In the afternoon of 25 May, the calls recorded by RH were played by MvW at the site where the species had been observed in March. The large trees in which the birds had been previously seen had very recently been felled, but two orioles responded and flew to surrounding trees, about 50 m apart and 25 m from the observer. The birds were not observed, however, but repeatedly responded to each other and to the tape recorder with exactly the same call as being played back. On 26 May at 10h30, one oriole responded to playback and moved from degraded forest with bamboo stands to a large tree, locally known as Dita Alstonia scholaris, in the company of three Blackish Cuckooshrikes. This bird was seen well for several seconds before disappearing into dense foliage: it was entirely green/yellow with heavily streaked wings, the bill was large and blackish (not reddish as in Whitelored Oriole which is endemic to northern Luzon) and the white lores were absent (again separating it from White-lored Oriole). It was later seen moving between large trees (Dracontomelon dao, Endospermum peltatum, Anisoptera thurifera and Alstonia scholaris), always in the canopy, in the company of Blackish Cuckooshrike, Bar-bellied Cuckooshrike, Black-and-white Triller Lalage melanoleuca and Balicassiao. It no longer responded directly to playback and did not approach. However, it did call from time to time and received a response from a second bird on the other side of the river, some 100 m away. Playback was used over 26–29 May in a variety of habitats near Ambabok, but no additional responses were received.

Three different sounds were recorded: (1) a clear, mournful, slightly descending whistle lasting 0.5 s, repeated irregularly every 1-2 s, resembling Fig. 1, sonagram 1; (2) a slightly higher, rising whistle, usually with a slight terminal inflection, lasting about 0.75 s, also irregularly repeated every 1-2 s, resembling Fig. 1, sonagram 2; (3) a harsh, rolling call, not unlike a cricket, lasting about 0.25 s, usually given at a rate of 2 calls/s and repeated every 2 s (resembling Fig. 1, sonagram 3), usually given as a long series in response to playback but occasionally interspersed with the other calls. In May 2003, an additional call was heard but not recorded: a clear three-tone whistle lasting 1 s with the second tone higher than the first, and the third lower than the first and second. These calls differ considerably from the varied vocabulary of White-lored Oriole in which the commonest two or three note calls are longer and flutier. The typical call of the closely related Philippine Oriole Oriolus xanthonotus, which is absent from Luzon, is a slightly longer, higher-pitched and flutier whistle than Isabela Oriole, and rises noticeably towards the end.

Ambabok was revisited during 1–4 April 2004 by MvW. A pair of Isabela Orioles, probably the same as those seen in 2003, was observed in exactly the same area. One individual responded strongly to whistled imitations of the calls described above. On 3 April it was captured in a mist-net. The bird was measured, documented (Plates 1–3) and released.The description in Kennedy et al. (2000) is accurate with the exception of the colouring of upperparts and underparts; we suggest describing the upperparts as dark olive-yellow (yellow-olive in Kennedy et al.) and the underparts as bright yellow with an olive wash (olive-yellow in in Kennedy et al. (2000) is accurate with the exception of the colouring of upperparts and underparts; we suggest describing the upperparts as dark olive-yellow (yellow-olive in Kennedy et al.) and the underparts as bright yellow with an olive

wash (olive-yellow in Kennedy et al.). The contrast between upperparts and underparts is rather strong.The main field characters distinguishing the species from the similar White-lored Oriole are the stout greyish bill and the lack of whitish lores. The pair was observed regularly, sometimes also with Black-naped Oriole in addition to the flockspecies
observed in 2003. On four occasions Isabela Orioles were observed eating caterpillars.

CONSERVATION

Ambabok (17o01’26”N 122o10’44”E) is the location of a former small settlement on the edge of the Sierra Madre mountain range on the banks of the Catalangan River.The village was deserted in 1994 and open grassy fields and small patches of degraded forest and bamboo remain. The elevation is around 200 m. The large open area with forest patches is bordered by selectively logged forest which gradually becomes less disturbed eastwards towards the mountains. Administratively Ambabok is part of barangay Dibuluan in the municipality of San Mariano, Isabela province. Ambabok is situated just inside the Northern Sierra Madre Natural Park (NSMNP), the largest of ten priority protected areas in the Philippines under the National Integrated Protected Area System (NIPAS). Although protected on paper, law enforcement in the Ambabok area (on the western side of the NSMNP) is generally lacking and the lowland forest on this side of the park is under heavy pressure from illegal small-scale logging. During the May 2003 and April 2004 visits, a large number of loggers were active in Ambabok, which is used as a campsite, and in the surrounding forest. Selective logging for the larger and most valuable timber trees has now disturbed all remaining lowland forest on this side of the Sierra Madre. Several internationally funded conservation projects are active in the area. A new project funded by the Netherlands and implemented by WWF Philippines intends to address the illegal logging in the park through a combination of community development, awareness-raising and strengthening of law enforcement.

The Isabela Oriole thus still occurs in Isabela province in the municipality (San Mariano) where the holotype was collected in 1894 (Dickinson et al. 1991) and near Disulap (8 km south-west of Ambabok) where 11 specimens were collected in 1961 when the species apparently was still quite common (Collar et al. 1999). The recording and description of the call will perhaps lead to more observations of this species. However, the facts that (1) many observers have been looking for the species during recent years without success, including a DENR/BirdLife survey team in 1991 and 1992 (Danielsen et al. 1994, Poulsen 1995) plus many keen birdwatchers brought to the area by Tim Fisher’s bird expeditions in the Philippines, and (2) the first author did not observe any other Isabela Orioles during three years of extensive fieldwork in the area, indicate that the species is probably genuinely rare rather than merely cryptic.

The reasons for the apparent rarity and patchy distribution of the species are unclear, but the known elevation range is 50–440 m (Collar et al. 1999), indicating that it could be a lowland forest specialist.