ROMER’S TREE FROG MANAGEMENT AND BREEDING 99 Int. Zoo Yb. (2008) 42: 99–108 DOI:10.1111/j.1748-1090.2007.00034.x Captive management and breeding of Romer’s tree frog Chirixalus romeri C. B. BANKS1, M. W. N. LAU2 & D. DUDGEON3 1Melbourne Zoo, Melbourne, Vic. 3052, Australia, 2Kadoorie Farm & Botanic Gardens, Tai Po, Hong Kong SAR, and 3Department of Ecology & Biodiversity, The University of Hong Kong, Pokfulam, Hong Kong SAR E-mail: [email protected] In the early 1990s, a significant portion of the range of the In 1991, C. romeri was only known from endemic Romer’s tree frog Chirixalus romeri was threa- three Hong Kong islands: Chek Lap Kok, tened by the construction of the new international airport in Hong Kong. Melbourne Zoo, Australia, partnered with Lantau and Lamma (Karsen et al., 1998; The University of Hong Kong to establish a captive- M. Lau, pers. obs). Subsequent extensive breeding programme as part of a broader conservation field surveys throughout Hong Kong con- initiative to secure the species in the wild. Large numbers firmed the frog’s presence on these three of frogs were successfully bred at both facilities, under- islands, plus Po Toi Island in Hong Kong pinning the success of subsequent releases. (Lau, 1992, 1998). Key-words: captive breeding; conservation partner- Concerns over the frog’s status were raised ships; growth; management; Romer’s tree frog. in 1991 by the World Wide Fund for Nature (Hong Kong) and other conservation bodies, arising from plans by the Hong Kong Gov- INTRODUCTION ernment to build a new international airport on Chek Lap Kok Island, with associated The rhacophorid genus Philautus comprises construction also having a significant impact many species found in tropical and subtropi- on C. romeri breeding sites on neighbouring cal Asia (Inger, 1999; Zhao, 1999; Frost, Lantau Island. Chek Lap Kok Island origin- 2004). Romer’s tree frog Philautus romeri ally covered an area of c.5km2 and the was discovered in 1952 living in a small, airport development required its expansion remote cave by the sea, on Lamma Island, to about 8 km2 of airport, with bridges joining Hong Kong (Smith, 1953). It was then ob- Chek Lap Kok Island to the neighbouring served a year later, but was subsequently not island of Lantau. A road and rail link was also seen again until it was rediscovered in 1984. to be constructed along Lantau’s northern In 2001, it was tentatively placed in another coast first to Tsing Yi Island and then to the genus, Chirixalus, based on its having a free New Territories. A new town was to be built tadpole stage, which members of the Philautus on northern Lantau, resulting in the loss of genus do not have (Bossuyt & Dubois, C. romeri breeding sites in that area. The 2001). concerns were communicated to the IUCN/ The maximum adult length of Chirixalus SSC Conservation Breeding Specialist romeri is 20 mm (Plate 1). The body is tan to Group, which called for support from the brown above, with a darker brown or black world’s zoos in implementing two key re- X-shaped marking on the back, made up of commendations in the Environmental Impact two lines that do not always meet in the centre Assessment for the project: (1) to establish (Karsen et al., 1998). Although referred to as secure off-shore captive populations; (2) to a tree frog, it is usually found at ground level undertake studies of C. romeri ecology and or sometimes in low bushes. identify suitable re-release sites. Int. Zoo Yb. (2008) 42: 99–108. Journal compilation c 2007 The Zoological Society of London No claim to original US government works 100 AMPHIBIAN CONSERVATION on the island were visited at least once every month during this period. A visit in April 1992 included staff from MZ and resulted in 30 frogs being captured and transferred to MZ 4 days later. Searching for C. romeri was undertaken in all appropriate microhabitats during the day and at night, although most frogs were found at breeding sites at night. Numbers were recorded and adults were sexed, with the presence of vocal sacs and smaller size of << used to distinguish the genders. Although Smith (1953) stated that << of C. romeri do Plate 1. Romer’s tree frog Chirixalus romeri. Michael not possess secondary sexual characteristics, Lau. << collected in this study had white nuptial pads on the first two fingers and a single Melbourne Zoo (MZ), Australia, agreed to external subgular vocal sac. However, the support the project in October 1991, focusing nuptial pads were not easily apparent to the on establishing a secure offshore captive naked eye and were not used as a diagnostic population at the Zoo. The population would feature in sex determination. be held until such time as appropriate release Males started calling soon after dark and sites had been identified and then the frogs were located using headlamps by slowly would be returned to Hong Kong. (Zoo scanning along the edges of shallow pools. Berlin also agreed to participate in the con- Most calling << were not collected initially, servation-breeding programme and received on the premise that they would entice the ,, 30 founder individuals. Unfortunately, all out. Many more << were encountered than these frogs died in the summer of 1994.) ,, and all ,, encountered were collected. Funds were subsequently provided by the From October to December 1992, all frogs Royal Hong Kong Jockey Club Charities Ltd found were collected, as it was approaching (RHKJCCL) to enable ecological studies and the end of the rescue operation. Egg clutches a translocation programme to be undertaken were collected in the early visits but this for the species by The University of Hong practice was discontinued because most eggs Kong (HKU) from 1991 to 1997. did not survive the long journey back to The species was not identified as threa- HKU. tened when this project commenced, but has since been listed under the Hong Kong Wild Captive management and breeding Animals Protection Ordinance (Chapter 170) and as Endangered by the IUCN (The World Housing Adults at HKU and MZ were main- Conservation Union) (Karsen et al., 1998; tained in acrylic terrariums (60 cm  40 cm  IUCN, 2006). 40 cm high at HKU and 61 cm  31 cm  31 cm high at MZ) with lids lined with plastic mosquito netting. The substrate was METHODS local soil (2–5 cm deep) with a top layer of leaf litter (Plate 2), which was changed every Collection 6 months. A container (35 cm  25 cm at Collection of C. romeri from Chek Lap Kok HKU and 15 cm diameter at MZ) of shallow Island by HKU researchers began in Novem- aged tap water (up to 3 cm deep) was placed ber 1991 and ended in December 1992, just in each tank for egg laying. Tap water was before the last patch of habitat was cleared for allowed to stand for 24–48 hours before use. airport construction. The remaining habitats A few dead leaves, twigs and stones were Int. Zoo Yb. (2008) 42: 99–108. Journal compilation c 2007 The Zoological Society of London No claim to original US government works ROMER’S TREE FROG MANAGEMENT AND BREEDING 101 lights at MZ corrected the problem of bone curvature in the frogs. 2. The depth of the substrate was increased to 9–10 cm. The composition was also chan- ged to a 4–5 cm-deep layer of palm peat on top of a 5–6 cm layer of aquarium pea gravel. All this rested on a piece of fine wiremesh and plastic egg-crate mesh. Below this, at the very bottom of each tank, was a 1–3 cm space where leached water and toxins were drained away through drainage holes in the base of each tank. The deeper substrate allowed for Plate 2. The diminutive Romer’s tree frog Chirixalus higher humidity to be maintained owing to romeri in leaf litter. Michael Lau. decreased moisture loss. It also accommo- dated the limited Zoo staff resources by placed in the water tray on which ,, could reducing or minimizing servicing of the in- attach their eggs. The terrariums were sprayed creasing number of holding tanks. The addi- with water daily to increase the humidity. Up tion of small grass tussocks allowed the frogs to 25 adults (more << than ,,) were housed the opportunity to get off the tank floors. Frog in each terrarium at HKU, while ten adults wastes were also leached away by daily were housed in each container at MZ. spraying. Anaerobic bacteria utilize nitrites At HKU, the terrariums were kept in air- and nitrates as they are leached through conditioned rooms illuminated by 40 W fluor- substrates and, hence, the deeper substrates escent tubes connected to a timer set with a acted as biological filters by converting or- 12 hour daylight cycle. Air temperature in the ganic compounds into organic wastes. rooms varied from 20–29 1C in summer to 16–25 1C in the winter, while water tempera- These two modifications were adopted at ture fluctuated from 19–27 1C in summer to HKU in July 1992. 15–25 1C in winter. At MZ, two of the tanks were placed on a heated, off-exhibit table in Diet Adult frogs at HKU and MZ were fed a the Zoo’s Reptile House at 20–27 1C. The staple diet of artificially raised young crickets third tank was placed in an adjacent off- Teleogryllus oceanus and wingless fruit flies exhibit area that could be viewed by Zoo Drosophila melanogaster once every 2 days.