Introduction

Information about the freshwater and estuarine species of fish of Buton and Kabaena is very limited. Little is known about their distribution or ecology. A field guide to the fishes of Western Indonesia and Sulawesi (Kottelat et al., 1993) contains keys and descriptions of more than 1000 species, but specific information about either Buton or Kabaena is not included. Other relevant publications include a preliminary checklist of 102 species of fish from Sulawesi that contains 13 new species of which 6 species are endemic to the region (Tjakrawidjaja & Hadiaty, 1996). A guide to the freshwater fishes of Australia that includes descriptions of species that may occur in Indonesia has been published recently (Allen, Midgley & Allen, 2002). The results obtained from the second year of a survey of the freshwater and estuarine fishes of these two islands are the subject of this report. Data were obtained on the fish populations present in some of more remote parts of these islands where relatively intact rainforest still remains. This year, four undergraduate projects were completed on the effects of deforestation on fish populations and on aspects of ecomorphology of some of the more abundant species. This work is both timely and important for the long-term management and protection of Indonesian rainforest.

Objectives

In this second year of the survey the objectives were to:

§ To complete a survey of freshwater and estuarine fish on Buton and Kabaena Islands.

§ To extend the database begun last year on the numbers of each species at each location.

§ To record standard lengths of all individuals and to preserve representative specimens for further taxonomic work at Balitbang Zoologi, Bogor and at the British Museum.

§ To investigate possible relationships between the distribution of fish in different rivers, the surrounding land use, the environmental characteristics of the water and the microhabitat.

Methodology

The survey was conducted over an 8 week period. The sites included a fast flowing river in forested hills on Kabaena, a number of different rivers on Buton in the vicinity of Labundo-Bundo, and a river system in a forested part of North Buton, inland from the coastal town of Maligano.

Fish samples were collected using portable electrofishing equipment capable of generating a potential of up to 400 volts between the fishing electrodes. The fish shocker was powered by a 24-volt battery and set to generate pulsed DC current to the electrodes. To standardise fishing effort between sites, three operators fished continuously for 60 minutes. At a few sites, where access was difficult or where battery life was a limiting factor, samples were collected over 30 minutes. In North Buton, electrofishing techniques were employed using similar equipment powered by a portable generator.

Seine netting in the larger rivers and where a suitable bank was available to land catches was used to obtain additional samples. The net was approximately 1200mm deep and 18m long, with a mesh diameter of 15mm (measured as the maximum distance between diagonally opposite corners with the mesh extended).

The position of each site was recorded using GPS. Temperature, pH, dissolved oxygen concentrations and water conductivity were measured using portable meters. A general description of each site was made that included typical depths and widths of the watercourse, the presence of deep pools and riffles and the degree of disturbance/land use of the surroundings. The substrate was characterised using the Wentworth Classification of substrate particle substratum size.

Fish were anaesthetised in benzocaine and the Standard Length of all the fish in each sample recorded. Fish were identified to species where possible and representative individuals photographed alive. Up to 6 individuals of each species were prepared as museum specimens. Before fixation, specimens were pinned to pieces of foam to extend the dorsal, anal and caudal fins. After approximately 7 days fixation in 10% formalin (4% formaldehyde), specimens were transferred to 70% alcohol.

Initial Results

Virtually all of the species collected during the 2001 season were also present in collections this year. For freshwater environments, a species accumulation plot over time indicates that the majority of the most common species in the region have now been recorded. In estuaries however, samples still tend to contain species that have not been seen previously so that a checklist of fish occurring in estuaries is presently incomplete. This probably reflects the occasional penetration of small numbers of marine species where they do not normally occur, and to the fact that many marine species use estuaries as nursery areas during the early part of their life cycle.

Species identification continues to be problematic, especially for members of the Gobiidae family. Almost half the species that have been recorded in the survey are gobies, but the taxonomy of the group is not well developed and a major revision will be needed before species can be identified with any certainty. It seems very likely that some of the gobies present in the samples, especially the smaller forms that may have been overlooked in the initial survey last year, will be new to science.

A recent taxonomic review of a family of fish commonly referred to as ‘halfbeaks’, the Hemiramphidae (Meisner, 2001), will help establish the identity of the two members of this family that occur in the region, one that apparently only occurs on Kabaena and the other only on Buton. Although correct identification is technically difficult, the author of the review has agreed to assist in identification. It appears from the results of the surveys of rivers flowing through relatively intact rainforest that this group of fish may contribute useful indicator species because they require overhanging vegetation as a source of the insects on which they feed.

This year, a number of samples were obtained from the upper regions of rivers on both islands. Such hill stream areas are often isolated from lower sections of rivers by extensive series of waterfalls on both islands. The fish assemblage present above these waterfalls is frequently distinct from those further downstream. One implication of this is that waterfalls are effective barriers that have isolated fish populations on these islands; conditions that will tend to facilitate speciation. A number of the potentially new species of gobies from these sites may have a restricted distribution limited to hill streams of just a few or perhaps only one river.

Initial results from project work on the effects of land use/deforestation adjacent to water courses suggests some differences in the fish fauna may be attributable to the removal of trees and/or the use of pesticides and fertilizers in areas where intensive agriculture occurs (e.g. rice paddies). Other projects have investigated the relationship between fish morphology and species distribution in different habitats. An extensive dataset has been collected on 5 common species. Further analysis may lead to the identification of important morphological characteristics that are characteristic of fish that live in specific microhabitats. This analysis may also assist in establishing key differences between closely related species that will be helpful in identification.