A system of suspending traps from a movable balloon has been developed and tested on pollinating and wood boring Coleoptera in a tropical rain forest canopy in Pasoh Forest Reserve, Peninsular Malaysia. The balloons, 2.5 m in diameter, were filled with helium gas for levitation and moored in different directions by three pieces of thin rope. Traps in which benzyl acetate and ethanol were used as attractants for the pollinators and wood borers respectively were suspended from the balloon at different heights. The trapping results show that the balloon system can be used for studying canopy insects and their vertical distribution on a spatial and temporal scale.

Pollinating and wood boring insects play important roles in regeneration and decomposition processes in tropical rain forests (Whitmore 1984, Yoneda et al. 1977). Many such insects inhabit the forest canopy (Basset 1990, Paoletti et al. 1991).

Methods such as large towers or cranes (Mitchell 1982, Parker et al. 1992), chemical fogging (Basset 1990, Hijii 1983, Ochi et al. 1968), and climbing techniques (Lowman 1992, Nadkarni & Matelson 1991, Perry 1978) have been used to study canopy insects. Halle (1990) developed a giant balloon air ship to bring researchers on a flexible raft up to the forest canopy in the South American tropical rain forest.

However, each method has disadvantages. Building a tower or crane in a tropical forest is extremely costly and disturbs the forest ecosystem. Tree climbing methods are dangerous and require climbing skills. Chemical fogging is inappropriate for determining the vertical distribution of insects and may disturb the natural environment. Halle's air ship system is expensive and requires a large number of staff.

We therefore developed an inexpensive and movable balloon system in combination with attractant traps and examined the effectiveness of the system in a lowland tropical rain forest.

Study sites

Two balloons were set up about 100 m apart in different sites (A and B) at Pasoh Forest Reserve, Peninsular Malaysia. The forest was a near virgin lowland dipterocarp forest dominated by Shorea and Dipterocarpus and the sites were both about 800m from the forest edge. The maximum height of the trees in the two sites was about 30 m.

Balloon system

The balloon was a sphere 2.5 m in diameter and made of white colored vinyl chloride film 0.12 mm in thickness. The weight of the balloon including ropes was about 3 kg. Pure lift of the balloon was about 4 kg when each balloon was filled to capacity with 7 cubic metres of helium. One or two attractant traps were suspended from the balloon by means of a thin rope. The balloon was moored with three pieces of thin rope in three different directions to prevent the traps from swaying (Fig. 1a).

Attractant traps

The plastic traps were of the water-trap type, equipped with a roof and two vertical collision plates 30 cm in width and 24 cm in height set at right angles to each other (Fig. 1b). The collecting vessel was 25 cm in diameter and 15 cm deep and was filled with 900 ml of water to which 1 ml of detergent and 0.5 g sorbic acid (an antiseptic) was added. The attractant was set in the center of the collision plates. The weight of the trap when filled was 1.5 kg.

White colored traps with benzyl acetate as an attractant were used for flower visiting insects (Ikeda et al. 1993) whilst black colored traps with ethanol as an attractant were used for wood boring Coleoptera (Ikeda et al. 1980).

Trapping schedule

Traps were suspended from the balloons at a height of 7 m, 15 m, and 21 m between 15 to 25 January 1992. Additionally, a set of traps unattached to the balloon were suspended at a height of 1.5 m from the ground. Insects trapped were generally collected daily. Since it was possible to suspend only two traps from the balloon at any one time, the height of the traps had to be varied over different periods to obtain different trapping heights. A comparison has therefore been made of beetles trapped on consecutive days in traps run simultaneously at different heights.

Trapping results

The traps yielded beetles from thirteen families but the dominant flower visiting genera attracted to benzyl acetate were Dasyvalgus, Macronota (Scarabaeidae) (now change to Mecinonota) and Endaeus (Curculionidae) whilst the dominant wood borer attracted to ethanol was Xylosandrus crassiusculus (Scolytidae). Table 1 shows the total number of individuals for each species from these dominant genera attracted to benzyl acetate in traps placed over consecutive days at three different heights. A much larger number of individuals were trapped at 15 m than at 1.5 m in nine parallel trapping days, particularly Dasyvalgus sp.1, whilst an intermediate number were obtained in a trap placed at a height of 7 m for two days during this period (Fig.2). In contrast to these flower visiting beetles, the reverse situation was seen in the wood borer X. crassiusculus in which nine specimens were trapped in ethanol traps placed at 1.5 m in comparison to only one at 15 m over five parallel trapping days.

A comparison of traps for flower visiting beetles suspended at a common height of 15 m in site A and B is also given in Table 1. Both plots had a similar species composition but the abundance of Endaeus sp.1 was much higher in site B. A further 6 specimens of this species, and no others, were obtained when the trap was raised for one day to 21 m. It is also worth noting that no beetles were obtained in a trap placed at 1.5 m in site B over the same period as the trap at 15 m in this site. This strengthens the result obtained in site A that the flower visiting beetles were much less frequently trapped near the forest floor.

Performance of the balloon system

The balloon system remained effective throughout the ten days that it was in use. Mooring the balloon with three cords is necessary and enabled it to remain reasonably stable during the study period (Photo 1). Although three people worked to set up this balloon system, it could probably be accomplished with just two. It was easy for one person to move the traps down and back up when collecting the insects in them.

The data demonstrates that balloon-suspended trap systems can be used to study canopy insects and the vertical distribution of insects on a spatial and temporal scale. Flower visiting beetles were more abundant in traps suspended at greater heights corresponding with the canopy levels of the forest where they feed. Likewise, the greatest activity of X. crassiusculus was nearer the forest floor where most dead wood material occurs.

This system proved to be inexpensive, portable and required fewer operators in comparison to a crane, a tower or canopy fogging methods. It therefore enables larger replication in carrying out ecological studies on the insects inhabiting forest canopies. Another advantage of the system is that it can be set up even in very small gaps in the forest.

The method can also be adapted for a variety of traps, such as sticky traps, light traps, suction traps and other attractant traps, as long as they are within the weight limit that the balloon system can support. It can also be used in capture-mark-release methods for the study of populations, dispersal and migration.

One disadvantage, however, is that the system does not enable direct observation of the canopy or the activity of insects in it. The balloon also requires regular refilling with helium gas when in long term use for more than two weeks, as the gas gradually escapes.

Strong winds could jeopardize the balloon system but mooring it with three cables enabled it to remain stable throughout the study period when the maximum wind speed was 14.4 km/h. Between November 23 1991 and February 16 1993 in Pasoh Forest Reserve, the wind speed exceeded 14.4 km/h on only 11 days, when the maximum recorded was 18.5 km/h. It is therefore unlikely that wind damage would occur regularly in a tropical forest, except perhaps in areas severely affected by a monsoon.

We are grateful to Drs. A. Nobuchi, M. Sato, K. Morimoto, H. Makihara and H. Kinuura for identifying insects collected in this study. We also thank Messrs. T. Matsumura and T. Hattori for their help in field work and Mr. Y. Maruyama for his assistance in carrying out this study. Dr. T. Ikeda and Mr. K. Ozaki critically assessed the manuscript. This study was part of a joint research project between the Forest Research Institute of Malaysia, Universiti Pertanian Malaysia and the National Institute for Environmental Studies of Japan under the Global Environment Research Programme funded by the Japan Environment Agency (Grant No. E-2).

Basset Y. (1990) The arboreal fauna of the rainforest tree Argyrodendron actinophyllum as sampled with restricted canopy fogging: composition of the fauna. Entomologist 109(3):173-183.

Fukuyama, K., Maeto, K. & Kirton, L.G. 1994. Field tests of a balloon-suspended trap system for studying insects in the canopy of tropical rainforests. Ecological Research, 9:357-360.

Halle F. (1990) A raft atop the rain forest. National Geographic. 178(4):128-138.

Hijii N. (1983) Arboreal arthropod fauna in a forest. I. Preliminary observations on seasonal fluctuations in density, biomass, and faunal composition in a Chamaecyparis obtusa plantation. Japanese Journal of Ecology 33(4):435-444.

Ikeda T., Enda N. Yamane A., Oda K. & Toyoda T. (1980) Attractants for the Japanese Pine Sawyer, Monochamus alternatus Hope (Coleoptera: Cerambycidae). Applied Entomology. & Zoology 15(3):358-361.

Ikeda T., Ohya E., Makihara H., Nakashima T., Saito T., Tate T. & Kojima K. (1993) Olfactory responses of Anaglyptus subfasciatus Pic and Demonax transilis Bates (Coleoptera: Cerambycidae) to flower scents. Journal of the Japanese Forestry Society. 75(2):108-112.

Lowman M. D. (1992) Leaf growth dynamics and herbivory in five species of Australian rain-forest canopy trees. Journal of Ecology. 80(3):433-447.

Mitchell A. W. (1982) Reaching the rain forest roof / A handbook on techniques of access and study in the canopy. UNEP.

Nadkarni N. & Matelson T. (1991) Fine litter dynamics within the tree canopy of a tropical cloud forest. Ecology 72(6):2071-2082.

Ochi K., Katagiri K. & Kojima K. (1968) Studies on the composition of the fauna of invertebrate animals in the canopy strata of pine forests. I. On the method of investigation and simple description of the fauna. Bulletin of the Government Forest Experiment Station. 217:167-170.

Paoletti M. G., Taylor R. A. J., Stinner B. R., Stinner D. H. & Benzing D. H. (1991) Diversity of soil fauna in the canopy and forest floor of a Venezuelan cloud forest. Journal of Tropical Ecology. 7(3):373-383.

Parker G. G., Smith A. P. & Hogan K. P. (1992) Access to the upper forest canopy with a large tower crane sampling the tree-tops in three dimensions. Bioscience 42(9):664-670.

Perry D. R. (1978) A method of access into the crowns of emergent and canopy trees. Biotropica 10(2):155-157.

Whitmore T. C. (1984) Tropical rain forests of the Far East, 2nd edn. Clarendon Press, Oxford.

Yoneda T., Yoda K. & Kira T. (1977) Accumulation and decomposition of big wood litter in Pasoh Forest, West Malaysia. Japanese Journal of Ecology. 27:53-60.

Fig. 1. Schematic representation of a) the balloon suspended trap system and b) an individual attractant trap.

Mecinonota regia sumatrana

Dasyvalgus vethii

D. niger

D. striatippennis

D. sumatranus

D. biguttatus

D. setipygus

D. dohli

Xylosandrus crassiusculus

Photo. 1 Scarabaeid beetles cought by balloon traps.

Table 1. The number of flower visiting beetles collected using benzyl acetate as an attractant in traps placed simultaneously at different heights or in different sites over consecutive days.

(Species Traps at different heights Taps in different sites in site A at 15 m )

	1.5m	15m	1.5m	7m	15m	Site A	Site B
Macronota sp.	1	6	1	1	2	1	2
Dasyvalgus sp.1	0	29	0	1	13	7	5
D. sp.2	0	1	0	1	1	0	0
D. sp.3	0	1	0	1	0	0	0
Endaeus sp.1	0	2	0	0	0	1	8
E. sp.2	0	2	0	0	1	1	0
Total individuals	1	41	1	4	17	10	15
Total species	1	6	1	4	4	4	3