condensed tannins, attine ants, and the performance of a ......inga spp. produce high concentrations...

Journal of Chemical Ecology, Vol. 17, No. 6, 1991

CONDENSED TANNINS, ATTINE ANTS, AND THE PERFORMANCE OF A SYMBIOTIC FUNGUS

C O L I N N I C H O L S - O R I A N S 1'2

fDepartment of Entomology Pennsylvania State University

University Park, Pennsylvania 16802

(Received November 26, 1990; accepted February 13, 1991)

Abstract--Field experiments indicate that the foliar concentration of condensed tannin affects the selection of leaf material oflnga oerstediana Benth., a tropical legume tree, by leaf cutter ants. In one study an increase in tannin concentration was correlated with a decrease in the acceptability of leaves to leaf-cutter ants, except at low tannin concentrations. Protein concentration was not correlated with acceptability nor was the ratio of protein to tannin. Results from a second study suggest that when the concentration of tannin was low the ants appear to select leaves on the basis of nutrient availability. Laboratory assays with the ants indicated that quebracho tannin, a commercially available condensed tannin, inhibits foraging ants. Again, at lower concentrations, quebracho tannin appeared to have little affect on the ants.The fungus the ants cultivate is a wood-rotting Basidiomycete that produces enzymes, such as polyphenol oxidase (PPO), that are capable of inactivating tannins. The activity of these PPOs may explain why leaf-cutter ants are undeterred by low concentrations of condensed tannins. I hypothesized that PPO activity would be absent from fungal cultures without tannin and that only high concentrations of tannin would inhibit the fungus. Cultures with and without tannin showed similar PPO activity. Thus PPO activity is constitutive. In fact, as fungal biomass increased, so did PPO activity. As hypothesized, only high concentrations of quebracho tannin inhibited PPO activity and fungal growth. However, it is not clear whether the ants can discriminate between concentrations that do and do not inhibit the fungus.

Key Words--Atta cephalotes, attine ants, Hymenoptera, Formicidae, Inga oerstediana, plant-herbivore interactions, host selection, condensed tannins, fungal performance, polyphenol oxidase.

2Present address: Department of Biology, Vassar College, Poughkeepsie, New York 12601.

1177

0098-0331/91/0600-1177506.50/0 �9 1991 Plenum Publishing Corporation

1178 NICHOLS-ORIANS

INTRODUCTION

The concentrations of nutrients and phenolics typically differ in leaves of plants growing under different environmental conditions and in leaves of different ages, and these chemical traits may influence patterns of herbivory (Fox and Ma- cauley, 1977; Chandler and Goosem, 1982; Coley, 1983a,b; Waterman et al., 1984; Meyer and Montgomery, 1987; Mole and Waterman, 1988; Denslow et al., 1990; Nichols-Orians and Schultz, 1990). Tannins are one class of phenolics that may influence selection of leaves by herbivores (Waltz, 1984; Mole and Waterman, 1988; Schultz, 1989). However, Wint (1983) and Mole and Waterman (1988) suggest that selection would not depend upon the concentration of nutrients or tannins, but upon the ratio of the two.

Leaf-cutter ants exhibit strong intraspecific preferences for host foliage (Fennah, 1950; Cherrett, 1968, 1972; Rockwood, 1976, 1977; Nichols-Orians and Schultz, 1990), and this appears to be related to the availability of nutrients and the concentrations of secondary chemicals (Barrer and Cherrett, 1972; Howard, 1990; Nichols-Orians, 1991a,b). Although terpenoids are widely rec- ognized for their ability to deter leaf-cutter ants (Hubbell and Wiemer, 1983), some plant species appear to contain tannins in sufficient quantities to deter foraging ants (Howard, 1990; Nichols-Orians and Schultz, 1990).

Leaves of sun plants growing in nutrient-poor soils are generally less sus- ceptible to leaf-cutter ants than leaves of plants growing in nutrient-rich soils (Nichols-Orians, 1991b; R. Marquis, personal communication). This may be because these growth conditions result in elevated concentrations in some plant species of carbon-based secondary compounds like tannins (Chandler and Goosem, 1982; Nichols-Orians, 1991b). However, leaves of plants growing in the sun and in nutrient-rich soil often have higher concentrations of both condensed tannins and foliar nutrients than the leaves of plants growing in the forest understory (Field and Mooney, 1986; Denslow et al., 1990; Nichols-Orians, 1991a), and these sun leaves have been found to be more acceptable to leaf- cutter ants than understory leaves (Nichols-Orians, 1991a; R. Marquis, personal communication). Thus, moderate concentrations of condensed tannins do not appear to deter foraging ants.

In addition to environmentally based differences in chemistry and acceptability, there may be phenological differences as well. Young leaves usually contain higher concentrations of nutrients and usually are more acceptable to leaf-cutter ants than mature leaves (Barrer and Cherrett, 1972; Waller, 1982). However, in one study (Nichols-Orians and Schultz, 1990), we found young leaves to be less acceptable chemically to leaf-cutter ants than mature leaves, despite higher concentrations of nutrients. These young leaves had much higher concentrations of condensed tannins. Young leaves of neotropical trees often have higher concentrations of tannins and other phenolics than mature leaves (Coley, 1983a,b), which suggests that leaf-cutter ants often may find young

TANNINS, LEAF-CUTTER ANTS, AND FUNGI | 179

leaves less acceptable. Recently, Coley and Aide (1989) found that as anthocyanin concentration increases leaves are less acceptable to leaf-cutter ants. Anthocyanin is structurally similar to tannins and, like tannins, has antibiotic properties. All these studies suggest that the chemical acceptability of leaves to leaf-cutter ants may depend upon both the concentrations of nutrients and tannins and perhaps upon the ratio of the two.

Here, I evaluate trade-offs in host plant acceptability with respect to levels of foliar protein and condensed tannin. I divided this study into two parts. First, using leaves from plants growing in different light and soil nutrient conditions, I investigated how quantitative differences in the concentrations of protein and condensed tannins influence the selection of leaves by leaf-cutter ants, and compare these results to those of previous studies on this system (Nichols-Orians and Schultz, 1990; Nichols-Orians, 1991a,b). Second, I determined how vary- ing the concentration of a commercially available condensed tannin, quebracho tannin, influences selection by the ants and performance of the symbiotic fungus the ants cultivate. The first part of the study was done with field colonies in Costa Rica and the second part with laboratory colonies at Pennsylvania State University.

The symbiotic fungus is a Basidiomycete fungus that produces enzymes, polyphenol oxidases and tannases, which are capable of inactivating tannins via polymerization (Cherrett et al., 1989). The ability of the fungus to inactivate tannins may explain why leaf-cutter ants appear insensitive to relatively low concentrations of tannin (Nichols-Orians, 1991 a). As with many fungal enzymes (Cooper and Wood, 1973, 1975), polyphenol oxidases (PPOs) can be induced by the presence of certain chemicals, such as gallic acid, a precursor of tannin (Mayer, 1987). The PPOs produced by the ants' fungus may be induced by the presence of condensed tannin and, once produced, polymerize and inactivate the tannins. Therefore, I hypothesized that PPO activity would only be present in fungal cultures containing tannin. Furthermore, I predicted that low concentrations of tannin would be quickly inactivated and so would not inhibit PPO activity or fungal growth. Since high concentrations of tannin may inhibit PPOs (Lyr, 1962), fungal growth may be inhibited as well. Because PPOs are produced extracellularly, as evident from the formation of colored oxidative prod- ucts following addition of fungus to media containing tannins and other phenolics (Bavendamm's test) (Kirk and Kelman, 1965; unpublished data), it is easy to quantify PPO activity.

M E T H O D S A N D M A T E R I A L S

Site

The study was conducted at the Organization for Tropical Studies' La Selva Biological Station (10~ 83~ in the Atlantic lowland near Puerto Viejo de Sarapiqui, Costa Rica. This forest is a tropical wet premontane forest

1180 NICHOLS-ORIANS

(Holdridge et al., 1971) and receives a mean annual rainfall of 4000 mm (Hart- shorn, 1983), with a short dry season from late January to April (La Selva, unpublished records). The field portion of this study was conducted from Jan- uary to July 1987.

Laboratory assays with the ants and the fungus were conducted in 1989- 1990 with colonies maintained at Pennsylvania State University. Colonies were collected from La Selva in January 1987 and maintained on a diet of frozen Forsythia and fresh cabbage during the experiments.

Leaf-Cutter Ants~Fungus

Colonies of a leaf-cutter ant, Atta cephalotes (L.) (Hymenoptera: Formi- cidae: Attini), are abundant within La Selva and naturally encounter plants growing under different environmental conditions (personal observation). Leaf- cutter ants use the harvested leaf material to support the growth of a symbiotic fungus, which is the sole food source of the developing larvae (Quinlan and Cherrett, 1979). Hydrolyzable tannins have been shown to inhibit the growth of an attine fungus (Seaman, 1984), but condensed tannins appear to be even more effective inhibitors of the fungi (Cherrett et al., 1989). This result is consistent with the observation that condensed tannins are very effective inhibitors of fungi and their enzymes (Zucker, 1983). Therefore leaf-cutter ants would be expected to avoid leaves with tannin levels sufficient to inhibit the fungus. If tannin levels are low, then nutrients would be expected to dictate acceptability (Berish, 1986; Waltz, 1984).

Plant Species

The tropical legume tree, Inga oerstediana Benth. [formerly Inga edulis var. minutula Schery, (Nichols-Orians and Schultz, 1990)] (Fabaceae: Mimo- soideae) is naturally attacked by leaf-cutter ants (personal observation). Yet lnga is regarded as being relatively resistant to leaf-cutter ants (Arckoll, 1984). Inga spp. produce high concentrations of tannins (Koptur, 1985; Nichols-Ori- ans, 1990, 1991a,b), so the low susceptibility of Inga to leaf-cutter ants may be due to these tannins.

Plant Treatments

In a previous study I showed that light and soil nutrient conditions influence foliar chemistry and the acceptability of leaves to leaf-cutter ants (Nichols- Orians, 1991b). For a subset of the plants used in the above study (N = 48), I measured foliar chemistry and conducted bioassays with the leaves to determine their acceptability to the ants. This study evaluates the relationship between chemistry and acceptability.

TANNINS, LEAF-CUTTER ANTS, AND FUNGI 1 18 ]

The following is a brief description of the growth conditions. Initially, field-collected seedlings were grown in pots containing either nutrient-rich allu- vial or nutrient-poor residual soils (Vitousek and Denslow, 1987) at 20%-of- full-sunlight. After 13 weeks half the seedlings were transferred to 2 %-of-full- sunlight and grown for an additional seven weeks. At the same time I altered light conditions, I initiated fertilization treatments. Thus, there were eight treatments with six seedlings in each treatment (Figure 1). I noted when leaves matured throughout the experiment so that at the end of the 20 weeks two different leaf types were harvested from each plant: (1) those that had matured during the first 13 weeks (PRE leaves) and (2) those that developed and matured during the last seven weeks (POST leaves).

Leaf Material

Both PRE and POST leaves were collected seven weeks (_+4.5 days) after light and nutrient manipulation. Seedlings were first transported in their pots to the study colony and a minimal amount of leaf material was used during the bioassays with the ants. L oerstediana leaves are compound. Bioassays to determine how leaves differed in acceptability were conducted with disks from the distal leaflets only in order to avoid confounding leaflet position effects (Gall, 1987). Although I have not detected position effects on acceptability (personal observation), Gall (1987) found that basal leaflets in a number of Juglandaceae species were less acceptable to various Catocala larvae (Noctuidae) than either lateral or terminal leaflets.

Once bioassays were complete (approximately 30 min), all remaining leaf material was removed, immediately frozen, freeze-dried, and ground to a powder, with a UDY Cyclone Mill, at La Selva before being transported back to Pennsylvania State University for chemical analyses. Because of the large amount of leaf material required for chemical analyses, chemistry was deter-

Soit Type: A l luv ia l Residual

Light : 20% 20%

13 weeks

2% 20% 2% 20%

F NF F NF F NF F NF

FIG. 1. Schematic diagram of treatments. F = fertilized, NF = not fertilized.

1182 NICHOLS-ORIANS

mined from the analysis of entire leaves minus the petiole. Although removal of leaf material during the bioassays could conceivably cause a change in chemistry, I found no change in chemistry 24 hr after damaging L oerstediana seedlings (Nichols-Orians, 1991a).

Unfortunately there was not enough leaf material from some plants to conduct bioassays and do chemical analyses. Therefore, the correlation between chemistry and acceptability was done on a subset of leaves (N = 37). Even with the reduced sample size, there was extensive variability in leaf chemistry. In fact, the range of values reported here very nearly expand the full range measured previously in L oerstediana leaves (Nichols-Orians, 1990, 1991a,b).

Chemical Analyses

For analysis of protein concentration, I extracted approximately 20 mg of leaf powder in 10 ml of 0.1 N NaOH for 2 hr in a boiling water bath and used Bio-Rad reagent to quantify protein content (Snyder and Desborough, 1978; Compton and Jones, 1985). This extraction technique is designed to measure the concentration of protein in leaves high in tannin (Jones et al., 1989).

For the analysis of the various phenolic traits, leaf powder (300-400 rag) was washed with ether for 30 min. Phenolics were extracted with 70% acetone for 3 hr at 40~ The acetone was removed under reduced pressure, and the extracts were diluted to 10 ml with distilled water. Extracts were analyzed for proanthocyanidin condensed tannins (butanol-HC1 method; Bate-Smith, 1977), leucoanthocyanin condensed tannins (vanillin method; Broadhurst and Jones, 1978; Butler et al., 1982), and protein-binding capacity using hemoglobin as the substrate (Schultz et al., 1981). Relative concentrations are expressed as wattle tannin (Acacia sp. from Leon Monnier, Inc., Peabody, Massachusetts) equivalents per milligram of dry weight (% WTE) for the two condensed tannins and as tannic acid (Sigma, Lot No. 11F-0559) equivalents per milligram of dry weight (%TAE) for protein-binding capacity. I also measured total phenolics (Folin-Denis assay; Swain and Hillis, 1959), but because they were over 90% correlated with the two condensed tannins I am restricting my discussion to the two tannins and protein-binding capacity. Thus far, surveys indicate that tannins are the only secondary compounds in Inga (Koptur, 1985), and only condensed tannins are abundant in L oerstediana (unpublished data).

Bioassays with Leaves

All bioassays were done with a single colony. Previous work has shown that the relative acceptability of leaf types of L oerstediana to leaf-cutter ants is similar among ant colonies (Nichols-Orians and Schultz, 1990; Nichols-Ori- ans, 1991a).

I conducted the "pickup" assay to assess the chemical acceptability of

TANNINS, LEAF-CUTTER ANTS, AND FUNGI 1183

each leaf type to leaf-cutter ants (Howard, 1987; Nichols-Orians and Schultz, 1990). The pickup assay is designed to determine preferences based on chemical differences, defensive or nutritional, between leaf types (Howard, 1987). Two leaf disks produced with a standard paper punch, were placed beside two disks of a highly preferred control, Hamelia patens (Rubiaceae). When one disk was removed it was replaced by a like disk. To ensure that the assays with the various leaf types were independent, I separated them by 9 in. and randomized the order of presentation each day. The chemical acceptability of a given leaf type was expressed as the number of leaf disks removed when 20 control disks had been removed. The final acceptability of each leaf type was the average number of leaf disks removed during two replicate trials.

The use of a control indicated whether ants were willing to pick up leaf disks on any given day and standardized for fluctuations in ant activity during each assay replication. H. patens was used as the control because of its high acceptability to leaf-cutter ants and because numerous plants can be found growing under similar conditions. It is abundant in a clearing near the study colony. The use of H. patens as a control yields very consistent results (Nichols- Orians and Schultz, 1990; Nichols-Orians 1991a,b).

Physical features of a leaf (e.g., trichomes) could influence disk removal, but I did not identify any such physical features in L oerstediana. Conceivably, leaf mass could influence disk removal as well. However, the leaf disks were small enough that all foraging ants could easily pick up these leaf disks. In fact, the toughest and heaviest leaves assayed to date were highly acceptable to the ants (Nichols-Orians and Schultz, 1990). Thus, leaf mass differences among these treatments would not have influenced the results.

Researchers often employ a "cut t ing" assay when testing whether physical features interfere with the harvesting of leaves by leaf-cutter ants (Howard, 1988; Nichols-Orians and Schultz, 1989, 1990). I did not employ this assay because I was only interested in determining the relationship between nutrients and tannins on the initial selection of leaves by leaf-cutter ants.

Laboratory Bioassays

Bioassays with Ants. I determined whether quebracho tannin influences foraging ants using two laboratory-maintained colonies (collected from La Selva, Costa Rica, in January 1987). Quebracho tannin was obtained from the Van Dyke Supply Co. (Woonsocket, South Dakota). Seven different tannin treatments (concentrations) were tested: 10, 5, 2.5, 1, 0.5, 0.25 and 0.1% (w/v). Distilled water was used at all times. Organic rye (1 g) flakes were soaked for 2 min in 2 ml of solution (one of the seven tannin treatments or a water only control). The flakes were drained and dried at 25~ overnight.

During each assay, 50 treatment flakes and 50 control flakes were placed

1184 NICHOLS-ORIANS

on a glass plate above a randomly generated grid on a foraging platform (design similar to Hubbell and Wiemer, 1983). Treatment and control flakes were placed on one of two patterned boxes. Ants were allowed access to the foraging platform and the number of control and treatment flake types remaining was determined every 2 min. The assay was terminated when the ants removed 25 of either flake type. After each run the access holes were plugged and the ants removed before initiating another assay. Because control flakes often were placed during subsequent assays where treatment flakes had been, and visa versa, the glass plate was wiped with acetone to prevent pheromone carry-over. The acceptability of each treatment is expressed as the number of treatment flakes removed when 25 control flakes had been removed. (Occasionally 25 treatment flakes were removed first. For these assays the number treatment flakes/25 control flakes was calculated algebraically.)

All treatments were tested on both colonies on the same day. The order in which the different concentrations were tested was randomized each day for each colony. Also treatment flakes were randomly assigned to one of the two patterned boxes for each assay. In all, there were five replicate assays for each colony. On the first day, only three treatments (10%, 1%, and 0.1%) were tested.

Fungal Bioassays. The fungus from one of the two colonies tested above was isolated, maintained on potato dextrose agar (PDA) plates, transferred to Sabaroud's dextrose broth (SDB), and later grown in a liquid medium designed to test for the effects of tannins on fungal performance (Seaman, 1984). This FPT medium contained the following components per liter of distilled water: (1) major components (g)--KH2PO 4 (1), MgSO 4 �9 7H20 (0.5), NaC1 (0.1), CaC12 �9 2H20 (0.1), NH 4 tartrate (5.0), and dextrose (20); (2) minor components (mg)--ZnSO4 �9 4H20 (4), MnSO4 �9 4H20 (4), and ferric tartrate (4); (3) organic supplement (mg)--4-methyl-5-thiazole ethanol (0.180), and 4-amino- 5-aminomethyl-2-methylpyrimidine dihydrochloride (0.250). In these experiments, 500 ml of distilled water was replaced with 500 ml of 0.05 M dimethyl glutaric acid buffer (pH 5.5) (Powell and Stradling, 1986).

Three basic in vitro experiments were conducted. First, at each tannin concentration I determined the effect of tannin on polyphenol oxidase (PPO) activity. Second, I measured changes in PPO activity in liquid cultures of fungi incubated with quebracho tannin for different lengths of time (10, 30, 60, or 120 min, and 2, 7, or 21 days). In a separate experiment, there were three replications per treatment at day 2. Finally, I measured differences in fungal biomass after 21 days of incubation and regressed PPO activity against biomass.

Fungal Growth Assays. Fungus obtained from PDA plates was placed in 65 ml SDB and blended for 1 min in a Vertis blender at low speed. Five milli- liters of the fungal suspension were pipettcd into Erlenmeyer flasks containing 45 ml of SDB. The cultures were then grown for 10 weeks in the dark. After


10 weeks, the fungus was filtered, resuspended in modified FPT (50 ml FPT/ flask of fungus), and blended for 1 min at low speed. The fungus was kept evenly suspended using a stir bar and magnetic plate.

One milliliter of fungal suspension was added to 8 ml of modified FPT in sterile test tubes and placed for four days in the dark at room temperature. After four days 1 ml of one of four quebracho tannin solutions was added to the suspension to give a final concentration of 0, 0.0025, 0.025, and 0.25% quebracho tannin. At various time intervals from 10 min to 21 days, the activity of polyphenol oxidase (PPO) was determined (using the protocol described below), and at 21 days the dry mass of the fungus in each treatment was determined.

PPO Activity. Catechol is commonly used as a substrate to measure PPO activity (Flurkey and Jen, 1978). Liquid cultures of fungus were centrifuged at 1000 rpm for 10 min and the supernatants placed on ice. In test tubes, 0.75 ml of the supematant from the fungal cultures were combined with 2 ml 0.1 M citric acid-sodium citrate buffer (pH 5.5). Following the addition of 0.25 ml of catechol (or distilled water for controls), the test tubes were vortexed and incubated at 30~ for 30 min, and then the final absorbance was determined (420 nm) with a spectrophotometer. After subtracting the absorbance of the controls, a final PPO activity was obtained.

Finally, because the concentration of tannin in solution influences pH, I also determined whether differences in pH would influence PPO activity. Using a 0.1 M citric acid-sodium citrate buffer, the effects of three different pH solutions (4.0, 5.0, 6.0) on PPO activity were tested. Solutions were incubated for two days before measuring PPO activity.

Statistical Analyses

All statistical analyses were done using SAS (SAS Institute, 1985). Chem- istry data were natural log transformed prior to analyses.

Bioassays with Leaves. The number of leaf disks removed during the pickup assay was regressed against each of the chemical traits measured. In the pro- cess, three outliers were identified (based on studentized residuals) and removed. Because Mole and Waterman (1988) suggested that the ratio of protein to tannin may be the key to the susceptibility of plants to herbivores, I also divided the concentration of protein by each of the three phenolic traits (the two condensed tannins and protein-binding capacity) and regressed the number of leaf disks removed against each of these ratios.

Laboratory Bioassays with Ants. Neither the order in which the treatments were tested nor the colony tested had a significant effect on leaf pickup assays. Therefore the effects of increasing tannin concentration on the acceptability of rye flakes was determined using the following one-way ANOVA model: flakes removed (X) = constant + In (tannin concentration). A Bonferroni test was

1186 NICHOLS-ORIANS

used to compare differences in acceptability among the different tannin concentrations.

Laboratory Fungal Bioassays. The effect of increasing tannin concentration on PPO activity also was analyzed with an ANOVA model with tannin concentration as the main effect. A Bonferroni test was used to compare treatment means. Unfortunately, the quebracho tannin bound to the fungus, so I was unable to determine the effects of tannin concentration on fungal growth. A regression analysis was used to determine the relationship between fungal biomass and PPO activity.

RESULTS

Bioassays with Leaves

Results from the regression analysis indicate that the acceptability of leaves to leaf-cutter ants was negatively correlated with protein-binding capacity (PBC) and the two condensed tannins (Figure 2A-C). The fact that all three traits gave similar negative correlations was because PBC and the two condensed tannins were all highly correlated (P > 0.91). For both condensed tannin traits, proanthocyanidin (PA) and leucoanthocyanin (VN) condensed tannins, a quadratic model provided the highest correlation. This suggests that lower concentrations of tannins did not deter the ants. However, linear and quadratic models provided identical correlation coefficients for PBC. Protein (PRT) concentration was not correlated with leaf acceptability (R 2 = 0.02, P = 0.651) and the ratio of PRT to PBC, PA, or VN did not improve the correlation (PRT/PBC, R 2 = 0.31; PRT/PA, R 2 = 0.23; PRT/VN, R 2 = 0.24).

Laboratory Bioassays

Bioassays with Ants. An analysis of variance indicates that the concentration of quebracho tannin (QT) affected the acceptability of rye flakes to leaf- cutter ants (Figure 3). Lower concentrations did not deter the ants, while higher concentrations did deter them.

Fungal Bioassays. Quebracho tannin did inhibit fungal PPO activity. Although the addition of QT to the fungal suspensions lowered pH, the inhib- itory effect was not due to a change in pH (Table 1). I found no difference in the PPO activity of fungal cultures buffered at pH 5.0 and 6.0. Only at pH 4 was there a reduction in PPO activity (PPO = 0.075).

Incubation of QT with the fungus culture for two days demonstrated that PPO activity is inhibited only at the highest concentration (Figure 4). Only high concentrations appeared to inhibit the fungus as well. After 21 days, fungal cultures in 0.25 % QT had died, whereas those at the lower concentrations were still growing vigorously.

A)

O

r

C~

1 5

10

0

0 . 5

[]

[]

[]

u i i , r - I ~ �9

1 . 0 1 . 5 2 . 0 2 .5

PROTEIN B I N D I N G CAPACITY (%TAE)

y = 12.29 - 5o00X R 2 = 0.31

3.0

B )

C~

[.T.1

qr~

C~

ql:

15

1 0

5'

0

1 . 0

[]

[] [] [] r<uul~-~.M [ ]

m N [ ] ~

i i a i u r- . u - "

1 . 5 2 . 0 2 . 5 3 . 0 3 . 5 4 . 0 4 .5

T A N N I N (PA) C O N C E N T R A T I O N ( % l ~ r E )

m

y = 9 . 6 6 - 0.17x - 0.51x 2 R 2 = 0.27

5.0

C)

C~ tz~ ;> O

r/]

15

1 0

y = 2.62 + 7.55x + 2.55x2 [] R 2 = 0.26 [] []

[] m p < 0.01 [] [] m

[] [] [] " ~

2 . 0 2 .5 3 . 0

T A N N I N (VN) C O N C E N T R A T I O N (%WTE)

0

1.5 3.5

F[6. 2. Regression analysis showing the relationship between the acceptability of leaf disks to leaf-cutter ants and three phenolic traits: (A) protein-binding capacity, (13) proanthocyanidin (PA) condensed tannin, (C) leucoanthocyanin (VN) condensed tannin. Chemical data were natural log transformed.

1188 NICHOLS-ORIANS

3~ k 25

a a

: : co_ o c

d 5-

0 I I I I I

-3 -2 -1 0 1 2

Ln(QUEBRACHO TANNIN CONCENTRATION)

FIG. 3. Analysis of variance results showing the relationship between the acceptability of rye flakes (_+ standard error) to leaf-cutter ants and the concentration of quebracho tannin. Chemical data were natural log transformed.

TABLE 1. EFFECTS OF QUEBRACHO TANNIN (QT) AND pH ON PPO ACTIVITY a

Concentration (%) pH PPO

A. Quebracho tannin 0 5.32 0.105 0.0025 5.28 0.114 0.025 5.28 0.093 0.25 5.16 0.015

B. Buffer 6.0 0.085 5.0 0.086 4.0 0.075

a PPO activity was measured as the change in absorbance following incubation of fungal suspension with catechol for 30 minutes. (A) The effects of QT on pH and PPO activity, and (B) the effects of pH on PPO activity. In (B), a 0.10 M citric acid-sodium citrate buffer was used to manipulate pH.

T A N N I N S , LEAF-CUTTER ANTS, A N D F U N G I ] 1 8 9

0 , 1 0

i 0,06 "

O 0,04"

0 . 0 2 -

o.oo

a ,,-t a

b

! i . . . . . . . . ! . . . . i " "

o% ,oo25% .o25~ .25~ QUEBRACHO TANNIN CONCENTRATION

FIG. 4, Polyphenol oxidase activity (_+ standard error) after two days in fungal cultures containing different concentrations of quebracho tannin, x axis is log scaled.

O

0 . 4 8

0.38

0 . 2 8

0.18

0 . 0 8

-0.02 ~

I C O N C ~ R A a I O N I o

I o I { A .oz~% {

@ x @ .@

+ + + § +

10 30 60 120 2

MINUTES

+ 4- i !

7 21

DAYS

FI~. 5. The effects of different concentrations of quebracho tannin on the polypheno/ oxidase activity through time.

1190 NICHOLS-ORIANS

0.48 [ ]

y = 0 . 2 7 + 0 . 1 0 x 0.46 ' R 2 = 0 . 8 2

0.44 '

i 0.42'

0.40 '

0.38 �9

0.36 , , , ,

1.0 1.2 1.4 1.6 1.8 2.0

FUNGAL BIOMASS (rag)

FIG. 6. The effects of fungal biomass on polyphenol oxidase activity. All measures taken from 21- day-old tannin-free fungal cultures.

There was no evidence that the presence of tannin was necessary for the production of PPO (Figures 4 and 5). PPO activity in supernatants without tannin was as high if not higher than in supernatants with tannin at several sampling periods (Figure 5). Rather, PPO activity is correlated with fungal biomass (Figure 6). In 21-day-old control (0% QT) cultures, the greater the biomass the higher the PPO activity.

DISCUSSION

Tannins do not deter herbivores uniformly (Bemays, 1978, 1981; Schultz, 1989). Some herbivores appear adapted for consuming high tannin leaves (see Schultz, 1989). For others, some dietary tannin may actually be beneficial (Ber- nays, 1978; Keating et al., 1988). Previous research on the interactions between L oerstediana and leaf-cutter ants suggests that only higher concentrations of tannin deter the ants (Nichols-Orians and Schultz, 1990; Nichols-Orians, 1991a,b). Since the fungus the ants cultivate produces enzymes that facilitate the inactivation of phenolic substances (Cherrett et al., 1989), the fungus may be adapted to handle lower concentrations of tannin. I had hypothesized that leaf-cutter ants may not be sensitive to lower concentrations of tannin because the fungus is capable of inactivating the tannins.


Results of the bioassays with leaves and rye flakes indicate that the condensed tannins of L oerstediana and QT inhibited leaf-fungal substrate selection by leaf-cutter ants (Figures 2 and 3). I should note that the protein-binding capacity of L oerstediana leaves was as high or higher than the flakes (unpublished data), which suggests that the concentrations of QT selected were appro- priate. Consistent with previous research (Nichols-Orians, 1991a), very low concentrations did not deter the ants (Figures 2 and 3).

I had hypothesized that only in the presence of QT would there be PPO activity (sensu Mayer, 1987). I further suggested that only higher concentrations of tannin would inhibit PPO activity and, therefore, the fungus. Contrary to my hypothesis, fungal cultures without QT showed PPO activity (Figures 4 and 5). In fact, PPO activity increases with fungal biomass (Figure 6). Thus, the production of PPO is constitutive and does not appear to be induced by QT. However, the results are not definitive because QT can inhibit the enzymes directly (unpublished data). This could have masked an increase in PPO activity above constitutive levels at the lower tannin concentrations. Studies of gene activation are needed to determine whether QT or other phenolics can induce PPO production.

As hypothesized, only a high concentration of quebracho tannin strongly inhibited PPO activity (Figures 4 and 5). This may explain why leaf-cutter ants were most inhibited by higher concentrations of condensed tannins. Unfortu- nately, it was not possible to compare directly the concentration of QT in the liquid media to the concentration applied to the flakes. Therefore, it is not clear from these results whether leaf-cutter ants can discriminate between a tannin concentration that does and does not inhibit the fungus. Future experiments should investigate this possibility.

Tannins are not the only chemicals that vary among leaves of different ages and among plants in different environments; nutrients vary as well. Both nutrients and secondary chemicals may be responsible for differential acceptability of leaves to herbivores (Fox and Macauley, 1977; Mattson, 1980; Wint, 1983; Waltz, 1984; Mole and Waterman, 1988). Wint (1983) and Mole and Waterman (1988) hypothesized that the selection of leaves by herbivores would not depend upon the concentration of protein or tannin but upon the ratio of protein to tannin. The data presented here do not provide support for their hypothesis.

In a previous study (Nichols-Orians, 1991a), leaves of seedlings growing in full sun had a higher concentration of protein and were more acceptable to leaf-cutter ants than were leaves of understory seedlings, despite a higher concentration of tannin (Figure 7). [Since the concentration of other nutrients, such as nonstructural carbohydrates, may also increase with increasing light availability (Nichols-Orians, 1991a), the preference for sun leaves may not have been due changes in protein concentration specifically.] In another study

1192 NICHOLS-ORIANS

FIG. 7. Relationship between the acceptability of leaf disks to leaf-cutter ants (_+ SE) and the protein and condensed tannin (leucoanthocyanin) concentration (_+ SE) in various leaf types. (Included are mature leaves of seedlings growing in different light environments, and young and mature leaves of trees growing in full sun. Chemical data were natural log transformed prior to statistical analyses. Bars (within category) with different letters are different at P _< 0.05. (Data from Nichols-Orians and Schultz, 1990; Nichols-Orians, 1991a).

(Nichols-Orians and Schultz, 1990), young leaves of juvenile trees growing in full sun had a higher protein concentration than mature leaves, yet were less acceptable to the ants. The rejection of the young leaves appeared due to the higher concentration of tannin in young leaves as compared to mature leaves (Figure 7). These studies suggest that, although the ratio of nutrients to tannin appears unimportant, both nutrients and condensed tannins do influence leaf selection by leaf-cutter ants.

Researchers have not always found tannins to deter leaf-cutter ants (Little- dyke and Cherrett, 1976; Kawanashi and Raffauf, 1986; Howard, 1987, 1988, 1990). Most of these studies have been qualitative. Since most woody plants produce tannins, yet many produce only a low concentration (Coley, 1983a), qualitative studies may not detect tannin effects on acceptability. Relative to other species (Denslow et al., 1990), L oerstediana produces a high concentra-


tion of tannin (Nichols-Orians, 1990), yet even in this species an increase in tannin does not necessarily lead to a decrease in the acceptability of leaves to leaf-cutter ants (Nichols-Orians, 199 la). Clearly, the inhibition of ants and their fungus by condensed tannins is concentration dependent.

Acknowledgments--H. Appel, J. Cazin, W. McCarthy, C. Mullin, and D. Wiemer provided suggestions for the experimental protocol. The Organization for Tropical Studies (OTS) helped make this research possible. A special thanks go to David and Deborah Clark and Victor Chavarria of OTS for making it possible to live in Puerto Viejo. Marc Pacchioli helped me conduct laboratory' bioassays with the ants. Early versions of this paper benefited from the comments of J. Schultz, H. Appel, J. Howard, S. Keating, and two anonymous reviewers. This work was supported by a Jesse Smith Noyes Fellowship Grant awarded by OTS, and by three National Science Foundation Grants [BSR-8605106 to J. Schultz, BSR-8815648 to C. Nichols-Orians (Dissertation Improvement Grant), and BSR-8917752 to R. Fritz.]

REFERENCES

ARKCOLL, D.B. 1984. A comparison of some fast growing species suitable for woodlots in the wet tropics. Pesquisa Agr. Brasil. 19:61-68.

BARRER, P.M., and CHERRETT, J.M. 1972. Some factors influencing the site and pattern of leaf- cutting activity in the ant Atta cephalotes (L.). J. Entomol. 47:15-27.

BATE-SMITH, E.C. 1977. Astringent tannins of Acer species. Phytochemistry 16:1421-1426. BERISH, C.W. 1986. Leaf-cutting ants (Atta cephalotes) select nitrogen rich forage. Am. Midl. Nat.

115:268-276. BERNAYS, E.A. 1978. Tannins: an alternate viewpoint. Entomol. Exp. Appl. 24: 44-53. BERNAYS, E.A. 1981. Plant tannins and insect herbivores: An appraisal. Ecol. EntomoL 6:353-

360. BROADHURST, R.B., and JONES, W.T. 1978. Analysis of condensed tannins using acidified vanillin.

J. Sci. Food Agric. 29:788-794. BUTLER, L.G., PRICE, M.L., and BROTHERTON, J.E. 1982. Vanillin assay for proanthocyanidins

(condensed tannins): Modification of the solvent for estimation of the degree of polymerization. J. Agric. Food Chem. 30:1087-1089.

CHANDLER, G., and GOOSEM, S. 1982. Aspects of rainforest regeneration. III. The interaction of phenols, light and nutrients. New Phytol. 92:369-380.

CHERRETT, J.M. 1968. The foraging behavior of Atta cephalotes L. (Hymenoptera, Formicidae). 1. Foraging pattern and plant species attacked in tropical rain forest. J. Anita. Ecol. 37:387- 403.

C~ERRETT, J.M. 1972. Some factors involved in the selection of vegetable substrate by Atta cephalotes (L.) (Hymenoptera: Formicidae) in tropical rain forest. J. Anim. Ecol. 41:647-660.

CHERRETT, J.M., POWELL, R.J., and STRADLING, D.J. 1989. The mutualism between leaf-cutting ants and their fungus, pp. 93-120, in N. Wilding, N.M. Collins, P.M. Hammond and J.F Webber (eds.). Insect-Fungus Interactions. Academic Press, London.

COLEY, P.D. 1983a. Herbivory and defensive characteristics of tree species in a lowland tropical forest. Ecol. Monogr. 53:209-233.

COLEY, P.D. 1983b. Intraspecific variation in herbivory on two tropical tree species. Ecology 64:426-433.

COLEY, P.D., and A~DE, T.M. 1989. Red coloration of tropical young leaves: A possible antifungal defence? J. Trop. Ecol. 5:293-300.

1194 NICHOLS-ORIANS

COMPTON, S.J., and JONES, C.G. 1985. Mechanisms of dye response and interference in the Brad- ford protein assay. Anal. Biochem. 151:369-374.

COOPER, R.M., and WOOD, R.K.S. 1973. Induction of synthesis of extracellular cell-wall degrading enzymes in vascular wilt fungi. Nature 246:309-311.

COOPER, R.M., and WOOD, R.K.S. 1975. Regulation of synthesis of cell wall degrading enzymes by Verticillium albo-atrum and Fusarium oxysporum f. sp. lycopersici. Physiol. Plant Pathol. 5:135-156.

DENSLOW, J.S., SCHULTZ, J.C., VITOUSEK, P.M., and STRAIN, B.R. 1990. Growth responses of tropical shrubs to treefall gap environments. Ecology 71 : 165-179.

FIELD, C., and MOONEY, H.A. 1986. The photosynthesis-nitrogen relationship in wild plants, pp 25-56, in T. Givnish (ed.). On the Economy of Plant Form and Function. Cambridge Uni- versity Press, Cambridge, England.

FENNAH, R.G. 1950. Parasol ants, their life history and methods for their control. Proc. Agric. Soc. Trinidad 50:312-326.

FLURKEY, W.H., and JEN, J.J. 1978. Peroxidase and polyphenoloxidase activities in developing peaches. J. Food Sci. 43:1828-1831.

Fox, L.R., and MACAULEY, B.J. 1977. Insect grazing on Eucalyptus in response to variation in leaf tannins and nitrogen. Oecologia (Berlin) 29:145-162.

GALL, L.F. 1987. Leaflet position influence caterpillar feeding and development. Oikos 49:172- 176.

HARTSHORN, G.S. 1983. Plants: Introduction, pp 136-140, in D.H. Janzen (ed.). Costa Rican Natural History. University of Chicago Press, Chicago, IL.

HOLDRIDGE, L.R., GREMEKE, W.C., HATHEWAY, W.H., LIANG, T., and TOSE, J. JR. 1971. Forest Environments in Tropical Life Zones: A Pilot Study. Pergamon Press, Oxford. 747 pp.

HOWARD, J.J. 1987. Leaf-cutting ant diet selection: The role of nutrients, water and secondary chemistry. Ecology 68:503-515.

HOWARD, J.J. 1988. Leaf-cutting ant diet selection: The relative influence of leaf chemistry and physical features. Ecology 69:250-260.

HOWARD, J.J. 1990. Infidelity of leafcutting ants to host plants: Resource heterogeneity or defense induction? Oecologia 82:394-401.

HUBBELL, S.P., and WEIMER, D.F. 1983. Host plant selection by an attine ant, pp. 133-154, in P. Jaisson (ed.). Social Insects in the Tropics. University of Paris Press, Paris, France.

JONES, C.G., HARE, J.D., and COMeTON, S.J. 1989. Measuring plant protein with the Bradford assay. 1. Evaluation and standard method. J. Chem. Ecol. 15:979-992.

KAWANASHI, K., and RAFFAUF, R.F. 1986. Caryocar microcarpum: An ant repellent and fish poi- son of the northwest Amazon. J. Nat. Prod. 49:1167-1169.

KEATING, S.T., YENDOL, W.G., and SCHULTZ, J.C. 1988. Relationship between susceptibility of gypsy moth larvae (Lepidoptera: Lymantriidae) to baculovirus and host plant foliage constituents. Environ. Entomol. 17:952-958.

KIRK, T.K., and KELMAN, A. 1965. Lignin degradation as related to the phenoloxidases of selected wood-decaying Basidiomycetes. Phytopathology 55:739-745.

KOPTUR, S. 1985. Alternative defenses against herbivores in Inga (Fabaceae: Mimosoideae) over an elevational gradient. Ecology 66:1639-1650.

LITTLEDYKE, M., and CHERRETT, J.M. 1976. Direct ingestion of plant sap from cut leaves by the leaf-cutting ants Atta cephalotes (L.) and Acromyrmex octospinosus (Reich) (Formicidae, Attini). Bull. Entomol. Res. 66:205-217.

LYR, H. 1962. Detoxification of heartwood toxins and chlorophenols by higher fungi. Nature 195:289-230.

MATTSON, W.J. 1980. Herbivory in relation to plant nitrogen content. Annu. Rev. Ecol. System. 11:119-161.


MAYER, A.M. 1987. Polyphenol oxidases in plants--recent progress. Phytochemistry 26:11-20. MEYER, G.A., and MONTGOMERY, M.E. 1987. Relationships between leaf age and the food quality

of cottonwood foliage for the gypsy moth, Lymantria dispar. Oecologia (Berlin) 72:527-532. MOLE, S., and WATERMAN, P.G. 1988. Light-induced variation in phenolic levels in foliage of

rain-forest plants. II. Potential significance to herbivores. J. Chem. Ecol. 14:23-34. NICHOLS-ORIANS, C.M. 1990. Interactions among plant environment, foliar traits, and leaf selec-

tion by an attine ant. PhD thesis. The Pennsylvania State University, University Park, Penn- syl.vania.

NICHOLS-ORIANS, C.M. 1991a. The effects of light on foliar chemistry, growth and susceptibility of seedlings of a canopy tree to an attine ant. Oecologia In press.

NICHOLS-ORIANS, C.M. 1991b. Environmentally induced differences in plant traits: Consequences for susceptibility to a leaf-cutter ant. Ecology, 72(5).

NICHOLS-ORIANS, C.M., and SCrtULTZ, J.C. 1989. Leaf toughness affects leaf harvesting by the leaf-cutter ant, Atta cephalotes (L.) (Hymenoptera: Formicidae). Biotropica 21:80-83.

NICHOLS-ORIANS, C.M., and SCHULTZ, J.C. 1990. Interactions among leaf toughness, chemistry and harvesting by attine ants. Ecol. Entomol. 15:311-320.

POWELL, R.J., and STRADLING, D.J. 1986. Factors influencing the growth of Attamyces bromati- fucus, a symbiont of attine ants. Trans. Br. Mycol. Soc. 87:205-213.

QUINLAN, R.J., and CHERRETT, J.M. 1979. The role of fungus in the diet of the leaf-cutting ant Atta cephalotes (L.) .Ecol. EntomoL 4:151-160.

ROCKWOOD, L.L. 1976. Plant selection and foraging patterns in two species of leaf-cutting ants (Atta). Ecology 57:48-61.

ROCKWOOD, L.L. 1977. Foraging patterns and plant selection in Costa Rican leaf-cutting ants. J. NY Entomol. Soc. 85:222-233.

SAS IINSTITUTE. 1985. SAS User's Guide: Statistics. SAS Institute, Cary, North Carolina. SCHULTZ, J.C. 1989. Tannin-insect interactions, pp. 417-433, in R.W. Hemingway and J.J. Kar-

chesy (eds.). Chemistry and Significance of Condensed Tannins. Plenum Press, New York. SCHULTZ, J.C., BALDWIN, I.T., and NOTHNAGLE, P.J. 1981. Hemoglobin as a binding substrate in

the quantitative analysis of plant tannins. J. Agric. Food Chem. 29:823-826. SEAMAN, F.C. 1984. The effects of tannic acid and other phenolics on the growth of the fungus

cultivated by the leaf-cutting ant, Myrmicocrypta buenzlii. Biochem. System. Ecol. 12:155- 158.

SNYDER, J.C., and DESBOROU~H, S.L. 1978. Rapid estimation of potato tuber total protein content with coomassie brilliant blue G-250. Theor. Appl. Genet. 52:135-139.

SWAIN, T., and HILLIS, W.E. 1959. The phenolic constituents of Prunus domestica. I. The quantitative analysis of phenolic constituents. J. Sci. Food Agric. 10:63-68.

VITOUSEK, P.M., and DENSLOW, J.S. 1987. Differences in extractable phosphorous among soils of the La Selva Biological Station, Costa Rica. Biotropica 19:167-190.

WALLER, D.A. 1982. Leaf-cutting ants and live oak: The role of leaf toughness in seasonal and intraspecific host choice. EntomoL Exp. Appl. 32:146-150.

WALTZ, S.A. 1984. Comparative study of predictability, value, and defenses of leaves of tropical wet forest trees (Costa Rica). PhD Thesis. University of Washington, Seattle, Washington.

WATERMAN, P.G., Ross, J.A.M., and McKEY, D.B. 1984. Factors affecting levels of some phenolic compounds, digestibility, and nitrogen content of the mature leaves of Barteriafistulosa (Passifloraceae). J. Chem. Ecol. 10:387-401.

WINT, G. R. W. 1983. The effect of foliar nutrients upon the growth and feeding of a lepidopterau larva, pp. 301-320, in J.A. Lee, S. McNeill, and I.H. Rorison (eds.). Nitrogen as an Eco- logical Factor. Blackwell Scientific, Oxford, England.

ZUCKER, W.V. 1983. Tannins: Does structure determine function? An ecological perspective. Am. Nat. 121:335-365.

condensed tannins, attine ants, and the performance of a ......inga spp. produce high concentrations...

Documents