Termite IPM in historic sites

Abstract

Given the traditional use of wood, historical structures are susceptible to termite damage which can irreversibly affect their historical significance. Methods like drilling into the foundation for termiticide injection are impractical and potentially harmful to the structural integrity of these structures. Moreover, the location of these structures in ecologically sensitive areas prohibits the use of soil termiticides due to potential contamination of groundwater, rivers, or oceans. Termite Integrated Pest Management (IPM) strategies include baiting programs that use minimal quantities of benign insecticides to reduce environmental impact. This approach is consistent with the risk management component of termite IPM. Bait systems are proven to be effective in eliminating termite colonies, contributing to the benefit component of IPM by preserving the value of historical sites. This article discusses successful baiting interventions at notable locations such as the Statue of Liberty National Monument, New York, USA, the Christiansted National Historic Site in St. Croix, US Virgin Islands, and the Historic Tzu-Su Temple in Taiwan.

Recognizing that pests mainly arise due to human activities such as manipulation of the environment, transportation, and lowering of crop’s acceptance level (or economic threshold), Stern et al. (1959) suggested using biological methods to keep pest numbers low and employing chemical insecticides only when pest numbers exceed the acceptable level. They introduced the concept of the Economic Injury Level (EIL), which is defined as “the lowest pest population that causes significant financial loss.” EIL is the unifying principle of Integrated Pest Management (IPM), with EIL = C/VIDK, where C: cost of management tactic per production unit, V: market value per production unit, I: injury unit per pest, D: damage per injury unit, and K: proportional reduction in pest attacks due to management action (Pedigo and Higley 1992). While the EIL concept is widely used in managing pests in agriculture, applying it to noncrop pests has been challenging.

IPM for Noncrop Pests

Instead of using EIL that is primarily based on economic parameters, Olkowski (1974) proposed Aesthetic Injury Level (AIL) for his street-trees IPM, in which the AIL was defined by the level of insect damages (such as insect secretions or droppings on cars parked beneath) tolerated by the public. There have been attempts to use AIL in developing IPM programs for ornamental plant pests (Hellman et al. 1982) and cockroaches (Wood et al. 1981, Zungoli and Robinson 1984), but due to the difficulty in quantifying AILs for these pests, it does not play as a robust role as EILs in agricultural IPM (Pedigo et al. 1986). For noncrop pests, the relationship between injury level and economic loss is difficult to establish, and any values assigned to calculate EILs tend to be too subjective to be used by others. For example, Pinto and Kraft (2000) proposed Action Thresholds (ATs) for some pests in school IPM, but AT values they assigned (5 ants per room, 2 cockroaches per room, etc.) are mostly educated guesses instead of data-based (Su 2023).

IPM for Subterranean Termites

Another group of pests whose damage is difficult to predict is structural pests such as termites. There are 2 types of treatments to control subterranean termites: preventive and remedial treatments. Preventive treatments, also known as preconstruction treatments, are mandated by building codes or by financial institutes that issue mortgages. Most preconstruction treatments apply a large quantity of soil termiticides beneath and around a house foundation regardless of the presence of termites, hence EIL ≤ 0. Remedial treatment is done usually after termites have caused substantial damage to a house, and the house is immediately treated without considering EIL. It is apparent that treatment decision in subterranean termite control has little to do with EIL, AIL, or AT.

In formulating termite IPM, Su and Scheffrahn (1998) argued that “… it is unrealistic to apply the cost-benefit model…for agricultural IPM to an urban pest management program…,” however, “the underline concept of IPM to use a cost-effective approach to solve problem … is the same.” They suggested that the main aim of a termite IPM program is to safeguard homes in the most cost-effective way and explored the cost-effectiveness of termite control businesses. Essentially, for a homeowner to justify the expense of termite control services, the potential termite damage (D) must surpass the service fee (F). This fee includes the control firm’s operational costs plus their profit (P). Operational costs comprise labor (l), materials (m), equipment (e), liability (b), and overheads (o); hence D > l + m + e + b + o + P. However, accurately estimating the potential damage (D) and liability (b) costs is challenging. Instead of attempting to derive accurate figures for D and b, Su and Scheffrahn (1998) suggested that these 2 variables could be expressed by using their proxies. Namely, termite damage potential D is proportional to termite population density near a house, i.e., D is higher in places with higher termite population density, and D can be decreased when termite density is reduced. The liability cost (b) can be related to the risks associated with the quality and quantity of insecticide application, i.e., b is higher when a more persistent and toxic insecticide is used at a larger quantity and conversely is lower when a smaller quantity of less persistent and toxic insecticide is used. It was concluded that a termite IPM could be formulated to use a smaller quantity of less persistent and less toxic insecticide to reduce termite population density near a target structure (Su and Scheffrahn 1998).

Soil termiticide applications and baiting programs are 2 major options for subterranean termite control. The aim of soil termiticide applications is to create an insecticide barrier to exclude soil-born termites from a structure. The amount of termiticide AI (active ingredient) needed for a preconstruction treatment of a home of 230 m² is 0.745 kg or 32.4 kg/hectare (when the minimum rate of 0.06% fipronil is used). A soil treatment may kill small colonies of the eastern subterranean termite Reticulitermes flavipes (Kollar) (Blattodea: Rhinotermitidae) in the vicinity of treated soil (Vargo and Vincent 2012), but it had little effect on larger colonies of R. flavipes (Ripa et al 2007) or much larger colonies of the Formosan subterranean termite, Coptotermes formosanus Shiraki (Chouvenc 2018, Su 2019). The objective of bait application, on the other hand, is to reduce or eliminate colonies of subterranean termites in the vicinity of a structure. Field studies have shown that only a small quantity (0.012–1.142 g) of chitin synthesis inhibitors (CSIs) is needed for successful colony elimination (Evans and Iqbal 2014). Assuming 1 g of CSI is used for a house of 230 m², the pesticide used rate for a baiting program amounts to be ~0.043 kg/ha, which is 753-fold less pesticide use than soil termiticide application. Moreover, the most used soil termiticide, fipronil (rat oral LD₅₀: 97 mg/kg), is 33-fold more toxic than a CSI such as noviflumuron (rat oral LD₅₀: > 3,160 mg/kg) used in a bait system. Consequently, the risks (b) associated with a baiting system are drastically reduced. Couple with the fact that a baiting system reduces the termite damage potential (D) by eliminating termite colonies near a structure, it is a better option than soil termiticide for use in a termite IPM program, and baits have been used in many environmentally sensitive historic sites.

Baiting Programs for Termite IPM in Historic Sites

Due to the traditional use of wood, historic structures are prone to termite infestations, leading to significant and often irreversible damage that diminishes their historical values. The cost of repairing or restoring such damage is considerable. These structures frequently exist in ecologically sensitive areas, like those adjacent to groundwater, rivers, or oceans, making the use of soil termiticides a prohibited practice. Moreover, injecting liquid termiticides into the soil by drilling through foundation floors is generally not feasible, as it compromises the integrity of the original structures and risks losing their cultural significance. Consequently, property managers of historic sites frequently seek alternative termite management solutions. One such solution involves the early application of bait systems, which have been tried in various historic sites to effectively eliminate subterranean termite colonies without damaging the original structure. Following are some of the examples.

Statue of Liberty National Monument, New York

The nation’s most iconic monument, the Statue of Liberty, was erected on a pedestal within the former Fort Wood on the 14-acre Liberty Island (Fig. 1A). There were no records of termite infestations on the island until 1994, when flying alates (winged reproductive termites) of the eastern subterranean termites (R. flavipes) were found in the museum area and boiler room. The presence of alates indicates that the colonies had been established on the island for several years, likely introduced during the 1984–1986 restoration project (Su et al. 1998).

A site survey in 1995 identified four clusters of infestations: boiler room (BOL), museum display cases (DIS), sally port (SAL), and sally port exit (EXT) (Fig. 1B). Of the over 70 wooden stakes placed in the soil outside of the structure, 2 detected termites and was replaced by in-ground (IG) monitoring stations as described by Su and Scheffrahn (1986). One IG station (T1) outside of the boiler room was included in BOL cluster, and the one outside of the sally port (T3) was considered part of EXT cluster (Fig. 1B). Above-ground monitoring stations (AGM) consisted of plastic boxes containing wood pieces as described by Su et al. (1996) were installed at indoor infestation sites. Both IG and AGM stations were checked every 1–2 months for termite activity. Additionally, an acoustic emission device (AED) (Dow AgroSciences, Indianapolis, IN, USA) was employed to survey termite activity within the museum’s wood display cases. The AED sensor was placed on a wood surface to detect acoustic signals of > 20 kHz, which are produced when termites tear wood fibers. A control sensor was situated on un-infested wood to cancel out the ambient noise. The AED monitored feeding incidents along the wood grain for distances up to 80 cm, with each incident documented as a single count (Scheffrahn et al. 1993). The rate of termite feeding was quantified by the number of counts per minute. AED count and number of active stations were used to represent termite activity during the baiting program.

Two commercial prototype AG bait stations, soft- and hard-style, were used indoors to deliver baits at active sites. The soft-style station (AGS) comprised a flexible plastic pouch containing 15 g of 0.5% hexaflumuron bait, with a resealable cover flap on the front and a removable flap on the back with flexible adhesive (Fig. 2A). The back flap was removed to allow termites access to the bait. The hard-style station (AGH) consisted of a plastic box containing 25 g of bait 0.5% hexaflumuron bait. One side of the box was open to provide termite access to baits when the AGH is attached over a termite active infestation (Fig. 2A, AGH). Additionally, commercial Sentricon IG bait stations (Dow AgroSciences) (Fig. 2A, S) were installed in soil near active IG monitoring stations of clusters BOL and EXT (Fig. 1B). Initially, these Sentricon IG stations used wooden monitors to detect termites, and when infestation was confirmed, the wood was replaced with a Recruit II bait tube (Dow AgroSciences) containing 20 g of 0.5% hexaflumuron bait.

Bait application started in July 1996 by applying Recruit II bait in one Sentricon IG station outside the boiler room (Fig. 1B, BOL, S1). After one soft-style AG bait, each was applied to AGS4 (Fig. 1B, BOL) in August and September, respectively, termite activity of cluster BOL ceased by October (Fig. 2B).

Between September and October, soft-style AG stations were placed over a highly active loci AGS13 behind the display case at the museum level (Fig. 1B, DIS), which were fully consumed. Despite consuming these baits, termites remained active at DIS (Fig. 2B), and more baits were applied via AG stations (AG16 and AG17) through November 1996 and January 1997. In February, termite activity began to decline, as indicated by the lower acoustic emission counts (Fig. 2B, DIS), and by March, activity ceased.

In the sally port area (Fig. 1B, SAL), 3 AG baits were placed over active infestations on the wall and floor, and after consuming a large quantity of baits (visual estimate of > 50%) between July and September 1996, termite activity declined in October with slight activity lingered on for another 8 months until July 1997 when no termite was found in any of the stations (Fig. 2B, SAL). Only the scavenger springtails (Insecta: Collembola) were found in one AG station, indicating the demise of the baited colony(s).

Near the sally port exit, termites were detected in an IG station (Fig. 1B, EXT, S13) in October 1996, leading to the placement of a Recruit II bait tube. Despite cold weather, termites remained active in the soil until the spring of 1997. More baits were added to additional IG stations (Fig. 1B, EXT, S11, S14), and by June 1997, termites in the IG monitoring station (Fig. 1B, EXT, T3) showed signs of hexaflumuron effects (lethargy and marbled abdomen), indicating colony decline. By September, termite activity had ceased in the Sentricon stations and one IG monitoring station (Fig. 2B, EXT).

During the 12-month baiting period, a total of 1.8 g of hexaflumuron bait was consumed, resulting in the elimination of four clusters of R. flavipes activity from Liberty Island. No termite swarming was observed in 1997 and 1998. Following the elimination, 77 Sentricon stations were installed in the soil around the wall foundation, and to date, no termites have been detected.

Christiansted National Historic Site, St. Croix, US Virgin Island

Fort Christiansvaern, constructed by the Danes from 1738 to 1749, served to protect commercial shipping and to deter slave insurrections at the time (Greene and Cissel 1988). Located in a tropical region known for termite prevalence, it is likely that the fort has faced termite infestations since its construction in the mid-18th century. During a refurbishment project in the early 1990s, termite activity was discovered at multiple locations within the fort, prompting the National Park Service to seek our assistance. A survey conducted from 1994 to 1996 revealed numerous termite infestation clusters within the fort and its stable, with additional infestations identified during the subsequent baiting program (Fig. 3). These termites were identified as Heterotermes sp. (Su et al. 2003). The project’s goal was to eliminate all active termite infestations from the historic site and to streamline the project operation. Nearby infestation clusters were grouped into zones regardless of termite colony origins. The survey identified a total of 7 zones within the main fort structure (Fig. 3, Zone 1–7) and 2 zones in the stable (Fig. 3, Zone 8–9).

Following the detection of termites using survey stakes, IG monitoring stations were installed, and AG monitoring stations (AGM) were placed over areas of live termite activity indoors. Additionally, an AED was employed to identify live termite presence behind wooden structures. The overall termite activity was represented by the number of active sites, which included active IG and AG monitoring stations, bait stations, newly infested survey stakes, and locations with positive acoustic emission signals.

Similar to the Statue of Liberty project, soft-style (AGS) and hard-style (AGH) above-ground stations, along with Sentricon IG stations (S), were deployed at sites of termite activity (Fig. 3). Baits that were significantly consumed were replaced, while baits were left unchanged if live termites were present in the stations. Baits were removed after no termite activity was observed for 2 consecutive monitoring periods but were reinstalled upon detection of new activity.

Among the 9 termite infestation zones, Zones 1, 4, and 7 were controlled with relative ease. Zone 1 spanned 3 clusters of termite activity across the 1st and 2nd floors (Fig. 3). In October 1996, with 7 active loci identified, 2 AGS, and 2 AGH bait stations were installed over foraging tubes of Zone 1 (Fig. 4). Additional baits were placed in the winter, leading to a significant reduction in termite presence by March 1997, with termite activity ceasing by May 1997. Baits were removed in July 1997. Zone 4 included termite activity on the north side of the fortress (Fig. 3), with an IG monitoring station installed in October 1996 showing minimal termite activity. In May 1998, termites infested 3 wooden stakes, which were then replaced with Sentricon IG stations. A Recruit II bait tube was added, but bait consumption was minimal. Despite this, no termite activity was observed in Zone 4 since July 1998 (Fig. 4). In Zone 7, several foraging tubes were found on the ceiling of the hallway leading to the stable yard (Fig. 3). An AGS was placed over a foraging tube, and after baits were consumed between June and October 1998, termite activity in Zone 7 was eliminated (Fig. 4).

Baiting of Heterotermes activity in Zones 2, 3, and 9 was more complex compared to other zones. In Zone 2, 3 AGHs were installed in March 1997 overactive foraging tubes on the ceiling wooden beams of the second floor (Fig. 3). Despite light bait consumption initially, termite activity stopped by May, and baits were removed in July. However, activity resumed in December, necessitating additional AGH stations. By the spring and summer of 1998, after significant (visual estimate of >50%) bait consumption, termite activity ceased in October (Fig. 4). The characteristic hanging tubes of Heterotermes were found extending from the ceilings of 3 rooms of Zone 3 (Fig. 3), and 2 AGHs plus one AGS were placed over these tubes in August 1996. Bait consumption through 1996 and 1998 was erratic, but more AG bait stations were added, and by May 1998, termite activity ceased, and baits were removed (Fig. 4). Although activity briefly resumed in September 1998, it was eliminated within a month, with no activity detected since October 1998. Zone 9, the stable area (Fig. 3), had foraging tubes on wooden rafters and walls. Four AGH stations placed in October 1996 saw no bait consumption until March 1997. With extensive bait consumption in the spring and summer 1997, termite activity ceased in August. However, this cessation was temporary, with activity rebounding several times until February 2000 (Fig. 4). After 39 months of baiting, termite activity in Zone 9 eventually ceased.

Baiting programs in Zones 5, 6, and 8 are characterized by a prolonged period of termite activity during which baits were consumed. Zone 5 (Fig. 3), known for the most severe termite infestation within the fort, had undergone unsuccessful treatments with liquid insecticides from 1991 to 1995. Monitoring starting in August 1996 identified 11 active termite sites by October, with 8 sites receiving AG bait stations (Fig. 4). By March 1997, this number increased to 15 active sites. A fluctuating pattern of activity was noted from June 1997 to May 1998, with active sites ranging from 7 to 15 (Fig. 4). Activity declined by June 1998, and from November 1998, no further termite presence was detected in Zone 5 (Fig. 4). Zone 6, similar to Zone 5, sustained significant termite damage and also received liquid insecticide treatments from 1991 to 1995 (Fig. 3). Survey stakes placed in the summer of 1996 indicated termite activity in soil, leading to the installation of Sentricon IG stations. From October 1996 to the end of 1998, termites consumed a substantial amount of bait, leading to a gradual decline in activity, and by January 1999, no termite activity was detected from Zone 6. In Zone 8 (Fig. 3), active foraging tubes were found in October 1996 on ceilings and walls of the stables. Despite intermittent bait consumption after the installation of AG bait stations in February 1997, activity continued with minimal bait consumption through 1998 and 1999. After a total consumption of only 63 g of bait over a 36-month period, termite activity ceased in February 2000, with no further activity observed thereafter (Fig. 4).

By the spring of 2000, all Heterotermes activities at the Christiansted National Historic Site were eliminated. The duration required for elimination varied from as short as four months in Zone 7 to as long as 39 months in Zone 9. In many zones, there were probably multiple colonies that had to be sequentially baited and eliminated over a long period. The quantity of bait consumed by the termites across different zones ranged from 21.0 g in Zone 1 to 325 g in Zone 5

Historic Tzu-Su Temple, Taiwan

The Tzu-Su Temple, originally built in 1769, has undergone several cycles of destruction (fire, war, and termite) and reconstruction, with the current structure’s construction starting in 1947. It is recognized for embodying southern Chinese architectural and artistic styles. The temple features a construction where roofs and ceilings are supported by wooden beams resting on columns of stone or copper. It is adorned with intricate sculptures (Fig. 5A), many of which are crafted from Japanese cedar (Cryptomeria japonica D. Don) or camphor (Cinnamomum camphora (L.) Sieb.), materials chosen for their insect-repellent properties. Despite these precautions, the Formosan subterranean termite, C. formosanus has persistently damaged these wooden elements. Records indicate that since 1950, teams of craftsmen have been employed to continuously replicate and replace the termite-damaged wooden artworks.

As with previous projects, IG monitoring stations and Sentricon IG stations were established following the positive detection of termites by stake survey. Commercial hard-style AG baits (Recruit AG, Dow AgroSciences) and Recruit II for Sentricon IG stations were used to deliver baits containing 0.5% hexaflumuron. Number of active stations, including both bait and monitoring stations, were used to represent termite activity.

In the summer of 1999, a survey of the site identified widespread damage, active foraging tubes, and termite infestations throughout the temple. For effective management of the treatment project, these infestations were grouped into four distinct areas: the Main Shrine, the Drum Tower on the west wing, the Bell Tower on the east wing, and the Auxiliary Building (Fig. 6).

Baiting began in the Main Shrine in October 1999 with the installation of 25 Recruit AG bait stations on stone columns supporting wooden structures (Fig. 5B). By early November, 19 AG stations showed significant (visual estimate of >50%) bait consumption and were replaced with fresh baits (Fig. 7). Additionally, 12 AG bait stations were placed in the attic in November, with many sustaining considerable bait consumptions within 2 wk. More stations were installed on the second floor of the Main Shrine (Fig. 6). After the consumption of 30 Recruit AG baits from October to December 1999, termite activity slightly declined in January 2000, likely due to lower temperatures, but increased again in spring 2000, with more baits added and consumed (Fig. 7). By April, activity declined to 17 stations, with visible signs of bait effect on termites, such as marbled coloring and sluggish movement. In May 2000, the C. formosanus activity, present since the 1950s, ceased abruptly from the Main Shrine.

In November 1999, Recruit AG baits were placed over active foraging tubes on the second floor and attic of the Drum Tower (Figs. 6 and 7). Similar to the Main Shrine, termite activity declined slightly in January 2000 but resumed with continuous bait consumption through spring. By April, there was a significant reduction in activity, and by May, termite activity ceased in the Drum Tower.

In the Bell Tower (Fig. 6), only old termite damage and inactive tubes were found across its 3 levels. No consumption of Recruit AG baits was observed, indicating the absence of termites in this part of the Temple.

Similar to the Main Shrine, Recruit AG baits were deployed in the Auxiliary Building in October 1999. Due to high feeding rates, new baits were added bi-weekly in November and December (Fig. 7). By January 2000, termite activity declined to a single active station, and by March 2000, no termites were detected in the Auxiliary Building. This cessation of activity continued until July 2000, when the subterranean termite Reticulitermes speratus (Kolbe) was found in one Sentricon IG station. Reticulitermes speratus is one of the pest termite species in Taiwan, and this species was probably excluded from the territory previously occupied by C. formosanus. Laboratory and field studies showed that when C. formosanus colonies were eliminated by baits, their colony territories could be re-occupied by R. flavipes (Messenger et al. 2005, Su et al. 2016, Bernard et al. 2017). The R. speratus colony detected in the Sentricon IG station was subsequently baited with Recruit II and eliminated (Su and Hsu 2003).

After successfully eliminating all termite activity in the Temple, Sentricon IG stations were installed in soil around the Temple and Auxiliary Building in August 2000, and within 2 wk, one Sentricon IG station on the East Perimeter detected C. formosanus (Fig. 7), indicating a high local population density of C. formosanus. Moreover, indoor termite activity, such as live foraging tubes, was found near the active Sentricon IG station in September, and Recruit AG baits were installed over these newly emerged activities. The new C. formosanus activity likely originated from an external colony, given the 5-month gap since the last elimination of C. formosanus in March 2000. This invading colony on the East Perimeter consumed baits intensively for 3 months before being eliminated by December (Fig. 7).

In November when C. formosanus activity on the East Perimeter was on the decline, one Sentricon IG station on the West Perimeter detected another C. formosanus colony (Fig. 7). Recruit II baits were applied from November 2000, and by May 2001, the baited colony was eliminated (Fig. 7).

As with many places in Taiwan, the Temple is located in a region with high populations of C. formosanus. Despite the successful elimination of termites from the Temple property and its surroundings, the risk of re-invasion by this highly destructive termite species remains. To safeguard this historic site from further damage, the management project was transferred to a professional pest control firm. This firm has been responsible for regular monitoring of Sentricon IG stations and conducting periodic inspections of the site.

Termite IPM in Environmentally Sensitive Historic Sites

Prior to the development of termite bait systems, management strategies at locations such as Fort Christiansvaern in the US Virgin Islands and the Cabildo and Presbytère in New Orleans involved drilling and injecting liquid insecticides, yet these methods failed to halt termite activity (Su et al. 2000, 2003). An intensive baiting effort, some may take up to 39 months, as shown in the above examples, can eliminate all termite populations in these historic sites with minimal impact on their historical value. Such programs utilize a small quantity of low-toxicity insecticide to minimize environmental and health risks, which aligns with the risk-related cost considerations of termite IPM as outlined above. The effectiveness of bait systems in eliminating termite colonies addresses the benefit aspect of the IPM cost-benefit equation. One shortcoming of a baiting program is the lengthy process it may take to accomplish the goal, especially when multiple colonies have to be baited, as shown in Fort Christiansvaern and Tzu-Su Temple, but the time required is well justified by the long-term protection offered by the elimination of the source of the infestation. Because many of these historic sites are in areas with a high density of termite populations, an on-going monitoring and routine inspection of the sites is essential for early detection of re-invading termites.

Acknowledgments

I am grateful to the property managers of these historic sites and many collaborators for their technical assistance and Dow AgroSciences for providing bait materials.

Funding

Partial funding of the projects was provided by the National Center for Preservation Technology and Training, National Park Service, and the Statue of Liberty-Ellis Island Foundation, Inc. Additional support was provided by the USDA National Institute of Food and Agriculture, Hatch projects number FLA-FTL-005865.

Author Contributions

Nan-Yao Su (Conceptualization [lead], Data curation [lead], Formal analysis [lead], Funding acquisition [lead], Investigation [equal], Methodology [lead], Project administration [lead], Resources [lead], Supervision [lead], Validation [lead], Visualization [lead], Writing—original draft [lead], Writing—review & editing [equal])

References