PINE TREE SURVEY AREA

The area selected for the survey (Fig. 1) was chosen for its urban forest canopy, with the presence of many old live oaks and slash pines in residential settings. In addition, this area is known for relatively high densities of C. gestroi (Chouvenc et al. 2016a, 2017). The survey was initially performed by driving the major streets and visually locating pine trees from a distance, as many of these trees are taller than most of the other trees forming the urban forest canopy.

Fig. 1.

Residential area surveyed (black outline) for the presence of Coptotermes gestroi in slash pines (Pinus elliottii) in Fort Lauderdale, Broward County, Florida (ca. 26.1000°N, 80.1800°W). A–C indicates the locations of parks with relatively high densities of pine trees, and which were extensively surveyed. The major road in the center of the image is Interstate I-95.

Two types of locations were surveyed in this study: residential (private properties) and parks. We made this distinction because the relative density of pine trees, the surrounding vegetation, the potential stress levels to trees, and street light intensities are different between the 2 types of locations. Three parks (Fig. 1, locations A-C) with relatively high slash pine densities were selected and fully surveyed. These parks initially were undeveloped areas that were eventually converted into recreational parks, while keeping the initial tree canopy. Parks are regularly maintained and mowed. The residential area (3.4 km2, highlighted area in Fig. 1) also was extensively surveyed, and pine trees were spotted while driving. Tree inspection was performed only after obtaining permission from residents. Trees were not inspected when a resident could not be contacted, or if permission was denied. Thus, the survey in the residential area represents a subsample of the pine trees.

TREE INSPECTION

Pine trees were visually observed for the presence of C. gestroi by inspecting the surface of the trunk from the soil surface to approximately 2 m above-ground. Loose, dry pieces of outer bark were gently peeled off by hand to reveal termite presence or signs of termite activity (termite fecal material, mud tubes, termite damage, or carton material, which is a sponge-like fecal material that fills cavities after wood consumption). Removal of outer bark was kept to a minimum in order to preserve the aesthetics of the trees. Each pine tree was visually inspected for at least 2 minutes by 2 observers. Trees were classified into 4 categories based on the presence of termites and tree status: “live non-infested,” wherein no termites or signs of termite activity could be detected; “live infested,” wherein C. gestroi or obvious signs of C. gestroi activity were detected; “dead infested,” wherein the inspected pine tree was dead and C. gestroi was present; and “dead, non-infested,” wherein a dead pine tree had no obvious signs of termite activity or termite feeding damage. All termites collected from pine trees were confirmed to be C. gestroi by using soldier morphology.

The distribution of the 4 tree categories was compared among the 3 parks to first determine if all 3 locations (A-C) had similar proportions of trees infested with C. gestroi, using a Chi-square test of goodness of fit for multiple groups, with a Yates correction applied (only 3 tree categories were found in parks, 3 groups to compare, df = 4). After confirming that all 3 parks had the same proportions of tree categories, all trees from the parks were combined in a single dataset and the distribution of tree categories between parks (combined, 185 trees) and the residential area (199 trees) were compared with a Chi-square test (4 categories, 2 groups, df = 3).

TERMITE DAMAGE IN TREES

One of the hypotheses for slash pines being differentially susceptible compared to hardwood species is that the location of feeding damage by C. gestroi is different. We tested this hypothesis using 2 different means: visual inspection of the cross-section of dead trees, and measuring the wood densities of live trees using a resistograph (IML Wood Testing Systems, LLC, Moultonborough, New Hampshire, USA).

Visual Inspection of Cross-Sections

Three dead pine trees found during the survey were selected and permission was granted from homeowners to have the trees cut down. Cross sections were made using a chainsaw, and were visually assessed for the presence of feeding damage and carton material. In addition, 3 live oak trees where termite presence had been detected also were selected, and infested limbs were cross-sectioned, with permission, to identify the location of termite damage and carton material in trees.

Wood Density of Infested Trees

In order to support the hypothesis that C. gestroi termites are a major stress to live slash pines, it is necessary to confirm that the heavy damage afflicted to slash pines occurs before the death of the tree and not post-mortem. Therefore, the visualization of feeding damage from cross-sections of dead slash pines that had been cut down also needs to be shown to occur in live trees. However, cutting down these live trees was not possible. Alternatively, slash pine trees that were still alive, but known to be infested with C. gestroi, were tested with a resistograph to indirectly locate termite damage within live trees, if any. This tool measures the resistance of the wood as the drill bit penetrates it and provides direct analog readings of the wood's relative density. The trunks of all trees tested in this study with the resistograph measured at least 40 cm in diameter at 1.5 m above-ground for both slash pines and live oaks. During the survey, 3 healthy pine trees with no visible signs of termite infestation (controls), and 3 live pine trees with confirmed termite activity and visible advanced feeding damage were selected. Each tree was tested with the resistograph at 3 heights: 0.5, 1.0, and 1.5 m above the soil. All relative resistance values from the 3 trees were digitally converted and averaged (n = 3 × 3 subsamples). The density measurements were repeated with 3 healthy live oak trees, and 3 C. gestroi-infested oak trees using the same protocol.

TREE INFESTATION

Although C. gestroi infestations were surveyed principally in association with P. elliottii, we also observed the presence of this species of termite in many other tree species in the surveyed area (Fig. 2): live oak (Q. virginiana); sabal palm, Sabal palmetto (Walt.) (Arecaceae); pindo palm, Butia capitata (Mart.) (Arecaceae); gumbo-limbo, Bursera simaruba (L.) (Burseraceae); royal Poinciana, Delonix regia (Boj.) (Fabaceae); Australian umbrella (Schefflera actinophylla), glaucous cassia, Senna surattensis (Burm.) (Fabaceae), and other plants that were not identified in this study. This indicates that C. gestroi can utilize a number of different trees in Florida, including many of the dominant native and non-native species, though whether or not they were primary or secondary invaders was not determined.

Slash pine trees were the easiest tree species to survey for detection of termites because the activity could readily be detected on the surface of the tree, or immediately underneath the loose outer bark (Fig. 3). In parks, we detected 32, 50, and 104 slash pine trees at locations A, B, and C, respectively. All 3 parks had similar termite infestation proportions, with only 3 of the 4 tree categories detected, as all dead trees were positive for termites (Fig. 4: X2 = 2.17; df = 4; P = 0.71). During our survey in the residential area, we detected 260 slash pine trees and were able to inspect 199 of them, from 42 separate private properties (Fig. 5A). Among the 199 inspected slash pine trees, 79 were live non-infested, 91 were live infested, 25 were dead infested, and 4 were dead non-infested, resulting in a 58% (N = 199) infestation rate (Fig. 6). In comparison, slash pines surveyed in parks (Fig. 5B, N = 186) had a lower rate (18%) of C. gestroi infestation (X2 = 73.9; df = 3; P < 0.001).

TERMITE DAMAGE ON TREES

The cross-sections of dead P. elliottii revealed that most of the damage was restricted to the outer layer of the wood, and that in no instances was the heartwood affected (Fig. 7A–C). Over time, damage to the dead trees progressively turned the outside layers of wood into carton material, eventually causing the bark to fall from the tree. As the carton material was washed off by the weather, only the heartwood was left standing from the ground, untouched by the termites (Fig. 7D).

Comparative observations of tree cross-sections between pine and oak trees revealed a striking difference in feeding damage patterns. The damage in pine trees was restricted to the outer tissues of the trunk, including the bark, phloem, and cambium, thus effectively girdling the trunk. In sharp contrast, with live oak trees and other hardwood species, the termite damage pattern was primarily restricted to the interior tissues of the tree's trunk (older xylem), forming a central cavity (Fig. 8). This shows that termite feeding damage in P. elliottii is directly adjacent to the living tissue of the inner bark, whereas in Q. virginiana the damage is much further inside the trunk, away from living tissues. This observation was supported by the resistograph measurements, which showed that for live infested slash pine trees, the damage was located only within the proximity of the bark (outer layers), acting to girdle the tree trunk, whereas in infested live oak trees, the damage was mostly restricted to a central cavity (Fig. 9).

IMPACT OF SUBTERRANEAN TERMITES TO LIVE TREES

In addition to being a major structural pest (Chouvenc et al. 2016a), this study shows that C. gestroi has the potential to become a major threat to live trees in southeastern Florida. Some of the native trees may be naïve hosts without sufficient adaptations to survive such termite pressure. Although some tree species may be able to sustain the damage from termites, our observation suggests that C. gestroi infestation in P. elliottii may ultimately be fatal. The resistograph data confirmed that live slash pines infested with C. gestroi sustained heavy damage corresponding to the girdling observed in cross sections of dead slash pines. Therefore, the damage that could potentially result in the death of slash pines can occur pre-mortem when termites are present. This observation and the high association rate of C. gestroi with declining slash pines suggest that C. gestroi is a major stress factor to slash pine in urban southeastern Florida. This may be partially because wounding of live tissues puts too much stress on the trees, which are relatively slow-growing by nature, and the trees may not be able to sustain the pressure of a large C. gestroi colony.

When comparing feeding damage in different tree species, the feeding damage in slash pines is located adjacent to the live tissue underneath the bark, which differs from hardwood trees, wherein the damage is located in a central cavity far from the live tissue. It is possible that the heartwood of pines is resistant to termite attack because of its density and concentration of allelochemicals (Scheffrahn et al. 1988; Scheffrahn 1991). Ultimately, this may serve to restrict C. gestroi to the outer layers of the tree trunk. One key observation we made is that termite-infested pine trees did not produce resin secretions to flush insects out of their trunk, like pines would do with some beetle infestations (Phillips & Croteau 1999). One of the possibilities for the absence of defensive secretion is that C. gestroi deposits fecal material in direct contact to the wood chamber that was excavated. We suggest that such fecal deposition may be analogous to a bandage to the wound, inhibiting subsequent secretions from the pine tree, keeping termites safe from the tree's defense. However, a chemical analysis is necessary to confirm if the fecal deposit in direct contact with the wood is saturated with resin. Alternatively, slash pine trees that are heavily infested with C. gestroi may have been stressed previously by other factors, weakening the tree and preventing it from producing the resin.

Fig. 2.

Feeding damage from Coptotermes gestroi on various tree species. (A) Carton material extracted from (B) a dead gumbo-limbo tree (Bursera simaruba); this tree was still alive less than 1 yr before the picture was taken. (C) Termite feeding damage on a Senna surattensis, still alive, but having lost most of its branches. (D) Old scars from feeding damage of C. gestroi on a live sabal palm (Sabal palmetto). (E) Active infestation and carton material in the trunk of a live sabal palm. (F) Carton material in the heart of a dead pindo palm (Butia capitata).

Some may argue that it is unknown if the termite damage is the cause or a consequence of the initial stress to P. elliottii, as other stress sources may ultimately be responsible for slash pine decline. Because of the long life cycle of this termite species (> 5 yr), and the challenge to infest slash pines with C. gestroi colonies and wait for feeding damage (it may take up to 10 additional yr) while preventing any other stress sources, it is logistically difficult to test for the Koch's postulate in this system. It is possible that P. elliottii in urban setups are already stressed by a wide range of factors, and that C. gestroi is taking advantage only of weakened trees. We noticed that many trees had what appeared to be numerous cerambycid (Coleoptera) exit holes, in absence of sap secretion. These trees were therefore stressed, but it is unknown if termites where the initial stress factor and if beetles came before or after the stress from termite damage. Urban stresses from additional sources contributing to P. elliottii decline cannot be ignored. Grade changes due to construction, soil compaction due to reoccurring landscape maintenance operations (frequent mowing), excessive irrigation and fertilization, potential diseases, and other insects all may take their toll. Urbanization, therefore, may be a confounding stress factor that can exacerbate the susceptibility of slash pines to wood destroying insects, including C. gestroi.

Fig. 3.

Active infestation of Coptotermes gestroi in slash pines. (A) Visible termite mud covering the surface of the bark of a pine tree. (B) Fecal deposition and carton material visible after peeling the dead scale of the bark from a pine tree. In this case, the activity was only noticeable after scraping the bark.

Fig. 4.

Details of slash pines infested by Coptotermes gestroi in three parks (A–C). The 3 locations are referred to in Figure 1. The prevalence of each tree category within each location was not different among locations (χ2 = 2.17; df = 4; P = 0.71).

It is remarkable that C. gestroi was detected more often in P. elliottii around residential areas (58%) than in the 3 parks (18% of the trees we surveyed). The difference in surrounding vegetation and urbanization composition, and in relative densities of slash pines, may be an important factor influencing the rate of C. gestroi infestation. One additional factor may be the relative absence of city lights in the direct proximity of pine trees in parks, as Chouvenc et al. (2017) showed that C. gestroi alates were attracted to light during dispersal flights at dusk in early Mar. As the city lights automatically turned on, alates are directly attracted to them. Therefore, the presence of city lights in the residential area may be responsible for a higher incidence of C. gestroi in the landscape around structures, in addition to other factors.

The Asian subterranean termite was first detected in 1996 in South Florida (Su et al. 1997a), and has been spreading since at an alarming rate, building high-density populations throughout urban southeastern Florida (Chouvenc et al. 2016a). With termite densities increasing over the past few years (Chouvenc et al. 2017), slash pine and other susceptible trees are at risk. More problematic, our survey suggests that in residential areas, 10% of the surveyed slash pines already were dead, potentially as a result of termite damage, but this figure may be overly conservative for 2 primary reasons. First, termite infestation in pine trees could be detected only if the damage was heavy enough and located at the base of the tree, where the inspection occurred. If the damage was relatively minor because the tree had become infested only recently, or if the damage was located higher in the canopy, then it might not have been detected. This means that our observed rate of infestation (58%) may be an underestimation. Second, we noticed that some of the dead trees that we surveyed, both in parks and in the residential areas, were cut down after we surveyed them. A common theme echoed by homeowners was that they used to have more pine trees on their property, but had to have them removed after they died within the past few years, before we had a chance to detect and survey them. Therefore, because the survey in the residential area relied on visually locating the pine trees within the urban canopy, it is possible that many of the trees killed by C. gestroi may have been removed prior our survey. Given this scenario, it is possible that our 10% death rate for pine trees in residential areas over the past few years may understate the actual impact of C. gestroi on the pine tree population.

Fig. 5.

Proportion of slash pine trees (Pinus elliottii) infested by Coptotermes gestroi in the surveyed area presented in Figure 1. (A) Residential areas, (B) Cumulative data from all 3 parks. The distribution among tree categories was different between the 2 types of area (χ2 = 73.9; df = 3; P < 0.001).

Fig. 6.

(A) Dead pine tree with (B) active Coptotermes gestroi infestation revealed after peeling the superficial scales of the bark.

EXCHANGES WITH RESIDENTS

During the survey of pine trees in the residential area, 61 trees from 18 properties were not inspected, as we were not able to contact the owners or obtain their permission to access the property. Fortunately, we encountered direct refusal to perform tree inspection only twice during the course of the survey. The vast majority of the residents were welcoming and interested. Therefore, throughout the survey process, we engaged homeowners in discussions about termite problems in southeastern Florida. We report here some of the exchanges that reflect some of the opinions and questions that the local residents had concerning termite issues. Comments presented here are subjective and solely reflect the authors' opinion.

Fig. 7.

Feeding damage from Coptotermes gestroi on dead pine trees. (A) Cross section of a dead pine tree, revealing the feeding occurred from the outsidein, leaving the heartwood intact. (B) All the space between the bark and the tree was replaced with carton material, effectively resulting in the detachment of the bark. (C) Advanced feeding damage, where the bark has fallen off and revealing extensive “shredding” of the wood with deposition of carton material 10–15 cm deep. (D) Final damage on a pine tree, where all the wood, except the heartwood was consumed, leaving a “pole” sticking out from the ground. Notice the bark ring at the base, showing the size of the tree before death, and all the in-ground space between the bark and the heartwood is exclusively carton material.

Fig. 8.

Cross section of tree trunks with Coptotermes gestroi feeding damage. (A) Pine tree (Pinus elliottii), dead, with damage from the outside-in, destroying all the wood material adjacent to the live tissue underneath the bark (bark absent here). (B) Oak tree (Quercus virginiana), still alive, with damage from the inside-out, leaving a central cavity, but sparing most of the material adjacent to the live tissue.

“ Termites? It's okay, I tented last year.” One of the problems is the confusion about termite treatments in Florida because most termite infestations relate to drywood termites, Cryptotermes brevis (Walker) (Isoptera: Kalotermitidae), for which structural fumigation is the most common remedial treatment. Such treatment is useless against subterranean termites and we repeatedly had to explain that subterranean termites are a different problem all-together.

“ Termites in my tree? I'll just cut it down.” This statement was probably the most disconcerting, and unfortunately far too common among residents, as it bypasses the rationale behind termite biology and the legality of tree removal. First, as colonies of Asian subterranean termites are connected to the ground, a large part of a colony may not be contained within the infested tree. If the tree is removed, the next available item for the remaining termite population to consume would be the house itself (or other trees), increasing the chance for structural damage. Second, Broward County has strict rules about tree removal, and most Broward jurisdictions require a permit and mitigation. Steep fines can be given for illegal tree removals. Finally, some of the trees are decades old, and simply cannot be replaced. Let us reiterate that the initial goal of this study was to provide data in order to set-up a long-term strategy to protect and save these trees in the first place.

“ I lost three pine trees this year, what can I do to save the other two?” Alternatively, some concerned homeowners shared their distress about witnessing the death of sometimes centennial trees on their properties. Many homeowners had strong motivation to do anything possible to prevent the death of their trees. We discuss potential remedial treatments to protect pine trees below.

“ Is this because of climate change? I've been here for 35 years and I see things now that I have never seen before.” Although C. gestroi became established in southeastern Florida because of human maritime activity and favorable environmental conditions (Chouvenc et al. 2016a), the peculiar damage on pine trees by this termite species may be inherent to the biology of the pine tree and the behavior of this termite species interacting with it. However, given the hypothesis that temperatures will rise in the foreseeable future, it is possible that this termite species will expand its distribution northward (Su et al. 2017) and that the problem may not be restricted to just southeastern Florida.

“ Termites? I'll just pour some chlordane. I still have some.” No comment, in light of the ban of this pesticide for all uses in the USA since 1988.

PROTECTING PINE TREES FROM TERMITE DAMAGE

Preventive Treatments

There are at least 2 different approaches that can be considered, and that reflect current protocols for preventative structural treatments against subterranean termites. The first would be the use of a chemical barrier using liquid termiticides, so that a foraging colony of C. gestroi in the soil would be denied access to the tree in the first place. However, the surviving termite population (Su 2005) may still affect nearby untreated areas (trees and structures). Second would be the use of termite baits, placed in the soil around the tree. Such termite baits would be fed upon by foraging termites, and result in colony elimination over time (Su 2005; Eger et al. 2012). Loose, shaggy pine tree bark appears to be a good entry point for alates during their dispersal flight (Chouvenc et al. 2017). Most pine trees that we surveyed within a few weeks after major dispersal flight events had C. gestroi wings present underneath the outer layer of the bark, suggesting that alates may have the opportunity to establish colonies directly in live trees. Although there are no research data to support trunk treatments with persistent insecticides to prevent termites from attacking live trees, this approach is commonly used to prevent attack by bark beetles in the urban forest, so it also might be an effective method to reduce the risk of infestation by these termites.

Fig. 9.

Relative resistance (density) of the wood on different live trees, categorized as healthy or infested with Coptotermes gestroi. Each line represents the average resistance value obtained from the resistograph, from 3 trees per category, with 3 subsamples per tree (n = 3). (A) Pine trees, (B) oak trees. Infested pine trees reveal damage right underneath the bark, while infested oak trees reveal a cavity at the center of the trunk. The “step” effect on the lines resulted from the average of 3 different trees, with different level of termite damage.

CONCLUSION

If C. gestroi termite colonies increase their population density in trees, it is possible for mature colonies to move to nearby structures from their underground foraging galleries and cause structural damage after they have fed upon the tree. This underscores the fact that IPM strategies to protect structures should not rely solely on temporarily keeping termites outside of structures (Su 2005), but rather, they should involve colony elimination within the surrounding environment to prevent future potential damage. Aside from structural damage, C. gestroi may have a strong negative effect on the current urban forest canopy of southeastern Florida. Many live trees seem to be directly affected by the feeding damage of this termite species, and the rate of infestation in pine trees and other trees may be critical in the near future for the overall survival of a diverse urban tree canopy. Many of the pine trees currently present in the urban landscape were part of the original environment when urban development occurred more than half a century ago and represent a legacy of the native southeastern Florida landscape. The rapid loss of such trees due to termite feeding activity may, therefore, have irreversible negative consequences for the health and sustainability of affected urban forest canopy.

The continuous feeding damage to live trees also may have an important indirect effect. Trees weakened by Formosan subterranean termite colonies may break or be uprooted during storms (Osbrink et al. 1999; Cornelius et al. 2007; Osbrink et al. 2008). As we advocate that the effects of C. gestroi on live trees in southeastern Florida are analogous to the impacts of C. formosanus on live trees in Louisiana, we raise major concerns about the fate of these weakened trees in the event of a major hurricane landfall in southeastern Florida. In 2005, hurricane Wilma was the last named storm to bring major damage to southeastern Florida, and all the while, many trees have been severely compromised by C. gestroi feeding damage over a span of 12 years. As hurricane Irma struck Florida in Sep 2017, the center of the storm spared Broward County and most oak trees lost only a few branches, at most. However, we monitored several large oak trees that completely collapsed during the storm. Out of 3 collapsed oak trees found, all were hollowed out by C. gestroi, confirming that the tree structures were compromised. Such falling and breaking trees resulted in property damage (Fig. 10). Therefore, in addition to potentially killing slash pines, C. gestroi also can affect other tree species indirectly and negatively affect the south Florida tree canopy as a whole, as previously observed in Louisiana with C. formosanus (Osbrink et al. 1999), and be a hazard to people and properties in urban landscapes.

Fig. 10.

Large oak tree hollowed out by Coptotermes gestroi, with presence of carton at the center of the tree (arrow), and property damage resulting from the tree structural collapse during hurricane Irma (Sep 2017

Therefore, we here suggest that the presence of C. gestroi in trees is a time-sensitive issue because in addition to its potential ability to directly kill trees, it also indirectly can increase the destructive power of major storms by weakening the structural integrity of the trees. There is an incentive to save large trees, including many native trees such as gumbo-limbo, live oak, sabal palm, slash pine and others, which comprise a large proportion of the native urban forest canopy of southeastern Florida. The loss of such urban forest would have direct negative effects on the local ecosystem, and a long term economic impact, as high densities of C. gestroi colonies will ultimately result in structural damage. Managing populations of C. gestroi in urban southeastern Florida through a comprehensive IPM program is very much needed.