Abstract: Background and Objective: So far, no scientific papers have reported on the termite population of subterranean termite Coptotermes curvignathus especially when they attack the wooden components in the building. This information is crucial for termite control such as system development, specifically capable of non-destructively detecting and predicting a termite population. A study was conducted to determine the termite population on the various volumes of pine boards. Materials and Methods: In this study, the exhaustive trapping technique was applied to the termite colonies. Thirty boards, infested with termites, were collected, after which the volumes were measured and dismantled carefully to determine termite population. A statistical analysis of linear regression is used to analyze the relationship between board volume (x) and termite population (y). Results: This survey showed that the board infested by C. curvignathus was dominated by termite workers (75.21±6.5101%) compared to termite soldiers (24.79±6.5101%). Based on the numerical analysis, the linear regression model (y = 40.09368+0.55031x) provides an overview of the number of termites present for a feeding test and provides an analysis of the extent to which the design of detection system will be able to predict a termite population. Conclusion: This research showed that the larger the board volume, the larger the termite population, since the larger board has an abundance of food sources containing large amounts of cellulose. This cellulose attracts termites to live inside it and also stops the process of further foraging.
INTRODUCTION
Termites play an essential role in the tropical ecosystem1. In their natural habitat, termites are considered as beneficial insects because they can decompose the dead or dying plant materials and, thus are an important part of the nutrient cycle2. However, when termites eat wooden building structures and cellulose products, they are considered as pests2-4. Drywood termites and sub-terranean termites are two major termite groups that have the most significant damage impacts worldwide. In Asian regions, Kuswanto et al.5 stated that 70% of damage to wooden buildings is caused by subterranean termites. The global damage caused by these termites was estimated at $32 billion in 20106. Indonesia itself has the endemic species of sub-terranean termites, i.e., Coptotermes curvignathus which is the most aggressive one7-12. Rilatupa et al.13 reported that this termite is capable of attacking a high-rise building's construction such as apartments and hotels in Jakarta-Indonesia up to the 33rd floor.
Termite species vary in their basic biology and ecology, including termite population, foraging territories, swarming, feeding, nesting and reproductive behavior. In terms of termite population, many researchers have investigated various natural habitats including forests14-16, urban trees17 and silvicultural plantations18. Until now, it is widely reported that the study of termite population in various regions has covered various families: Kalotermitidae (Incitermes nr. bequaerti, Neotermesmona, Glyptotermes satsumensis, Glyptotermes fuscus)14,19, Rhinotermitidae (Reticulitermes flaviceps, Reticulitermes flapives, Reticulitermes virginicus)15,19,20 and Termitidae (Nasutitermes parvonasutus, Odontermes formosanus, Pericapriter mesnitobei)19. However, to date, no scientific papers have reported on the termite population of subterranean termite Coptotermes curvignathus. Therefore, the results of this study will provide critical information regarding its population biology. In addition, according to the literature review, researchers only focus on surveying the termite population in particular areas of the tree such as the branch or trunk14,16,18. Keep in mind that besides termites attacking the tree, they also attack the wooden component on various volumes in the building. In this study, the pine boards in various volumes were used as a representation of the wooden component in the building.
This study is part of a substantial series of research, which is the system development specifically capable of non-destructively detecting and predicting the termite population. As it relates to both the scientific reports21-23 and the tools available in the market24,25, so far those are only capable of detecting the presence or absence of termites and have not been able to predict a population size. Basically, the proposed termite detection system is designed based on the board volume to be able to sense the signals generated by termites but each board volume has the different carrying capacity correspond to termite population size. Therefore, the knowledge of termite population biology in various board volumes is the key information that must be disclosed. This what extent the design of the detection system will be able to predict a termite population. Additionally, information can be used to estimate the wood damage rate and the termite control techniques.
MATERIALS AND METHODS
Study location: This survey was completed at the Termite Laboratory, Faculty of Forestry, Bogor Agricultural University, which has a facility for subterranean termite rearing. To keep the laboratory condition as a natural termite habitat, the temperature and humidity of this laboratory were arranged, i.e., 28°C and 89%, respectively. Historically, since 2012, the laboratory has successfully cultivated the subterranean termite Coptotermes curvignathus obtained from Yanlappa Experimental Forest, Bogor, Indonesia. In this laboratory, there are six concrete boxes that consist of an actively reared colony of C. curvignathus (used as a survey location). The visualization of the concrete box is given in Fig. 1. A brief description is as follows: (a) In composition, each box from bottom to top consists of the brick, tree trunk, soil, sawdust and pine boards, (b) Each box has a dimensional size of 180 (l)×120 (w)×130 (h) cm with the edge surrounded by stagnant water to avoid termites moving to another area, (c) To maintain the termite survival, every month the air-drying pine boards were served to the each box. Regarding the population size of the termite colony in each concrete box, so far, it has not been reported yet. However, this colony is always seen attacking the served pine boards aggressively. One way it that can be applied to know its population size is through mark-release-recapture technique26. However, this issue is not a concern in this study and that technique is only able to provide predictive value that may exceed the exact of the colony population size. In the end, the primary concern of this study was to survey the termite population size on various volumes of the pine boards, i.e. as many as five pine boards with signs of termite activity in each concrete box were selected as a sample (Fig. 1).
| Fig. 1: | Visualization of the concrete box for termite cultivation |
Assessment of infested boards, population size and caste composition: To estimate the population size, the exhaustive trapping technique given by Thorne et al.27 was applied to this study. This method is appropriate to apply to this investigation as the location of the study is not carried out in the natural habitat of the termite but rather in the laboratory-reared colonies. The advantage of this technique is that the number of termites that are isolated will not overestimate an actual population. Thus, it is conservative and beneficial in providing highly credible minimum estimates of termite population (an assumption only is made that termites are collected from a single food source or from food source(s) in close proximity).
To obtain credible information regarding the population size and caste composition, the termite gallery structure inside the board is dismantled carefully. Furthermore, all termites were sorted by member categories28: (1) Soldiers, (2) Worker, (3) Nymph and (4) Alate. The exact counts for each category were made. Equation 1 is used to calculate the caste composition (Ccom), where, T is member categories and Psize is a termite population in each board. In addition, the board volume (V) is defined in Eq. 2:
| (1) |
| (2) |
where, l, w and h represent the length, width and height of the board:
Statistical analysis: The core of this research is to analyze the relationship between the board volume and the termite population. To explore this, a linear regression method was performed on XLSTAT ©Addinsoft 1995-2014 (version 2014.5.03). This purpose of this method is to discuss the relationship between independent variable (x, board volume) and dependent variable (y, termite population). This method produces the linear regression model in the form Montgomery et al.29 as shown in Eq. 3:
| (3) |
where, β0 is the y intercept, β1 is the slope of the linear regression line. For assessing the overall accuracy of the model, we compute the root mean squared error (RMSE), which is the standard deviation of residuals (error prediction). The coefficient of determination (R2) is also given in this analysis, if the value of R2 is close to 1, then both variables (x and y) have very close correlation and the resulting model is more accurate. In addition, the quantile approach is used to identify deeply about the distribution of termite population on each quantile30. This approach graphed the termite population relation to the cumulative relative frequency (Crf). Based on the Crf value with the range of 0 to 1, the termite population on each quantile (Q) can be generated, i.e., the minimum (0 Crf), Q1 (0.25 Crf), Q2 (0.5 Crf), Q3(0.75 Crf) and maximum (1 Crf).
RESULTS
A study was conducted to survey the termite population within the 30 pine boards in various volumes. Based on the following parameter (i.e., board volume and termite population), the mathematics model was successfully developed by a linear regression approach to analyze the relationship of both parameters. As can be seen in Fig. 2, linear regression indicates a positive relationship between board volume and termite population in Eq. 4 as follows:
| Fig. 2: | Fitted relationship between board volume and termite population |
| (4) |
With a RMSE of 79.6045, explains the error predictions generated from the model. In addition, based on numerical analysis, the board volume and the termite population have a very close correlation, which is indicated by the R2 value of 0.9499.
In the practical framework, the population size of C. curvignathus, which attacks the wooden components in the building can be estimated with the input parameter x (board volume).
For example, if the board volume is 1000 cm3, then the population size is about 590 individuals (round to nearest decimal, the exact calculation is 590.40). After their population is known, the next step is the termite control technique, especially the spraying method31, with chemicals such as termiticide. However, the effectiveness of the termiticide dose to the termite population requires further identification. Of course, if this is practiced, it will greatly save the use of termiticide.
Figure 3 showed an overview of the distribution of termite population on each quantile according to cumulative relative frequency. It can be explained that the minimum, Q1, Q2, Q3 and maximum of the termite population are 61, 148, 250, 676 and 1187 individuals, respectively.
| Fig. 3: | Distribution of termite population on each quantile |
This survey found only the termite worker and soldier, the other members such as nymph and alate are not found. Therefore, based on the calculation of caste composition in Eq. 1, the subterranean termite C. curvignathus has the caste composition of 75.21±6.5101% for termite worker and 24.79±6.5101% for termite soldier (Fig. 4).
| Fig. 4: | Caste composition of C. curvignathus |
The caste proportions clearly show that a colony was dominated by termite workers. This, is in accordance with the statement of Kuswanto et al.5 stated that the workers constituted the largest proportion in the colony. The excessive number of soldiers will burden the colony, because they have to be fed by workers from mouth-to- mouth or anus-to-mouth, this is known as trophallaxis32. If the number of soldiers is excessive, then this might break down the social structure of a colony and could be used to control termite population.
Once the termite population has been identified through the board volume, then the range of termite caste composition can also be estimated. For example, if the termite population attacking board is 590 individuals, refer to the caste composition results in Fig. 4, then there are at least 444±38 termite workers and 146±38 termite soldiers. The predicted results of that caste composition both termite workers and soldiers are round to nearest decimal. Information on the composition of this caste is very useful for standardizing wood and wood products’ resistance test against wood destroying pest, C. curvignathus.
DISCUSSION
One of the main goals of this experiment was to survey the termite population on various volumes of the pine boards. The results showed that the larger the board volume, the larger the termite population. The selection of larger board by subterranean termites was also observed by Wang and Powell in a forest in Mississippi33, Nobre et al.18 in a managed silvicultural plantation and Manzoor et al.16 in a forest in Pakistan. Their findings supported the results of this survey that the larger board tends to be more attractive for termites to feed on. Lenz et al.34 and Togola et al.35 explained that where there is a larger board, then it has an abundance of food sources containing large amounts of cellulose. This cellulose attracts termites to live inside it and also stop the process of further foraging. In addition, according to Kamaluddin et al.36, their classic theory of the termite ecology, explains that the subterranean termite tends to prefer moist board and is mainly associated with board partially decomposed by fungi. The moisture availability is crucial to subterranean termites as they are thin-skinned insects that quickly dry up when exposed to the wind and strong sunshine, consequently, they require a constant supply of moisture18. Also, generally, the termites will follow the direction of board fibers (longitudinal) in foraging and excavating to build their gallery system37.
This study confirmed that the termite caste attacking the board consists of termite workers and soldiers only. The other members such as nymph and alate were not found. This study contradicts with previous research conducted by Manzoor et al.16 and Jones et al.14, where they have surveyed the termite populations in their natural habitat and found the members such as nymph and alate. The difference in the survey location is considered as a reason why in this study did not see those members. In addition, regarding termite ecology, both termite workers and soldiers are foraging population that have a role in finding food sources. Keep in mind that a termite colony has three castes system based on their specific roles5,38,39. The castes include: (1) The reproductive termite or sexual adults, consists of king and queen that are tasked to inseminate and to spawn respectively, (2) The termite worker, being a member of the colony that has the primary task of finding food sources, (3) The termite soldier, whose role is to protect the colony, both workers and reproductive counterparts, from their natural enemy or other predators. This provides a clear reason why when searching for food sources, the colony represents only the workers and soldiers because it appropriates their role. In addition, common subterranean termite attacks the board to forage food sources only, they do not make the board a nest location. These food is brought to their nest in the soil, where the termite workers keep a certain liquid food and serve it to the other colony members such as reproductive caste, soldier, nymph and alate. Compared with the drywood termite, the subterranean termite such as C. curvignathus has a larger population when they explore for food. Therefore, overall the drywood termite is less economically important and has a lower attack than a subterranean termite. In various regions of the world such as north America, Malaysia, China and Singapore, the same impact is seen where the sub-terranean termite in the genera of Coptotermes is the most economically important species5,40.
As explained in introduction section, the results of this survey are basic information to supplement following research, i.e., the development of termite detection system, which is able to detect and predict precisely a termite population. In the previous study reported by Nanda et al.21, they have successfully developed a termite detection system but it was only able to detect the presence or absence of termites. The main finding of this study is the linear regression model in Eq. 4, this is a crucial property which provides an overview of the number of termites for feeding test and determine the extent of the design of detection system to predict a termite population. For example, in practical application, the termite detection system is designed for sensing a termite signal on 10×10 cm board surface area and a depth of 4 cm, so its volume is 400 cm3. Based on Eq. 4, this board is capable of being inhabited by at least 260 individuals (round to nearest decimal, the exact calculation is 260.22). This value estimate is significant as a reference to the minimum or maximum boundary of the termite population to be detected. If the maximum boundary is set to 260 individuals, then the feeding test will be performed on various populations such as 20, 40, 60, ..., 260 individuals and acquire the termite signals such as acoustic emission, temperature, moisture content and so on. Through a combination of distinctive termite signal and machine learning algorithms such as discriminant analysis, support vector machine, artificial neural network and deep learning41-44, the system will able to predict precisely the termite population. Also, hopefully, the results of this investigation will attract future studies that involve sub-terranean termite in foraging and excavating behaviors as well as novel methods and techniques that will lead to more effective and efficient integrated pest management.
CONCLUSION
This research revealed the critical information regarding the termite population of C. curvignathus on various volumes of the pine boards. The results showed that the larger the board volume, the larger the termite population, since the larger board has an abundance of food sources containing large amounts of cellulose. This cellulose attracts termites to live inside it and also stops the process of further foraging. Also, this survey showed that the board infested by C. curvignathus was dominated by termite workers (75.21±6.5101%) compared to termite soldiers (24.79±6.5101%). The linear regression model (y = 40.09368+0.55031x) is an important property for developing the termite detection system, this models an overview of the number of termites for a feeding test and a perspective of what extent the design of detection system will be able to predict a termite population.
SIGNIFICANCE STATEMENT
The larger the board volume, the larger the termite population of C. curvignathus, since the larger board has an abundance of food sources containing large amounts of cellulose. The main finding of this study, i.e., the linear regression model (y =40.09368+0.55031x), will help the researchers for developing the termite detection population within board.
ACKNOWLEDGMENT
This study is a portion of a dissertation submitted by the first author in partial fulfillment of requirements for the Ph.D. degree in the Department of Agricultural Engineering and Biosystem, Bogor Agricultural University, Indonesia. This study was performed with the financial support from the Ministry of Research, Technology and Higher Education of the Republic of Indonesia (project no. 129/SP2H/PTNBH/DRPM/2018) through the accelerated master program leading to doctorate research grants (PMDSU).