The WWCs were manufactured following the technology developed by Forest Products Research and Development Institute (FPRDI) [1]. Small diameter (12–40 cm) G. arborea logs were cut into 40 cm long billets. The billets were debarked and converted into 0.5 mm × 5.0 mm × 300–400 mm wood excelsior. The excelsior were soaked for 48 h in water and air-dried prior to mixing with the required amount of cement, water and accelerator. The mixed materials were spread evenly on marine plywood caul plate and cold pressed to form 10 mm × 600 mm × 2400 mm WWCBs. The WWCBs were piled horizontally with stickers in between and conditioned for 3 to 4 weeks prior to trimming.

A nest of M. losbañosensis collected from the field was maintained in a termite chamber. The nest members were allowed to establish and increase their population. Fifty pieces of conditioned WWCB measuring 12 mm × 20 mm × 60 mm and G. arborea solid wood blocks were randomly placed over a concrete foundation and exposed to termites in the termite chamber (Figure 1a). The degree of damage on the test wood blocks and board samples was evaluated after 4 months.

Drywood termites C. dudleyi were collected from infested lumber in a culture chamber and carefully transferred into an aluminum tray with wood chips using a smooth camel’s brush. Theconditioned WWCB specimens and solid wood blocks of G. arborea were individually exposed to 100 workers and 2 soldiers inside a French-square bottle (Figure 1b). The assembled experimental bottle was incubated at room temperature and the rate of damage on the board samples was observed and recorded after 6 months.

The degree of termite damage on each board sample was evaluated based on the board’s loss in volume. The resistance of WWCB to termites was classified according to the following arbitrary ratings (Table 1).

WWCBs with dimensions of 10 mm × 300 mm × 600 mm were used in the field test. Fifteen (15) each of painted and unpainted boards were installed as siding, ceiling and paneling of a pre-constructed exposure shed (Figure 2a–c). These were tested while in service, and the degree of termite damage was visually examined every 6 months for 8 years according to the above-mentioned rating scales.

M. losbañosensis extended their mud tubes from the soil in the termite chamber to the WWCB samples placed on the concrete foundation after one week of exposure. The number of tubes increased with time—a good sign of the worker termites’ foraging activities. Nibblings were observed on the board surfaces after 6 months of exposure. The mean percent damage in WWCB was 13.2% corresponding to a rating of “resistant” (Table 2). On the other hand, G. arborea wood blocks were moderately resistant, sustaining 37% termite damage. High termite activity was still noted in termite chamber even after the removal of board samples. The G. arborea wood block was moderately durable to subterranean termites, though some cases of termite attacks have been reported. The results obtained were in conformity with the G. arborea woodblocks exposed to subterranean termites under laboratory conditions [11]. The solid wood blocks were found moderately resistant to subterranean termites after 4 months of exposure.

It appears that termite resistance of WWCB depends on the raw materials’ inherent natural durability, the additional protection from soaking wood/wool particles in water, and the binding cement that conceals wood fibers.

There were no comparative quantitative data on the chemical or starch content analysis of the soaked and unsoaked G. arborea excelsior. But soaking in water is known to reduce starch content of some materials. For instance, soaking bamboo strips in water for one week decreased their starch content from 5% to 3% and prevented beetle attack [12]. Furthermore, the cement coating of the wood excelsior seemed to provide WWCBs with supplemental protection against termite attack.

Although the WCCBs were rated as resistant composites, slight damage was seen in the interior portion, and termites entered through the gaps between wood and wool particles after the final trimming of the boards. On the other hand, termites readily attacked the surface of G. arborea solid wood blocks.

Interior feeding activity was demonstrated by the discharge of pellet-like materials (fecal pellets) from the board surface in the accelerated test against C. dudleyi. About 61% of WWCB samples and 100% of G. arborea solid wood blocks manifested initial termite attack after a month of exposure. However, no new pellets were discharged after a 3-month exposure, which meant the cessation of termite feeding. The WWCB and G. arborea solid wood blocks sustained mean damage of 2.1% and 4.8%, respectively, after 6 months, (Table 2). These results suggest that although drywood termites could bore into the WWCBs, they were unable to subsequently penetrate the substrate due to insufficient food. It was roughly estimated that 50% of WWCB samples sustained incipient damage and the rest, only 5% termite attack. The solid wood blocks were attacked for a short period, but the attack did not persist due to G. arborea’s inherent durability. Based on the degree of damage, the WWCBs and G. arborea wood blocks were classified as resistant to drywood termites.

No signs of termite incidence were noted in the first 5 years of field exposure. Subterranean termites were first observed on WWCB 5½ years after field exposure (Table 3). Both N. luzonicus and M. losbañosensis extended their mud galleries from the ground line to the base of painted and unpainted WWCBs installed as exterior and interior walls. The galleries were present in approximately 50% and 33.3% of painted and unpainted exterior walls, respectively, and in 8.3% of interior walls regardless of whether painted or unpainted. Observations beneath the galleries revealed nibbling only or no termite damage on any board.

Termite incidence in WWCBs slightly decreased from 50% to 45.4% and from 33.3% to 27.3% in painted and unpainted exterior walls, respectively. Only 9.1% of painted and unpainted interior walls showed signs of termite incidence after 6 years. After 6½ years, N. luzonicus remained active at the base of painted and unpainted exterior walls, whereas M. losbañosensis became inactive on painted and unpainted interior walls.

Incidence of N. luzonicus was observed on painted and unpainted exterior walls only after 7 years, and that 40% of each board type did not show any damage. Termites completely disappeared thereafter. The painted and unpainted ceiling WWCBs had no termite damage for the duration of the field test.

The results showed that the presence of mud termite galleries did not always indicate termite attacks on WWCBs. The present study, therefore, demonstrated that WWCBs were not suitable food substrates for termites and were resistant to Philippine subterranean termites. These findings were relatively similar to the resistance of Indonesian composite boards under field conditions but the species of subterranean termites were not identified [13]. Subterranean termites were not able to cause infestation to locally produced WWDB due to G. arborea’s inherent natural durability and the cement that served as protective shield of wood excelsior.