Open access journal: http://periodicos.uefs.br/ojs/index.php/sociobiology ISSN: 0361-6525 Sociobiology 60(1): 69-76 (2013) Utility of Acoustical Detection of Coptotermes formosanus (Isoptera: Rhinotermitidae) W. OSBRINK,1 M. CORNELIUS2 Introduction Total economic loss due to termites in the United States has been estimated at $11 billion per year, and where they occur, the Formosan termite, Coptotermes formosanus Shiraki (FST), is the most devastating termite pest in the world (Su, 2002). The FST was introduced into the United States from Asia when troops and equipment returned from World War II (Su & Tamashiro, 1987). In addition to struc- tural infestations, C. formosanus infestations of living trees are common (Osbrink et al., 1999; Osbrink & Lax, 2002; Ring et al., 2002; Osbrink & Lax, 2003). Development of techniques for detecting hidden ter- mite infestations have produced only a few successful al- ternatives to traditional visual inspection methods (Lewis, 1997). Efficient non-invasive detection of termite activity can provide timely location of an infestation thereby reduc- ing economic impact. Non-invasive detection is also ideal for evaluating the efficacy of control efforts because non- invasive monitoring has no effect on population dynamics. Abstract The AED 2000 and 2010 are extremely sensitive listening devices which can effectively detect and monitor termite activity through a wave guide (e.g. bolt) both qualitatively and quantitatively. Experiments conducted with one to ten thousand termites from differing colonies infesting wood in buckets demonstrated that acoustical emission de- tector readings significantly increased when number of termites increased. Termites were also detected in infested trees with the installation of several wave guides into their trunks. These devices can detect termites and changes in termite activity caused by changes in termite numbers, making it an effective pest management professional and research tool for finding and evaluating termite infestations and efficacy of treat- ments in specific locations. Sociobiology An international journal on social insects 1 - USDA-ARS-SPA, Kerrville, Texas, USA. 2 - USDA-ARS-BARC, Beltsville, Maryland, USA. Research article - Termites Article History Edited by: Evandro N. Silva, UEFS - Brazil Received 29 November 2012 Initial acceptance 02 January 2013 Final acceptance 14 February 2013 Keywords AED 2000, AED 2010, Formosan subterranean termite, tree, monitoring, pest control Corresponding author Weste Osbrink USDA-ARS-SPA Knipling-Bushland U.S. Livestock Insects Research Lab 2700 Fredericksburg Road, Kerrville, Texas 78028 - Phone: 504.330.41326 E-Mail: weste.osbrink@ars.usda.gov Conversely, invasive monitoring techniques can drive ter- mites away from the monitor, creating an artifact of appar- ent control because of relocation of the termites (Aluko & Husseneder, 2007). Alternatives to visual inspection include monitoring devices with sensors that detect acoustic emis- sions of termites in wood (Fujii et al., 1990; Lewis & Le- master, 1991; Noguchi et al., 1991; Robbins et al., 1991). Acoustic emission sensors are successful because they are nondestructive and operate at high frequencies (ca. 40 kHz) where there is negligible background noise to interfere with detection and interpretation of insect sounds (Lewis & Le- master, 1991; Robbins et al., 1991). Acoustic emission sys- tems have been applied as research tools to estimate termite population levels (Fujii et al., 1990, Lewis & Lemaster, 1991; Scheffrahn et al., 1993; Osbrink et al., 2011). Acous- tic emission systems are also ideal for detection of termites in trees (Osbrink et al., 1999; Kramer, 2001; Mankin et al., 2002; Osbrink et al., 2011). Understanding the efficacy and dynamics of acousti- cal detection is critical to it being successfully integrated W. Osbrink, M. Cornelius - Acoustical Detection of Coptotermes formosanus70 into an effective pest management strategy. The central objective of this research was to determine the efficacy of using the AED 2000 acoustical emissions detector (Acous- tical Emissions Consulting, Inc Fair Oaks, CA) to detect and quantify termite infestations. To meet this objective, studies were conducted to monitor C. formosanus through acoustical emission detection both in the laboratory and in trees outdoors. These studies provide evidence that AED significant potential for application in termite management efforts. Materials and Methods Acoustical Emission Detector (AED) An AED 2000 acoustical emissions detector (Acous- tical Emissions Consulting, Inc Fair Oaks, CA) was used to quantify termite activity. Lag bolt wave guides (76.2 or 150 x 9 mm) were screwed horizontally into pre-drilled pilot holes in wood substrates. Acoustical emissions were detect- ed with a sensor probe (Model SP-1L with Model DMH-30 high force magnetic accessory attachment, Acoustic Emis- sion Consulting, Inc). AED counts were acquired for 60 s with accompanying software, which converts termite sounds to counts per second and enters them into Excel (Microsoft, Redmond WA). Only the numbers of counts in the first 10 s of the 60 s recording were used to represent each unique in- dividual recording. If the first 10 s of recording was contam- inated with interference noise (elevated spiked counts), the first 10 s of recording following the cessation of interference noise were used to represent the unique individual record- ing. Comparisons also were made between the AED 2000 and the more recently manufactured AED 2010 (Acoustical Emissions Consulting, Inc Fair Oaks, CA). Laboratory Bucket Tests A vertically oriented section of spruce (Picea sp.) 38 x 89 mm (2x4 inch) dimensional lumber 17 cm in length was attached to the inner side of a lidded plastic buckets (3.79 l, 18.5 cm height x 20 cm diam.) with 2 horizontally applied drywall screws (high and low) and central lag bolt wave guide (76.2 x 9 mm). The head of the lag bolt was accessible from the outside of the lidded bucket (Fig. 1). The bucket was filled to within 4 cm of top with a moist (≈ 20% water wt/wt) mixture of sand and vermiculite (50:50 by volume). Ten holes were created in the sand-vermiculite substrate with a 5 ml pipette to increase surface area and accelerate acclimatization of termites. Four buckets were prepared for each of 4 termite densities (0, 1,000, 5,000, and 10,000 ter- mites per bucket), in which each density level represents 4 distinct colonies (A, B, C, and D), one colony per bucket. In laboratory bucket tests there were 4 replicates (bucket A, B, C, D), each replicate consisting of a 10 s recording from a specific bucket. Termites were obtained from bucket trap monitors (Su & Scheffrahn, 1986) and termite numbers de- termined by weight. Soldier proportions were about 10%, unchanged from when collected. Termite Density Response at 7 and 14 d. Formosan termites were placed in buckets on day 0 (0 d) as described above. Buckets were held in the labora- tory (≈ 26.7º C). On 7 d and 14 d AED readings were taken from each bucket. In laboratory bucket tests there were 4 replicates (bucket A, B, C, D), each replicate consisting of a 10 s recording from a specific bucket. Termite Density Response at Three Temperatures After completion of readings at 14 d for dose re- sponse at 1 temperature, buckets were placed in 3 incubators stabilized at 15, 20, and 25º C, respectively, and evaluated according to the schedule indicated in Table 1. After read- ings, buckets were rotated to a new temperature (incubator) and allowed 24 h to acclimate before acoustic readings were again taken. Incubators space limitations required the D samples to be split to fit the 12 buckets into three incubators. In laboratory bucket tests there were 4 replicates (bucket A, B, C, D), each replicate consisting of a 10 s recording from a specific bucket. Disturbance Test AED recordings were taken before and after the ap- plication of three sharp strikes with a screwdriver to the high density laboratory buckets. In laboratory bucket tests there were 4 replicates (bucket A, B, C, D), each replicate consist- ing of a 10 s recording from a specific bucket. Field Test on Trees Nine wave guides in the form of lag bolts (150 x 9 mm) were screwed horizontally into pre-drilled pilot holes in the trunk of test trees facing north, east, south, and west (Fig 1). Four wave guides were installed at ground level, four at 20 cm above ground level, and one into the east side of the trunk at a height of approximately 122 cm from the ground. Test trees consisted of four southern live oak trees (Quercus virginiana Philip Miller) with a diameter at breast height (dbh) of ≈ 90 cm, adjacent to Su-bucket-trap-monitors active with C. formosanus (Su & Scheffrahn 1986) located on the City Park campus of the Southern Regional Research Center, New Orleans, LA. In field tests on trees, only the numbers of counts in the first 10 s of the 60 s recording were used to represent each unique individual recording. If the first 10 s of recording was contaminated with interference noise (elevated spiked counts), the first 10 s of recording Sociobiology 60(1): 69-76 (2013) 71 following the cessation of interference noise were used to represent the unique individual recording. In field tests on trees, ten consecutive counts (10 s) were used to calculate mean (± SE) counts per second to quantify termite activity associated with each unique AED tree bolt attachment. Comparison of AED 2000 with AED 2010. Eight different recordings from trees were conducted with each model of acoustical emissions detector and results were compared between the AED 2000 and the AED 2010. In field tests on trees, only the numbers of counts in the first 10 s of the 60 s recording were used to represent each unique individual recording. If the first 10 s of recording was con- taminated with interference noise (elevated spiked counts), the first 10 s of recording following the cessation of inter- ference noise were used to represent the unique individual recording. In field tests on trees, ten consecutive counts (10 s) were used to calculate mean (± SE) counts per second to quantify termite activity associated with each unique AED tree bolt attachment. Data Analysis. Ten consecutive count values (10 s) were used to represent termite activity associated with each unique AED attachment. In laboratory bucket tests there were 4 repli- cates (bucket A, B, C, D), each replicate consisting of a 10 s recording from a specific bucket. In field tests on trees, ten consecutive counts (10 s) were used to calculate mean (± SE) counts per second to quantify termite activity associated with each unique AED tree bolt attachment. Acoustical data were analyzed using one way analysis of variance (ANOVA) with means separated using the protected Tukey test, P < 0.05 (Systat, 2008). Results Termite Density Response at 7 and 14 d. Buckets with no termites produced AED readings of zero (control). There were highly significant differences in termite activity between termite colonies, and acoustical emission activity increased concomitantly with increased termite density (Table 2). The highest density always had significantly greater activity than the lowest. Overall, there was no consistent change in termite acuity between 7 d and 14 d, however at the lowest density there was non-significant but numerically consistent increase in activity (Table 2). Termite Density Response at Three Temperatures. Buckets with no termites produced AED readings of zero (control). At low termite density, there was no signifi- cant difference in inter-colony activity at all 3 temp (Table 3), but there was a significant increase in termite activity at the highest density with two colonies and combined colonies (Table 4). At 20 and 25º C there was always a significant activity dose response except with colony D which did not statistically but did numerically separate 5k from 10k (Table 3). Combined colonies demonstrated highly significant den- sity dose response at all temps. At the highest density there was always significantly less termite activity at the lowest temp (Table 4). At lower density this temperature separation was not as clearly defined. Disturbance Test Three of the four colonies displayed a significant de- crease in termite activity, and one colony had a numerical but non-significant increase in recorded activity (Table 5). Qualitatively, termite activity could be heard though the ear- phones to increase for a brief time before the recording began. Field Test on Trees Out of the nine bolts per tree, generally only one or two had significantly high termite activity, with the remain- der of the bolts displaying low termite activity (Table 6). Comparison of AED 2000 with AED 2010 Of eight different recordings of trees, there was little difference observed between the AED 2000 and AED 2010. The AED 2010 had consistently higher readings that may indicate that it may be slightly more sensitive (Table 7). Table 1. Incubator temperature and rotation of termite densities. Temp (º C) Colony # Termites 10 15 20 25 0 1000 A 5000 Day1 Day 2 Day 3 10000 0 1000 B 5000 Day 3 Day 1 Day 2 10000 0 1000 C 5000 Day 2 Day 3 Day 1 10000 0 D Day 1 Day 2 Day 3 1000 5000 D Day 3 Day1 Day 2 10000 W. Osbrink, M. Cornelius - Acoustical Detection of Coptotermes formosanus72 Discussion Termite Density Response at 7 and 14 d. Having highly significant differences in termite ac- tivity between colonies is consistent with the generally ac- cepted understanding that there can be profound inter-colo- ny differences. These findings support the suggestion of Su and La Fage (1984) to use multiple colonies when conduct- ing bioassays. A possible explanation of the non-significant but numerically consistent increase in activity at the lowest density is that it takes longer for fewer termites to create a gallery system in the wood. Increased size of galleries in- creases the surface area occupied by termites creating an op- portunity for increased generation of acoustical emission. Termite Density Response at Three Temperatures. Dose responses to density and temperature were demonstrated most clearly with the combined colony data due to the increased number of samples. These results dem- onstrate the efficacy of using an acoustical emission detector to detect and monitor termite activity. Because there were significant differences in the AED readings based on termite density, the detector can be useful not only in detecting the presence of termites but also in estimating population den- sity in infested trees or structures. Disturbance Test. Though a post-disturbance decrease in activity oc- curred, a substantial amount of termite activity remained (Table 5). Qualitatively, earphone monitoring indicated an immediate, brief increase in termite sounds in all instances that is consistent with absconding. Unpublished video has shown FST to cease feeding and abscond following a distur- bance, and soldiers (incapable of chewing wood) produce characteristic termite sounds monitored with the AED 2000 (WO personal observation). Additionally, the presence of red imported fire ant colonies, Solenopsis invicta Buren, at tree study sites have produced sounds similar to termites (WO personal observation). Thus, results indicate that AED 2000 recordings are created by termite movement and not feeding activity, possibly a result tarsal claw-substrate inter- action. This is inconsistent with reports of Scheffrahn et al. (1993) and Fujii et al. (1990) who attribute signals detected by their devices specifically to termite feeding. This differ- ence in interpretation of results may reflect differences in the nature of the disturbance or in the specifics of the detection mechanism. Field Test on Trees. Of the nine bolts per tree, generally only one or two transmitted high termite activity while the remainder of the bolts displayed low termite activity (Table 6). Energy atten- uates much more rapidly horizontally across the trunk than vertically up and down the trunk (Mankin et al., 2002), sug- gesting that termite activity is oriented vertical to the bolt. Thus, a single bolt, or readings from a single point cannot determine that a tree is not infested with termites. Limitations of AED 2000 and AED 2010. Certain events can interfere with successful recording of termite activity including wind noise, trucks with squeak- ing breaks, generators, crowd noise, etc. Wind speeds > 14 km/h interfere with recording activity in trees because of leaf flutter, and Excel recordings do not distinguish termite events from unrelated sound events, therefore maintaining a log with qualitative notes is advised. Elevated wind can be a common cause for cancellation of field tests, and demands flexibility in scheduling. Radio interference can also be- come an issue that may be mitigated by incorporating ferrite chokes and the shortest cord possible. In conclusion, the AED 2000 and 2010 are extremely sensitive devises which can detect termite activity channeled through a wave guide. Because of the significant increases in AED readings with increasing group size, a trained pest management professional would be able to use the acousti- cal detectors to estimate the severity of an infestation, in addition to merely determining the presence or absence of termites. Use of this technology may be quite valuable in specific applications such as pre- and post-treatment evalu- ations of termite activity. In applications where multiple lo- cations are to be evaluated, interference from external noise can become an issue. Acknowledgements We thank R. Davey, K. Lohmeyer, J.M. Pound, and S. Skoda for agreeing to review this manuscript and valuable improvements contributed by their reviews. References Aluko, G. & Husseneder, C. (2007) Colony dynamics of the Formosan subterranean termite in a frequently disturbed ur- ban landscape. J. Econ. Entomol., 100: 1037-1046. doi: 10.1603/0022-0493(2007)100[1037:CDOTFS]2.0.CO;2 Fujii, Y., Noguchi, M., Imamura, Y. & Tokoro, M. (1990) Using acoustic emission monitoring to detect termite activity in wood. For. Prod. J., 40: 34-36. Kramer, R. (2001) Detector for termites in soil? Pest Control Tech., 29: 130-131. Lewis, V. R. (1997) Alternative control strategies for ter- mites. J. Agric. Entomol., 14: 291-307. Sociobiology 60(1): 69-76 (2013) 73 Lewis, V. R. & Lemaster, R. (1991) The potential of using acoustical emission to detect termites within wood. In: M. I. Haverty & W. W. Wilcox (Eds.), Proceedings of the sym- posium on current research on wood-destroying organisms and future prospects for protecting wood in use (pp. 34-37). Washington, DC: USDA For. Serv. Gen. Tech. Rep. PSW- 128. Mankin, R. W., Osbrink, W, Oi, F. & Anderson, J. (2002) Acoustic detection of termite infestations in urban trees. J. Econ. Entomol., 95: 981-988. doi: 10.1603/0022-0493- 95.5.981 Noguchi, M., Fujii, Y., Owada, M., Imamura, Y., Tokoro, M. & Tooya, R. (1991) AE monitoring to detect termite attack on wood of commercial dimension and posts. For. Prod. J., 41: 32-36. Osbrink, W. & Lax, A. (2002) Termite (Isoptera) gallery char- acterization in living trees using digital resistograph tech- nology. In: W. C. Jones, J. Zhai, & W. H. Robinson (Eds.), Proceedings, 4th International Conference on Urban Pests (pp. 251-257). Charleston, SC, USA. Pocahontas Press, Inc. Blacksburg, Virginia, U.S.A. Osbrink, W. & Lax, A. R. (2003) Effect of imidacloprid tree treatments on the occurrence of Formosan subterranean ter- mites, Coptotermes formosanus Shiraki (Isoptera: Rhinoter- mitidae). J. Econ. Entomol., 96: 117-125. doi: 10.1603/0022- 0493-96.1.117 Osbrink, W., Woodson, W. & Lax, A. R. (1999) Population of Formosan subterranean termite, Coptotermes formosanus (Isoptera: Rhinotermitidae), established in living urban trees in New Orleans, Louisiana, U. S. A., pp. 341-345. In: W. H. Robinson, F. Rettich, & G. W. Rambo (Eds.), Proceedings, 3rd International Conference on Urban Pests (pp. 341-345). Prague, Czech Republic. Graficke zavody Hronov, Czech Republic. Osbrink, W., Cornelius, M. & Lax, A. (2011) Area wide field study on effects of three chitin synthesis inhibitor baits on populations of Coptotermes formosanus and Reticulitermes flavipes (Isoptera: Rhinotermitidae). J. Econ. Entomol., 104: 1009-1017. doi: 10.1603/EC10217 Ring, D., Henderson, G. and McCown, C. (2002) Evaluation of the Louisiana state program to treat trees infested with Formosan subterranean termites (Isoptera: Rhinotermitidae) in Louisiana, pp. 259-266. In: S. C. Jones, J. Zhai, & W. H. Robertson (Eds.), Proceedings of the 4th International Con- gress on Urban Pests (PP. 259-266). Blacksburg, VA: Poca- hontas Press. Robbins, W. P., Mueller R., Schaal, T. & Ebeling, T. (1991) Characteristics of acoustic emission signals generated by termite activity in wood. In: Proceedings, IEEE Ultrasonics Symposium (pp. 1047-1051). Orlando, FL. Conference Pub- lications. Scheffrahn, R., Robbins, W., Busey, P., Su, N.-Y. & Mueller, R. (1993) Evaluation of a novel, had-held, acoustic emissions detector to monitor termites (Isoptera: Kalotermitidae, Rhi- notermitidae) in wood. J. Econ. Entomol., 86: 1720-1729. Su, N.-Y. (2002) Novel technologies for subterranean termite control. Sociobiology, 40: 95-101. Su, N.-Y. & La Fage, J. (1984) Differences in survival and feeding activity among colonies of the Formosan subterra- nean termite (Isoptera: Rhinotermitidae). Z. Angew. Ento- mol., 94: 134-138. Su, N.-Y. & Scheffrahn, R. (1986) A method to access, trap, and monitor field populations of the Formosan subterranean termite (Isoptera: Rhinotermitidae) in the urban environment. Sociobiology, 12: 299-304. Su, N.-Y. & Tamashiro, M. (1987) An overview of the For- mosan subterranean termite (Isoptera: Rhinotermitidae) in the world. In: M. Tamashiro & N.-Y. Su (Eds.), Biology and Figure 1. AED 2000 attached to bucket and tree. W. Osbrink, M. Cornelius - Acoustical Detection of Coptotermes formosanus74 control of the Formosan subterranean termite. Hawaii Insti- tute of Tropical Agriculture and Human Resources Research Extension Series 083 (pp. 3-15). Honolulu, HI: University of Hawaii and Manoa. Systat Software. (2008) SigmaPlot users guide: statistics, version 11. Systat Software, Inc. San Jose, CA. Table 2. Acoustical emission dose response (mean ± SE) by termites (10 s). Number termites (x1,000) 7 d 14 d _____________________________________ _______________________________________ Colony 1 5 10 1 5 10 A 23.8 ± 4.4cB 189.9 ±25.5 aA 235.6 ± 12.3aB 65.8 ± 1.3cA 105.0 ± 8.9bC 129.5 ± 12.4bB F = 34.703 df = 5, 59 P < 0.001 B 13.3 ± 2.7cBC 96.0 ± 11.9bB 146.8 ± 29.9bC 24.0 ± 3.0cB 152.7 ± 14.2abA 210.8 ± 14.7aC F = 26.757 df = 5, 59 P < 0.001 C 6.5 ± 1.9dC 39.1 ± 7.4cdB 124.0 ± 16.2bC 19.7 ± 2.6dB 56.9 ± 7.3cB 176.0 ± 16.4aBC F = 40.423 df = 5, 59 P < 0.001 D 46.4 ± 4.1dA 166.0 ± 9.5cA 744.3 ± 23.4aA 60.7 ± 6.7dA 149.7 ± 16.0cAC 320.2 ± 29.1bA F = 229.179 df = 5, 59 P < 0.001 F=25.790 F=20.158 F=195.680 F=36.223 F=13.753 F=17.787 df=3.39 df=3.39 df=3.39 df=3.39 df=3.39 df=3.39 P<0.001 P<0.001 P<0.001 P<0.001 P<0.001 P<0.001 Means within a row (lower case) or column (upper case.) with same letter are not significantly different, protected Tukey Test (P > 0.05). Table 3. Varied temperature with AED dose response (mean ± SE) by termites (10 s) . Colony Temp. # Termites (º C) (x1,000) A B C D Combined 15 1000 1.5 ± 0.5aB 3.2 ± 0.7aB 2.9 ± 0.8aB 2.9 ± 0.9aB 2.6 ± 0.4aC F = 0.808 df = 4, 79 P = 0.524 5000 17.4 ± 2.8abAC 28.0 ± 4.6aAC 2.7 ± 0.6cB 10.4 ± 2.7bcB 14.6 ± 2.1bB F = 6.463 df = 4, 79 P < 0.001 10000 15.5 ± 3.4bC 18.4 ± 2.3bC 34.3 ± 6.0abA 57.3 ± 6.0aA 31.3 ± 3.5bA F = 7.678 df = 4, 79 P < 0.001 F = 11.7 F = 17.418 F = 26.681 F = 59.908 F = 37.647 df = 2, 29 df = 2, 29 df = 2, 29 df = 2, 29 df = 2, 119 P < 0.001 P < 0.001 P < 0.001 P < 0.001 P < 0.001 20 1000 3.0 ± 1.2aC 2.1 ± 0.7aC 1.1 ± 0.4aC 2.2 ± 0.4aB 2.1 ± 0.4aC F = 0.767 df = 4, 79 P = 0.550 5000 25.2 ± 6.1abB 28.1 ± 5.1abB 23.9 ± 4.1bB 51.1 ± 9.4aA 32.1 ± 3.6abB F = 2.623 df = 4, 79 P = 0.041 10000 45.3 ± 5.4aA 57.8 ± 6.2aA 54.1 ± 4.9aA 63.9 ± 7.6aA 55.3 ± 3.1aA F = 1.204 df = 4, 79 P = 0.316 F = 19.643 F = 36.039 F = 57.422 F = 21.872 F = 93.620 df=2.29 df=2.29 df=2.29 df=2.29 df=2.29 P < 0.001 P < 0.001 P < 0.001 P < 0.001 P < 0.001 25 1000 17.0 ± 5.2aC 2.4 ± 0.6aC 14.1 ± 3.9aC 5.8 ± 1.7aC 13.5 ± 2.2aC F = 2.751 df = 4, 79 P = 0.034 5000 53.4 ± 4.6aB 42.7 ± 5.6abB 30.9 ± 4.3bB 32.3 ± 4.5bB 42.5 ± 2.8abB F = 3.195 df = 4, 79 P = 0.018 10000 159.3 ± 12.9aA 65.9 ± 6.2bcA 69.7 ± 4.9bcA 46.3 ± 3.3cA 108.7 ± 9.4bA F = 10.226 df = 4, 79 P < 0.001 F = 76.815 F = 44.548 F = 42.584 F = 37.628 F = 70.673 df = 2, 29 df = 2, 29 df = 2, 29 df = 2, 29 df = 2, 29 P < 0.001 P < 0.001 P < 0.001 P < 0.001 P < 0.001 Means within a row (lower case) or column (upper case) with same letter are not significantly different, protected Tukey Test (P > 0.05). Sociobiology 60(1): 69-76 (2013) 75 Table 4. Temperature and termite density; mean (± SE) number of AED counts (10 s). Colony Temp. Number termites ________________________________________________________________ (º C) (x1,000) A B C D Combined 15 10 15.5 ± 3.3c 18.4 ± 2.3b 34.3 ± 6.0b 57.3 ± 6.0a 31.4 ± 3.5c 20 10 45.3 ± 5.4b 57.8 ± 6.2a 57.8 ± 5.1a 63.9 ± 7.6a 55.3 ± 3.1b 25 10 159.3 ± 12.9a 65.9 ± 6.2a 69.7 ± 4.9a 46.3 ± 3.3a 108.7 ± 9.4a F = 83.805 F = 23.678 F = 26.681 F = 2.290 F = 42.547 df = 2, 29 df = 2, 29 df = 2, 29 df = 2, 29 df = 2, 119 P < 0.001 P < 0.001 P < 0.001 P = 0.121 P < 0.001 15 5 17.4 ± 2.8a 28.0 ± 4.6a 2.7 ± 0.6b 10.4 ± 2.7b 14.6 ± 2.1c 20 5 3.0 ± 1.2b 28.1 ± 5.1a 23.9 ± 4.1a 51.1 ± 9.4a 32.1± 3.6b 25 5 17.0 ± 5.2a 42.7 ± 5.6a 30.9 ± 4.3a 32.3 ± 4.5a 42.5 ± 2.8a F = 5.566 F = 2.754 F = 18.182 F = 10.796 F = 23.928 df = 2, 29 df = 2, 29 df = 2, 29 df = 2, 29 df = 2, 119 P = 0.009 P = 0.082 P < 0.001 P < 0.001 P < 0.001 15 1 1.5 ± 0.5b 3.2 ± 0.7a 2.9 ± 0.9b 2.9 ± 0.9a 2.6 ± 0.4b 20 1 3.0 ± 1.2b 2.1 ± 0.7a 1.1 ± 0.4b 2.2 ± 0.4a 2.1 ± 0.4b 25 1 17.0 ± 5.2a 2.4 ± 0.6a 14.1 ± 3.9a 5.8 ± 1.7a 13.5 ± 2.2a F = 7.659 F = 0.699 F = 9.171 F = 2.859 F = 24.819 df = 2, 29 df = 2, 29 df = 2, 29 df = 2, 29 df = 2, 119 P = 0.002 P = 0.506 P < 0.001 P = 0.075 P < 0.001 Means within a row with same letter not significantly different, protected Tukey Test (P > 0.05). Table 5. Pre - and post - disturbance mean (± SE) number AED counts (10 s) of termites. Colony Termite disturbance ___________________ Pre- Post- A 126.0 ± 9.5a 222.8 ± 109.5a F = 0.776 df = 1, 19 P = 0.390 B 141.0 ± 8.3a 70.6 ± 10.5b F = 27.742 df = 1, 19 P < 0.001 C 255.3 ± 17.4a 180.8 ± 27.9b F = 5.137 df = 1, 19 P = 0.036 D 272.5 ± 19.1a 91.4 ± 18.9b F = 45.558 df = 1, 19 P < 0.001 Means within a row with same letter not significantly different, protected Tukey Test (P > 0.05). W. Osbrink, M. Cornelius - Acoustical Detection of Coptotermes formosanus76 Table 6. AED counts (10 s) of termites in trees (mean ± SE). Wave guide location on tree Oak Trees North East South West ______________________ ______________________________________ ______________________ __________________________ Base 20 cm Base 20 cm 122 cm base 20 cm base 20 cm 1 3.0 ± 0.1c 4.2 ± 0.8c 2.8 ± 0.5c 14.3 ± 4.4bc 2.8 ± 2.8c 23.5 ± 5.7b 2.1±0.4c 3.5±0.9c 58.2±5.2a F = 37.895 df = 8, 89 P < 0.001 2 3.7 ± 0.1b 3.2 ± 0.1b 1.9 ± 0.5b 2.4 ± 0.6b 4.2 ± 0.6b 3.2 ± 0.7b 354.6±55.4a 35.1±6.1b 4.1±1.3b F = 39.070 df = 8, 89 P < 0.001 3 276.0 ± 9.1a 4.3 ± 1.0b 3.9 ± 0.9b 5.0 ± 0.5b 5.0 ± 0.7b 8.9 ± 2.7b 3.2±0.5b 4.3±0.9b 1.2±0.9b F = 777.511 df = 8, 89 P < 0.001 4 25.6 ± 4.1b 0.0 ± 0.0b 3.9 ± 0.6b 149.7 ± 19.1a 5.9 ± 0.8b 3.3 ± 0.9b 6.7±1.2b 0.0±0.0b 4.6±0.8b F = 54.887 df = 8, 89 P < 0.001 Means within a row with same letter not significantly different, protected Tukey Test (P > 0.05). Table 7. Comparison of AED 2000 with AED 2010 with AED counts (10 s) of termites in trees (mean ± SE). Wave guide location on tree Oak Trees North East South West 6 AED 2010 2.6 ± 0.9a 8.9 ± 2.0a 17.9 ± 3.3a 3.5 ± 1.5a AED 2000 0.3 ± 0.2b 1.9 ± 0.8b 12.7 ± 2.5b 0.9 ± 0.6a F = 6.065 F = 10.357 F = 1.535 F = 2.500 df = 1, 19 df = 1, 19 df = 1, 19 df = 1, 19 P = 0.024 P = 0.005 P = 0.231 P = 0.131 South (March) South (June) 350 AED 2010 1.9 ± 1.0a 0.5 ± 0.3a 13.7 ± 3.6a 47.2± 6.0a AED 2000 0.9 ± 0.4a 0.0 ± 0.0a 8.2 ± 1.3a 27.2 ± 2.3b F = 0.820 F = 2.143 F = 2.041 F = 9.571 df = 1, 19 df = 1, 19 df = 1, 19 df = 1, 19 P = 0.377 P = 0.160 P = 0.170 P = 0.006 Means within a row with same letter not significantly different protected, Tukey Test (P > 0.05).