Inhibition of Fungal Spore Germination by Nasutitermes - Springer Link

Journal of Chemical Ecology, Vol. 26, No. 1, 2000

INHIBITION OF FUNGAL SPORE GERMINATION BY Nasutitermes: EVIDENCE FOR A POSSIBLE ANTISEPTIC ROLE OF SOLDIER DEFENSIVE SECRETIONS REBECA B. ROSENGAUS,* MICHELE L. LEFEBVRE, and JAMES F. A. TRANIELLO Biology Department, Boston University Boston, Massachusetts 02215 (Received August 26, 1998; accepted August 16, 1999) Abstract—The antifungal property of two of the principal components of the terpenoid frontal gland secretions of nasute termite soldiers was studied by incubating spore suspensions of the fungus Metarhizium anisopliae with a-pinene and limonene singly or in combination at different concentrations. In vitro assays showed that these substances reduced spore germination through direct and indirect (vapor) contact. To determine if the frontal gland secretions protected termites from fungal infection in vivo, the effect of M. anisopliae on the time course of survival of Nasutitermes costalis and N. nigriceps was studied by exposing termites to either a 4.3 × 107 spores/ ml or a control sporeless suspension. The caste composition of experimental groups was manipulated to create mixed-caste subcolonies and monocaste groups. Relative to Coptotermes formosanus, a species that relies on the mechanical defenses of soldiers, N. costalis and N. nigriceps were less susceptible to fungal infection. Spore-exposed N. costalis and N. nigriceps termites had 1.2 times the hazard ratio of death of controls, while the hazard ratio of death of spore-exposed C. formosanus was 11.4 times that of controls. Although the lower susceptibility to infection in Nasutitermes may be explained in part by the antifungal properties of a-pinene and limonene, group composition also played a major role in the survival of spore-exposed termites. Mixed-caste and soldier monocaste groups had 3.4 and 4.7 times the hazard ratio of death, respectively, relative to the worker monocaste treatment. These results suggest that although Nasutitermes terpenoid secretions may have antifungal properties, the caste composition of groups and the social interactions of termites also play a role in determining susceptibility to fungal infection. Key Words—Disease, social behavior, fungistatic, limonene, a-pinene, Nasutitermes costalis, Nasutitermes nigriceps, Metarhizium anisopliae.

* To whom correspondence should be addressed.

21 0098-0331/ 00/ 0100-0021$18.00/ 0  2000 Plenum Publishing Corporation

22

ROSENGAUS, LEFEBVRE,

AND

TRANIELLO

INTRODUCTION

Termite soldiers have evolved defenses that range from mechanical to chemical mechanisms of deterring invertebrate and vertebrate predators (Deligne, 1971; Quennedey, 1975; reviewed in Coles and Howse, 1983; Prestwich, 1984). The phylogenetically lower termite soldiers rely mainly on mechanical defense and have heavily sclerotized head capsules and powerful mandibles used for crushing or slashing (Prestwich, 1984), whereas soldiers of the higher termite species use chemical defenses or a combination of mechanical and chemical defenses (Prestwich, 1983, 1984). The ampule-shaped head of Nasutitermes soldiers houses the frontal gland from which viscous secretions composed primarily of terpenoids are ejected (Moore, 1964, 1968). These secretions function in alarm communication and defensive recruitment (Eisner et al., 1976; Traniello, 1981; Prestwich, 1984) but due to their toxicity might also be involved in the control of nest microbes. Although there is evidence that some termites use chemical secretions to control the growth of microorganisms in the nest (Lavie, 1960a; Batra and Batra, 1966, 1979; Sannasi and Sundara Rajulu, 1967; Bouillon, 1970; Lutikova, 1990; Ljutikova and Judina, 1996; Rosengaus et al., 1998a), the antimicrobial function of frontal gland secretions has not been explored. In this paper, we describe the results of in vitro and in vivo experiments on the fungistatic role of the frontal gland secretions of Nasutitermes costalis (Holmgren) and N. nigriceps (Halderman) soldiers.

METHODS AND MATERIALS

Colony Collection and Maintenance. Mature colonies of Nasutitermes nigriceps and N. costalis were collected in Puerto Rico in 1982 and the Lesser Antilles in 1972, respectively. Both colonies have thrived in the laboratory at 228 C under constant light conditions. Rearing conditions are described in Traniello (1981). Soldiers of both species have comparable body size and mass; the average weight ± SD of soldiers of N. nigriceps and N. costalis is 1.3 ± 0.3 and 1.5 ± 0.2 mg, respectively (P > 0.05; Mann-Whitney U test) (SPSS, 1990). Workers of N. nigriceps and N. costalis are also similar in size, weighing on average 2.9 ± 0.4 and 2.9 ± 0.3 mg, respectively (P > 0.05, Mann-Whitney U test) (SPSS, 1990). Any differences in disease susceptibility within the soldier or worker castes are, therefore, unlikely due to body size. Fungal Spore Preparation. The entomopathogenic fungus Metarhizium anisopliae (Metschnikoff) is known to be naturally associated with termites (Toumanoff and Rombaut, 1965; Sands, 1969; Ko et al., 1982; Lutikova, 1990; Zoberi, 1995; Milner et al., 1998). Spores of M. anisopliae var. anisopliae were harvested from cadavers of Zootermopsis angusticollis Hagen that had been pre-

FUNGISTATIC PROPERTIES OF TERMITE DEFENSIVE SECRETIONS

23

viously exposed to spores originally obtained from the American Type Culture Collection (batch 93-09, media 325; ATCC 90448). A detailed description of the preparation of stock spore suspensions is provided by Rosengaus and Traniello (1997). Control treatments involved a 0.1% Tween 80 solution that contained no fungal spores. Conidia viability was determined by plating 150 ml of a spore suspension on a thin layer of previously solidified potato dextrose agar; germination was recorded at a 400× magnification after incubation (18 hr at 258 C and 61.5% relative humidity) under a 12L : 12D light cycle. The average percent germination across experiments (± SD) was 93.9% ± 3.8% (range 88.7–98.8%). Effect of Terpenes on Spore Germination Rates. The principal terpene component found in the frontal gland of Nasutitermes soldiers is a-pinene (comprising 30–97% of the secretion) with b-pinene (1–64%) and limonene (0–30%) as secondary and minor components, respectively (Moore, 1964, 1968; Grasse´ , 1982). To determine if these terpenoids inhibit fungal germination and/ or development, a 5.7 × 105 spore/ ml solution of the fungus M. anisopliae was incubated with different concentrations of a-pinene and limonene, which were obtained from Sigma Chemical (St. Louis, Missouri). b-Pinene was not tested because it was not readily available from commercial suppliers. The first set of spore–terpene suspensions was prepared by adding a volume of a terpenoid that was proportional to the lowest amount reported in the frontal gland secretions for a variety of Nasutitermes species. Therefore, a 30% a-pinene–spore solution was prepared by adding 300 ml of a-pinene (98% purity) to 700 ml of a 5.7 × 105 spore/ ml concentration. Similarly, 20 ml of limonene (97% purity) were added to 980 ml of a 5.7 × 105 spore/ ml solution to prepare a 2% limonene–spore suspension. The synergistic effect of the compounds was tested by adding 300 ml of a-pinene and 20 ml of limonene to 680 ml of the 5.7 × 105 spore/ ml suspension. To determine the threshold inhibitory effect of both terpenoids, 5.0% and 0.5% a-pinene–spore suspensions and 0.5% and 0.05% limonene–spore suspensions were prepared. The synergistic inhibitory effect of both terpenoids at concentrations of 5.0% a-pinene + 2% limonene and 0.5% a-pinene + 0.5% limonene was also tested. Approximately 40 ml of each spore–terpene suspension was plated immediately after the suspension was made on a microscopic slide that contained a thin layer of solidified potato dextrose agar (PDA, 1 ml before the agar solidified). Antifungal assays of each spore–terpene suspension were replicated three times. As a control, microscope slides were seeded with 40 ml of a 5.7 × 105 spore/ ml suspension to which no terpenoids were added. To determine if the incubation time of spores and terpenoids had an effect on spore germination rates, the same protocol was repeated after allowing the 30% a-pinene–spore suspension, the 2% limonene/ spore suspension, and the combined 30% a-pinene and 2% limonene–spore suspension to stand for 10 hr at 13.58 C before plating on solidified PDA slides.

24


AND

TRANIELLO

Germination rate, which was our measure of conidia viability, was determined by recording the proportion of the total number of spores observed in a field of vision that had a visible germ tube. Germination rates were recorded at 400× magnification after the slides were seeded with a spore–terpene suspension and incubated at 258 C and 61.5% relative humidity for 18 hr under a 12L : 12D light cycle. Each slide was scanned from end to end and germination rates were recorded across 10 consecutive fields of vision for each slide. To test if the volatility of a-pinene and limonene separately and in combination affected spore germination, solidified PDA plates (60 mm × 15 mm) were seeded with 200 ml of a 2.1 × 107 spores/ ml solution stained with blue food coloring to allow us to assess visually that the agar medium was completely covered with the spore solution. The food coloring had no effect on conidia viability. A 0.5-mm × 0.5-mm piece of filter paper (Whatman Qualitative No. 5) was secured with tape to the inner side of a Petri dish lid. The filter paper was impregnated with 10 ml of either a-pinene (N c 15), limonene (N c 15), or a mixture of 30% a-pinene + 2% limonene (N c 10). The mixture was prepared by adding 75 ml of a-pinene and 5 ml of limonene to 170 ml of sterile 0.1% Tween 80 suspension. Controls (N c 20) had filter paper secured in the same manner but treated only with 10 ml of a 0.1% Tween 80 suspension. The lid was immediately placed over the PDA plate; the filter paper in experimental and control dishes was suspended approximately 5 mm above (and not in direct contact with) the agar surface. All plates were stacked inside a covered plastic box and maintained at 258 C and 61.5% relative humidity under a 12L : 12D light cycle for 12 days. The growth of M. anisopliae on each of these PDA plates was estimated by using relative planimetry on days 6 and 12 after seeding. The average area of zones free of fungal growth for the a-pinene-, limonene-, and the combined a-pinene and limonene-treated dishes was compared to that of control plates. Exposure of Termites to Spores. To determine if the fungistatic effects of frontal gland secretions protected termites from fungal infection, workers, and soldiers of N. nigriceps and N. costalis were removed from their parental colonies by using soft forceps. Three social groupings were established: (1) worker monocaste groups composed of 5 workers, (2) soldier monocaste groups composed of 15 soldiers, and (3) mixed-caste groups consisting of 15 soldiers and 5 workers. The manipulation of group composition allowed us to determine the survival of individuals nesting with members of the same and different castes. Termites in each group were exposed to spores in solution by allowing them to walk freely for 1 hr inside a plastic Petri dish (100 mm diam. × 15 mm high), the bottom of which was lined with a filter paper disk (Whatman Qualitative No. 5, particle retention >2.5 mm, fine porosity) that had been moistened with 1 ml of a lethal spore solution (4.3 × 107 spores/ ml) (Rosengaus and Traniello, 1997). Ten replicates per social group were established for each species. Follow-


25

ing exposure, termites were transferred to new sterile Petri dishes (100 mm diam. × 15 mm high) lined with filter paper (Whatman Qualitative No. 1) moistened with 1 ml of sterile water by tapping lightly the exposure dish. In this way we avoided manipulating termites with forceps and/ or removing spores from their cuticle during the transfer. All Petri dish nests were stacked in covered plastic boxes (30 × 23 × 10 cm) lined with a moist paper towel and maintained at 258 C and 61.5% relative humidity under a 12L : 12D light cycle. In addition, control replicates were established for N. costalis (N c 10 replicates/ social grouping) and N. nigriceps (N c 10 replicates/ social grouping) in which termites were exposed for 1 hr to a control Tween 80 solution lacking fungal spores. Due to the fragility of nasute termites, an initial census was performed 10 hr aftr exposure to spores. Any individual that appeared injured in any way or behaved abnormally was removed from the subcolony and deleted from the sample. To avoid contamination of controls with spores, control replicates were consistently set up before the spore-exposure replicates, and forceps were submerged in alcohol and then flamed between treatments. The degree of susceptibility to fungal infection was measured by recording daily the survivorship of soldiers and workers for a period of 10 days after exposure to either the spore or control suspension. Dead termites were removed immediately to ensure that the spores provided during the exposure treatment were the only source of disease inoculum. To assess if the survival of the chemically protected nasute soldiers was representative of termite species that rely on a mechanical defense, we ran concurrently with the Nasutitermes replicates a set of experiments in which 15 soldiers and 5 workers of the Formosan termite Coptotermes formosanus Shiraki were exposed to either the same 4.3 × 107 spores/ ml or control Tween 80 suspension. To confirm that the cause of mortality was due to M. anisopliae infection, dead termites were surface sterilized with a 5.2% sodium hypochlorite solution and rinsed twice with sterile water before they were plated on PDA plates (60 × 15 mm). The plates were stacked inside a covered plastic box and incubated at 258 C and 61.5% relative humidity under a 12L : 12D light cycle until the spores diagnostic of a M. anisopliae infection appeared on termite carcasses (Rosengaus and Traniello, 1997). Confirmation rates of dead control termites were similarly determined. Statistical Analysis. Spore germination rates were not normally distributed. Therefore, differences in germination between control spore suspensions and terpenoid-treated spore solutions were analyzed with either a Mann-Whitney U test or a Kruskal-Wallis test (SPSS, 1990). The volatile effect of terpenoids on fungal development was analyzed by comparing the average area lacking fungal growth in the a-pinene-, limonene-, and combined a-pinene + limonene-treated dishes to that of control PDA plates through relative planimetry (Kruskal-Wallis test or a Mann-Whitney U test; SPSS, 1990).

26


AND

TRANIELLO

A Cox proportional hazard regression analysis was performed on the survival data by including in the model the variables “exposure treatment,” “group composition,” “caste,” and “species.” The hazard function characterizes the instantaneous rate of death at a particular time, given that the individual survived until that point, while controlling for the effect of the various variables on survival (Smith et al., 1994). This estimate provides a relative measure of the degree of susceptibility of termites exposed to different treatments while controlling for the effect of all other variables included in the model. To determine if the two Nasutitermes species were susceptible to infection by M. anisopliae, we estimated the median survival time of termites (LT50 ) and statistically compared the survival distributions between spore-exposed and control soldiers and workers (Breslow statistic, Kaplan-Meier survival test) (SPSS, 1990).

RESULTS

Fungistatic Effects of a-Pinene and Limonene. The average germination rate of spores was influenced differentially by the terpenoid compounds: apinene had a greater inhibitory effect on fungal growth than limonene (Figure 1). A synergistic antifungal effect was evident when both 30% a-pinene and 2% limonene were added to a spore solution. The average percent germination (± SD) of spores in the combined terpenoid–spore solution was 1.4 ± 5.0%, a reduction that was lower than that of either terpenoid alone (8.0 ± 15.8% germination in the 30% a-pinene-treated spore suspension, P c 0.06, Mann-Whitney U test; 82.5 ± 17.8% germination in the 2% limonene treated spore suspension, P < 0.001, Mann-Whitney U test). The inhibition of fungal germination by a-pinene did not require prolonged incubation with spores. The germination rates of spores in a 30% apinene–spore suspension seeded immediately were not significantly different from those seeded after a 10-hr incubation period (Mann-Whitney U test with Bonferroni correction due to four comparisons, P > 0.01; Figure 1). The germination rate for limonene-treated spores, in contrast, was reduced only after a 10-hr incubation (P c 0.004, Mann-Whitney U test; Figure 1). The inhibitory effect on spore germination of a-pinene was also evident at a 5% concentration (Figure 2A). A further reduction in the concentration of a-pinene to 0.5% eliminated the inhibition of germination. Limonene at concentrations of 0.5% and 0.05% did not lower germination rates relative to controls (Figure 2B). The effect of a-pinene and limonene appeared to be synergistic at reduced concentrations of 5% a-pinene + 2% limonene. The addition of this mixture to the spore solution reduced the average germination of spores to 4.6 ± 11.6%, which was significantly lower than the reduction produced by either a 5% a-pinene–spore suspension (46.8 ± 27.8%, P < 0.001, Mann-Whitney test) or a 2.0% limonene


27

FIG. 1. Average (± SD) percent germination of a control 5.7 × 105 spores/ ml suspen⯗⯗ ), 2% limonene-treated (R /// ), and 30% a-pinene + 2% sion (R), 30% a-pinene treated (R limonene-treated spore solution (Q) seeded immediately and after a 10-hr incubation period. *Significance at P < 0.001 (Kruskal-Wallis test) (SPSS, 1990).

suspension (82.5 ± 17.8%, P < 0.001, Mann-Whitney Test). No synergistic effect was observed when terpenoid concentrations were reduced to 0.5 a-pinene and 0.5% limonene concentrations (Figure 2C). The vapors of a-pinene had a strong inhibitory effect on the growth of fungus (Table 1). An area of inhibition was visible directly beneath the suspended paper treated with a-pinene within 24 hr. Over time, the zone of inhibition radiated outward in an quasicircular shape. The volatile components of limonene, on the other hand, had no effect on spore development relative to controls (Table 1). The average area free of fungal growth due to the volatile effect of a-pinene was approximately twice that of the control PDA plates 6 and 12 days after spores were seeded (Table 1). Thus, a-pinene, the principal component of the soldier’s defensive secretion, appeared to exert an antibiotic effect through both volatility and direct contact with spores, whereas limonene at a 2% concentration influenced germination only after prolonged physical contact with the spores (Figure 1). No synergistic volatile effect was observed when a-pinene and limonene were tested in combination. The inhibitory effect of a-pinene + limonene on days 6 and 12 after seeding was not different from the inhibitory effect of a-pinene alone (P > 0.2, Mann-Whitney U test), although it did differ

28


AND

TRANIELLO

FIG. 2. Average (± SD) percent germination of a 5.7 × 105 spore suspension treated with various concentrations of a-pinene (A), limonene (B), or a combination of a-pinene and limonene (C). R c control spore solution to which no terpenoids were added. *Significance at P < 0.001 (Kruskal-Wallis test) (SPSS, 1990). For reference, the average percent germination at 30% a-pinene (A), 2% limonene (B), and a combination of 30% a-pinene + 2% limonene (C) is repeated from Figure 1.

from that of limonene alone (after Bonferroni correction due to multiple comparisons, P ≤ 0.009, Mann-Whitney U test). Factors Affecting Termite Survival. A Cox proportional regression model based on 1600 individuals (800 N. nigriceps and 800 N. costalis) showed that exposure to spores was a significant and independent predictor of termite survival (Wald statistic c 73.2, df c 1, P c 0.001). After controlling for the effects of caste, species, and group composition, spore-exposed termites had a hazard ratio of death 1.2 times higher than that of control termites (P c 0.001, 95% confidence interval 1.2–1.3). Although this difference was statistically significant, the survival distributions of spore-exposed and control nasute soldiers and workers followed similar patterns (Figure 3A, 3B). In sharp contrast to the time course of survival of nasute termites, the survival distributions of spore-exposed

29


TABLE 1. AREA FREE OF FUNGAL GROWTH AFTER TREATMENTa Area free of fungal growth (mm2 , mean ± SD) Treatment a-Pinene Limonene a-Pinene + limonene Control P†

6 days after seeding

12 days after seeding

P‡

4.5 ± 2.6 (15) 2.1 ± 2.7 (15) 6.4 ± 3.8 (10) 1.9 ± 1.7 (20)

6.3 ± 4.0 (15) 3.1 ± 4.7 (15) 6.1 ± 5.4 (10) 3.1 ± 3.2 (20)

NS

***

***

NS NS NS

aThe

average (± SD) area free of fungal growth is shown after a-pinene-, limonene-, a-pinene + limonene- and Tween 80-treated filter paper was suspended over PDA plates seeded with a 2.1 × 107 spores/ ml concentration of the fungus M. anisopliae. The numbers in parenthesis indicate numbers of replicates. ***Significant differences at P < 0.001 by †Kruskal-Wallis test or ‡Mann-Whitney U test. NS indicates no significant differences.

and control C. formosanus soldiers and workers and their corresponding hazard ratios of death were dramatically different (Figure 3C). Therefore, Nasutitermes soldiers and workers appeared to be less susceptible to fungal infection by M. anisopliae than C. formosanus. Group composition was the most significant independent predictor of termite survival (Wald statistic 143.8, df c 2, P < 0.001). The social environment of termites influenced survival to a greater extent than exposure treatment (Wald statistics 143.8 vs. 73.2, respectively). After controlling for the effect of all other variables, the following trend of disease susceptibility was evident: worker monocaste groups < mixed-caste groups < soldier monocaste groups (Figure 4). Termites in mixed-caste groups had a hazard ratio of death 3.4 times higher than those in worker monocaste groups (Wald statistic c 115.9, df c 1, P < 0.001, 95% confidence interval 2.7–4.3). Soldier monocaste groups had 4.7 times the hazard ratio of death of worker monocaste groups (Wald statistic c 143.6 df c 1, P < 0.001, 95% confidence interval 3.6–6.1). Caste appeared to influence termite disease susceptibility, although it was not a predictor of survival (Wald statistic c 3.5, df c 1, P c 0.06). After controlling for the effect of all other variables, a trend toward higher soldier survival was observed: relative to workers, soldiers were 14% less susceptible to infection. Species was not a predictor of termite survival (Wald statistic c 1.3, df c 1, P c 0.24). Survivorship in Mixed-Caste Groups (15 Soldiers and 5 Workers). Nasute soldiers and workers had low survivorship independent of their contact with spores (Figure 3A, 3B). The LT50 values of soldiers and workers in control

30


AND

TRANIELLO

FIG. 3. Survival distributions for N. costalis and N. nigriceps in mixed-caste groups following an exposure to a 4.3 × 107 concentration of M. anisopliae conidia or a control solution (Tween 80 solution with no conidia). O—O c workers exposed to conidia; ˚—˚ c control unexposed workers; Q—Q c soldiers exposed to conidia; R—R c control unexposed soldiers. Significance of survival distributions is indicated in Table 1. For reference, the survival distributions of C. formosanus soldiers and workers exposed to spores or to a control suspension is provided. Note that soldiers and workers of C. formosanus exposed to spores had a hazard ratio of death 11.4 times higher than the controls (P < 0.001, 95% confidence interval c 7.8–16.6; Cox proportional regression). Caste in C. formosanus was not a significant predictor of termite survival (P c 0.8, 95% confidence interval c 0.8–1.4), and the survival distributions of spore-exposed soldiers and workers did not differ significantly (Breslow statistic c 0.1, df c 1, P c 0.9). The confirmation rates for exposed soldiers and workers was 80% and 64%, respectively. None of the control C. formosanus confirmed infection by M. anisopliae.

groups ranged from 3.0 to 4.0 days, similar to those of termites exposed to fungal spores (Table 2). Differences in the survival distributions of spore-exposed and control termites were only observed for workers of N. nigriceps (Table 2). Given the similarity in survival distributions between spore-exposed and control


31

FIG. 4. Survival distributions for Nasutitermes as a function of group composition. O—O c worker monocaste (groups of 5 workers); J—J c soldier monocaste (groups of 15 soldiers); Q—Q c mixed-caste groups (groups of 15 soldiers + 5 workers). Relative to the worker monocaste groups and after controlling for the effect of all other variables in the model, mixed-caste and soldier monocaste groups had 3.4 and 4.7 times the hazard ratio of death, respectively (P < 0.001, Cox proportional regression model).

nasute soldiers and workers (Figure 3A, 3B) and the results of the Cox proportional regression analysis, we conclude that both castes have equivalent susceptibility to infection by M. anisopliae, in spite of the presence of soldier frontal gland secretions, although soldiers appear to be somewhat less vulnerable (P c 0.06, Cox proportional regression). Survivorship in Worker Monocaste Groups. Nasutitermes workers in monocaste groups were the least susceptible to infection of all three social groupings (Figure 4). Workers nesting with soldiers did not appear to benefit from interacting with their chemically defended nestmates. On the contrary, the hazard ratio of death of mixed-caste groups was 3.4 times higher than worker monocaste groups. Nesting with soldiers thus had a survivorship cost for workers independent of whether termites were exposed to spores. For example, N. nigriceps and N. costalis workers in control monocaste groups had LT50 values of 5 and >10 days, respectively, whereas the LT50 values of N. nigriceps and N. costalis workers in control mixed-caste groups were 3 and 4 days, respectively (P < 0.001, Kaplan-Meier survival test). Survivorship in Soldier Monocaste Group. Soldier monocaste groups were


32

AND

TRANIELLO

TABLE 2. MEDIAN SURVIVAL TIMES (LT50 ), CONFIRMATION RATES OF FUNGAL INFECTION, AND SIGNIFICANCE OF SURVIVAL DISTRIBUTIONS OF Nasutitermes SOLDIERS AND WORKERS HOUSED IN MIXED-CASTE GROUPS FOLLOWING EXPOSURE TO 4.3 × 107 SPORES/ ml OR CONTROL SUSPENSION OF FUNGUS M. anisopliaea LT50

Soldier caste (5 worker, 15 soldier) N. nigriceps

N. costalis

P across species Worker caste (5 worker, 15 soldier) N. nigriceps

N. costalis

P across species

Percent confirmation

Exposed to 4.3 × 107 spores/ ml

Exposed to a control Tween 80 suspension

3.0 [2.8–3.2] (150) 4.0 [3.8–4.2] (150)

3.0 [2.7–3.3] (150) 4.0 [3.6–4.3] (150)

***

***

2.0 [1.7–2.3] (50) 3.0 [2.6–3.4] (50)

3.0 [2.1–3.9] (50) 4.0 [3.1–4.9] (50) NS

***

P within species

Of fungusinfected termites

Of Tween 80 control termites

NS

0

0

NS

0

0

***

0

0

NS

0

0

a***Denotes

overall significant differences in survival distributions (Breslow Statistic, Kaplan-Meier survival test, at P < 0.008 after a Bonferroni correction). The numbers in brackets represent the 95% confidence interval, the numbers in parentheses indicate the total number of individuals exposed to the spore or control suspensions. Within each row, ***indicates significant differences in the survival distributions of termites exposed to spores and the corresponding control. Within each column, ***indicates significance between the survival distributions between the two species. NS indicates no significant difference.

the most susceptible of all three social groupings (Figure 4). Relative to the mixed-caste groups, soldier monocaste groups had 1.4 times the hazard ratio of death (Wald statistic c 28.8, df c 1, P < 0.001; 95% confidence interval 1.2– 1.6). Confirmation Rates of Infection by M. anisopliae. There was a lack of confirmation of M. anisopliae infection for N. costalis and N. nigriceps when dead termites from the mixed-caste groups were plated (Table 2). This may have resulted from either the volatile effect of soldier terpenoids or the leaching of the secretion from the soldier’s head on the agar medium. The inhibitory effect of the soldier secretions on sporulation of worker carcasses in the mixed-caste groups


33

is further demonstrated by the comparatively high confirmation rates of workers in monocaste groups. Approximately 88.0% and 50.0% of dead N. nigriceps and N. costalis workers, respectively, plated in the absence of soldiers confirmed an infection of M. anisopliae. N. nigriceps and N. costalis soldiers in the monocaste group had 14.6% and 2% confirmation rates, respectively. These patterns by M. anisopliae across the three social groupings suggests that allogrooming by workers reduces the number of spores on the cuticle of living soldiers, and secretions from the frontal gland suppress sporulation of the fungus on dead termites.

DISCUSSION

The defensive secretions of nasute soldiers had a fungistatic effect in in vitro experiments. a-Pinene at concentrations similar to the minimum concentrations found in the frontal glands of Nasutitermes soldiers inhibited the germination of M. anisopliae spores without prolonged contact. Spore germination by a-pinene was reduced even at the relatively low concentration of 5%. Limonene, on the other hand, did not influence spore germination at 2%, 0.5%, and 0.05% concentrations. The presence of a-pinene, therefore, may explain in part the modest reduction in disease risk that soldiers appeared to have relative to workers. However, the frontal gland secretions of Nasutitermes may explain their lower susceptibility to fungal infection relative to C. formosanus soldiers (Figure 3) and other Coptotermes species (Milner and Staples, 1996), although we cannot rule out the possibility that other physiological or anatomical differences were responsible. Other observations support the notion that chemicals produced by soldiers of higher termite species may prevent the development of microorganisms in the nest. For example, large numbers of cadavers of N. infuscatus may be found in humid conditions without being invaded by fungi (Bouillon, 1970). The nest carton of many arboreal Nasutitermes species bears a strong odor of terpenoids, suggesting that the volatile components of nasute soldier secretions may impregnate nest material and render it unsuitable for fungal growth. Laboratory and field experiments in which M. anisopliae has been used as an agent for the biological control of N. exitosus Hill show that colonies do not necessarily succumb to infection, and many colonies in the field recover after fungal exposure (Ha¨ nel and Watson, 1983; Ljutikova, 1990; Milner and Staples, 1996). Workers collected from these colonies carried spores that failed to germinate even after 15 weeks from the original treatment (Ha¨ nel and Watson, 1983; Watson, 1990). The dormant state of spores on the termite hosts in these studies and the lower microbial loads isolated from the cuticle of nasute soldiers (Cruse, 1998) could reflect an antifungal role of the frontal gland secretions. The fungistatic effects of frontal gland secretions do not translate in a straightforward manner into higher survival for nasute soldiers. We hypothe-

34


AND

TRANIELLO

sized that if terpenoids reduced fungal susceptibility, monocaste worker groups would be the most susceptible to infection given their lack of defensive chemistry. Instead, our in vivo results suggest that mechanisms other than chemical protection are important in determining termite survival. Previous experiments have demonstrated that survival rates in Nasutitermes and Zootermopsis are positively correlated with group size (Lenz and Williams, 1980; Lenz and Barrett, 1984; Lenz, 1985; Rosengaus et al., 1998b), perhaps as a result of rates of social interaction (Williams and Perez Morales, 1980; Miramontes and DeSouza, 1996; Rosengaus et al., 1998b). In addition to group size, it appears that group composition is also an important determinant of Nasutitermes survival. Groups composed solely of workers, albeit smaller in size (N c 5), perhaps benefited from higher rates of allogrooming and/ or mutual feeding than when 25% of the colony members (N c 5 workers) were responsible for performing these tasks for the dependent soldiers (N c 15) that were incapable of allogrooming or feeding themselves due to their reduced mouthparts and head morphology (Moore, 1964). Thus, the low survival through time of termites in mixed-caste groups relative to worker monocaste groups reflects the extent to which soldiers tax the ability of workers to reduce disease risk through allogrooming and provide nutritional sustenance for soldiers through trophallaxis. However, the worker–soldier ratio in our experimental colonies was lower than that which characterizes Nasutitermes colony demography (Haverty, 1977). Thus, at natural worker–soldier ratios, workers may be capable of sufficient allogrooming to remove conidia, as well as feeding the dependent soldiers. The high susceptibility of soldier monocaste groups relative to the other social groupings was likely the combined result of soldier starvation and lack of social interactions, including allogrooming by workers. Differences in the rates of M. anisopliae confirmation as a function of group composition demonstrate the combined effects of defensive secretions and allogrooming on spore germination and/ or number of spores on the cuticle of nestmates. For example, no worker or soldier from the mixed-caste groups confirmed an infection of M. anisopliae (Table 2), but workers from the monocaste groups plated in the absence of soldiers had confirmation rates ranging from 50.0 to 88.0%. Dead soldiers from the monocaste group treatment confirmed an infection in 2.0–14.6% of cases. These values are higher than those of soldiers in the mixed-caste group, again suggesting that allogrooming by workers is necessary to reduce the number of spores on the cuticle that would otherwise remain on the soldiers and potentially sporulate after death. This interpretation is consistent with previous work demonstrating that allogrooming in nymphs and dealates of Z. angusticollis increases following contact with fungal spores, apparently reducing susceptibility to fungal infection (Rosengaus et al., 1998b). The dichotomy in defense strategies between the lower and higher termites appears to reflect the different predatory pressures encountered by these two


35

groups, given their different nesting and foraging biology (Coles and Howse, 1983; Traniello and Leuthold, 1999). The nesting and feeding habits of termites may also predispose lower and higher termites to encounter different pathogens. The high cuticular microbial loads of Mastotermes, together with the high density of microorganisms in the decayed wood and soil surrounding the nest, relative to those of Coptotermes and Nasutitermes (Cruse, 1998) suggest a correlation between microbial density and nesting/ feeding ecology. This correlation, however, cannot explain why Nasutitermes soldiers have lower cuticular microbial loads than worker nestmates (Cruse, 1998) or why soldiers of Odontotermes obesus are less susceptible to fungal infection than workers (Khan et al., 1993). We believe these differences may be explained in part by the presence of terpenoid-based frontal gland or quinone-based labial gland secretions of soldiers. The use of antibiotics is widespread in the social Hymenoptera. Terpenes present in honeybee propolis appear to reduce bacterial growth (Michener, 1974; Bailey and Ball, 1991), and secretions from the metapleural, Dufour’s, and venom glands, as well as rectal fluids, salivary secretions, and body exudates have fungistatic, bacteriostatic, and nematocidal properties (Lavie, 1960a–c; Maschwitz et al., 1970; Maschwitz and Tho, 1974; Michener, 1974; Vander Meer, 1983; Ho¨ lldobler and Engel-Siegel, 1984; Beattie et al., 1985, 1986; Obin and Vander Meer, 1985; Kermarrec et al., 1986b; Attygalle et al., 1989; Bailey and Ball, 1991; Veal et al., 1992; Gambino, 1993; Knapp et al., 1994; DoNascimiento et al., 1996). Evidence of the presence of glands devoted to the production of antibiotic secretions in termites is relatively sparse in spite of intense histological studies (Noirot, 1969). Still, termites appear to chemically control the development of pathogens within their nests by using their own secretions (Batra and Batra, 1966, 1979; Sannasi and Sundara Rajulu, 1967; Rich, 1969; Maschwitz and Tho, 1974; Olagbemiro et al., 1988; Ljutikova, 1990; Chen et al., 1998; Rosengaus et al., 1998a) or metabolites produced by microorganisms inhabiting the nest (Clerk, 1969; Watson and Ford, 1972; Maschwitz, 1974; Kermarrec et al., 1986a,b; Keller and Zimmermann, 1989; Wood and Thomas, 1989; Dykes, 1995). Although defensive secretions in termites evolved as a result of predation, defensive chemicals may have attained a secondary antimicrobial function. Our results suggest that while the frontal gland secretion of nasute soldiers has fungistatic properties, group demography and allogrooming behavior also play a role in disease reduction. Acknowledgments—We thank Mary Lesniak and Dr. Samuel Nurko for assistance, Dr. Gregg Henderson (LSU Agricultural Center, Baton Rouge, Louisiana) kindly provided us with a portion of a Coptotermes formosanus colony. The comments and suggestions of two anonymous reviewers were greatly appreciated. This research was funded by NSF grant IBN-9632134 (J. F. A. Traniello and R. D. Karp, sponsors).

36


AND

TRANIELLO

REFERENCES ATTYGALLE, A. B., SIEGEL, B., VOSTROWSKY, O., BESTMANN, H. J., and MASCHWITZ, U. 1989. Chemical composition and function of metapleural gland secretion of the ant Crematogaster deformis Smith (Hymenoptera: Myrmicinae). J. Chem. Ecol. 15:317–327. BAILEY, L., and BALL, B. V. 1991. Honeybee Pathology. Academic Press, San Diego. BATRA, L. R., and BATRA, S. W. T. 1966. Fungus-growing termites of tropical India and associated fungi. J. Kans. Entomol. Soc. 39:725–738. BATRA, L. R., and BATRA, S. W. 1979. Termite-fungus mutualism, pp. 117–163, in L. R. Batra (ed.). Insect–Fungus Symbiosis, Nutrition, Mutualisms, and Commensalism. Wiley, New York. BEATTIE, A. J., TURNBULL, C., HOUGH, T., JOBSON, S., and KNOX, R. B. 1985. The vulnerability of pollen and fungal spores to ant secretions: Evidence and some evolutionary implications. Am. J. Bot. 72:606–614. BEATTIE, A. J., TURNBULL, C., HOUGH, T., JOBSON, S., and KNOX, R. B. 1986. Antibiotic production: A possible function of the metapleural glands of ants (Hymenoptera: Formicidae). Ann. Entomol. Soc. Am. 79:448–450. BOUILLON, A. 1970. Termites of the Ethiopian region, pp. 154–280, in K. Krishna and F. M. Weesner (eds.). Biology of Termites. Academic Press, New York. CHEN, J., HENDERSON, G., GRIMMS, C. C., LLOYD, S. W., and LAINE, R. A. 1998. Termites fumigate their nests with naphthalene. Nature 392:558–559. CLERK, G. C. 1969. Influence of soil extracts on the germination of conidia of the fungi Beauveria bassiana and Paecilomyces farinosus. J. Invrtebr. Pathol. 13:120–124. COLES, H. R., and HOWSE, P. E. 1983. Chemical defence in termites: Ecological aspects, pp. 21–29, in P. Jaisson (ed.). Social Insects in the Tropics. Universite´ Paris-Nord, Cocoyoc, Me´ xico. CRUSE, A. 1998. Termite defences against microbial pathogens. PhD thesis. Macquarie University, Australia. 175 pp. DELIGNE, J. 1971. Mecanique du comportament de combat chez les soldats de termites (Insectes Isopte` res). Forma Functio 4:176–187. DO NASCIMIENTO, R. R., SHOETERS, E., MORGAN, E. D., BILLEN, J., and STRADLING, D. J. 1996. Chemistry of metapleural gland secretions of three attine ants, Atta sexdens rubropilosa, Atta cephalotes and Acromyrmes octospinosus (Hymenoptera: Formicidae). J. Chem. Ecol. 22:987–1000. DYKES, G. A. 1995. Bacteriocins: Ecological and evolutionary significance. Tree 10:186–189. EISNER, T., KRISTON, I., and ANESHANSLEY, D. J. 1976. Defensive behavior of a termite (Nasutitermes exitosus). Behav. Ecol. Sociobiol. 1:83–125. GAMBINO, P. 1993. Antibiotic activity of larval saliva of Vespula wasps. J. Invertebr. Pathol. 61:110. GRASSE´ , P. P. 1982. Termitologia. Anatomie–Physiologie–Reproduction des Termites. Masson, Paris. ¨ HANEL , H., and WATSON, J. A. L. 1983. Preliminary field tests on the use of Metarhizium anisopliae for the control of Nasutitermes exitosus (Hill) (Isoptera: Termitidae). Bull. Entomol. Res. 73:305–313. HAVERTY, M. I. 1977. The proportion of soldiers in termite colonies. A list and a bibliography. Sociobiology 2:199–216. ¨ , B., and ENGEL-SIEGEL, H. 1984. On the metapleural gland of ants. Psyche 91:201–224. HOLLDOBLER KELLER, S., and ZIMMERMANN, G. 1989. Mycopathogens of soil insects, pp. 239–270, in N. Wilding, N. M. Collins, P. M. Hammond, and J. F. Webber, (eds.). Insect–Fungus Interactions. Academic Press, San Diego. KERMARREC, A., DECHARME, M., and FEBVAY, G. 1986a. Leaf-cutting ant symbiotic fungi: A synthesis of recent research, pp. 231–266, in C. S. Lofgren and R. K. Vander Meer, (eds.). Fire Ants and Leaf-cutting Ants. Westview Press, Boulder, Colorado.


37

KERMARREC, A., FEBVAY, G., and DECHARME, M. 1986b. Protection of leaf-cutting ants from biohazards: Is there a future for microbiological control?, pp. 339–356, in C. S. Lofgren and R. K. Vander Meer, (eds.). Fire Ants and Leaf-cutting Ants. Westview Press, Boulder, Colorado. KHAN, H., JAYARAJ, S., and GOPALAN, M. 1993. Muscardine fungi for the biological control of agroforestry termite Odontotermes obesus (Rambur). Insect Sci. Appl. 14:529–535. KNAPP, J. J., JACKSON, C. W., HOWSE, P. E., and VILELA, E. F. 1994. Mandibular gland secretions of leaf-cutting ants: role in defence against alien fungi, p. 109, in A. Lenoir, G. Arnold, and M. LePage, (eds.). Les Insectes Sociaux. Universite´ Paris Nord, Paris. KO, W. H., FUJII, J. K., and KANEGAWA, K. M. 1982. The nature of soil pernicious to Coptotermes formosanus. J. Invertebr. Pathol. 39:38–40. LAVIE, P. 1960a. Les antibiotique chez les insectes sociaux. Ann. Abeille 3:270–276. LAVIE, P. 1960b. Les substances antibactriennes dans colonie d’abeilles (Apis mellifica) (fin). Ann. Abeille 3:201–306. LAVIE, P. 1960c. Les substances antibactriennes dans colonie d’abeilles Apis mellifica) (Ire partie). Ann. Abeille 3:103–184. LENZ, M. 1985. Variability of vigour between colonies of Coptotermes acinaciformis (Froggatt) (Isoptera: Rhinotermitidae) and its implications for laboratory experimentation. Bull. Entomol. Res. 75:13–21. LENZ, M., and BARRETT, R. A. 1984. Implications for comparability of laboratory experiments revealed in studies on the effect of population density on vigoour in Coptotermes lacteus (Froggatt) and Nasutitermes exitosus (Hill) (Isoptera: Rhinotermitidae and Termitidae). Bull. Entomol. Res. 74:477–485. LENZ, M., and WILLIAMS, E. R. 1980. Influence of container, matrix volume and group size on survival and feeding activity in species of Coptotermes and Nasutitermes (Isoptera: Rhinotermitidae, Termitidae). Mater. Org. 15:25–46. LJUTIKOVA, L. I. 1990. Entomogenous fungi as control agents for termite Anacanthotermes ahngerianus Jacobson (Isoptera: Hodotermitidae), pp. 634–635, in G. K. Veeresh, B. Mallik, and C. A. Viraktamath, (eds.). Social Insects and the Environment. Oxford and IBH Publishing Company, New Delhi. LJUTIKOVA, L. I., and JUDINA, T. G. 1996. Antibacterial activity of some exudates of Anacanthotermes ahngerianus Jacobson (Isoptera: Hodotermitidae), pp. 13–105, in XX International Congress of Entomology, Firenze, Italy. LUTIKOVA, L. I. 1990. The influence of soil substrates from the Anacanthotermes ahngerianus Jac. Termite-house on the development of enthomorathogenic fungi. Mycol. Physiol. 24:520–528 (in Russian). MASCHWITZ, U. 1974. Vergleichende Untersuchungen zur Funktion der Ameisenmetathorakald Ruse. Oecologia 16:303–310. MASCHWITZ, U., KOOB, K., and SCHILDKNECHT, H. 1970. Ein Beitrag zur Funktion der Metathoracaldruse der Ameisen. J. Insect Physiol. 16:387–404. MASCHWITZ, U., and THO, Y. P. 1974. Chinone als Wehrsubstanzen bei Einigen Orientalische Macrotermitinen. Insectes Soc. 21:231–234. MICHENER, C. D. 1974. The Social Behavior of Bees. A Comparative Study. Belknap Press, Cambridge, MA, USA. MILNER, R. J., and STAPLES, J. A. 1996. Biological control of termites: Results and experiences within a CSIRO project in Australia. Biocontrib. Sci. Tech. 6:3–9. MILNER, R. J., STAPLES, J. A., HARTLEY, T. R., LUTTON, G. G., DRIVER, F., and WATSON, J. A. L. 1998. Occurrence of Metarhizium anisopliae in nests and feeding sites of Australian termites. Mycol. Res. 102:216–220. MIRAMONTES, O., and DESOUZA, O. 1996. The nonlinear dynamics of survival and social facilitation in termites. J. Theor. Biol. 181:373–380.

38


AND

TRANIELLO

MOORE, B. P. 1964. Volatile terpenes from Nasutitermes soldiers (Isoptera: Termitidae). J. Insect Physiol. 10:371–375. MOORE, B. P. 1968. Studies on the chemical composition and function of the cephalic gland secretion in Australian termites. J. Insect Physiol. 14:33–39. NOIROT, C. 1969. Glands and secretions, pp. 89–123, in K. Krishna and F. M. Weesner, (eds.). Biology of Termites. Academic Press, New York. OBIN, S. M., and VANDER MEER, R. K. 1985. Gaster flagging by fire ants (Solenopsis spp.): Functional significance of venom dispersal behavior. J. Chem. Ecol. 11:1757–1768. OLAGBEMIRO, T. O., LAJIDE, L., SANI, K. M., and STADDON, B. W. 1988. 2-Hydroxy-5-methyl1,4-benzoquinone from the salivary gland of the soldier termites Odontotermes magdalenae. Experientia 44:1022–1025. PRESTWICH, G. D. 1983. Chemical systematics of termite exocrine secretions. Annu. Rev. Ecol. Syst. 14:287–311. PRESTWICH, G. D. 1984. Defense mechanisms of termites. Annu. Rev. Entomol. 29:201–232. QUENNEDEY, A. 1975. La guerre chimique chez les termites. Recherche 54:274–276. RICH, S. 1969. Quinones, pp. 647–648, in D. C. Torgeson (ed.). Fungicides: An Advanced Treatise. Academic Press, New York. ROSENGAUS, R. B., and TRANIELLO, J. F. A. 1997. Pathobiology and disease transmission in dampwood termites [Zootermopsis angusticollis (Isoptera: Termopsidae)] infected with the fungus Metarhizium anisopliae (Deuteromycotina: Hypomycetes). Sociobiology 30:185–195. ROSENGAUS, R. B., GULDEN, M. R., and TRANIELLO, J. F. A. 1998a. Inhibitory effect of termite fecal pellets on fungal spore germination. J. Chem. Ecol. 24:1697–1706. ROSENGAUS, R. B., MAXMEN, A. B., COATES, L. E., and TRANIELLO, J. F. A. 1998b. Disease resistance: A benefit of sociality in the dampwood termite Zootermopsis angusticollis (Isoptera: Termopsidae). Behav. Ecol. Sociobiol. 44:125–134. SANDS, W. L. 1969. The association of termite and fungi, pp. 495–524, in K. Krishna and F. M. Weesner (eds.). Biology of Termites. Academic Press, New York. SANNASI, A., and SUNDARA RAJULU, G. 1967. Occurrence of antimicrobial substance in the exudate of physogastric queen termites, Termes redemanni Wasmann. Curr. Sci. 16:436–437. SMITH, S. J., WAIWOOD, K. G., and NEILSON, J. D. 1994. Survival analysis for size regulation of Atlantic halibut, pp. 125–144, in N. Lange, L. Ryan, and L. Billard, (eds.). Case Studies in Biometry. Wiley, New York. SPSS. 1990. SPSS/ PC+ 4.0 Advanced Statistics Manual. Chicago. TOUMANOFF, C., and ROMBAUT, J. 1965. Action de certains champignons entomophages, cultives sur les milieux appropries attractifs sur le termite de Saintonge, Reticulitermes santonensis Feytaud. Ann. Parasitol. Hum. Comp. 40:605–609. TRANIELLO, J. F. A. 1981. Enemy-deterrence in the recruitment strategy of a termite: Soldier-organized foraging in Nasutitermes costalis. Proc. Natl. Acad. Sci. U.S.A. 78:1976–1979. TRANIELLO, J. F. A., and LEUTHOLD, R. 2000. The behavioral ecology of foraging in termites, in T. Abe, T. Higashi and D. Bignell, (eds.). Termites: Evolution, Sociality, Symbiosis, Ecology. Kluwer, The Hague. VANDER MEER, R. K. 1983. Semiochemicals and the red imported fire ant (Solenopsis invicta Buren) (Hymenoptera: Formicidae). Fla. Entomol. 66:139–161. VEAL, D. A., TRIMBLE, J. E., and BEATTIE, A. J. 1992. Antimicrobial properties of secretions from the metapleural glands of Myrmecia gulosa (the Australian bull ant). J. Appl. Bacteriol. 72:188–194. WATSON, J. A. L. 1990. Biological control of termites in Australia, pp. 632–633, in G. K. Veeresh, B. Mallik, and C. A. Viraktamath, (eds.). Social Insects and the Environment. Oxford and IBH Publishing Company, New Delhi.


39

WATSON, A. G., and FORD, E. J. 1972. Soil fungistasis—a reappraisal. Annu. Rev. Phytopathol. 10:327–348. WILLIAMS, R. M. C., and PEREZ MORALES, J. V. 1980. The effect of group size on the survival and feeding economy of pseudoworkers of building damaging Cryptotermes spp. (Isoptera: Kalotermitidae), pp. 219–234, in P. Jaisson (ed.). Social Insects in the Tropics. Universite´ ParisNord, Cocoyoc, Me´ xico. WOOD, T. G., and THOMAS, R. J. 1989. The mutualistic association between Macrotermitinae and Termitomyces, pp. 69–92, in N. Wilding, N. M. Collins, P. M. Hammond, and J. F. Webber (eds.). Insect–Fungus Interactions. Academic Press, San Diego. ZOBERI, M. H. 1995. Metarhizium anisopliae, a fungal pathogen of Reticulitermes flavipes (Isoptera: Rhinotermitidae). Mycologia 87:354–359.