Evaluation of the Natural Antitermitic Properties of

Aleurites fordii (Tung Tree) Extracts

Rachel A. Hutchins

Poplarville High School, Poplarville, MS 39470

[Rachel A. Hutchins received the Mississippi Junior Academy of Sciences Overall Award for 1996. Special thanks to Georgia Pacific Corporation's Leaf River Pulp Operation in New Augusta, Mississippi, for underwriting the publication of this research paper.]

While many commercial termiticides are available to combat the destructive termite, none are entirely natural. The risk of environmental contamination can be reduced by using natural antitermitics. Through five phases of experimentation, extracts from Aleurites fordii (tung tree) have been found to have antitermitic potentials. Effective termite control methods using tuna tree extracts included termite toxicants, repellents, attractants, bait toxicants, soil barrier treatments, and topical lumber treatments.

KEY WORDS: termiticides, Aleurites fordii, tung tree, toxicology

Termites cause millions of dollars in damages in the United States each year. This destruction is not limited to wooden structures alone. Many commercial termiticides are potential environmental contaminants and even carcinogens. For example, the most common method of subterranean termite infestation prevention involves applying a chemical termiticide to a structure's peripheral grounds. This often leads to soil and water contamination when applied improperly. One popular termiticide, chlordane, is known to cause cancer and has been banned from above ground use. Though past environmental contamination caused by termiticides is irreversible, future damage is not inevitable. Termite treatment methods must be reconsidered. The use of natural antitermitics is a potential alternative. This research examines one possible solution: extracts of the tung tree (Aleurites fordii).

The tung tree, native to China, was widely cultivated in Mississippi from 1925 until 1967, when the major orchards were destroyed by Hurricane Camille. The tung oil procured from the tung tree's fruit is marketed as a waterproofing agent for use in products such as printer's ink and wood varnishes. Resistance to various insects is an inherent property of the tung tree which, prior to this experimentation, was relatively unresearched. This study, by evaluating the tung tree for antitermitic properties, seeks to develop a natural alternative to current environmentally-detrimental termiticides.

METHODS

PHASE I. Antitermitic Bioassay of the Tung Tree--A preliminary test was performed to detect the presence of natural antitermitic properties in various parts of the tung tree. Survival of Eastern subterranean termites (Reticulitermes flavipes), when isolated and exposed to ground tung products (leaves, bark, wood, kernels, and nut hulls) was monitored over a 10 day period. Controls consisted of cellulose pads (0.5 g, 47 mm diameter) moistened with 1.0 ml portions of distilled water. Total mortality occurred in bark, wood, and kernel test groups; further bioassays were restricted to these three tung materials.

Tung wood and bark were each ground using a Wiley mill. A 300 g portion of ground wood was extracted through six cycles in a sequential soxhlet extractor with 1000 ml of acetone-hexane that was 50:50 by volume. The extraction was repeated using ground bark. Tung kernel extract was derived from ground kernels using a sonic disrupter with acetone-hexane solvent. The solvent was removed from extracted components using a rotary evaporator. From the bark extraction, two extracts were obtained and were separated into bark 1 and bark 2. Extract dilutions of 0.1%, 0.5%, 1.0%, 2.5%, and 5.0% (weight/weight) were made using acetone. To identify the specific termite behavior alterations caused by each tung tree extract, three bioassays were performed (McDaniel, 1992), each in triplicate, using externally undifferentiated workers beyond the third instar. The colonies were collected in Pearl River County, Mississippi, and were separated from wood and debris by published methods (Tamashiro et al, 1973; Su and Scheffrahn, 1986).

No-choice Test (evaluated extract toxicity in a force-feed environment): One milliliter aliquots of tung tree extracts were topically applied to 0.5 g cellulose pads. The solvent was allowed to evaporate, and the pads were then moistened with 1.0 ml of distilled water and placed into the bottom of cylindrical plastic containers 4.0 cm high by 5.3 cm in diameter. Cellulose pads that had been treated with solvent alone, air-dried, and then moistened served as controls. Twenty-five termites were then added to each container. Termite mortality rates were monitored for 25 days.

Repellency Test (assayed termite attractancy or repellency to extract): Cellulose pad halves were treated with 0.5 ml aliquots of tung extract. These treated halves were placed beside untreated pad halves into small containers. Ten termites were added to each. The locations of the termites were noted at eight time intervals: 15, 30, 45, 60, 90, 120, 180, and 240 minutes. Based on the number of termites which chose to stay on the extract-treated pad halves, each extract was designated to be an attractant or a repellent. Extract attractancy is evinced by more than 50% of the termites remaining on extract-treated pad halves, while extract repellency is shown by less than 50% of termites staying on treated pad halves.

Choice Test (examined feeding of termites when a choice between extract-treated and untreated cellulose pads was given): The choice test apparatus was constructed as shown in Figure 1. One chamber contained an untreated pad. A pad that had been dyed blue and topically treated with 1.0 ml tung tree extract was placed in the other chamber. (The blue dye serves as a feeding indicator, as it becomes visible in termites' hindguts when consumption of the treated pad has occurred.) Twenty-five termites were introduced via the treated chamber, though their movement throughout the test apparatus was not restricted. Daily throughout the 8-day test period, a visual feeding rating was assigned to the termites based on the intensity of the blue color of their hindguts.

Each bioassay was subjected to analysis of variance (ANOVA). Main effects were tested using the General Linear Model procedure where Alpha=0.5. Significant differences among means were compared using the Least Significant Difference test.

PHASE II. Wood and Kernel Extract Combinations as Termite Bait Toxicant--While the tung wood extract is a very effective toxicant, it is also a repellent. The termites' avoidance of the wood extract reduces their death rate. To solve this problem, this stage focused on combining the toxic wood extract with the attractant kernel extract to create a termite bait toxicant. Extracts from the tung wood and kernels were obtained as in Phase I, 0%, 0.1%, 1.0%, 2.5%, and 5% (w/w) concentrations were used. These dilutions of wood and kernel extracts were united to yield 25 different wood/kernel extract combinations (Table 1), which were used in two tests. A repellency test was performed for one hour, and a choice test was monitored for 31 days, with three replications of each (McDaniel, 1992).

Table 1. Wood/kernel extract combinations.

PHASE III. Field Tests--Previous laboratory findings were assessed through field testing. Southern yellow pine stakes (10.16 cm x 2.54 cm) were separately vacuum-impregnated with tung wood and kernel extract (2% weight/weight concentrations), as well as with combinations of the two (0.125%, 0.25%, 0.5%, 1%, and 2% weight/weight concentrations). Control stakes were treated with the various solvents used in extraction and dilution processes. Untreated controls were also included. Additionally, stakes cut from a tung tree were tested to determine if termites would feed on actual tung lumber. Five replications were performed. Stakes were dried, weighed, labeled, and then randomly inserted one foot apart into the shaded, moist ground of the field test site. After 5 months, stakes were removed, cleaned, dried, and reweighed. Results were analyzed using the General Linear Model procedure and the Least Significant Difference Test at Alpha=0.10.

PHASE IV. Practicable Applications--Various practical uses of the previous findings were explored. The possible use of the tung kernel extract as a termite bait was evaluated in the first experiment. To test the hypothesis that termites' attraction to kernel extract-treated pine blocks would result in increased feeding on any untreated blocks in the immediate vicinity, a choice consumption test (Carter and Smythe, 1974) was performed in a laboratory colony of 2,000 termites. The test involved placing paired blocks (0.635 cm x 1.905 cm x 1.905 cm) of Southern yellow pine into a sand medium in a large plastic container. One block of each pair was left untreated, while the other block was topically treated with kernel extract in three concentrations (0.2%, 1.0%, and 5% w/w) plus a control (0.0%). Five replications were used. After 31 days, differences in consumption of the untreated blocks were examined by comparing pre- and post-test block weights. Consumption of the treated blocks was not a major concern, as Phase I tests had shown that the kernel extract was a physical attractant only.

A soil barrier test (Amburgey and Smythe, 1977) which simulated the widely used soil treatment method of termite infestation prevention was performed in triplicate using the highly toxic tung wood extract. One hundred termites were placed at one end of a three-chambered container with a food source (an untreated pine block) at the opposite end (Fig. 2). In order to reach the food source, the termites had to tunnel through the center chamber, which contained sand to which 5 ml wood extract had been applied. Wood extract concentrations of 0.0% (control), 0.2%, 1.0%, and 5.0% (w/w) were tested, with three replications of each. Plastic barriers prevented termites from crossing over the treated sand.

The second practical test using the wood extract tested its effectiveness as a topical spray to prevent termite lumber infestation. Stacked blocks of Southern yellow pine were placed onto sand in individual containers which held 100 termites each. A wood extract-treated block (1.270 cm x 1.905 cm x 1.905 cm) was placed on the sand, and a thinner, untreated block (0.635 cm x 1.905 cm x 1.905 cm) was placed on top of the treated block (Fig. 3). After 31 days, in order to ascertain if the topical wood extract treatment of the bottom blocks prevented or reduced termite consumption of the top untreated blocks, untreated block consumption was calculated by comparing pre- and post-test block weights.

PHASE V. Extract Identification--An identification of the chemical constituents of the tung tree extracts was sought. The kernel extract closely resembles tung oil, which is derived by machine pressing the nuts of the tung tree, in odor, appearance, and consistency. To determine if these two substances are structurally similar as well, thin-layer chromatography (TLC) of both tung oil and tung kernel extract was performed using chromatogram sheets coated with silica gel. Sheets were developed using a 75% hexane/ethyl acetate solution. Resulting bands were visualized with anisaldehyde-sulfuric acid spray reagent.

RESULTS

PHASE I. Antitermitic Bioassay of the Tung Tree--After the no-choice test termite mortality responses (Fig. 4) were analyzed, the strongest termite toxicant overall was found to be the wood extract. Total termite mortality occurred in all replications after only 48 hours exposure. While the greater concentrations of the kernel and bark 2 extracts also caused high termite mortality, this toxicity occurred over a ten-day period. The bark 1 extract was a mild toxicant at the 0.5% and 2.5% concentrations. The weaker dilutions produced termite mortality rates of less than 50% and thus were only mildly toxic.

To assess potential effects of vapor-phase toxicant in the wood extract, as suggested by the rapid toxicity of the wood extract in the no-choice test, a ten-day fumigant test was performed. Termites were enclosed with the wood extract but were restricted from touching it. Few deaths occurred. Thus, no apparent fumigant effects were generated by the wood extract. Contact or stomadeal poisoning probably occurred.

Comparative locations of termites in the repellency tests (Fig. 5) designated the wood and bark extracts as repellents since most termites tended to avoid these extract-treated pad halves. In contrast, the kernel extract was a strong attractant. An average of 95% of the termites tested consistently preferred kernel extract-treated pad halves over untreated pad halves. This attractancy propensity of the kernel extract could render it extremely valuable as a termite bait.

According to visual feeding ratings given to termites in the choice tests (Fig. 6), no extract-treated pads received greater feeding than the controls. In the kernel extract choice test, termites consistently remained in the treated chamber throughout the entire test, thereby confirming repellency test results of the extract being an attractant. While the kernel extract acquired the most notable consumption among the extract-treated pads, this amount did not surpass that of the controls. Therefore, the kernel is a physical attractant only. It does not augment the amount of feeding on the treated substrate.

PHASE II. Wood and Kernel Extract Combinations as Termite Bait Toxicant--In the repellency test, the attractancy of the wood/kernel combinations varied inversely with the concentration of the wood extract component. As the concentration of the wood extract component increased, the termite attractancy response to the combination decreased (Fig. 7). Therefore, the stronger concentrations of the wood extract negated the kernel extract's attractancy. As implied by these results, in order for a wood/kernel extract combination to be an effective bait, the wood extract component must have a concentration of 1% or less, otherwise, attractancy will not prevail.

In the choice tests, as the concentration of the wood extract component of the combinations increased, the termite mortality rates consistently increased while feeding decreased. Based on choice test and repellency test results, the most effective bait toxicant was chosen. Because a minimally-consumed bait toxicant will last much longer than one on which termites feed heavily, the general criterion for an efficient short-term bait toxicant requires high termite mortality rates with only slight feeding on the treated substrate. The wood/kernel extract combination which best fit this criterion was the 1% wood/1% kernel combination. Figure 8 shows the feeding and mortality of the 1% kernel extract, the 1% wood extract, and the 1% wood/kernel combination. The combination is remarkably more effective than the individual extracts. The kernel extract boosts the toxicity of the wood extract by 51%. This extensive increase of termite mortality occurs as a result of minimal feeding on the wood/kernel combination-treated substrate, thereby fulfilling the bait toxicant criterion.

As a toxic termite short-term bait, 1% wood/kernel extract combinations could effectively eliminate termite colonies without applying potentially-harmful chemicals to the soil.

PHASE III. Field Test--Stakes with the greatest termite feeding as determined by comparing pre- and post-test stake weights were the controls (Fig. 9). Neither the wood nor kernel extract treated stakes received greater feeding than the controls, thereby reaffirming Phase I choice test lab results in which feeding on extract-treated cellulose pads did not exceed that of untreated pads. The 1% wood/1% kernel combination stakes lacked damage. This further confirmed previous findings (Phase II choice test). The slight damage to the tung tree wood stakes indicated an obvious termite repulsion due to the presence of antitermitic properties which have been previously assessed in this study. Collective field test results correlated well with laboratory assessments.

PHASE IV. Practicable Applications--The results of the kernel extract feeding attractancy test, as verified by statistical analysis, confirmed the hypothesis that the kernel extract would increase feeding on untreated neighboring blocks. Termites consumed 10.9 times more wood in the blocks paired with 1% kernel extract-treated blocks than in those paired with control (solvent-treated) blocks (Fig. 10). In contrast, termites' feeding on the extract-treated blocks was not significant at Alpha=0.05 (Fig. 11). Additionally, tunneling activity was visually analyzed and was found via statistics at Alpha=0.05 to be more prevalent near the 1% kernel/untreated pairs than near the control/untreated pairs (Fig. 12). Thus, the 1% kernel extract-treated blocks effectively attracted termites to feed on the neighboring untreated blocks.

These results indicate that a concentration of 1% kernel extract, when used as part of a bait station, but not as the food source, can cause a significant increase in termite consumption of the food source. This effect is desirable with long-term, slow elimination bait toxicant.

In the wood extract soil barrier test, no termites in any of the 1% or 5% wood extract containers ever reached the food source. They only briefly entered the treated barrier. All of these termites were dead within 10 days. Meanwhile, all termites in the control, which had a solvent-treated sand barrier, reached the food source within the first 6 hours of the test and, therefore, survived.

In the wood extract topical treatment test, consumption of the control blocks was significantly higher (at Alpha=0.05) than that of the untreated blocks placed on top of the 1% and 5% wood-extract treated blocks (Fig. 13). The wood extract was efficacious in reducing termite infestation of the upper untreated blocks. This finding suggests that the tung wood extract could be used as a topical lumber treatment for termite infestation reduction.

PHASE V. Extract Identification--When bands resulting from TLC of tung oil and tung kernel extract were compared, only slight variations were noted. Therefore, these two substances are closely related. Future bioassays will determine if tung oil has the same antitermitic effects as the kernel extract. Identification of the wood extract's components is in rudimentary stages. Further studies aim to isolate the toxic compounds.

DISCUSSION

Although the tung tree has been a versatile plant for hundreds of years, its antisemitic potentials were largely unrealized before this research was conducted. The tung tree's list of many uses could now be extended to include natural antitermitics. This study has discussed numerous practical applications of tung tree extracts: termite repellents, toxicants, attractants, bait toxicant, soil barrier treatments, and topical wood treatments. Each of these termite elimination methods aims to eliminate current environmentally-deleterious termite suppression practices. A natural antisemitic developed from the tung tree could prove to be not only environmentally benign, but economically valuable as well. The tung industry is currently rebuilding in the United States. However, increased competition from other tung oil-producing countries is suppressing this effort. A new use for the tung tree, such as natural antitermitics, developed and marketed in the USA would be an incentive for our tung tree growers to augment their orchards.

Future plans of study include isolating specific antisemitic compounds of the tung tree extracts and further examining the extracts' effects of the physiology of termites. This study's ultimate goal is to initiate the widespread use of tung tree extracts or related derivatives as natural, environmentally-safe termiticides. It is hoped that termite suppression using natural products will proliferate as the focus of research aimed at finding safer, yet effective and efficient compounds to control termites.

ACKNOWLEDGMENTS

Thanks to Dr. Clarence McDaniel, formerly of the USDA Forest Service Southern Forest Research Experiment Station in Gulfport, MS; Dr. Terry Amburgey and Susie Parikh of the Mississippi State University Forest Products Lab; Dr. Julie Michel of the University of Mississippi School of Pharmacy's National Center for the Production of Natural Products; Dr. Clarence Watson of the Mississippi State University Experimental Statistics Lab; and Mrs. Angela Jones of Poplarville High School for assistance in this study.

LITERATURE CITED

Amburgey, T.L., and R.V. Smythe. 1977. Factors influencing termite feeding on brown-rotted wood. Sociobiology 3:3­12.

Carter, F.L., and R.V. Smythe. 1974. Feeding and survival responses of Reticulitermes flavipes (Kollar) to extractives of wood from 11 coniferous genera. Holzforschung 28:41­45.

McDaniel, C.A. 1992. Major antisemitic components of the heartwood of southern catalpa. Journal of Chemical Ecology 18:359­369.

Su, N.-Y., and R.H. Scheffrahn. 1986. A method to access, trap, and monitor field populations of the Formosan termite (Isoptera: Rhinotermitidae) in the urban environment. Sociobiology 12:299­304.

Tamashiro, M., J.K. Fujii, and P.-Y. Lai. 1973. A simple method to observe, trap, and prepare large numbers of subterranean termites for laboratory and field experiments. Environmental Entomology 2:721­722.