleistocene Aggregate Island Complex (PAIC) and is considered as one of the country’s major centers for species endemism (Diesmos et al., 2000). Its eastern side is reported to harbor one of the largest remaining forest blocks in the country today. Recent biodiversity conservation priority-setting efforts have identified 34 priority conservation sites within this region (Ong et al., 2002). Because of the presence of these important biodiversity sites and the continuous threat of habitat destruction, Eastern Mindanao is considered an area noteworthy of immediate conservation action. The site was declared an important biological corridor, here referred to as Eastern Mindanao Corridor, bounded by Siargao and Dinagat Island in the north while the western portion encompasses Agusan Marsh and the southern-most portion includes Mount Hamiguitan (Critical Ecosystem Partnership Fund, 200�). We undertook this project to provide scientific baseline information on biological diversity of two proposed protected areas within Eastern Mindanao (Mt. Hilong-Hilong, and Mt. Tagub-Kampalili), and the existing protected area Mt. Hamiguitan. This study provides baseline information on species richness and composition of the herpetofauna of Mt. Hamiguitan. Additionally, we include new information on natural history and microhabitat preference for several species.

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

Materials and Methods

Study area Mt. Hamiguitan is located in southeastern Mindanao. This range is a combination of rugged mountains and plains including Agustin Peninsula with a peak of about �650 m.a.s.l. (Mallari et al., 200�). Aside from having lowland, montane and mossy habitat types, this range also includes an extensive “bonsai forest,” a product of ultramafic soils. The range receives adequate rain throughout the year with the highest rainfall between November and January. Average annual rainfall is between �500 to 2500 mm. By virtue of Republic Act 9�0� of 2004, a total of 6954 ha of Mt. Hamiguitan had been declared as a Protected Area under the category Wildlife Sanctuary (Supreme Court E-Library, 2004). Our surveys were conducted in the municipalities of San Isidro and Mati, Davao Oriental Province (Figure �). Surveys were conducted in the same areas where faunal (birds and mammals) and floral surveys were conducted. Descriptions of each site are given below. Site 1 The site (N 06° 44’�.4”, E �26° 09’�.4”) is a secondary growth dipterocarp forest with an elevation of 545–785 m a.s.l. located at Purok Palo X, Sitio Tumalite, Barangay La Union, Municipality of San Isidro, Davao Oriental. The site is dominated by “malabayabas,” trees of the family Myrtaceae which have a diameter at breast height (DBH) range of 6 to 86 cm. The forest floor had a thick layer of dry leaf litter with tree stumps and abundant rotting logs; streams and creeks were also present (both dried-up and active). Habitat disturbance includes selective logging in the past and more recently slash and burn agriculture. Our sampling plot was outside the wildlife sanctuary and is part of a Community Based Forest Management site (CBFM), accessible by a � to 4 h hike through a forest trail. Sampling for this site was conducted from �7 to 24 May 2005.Site 2 The site (N 6° 4�’56”, E �26° �0’4�”) is a transitional dipterocarp-montane forest near Tinagong Dagat (a wide area covered by cogonales that accumulates rain and ground water) with an elevation of 950 to �200 m.a.s.l. and straddles right at the political boundaries of Barangay La Union of San Isidro and Barangay Macambol of Mati, Davao Oriental. Vegetation was dominated by species belonging to the Families Podocarpaceae (Dacrydium elatum), Sapotaceae (Palaquium sp.), Araucareaceae (Agathis philippinensis or Almaciga) and Myrtaceae

�0 ��

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

(Syzigium sp.). Climbing vines such as Freycinetia sp., tree ferns and pitcher plants were abundant and fruiting Medinilla sp. was observed. The forest floor had a thick leaf litter with a substrate consisting of a layer of humus. The area is accessible by a ca. 6 hour hike through a forest trail that passes the Alog river from Sitio Magum. The site was visited from �8 to �� July 2005. Site 3 The site (N 6° 4�’�”, E �26° ��’�.9”) is a mossy-bonsai forest on the eastern slope of Tumadgo peak with an elevation of ��28 to �4�5 m a.s.l., locally referred to as Camp � in Sitio Tumalite, Barangay La Union, Municipality of San Isidro, Davao Oriental. Vegetation was dominated by Agathis stands, with moss covering tree branches and trunks extending even down to the ground. Forest floor was covered by a thin layer of leaf litter; substrate was clay with a reddish color possibly due to the high mineral content. Flowering and fruiting trees were observed. Sampling for this area was from 4 to �4 May 2006.

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

Sampling techniques We employed belt transect sampling, pitfall trapping and microhabitat searches (Heyer et al., �994). A total of 50 transect lines and 50 pitfalls were established in all sites. With the exception of Site �, 20 transect lines and 20 pitfalls were stationed per site. Each transect was �00 m long and �0 m wide and each �0 m interval was marked with ribbons and was walked twice: morning (0900 to �200 hours) and evening (�800 to 2�00 hours). In order to minimize observer disturbance, a one-day interval allowed to elapse between transect walking (transects that were traversed in the morning were not walked in the evening for the same sampling day). Searchers relied on sight and calls to locate species. Samples were then hand-captured whenever possible. An average of 0.57 km per sampling day (10.5 person-h/day) was sampled. For pitfall trapping, dry-type pitfall traps with drift fences utilizing both straight and array designs (Heyer et al., �994) were used; these were checked thrice daily. Transects and pitfalls were randomly stationed in each of the sites in an effort to cover all possible areas where amphibians and reptiles reside. To evaluate adequacy of sampling effort employed, species effort curve plots were made (Brower et al., 1990). Seven sampling days were allotted for each of the sites. Five samples per species per transect were collected as vouchers for identification purposes. Amphibian and reptile species were keyed out using Alcala and Brown (�998), Brown and Alcala (�978), Brown and Alcala (�980) and Inger (�954). Liver samples of preserved specimens were also taken. Tissue samples (liver) are currently deposited at the Molecular Biology Laboratory of the University of the Philippines in Mindanao, Davao City and voucher specimens were deposited at the National Museum of the Philippines and Central Mindanao University Museum of Natural History.

Ecology and life history Information on altitudinal distribution, relative abundance and microhabitat preferences of species was recorded. Activity of individual species was also noted. For relative abundance, encounter rates were calculated by dividing the total number of captures of all the searchers by the total number of hours spent for sampling. Encounter rates were then evaluated using an ordinal scale formulated by Lowen et al. (�996 as cited in Bibby et al., �998). Microhabitats where specimens were captured/observed were also classified into three broad categories: aquatic, arboreal and terrestrial. The microhabitats were differentiated based on vertical

Figure 1. Map of Mt. Hamiguitan showing sampling sites

�2 ��

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

distance from the ground and proximity to water bodies. Aquatic microhabitats referred to those areas in or near banks of water bodies and pools. Terrestrial microhabitats included locations 0 to 5 m above ground and below the ground (burrowing species). Areas of more than 5 m above ground were treated as arboreal microhabitats.

Results and Discussion

Species richness, composition and relative abundance Thirty four species of amphibians and reptiles were recorded for the first time on Mt. Hamiguitan, including 15 frogs, 14 lizards and five snakes (Table 1). Endemism appears high with 25 species (74%) classified as Philippine endemics, and 38% (13 species) are confined to Mindanao PAIC. The data also account for the presence of poorly-known and site-specific endemic species previously reported from parts of Davao (Tropidophorus davaoensis); and Mt. Hilong-Hilong (Philautus poecilius [Alcala and Brown, �998; Brown and Alcala, �994], and Brachymeles gracilis hilong [Brown and Alcala, �980]). Samples were collected appearing to be representative of two species known to occur in Mt. Hilong-hilong: Limonectes cf diautus (Brown and Alcala, �977) and Sphenomorpus cf diwata (Brown and Alcala, 1980). Six of the anuran species are classified as Vulnerable and one is considered Near Threatened (Global Amphibian Assessment, 2004). Two of the species sampled will require further taxonomic study; they may either represent new species or new country records. One is a lizard that has close resemblance to Sphenomorphus diwata, a scincid recorded from Mt. Hilong-hilong. Our samples are distinct from published data for S. diwata by having smaller snout-vent-length (SVL) measurements and our samples were caught at lower elevations, contrasting with published data (Brown and Alcala, �980). Future study will be required to determine if our specimens are new species or simply an extension of the normal range of variation in S. diwata. Interestingly, about 20 specimens of ranid frogs initially identified as Starois natator appear to be representative of Meristogenys, a genus not currently recognized in the country (Alcala and Brown, �998; Inger, �954). Our tentative claim on the samples as potential Meristogenys representatives rely on the habitat as well as morphological similarities of established species. The clear rocky stream with swift flowing water where we took our samples is similar to the habitat of most Amolops (=Meristogenys) species (Inger and Kotellat, �998). The slender body, long head as well as

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

Table 1. The amphibians and reptile species documented on three sampling sites within Mt. Hamiguitan Range from May 2005 to May 2006

Species Endemisma Siteb Dietc Relatived Microhabitat/se Status/Remarkf

activity abundance

Anurans (Bufonidae) Ansonia muelleri (Mueller’s toad) ** � N, D Rare Rocks of clear (Aq) Vulnerable mountain streamsMegophryidae Leptobrachium 1, 3 N, D Rare Atop leaf litter (G); Least Concern cf hasselti creek bank (Aq) Megophrys stejnegeri ** 1, 2 D, N Rare Atop leaf litter (G) Vulnerable (Mindanao horned frog)Ranidae Limnonectes cf ) ** � N Rare Swimming on Vulnerable diuatus (Tagibo river (Aq) Identification needs warty frog further verification Limnonectes magnus ** �, � N, D Rare Creek bank (Aq); Near Threatened (Mindanao fanged Consumed as food frog) Platymantis corrugatus * � N, D Rare On heap of forest Least Concern (Rough-backed leaf litter (G) forest frog) Platymantis guentheri ** 2 N Rare Interior of branches Vulnerable (Guenther’s forest that crossed (A) frog) Rana everetti * 2 N Rare Pool of water (Aq) Data Deficient (Everett’s frog) Rana grandocula * �, 2 N Rare Pool of water and Least Concern (Big eyed frog) river (Aq)

Starois natator 1 D, N Rare Rocks on stream Identification needs (Rock frog) (Aq) verification as 20 of the samples appear to be representative of the genus Meristogenys, which is not from the PhilippinesRhacophoridae Philautus acutirostris ** 2, � N Rare One sample on Vulnerable (Pointed-snouted forest floor (G); tree frog) atop fern fronds (G) Philautus surdus * 2, � N Rare Axils of tree leaves Least Concern (Common forest (A) tree frog) Philautus poecilius ** 2 N Rare Axils of tree leaves VulnerablePotentially (Mottled tree frog) (A) new distribution record Philautus sp. 2, 3 N Rare Axils of tree leaves Identification needs (A) further verification Polypedates leucomystax �, 2 N, D Rare Axils of leaves (A) Least Concern (Common tree frog)Lizards (Agamidae) * � D Rare Basking on tree Draco bimaculatus branches (A) (Two-spotted flying lizard) Gekkonidae Cyrtodactylus * � D Rare Bark of a dead tree annulatus (Small (<5 m from the bent-toed gecko) ground) (G) Cyrtodactylus * 1 D Rare Bark of rotting tree philippinicus (G); forest duff near (Philippine bent- river (G) toed gecko)

�4 �5

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

presence of structures inside the mouth appearing as vomerine teeth, which are typical of Meristogenys species from Borneo (Malkmus et al., 2002), were also found in our samples. If future comparison with specimens from Borneo validates this tentative finding, the samples from our study may represent a new country record of Meristogenys. A distinct plateau in the species accumulation curve was not reached for all sites (Figure 2) and encounter rates for the species appear too low so they were graded as either uncommon or rare. This is probably a product of the non-asymptotic effort curve. Future studies should extend the days of sampling in an effort to increase the number of species and individuals for each species. With the observed pattern for species effort curve and low encounter rate, it is highly possible that the species list presented here is incomplete. Thus, all three study sites certainly warrant additional survey works before one can say that all resident species have been identified.

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

Species distribution Inverse relationship between species richness and increasing elevation was observed as 25 species of the �4 encountered were observed in Site � (545 to 785 m a.s.l.). This was followed by �4 species for Site 2 (950 to �200 m a.s.l.) and lastly, �� species recorded in Site � (��28 to �4�5 m a.s.l.). Only three species (Sphenomorphus coxi, Rhabdophis auriculata and Psammodynastes pulverulentus) were encountered in all of the sites and only P. pulverulentus is non-

Table 1. con’t

Figure 2. Relationship of sampling effort (days) and the number of species encountered at Mt. Hamiguitan Range.

Lowland(545 - 785 m.a.s.l.)

Montane(950 - 1200 m.a.s.l.)

Mossy(1128 - 1435 m.a.s.l.)

30

25

20

15

10

5

0

0 1 2 3 4 5 6 7

Sampling Days

No.

of S

peci

es

Species Endemisma Siteb Dietc Relatived Microhabitat/se Status/Remarkf

activity abundance

Scincidae Brachymeles gracilis ** � D Rare Caught in pitfall in Potentially new hilong(Common an area with thick distribution record an area with thick distribution recordan area with thick distribution record distribution record burrowing skink) leaf litter and loose soil (G) Brachymeles ** � D Rare Caught in pitfall in schadenbergi areas with logs and orientalis (Southern thick leaf litter (G) burrowing skink) Lamprolepis smaragdina � D Rare Thicket of vegetation (Green tree skink) (G) Lipinia pulchella * 1 D Rare Leaf litter on forest (Yellow-striped floor (G) slender tree skink) Lipinia quadrivitatum � D Rare Grounds near the (Black-striped slender camp (G) tree skink) Sphenomorphus coxi * 1, 2, 3 D, N Rare Duff on forest floor (Cox’s sphenomorphus) (G) Sphenomorphus * 1 D Rare Duff on forest floor decipiens (Black- (G) sided sphenomorphus) Sphenomorphus cf ** 1, 3 D Rare Duff on forest floor Identification needs diwata (Diwata further verification sphenomorphus) Potentially new distribution record Sphenomorphus * 2 D Rare Rotting logs near fasciatus (Banded camp (G) sphenomorphus) Sphenomorphus ** � D Rare On tree branches (A); variegatus (Black forest litter on floor spotted (G) sphenomorphus) Tropidophorus ** 2 D Rare Caught in snap traps Potentially new davaoensis located on dead distribution record (Waterside skink) creek with boulders (G)Snakes Colubridae Psammodynastes �, 2, � D, N Rare Coiled in a shrub or pulverulentus (Dark- near the ground on spotted mock viper) low vegetation (G) Rhabdophis ** �, 2, � D Rare On low lying auriculata (White- vegetation (G); creek lined watersnake) (Aq) Rhabdophis lineata � D Rare Crawling atop leaf (Zigzag-lined water litter on floor (G) snake)Elapidae Maticora intestinalis � D Rare Crawling near low philippina (Striped lying vegetation (G) coral snake) ViperidaeTrimeresurus * �, � D, N Rare Ground near creek (Aq) flavomaculatus (Philippine pit viper)

a Endemism: ** - Mindanao Faunal Region Endemic; * - Philippine Endemicb Site: � – Tumalite; 2 – Tinagong Dagat; � – Camp �c Diel Activity: D – diurnal; N – Nocturnald Relative Abundance is based on the calculated encounter ratee Microhabitats: A – arboreal; G – terrestrial, Aq – Aquaticf Status conforms with CITES 2006, IUCN 2006 and Global Amphibian Assessment (globalamphibiansassessment.org)

�6 �7

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

endemic. Results present no distinct pattern for endemism as more than 50% of the species recorded for each site are endemics. Non-direct developing frogs were abundant in Site � whereas most direct developing frogs where recorded in higher elevations sites (Sites 2 and �). Reptiles (mostly lizards) also appear abundant in Site �. Parallel results were observed in Luzon (Brown and Alcala, �96�; Brown et al., 2000; Diesmos et al., 200�). A congregation of factors best explains these although effects of temperature and humidity affecting egg development and thermal physiology may also be important, especially for reptiles (Navas, 2002).

Ecology and life history Eighty-six percent of the species recorded are forest dwellers. Many are endemics and were encountered in the lowest elevation site (Site �). Species were found to occupy various types of microhabitats (Table �) though the majority was observed on the ground (leaf litter microhabitat was favored by most). For aquatic microhabitats, more species and individuals were found in rivers and streams that have constant, free flowing and clean waters as evidenced by our captures of A. muelleri, L. magnus, and L. cf diautus. In contrast, few individuals of R. everetti and R. grandocula were captured in small standing water pools near our camp. Arboreal frog species were found in leaf axils of trees in deep forest quite far from disturbances. The observed abundance of species (mostly endemics) and corresponding low number of individuals in slightly disturbed or undisturbed sites suggests that species abundance is influenced by forest disturbance. Endemics are likely found in areas with minimal disturbance (see also Diesmos et al., 200�; Brown et al., 2000). Microhabitat data reveal that most reptiles stay on the ground, associated with leaf litter. Presence of reptile species (especially lizards) in leaf litter has been associated with thermoregulation (Alcala and Brown, �966) and foraging (Inger, �980). Thus, the dominance of ground microhabitat preference in most reptiles is not surprising. For anurans, microhabitat preferences varied among sites. In Site �, species accumulated either near a creek or a river, many of which were non-direct developers while arboreal (mostly direct developers) species abound in high elevation forests of Site 2 and Site �. Non-direct developers (Alcala and Brown, �998; Inger, �954) were found to congregate within close proximity to areas with constantly filled bodies of water (water either flowing such as

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

rivers and creeks or isolated water pools). Because of their constant need for water, species recorded in Site � were found in or close to water. In contrast, most species recorded for Sites 2 and � exhibits direct development of their young (Alcala and Brown, �998; Inger �954). This developmental mode eliminates the reliance for constant water source and thus most of the species accounted were found in arboreal microhabitats such as leaf axils and branches, which are quite far from water bodies. Direct developing species in these sites were recorded in small, isolated water pools and narrow creeks, which were filled with water only during rainy days. The current effort in archiving the herpetofaunal assemblage of Mt. Hamiguitan highlights formerly unknown species of amphibians and reptiles that inhabit its forest. Interestingly, the study revealed the presence of species known only from the highlands of Northeastern Mindanao and portions of Southern Mindanao. The majority of these species reside in dipterocarp forest at lower elevations. It is hoped that our data will contribute to ongoing efforts to formulate schemes for conservation efforts. The herpetofauna of Mt. Hamiguitan is far from completely known but our work will contribute to the slowly accumulating pool of available data for conservation and management purposes.

Conclusion

The �4 species of frogs and reptiles accounted in this study highlights the herpetological richness of Mt. Hamiguitan. Our non-asymptotic species accumulation curve implies that more species await discovery especially if other areas can be surveyed in the future such as those located in the Municipality of Governor Generoso. The presence of more than 50% endemism and numerous potentially threatened species further expressed the herpetological importance of this mountain range. Moreover, because diversity is highest at lower elevations, we urge evaluation of new conservation strategies aimed at conserving the unique forests of Mt. Hamiguitan.

Acknowledgements

The authors would like to express their gratitude to the Critical Ecosystem Partnership Fund for the financial support of the Eastern Mindanao Corridor Biodiversity Archiving and Assessment Project (EMCBAAP) and also acknowledge the support given by the board and managers of the Philippine Eagle Foundation. We also thank

�8 �9

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

the Department of Environment and Natural Resources XI and Community, Environment and Natural Resource Office of Lupon for granting the necessary permits. Our entry to the sites would not have been possible without the partnership of the local government units of San Isidro, Mati and Governor Generoso. The invaluable help of our parabiologists especially to Mr. Valentin Jimenez, Alfredo Volante Sr., and Amay Jimenez who eagerly devoted their time searching for herps is also acknowledged. Permission to store tissue samples was given by Dr. Severo Bastian of UP Mindanao. Special compliments are also given to Giovanne Tampos, Joshua Donato, and Roselyn Quidlat who patiently helped in specimen preparation. And to all who have helped shape this paper especially our reviewers Dr. Rafe Brown and Sabinne Schoppe and to the editorial board, thanks a lot!

References

Alcala, A. and W. Brown. �966. Thermal relations of two tropical lizards on Negros Island, Philippine Islands. Copeia �:59�-594.

Alcala, A. C. and W. Brown. �998. Philippine amphibians: An illustrated field guide. Angel C. Alcala, Walter C. Brown and Bookmark, Philippines.

Bibby, C., M. Jones, and S. Mardsden. 1998. Expedition field techniques bird surveys. Expedition Survey Advisory Center, London.

Brower, J., J. Zar and C. von Ende. �990. Field and laboratory methods for general ecology �rd ed. Wm. C. Brown Publishers, USA.

Brown, R., J. McGuire, J. Ferner, N. Icarangal Jr., and R. Kennedy. 2000. Amphibians and reptiles of Luzon Island II: Preliminary report on the herpetofauna of Aurora Memorial National Park, Philippines. Hamadryad 25(2):�75-�95.

Brown, W. and A. C. Alcala. �96�. Population of amphibians and reptiles in the submontane and montane forests of Cuernos de Negros, Philippine Islands. Ecology 42(4):628–6�6.

Brown, W and A. Alcala. �977. A new frog species of the genus Rana from the Philippines. Proc. Biol. Soc. Wash. 90(�0):669–675.

Brown, W. and A. Alcala. �978. Philippine lizards of the family Gekkonidae. Silliman University Press, Dumaguete City.

Brown, W. and A. Alcala. �980. Philippine lizards of the family Scincidae. Silliman University Press, Dumaguete City.

Brown, W. and A. Alcala. �986. Comparison of the herpetofaunal species richness of Negros and Cebu Islands, Philippines. Silliman J. ��(�-4):74–86.

Brown, W. and A. Alcala. �994. Philippine frogs of the family Rhacophoridae. Proc. Calif. Acad. Sci. 48(�0):�85-220.

Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES). 2006. Appendices I, II, and III. �� June 2006 <www.cites.org>.

Critical Ecosystem Partnership Fund. 2001. Ecosystem profile: Philippines. 20 June 2006. <http://www.cepf.net/xp/cepf/where_we_work/philippines/full_strategy.xml>.

Diesmos A. and the Herpetofauna Working Group. 2000. Philippine amphibians and reptiles: An overview of diversity, biogeography and conservation. Paper presented at the National Biodiversity Conservation Priority Setting Workshop. White Rock Hotel, Subic, Philippines.

Diesmos, A., R. Brown and G. Gee. 200�. Preliminary report on the amphibians and reptiles of Balbalasang-Balbalan National Park in Luzon Island, Philippines. Slyvatrop ��(�&2):6�–80.

Heyer,W.R., M.A. Donnelly, R.W. McDiarmid, L.A.C. Hayek, and M.S. Foster. �994. Measuring and monitoring biological diversity: Standard methods for amphibians. Smithsonian Institution Press, Washington, D.C.

Inger, R.F. �954. Systematics and zoogeography of Philippine amphibia. Fieldiana ��:�8�-5��.

Inger, R.F. 1980. Densities of floor-dwelling frogs and lizards in lowland forests of Southeast Asia and Central America. The American Naturalist ��5(6):76�-770.

Inger, R and M. Kottelat. 1998. A new species of ranid frog from Laos. Raffles Bul. Zool. 46(1):29-34.

International Union for the Conservation of Nature, Conservation International and Nature Reserve. 2004. Global amphibian assessment. 8 June 2006. <http://www. Globalamphibians.org>.

Kennedy, R.S., P.C. Gonzales, and H.C. Miranda Jr.�997. New aethopyga sunbirds (Aves: Nectariniidae) from the Island of Mindanao, Philippines. The Auk ��4(�):�–�0 + frontispiece.

Malkmus, R., U. Manthey, G. Vogel, P. Hoffmann, and J. Kosuch. 2002. Amphibians and reptiles of Mount Kinabalu (North Borneo). Koeltz Scientific Books, Germany.

Mallari, A., B. Tabaranza Jr., and M. Crosby. 200�. Key conservation sites in the Philippines. Bookmark, Manila.

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40.

40 4�

Navas, C.A. 2002. Herpetological diversity along Andean elevational gradient: Links with physiological, ecology and evolutionary physiology. Comparative Biochem. Physiol. Part A: Molecular and Integrative Physiology ���(�):469–485.

Ong, P.S., L.E. Afuang, and R.G. Rosell-Amball (eds). 2002. Philippine biodiversity conservation priorities: A Second Iteration of the National Biodiversity and Action Plan. DENR-Protected Areas and Wildlife Bureau, Conservation International Philippines, Biodiversity Conservation Program-UP CIDS, and the Foundation for the Philippine Environment, Quezon City, Philippines.

Smith, B. �99�a. Notes on a collection of squamate reptiles from Eastern Mindanao, Philippine Islands Part �: Lacertilia. Asiatic Herpetological Res. 5:85-95.

Smith, B. �99�b. Notes on a collection of squamate reptiles from Eastern Mindanao, Philippine Islands Parts: Serpentes. Asiatic Herpetological Res. 5:95-�02.

Supreme Court E-Library. 2004. Republic Act No. 9�0� An act declaring Mt. Hamiguitan range and its vicinities as Protected Area under the category of Wildlife Sanctuary and its peripheral areas as buffer zone and appropriating funds therefore. 20 September 2007. <http://elibrary.supremecourt.gov.ph.>.

Taylor, E.H. �922. The lizards of the Philippine Islands. Philippine Bureau of Science, Monogr. �7. Manila, Philippines.

Delima, Diesmos, and Ibañez. 2007. Banwa 4(�):27-40. Banwa. 2007. 4(�):4�-68.

Report on a Survey of Mammals of the Sierra Madre Range, Luzon Island, Philippines

Mariano Roy M. Duya1,3 Philip A. Alviola2, Melizar V. Duya3, Danilo S. Balete4, and Lawrence R. Heaney5

� Corresponding author. mduya@conservation.org2 UPLB Museum of Natural History, University of the Philippines Los Baños,

College, Los Baños, Laguna. flip_alviola@yahoo.com.� Conservation International Philippines, 6 Maalalahanin St., Teachers Village,

Diliman Quezon City, Philippines. mvduya@conservation.org.4 Laksambuhay Inc., College, Laguna. dsbalete@gmail.com.5 The Field Museum, �400 S Lakeshore Drive, Chicago, Illinois 60605 USA.

heaney@fieldmuseum.org.

This study was undertaken as part of the project entitled “Conservation of Key Biodiversity Areas within the Sierra Madre Mountain Range – Luzon Islands Philippines” funded by the Critical Ecosystem Partnership Fund.

Abstract

An inventory of mammals was undertaken at �� localities along the Sierra Madre range in 2002 to 2005, in areas where few or no data were available previously. The inventory included lowland and montane forest habitats, at elevations from �00 to �500 m. Thirty-eight species of mammals were recorded, including nine new records for the mountain range. One species, Kerivoula cf. papillosa, had not been recorded previously from the Philippines, and one, Coelops hirsutus, was known previously only from Mindanao and Mindoro Islands. Two species, in the genera Apomys and Chrotomys, may represent previously unknown species. We captured Archboldomys musseri only on Mt. Cetaceo, supporting previous evidence that it is endemic only to that mountain. A modified mist-netting technique (V-net) for insectivorous bats was effectively used to capture these species. The new records clearly demonstrate that the mammalian fauna of the Sierra Madre is poorly known. Surveys of many additional areas are needed in all known habitat types along the Sierra Madre, especially karst, ultrabasic, and mossy forest, to fully document its diversity.

Keywords: Biodiversity, conservation, endemic, Luzon, mammals, Philippines, Sierra Madre

42 4�

Duya et al. 2007. Banwa 4(�):4�-68.

Introduction

Although the mammals of Luzon Island were among the first in the Philippines to be surveyed, nearly all of the studies were conducted in the Southern Central Cordillera, on or near Mt. Data (Hoogstraal, �95�; Mearns, �905; Sanborn, �952; Thomas, �898). In the �960s to �980s, some highly distinctive murid rodents were subsequently recorded in surveys of Mt. Isarog, on the Bicol Peninsula of Southern Luzon (Balete and Heaney, �997; Heaney et al., �999; Musser, �982; Musser and Freeman, �98�; Rickart and Heaney, �99�; Rickart et al., �99�, �998;). Extensive surveys were conducted in Kalinga Province in the northern Central Cordillera (Heaney et al., 2005), and on Mt. Tapulao in Zambales Province (Balete et al., 2007). Several studies of bats have been conducted on Mt. Makiling, Laguna Province (Ingle, �992; Sedlock, 200�); while other parts of Luzon Island remain poorly documented (Heaney and Mallari, 2002). The mammals of the Sierra Madre range of Eastern and Northeastern Luzon were almost entirely unknown until the �980s when surveys were initiated. These include a brief survey of bats along the coast of Isabela and Cagayan provinces (Mudar and Allen, �986), a brief survey of bats and rodents on Mt. Cetaceo (Danielsen et al., �994), and a hunter-based survey of cloud rats (Oliver et al., �99�), which provided a majority of the records (Heaney et al., �998). An assessment of the adequacy of mammalian diversity data in the Philippines showed that little information exists for any portion of the Sierra Madre, and no information at all on the mammals of Quirino and Quezon provinces (Heaney et al., 2002). Given the size and the relatively intact forest of the Sierra Madre and the discovery of a new species of small mammal, Archboldomys musseri, that is likely to be endemic to the Sierra Madre (Balete et al., 2006; Rickart et al., �998), the need for additional information is evident. Interest in the Sierra Madre as a globally important area for biodiversity conservation increased with its recognition as a conservation priority area during the Philippine Biodiversity Conservation Priorities Program (Ong et al., 2002; see also Heaney and Mallari, 2002). The program focused on building consensus on the location of conservation priority areas in the country (Brooks et al., 2004). In this paper, we present the results of our surveys of mammals of the Sierra Madre in 2002 to 2005.

Duya et al. 2007. Banwa 4(�):4�-68.

Materials and Methods

Mist-netting of bats Mist-nets with an average mesh size of �6 mm and height and length of 2.5 m and 12 m were used to catch bats. Mist-netting stations consisted of a series of 5-�0 mist nets (60-�20 m long) that were left open for three to four consecutive nights, then transferred to another mist-netting location. Nets were set with the bottom edges of the nets about 0.� meters above the ground. Mist-nets were placed along ridges and streams, and in forest and clearings, especially in possible flyways of bats. Nets were checked in early morning for netted bats and continuously in the evening for two hours (�800-2000 h), and at �0 min to � h intervals until 2200 h in the evening. Specimens captured were identified to species level in the field, and we recorded sex, age, individual reproductive condition, mammae size, tail length, forearm, hind foot, ear, total length, and noseleaf width (for relevant bats) as defined in Ingle and Heaney (�992). In addition to the traditional mist-netting technique, we also employed a novel mist-netting technique designed solely for insectivorous bats, which we called the “V-net”. We arranged two mist-nets in a V configuration wherein one end of each net is attached to a common secured pole. One arm of the V-net is fixedly positioned and serves as a wall whereas the other arm is mobile. The V-net works along a principle similar to tunnel trap with both arms of the V serving as the opening (Alviola, 2000; Sedlock, 200�). As the V-net is placed along a trail or stream, one person is assigned at the mobile arm of the V and another person stands at its side along the wall. When a bat enters, the mobile arm is shut quickly, causing the bat to be trapped and entangled along the interior of the wall.

Trapping of rodents A mixture of locally manufactured cage traps and Victor rat traps was used at Sites 2, �, 5, 7, and 8. Trapping of non-volant mammals was not done at Site �, as no traps were available at the time, and we used only Victor rat traps at Sites 4, 6 and 9. A mixture of Victor rat traps and Museum Specials was used at Site �0 and ��. Traps were usually placed on the ground near fallen logs or holes, along suspected runways, and at openings among roots of trees and stumps. Some were placed on the branches of trees above the ground to capture arboreal murids. Traps were spaced at 5-�0 m intervals;

44 45

Duya et al. 2007. Banwa 4(�):4�-68.

each trap line was operated for four to five days. Traps were baited twice each day, during early morning and late afternoon, using thin slices of roasted coconut coated with peanut butter. Live earthworms were also used as bait on some traps at Sites �, 4, 6, 7, 9, �0 and ��. Animals caught in live traps were released at the site of capture unless voucher specimens were needed. For captured individuals, we recorded the: length of tail vertebrae, hind foot, head and body, total length, weight, mammae number, and reproductive condition. Nomenclature was based on Heaney et al. (�998) and Musser and Heaney (�992). Voucher specimens (listed below as “specimens examined”) of rodents and bats were prepared for study, cataloged, and identified at the Field Museum, with half to be returned to the National Museum of the Philippines.

Additional methods Direct observation of snares called silo in Tagalog set by local people, footprints, fecal droppings, nests, and animal remains were noted. Animals traded in nearby villages, especially if these animals were taken within the survey site, were also noted. Interviews with local inhabitants (e.g. local guides, hunters, and residents) were conducted using color photos of mammals; we noted local names for animals, perceived abundance and frequency, month and season observed, behavior, and economic importance. Extent of hunting, trapping, and other forms of gathering were noted as well. Distributions, habitat associations, and English common names are from Heaney et al., (�998) unless noted otherwise. Locations are given in order by elevation, from low to high (Figure �). Longitude and latitude were determined with using a GPS Garmin eTrex Summit. Specimen examined per site were also indicated in the species accounts.

Results and Discussion

Study sites

Site 1. Blue Waters, Sitio Pallagao, Barangay Sta Margarita, Baggao Municipality, Cagayan Province (�8° 0�’ ��” N, �2�° 59’ 26” E, 4-6 June 200�). Mist netting was done at the edge of a moderately disturbed limestone forest at �00 m. Several shallow caves located within the vicinity were checked for presence of bat colonies. Only the Blue Water Cave was inhabited by a colony of bats (Hipposideros diadema),

Duya et al. 2007. Banwa 4(�):4�-68.

Figure 1. Map showing the location of the different survey sites along the Sierra Madre mountain range

46 47

Duya et al. 2007. Banwa 4(�):4�-68.

numbering about �50 individuals. Slash and burn farming was common in the area. Commercial logging was cancelled in the area in �989 but some timber poaching occurred subsequently.

Site 2. Botanical Garden, Sitio Mansarong, Barangay Sta. Margarita, Baggao Municipality, Cagayan Province (�7° 56’ 24.05” N, �22° 0�’ 29.�2” E, �-� June 200�). Netting and trapping were done in moderately disturbed secondary lowland forest and along the forest edge at �00 m elevation. Forest trees were predominantly Dipterocarpaceae. Shrubs and saplings were mostly Euphorbiaceae, Dipterocarpaceae, Myrsinaceae, Sapindaceae and Rubiaceae. The area was logged �5 years ago and is currently protected by the local community and a small band of Agta. The site is adjacent to a slash and burn farms maintained by the Agta and an old logging road used by the community.

Site 3. Mt. Twin Peaks, Sitio Matulang, Barangay Sta. Margarita, Baggao Municipality, Cagayan Province (�7° 52’ 5�.2” N, �22° 06’ �4.5” E, 2-�� May 200�; �8-22 September 2004). We surveyed 2 sub-sites along the Tabuan river. The first sub-site was in slightly disturbed secondary lowland forest at about �00 m elevation. Forest trees were dominated by Dipterocarpaceae with emergents reaching maximum heights of �5 to 40 m, and the canopy was estimated at �5 to 20 m. Diameter breast height (DBH) ranged from �5 to 50 cm diameter. Undergrowth consisted mainly of ferns and gingers. Epiphytes were chiefly Araceae and ferns. Lianas consisted mostly of Freycinetia spp. and Ichnocarpus spp. Shrubs and saplings included Sapotaceae, Dipterocarpaceae, Gutifferae and Euphorbiaceae. At the second sub-site, we conducted only mist-netting along the forest edge at about 300 m elevation. Timber poaching and patches of slash and burn farms were observed at both locations. Logging in the area was cancelled in �99�.

Site 4. Mt. Mungiao, Sitio Mangitagud, Barangay Matmad, Nagtipunan Municipality, Quirino Province (�6° 0�’ 22.5” N, �2�° 28’ �9.7” E, �-�2 June 2004). We surveyed 2 sites at 400 m and 700 m elevation. Netting was done at 400 m and trapping was done at 400 m and 700 m. The area was a selectively logged primary lowland forest. Emergent trees were predominantly dipterocarp species reaching maximum height of about �0 m with DBH reaching about ��0 cm. Height of canopy tree species was estimated at about 20 to 25 m with DBH

Duya et al. 2007. Banwa 4(�):4�-68.

about �2-40 cm. Understory vegetation was minimal, mostly ferns, rattans, ginger, and Begonia species. Vines were uncommon. The forest floor was rocky, with large rocks and boulders common, especially near the creek. The site is the hunting ground of the Bugkalot. Timber poaching, mainly for the premium species (such as narra and kamagong) was common. Hunting and some previous forest clearings were observed near our study area.

Site 5. Mt. Binuang, Barangay Minahan, General Nakar Municipality, Quezon Province (�4° 45’ 52” N, �2�° �5’ 0�” E, 24 June – � July 200�). Mt. Binuang rises a little over 1000 m. Trapping and netting was done in a highly disturbed secondary lowland forest between 500 and 700 m. Emergent and canopy trees are predominantly Dipterocarpaceae and Dilleniaceae. Emergent trees reached a maximum height of 25 m and canopy trees reached 20 to 25 m, with DBH of about �0 to 40 cm. The herb layer consisted mainly of pteridophytes. Epiphytes were mostly aroids and lianas, mainly Freycinetia spp. and Ficus spp. Shrubs and saplings consisted mostly of Dipterocarpaceae, Euphorbiaceae, Proteaceae, and Celastraceae. Slash and burn farming was common below our study area. Commercial logging was cancelled in the area in the early �990s, but timber poaching was rampant with freshly cut trees observed along the trail and along the slopes.

Site 6. Mt. Cagua, Barangay Magrafil, Gonzaga Municipality, Cagayan Province (�8° ��.�2’ N, �22° 06.66’ E, 6-�4 August 2004). Netting and trapping was done in a slightly disturbed secondary lowland forest at 7�0 m. Emergent and canopy trees were predominantly Myrtaceae, Melastomataceae and Dipterocarpaceae reaching a maximum height of �8 m and canopy height of �4 meters with DBH ranging from �0 to 40 cm. Epiphytes were ferns and lianas, mostly Freycinetia spp. Understory plants were mostly saplings, most belonging to Dipterocarpaceae, Sapindaceae and Lauraceae. Terrain was moderately flat. Mt. Cagua is a dormant volcano; its last eruption, in �860, created a circular summit crater about �.5 km in diameter, with steep, 60-m-high walls; the highest peak reaches 1133 m (http://www.volcano.si.edu/world/volcano). Hunting was observed in the area and timber poaching was evident outside the crater at 800 m due to the presence of newly cut trees and tree stumps. Commercial logging in the area was cancelled in �99�.

48 49

Duya et al. 2007. Banwa 4(�):4�-68.

Site 7. Sitio Lowak, Barangay Minanga, Peñablanca, Cagayan Province (�7o �6’ 49.�” N, �2�o 56’ 2�.8” E, ��-�9 November 2002). Mist netting and trapping were conducted between 960 and 1200 m. The vegetation was transitional lowland to montane forest, with typical lowland plant species (e.g. Dipterocarpaceae) uncommon relative to montane tree species. Most emergent trees were Shorea polysperma. Canopy trees were predominantly Lithocarpus and Syzygium species. The subcanopy included laurels (Lauraceae), oil fruits (Elaeocarpus spp.), and tea (Theaceae). Common understory plants included Discocalyx, Symplocos, Olea, Macaranga and Canthium species. Trees were predominantly low-branching and multi-stemmed. Emergent trees reached a maximum of 25 m. Prevailing canopy species reached heights of about �2 to �8 m. Terrain consisted of moderate to rather steeply ascending ridges. We observed about �50 snares for wild pigs and deer in the area. Commercial logging was cancelled in late �990s, but timber poaching was apparent as we often encountered people with their carabao pulling freshly cut timber near our study area.

Site 8. Mt. Lataan, Barangay Disimungal, Nagtipunan, Quirino Province (�6° �0’ 20.6” N, �2�° 44’ 22.7” E, 5-�4 March 200�). Mist netting and trapping was done from 900 to 1000 m in lower montane forest. Emergent and canopy trees were predominantly Fagaceae, Dilleniaceae, Myrtaceae and Sapotaceae. Canopy height was estimated at 8 m, and emergents reached a maximum of �5 m. Trees were predominantly low branching and multi-stemmed, �0 to �5 cm diameter. Shrubs and saplings consisted of Rubiaceae, Cornaceae, Myrtaceae, Rubiaceae species. The herb layer commonly consisted mostly of ferns and gingers. Epiphytic plants were mostly seen on tree boles, dominated by orchids, ferns and bryophytes. Lianas were mostly Freycinetia spp. The slope was moderately ascending to flat. According to our local guides, commercial logging did not reach the area. Timber poaching and hunting were observed.

Site 9. Mt. Cetaceo, Sitio Baua, Barangay Lapi, Peñablanca Municipality, Cagayan Province (�7º 42’ 28.8” N, �2�º 59’ 49” E, 5-�9 May 2004). We conducted netting and trapping in old-growth montane forest at ��00 m dominated by Tristaniopsis spp. and Lithocarpus spp. Tree trunks were covered with moss. Emergent trees were approximately 20 m (Agathis philippinensis) and average height of dominants was

Duya et al. 2007. Banwa 4(�):4�-68.

approximately 10 to 12 m. Undergrowth plants included rattans, ferns, and Cyperaceae. Forest litter was about 3 to 5 cm deep. Exposed rocks were common measuring about half a meter to a meter in diameter and were mostly covered with thick moss.

Site 10. Mt. Cetaceo, Sitio Baua, Barangay Lapi, Peñablanca Municipality, Cagayan Province (�.5 km SW of Mt. Cetaceo Peak, �7º 4�’ 52” N, �22º 0�’ 08.9” E, 9-�8 June 2005). We conducted netting and trapping in montane forest at 1400 m. Emergent trees were Agathis philippinensis approximately �7 m in height. Height of canopy species was approximately �2 to �5 m. Average height of dominant trees species was approximately �0 to �5 m dominated by Syzygium and Lithocarpus species. Ground cover was mainly composed of ferns and mosses.

Site 11. Mt. Cetaceo, Sitio Baua, Barangay Lapi, Peñablanca Municipality, Cagayan Province. (�.5 km SW of Mt Cetaceo peak, �7º 42’2�.80” N, �22º �5.80’ E, �4-�7 June 2005). We conducted trapping at �500 m in mossy forest dominated by Leptospermum species. Emergent trees were approximately 6 to 7 m and average height of canopy tree species was 5 m. DBH of canopy cover was 5 to �5 cm. Canopy epiphytes included orchids and vines. Pitcher plants were relatively common. Pandan and Ficus species were rare. Understory plants were dominated by Drymis viperata and ground cover plants were dominated by ferns and mosses.

Accounts of speciesOrder Insectivora�. Family Soricidae – Shrews a. Crocidura grayi (Dobson, �890). The Luzon shrew is endemic

to the Philippines and has been recorded in Catanduanes, Aurora, Camarines Sur, Laguna, and Rizal Provinces and on Mindoro (Heaney et al., �998). We captured this species only in montane forest at �400 m and mossy forest at �500 m (Table �). Specimens examined: Site �0; �; Site ��, �.

Order Chiroptera�. Family Pteropodidae – Fruit bats a. Acerodon jubatus (Eschscholtz, �8��). The golden-crowned

flying fox, a Philippine endemic, has been documented in the Sierra Madre, in Divilacan and Dinapigue, Isabela Province (Danielsen et al., �994; Mudar & Allen, �986). We did not capture this species, but at Site 9 local people reported

50 5�

Duya et al. 2007. Banwa 4(�):4�-68. Duya et al. 2007. Banwa 4(�):4�-68.

Table �. Number of individuals captured by trap or net (+ number caught by hand) during the biological surveys conducted along the Sierra Madre in 2002, 200�, and 2004. Scientific name in bold face indicates species endemic to the Philippines

Species Site � Site 2 Site � Site 4 Site 5 Site 6 Site 7 Site 8 Site 9 Site �0 Site ��

Moderately Moderately Lightly Selectively Highly logged Lightly logged Transition Old growth Old growth Old growth Mossy disturbed disturbed disturbed logged secondary secondary secondary montane montane montane forest limestone secondary secondary primary lowland lowland lowland to forest forest forest forest lowland lowland lowland forest forest montane forest forest forest forest

�00 m �00 m �00 m 400 m 500-700 m 7�0 m 960-�200 m 900-�000 m ��00 m �400 m �500 m

�. Crocidura grayi 0 0 0 0 0 0 0 0 0 � � 2. Acerodon jubatus R 0 0 0 0 0 0 0 0 0 0 �. Cynopterus brachyotis 2 9 � � � � 20 0 0 0 0 4. Haplonycteris fischeri � � �� 0 8 � �0 0 0 0 0 5. Macroglossus minimus � � �2 0 � � 2 0 0 0 0 6. Otopteropus cartilagonodus 0 0 4 0 4 � � �8 �� 0 0 7. Ptenochirus jagori 0 � � 2 6 0 29 �2 � 0 0 8. Pteropus hypomelanus 0 0 S 0 0 0 0 0 0 0 0 9. Pteropus leucopterus 0 0 0 0 S 0 0 0 0 0 0 �0. Pteropus vampyrus 0 0 S 0 O 0 0 0 0 0 0 ��. Rousettus amplexicaudatus � � 0 0 0 0 0 0 0 0 0 �2. Coelops hirsutus2 0 0 0 0 0 0 0 � 0 0 0 ��. Hipposideros ater 0 0 0 � 0 0 0 0 0 0 0 �4. Hipposideros diadema � 0 � � (+�) 0 � 0 0 � 0 0 �5. Hipposideros obscurus2 0 0 0 2 � (+�) 0 0 0 0 0 0 �6. Rhinolophus arcuatus 0 0 0 7 2 5 0 0 0 � 0 �7. Rhinolophus inops2 0 0 0 0 0 � 9 0 2 � 0 �8. Rhinolophus philippinensis2 0 0 0 0 0 0 5 0 0 0 0 �9. Rhinolophus virgo 0 0 0 � 0 � 0 0 0 0 0 20. Harpiocephalus harpia2 0 0 � 0 0 0 0 0 0 0 0 2�. Kerivoula cf. papillosa2,� 0 0 0 � 0 0 0 0 0 0 0 22. Murina cyclotis2 0 0 � � 0 0 0 0 0 0 0 2�. Myotis horsfieldi 0 0 0 � 0 0 0 0 0 0 0 24. Myotis muricola 0 0 0 0 0 � 0 0 0 0 0 25. Pipistrellus javanicus 0 0 0 � 0 0 0 0 5 � 0 26. Chaerophon plicata� 0 0 0 0 0 0 0 0 0 0 0 27. Macaca fascicularis R R R S R R R R R 0 0 28. Apomys sp. 0 0 0 9 0 �9 8 � 67 54 85 29. Apomys cf microdon 0 0 0 0 0 0 0 0 0 2 � �0. Bullimus sp. 0 � � �6 �8 �4 � 4 �� �4 8 ��. Archboldomys musseri 0 0 0 0 0 0 0 0 0 0 4 �2. Chrotomys sp. 0 0 +� 9 0 0 0 0 0 0 0 ��. Phloeomys pallidus S R R R 0 R R S R 0 0 �4. Rattus everetti 0 � 5 2� �9 7 5 4 4 8 � �5. Rattus exulans 0 0 0 � 0 0 0 0 0 0 0 �6. Paradoxurus hermaphroditus 0 0 0 0 0 S 0 S 0 0 0 �7. Sus philippensis R R S S R S R R S R 0 �8. Cervus marianus R R R S R R S R S S 0

Total trap-nights (Cage traps) 0 6 5�0 4 0 245 0 50� x 0 0 0 Total trap nights (Victor Rat Traps) 0 0 �04 4 ��88 �454 255 �744 x 7�4 9�� 68� Total trap nights (Museum specials) 0 0 0 0 0 0 0 0 0 76 9� Total net nights �.5 �.5 �2 �� 64 �7 78 x 98 �2 0 Total V-nets (2 hours/night) 0 0 0 40 2 8 0 0 0 0 0

S = sighted; R = reliable reports; 0 = absent, 1 = Specimen given by a local spelunker; 2 = new record for Sierra Madre; 3 = new country record; 4 = locally made snap traps; x = no data were kept.

Table �. contd.

52 5�

Duya et al. 2007. Banwa 4(�):4�-68.

a roost of very large bats that we were not able to verify; these might have been this species or Pteropus vampyrus, or both together (Mildenstein et al.. 2005). The large bats are frequently hunted in the area.

b. Cynopterus brachyotis (Mueller, �8�8). We captured the common short-nosed fruit bat in lowland agricultural areas and secondary forest from �00 to 7�0 m, and in transitional lowland-montane forest at 960 m (Table �). Despite extensive netting, we did not capture them in montane forest between 900 to ��00 m at Site 8 and 9 and �400 m at Site �0. Specimens examined: Site 4, �; Site 6, �.

c. Haplonycteris fischeri (Lawrence, �9�9). We caught the endemic Philippine pygmy fruit bat in secondary lowland forest from �00 to 7�0 m, and in transitional lowland-montane forest at 960 to 1200 m. Despite extensive netting at Site 8, 9 and 10 in montane forest between 900 to ��00 m, we captured none (Table �). All individuals were released.

d. Macroglossus minimus (E. Goeffroy, 1810). We captured the dagger-toothed flower bat, a widespread Southeast Asian species, in agricultural areas, forest edge and secondary lowland forest from �00 to 7�0 m, and in transitional lowland-montane forest at 960 m. They apparently were absent in montane forest between 900 to �400 m at Sites 8, 9 and �0, and mossy forest at Site �� at �500 m (Table �). Specimens examined: Site 6, �.

e. Otopteropus cartilagonodus (Kock, �969). The Luzon pygmy fruit bat is a monotypic genus endemic to Luzon Island (Heaney et al., �998); previous records were from secondary and primary lowland, montane, and mossy forest at 200 to �900 m (Mudar and Allen, �986; Ruedas et al., �994). We captured this species in secondary lowland forest from �00 to 7�0 m, in transitional lowland-montane forest at 960 to �200 m, and in old-growth montane forest at 900, ��00 and �400 m. As seen on Mt. Isarog (Heaney et al., �999), O. cartilagonodus overlapped with H. fischeri at some sites, but O. cartilagonodus was less common than H. fischeri in lowland forest, and H. fischeri was absent at the montane forest sites where O. cartilagonodus was common (Table �). Specimens examined: Site 9, �.

f. Ptenochirus jagori (Peters, �86�). The musky fruit bat is a widespread endemic within the Philippines, often common in primary and secondary forest and agricultural areas

Duya et al. 2007. Banwa 4(�):4�-68.

adjacent to forest (Heaney et al., �999; Heideman and Heaney, �989; Ingle, �992; Mudar and Allen, �986; Lepiten, �995). We found it to be common in secondary lowland forest from �00 to 500 m and in transitional lowland-montane forest at 960 m; it was uncommon in old-growth montane forest at 900 and ��00 m (Table �). Specimens examined: Site 4, 2; Site 9, �.

g. Pteropus hypomelanus (Temminck, �85�). The common island flying fox is distributed throughout much of Indo-Australia, including the Philippines, where it is common in agricultural areas and absent in primary forest (Rickart et al., �99�; Utzurrum, �992). We captured none, but observed the island flying fox feeding in groups on a fig tree (tangisang bayawak – Ficus variegata) at Site �, together with the larger Pteropus vampyrus. Groups of flying foxes were also seen high in the sky at dusk (�800 h) and consistently observed for the duration of the survey period. Local people call this species bayakan in Tagalog, and report that they regularly observe this species visiting their orchards.

h. Pteropus leucopterus (Temminck, 1853). The mottled-winged flying fox is endemic to the Luzon faunal region and Dinagat Island (Heaney et al., �998). In the Sierra Madre, it was reported in Dinapigue, Isabela Province (M. Van Weird, personal communication). We did not capture this species at our sites, but we used binoculars to observe one individual (with distinctive white spots on its wings) flying over our camp at Site 5. Local people at all of our sites reported the presence of this species, but further survey is needed to verify these reports.

i. Pteropus vampyrus (Linnaeus, �758). This species is found from Indochina to the Lesser Sundas and is widespread within the Philippines, in lowland forest up to �250 m (Heaney et al., �998; Rabor, �955, �986; Rickart et al., �99�). We did not catch this species, but saw them feeding on figs (tangisang bayawak – Ficus variegata) at Site � together with P. hypomelanus. Some individuals were observed to have a large reddish brown “cape” from the head to the nape of the bat that typifies P. vampyrus. Local people call this species bayakan in Tagalog, and reported that they regularly observed this species visiting their orchards. We also observed several individuals over our camp at Site 1; the identification was not certain.

54 55

Duya et al. 2007. Banwa 4(�):4�-68.

j. Rousettus amplexicaudatus (É. Geoffroy, 1810). The common rousette occurs from Thailand to Solomon Island including the Philippines, usually in open, heavily disturbed habitats (Heaney et al., 1998). We netted the species in open and agricultural areas at Sites � and 2 at �00 m elevation; they were absent at other sites, where all of the netting was done in the forest. All individuals were released.

2. Family Rhinolophidae - Horseshoe Bats a. Coelops hirsutus (Miller, �9��). The Philippine tailless

roundleaf bat occurs only in the Philippines (Heaney et al., �998). We captured one individual of this poorly known species at Site 2 in secondary lowland dipterocarp forest at about 900 m. The species was known previously from Mindoro and Mindanao (Heaney et al., 2006). The species has also been captured on Mt. Makiling, Laguna Province (J. Sedlock, personal communication). Specimens examined: Site 8, �.

b. Hipposideros ater (Templeton, �848). The dusky roundleaf bat occurs from India to Australia, including the Philippines (Heaney et al., �998). We captured the species in secondary lowland forest at about 400 m at Site 4 using a V-net. Specimens examined: Site 4, �.

c. Hipposideros diadema (E. Goeffroy, 1813). The diadem roundleaf bat occurs from Burma to the Solomon Islands and common throughout the Philippines (Heaney et al., �998). We caught the species in almost all of the habitat types we sampled from �00 m to ��00 m (Table �). A colony of about �50 individuals was observed inside a large cave near Site � and we observed the species emerging from tree hollows at Site �. One individual roosting under a large fern at Site 4 was caught by hand. Specimens examined: Site �, �; Site �, �; Site 4, 2; Site 6, �; Site 9, �.

d. Hipposideros obscurus (Peters, �86�). The Philippine forest roundleaf bat is endemic to the Philippines and occurs in primary and disturbed forest up to ��00 m (Heaney et al., �998, 2006; Sedlock, 200�). We captured two individuals using a V-net set across a creek at Site 4 and three individuals by mist nets and one by hand in an abandoned mineshaft at Site 5 at 400 m and 500 m elevation, respectively. Specimens examined: Site 4, 2; Site 5, �.

Duya et al. 2007. Banwa 4(�):4�-68.

e. Rhinolophus cf. arcuatus (Peters, �87�). The arcuate horseshoe bat is a widespread endemic within the Philippines (Esselstyn et al., 2004; Heaney et al., �998; Sedlock, 200�). The species currently referred to as Rhinolophus arcuatus in the Philippines may consist of two or more species (Heaney et al., �99�, �999, 2006; Ingle and Heaney, �992). We captured the species in lightly disturbed lowland forest at Sites 4 and 6, and highly disturbed secondary lowland forest at Site 5 (Table �). Initial examination of our specimens suggests two different species, corresponding to the “large” and “small” morphs (Ingle and Heaney, �992). Further taxonomic study is needed. Specimens examined: Site 4, �0; Site 5, 2; Site 6, 5; Site �0, �.

f. Rhinolophus inops (K. Andersen, �905). The Philippine forest horseshoe bat is endemic to the Philippines; it is often abundant in primary lowland and montane forest up to 2250 m (Heaney et al., �998, 2006). We captured this species in secondary lowland forest and old growth montane forest between 7�0 m and ��00 m (Table �). Specimens examined: Site 6, 2; Site 7, 9; Site 9, 2; Site �0, �.

g. Rhinolophus philippinensis (Waterhouse, �84�). The enormous-eared horseshoe bat is widely distributed from Borneo to Australia and has been captured only in primary and secondary forest (Heaney et al., �998; Lepiten, �995; Ruedas et al., �994). We captured this species only in transitional lowland forest and montane forest between 960 m and �200 m (Table�). Specimens examined: Site 7, 5.

h. Rhinolophus virgo (K. Andersen, �905). The yellow-faced horseshoe bat is endemic to the Philippines and inhabits secondary and primary lowland forest from 250 to ��00 m (Heaney et al., �99�, �998, 2006; Ingle, �992; Rickart et al., �99�; Sedlock, 200�). We captured this species in selectively logged primary lowland forest and slightly disturbed secondary lowland forest using a V-net at 400 m and 7�0 m (Table �). Our specimens appear to represent two different morphs; the specimen from Site 6 (FMNH �76548) is substantially smaller than the others, which appear to represent typical R. virgo, and some cranial features differ. Further taxonomic study is needed. Specimens examined: Site 4, �; Site 6, �.

56 57

Duya et al. 2007. Banwa 4(�):4�-68.

�. Family Vespertilionidae - Common bats a. Harpiocephalus harpia (Temminck, �840). The hairy-winged

bat is widespread in southern Asia and previously recorded in the Philippines only in Camarines Sur and Laguna provinces on Luzon (Heaney et al., �998; Ingle, �99�). It was previously documented in primary and disturbed lowland forest between 475 m to 750 m (Rickart et al., �99�). We captured an individual at Site � in secondary lowland forest at �00 m. Specimen examined: Site �, �.

b. Kerivoula cf. papillosa (Temminck, �840). Three species of Kerivoula have been documented in the Philippines previously; all three are fairly widespread in Southeast Asia, principally in lowland forest (Esselstyn, 2004; Heaney et al., �998; Ingle and Heaney, �992). We captured an individual of a large Kerivoula at Site 4 using a V-net set along an existing trail in selectively logged lowland forest. Compared to the known species in the Philippines, our species is much larger, with forearm of 44 mm and a total length of �08 mm. Direct comparison with a series of K. papillosa from Vietnam (FMNH �2209, 46564, 46565, 46567, 46568, 4662�) show the external morphology and crania to be very similar. Kerivoula papillosa occurs from eastern India to Indochina, Borneo, and Sulawesi (Corbet and Hill, �992; Payne et al., �985); further study is needed. Specimen examined: Site 4, �.

c. Murina cyclotis (Dobson, �872). The round-eared tube-nosed bat occurs from Sri Lanka to Hainan and Borneo and is widespread in the Philippines. It occurs in primary and lightly disturbed lowland and montane forest between 250 m and �500 m (Corbet and Hill, �992; Heaney et al., �99�; Heaney et al., �998, Lepiten, �995; Rickart et al., �99�;). We captured an individual at Site � using mist nets and one individual at Site 4 using a V-net. On Luzon, the species has previously been reported from Camarines Sur (Heaney et al., �998, �999), Laguna (Sedlock, 200�) and Kalinga provinces (Heaney et al., 2004). Specimens examined: Site �, �; Site 4, �.

d. Myotis horsfieldii (Temminck, �840). The common Asiatic myotis occurs from southeastern China to the Malay Peninsula, Bali, and Sulawesi, and is widespread in the Philippines. It inhabits agricultural areas and lowland forest (Heaney et al., �998; Sedlock, 200�). We captured one individual in selectively logged lowland forest at Site 4 using

Duya et al. 2007. Banwa 4(�):4�-68.

a V-net. Specimen examined: Site 4, �. e. Myotis muricola (Gray, �846). The whiskered myotis is

widespread in Asia and throughout the Philippines (Heaney et al., �998). It inhabits primary and lowland forest from near sea level to ��25 m (Heaney et al., �999; Rickart et al., �99�). We captured one individual at Site 4 using a V-net in selectively logged lowland forest. Previous Luzon records are from Camarines Sur (Heaney et al., �999), Laguna (Sedlock, 200�), and Kalinga provinces (Heaney et al., 2005). Specimen examined: Site 4, �.

f. Pipistrellus javanicus (Gray, �8�8). The Javan pipistrelle is widespread in eastern Asia (Corbet and Hill, �992). In the Philippines, it occurs from sea level to �750 m where it is common in montane forest and uncommon in lowland forest and mossy forest (Heaney et al., �998). We captured three individuals at Site 4 in selectively logged lowland forest at 400 m using V-nets, five individuals in primary montane forest at site 9 at ��00 m and three individuals at Site �0 at �400 m. Our species seem to consist of two morphs based on some differences on external morphologies noted by Sedlock (200�) and Heaney et al. (2006). Further taxonomic study is needed. Specimens examined: Site 4, �; Site 9, �; Site �0, �.

4. Family Molossidae - Free-tailed bats a. Chaerephon plicata (Buchanan, �800). The wrinkled-lipped

bat occurs from India to Bali, Hainan, and the Philippines (Corbet and Hill, �992). The species is believed to be declining in the Philippines due to heavy disturbance of caves (Heaney et al., �998; Rickart et al., �99�). One dead specimen given to us on �4 June 2004 by a local spelunker was taken from a rock near a cave in Barangay Aggugadan, Peñablanca Municipality, Cagayan Province. Local people turned over three more dead specimens from the same cave where the previous one was found. A colony of this species may be present; assessment of the cave is urgently needed. Specimens examined: Near Site 7, �.

Order Primates�. Family Cercopithecidae – Monkeys a. Macaca fascicularis (Raffles, 1821). The long-tailed macaque

occurs from Burma to Timor and is widespread in the Philippines (Heaney et al., �998). We observed this species at

58 59

Duya et al. 2007. Banwa 4(�):4�-68.

most of our sites feeding on fruits in lowland forest canopy. Based on our interviews, the species is frequently seen in groups of 20 to �0 individuals and feeds on corn and root crops. Our local guide caught one adult male at Site 4. The species is regularly hunted for local consumption.

Order Rodentia�. Family Muridae – Rats and mice a. Apomys sp. The genus Apomys is the most speciose genus in

the Philippines, including a large (and increasing) number of recently discovered and still undescribed species (Heaney et al., �998; Steppan et al., 200�). Our specimens are the first large series from the Sierra Madre, and preliminary comparisons indicate that they represent one or more previously unknown large-bodied species; investigations are on-going (Heaney et al. in prep.). We captured these large forest mice in secondary lowland forest montane and mossy forest. None was captured in forest adjacent to agricultural areas at ca. �00 m elevation, but they were common at sites from 400 to �500 m (Table �). Specimens examined: Site 4, 9; Site 6, �5; Site 7, 7; Site 8, �; Site 9, 67; Site �0, 54; Site ��, 85.

b. Apomys cf microdon (Hollister, �9��). This small Luzon forest mouse is endemic to the Philippines and known only from Catanduanes Island, and from Camarines Sur, Isabela and Kalinga provinces on Luzon (Heaney et al., �998, 2005). We captured the species at our two highest sites, in montane forest at Site �0 at �400 m and mossy forest at Site �� at �500 m using Victor rat traps and Museum Specials placed on tree branches. We did not catch the species on the ground or at other sites that we surveyed. Specimens examined: Site �0, 2; Site ��, �.

c. Archboldomys musseri (Rickart et al., �998). This species is endemic to the Sierra Madre Mountains and is known only from Mt. Cetaceo (Balete et al., 2006). We captured four individuals at Site ��, which is also the locality of holotype that was taken in �992 (Danielsen et al., �994; Heaney et al., �998). We captured the species only in mossy forest at �500 m despite intensive trapping in montane and lowland forest. They should be sought at the peaks of other high mountains in the Sierra Madre. Specimens examined: Site ��, 4.

d. Bullimus sp. (Thomas, �895). This genus of large forest rat is endemic to the Philippines. It inhabits primary and

Duya et al. 2007. Banwa 4(�):4�-68.

disturbed lowland, montane, and mossy forest (Heaney et al., �998; Heaney et al., 2005; Musser, �982; Sanborn, �952). We captured this species at all of our sites except at Site �, where no trapping was done, and at Site 2, in forest adjacent to agricultural areas (Table 1). Morphological differences were observed between our specimens and those from the Central Cordillera; taxonomic investigations are on-going. Specimens examined: Site 2, �; Site �, �; Site 4, �6; Site 5, 4; Site 6, ��; Site 7, �; Site 8, 2; Site 9, ��; Site �0, �4; Site ��, 6.

e. Chrotomys sp. This genus of murid rodents is endemic to the Philippines. Currently, five species are known: C. gonzalesi is restricted to Mt. Isarog, C. whiteheadi in the Central Cordillera, C. mindorensis in the lowlands of central Luzon and Mindoro (Heaney et al., �998), C. sibuyanensis from Sibuyan, and C. silaceus, formerly placed in the genus Celaenomys, from the Central Cordillera (Rickart et al., 2005). Present ecological data indicates that C. whiteheadi occurs at high altitudes, ca. �000 to 2500 meters (Heaney et al., �998, 2005), as does C. gonzalesi (Rickart et al., 1991). Our first specimen of Chrotomys at Site � (�00 m elevation) was captured by a pet cat. We captured a series at Site 4 between 400 to 700 m; both sites were in secondary lowland forest. Preliminary comparison with the known species indicates that this may represent a sixth species; taxonomic studies are on-going. Specimens examined: Site �, �; Site 4, 9.

f. Phloeomys pallidus (Nehring, �890). The slender-tailed cloud rat is endemic to the Philippines and occurs only in the central and northern part of Luzon Island (Heaney et al., �998; Oliver et al., �99�). Most of our information on this species was based on interviews, though we encountered them at two of our sites. An adult male foraging in a cornfield was caught by a farmer at Site 1. At Site 8, our local guide shot an individual a kilometer away from the camp in secondary lowland forest at about 500 m. Based on our interviews, a dark brown color-morph was also reported in the area, which resembles Phloeomys cumingi that occurs in the central and southern part of Luzon. These large (ca. 2.5 kg) rodents are hunted regularly in the Sierra Madre. Local residents often reported capturing this species on their farms or in coconut plantations, and some were caught by dogs; we confirmed the species to be Phloeomys pallidus by showing them pictures.

60 6�

Duya et al. 2007. Banwa 4(�):4�-68.

g. Rattus everetti (Gunther, �879). The common Philippine forest rat is endemic to the Philippines (Heaney et.al. �998), in secondary and primary lowland and montane forest and less often at high elevations (Rickart et.al., 1991, 1993). We captured the species at all of our study sites except at Site � where no trapping was done. They were relatively common in secondary lowland forest and declined in relative abundance as elevation increased (Table �). Specimens examined: Site 2, �; Site �, 4; Site 4, 2�; Site 5, 5; Site 6, 7; Site 7, 2; Site 8, �; Site 9, 5; Site �0, 8; Site ��, �.

h. Rattus exulans (Peale, 1848). The spiny rice-field rat occurs from Bangladesh to Easter Island and throughout the Philippines (Heaney et al., 1998), usually in open agricultural or fire-associated grassland habitats, and rarely in forest (Heaney et al., �989; Heideman et al., �987; Rickart et al., �99�). We captured only one individual at Site 4 in forest edge adjacent to an agricultural area. Specimen examined: Site 4, �.

Order Carnivora�. Family Viverridae-Civets a. Paradoxurus hermaphroditus (Pallas, �777). The common palm

civet occurs from Sri Lanka to Hainan and Lesser Sunda Islands and widespread in the Philippines (Heaney et al., �998). We captured one adult male in secondary lowland forest at Site 5 using a local snare set by our guides, and one young individual in a cage trap in montane forest at Site 8. Both individuals were later released. Based on our interviews, the species is hunted for local consumption.

Order Artiodactyla�. Family Suidae – Pigs a. Sus philippensis (Nehring, �886). The Philippine warty pig

is endemic to the Philippines (Heaney et al., �998). We documented this species at all of our study sites. We often observed tracks in lowland and montane forest between �00 m and ��00 m. An adult male was captured in a local snare at Site 9, and a young individual was seen in the house of a local resident near Site 4. Based on our interviews with the local people, the species is regularly hunted in the Sierra Madre both for local and commercial consumption.

2. Family Cervidae – Deer a. Cervus mariannus (Desmarest, �822). The Philippine brown

Duya et al. 2007. Banwa 4(�):4�-68.

deer occurred originally only in the Philippines but was later introduced to the Marianna Islands (Heaney et al., �998). We observed one captive individual at a house near Site �. Another individual was seen in the back yard of a local resident at Site �. The owners told us that they intended to raise them in captivity as they fetch a high price in the market. At Site 4 a local resident offered to sell a skin and head of a deer. We also encountered hunters at Site 9 carrying a newly butchered deer and we also observed a deer foraging near our camp at Site �0. According to local hunters, they were having difficulty catching the species, unlike in previous years; hunters now must spend more time in the forest and further away from settlements. The species is hunted both for local and commercial consumption.

Mammalian diversity It has been about two decades since the first moderately comprehensive biological surveys of mammals were conducted in the Sierra Madre, each at a single site or in limited areas (Danielsen et al., �994; Mudar and Allen, �986). Although several surveys have followed, very few have been published and few specimens are available. Our 2002 to 2005 surveys have generated information on the importance of the mountain range in terms of mammalian diversity. We documented a total of �7 species of mammals during our surveys (Table 1). Among these, one shrew, five fruit bats, four insectivorous bats, seven murid rodents and two ungulates are endemic to the Philippines. Seven species of insectivorous bats were documented for the first time in the Sierra Madre and two species of rodents are potentially new species. We captured Kerivoula cf. papillosa, documented within the Philippines for the first time. Coelops hirsutus, captured in Quirino Province, was previously recorded only from Mindoro and Mindanao Islands (Heaney et al., �998), but this elusive bat was also captured in 2004 on Mt. Makiling, (J. Sedlock, personal communication), suggesting that this species is widely distributed on Luzon and throughout the country. In addition, Rhinolophus inops, R. philippinensis and Hipposideros obscurus have not been previously recorded in the mountain range (Heaney et al., �998). Similarly, the only previous record of Harpiocephalus harpia from Luzon is from Camarines Sur despite the fact that it is widespread in southern Asia (Heaney et al., �998; Nowak, �99�). Murina cyclotis has been documented at scattered localities in the Philippines, but not in the Sierra Madre.

62 6�

Duya et al. 2007. Banwa 4(�):4�-68.

Additionally, species of insectivorous bats that we have tentatively identified as Rhinolophus arcuatus, Rhinolophus virgo, and Pipistrellus javanicus need further investigation. Like any other potential new species, nothing is certain unless a thorough taxonomic study has been made and specimens compared with other known species. Among the murid rodents, Chrotomys mindorensis was reported from Isabela (NORDECO-DENR, �998), but this needs to be reexamined in the light of our large series of what appears to be a distinct, new species. Two additional rodents that we have documented may represent new species, Bullimus sp. and Apomys sp.; the existence of the series we collected makes taxonomic studies possible. We observed that majority of the native species of rodents, such as Apomys sp, Bullimus sp, Chrotomys sp, and Rattus everetti, were captured in primary and disturbed forest, while commensal species such as Rattus exulans were caught only in agricultural areas adjacent to forest. As observed in the past, surveys of insectivorous bats by means of standard mist-netting methods is inadequate, as these bats can echolocate and thus evade mist-nets easily (Alviola, 2000, Sedlock, 2001). We employed a novel mist-netting technique (V-nets) at Site 4, 5 and 6. This resulted in �� species of insectivorous bats captured, a relatively high number for microchiropterans (Table �). Tunnel-trapping on Polillo Island (Alviola, 2000) and Mt. Makiling (P. Alviola, personal communication; Sedlock, 200�) have resulted in �0 and �� new records for the two localities, respectively.

Conservation priorities Many of the insectivorous bats documented in the survey are cave dwellers. Unfortunately, no study has been done on caves in the Sierra Madre. Many of these caves harbor large colonies of insectivorous bats, such as the “Bat Cave” in the municipality of Peñablanca, Bolos Point Cave in Gattaran, Kapanikian caves in Sta. Ana, Cagayan Province, Aglipay Caves in Quirino Province, and several other large caves reported in our study sites. We documented specimens of Chaerophon plicata near Site 7; although they are not endemic or officially listed as threatened, several caves that were known to support the typically very large colonies of this species have been destroyed (Heaney et al., �998, Rickart et al., �99�). Survey of these caves should be done and protection measures should be put in place, particularly those that support large colonies of insectivorous bats that are located adjacent to human settlements.

Duya et al. 2007. Banwa 4(�):4�-68.

The Philippine warty pig and Philippine brown deer are the most frequently hunted mammals in the Sierra Madre as they fetch a high price in the market. Although these species were reported to be relatively common in our study sites, many of the local people we interviewed noted that they were becoming scarce. The Philippine macaque and cloud rats were frequently encountered in farm lots adjacent to forest areas. They were hunted by the local people and were sometimes considered as pests of agricultural crops. Colonies of flying foxes were also reported in our study sites but we did not have the chance to visit these areas. Existing roosts comprising A. jubatus and P. vampyrus in Divilacan and Dinapigue, Isabela numbering to about 60,000-�20,000 individuals have been reported (Danielsen et al., �994). Most of these roost sites are located in the lowland forest and accessible to local people, making them vulnerable to hunting and destruction of roosting areas. A survey should be conducted to locate these roosting sites, conduct population and ecological studies and put management and protection measures in place. In addition, many of the endemic mammals we documented in our study use both secondary and primary lowland forest. This includes four species whose ranges are confined to Luzon: Otopteropus cartilagonodus, Apomys sp., Bullimus sp., and Phloeomys pallidus. As mentioned above, O. cartilagonodus was recorded at almost all of our survey sites from �00 m to ��00 m. (Heaney et al., �998; Ruedas et al., �994). Haplonycteris fischeri, which is abundant in the southern Philippines, is less common on Luzon and is generally dependent on primary or good quality secondary forest at low elevation (Heaney et al., �989, �998; Ingle, �992). We trapped Bullimus sp. and Apomys sp. at nearly all of our sampling sites. This suggests that both species have wide habitat preference and can tolerate disturbed lowland forest (Table �). We recorded P. pallidus at all sites except Site 5, thus tentatively concurring with its reported distribution, which is confined to northern and central Luzon (Oliver et al., 1993). We received reports from local communities in the Sierra Madre that they have encountered a darker version of the species in areas where P. pallidus is known to occur. This observation needs further investigation. Despite the cancellation of Timber License Agreements in the �990s, destruction of forest in the Sierra Madre remains unabated. Almost all of the lowland forest in the western side of Sierra Madre from �00-800 m is accessible because of old logging roads. This has led to the transformation of many of these areas into farmlands and has provided easy access to the remaining forest for timber poachers.

64 65

Duya et al. 2007. Banwa 4(�):4�-68.

Management and protection of the lowland forest should be a high priority if we are to protect our endemic species, and the watershed forests where they live. Very few have been declared as protected areas. Although some have been place under the Community based Forest Management Agreement (CBFMA) and Certificate of Ancestral Domain Claim (CADC), the lowland forest still does not receive enough protection. New protected areas should be large enough to include all known habitat types, even the degraded lowland forests. Existing protected areas should be expanded to include adjacent forest areas to maintain their watershed functions. Our data show that there is a need for more research, especially in areas not previously studied, particularly the caves, high elevations and the canopy of lowland forest, and on the still–unsurveyed high peaks where more endemic mammal species might be found. Filling in these gaps will help us better understand the mammalian diversity of the Sierra Madre and to formulate better conservation measures to protect the wildlife of the mountain range.

Acknowledgement

The authors would like to thank the following individuals and organizations for their contributions during the conduct of the fieldwork: N. Antoque, W. Reyes, J. Sarmiento, R. Fernandez, G. Bueser, E. Rico, N. Bartolome, F. Dalin, G. Gee, R. Ambal, L. Co, I. Osbucan, M. De Guia, M. Abuel-Daclan, G. Pallaya, B. Soriano, our guides and porters from the community, and our partner non-government organizations in the Sierra Madre: PROCESS Luzon, CAVAPPED and EWW. We would like to acknowledge R. Trono, A. Antolin, T. Brooks, P. Langhammer and N. De Silva for their support and encouragement. We also thank the Department of Environment and Natural Resources Region 02 and 04 and the Local Government Units for their continued support and encouragement. For their assistance with laboratory work at the Field Museum, we thank J. Phelps, M. Prondzinski, M. Schulenberg, and W. Stanley. Funding for the field work was provided by Critical Ecosystem Partnership Fund and Conservation International. Funding for museum studies was provided by the Barbara Brown Fund for Mammal Research.

References

Alviola, P.A. 2000. The distribution and ecology of bats in the Polillo Islands, Philippines. In: Bennett, D. [ed]. Wildlife of Polillo Island,

Duya et al. 2007. Banwa 4(�):4�-68.

Philippines. University of Oxford-University of the Philippines at Los Banos Polillo’99 Project: Final Report. Viper Press, Glossop.

Balete, D.S. and L.R. Heaney. �997. Density, biomass, and movement estimates for murid rodents in mossy forest on Mount Isarog, southern Luzon, Philippines. Ecotropica �: 9�-�00

Balete, D.S., E.A. Rickart, and L.R. Heaney. 2006. A new species of the shrew-mouse, Archboldomys (Rodentia: Muridae: Murinae), from the Philippines. Systematics and Biodiversity 4: 489-50�.

Balete, D.S., E.A. Rickart, R.G.B. Rosell-Ambal, S. Jansa, and L.R. Heaney. 2007. Descriptions of two new species of Rhynchomys Thomas (Rodentia: Muridae: Murinae), from Luzon Island, Philippines. J. Mammalogy 88:287-�0�.

Brooks, T., M. Abuel, L. Co, O. Coroza, M.V. Duya, M.R. Duya, P. Langhammer, A. Mallari, C. Morales, N. Palomar, R. Rodriguez, B. Tabaranza, and R. Trono. 2004. Targets and priorities for biodiversity conservation globally, and in the Philippines. Agham Mindanaw 2:�-�0.

Corbet, G. and J.E. Hill. �992. The mammals of the Indomalayan Region. Oxford University Press, Oxford.

Danielsen, E, D.S. Balete, T.D. Christensen, M. Heegaard, O.F. Jakobsen, A. Jensen, T. Lund, and M.K. Poulsen. �994. Conservation of biological diversity in the Sierra Madre mountains of Isabela and southern Cagayan province, the Philippines. BirdLife International, Manila and Copenhagen.

Esselstyn, J.A., P. Widmann, and L.R. Heaney. 2004. The mammals of Palawan Island, Philippines. Proc. Biological Soc. Washington ��7:27�–�02.

Heaney L.R., P.C. Gonzales, R.C.B. Utzurrum, and E.A. Rickart. �99�. The mammals of Catanduanes Island: Implications for the biogeography of small land bridge islands in the Philippines. Proc. Biological Soc. Washington �04:�99–4�5.

Heaney LR, D.S. Balete, L. Dolar, A.C. Alcala, A. Dans, P.C. Gonzales, N.R. Ingle, M. V. Lepiten-Tabao, W. Oliver, P.S. Ong, E.A. Rickart, B.R. Tabaranza Jr., and R.C.B. Utzurrum. �998. A synopsis of the mammalian fauna of the Philippine Islands. Fieldiana Zoology, new series 88:�–6�.

Heaney, L.R., P.D. Heideman, E.A. Rickart, R.C.B. Utzurrum, and J.S.H. Klompen. �989. Elevational zonation of mammals in the central Philippines. J. Trop. Ecol. 5:259-280.

Heaney, L.R. and N.A.D. Mallari. 2002. A preliminary analysis of current gaps in the protection of threatened Philippine terrestrial mammals. Sylvatrop �0:28–�9.

66 67

Duya et al. 2007. Banwa 4(�):4�-68.

Heaney, L.R., D.S. Balete, E.A. Rickart, R. C.B. Utzurrum, and P.C. Gonzales. �999. Mammalian diversity on Mount Isarog, a threatened center of endemism on Southern Luzon Island, Philippines. Fieldiana: Zoology, new series 95:�–62.

Heaney, L.R., D.S. Balete , G.A. Gee, MV. Lepiten-Tabao, E.A. Rickart, and B.R. Jr. Tabaranza. 2005. Preliminary report on the mammals of Balbalasang, Kalinga Province, Luzon. Sylvatrop ��:5�-62.

Heaney, L.R., B.R. Jr. Tabaranza, D.S. Balete, E.A. Rickart, and N.R. Ingle. 2006. The mammals of Mt. Kitanglad Nature Park, Mindanao, Philippines. Fieldiana: Zoology, new series ��2:�-6�.

Heaney, L.R., E.K. Walker, B.R. Jr. Tabaranza, and N.R. Ingle. 2002. Mammalian diversity in the Philippines: an assessment of the adequacy of current data. Sylvatrop �0:6-27.

Heideman, P.D. and L.R. Heaney. �989. Population biology of fruit bats (Pteropodidae) in Philippine submontane rainforest. J. Zoology (London) 2�8:565–586.

Heideman, P.D., L.R. Heaney, R.L. Thomas, and K.R. Erickson. �987. Patterns of faunal diversity and species abundance of non-volant small mammals on Negros Island, Philippines. J. Mammalogy 68:884-888

Hoogstraal, H.H. �95�. Philippine zoological expedition, �946-�947. Narrative and itinerary. Fieldiana: Zoology ��:�-86.

Ingle, N.R. �992. The natural history of bats on Mt. Makiling, Luzon Island, Philippines. Silliman J. �6:�-26.

Ingle, N.R. 1993. Vertical stratification of bats in the Philippine rainforest. Asia Life Science 2:2�5-222.

Ingle, N.R. and L.R. Heaney. �992. A key to the bats of the Philippine Islands. Fieldiana: Zoology, new series 69:�–44.

Lepiten, M.V. 1995. The mammals of Siquijor Island, central Philippines. Sylvatrop J. 5:�-�7.

Mearns, E.A. �905. Descriptions of new genera and species of mammals from the Philippine Islands. Proc. U.S. National Museum 28:425-460.

Mildenstein, T.L., S.C. Stier, C.E. Nuevo-Diego, and L.S. Mills. 2005. Habitat selection of endangered and endemic large flying-foxes in Subic Bay, Philippines. Biological Conservation �26:9�–�02

Mudar, K.M. and M.S. Allen. �986. A list of bats from Northeastern Luzon, Philippines. Mammalia 50:2�9-225.

Musser, G.G. �982. Results of the Archbold Expedition No. �08. The definition of the genus Apomys, a native rat of the Philippine Islands. American Museum Novitates 2746:�-4�.

Duya et al. 2007. Banwa 4(�):4�-68.

Musser, G.G. and P.W. Freeman �98�. A new species of Rhynchomys (Muridae) from the Philippines. J. Mammalogy 62:�54-�59.

Musser, G.G. and L.R. Heaney. 1992. Philippine rodents: Definitions of Tarsomys and Limnomys plus a preliminary assessment of phylogenetic patterns among native Philippine murines (Murinae, Muridae). Bulletin of the American Museum of Natural History 2��:��5-��8.

NORDECO-DENR. �998. Technical Report. Integrating conservation and development in protected area management in the Northern Sierra Madre Natural Park, the Philippines. NORDECO, Copenhagen and DENR, Manila.

Oliver, W.L.R., C.R. Cox, P.C. Gonzales, and L.R. Heaney. �99�. Cloud rats in the Philippines: Preliminary report on the distribution and status. Oryx 27:4�-48

Ong, P., L.E. Afuang, and R.G. Rosell-Ambal (eds.). 2002. Philippine biodiversity conservation priorities: A second iteration of the national biodiversity strategy and action plan. Philippine Department of Environment and Natural Resources, Quezon City.

Payne, J., C.M. Francis, and K. Phillips. 1985. A field guide to the mammals of Borneo. Sabah Society, Kota Kanabalu.

Rabor, D.S. �955. Notes on mammals and birds of the Central Northern Luzon Highlands, Philippines. Part �. Notes on mammals. Silliman J. 2:�9�-2�8.

Rabor, D.S. 1986. Guide to the Philippine flora and fauna, Vol. XI: Birds, Mammals. Natural Resources Management Centre, Ministry of Natural Resources and University of the Philippines, Quezon City.

Rickart, E.A. and L.R. Heaney. �99�. A new species of Chrotomys (Muridae) from Luzon Island, Philippines. Proc. Biological Soc. Washington �04:�87-�98.

Rickart, E.A., L.R. Heaney, and R.B. Utzurrum. �99�. Distribution and ecology of small mammals along an elevational transect in Southeastern Luzon, Philippines. J. Mammalogy 72:458-469.

Rickart, E.A., L.R. Heaney, D.S. Balete, and B.R. Tabaranza, Jr. �998. A review of the genera Crunomys and Archboldomys (Rodentia, Muridae, Murinae) with descriptions of two new species for the Philippines. Fieldiana, Zoology, new series 89:�-24.

Rickart, E.A., L.R. Heaney, P.D. Heideman, and R.C.B. Utzurrum. �99�. The distribution and ecology of mammals on Leyte, Biliran, and Maripipi islands, Philippines. Fieldiana: Zoology, new series 72:�-62.

68 69

Duya et al. 2007. Banwa 4(�):4�-68.

Rickart, E.A., L.R. Heaney, S.M. Goodman, and S. Jansa. 2005. Review of the Philippine genera Chrotomys and Celaenomys (Murinae) and description of a new species. J. Mammalogy 86:4�5-428.

Ruedas, L.A., J.R. Demboski, and R.V. Sison. �994. Morphological and ecological variation in Otopteropus cartilagonodus Kock �969 (Mammalia: Chiroptera: Pteropodidae) from Luzon, Philippines. Proc. Biological Soc. Washington �07:�-�6.

Sanborn, C.C. �952. Philippine Zoological Expedition �946–�947: Mammals. Fieldiana: Zoology ��:89–�58.

Sedlock, J.L. 200�. Inventory of insectivorous bats on Mount Makiling, Philippines using echolocation call signatures and a new tunnel trap. Acta Chiropterologica �(2): �6�-�78.

Steppan, S.J., C. Zawadzki, and L.R. Heaney. 200�. Molecular phylogeny of endemic Philippine rodents Apomys (Muridae) and the dynamics of diversification in an oceanic archipelago. Biological J. Linnean Soc. 80:699-7�5.

Thomas, O. �898. On the mammals obtained by Mr. John Whitehead during his recent expedition to the Philippines. Trans. Zool. Soc. London �4:�77-4�4.

Utzurrum, R.C.B. �992. Conservation status of Philippine fruit bats (Pteropodidae). Silliman J.�6:27-45.

Banwa. 2007. 4(�):69-82.

Density and Feeding Preference of the Polillo Tarictic Hornbill Penelopides manillae subnigra in Fragmented Forests of Polillo Island

Ana Katrina Mamangun1 and Juan Carlos Gonzalez2

� Corresponding author. University of the Philippines Los Baños, BS Biology. keimamangun@yahoo.com.

2 Animal Biology Division, University of the Philippines Los Baños. jctgonzalez@yahoo.com.

A portion of a thesis submitted to the Institute of Biological Sciences, University of the Philippines, Los Baños, October 2007.

Abstract

The density and feeding preference of the Polillo hornbill Penelopides manillae subnigra were compared between a heavily disturbed secondary forest and a residual lowland forest. Using transect analysis, the density of tarictics in each forest site was computed. The disturbed forest site exhibited higher tarictic density (4.75 individuals per km²) than the residual forest site (�.25 individuals per km²). The difference in tarictic density between the two sites was affected by several factors, such as the abundance of fruiting trees, the presence of nesting trees and also the degree of anthropogenic disturbance. The disturbed forest site was more fragmented (20% forested area) than the residual forest site (26% forested area). Both forest sites were suitable for sustaining tarictic populations. The disturbed forest site contained more fruiting trees foraged and dispersed by tarictics, thus a higher density was observed. In the residual forest, large trees were observed that are essential for nesting during the breeding season. Male tarictics are territorial especially during this time thus they drive away other tarictics to protect their nests and as a result, a lower density was observed. Forests of Polillo have undergone fragmentation. Being endemic, tarictics tend to tolerate anthropogenic disturbances to the forest habitat.

Keywords: feeding preference, fragmented forests, Polillo Island, population density, Tarictic hornbill

70 7�

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

Introduction

The Polillo tarictic hornbill Penelopides manillae subnigra is a species of small hornbill exclusively found in the Polillo Islands and are locally referred to as the tariktik or taliktik. Generally, hornbills are considered as an important biological species, which can indicate the presence of a healthy forest within an area (Rabor, �977). The tarictics are especially important having been adopted as a flagship species for the Municipality of Polillo because of its endemism within the island group. Although the tarictic is much celebrated by the people of Polillo, it was once categorized as a threatened species due to the decline in its population and restricted range in two small islands totaling only 802 square kilometers. Apparently, the population of tarictics is inherently small due to the limited forest habitats available in the islands (Gonzalez, �997). Tarictics are commonly found residing in secondary lowland evergreen rainforest scattered along the central mountain spine of Polillo and Patnanungan Islands. It has been classified by the Asian Hornbill Network as critically endangered due to its narrow range and limited forest habitats currently available (Asian Hornbill Network, �99�). They were also once poached and sold in the illegal pet trade at a very high price, but are now given special attention by local people of the island. The sustainability of the current remaining fragments of lowland forests on Polillo is of great importance in the conservation of these unique birds. The Polillo Islands has been recognized as one of the essential global hotspots for conservation of biological diversity in the Philippines. It is a cluster of about 27 islands and islets still known to contain patches of secondary and old growth lowland evergreen forests that harbor different species of indigenous flora and fauna. One of the recognized problems in the island, and generally throughout the country, is the fragmentation of habitats being used mostly by different important species. This fragmentation of habitat is brought about by both natural and human activities, which has resulted to the destruction of terrestrial faunas (Catibog-Sinha and Heaney, 2006). Understanding the impacts of forest disturbance and forest fragmentation on the insular ecology of island endemic birds like the Polillo tarictic hornbill would provide essential baseline information to develop more effective conservation and management plans. This comparative survey of hornbill densities found between disturbed

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

and residual forests can serve as a quantitative measurement of the effects of anthropogenic disturbance. Disturbed forests are fragmented due to the conversion of forests into agro-ecosystems while residual forests have extensive cover due to lesser human disturbance. The differences in feeding preference of these hornbills noted between the two forest sites can represent a qualitative measure of these impacts. Generally, this study was conducted to determine whether the degree of forest disturbance and forest fragmentation have an influence on the density and selection of fruits of the tarictic hornbills in Polillo Islands. The study was conducted within the reported breeding period of the Polillo tarictic hornbill from April to May 2006 on the island of Polillo in Quezon Province, Republic of the Philippines. Data collection and fieldwork was conducted on two sites in the northern part of Polillo Island selected to represent a disturbed forest and a less disturbed residual forest. The survey of tarictic hornbills on the disturbed forest site was conducted from �9 April to 2 May 2006. It represented a fragmented secondary lowland evergreen forest interspersed with man-made clearings and agro-ecosystems along the Anibawan River in sitio Salapakan, barangay Anibawan, Municipality of Burdeos in Polillo Island. The succeeding survey for the residual forest site was conducted from 2� May to � June 2006. It represented a less disturbed secondary lowland evergreen forest within sitio Abaca, barangay Lipata, Municipality of Panukulan in Polillo Island. Seasonal and temporal variation did not affect the results of the study.

Materials and Methods

Site selection There were two sites considered in the study, (A) a heavily fragmented, disturbed forest where remnant patches of secondary forest was observed to be interspersed with agro-ecosystems; and (B) a relatively less fragmented, residual forest where extensive forest cover is present and minimal disturbance was observed. Other factors noted for study site selection include the knowledge of the local guide in identification of trees and existing foot trails, freshwater sources like a stream or river present, elevation is within �00 to 200 m above sea level and accessibility and proximity to settlement area is less than 5 km.

72 7�

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

Transect method Line transects � km long were surveyed for hornbills in two sites. Several radiating transects in different directions covered nearly all of the study area for a cumulative total of ten sampled transect lines. The selected transect line was observed twice a day, one in the morning and another in the afternoon. The observations made during morning and afternoon transect counts were considered as separate data sets. With the use of a global positioning system (GPS) recorder (Garmin Etrex), the coordinates and the position of each of the transect lines were recorded. Presence of hornbills in the transect line were noted both as calls heard and individuals seen with notes on their behavior, sex, and age group. Ocular observations were made through the use of a field binocular (10x50 Andubon/Pentax) and calls were recorded using a hand held cassette tape recorder (Sony) with built-in microphone. From the data collected in the implementation of the line transect method, the relative density per km² was computed.

Vegetation analysis The Point Center Quarter Method (PCQM) was used wherein the transect lines of 50 m were established at each study site, one on the sloping hill forest on ridge top, another on the forest near a freshwater body and finally in a disturbed area. Five sampling points were noted and marked and each sampling point was then divided into quarters. The nearest large tree was noted in each quarter. To qualify as a tree, the trunk should be at least �0 cm in diameter. The tree should be within 5 m from the sampling point and the distance, diameter at breast height (DBH) and height of the tree were measured. A sample of the tree was collected, pressed and dried for identification later in the herbarium.

Forest fragmentation analysis The coordinates of the site were obtained through the use of a GPS recorder. From these coordinates, the study site was established on the existing land-cover map then the pixels were established and positioned. The percentage of the forest pixels were counted and computed to compare the extent of forest fragments between the two sites. The percentage of other overlays on the land-use map such as coconuts, mangroves, cultivated areas, water, bare soil, boundaries and clouds were also computed in order to compare the extent of forest cover on the map.

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

Results and Discussion

Site selection The survey was conducted within the large forest block located in the northern part of Polillo Island (�4°49’N�2�° 54’E), the largest among the Polillo island group (Figure 1). The first study site was described to be the disturbed site, since anthropogenic disturbance was evident and a large part of the forest was already converted into agricultural lands. This site was located in sitio Salapakan, barangay Anibawan, Municipality of Burdeos, Polillo Islands, Quezon Province (Figure 2). Several local residents were living on the site and were dependent on the adjacent forest for food and shelter. The local guide for the anthropogenic site was very knowledgeable on movements of the tarictics in his area and also on the local names of the fruiting trees where birds were observed feeding or perching. The local villagers along the Anibawan River were already well informed on conservation issues needed by the different species found in this study site and had designated their own forest rangers. The second study site represents the residual forest site. It is located in sitio Abaca, barangay Lipata, Municipality of Panukulan, Polillo Islands, Quezon Province (Figure 2). It was relatively less disturbed and was observed to have extensive cover of lowland evergreen forest. The forest had minimal human disturbance and also less forest area that was converted for agricultural and human use. Unlike the first site, the residents living near the second study area had little knowledge on the importance of forest conservation, since illegal hunters were sometimes observed.

Relative density in disturbed and residual forest sites Individual transect lines were observed twice a day, with observations done in the morning and afternoon. These two counts done per transect line on each site were considered as different data sets. The approximate area covered by each ocular observation done on each transect line is 40 ha (�km x 400 m) since 200 m are covered on each side of the transect line. Each transect observation is equivalent to about 0.4 sq km. Since �0 transect lines were established in different directions radiating from the central camp, this approximates a near total coverage of the designated study site of 6.6 km². The radial arrangement was done so as not to use the same transect more than once and thus avoid double counting of individuals.

74 75

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

A total of �2� individuals of tarictic hornbills were observed in the disturbed forest site. From the total, only �8 (�0.8%) were seen, while 85 (69.�%) were heard (Table �). Of the �8 individuals seen, 26 (2�.�%) were observed in the morning transect counts and �2 (9.8%) were observed in the afternoon transect counts. Likewise, there were 67 (54.4%) individuals heard in the morning transect counts and 18 (14.6%) in the afternoon. In the residual forest site, there were less tarictic hornbills counted with a total of 28 individuals. From the total, only �0 (�5.7%) were seen directly while �8 (64.�%) were heard (Table �). The �0 individuals seen directly were observed during the morning transect counts and none was observed during the afternoon transect counts. There were �0 (�5.7%) individuals heard during the morning transect counts and another eight (28.5%) heard during the afternoon transect counts. The relative density of hornbills was computed from these transect counts. For the disturbed forest site, there were 6.5 individuals seen directly per km² in the morning and � individuals per km² seen directly in the afternoon. An average of 4.75 individuals per km² was recorded from the disturbed forest site. For the residual forest site, 2.5 individuals per km² were seen directly during the morning transects but none was seen during the afternoon transects. A mean of �.25 individuals per km² occupied the residual forest site (Table 2).

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

Table �. Percentage of tarictic hornbills heard and seen at the two study sites on Polillo Island surveyed from April-June 2006

Study site Total % Individuals % Number % number of seen of calls individuals directly heard

Disturbed forest �2� 8�.5 �8 79.2 85 82.5 Residual forest 28 �8.5 �0 20.8 �8 �7.5 Total �5� �00.0 48 �00.0 �0� �00.0

Table 2. Relative density of tarictic hornbills at the two study sites on Polillo Island at different time of the day

Disturbed forest Residual forest

AM PM AM PM

Total no. of individuals observed 9� �0 20 8 Individuals seen directly 26 �2 �0 0 Relative density per km² 6.5 � 2.5 0 Mean density per km² 4.75 �.25

Figure 1. Map of the Polillo Islands, situated off eastern Luzon Island, Philippines (Source: Gonzalez et al., 2004)

Figure 2. Site �: The northern part of Polillo Island showing the disturbed forest site barangay Anibawan, Burdeos and the residual forest site barangay Lipata, Panukulan (Derived from a GIS map provided by Don De Alban, 2007)

76 77

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

There were more tarictics heard than seen. During the breeding season, male tarictics drive away other tarictics through announcing its territory where nests of the female tarictic and its nestlings are. Calls were frequently made during the early morning and late afternoon by male tarictics to announce and secure their territory from other tarictics and other species of birds. More hornbills were observed in the morning because this is the time when tarictics go out to forage for food and avoid extreme heat from the sun. Daylight temperatures were often high, since field observations were done during the dry summer months of April-June. Tarictics were also observed in late afternoon when it is cooler. There were more individuals observed at the disturbed forest site than in the residual forest site, despite the higher degree of habitat destruction at the disturbed forest site. This is likely because tarictic hornbills tend to tolerate the presence of human disturbance (Gonzalez, 2007). Hornbills are opportunistic frugivorous; they tend to try different fruits available at the time, but still appropriate for their diet and mode of feeding. The residual forest site had a lower population of tarictics because fruiting trees in the forest are also foraged by species of birds and bats that cannot adapt to anthropogenic disturbances. Thus, these species of fruigivorous birds and bats tend to look for a forest habitat where they can forage for fruits and be subjected to fewer disturbances. This results into tarictics’ involuntary elimination of foraging and feeding competition especially during the breeding season. Another factor that could account for the higher population density in the disturbed forest is the greater diversity and higher number of fruiting trees preferred by and foraged by the tarictics. In the residual forest site, there were less fruiting trees where tarictics can feed on but more dipterocarps where tarictics can set up their nests in. Indeed, the nesting issue was also a factor that affected the density of tarictic hornbills in the two sites. In the disturbed forest site in Anibawan, there were smaller trees and few trees were suitable as a nesting site. Based on the interviews with the local residents of Anibawan, large trees and dipterocarps were dominant in the forest several decades ago but were cut down for human use. The trees found in the forest at the time of the survey were early successional and small fruiting trees where tarictics feed on. Also, the tarictics were using the disturbed forest for its foraging activities since no large trees were present for nesting. In the residual forest site

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

however, relatively taller trees and dipterocarps were present that tarictics use for nesting. Territorial males tend to drive away other tarictics to protect their nests thus lesser individuals were observed in the deep forest site.

Analysis of vegetation In the disturbed site, the dominant tree is the Ficus nota, a member of the family Moraceae, reported to be the main fruit foraged by the Polillo tarictic hornbill. This species was dominant in the two sampling areas in the site; the secondary forest at the valley and the forest at the ridge top (Table �). Fruits of F. nota is known to be eaten by bats (Chiroptera) but there were also reports from the local guide that F. nota is also foraged by the Polillo tarictic hornbill. For the residual forest site, Macaranga grandifolia was the dominant tree species. This species is a member of the family Euphorbiaceae. It was dominant in the two sampling areas of the site, the secondary forest at the valley and the forest near a water body (Table �). The tree is often used for timber and wood derivatives. Generally, the forest habitat of the disturbed forest site consisted of small fruiting trees (with mean DBH of �8.52 cm). These trees provide the hornbill with fruits but are not suitable nesting sites for birds such as tarictic hornbills. Although anthropogenic disturbances and activities were present, hornbills adapted to their presence. In the residual forest site, relatively larger trees were present (with mean DBH of �2.05 cm). These trees supply the forest birds with their need for shelter and nesting activities. Protection of the nests

Table �. Dominant tree species noted from the vegetation analysis of the Disturbed forest and residual forest sites

Study site Habitat Dominant tree Importance Family Value

Disturbed forest Forest at Lithocarpus sp. 54.7 Fagaceae ridge top (Babaysakan)* 2° forest at Ficus nota 70.4 Moraceae valley (Tibig) 2° forest near Ficus nota 40.0 Moraceae water body (Tibig) Residual forest Forest at Ficus minahassae �06.6 Moraceae ridge top (Ayimit) 2° forest at Macaranga 84.4 Euphorbiaceae valley grandifolia (Takip-asin) 2° forest near Macaranga 80.� Euphorbiaceae water body grandifolia (Takip-asin)

* names in parenthesis are vernacular names

78 79

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

cause the male tarictic to be territorial, driving away other tarictic hornbills. Consequently, we observed a lower tarictic density in the residual forest site.

Analysis of forest fragments On the magnified map of the specific forest sites, we computed that one pixel is 0.� x 0.� mm, which is equivalent to a land area of 5�2.54 m2 (2�.08 x 2�.08 m) (see Table 4). A total area of 6.5 sq km or equivalent to �2,�2� pixels was covered per study site, inclusive of all transects. The disturbed forest site has lesser and more dispersed forest pixels than the residual forest site. There were 2,459 forest pixels, which is equivalent to �.� sq km of the actual land area of the study site or 20% of the map (Figure �).

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

The forest pixels in the residual forest site were more compact, which indicate a relatively more extensive forest cover. The forest pixels occupied 26% of the total study area. There were �,252 pixels, equivalent to �.7 sq km of the actual land area of the study site (Figure 4).

Percentage of other overlays on the map was also computed. For both disturbed and residual forest sites, the dominant pixel is the area for coconut plantation, which occupies more than half of the total study area (Table 5). This indicates the conversion of forests into agro-ecosystems and kaingin sites. Fragmentation of forested areas has a great effect on the population of tarictics. Fragmentation occurs when forested areas are converted to areas for cultivation and agriculture. When these areas are abandoned and cleared for slash and burn, succession occurs. Plant species succession in the forests are initiated by seed dispersal vectors like tarictics. Seeds dispersed by tarictics into abandoned areas (e.g. abandoned kaingin) have a higher survival and germination rate because of better light penetration and greater accessibility of soil nutrients (Molles, 2006). In less fragmented area, the extensive forest canopy would reduce light penetration and

Table 4. Scale conversion of one pixel established in the map into square meters of the actual study area

Scale Conversion

Length Width Length Width Area (mm) (mm) (m) (m) (m2)

Pixel 0.� 0.� 2�.08 2�.08 5�2.54 Study Area ��� ��� 2,56� 2,56� 6,56�,48�

Figure 3. Disturbed forest site (6.5 km²) located at barangay Anibawan, Burdeos. The map of forest site was obtained from the land cover map (provided by Don de Alban, 2007) with positioned pixels for the analysis of forest pixel percentage

Figure 4. Residual forest site (6.5 km²) located at barangay Lipata, Panukulan. The map of the study site was obtained from the land cover map (provided by Don de Alban, 2007) with positioned pixels for the analysis of forest pixel percentage

80 8�

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

activities, while the less disturbed site that harbor more residual hardwoods was probably used for nesting. Much like the forest structure in these two study sites, various degrees of disturbance has left much of the island’s forest fragmented. These forests tend to undergo ecological succession, wherein seeds from fruits dispersed by the tarictics participate as pioneer species. Consistent movement of tarictics between these forest fragments contributed selective seed dispersal, which can lead to eventual dominance of fruiting trees in secondary forest that are more preferred by Polillo tarictic hornbills. Other factors may have influenced the outcome of the fieldwork in May-June 2006. During transect counts, the absence of tags or markings on each particular individual observed foraging in the disturbed site made it difficult to distinguish them from individuals counted on the less disturbed site, color bands or radio-collars may help identify individual birds and avoid double-counting. Studies on the home range to determine the required minimum forest area needed to sustain an individual tarictic may prove useful in future conservation management. Studies on the preferred dimensions and taxa of nesting trees can also support the management of specific forest habitats used by these tarictic hornbills. Quantitative analysis of fragmentation in particular forest sites can be further compared using more standardized protocols and indices of forest fragments, based on the computation for distance, shape and size of each forest patch. The relative density and feeding preference of tarictic hornbills in variably disturbed forest sites indicate that despite impacts of fragmentation in much of the forests on Polillo Island, these still provide suitable habitat to allow continued persistence of these uniquely adaptable and tolerant forest birds. The ability of this endemic species to adapt to fragmented forests and tolerate anthropogenic disturbances provides them with a natural advantage among other forest endemics to survive in limited forest habitats.

References

Asian Hornbill Network. �99�. Asian hornbill assessment and management plan. Asian Hornbill Network Annual Report (�99�), Singapore.

De Alban, J.D.T. 2007. Spatial profile of local conservation areas in the Municipalities of Burdeos and Panukulan, Quezon Province, Philippines. Unpublished final report for Pioneering Community-

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

provide greater competition for soil nutrients; resulting in lower survival of dispersed seeds. In this manner, plant species depend on tarictic, making it a keystone species of the forest. Anthropogenic disturbance also has a great influence on the significant difference in abundance of tarictics between sites. The tarictic are well known in the islands of Polillo. They are considered as a flagship species in the island because of their endemicity and popularity. Because residents of the island know how important each tarictic individual is to their island, they tend to conserve the remaining populations of this species. They have stopped hunting and poaching, and then prevent the cutting of fruit trees since they need these trees as much as tarictics need them. Another possible factor is the influence of traditional practices of local inhabitants on forest use, such as the avoidance of cutting balete trees due to superstitious beliefs. Given the limitation of forests in Polillo Island, to survive tarictics need to adapt to various degrees of anthropogenic disturbance (e.g. forest degradation). Generally, island endemic taxa are inevitably forced to adapt to fragmentation and tolerate anthropogenic disturbances in order to persist in small islands.

Conclusions

Two fragmented forest sites surveyed in Polillo Island represent different habitats that appeared suitable for the persistence of island-endemic tarictic hornbills. The more disturbed forest site dominated by indigenous pioneer fruit trees was used for foraging

Table 5. Equivalent of different pixel classes to actual area and percentage covered per study site

Classes Site � – Disturbed forest Site 2 – Residual forest

Pixels Area (m2) % Share Pixels Area (m2) % Share

Forest 2,459 �,�09,526 20 �,252 �,7��,8�4 26 Coconuts 8,774 4,672,545 72 8,�46 4,444,6�6 68 Cultivated areas 462 246,0�5 � ��� 60,�77 � Water ��2 70,295 � 99 52,72� 0.7 Mangroves 57 �0,�55 0.55 - - 0.00 Bare soil 6 �,�95 0.05 - - 0.00 Clouds �� �7,57� 0.�5 - - 0.00 Ancestral domain boundary * �85 205,029 4 47� 250,828 4 Barangay boundary * �� 6,92� 0.�5 40 2�,�0� 0.�

TOTAL �2,�2� 6,56�,48� �00.00 �2,�2� 6,56�,48� �00.00

*present boundary line shown in the map

82 8�

Mamangun and Gonzalez. 2007. Banwa 4(�):69-82.

based Conservation Sites in the Polillo Islands Project. The Darwin Initiative, Fauna & Flora International and Polillo Islands Biodiversity Conservation Foundation, Inc.

Catibog-Sinha, C. and L.R. Heaney. 2006. Philippine biodiversity: Principles and practice. Haribon Foundation for the Conservation of Natural Resources, Quezon City.

Clements, T. 2002. Inventory of forest fragments in the Polillo Islands. In: D. Bennett (editor). 2002. Wildlife of Polillo Island, Philippines. Viper Press. Glossop, U.K. (updated online version at www.mampam.com).

Gonzalez, J.C.T. and A.T. Dans. �996. Distribution and ecology of the Polillo tarictic hornbill Penelopides manillae subnigra and notes on threatened fauna in the Polillo Islands, Philippines. Unpublished report. Institute of Biological Sciences, U.P. Los Baños (Philippines); Vogelpark Avifauna (Netherlands); Fauna and Flora International (U.K.); Zoological Society for the Conservation of Species and Populations (Germany); and Oriental Bird Club (U.K.).

Gonzalez, J.C.T. �997. The ecology and distribution of birds in the Polillo Islands, Philippines. University of the Philippines, Los Baños. MS Thesis.

Gonzalez, J. C. T. 2007. Fragmented forests and isolated islands: The distribution and status of Philippine endemic hornbills. In: The active management of hornbills and their habitats for conservation, CD-ROM. Proc. 4th Intl Hornbill Conf. Mabula Game Lodge, Bela-Bela, South Africa. Naturalists and Nomads, Pretoria.

Kemp, A. �995. The hornbills (Bucerotiformes). Oxford Univ. Press, New York.

Molles, M.C. Jr. 2006. Ecology: Concepts and applications. �rd edition. McGraw-Hill, USA.

Rabor, D. �977. Philippine birds and mammals. University of the Philippines, Quezon City.

Tsuji, A. �996. Hornbills: Masters of tropical forests. Hornbill Research Foundation, Bangkok.

Banwa. 2007. 4(�):8�-97.

Community-Based Wetland Conservation Protects Endangered Species in Madagascar: Lessons from Science and Conservation

Richard T. Watson1,2, Lily Arison René de Roland2, Jeanneney Rabearivony2, and Russell Thorstrom2

� Corresponding author. rwatson@peregrinefund.org 2 The Peregrine Fund, 5668 West Flying Hawk Lane, Boise, Idaho 8�709, United

States of America

Abstract

Survival of the Madagascar fish eagle (Haliaeetus vociferoids) is threatened by habitat loss. Of a population estimated at �00-�20 breeding pairs, �0 pairs breed on three adjacent lakes in western Madagascar. Fishing on the lakes is the main livelihood of local Sakalava people. From 1991 through 1995 we documented a massive influx of migrant fishermen who abused local traditional resource extraction rules and threatened the livelihood of local inhabitants, as well as the survival of one of the world’s most endangered eagles. Migrants’ economic incentive was strong. In 1995 per capita income from fishing was about $1500 for the six-month season, about 7.5 times the national annual average. Fish stocks were rapidly diminished through the fishing season as catches diminished to the point where fishermen gave up fishing before the end of the season. Fish stocks were lowest when Madagascar fish eagle nestlings fledged, affecting annual productivity. The most serious impact of fishermen may be on the lake-side forest, which was used as a source of dugout canoes and wood to fuel fish-drying fires. To conserve this important breeding site we worked with the local community to enhance and enforce traditional resource utilization rules that helped prevent loss of fish eagle breeding habitat, reduce nest site disturbance, and sustain prey availability. We used a �996 law to empower communities to control natural resource use by creating two community associations with authority to enforce local rules. We helped the associations become effective through training, advice, logistical, and scientific support.

Keywords: Community, conservation, habitat, law, Madagascar fish eagle, persecution, raptors, Sakalava.

84 85

Watson et al. 2007. Banwa 4(�):8�-97.

Introduction

The island endemic Madagascar fish eagle (Haliaeetus vociferoides) is critically endangered (Stattersfield and Capper, 2000) with a small population limited to wetland habitats on Madagascar’s western seaboard (Rabarisoa et al., �997). Habitat degradation and human persecution are the most likely causes for the species rarity (Watson et al., 2000a; Watson and Rabarisoa, 2000). Survival of the Madagascar fish eagle requires conservation of remaining suitable habitat and the natural resources on which the eagle depends, and control of human persecution. This paper describes The Peregrine Fund’s (TPF) efforts to conserve critical fish eagle habitat and reduce persecution of eagles by empowering the local Sakalava community of the Manambolomaty Lakes complex, western Madagascar, to manage and sustainably harvest the lakes and forest on which they depend, and which they share with about �0% of the Madagascar fish eagle breeding population.

Materials and Methods

We studied the ecology and natural history of the Madagascar fish eagle, the impact of humans on key natural resources that people shared with the Madagascar fish eagle, and we developed and applied new conservation methods to effect a change in human behavior to benefit fish eagles and their habitat. Each of these methods will be briefly described.

Madagascar Fish Eagle studies Studies on the Madagascar fish eagle were designed to measure the species’ distribution and abundance, and determine what factors affect them. We began with surveys within the species’ known range in Madagascar. Surveys were completed on foot, and by car and boat, in all suitable habitats along the coast and on lakes and rivers within about �00 km inland of the west coast of Madagascar. Surveys were conducted annually during the breeding season over at least five years from 1991 through 1995 (Rabarisoa et al., 1997), and every fifth year since then to detect change in the distribution and abundance of the species. To understand what factors affect the species’ distribution and abundance, we measured population parameters in a sample of breeding pairs located in and around the Manabolomaty Lakes complex, about �00 km due west of the capital (Antananarivo). Parameters measured included nesting density, annual nest

Watson et al. 2007. Banwa 4(�):8�-97.

occupancy, breeding productivity, survival, and turn-over at the nest. Causes of breeding failure and mortality were determined whenever possible (Watson et al., 1999; Watson and Razafindramanana, 1999; Watson et al., 2000a). Behaviors, such as breeding behavior, dispersal, habitat selection, and migration, can have a significant impact on a species’ distribution and abundance. Behavior was observed, documented, and interpreted in the context of when and where it occurred, and which individuals were involved (Berkelman et al., �999a, �999b, 2002; Rafanomezantsoa et al., 2002; Tingay et al., 2002, 2004). Studies on the genetic relationship between individuals of a pair, their progeny, and extra-pair adults at the nest were used to interpret behavior which included the unusual participation of more than two adults at the nest. Molecular genetics were also used to understand the species phylogeography and the genetic consequences of its rarity (Tingay et al., 2002, 2004).

Studies of the impact of humans on natural resources Studies to measure human use of natural resources which people share with the Madagascar fish eagle and its impact on the fish eagles, were initiated in �99� with a major study concluded in �996 and annual monitoring occurring since then. Observations of increasing numbers of fishermen active on the three main lakes of the Manambolomaty complex occurred annually from �99� through 1993. Systematic counts began in 1996 of the number of fishermen fishing, the number of dug-out canoes in use, the number and location of fishermen camps, the number of fish-drying fires, and the number, size, and distance to shore of cut trees (for either firewood or dug-out canoe construction). Counts were done simultaneously by three teams of two observers working around the perimeter of each lake, and were repeated at the beginning, middle and end of the fishing season. Fishermen dialogue surveys were done by one team of two people who answered 22 questions from dialogue with the head of household and from direct observation in each fisherman camp. Questions provided data on fishing effort and success, fishing nets (type, length, mesh size), income from fishing and market forces, and utilization of wood from the surrounding forest (Kalavah and Razanrizanakanirina, �997; Razanrizanakanirina and Watson, �997; Watson and Rabarisoa, 2000; Watson et al., 2000b). Measuring the human impact on Madagascar fish eagle productivity has been accomplished annually since �99� by observing territory and nest occupancy of banded, individually recognizable,

86 87

Watson et al. 2007. Banwa 4(�):8�-97.

fish eagles at all 10 territories that exist on the Manambolomaty lakes complex. Breeding success was measured as the proportion of eggs laid that hatched and fledged young (Watson et al., 1999).

Conservation implementation methods Conservation centered on the premise that, first, local residents who were indigenous to the area and who depend upon the availability of fish in the lakes and wood in the forest for their livelihood would have an incentive to control the use of these natural resources to guarantee their future, and second, that if there were enough fish in the lakes and trees in the forest for people then there would be enough for Madagascar fish eagles. We utilized a new (in �996) law that was designed to decentralize the control of natural resources away from government by empowering local communities to be responsible. The local community was unaware of this law when we began work in 1996, and had insufficient funding, transport, or communication to learn about the law or follow the process for implementing it. The Peregrine Fund’s primary role was to gather the information needed, share it with the local community, and provide the logistical resources needed to implement the law, mainly transport, communication, and a small amount of funding. In �996 the government of Madagascar, in compliance with the second phase of Madagascar’s Environmental Action Plan (PE-II), approved Law No. 96-025 to decentralize natural resource management by encouraging local communities to manage their own natural resources under a “management charter” with the government, known as Gestion Locale Securisée (GELOSE). The process adopted by the local communities to achieve the GELOSE charter was as follows: (�) The local population made a request to transfer authority for

management of one or more natural resources to the mayor of the community, which was eventually agreed under written contract.

(2) With expert services provided by TPF and others, the community demonstrated the technical foundation for this request.

(�) With TPF’s help, the community selected an “environmental mediator” to facilitate discussions and negotiations to:

(a) understand the respective points of view of stakeholders on the natural resources to be managed,

(b) elaborate a common vision of the long-term future for natural resource management, and

Watson et al. 2007. Banwa 4(�):8�-97.

(c) define the procedure to permit the effective management of natural resources.

(4) With the mediator’s help the community established the requirements of an adequate system of management that responds to the “Management Contract” with goals of conservation, sustainable development, and development of resources.

(5) The community finalized the contract by an authorized method.

Results and Discussion

Madagascar fish eagle ecology Surveys for breeding Madagascar fish eagles at their nests from �99� through �995 detected at least 222 adult individuals including 6� pairs, �6 probable pairs, 24 single adults, and �8 immature birds (Rabarisoa et al., �997). Assuming that all probable pairs were breeding, we estimated the fish eagle breeding population was 99 pairs (95% confidence interval = 78 to 120 pairs) in the area searched. This estimate was about twice the number previously estimated in the period �980-�985 (Langrand and Meyburg, �989), due mainly to our greater search effort. The number of breeding pairs in some localities had declined since �985, suggesting either a general population decline or movement of these pairs to other sites. Three major areas of concentration of the species were located: (�) Tsiribihina River near the southern extent of the species’ range, (2) Manambolomaty Lakes complex (Antsalova region) to the north of the Tsiribihina River, and (�) the northernmost coastal and estuarine region between Mahajamba Bay and Nosy Hara (an island) near Madagascar’s northern tip (Rabarisoa et al., �997). Our best guess of the total population size, including the area of the species’ range that we had not thoroughly searched, was about �20 pairs. Monitoring samples of known nest sites from �996 through 2006 suggests little or no change in the species’ distribution and abundance in the following decade. During the study period from 1996 through 2006, the number of territorial pairs of fish eagles around the Manambolomaty lakes complex varied from eight to ��, with generally higher numbers active (laying eggs), and significantly higher numbers successful (fledging young) since community-based conservation took effect in 2001 (mean number of young fledged per occupied nest 1996 to 2000 = 4.4 ± 0.55 young/pair; mean number of young fledged per occupied nest 2001 to 2006 = 6.5 ± 1.22 young/pair; t = 3.53, df = 9, P=0.006).

88 89

Watson et al. 2007. Banwa 4(�):8�-97.

Fish eagles nesting in the Manambolomaty Lakes complex utilized alternate nests, some building new nests annually while others occupied the same nest, averaging a 78% annual relocation rate. Despite the use of alternate nests within their territory, nest spacing between adjacent pairs was fairly constant at �.68 ± 0.66 km (n = 49). Home range areas ranged from 244 to 487 ha, with a mean of �50 ha ± ��9 ha. Pairs with the smallest home range were located on islands in the lakes, where ranging behavior was probably reduced by abundant shoreline foraging habitat and/or the easier defense of territories surrounded by exposed water (Watson and Razafindramanana, 1999). Madagascar fish eagle nesting and foraging habitat parameters including nest, nest tree, surrounding vegetation, and adjacent water parameters, were measured at 56 nests found along the western seaboard of Madagascar. Descriptive statistics were used to look for consistent patterns among habitat parameters. While certain trends were apparent, such as always nesting within sight of water and in the largest trees, there was little evidence that would suggest a negative human impact on nest site or foraging habitat availability exists wherever large trees and water-woodland ecotone remain, yet many such apparently suitable sites were unoccupied (Watson et al., 2000a; Berkelman et al., �999a, �999b, 2002). Fish eagles nested further from water than the African fish eagle Haliaeetus vocifer (Brown, �980) and bald eagles H. leucogaster in north America (Corr, 1974; Grubb, 1976; Kralovec et al., 1992), probably reflecting the effect of harvesting by fishermen of tall trees close to water for construction of dug-out canoes (Watson et al., 2000a; Watson and Rabarisoa, 2000). Cutting of trees for canoe construction may limit availability of suitable nest sites if all large trees within sight of water are removed. Introduced Tilapia spp. were the most common fish species available to fish eagles, and were the dominant prey species selected. Introduction of Tilapia may have benefited the Madagascar fish eagle by providing abundant and easily captured prey (Berkelman, �999a). Direct human persecution (collecting chicks from the nest and trapping adults) was observed to occur with regularity in the Manambolomaty area. Chicks were either eaten or were sold as pets, rarely surviving long. Adults were trapped, a foot removed, and then released. Only one adult has been seen to survive this abuse (Tingay et al., 2004), while about ten adults with a single foot missing have been found dead. The persecution of adult fish eagles

Watson et al. 2007. Banwa 4(�):8�-97.

stems from a local superstitious belief that the foot of a living eagle can act as a powerful talisman (Kalavah and Razanrizanakanirina, 1997). The population effect of increased adult mortality from this persecution is more significant than an equivalent level of persecution of nestlings, but the combined increased adult mortality and reduced recruitment is harmful to the species’ survival and contributes to its rarity and absence from suitable habitat.

Natural resource use by humansTree cutting for canoes and firewood In �99�, when we began studying Madagascar fish eagles, there were about 30 fishermen active on the lakes. By 1996 when we did the first quantitative survey of fishermen, we counted 300 fishermen and 275 dug-out canoes active on the lakes. There were 42 temporary fishermen camps and five permanent fishing villages. At night we counted a minimum of 32 fish-drying fires burning after 2200 hours. The density of cut trees in the forest ranged from 15 to 290 trees/ha. Trees used for canoe construction were large in diameter (mean = 6�.� cm in diameter) and averaged �40 m from the shore. Trees used for firewood were 18.4 cm in diameter and averaged 65 m from shore. A significant increase in the number of trees cut since fishermen numbers began to increase after 1991 was evident from estimated cut date, based on decomposition since cutting (Watson and Rabarisoa, 2000). Fishermen and fishing Fisherman dialogue surveys at a sample of 18 temporary fishing camps and one village revealed that fishermen came from 14 villages, the most distant of which was 50 km from the lakes. Extrapolating numbers, we estimated there were around 300 fishermen and 600 family members, totaling 900 people, about ten times the number present at the lakes when we first began in 1991. Migrants arrived at the lakes in June when the fishing season opened, and left again in November when fish catch was almost nil or December when the season officially closed. All fishermen agreed that fish catch diminished through the season, indicating a major impact on fish numbers. On average, each fisherman’s camp burned five fish-drying fires and we estimated by extrapolation that 200 fish-drying fires existed around the lakes. Fish were dried in front of fires on sticks holding three fish. Fishermen’s estimates of time needed to dry fish averaged 1.12 h, and each fire dried an average of 23 sticks of fish at a time. Fish were sold for cash or bartered for goods, such

90 9�

Watson et al. 2007. Banwa 4(�):8�-97.

as rice, coffee, oil, and batteries. Fish buyers came from 11 villages, mostly within 100 km of the lakes, and carried fish to commercial centers for resale (Watson and Rabarisoa, 2000).

Fish harvest Using the data above and making several assumptions, we estimated the number and weight of fish extracted from the lakes each season, the income derived from fishing, and the amount of time fires must burn to smoke and dry all the fish. The last estimate was used to gauge the impact of wood collecting on the surrounding forest and availability of nesting sites for Madagascar fish eagles. Assuming the number of fishermen was constant through the season, and there was a linear change in catch rates through the season, we modeled the relationship between time (days) from the beginning of the season and daily fish catch (fish per day) with the equation: daily fish catch = 84,578 - 460.5 x time (r2 = 0.92, P<0.05, df=2). Using this equation, we estimated total catch from the three lakes during an average 5.6 month fishing season to be 7,671,930 fish, or about 1,918 metric tons assuming each fish weighed an average of 250 g (Watson and Rabarisoa, 2000).

Wood consumption The number of hours in front of a fire required to dry the daily catch (fish-hours/day) was estimated by dividing the reported daily catch by three (number of fish per stick) and again by 23 (number of sticks per fire) and multiplying by the average time to dry the fish of 1.12 h. Using reported daily catch of fish at the beginning, middle and end of the season, and assuming a linear relationship through the season, then total number of fire-hours per day = 1,372.9 - 7.475 x time (r2 = 0.92, P<0.05, df=2). Using this equation, we estimated 124,528 fire-hours to dry the entire seasons’ catch. Based on local experience, we estimated that it would take about 8�,000 m of �0 cm diameter log to fuel these fires, all of which was collected from the forest surrounding the lakes (Watson and Rabarisoa, 2000).

Economic incentive The price of fish varied with demand from 500 to 1000 Francs Malgache (Fmg) in �996. Assuming the price averaged 750 Fmg then the total catch for the season was about �,9�7,982,500 Fmg (then about US$ 479,495) and each fisherman made about $1,562 for the season’s work. Annual average per capita income in 1996 was $225 to 250 in Madagascar, so fishing at these lakes provided an income

Watson et al. 2007. Banwa 4(�):8�-97.

about 7.5-times greater, a strong incentive to endure the hard work and hardship of camping on the lakes away from home for several months (Watson and Rabarisoa, 2000).

Conservation results In 1993 TPF first proposed the idea of a community-based conservation project to protect wetlands and natural resources used by local Sakalava people living in the Manabolomaty Lakes complex around Lakes Befotaka, Soamalipo, and Ankerika and shared with endangered Madagascar fish eagles (Watson and Rabarisoa, 2000; Watson et al., 2000b). Discussions with the tompondrano (traditional keeper of the lakes) of Lakes Soamalipo and Befotaka began at that time to better understand the existing traditional fisheries rules. The idea was based on the simple concept that, provided people left fish eagles alone, then if there were enough fish in the lakes for people to catch and enough trees in the forest for people to use, there should be enough of both these limiting resources for Madagascar fish eagles to survive also. Nest sites (trees) and food (in this case, fish) are the two main ecological resources that limit raptor population density and distribution (Newton �979). By �996 the local population and authorities at the villages of Soatana and Masoarivo (tompondrano, mayors, and elders) agreed that there were problems of over-fishing the three lakes and over-use of forest resources around the lakes, and wanted to do something about them by enforcing existing laws, traditional edicts, and dina (taboos). In June �996, TPF helped the community leaders to write these existing traditional laws and dina, and announce them at public meetings on 29 June �996. However, the writing and announcing of the laws proved insufficient to alter the behavior of immigrant fishermen from other parts of Madagascar who were the main cause of over-exploitation of fish and forest resources for profit. It didn’t help that the authorities and local elders avoided their responsibilities, did not communicate among themselves, and participated in the fisheries exploitation for profit. In response to these problems TPF selected Mr. Ravo as a mediator to begin the GELOSE process with assistance from TPF sociologist Daurette Razandrizananakanirina, local technicians Loukman Kalavah and Jules Mampiandra, and biologist Rivo Rabarisoa. Their acceptance in the local community was extremely important to be able to communicate with, and collect and pass on comments and information to the local people, stakeholders, and communities. Meetings and presentations were held to identify

92 9�

Watson et al. 2007. Banwa 4(�):8�-97.

local and regional authorities and other stakeholders. These were followed by informing the local authorities and stakeholders of the existence of Law 96-025, what it could do to help solve the problems they faced, and how to proceed with establishing the community charter (GELOSE). Two community associations (FIZAMA for Lakes Soamalipo and Befotoka, and FIFAMA for Lake Ankerika) were established with the help of TPF staff to take responsibility for natural resource management and control, and following through the GELOSE process. A mission statement was written by the associations with TPF’s guidance, and general agreement by all authorities and parties to proceed with developing the GELOSE was gained at public meetings and workshops, after which the authorities publicly announced the start of the GELOSE process during the ceremonies to open the fishing season. The GELOSE community management charter was developed by community leaders, written down by TPF staff, and then revised several times until a consensus by all authorities and stakeholders was met and finally voted-on in public. The community, represented by the associations, then applied to the Malagasy government for official recognition of the GELOSE under Law 96-025. An important element in managing the natural resources, and in obtaining the acceptance of the associations by the Malagasy government, was to establish methods for measuring and monitoring change in resource use and availability. Fishing and tree harvest surveys were established with TPF expertise to document fishing and tree harvesting impacts, origin of fish buyers and their markets, fishing camp locations on the three lakes, and land-use around the lakes. In 1997, TPF also supported student studies on fish, lemurs, and botanical resource-use to gain a better understanding of the effects of resource use on other fauna and flora in the area for support of the conservation effort on fish eagles, their habitat and other biodiversity. Throughout �997 local community dialogues, meetings, and presentations continued in collaboration with other non-governmental organizations to provide information and help resolve problems related to the GELOSE process. During this period, TPF and other NGOs had been working with the Malagasy government to designate the three lakes as one of the country’s first Ramsar wetland sites of international importance. On 2 March �998, the Manambolomaty Lakes Complex, which includes the three lakes (Befotaka, Soamalipo, and Ankerika), the smaller Antsamaka Lake, and a 500 m band of the Tsimembo forest around the lakes, were designated as one of the first two Ramsar sites in

Watson et al. 2007. Banwa 4(�):8�-97.

Madagascar. This international designation gave more importance to the protection of this area under a strategy aimed at management of resource use and conservation of the wetlands, maintenance of the ecological value of the site, continued research, and local capacity building in research, monitoring, and management of natural resources. The designation of the three lakes as a Ramsar site gave more importance and value to the GELOSE process, and for supporting management and resource control by the two community associations, FIZAMA and FIFAMA. From �999 to 200�, TPF continued supporting the GELOSE process by resolving problems and other issues with FIZAMI and FIFAMA, and helping them to enforce their own management guidelines and policies on persons who disobeyed the rules. TPF also assisted the associations’ requests to transfer natural resource management from central government to the local community. On 29 September 200� the two associations, FIZAMI and FIFAMA, were given a three-year probationary period to prove to the government that they could manage their natural resources and enforce resource use policies. In 2002 community meetings continued and TPF continued supporting the associations financially, logistically, and with training and equipment. The associations opened bank accounts in Morondava by depositing money they collected from issuing fishing and fish-buyer permits. Fishing limits and tree harvest limits were successfully enforced and limited to sustainable rates for the first time in over ten years. In 200�, the community associations continued their work with financial and logistical support from TPF by demarking the GELOSE management boundary, a community effort that took three months (August to October) of hard work to accomplish. The boundary was marked and labeled with cement blocks at trail and road crossings and the line was a cut swath of �.5 to 2 m in width. TPF paid for the work associated with this boundary delimitation. A tree nursery was established and operated by TPF technicians. About �,2�4 tree seedlings were raised, of which �,�84 were transplanted to several denuded forest areas around the three lakes. The two local associations made marked progress in their control and management of the fishery and forest resources. In 2004 the associations completed their three-year probationary period and applied for approval and authorization by the Malagasy government. Offices were built for each association in the village of Ankiranagato for FIZAMI and the village of Bejea for FIFAMA with funding from Ramsar and logistical assistance from TPF.

94 95

Watson et al. 2007. Banwa 4(�):8�-97.

On 30 June 2005 the two associations received the official government authorization and contract for a �0-year period to manage their natural resources. In June 2005 the two associations also received the World Wildlife Fund “Gift to the Earth” award for their pioneering role in developing the GELOSE process for resource management and conservation. The bank accounts for both associations continued to grow from the issue of resource use permits and with some of this money the associations bought rice to sell to local community members at a reduced rate during the annual rice shortage period, thus providing another tangible benefit to the community for limiting the fishery. Local personnel received training in tree nursery operation. The associations continued to receive increasing support from local authorities: police, judicial, and forestry and fishery departments. In 2006 a fishing permit covering a 4-year period (September 2006 to September 20�0) was issued by the Regional Fishery Representative, which became another important milestone for the two associations. The community associations have successfully limited the number of fishermen on the lakes, both local and migrants, and limited the fish catch, fishing season, net mesh-size, fish-drying methods and fuel wood consumption, numbers of canoes built and trees cut for construction, and implemented reforestation to restore tree abundance on the lakeshore.

Conclusions

This paper describes a conservation process that began with research to measure the distribution and abundance of the Madagascar Fish Eagle and understand what factors limit them, and expanded into a community-based wetland conservation project to protect fish eagles in their stronghold, the Manambolomaty Lakes complex, which supports about �0% of the species’ global population. In the first three years of work, the research documented the low fish-eagle population size (about �20 breeding pairs globally), its distribution along the western seaboard of Madagascar, the population’s largest stronghold, and the occurrence of human persecution. This knowledge was enough to justify conservation effort focused on the species’ stronghold, but studies since then have been important for improving and refining our understanding of the species’ behavior and its population and genetic consequences, and for detecting change in population size, density, distribution and productivity in response to conservation interventions.

Watson et al. 2007. Banwa 4(�):8�-97.

The community-based wetland conservation project was based on the simple premise that if there were enough fish in the lakes and trees in the forest to sustain the fishing community, then there should be enough of both resources to sustain fish eagles, provided people stopped persecuting fish eagles. Local residents had a traditional “keeper of the lakes,” the tompondrano, who established rules and taboos that limited fishing. Following his death in 1991, by 1993 his heir faced overwhelming numbers of migrant fishermen invading the lakes, and fishing, camping, and using the forest in disregard for local traditions. The tompondrano, mayors, and elders felt powerless to protect their livelihood. The intervention by TPF began by rallying community leaders to work together to take action, and by providing information on a new (�996) law designed to decentralize control of natural resources from government to village level. With awareness, strength in numbers, and logistical and moral support from TPF, the local community began a guided process to institutionalize mechanisms to control fishing and receive government authority through a “natural resource use charter” (GELOSE). The process required stakeholder participation, buy-in, and commitment which wavered at times but was always restored with encouragement and persistence of TPF staff. Over the decade-long process the community saw tangible results of their efforts, experienced the benefits of taking control, and underwent a transformation from helplessness to empowerment and success. In addition to facilitating community-empowerment, TPF’s intervention consistently explained the message that Madagascar fish eagles were exceptionally rare and unique to Madagascar, they were a valuable part of the community’s cultural and natural heritage, and that persecution of eagles was harmful to the species. We made no attempt to strike bargains with the community to protect the eagles, but through awareness they came to accept that persecution was not acceptable and its prohibition should be included among their taboos. Among the criteria for successful implementation of this community-based conservation strategy, we believe that employment and training of technician-level staff from the local community helped build important links and trust between TPF and the local community. Skepticism, fear, and distrust among the local community were most effectively handled by community members who worked for and got to know us and understand our motives. Second, although funding commitments tend to be offered in finite cycles of just two or three years during which measurable

96 97

Watson et al. 2007. Banwa 4(�):8�-97.

results are expected to be achieved, the success of this project depended on taking time (many years) to develop trust with and among community members, an outcome that can not be rushed or measured but we believe was critical.

Acknowledgments

We are grateful to the Liz Claiborne and Art Ortenberg Foundation for their long-term support of this project, and also to the John D. and Catherine T. MacArthur Foundation, and other important donors. We are especially grateful to Loukman Kalavaha and other field technicians from the local communities whose knowledge, patience, and dedication of effort helped ensure this project’s success.

References

Brown, L.H. �980. The African fish eagle. Bailey Bros. & Swinfen, Folkstone.

Corr, P.O. �974. Bald eagle Haliaeetus leucocephalus alascanus nesting relating to forestry in southeastern Alaska. University of Alaska, Fairbanks. Unpublished MS Thesis.

Berkelman, J., J.D. Fraser, R.T. Watson. 2002. Nesting and perching habitat use of the Madagascar fish eagle. J. Raptor Res. �6(4):287-29�.

Berkelman, J., J.D. Fraser, and R.T. Watson. �999a. Madagascar fish eagle prey preference and foraging success. Wilson Bul. ���(�):�5-2�.

Berkelman, J., J.D. Fraser, and R.T. Watson. �999b. Lake selection by Madagascar fish eagles. The Auk ��6(4):976-98�.

Grubb, T.G. �976. A survey and analysis of bald eagle nesting in western Washington. University of Minnesota, Minneapolis. Unpublished MS Thesis.

Kalavah, L. and D. Razanrizanakanirina. �997. Socio-economic status in �996 of the region around three lakes, Befotaka, Soamalipo, and Ankerika, in western Madagascar. The Peregrine Fund, Boise, ID.

Kralovec, M.K., R. Knight, G.R. Craig, and R.G. McLean. �992. Nesting productivity, food habits, and nest sites of bald eagles in Colorado and southwestern Wyoming. Southwestern Naturalist �7(4):�56-�6�.

Langrand, O., B.-U. Meyburg. �989. Range, status, and biology of the Madagascar sea eagle Haliaeetus vociferoides. In: Meyburg, B.U. and R.D. Chancellor (eds.) Raptors in the modern world. World Working Group on Birds of Prey, Berlin. pp. 269-278.

Watson et al. 2007. Banwa 4(�):8�-97.

Newton, I. �979. Population ecology of raptors. Poyser, London.Rabarisoa, R., R.T. Watson, R. Thorstrom, and J. Berkelman. �997.

Status of the Madagascar fish eagle Haliaeetus vociferoides in �995. Ostrich 68(�):8-�2.

Rafanomezantsoa, S., R.T. Watson, and R. Thorstrom. 2002. Juvenile dispersal of Madagascar fish eagles tracked by satellite telemetry. J. Raptor Res. �6(4):�09-��4.

Razandrizanakanirina, D., and R.T. Watson. �997. The process of developing the community-based wetland conservation project: �997-�998. In: Watson, R.T (ed.). Madagascar Wetlands Conservation Project: Developing community-based conservation in a proposed wetland biosphere reserve in Madagascar. Prog. Rpt. III, �995-�996. The Peregrine Fund, Boise, ID. pp. ��-44.

Stattersfield, A.J. and D.R. Capper. 2000. Threatened birds of the world. Lynx Edicions, Barcelona.

Tingay, R.E., M.L. Clarke, R.T. Watson, R. Thorstrom, and L. Kalavah. 2004. Survival and behavior of a one-footed Madagascar fish eagle in the wild. J. Raptor Res. �8(�):85-88.

Tingay, R., M. Culver, E.M. Hallerman, J.D. Fraser, and R.T. Watson. 2002. Subordinate males sire offspring in Madagascar fish eagle (Haliaeetus vociferoides) polyandrous breeding groups. J. Raptor Res. �6(4):280-286.

Watson, R.T. and R. Rabarisoa. 2000. Sakalava fishermen and Madagascar fish eagles: Enhancing traditional conservation rules to control resource abuse that threatens a key breeding area for an endangered eagle. Ostrich 7�(� & 2):2-�0.

Watson, R.T. J. Berkelman, R. Rabarisoa, R. Thorstrom, and C.R.B. Watson. 2000a. Description of nesting and foraging habitat of the Madagascar fish eagle Haliaeetus vociferoides: A conservation initiative. Ostrich 7�(�&2):��6-�40.

Watson, R.T., A. Andrianarimisa, D. Razandrizanakanirina, and L. Kalavaha. 2000b. Madagascar fish eagle Haliaeetus vociferoides and community-based wetland conservation. In: Chancellor, R.D. and B.U. Meyburg (eds.) Raptors at Risk. WWGBP/Hancock House, Berlin. pp. ���-�40.

Watson, R.T. and S. Razafindramanana. 1999. Nearest neighbor nest distances, home range and territory area of the Madagascar fish eagle (Haliaeetus vociferoides). J. Raptor Res. ��(4):��5-��8.

Watson, R.T., S. Razafindramanana, R. Thorstrom, and S. Rafanomezantsoa. �999. Breeding biol ogy, extra-pair birds, productivity, siblicide and conservation of the Madagascar fish eagle. Ostrich 70(2):�05-���.

98 99

BanwaInstructions for Authors

General Information

Banwa is a peer reviewed inter- and multi-disciplinary journal. It was originally intended for the faculty, students, and research, extension, professional services (REPS) staff of the University of the Philippines Mindanao and for researches on Mindanao or Southern Philippines and its environs. With two issues every year, it publishes scholarly articles within the research thrusts of the University—biosystems and environment, culture and the arts, and policy and management. External contributors, contributing researchers, or non-UP constituents are welcome to submit papers whose topic falls within the above-mentioned research areas. Banwa was realized through the collective effort of the College of Humanities and Social Sciences in 2004. The journal was later adapted by the University as the its lone, refereed academic journal.

Type of Contribution(s) to the Journal

�. Original papers not previously published elsewhere, except in a preliminary form, or are not being considered for publication anywhere.

2. Short communications should be as completely documented, both by reference to the literature and description of the research procedures employed, as a regular paper. They should not occupy more than eight (8) printed pages. a. Report of preliminary results of important research; andb. Newly developed methodology or modification of existing methodology, possibly

description of first test

The Author or Contributor

1. The author must issue a certification that the article has not been published and is not under consideration for publication elsewhere.

2. He/she must obtain permission to use copyrighted materials.�. The author may be a faculty, student, research personnel, or a non-UP constituent.

Preparation and Submission of the Manuscripts

1. Manuscript/s to be submitted to the Editorial Board of Banwa should be written in English or Filipino and must be in MS Word (Saved as Word 97-200� document) or other compatible format. Articles written in Filipino must be accompanied by an informative abstract and summary written in English.

2. Manuscript/s should be submitted with certifications that: (a) it has undergone English language editing, (b) it has been reviewed by two colleagues and revised accordingly. In addition, the names and contact details of three additional suggested reviewers should be provided.

3. Manuscript/s should be written double-spaced in 12-point type, Times New Roman font, 2.5 cm (� in) margin on all sides, and provided with line numbers by enabling the line numbering features of the word processing program. The paragraph indentation is five spaces. Manuscripts should be submitted online as digital/soft copies e-mailed to banwa@upmin.edu.ph.

4. Manuscripts in general should be organized according to the disciplinal requirements of the paper. However, it is suggested that the order of presentation outlined below be followed:

Title Page - The title should be printed as Title Case. The first letters of each word should be capitalized except for articles (e.g. a, an, the), coordinate conjunctions (e.g., and, or, not, but) or preposition (e.g. of, in, on, between, after, before). However, the first and last words of the title should be capitalized regardless of part of speech. In titles with a marked break in the form of a punctuation, capitalize the word that immediately follows the break, regardless of part of speech.

Name(s) of author(s) - Authors should identify themselves only in the title page that should precede the article for ease in undertaking the review process and ensure anonymity. Write the complete name with middle initials. Each author’s mailing address and/or institutional affiliation must be indicated in the title page as well as the author to whom correspondence should be addressed including his/her e-mail address. Indicate whether the research is the portion of a thesis or dissertation and the sources of fund of the research.

For the running title, include a shortened version of the title of the article, not more than 40 letters in length, on the upper left-hand corner of each page

Abstract - After the title page, print a 150 to 200 word informative digest of the objectives, methods used and significant findings of the article. The title preceding the abstract should also be printed in Title Case.

Include also a Summary, not more than 500 words and written in language for a general audience, of the findings and their significance. This summary will be incorporated in the Article Digest section of Banwa. It should be understandable to an interested reader here and abroad. New techniques, unfamiliar methods, and terms cannot be defined should be explained.

Keywords - The keywords and phrases following the Abstract should be alphabetically arranged and should reflect the contents of the paper. Give not more than eight keywords/phrases.

Abbreviations - For easy reference, an alphabetically arranged sequence of abbreviations and acronyms and their meanings are printed after the keywords. Avoid abbreviations in the title and abstract, although they may be used in graphs, tables, figures and legends. Acronyms are to be spelled out first and then enclosed in parenthesis at first mention. Avoid using acronyms and abbreviations as the first word of a sentence or a heading. Rewrite the sentence or spell out the term. Use abbreviations only for terms used at least three times.

Text - The following sequence is suggested for the body: Introduction, Materials and Methods (not Methodology), Results and Discussion, Conclusions, Acknowledgment and Literature Cited.

4. Tables, graphs, photographs and illustrations should be submitted in separate files. Photographs, illustrations, and graphics should be of publishable quality (TIFF, or maximum quality in JPEG), and should be in �00 to 600 dpi with dimensions of at least �0 x �5 cm (4 x 6 in). Figures included in the article should be in black and white, or grayscale only.

5. Statistics: All statistical procedures, including methods of analysis, should be identified.

Number of replications and sub-samples, transformations, and statistical tests should be stated. Tables of analysis of variance are not normally reported. Present the results of analyses in the appropriate tables or figures.

6. Following the format and the order of presentation, the article should be no more than 5000 words or 20 pages, inclusive of photos, graphs, tables and illustrations.

The Editorial Board reserves the right to charge specific fees to authors of articles with colored photographs, graphs and illustrations.

7. Style: Banwa adapted the ASHS Publications Style Manual (ASHS, 2002). Additional

guidance can be obtained from the Scientific Style and Format: The CSE Manual for Authors, Editors and Publishers (CSE, 2006).

References

American Society for Horticultural Science (ASHS). 2002. ASHS publications style manual. ASHS, Alexandria, VA. 27 May, 2008. <http://www.ashs.org/downloads/style_manual.pdf>

Council of Science Editors, Style Manual Committee (CSE). 2006. Scientific style and format: The CSE manual for authors, editors and publishers. 7th ed. The Council, Reston, VA.

�00