Mycobiota Associated with the Coffee Berry Borer (Coleoptera: Scolytidae) and Its Galleries in Fruit

Transcription

1 ECOLOGY AND POPULATION BIOLOGY Mycobiota Associated with the Coffee Berry Borer (Coleoptera: Scolytidae) and Its Galleries in Fruit GLORIA CARRIÓN 1 AND ARTURO BONET 2 Ann. Entomol. Soc. Am. 97(3): 492Ð499 (2004) ABSTRACT Fungi associated with the coffee berry borer, Hypothenemus hampei (Ferrari), are determined and galleries in coffee fruit, Coffea arabica L., are described. In total, 13 fungus species were found to be associated with the coffee berry borer and its galleries. These fungi were divided into four functional groups: 1) saprobic fungi that the borer uses as food: Fusarium heterosporum Nees ex Fries, Cladosporium sp., Cladosporium oxysporum Berkeley & Curtis, and Penicillium echinulatum Fassatiová; 2) saprobic fungi that degrade borer feces: Aspergillus flavus Link, Aspergillus niger Tieghem, Mucor luteus Linnemann, Penicillium sp. 1, Humicola grisea Traaen, and Gliocladium penicilloides Corda; 3) fungi that are coffee parasites: Fusarium oxysporum Schlechetend and Fusarium solani (Martius) Saccardo; and 4) the entomopathogenic fungus Beauveria bassiana (Balsamo-Crivelli) Vuillemin. The same fungi were found on and inside the insect as well as in galleries and were components of the insectõs diet. Apparent interference of B. bassiana by F. solani was observed. KEY WORDS associated mycobiota, Beauveria bassiana, coffee borer galleries, Hypothenemus hampei, insectðfungus interaction THE COFFEE BERRY BORER, Hypothenemus hampei (Ferrari), uses coffee fruits, Coffea arabica L., as a suitable habitat for growth of immature stages (eggs, larvae, and pupae), as do teneral adults for their progeny. This species has become the most important cosmopolitan pest on coffee crops (Le Pelley 1968). Insecticides have been used to control the insect, although with limited success and increased resistance problems (Mejõ a and López 2002). Various cultural and sanitary measures have been used to prevent the borer from reproducing, such as removal of infested fruit after harvest and before the production of new fruits. Such measures have been only partially successful, because not all producers follow them systematically (Benavides et al. 2002). Research on its control has focused on the speciesõ natural enemies, such as the hymenopteran parasitoids Cephalonomia stephanoderis Betrem and Phymastichus coffea La Salle, and the entomopathogenic fungi Beauveria bassiana (Balsamo- Crivelli) Vuillemin and Metarhizium anisopliae (Metschnikoff) Sorokin (Barrera et al. 1990; De la Rosa et al. 1995, 1997a,b; Villacorta and Barrera 1996; López-Vaamonde and Moore 1998; Haraprasad et al. 2001). However, neither parasitoids nor entomopathogens have succeeded in limiting coffee berry borer population growth in Mexico. 1 Departamento de Hongos, Instituto de Ecologõ a, A.C., Apartado Postal 63, Xalapa, Veracruz, México. gloriac@ ecologia.edu.mx. 2 Departamento de Entomologõ a, Instituto de Ecologõ a, A.C., Apartado Postal 63, Xalapa, Veracruz, México. bonetart@ ecologia.edu.mx. H. hampei was introduced from Guatemala and Þrst detected in 1978 in Soconusco, Chiapas, Mexico (Baker 1984). It has spread to every coffee-producing region in the country and has increased in prevalence since Many small coffee growers have abandoned their crops due to low market prices, leaving fruits unharvested. This has left fruits available for the borer and has made continual colonization possible, thus increasing borer population density. We know little about this pestõs biology, the organisms with which it interacts, and the environment where it develops. Baker (1984) and Villacorta and Barrera (1996) described the biological cycle of H. hampei in Mexico and indicated that damage begins when the adult female starts using unripe fruit to construct galleries where she oviposits from 20to 60 eggs. Larvae remain inside the fruit until they become adults. The life cycle lasts 30d. Recently, fertilized females actively seek new fruit. Certain fungi also have been detected on ripe coffee fruits, including natural enemies of this insect, such as B. bassiana. López-Vaamonde and Moore (1998) have recorded fungi such as Penicillium and have described them as contaminants without exploring possible interactions between them and the medium in which they develop, interactions that may inßuence the reproductive success of H. hampei. Rojas et al. (1999) and Morales-Ramos et al. (2000) reported the symbiotic association of Fusarium solani (Martius) Saccardo with the coffee berry borer wherein the insect feeds on ergosterol produced by the fungus, leading to an increase in borer progeny. Peterson et al. (2003) found /04/0492Ð0499$04.00/ Entomological Society of America

2 May 2004 CARRIÓN AND BONET: MYCOBIOTA ASSOCIATED WITH H. hampei 493 Penicillium brocae SW Peterson, Pérez, Vega et Infante to be associated with H. hampei. Preliminary tests indicated the presence of several fungi within the coffee fruit galleries. This led to the current study in which these mycobiota are identiþed, and the succession of fungal species in the galleries is described. We hypothesize that many fungal species are present in the galleries and in/on the borer throughout the infestation process of the coffee fruit. Also, the growth of the fungi in the fruit is due to the presence of the borer. Evidence is presented indicating an association or interaction involving fungi, borers, and coffee fruit, and the importance of galleries as a fungal microhabitat. Materials and Methods Random collection of unripe coffee fruits with and without borers was conducted from 1999 to 2002 on the El Tizal coffee plantation, located in the municipality of Emiliano Zapata, Veracruz, Mexico. Microscopic observations of longitudinal and transversal cuts were done in borer galleries as well as in healthy fruits, allowing description of the borerõs gallery formation process. Borers were dissected and their guts removed for microscopic examination. Other specimens, recently collected in the Þeld, were extracted from coffee fruits and killed with ethyl acetate for ease manipulation. One portion of the sample was washed with 96% alcohol for 1 min., 3% sodium hypochlorite for 3 min., and 70% alcohol for 30 s to eliminate possible fungi on the borerõs body surface. Finally, samples were rinsed three times with sterile distilled water. The remainder of the sample was left unwashed live to serve as a control. Both samples were covered with paladium gold and observed under a scanning electron microscope (JSM-5600LV, JEOL, Tokyo, Japan). Fruits with borers were cleaned externally with 96% alcohol and cut longitudinally to expose the inner gallery and eliminate any borer individuals. Afterwards, fruits were placed in a moisture chamber to enhance fungal growth. The same was done with healthy fruits, and the entire sample was kept under observation while fungi grew and sporulated. Detection of fungal mycelia was accomplished in petri dishes that contained agar with dextrose and potato plates (PDA) incubated at 21 2 C in10 treatments (n 50borers or fruits per treatment). Once specimens had been washed, three groups of insects were formed and given different treatments: 1) each washed borer specimen was placed on a PDA plate; 2) the gut of each borer was extracted, care being taken to avoid any contact with external body parts, and immediately placed in 0.5 ml of sterile distilled water where it was opened and its contents homogenized; Þve drops of this suspension was placed on a PDA plate; and 3) individual borers were macerated in sterile distilled water and Þve drops of the macerated product was placed on PDA plates. 4) As a control unwashed borers were placed in petri dishes with PDA plates. To observe fungi from infested fruits, we considered three groups: two groups with unripe fruits, 5) uninfested as a control, and 6) infested, and 7) a third group with ripe uninfested fruit recently collected from the Þeld. Fruit exteriors were cleaned as described previously, and fruits were cut to expose the tunnel. To detect fungal growth in the gallery, 8) recent and 9) mature gallery fragments of 1 mm 2 were extracted from each unripe fruit to isolate fungi. Five fragments per fruit were placed on a PDA plate. Finally, as a control (10) endosperm fragments from healthy fruits were inoculated. All fungi obtained from each petri dish were identiþed and recorded. Observation of the dishes continued for 30d. When fungi were obtained, they were isolated in pure cultures for morphological determination with the help of specialized keys (Booth 1971, Ellis 1971, Domsch et al. 1980). The presence of each fungal species was determined as the number of all mycelia obtained in the sample (Table 1). Data on mycelia in all treatments were not all normally distributed and could not be transformed effectively; hence, they were analyzed with nonparametric tests. For borer treatments we used a KruskalÐWallis one-way analysis of variance (ANOVA) on ranks and MannÐWhitney U tests for fruit and gallery data (StatSoft, Inc. 2000). Results The borer perforates fruits that have reached their maximum size. Penetration occurs at the apical tip, just at the scar left by the ovary style (n 100). There can be one (90%) or two (10%) possible entries (Fig. 1A), which can be situated at the center (52%) or slightly closer to the scar (48%). The entrance is a straight 2Ð3-mm-long tunnel that cuts through the pericarp and pulp and provides access to either of the fruitõs two seeds (Fig. 1B). Next, the borer perforates the testa that protects the endosperm of one of the seeds; here, it will form its gallery and oviposit (Fig. 1C and D). When the gallery has been completed (with a tunnel diameter of 1Ð1.2 mm), the endosperm that is exposed toward the tunnelõs interior (250Ð400 m in width) acquires a green and then a dark green tone as the gallery matures; bacteria and fungi are found in this layer. In cuts of healthy endosperm, a green coloration also is noted, which could be due to the mechanical action of cutting and subsequent oxidation. Unlike healthy fruits, infested ones tend to exhibit cotyledon growth; probably induced by fungal giberellins. At the gallery entrance, tiny brown or dark green fragments are visible. These are part of the waste and feces that result from the femaleõs activity as she removes them from the gallery. In early stages of larval development, the gallery is kept clean. During the last stages of larval development, the main gallery contains fecal residue, endosperm walls are darker, and galleries generally have been bored in the other seed of the same fruit. Larvae eat and develop within the fruit until they become adults, by which time the fruit resources have been exhausted and the young adults look for a new fruit. Thus, the gallery formed by the female borer and later modiþed by larvae constitutes

3 494 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 97, no. 3 Table 1. Number of mycelia obtained from the borer, fruit, and galleries that developed within coffee fruit after 30 d of incubation Fruits Galleries a Borers Control healthy Recent Mature endosperm a Without borer ripe b Infested unripe Control unripe Fungal presence after 30d Washed body Functional group Fungal species Gut a Macerated a unwashed Control live Coffee parasites F. solani F. oxysporum Introduced and possibly grown by the borer P. echinulatum F. heterosporum Cladosporium sp C. oxysporum Saprobes and degraders of borer waste A. niger A. flavus 1 1 M. luteus Penicillium sp.(1) 2 1 H. grisea 1 G. penicilloides 5 Entomopathogens B. bassiana For each treatment, n 50. a Inoculated with Þve gallery fragments or fragments of borer gut per fruit. b Very ripe fruit that had fungi. a microhabitat suitable for the establishment of microorganisms such as fungi. Unwashed borers observed under an scanning electron microscope show clusters of bacteria and fungal yeast spores on the exoskeleton. These clusters were located on the head, with conical adherences on asperites (Fig. 2A and B), at the base of serrate setae (Fig. 2C), and on the ventral part of the adult abdomen (Fig. 2D). Growth of these clusters was not detected on the exoskeleton. In total, 13 fungi were identiþed from the various experiments conducted with coffee borers, fruits, and the galleries formed by the insects within them (Table 1). Ten of these fungi are saprobic [Penicillium echinulatum Fassatiová, Fusarium heterosporum Nees ex Fries, Cladosporium sp. 1, Cladosporium oxysporum Berkeley & Curtis, Aspergillus niger Tieghem, Aspergillus flavus Link, Mucor luteus Linnemann, Penicillum sp. 1, Humicola grisea Traaen, and Gliocladium penicilloides Corda], two are considered parasites of the coffee grain (F. solani and Fusarium oxysporum Schlechtend), and one is a borer entomopathogen (B. bassiana). Fungi detected in the endosperm of infested fruits come from the insect and are, therefore, associated with the borer and its gallery. Although all of these fungi are consumed by H. hampei, we have separated them into four functional groups in relation to the borer: 1) the most abundant saprobic fungi, which we assume the borer uses as food [F. heterosporum, Cladosporium sp (1), C. oxysporum, and P. echinulatum]; 2) saprobic fungi that probably degrade borer waste (A. flavus, A. niger, M. luteus, Penicillium sp. 1, H. grisea, and G. penicilloides]; 3) fungi that are coffee parasites and can be found in the absence of borers where they colonize the coffee grain (F. oxysporum and F. solani), and 4) the entomopathogenic fungus B. bassiana. In the group of unwashed and macerated borers, 11 fungi were present. Their presence reßects the insectõs importance in transporting fungal spores (Fig. 2AÐD). Seven fungi were found in the gut, corresponding to the groups used as food and fruit parasites (Table 1). No fungal growth occurred in any of the washed borer specimens placed in the moisture chamber, demonstrating the efþcacy of the washing protocol. However, two borer individuals secreted a minute drop from the rectum, which fell directly onto the culture. Three fungi subsequently developed: A. niger, P. echinulatum, and F. solani. Seven fungi yeasts and bacteria were found in the borerõs gut and recorded as the number of mycelia per fungal species present in each treatment: F. oxysporum (9), F. heterosporum (3), Cladosporium sp. 1 (1), and C. oxysporum (12). F. solani was only found in 6% of the specimens, P. echinulatum in 16%, and A. niger in 4%. Only one fungus was obtained from 44% of the samples, two from 8%, three from 4%, and four from 2% (Table 2). Eleven fungal species were obtained from macerated borer cultures. These included the same taxa found in the guts, along with the following fungi de-

4 May 2004 CARRIÓN AND BONET: MYCOBIOTA ASSOCIATED WITH H. hampei 495 Fig. 1. Infested coffee fruits. (A) Entrance to borer gallery, at apical tip of fruit, just at the scar left of the ovary style. (B) Longitudinal cut at gallery entrance, with pericarp layers (exocarp and mesocarp) and testa perforated to endosperm. (C) Transverse cut of fruit, the healthy left cotyledon and the right one with galleries made by H. hampei. (D) Exposed gallery with recently oviposited borer eggs present. tected in very low quantities: Penicillium sp. 1 (2), A. flavus (1), M. luteus (1), and B. bassiana (1) (Table 1). In macerated borers, a single bacterium (4%) was found. Only one fungus was obtained from 30% of the borers (Table 2) and two fungi from 18%. Eleven fungal species grew on PDA plates (Table 1). The most common fungi (number of mycelia in brackets) were C. oxysporum (46), Cladosporium sp. 1 (43), F. heterosporum (40), B. bassiana (37), P. echinulatum (32), F. solani (29), and F. oxysporum (22). Species that occurred on only 2% were A. flavus, M. luteus, H. grisea, and Penicillium sp. 1. In dishes with fungal cultures, from two to eight fungi occurred simultaneously in various combinations, with two fungi on 6% of the sample, three on 12%, four on 24%, Þve on 28%, six on 22%, seven on 4%, and eight on 4% (Table 2). There were no signiþcant differences in the number of species of fungal mycelia found in the three groups of borers (KruskalÐWallis one-way ANOVA on ranks, H 1.869, df 2, P 0.39). The most abundant fungi belong to the two groups that we considered as insect food and fruit parasites. As in previous experiments, more than one fungus developed in petri dishes (Table 2). These fungi occurred sequentially, with some predominating over others. After 10d of growth in dishes, interactions between species began. F. oxysporum was observed to limit the growth of Cladosporium species, whereas F. heterosporum had the same effect on Cladosporium sp. 1. P. echinulatum grew on Cladosporium sp. 1 and C. oxysporum, although in limited quantities. H. grisea grew on F. heterosporum and P. echinulatum, and F. solani on Cladosporium sp. 1, C. oxysporum, F. oxysporum, F. heterosporum, P. echinulatum, and B. bassiana. In some dishes, B. bassiana grew on Cladosporium sp. 1 and C. oxysporum, and F. heterosporum grew on P. echinulatum. Nine B. bassiana strains limited the growth of F. solani, and sporulation of B. bassiana was completely free of the fungus. Two strains of P. echinulatum were observed to invade B. bassiana mycelia and interfere with sporulation. The dominant fungal species F. solani developed in 29 culture dishes. After 21 d, it was found in 11 more dishes. Six fungal species were obtained from borer-infested fruits, both from the pericarp (exocarp and mesocarp) and endosperm (seeds): A. niger (2), Cladosporium sp. 1 (17), F. oxysporum (16), F. solani (7),

5 496 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 97, no. 3 Fig. 2. Surface of unwashed borer bodies. (A) Dorsal view of head with conical adherences on plates. (B) Close-up (500 ) of asperites with conical adherences on head. (C) Clusters at the base of serrate setae. (D) Spore clusters in medium ventral region. G. penicilloides (5), and P. echinulatum (3). In uninfested unripe fruits, no fungal growth was observed in endocarp or endosperm. Two fungi that are coffee fruit parasites were present in ripe fruits, F. oxysporum (43) and F. solani (2), growing Þrst in the exocarp and mesocarp and later in the endosperm. No signiþcant difference was observed in fungi found on infested unripe fruits and ripe fruits without borers (MannÐ Whitney U test, U 5.50, Z 0.167, P 0.87). On unripe attacked fruits, P. echinulatum, Cladosporium sp. 1, A. niger, and G. penicilloides also were found. In recently formed galleries, four fungal species were found: F. oxysporum (6), F. solani (3), F. heterosporum (6), and P. echinulatum (5) (Table 1). A single fungus was obtained from 30% of the fruits, and two fungi from 6% (in F. oxysporumðf. solani and F. oxysporum F. heterosporum combinations). In 30% of recent galleries, only bacteria were found; in 34% no Table 2. Percentage of coffee berry borer specimens, fruit, and borer galleries with more than one fungal species after 30 d of growth No. of fungal species Washed body Unwashed live Borers Fruits Galleries Gut Macerated Uninfested Infested Recent Mature Bacteria No bacteria or fungi

6 May 2004 CARRIÓN AND BONET: MYCOBIOTA ASSOCIATED WITH H. hampei 497 bacteria or fungi were observed although the gallery wall was already green due to oxidation. All samples were observed for over 30d (Table 2). No fungus grew in the endosperm of healthy fruits used as controls. Seven species of fungi were found in ripe galleries: Cladosporium sp. 1 (2), C. oxysporum (9), and B. bassiana (2), F. oxysporum (17), F. solani (9), F. heterosporum (7), and P. echinulatum (11) (Table 1). These fungi were observed in various combinations in inoculated fragments, with one (26%), two (36%), and three fungi (6%). In 20% of the galleries, no growth was recorded, and bacteria were obtained from only 8% of the fruits. As in the previous experiment, the endosperm of healthy fruits used as controls was devoid of fungus. No signiþcant difference in fungi was found between recent and mature galleries (MannÐ Whitney U test, U 8.5, Z 1.039, P 0.298). Discussion A gallery formed by H. hampei is a microhabitat characterized by environmental conditions that favor the growth and succession of fungal communities until the fruit is gone (e.g., F. solani). Fungal growth in coffee fruits was observed within galleries made by insects, including yeast fungi from body surfaces and gut contents. It is through insect contact with gallery walls and body excretions that bacteria and yeast are introduced. Of the 13 fungi found associated with the coffee borer and its galleries, 10are saprobic, two are fruit parasites, and one is an entomopathogen. Although fungal associations are well known in other scolytids (Beaver 1989, Berryman 1989), previous studies of the coffee borer have cited only two: F. solani (Rojas et al. 1999; Morales-Ramos et al. 2000) and P. brocae (Peterson et al. 2003). Upon piercing the fruit, the borer initiates structural changes in the fruit that facilitate fungal colonization. When H. hampei bores into healthy fruit, the Þrst four fungi (F. oxysporum, F. heterosporum, P. echinulatum, and F. solani) establish in the endosperm within the gallery. Over time, the number of fungi increases to seven, including an entomopathogen. Most of them serve as food for the insect. Fungi obtained from the insectõs gut are all found in the ripe gallery except A. niger. A. flavus and M. luteus have been associated with the gut and have been reported in stomach content and maceration analyses. Beti et al. (1995) found that the maize weevil, Sithophilus zeamais Motschulsky, carries A. flavus spores both on the outside of its body and in its mesenteron; the fungus has also been associated with its feces. When the new borer generation has grown, after having fed on the fruit, the new adults disperse and introduce the fungi to healthy fruit. All fungi identiþed, including parasitic fungi, can be found in the coffee plant environment without necessarily causing fruit damage. Even without borer involvement, the latter can establish on the outside of the pericarp, causing it to rot. Thus, they are not introduced into the fruit endosperm as efþciently as with the aid of borer activity. The functional group formed by the coffee parasites F. solani and F. oxysporum was found in all treatments except that of the healthy endosperm, in which no fungal presence was noted. Most of the fungi identiþed in this study that are associated with the coffee borer have been registered as contaminants of different stored grains and are capable of producing mycotoxins (Frisvad 1988). However, we do not know whether processed coffee grains still contain these compounds. Robledo et al. (2001) have found ochratoxin A in unripe coffee grains from the state of Nayarit (Mexico), but contamination with this toxin is associated with high moisture and temperature during storage. Booth (1971) found F. oxysporum f.sp. coffeae in coffee, whereas Northolt and Soentoro (1988) recorded A. versicolor contaminating coffee grains. F. heterosporum is principally known as a cereal and grass parasite that is associated with rotting roots; it affects the germination of grass seeds and leguminous trees (Booth 1971). Wellman (1977) recorded the following Fusarium species as causing disease in the coffee plant: F. lateritium (cankers), F. moniliforme (moldy coffee bean shells), F. oxysporum (wilt), F. roseum (isolated from patiodried whole cherries and from enlarged leaf spots), F. solani (dieback of fruiting branches), and Cladosporium sp.; the latter is considered to be a species that damages the seed when it is placed on drying patios under the sun. In Colombia, F. oxysporum and F. solani were found to grow on the coffee berry borer, yet pathogenic tests were negative (Pérez-López et al., 1996). Along with being parasites of different plants, these two fungal species have been reported as the parasites of different insects in China (Feng-Yan and Qing-Tao 1991). In our experiments, the only entomopathogenic fungus present was B. bassiana, found in samples taken from ripe galleries, borer macerations, and live borers. The amount of fungi within borer galleries suggests that fungi help degrade the endosperm and serve as food for the female and her brood. It has been demonstrated that F. heterosporum produces a lipase composed of a very large N-terminal peptide with 275 amino acids and a C-terminal peptide with 26 amino acids (Nagao et al. 2000, 2001, 2002) that the borer could be using as food. When fruit production is over, the borer remains inside mummiþed fruits that are attached to the coffee plant or lying on the ground. Even with no endosperm remaining, the insect awaits the formation of new fruits. The fungi that experience the greatest growth, those which are thus considered to have an advantage over others, are Cladosporium spp., F. oxysporum, and F. solani, because sporulation on the surface of the fruits is extensive. Many scolytids are called ambrosial beetles because they feed almost exclusively on fungi grown in the galleries that they build (Berryman 1989). Furthermore, the term ambrosial fungus is used for those fungi that grow in tunnels constructed by ambrosial beetles and on which larvae and adults feed (Hawksworth et al. 1995). The fungi use the insect as transportation and can even feature a special organ that facilitates attachment to the insect and eventual

7 498 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 97, no. 3 contact with the gallery walls in the fruit endosperm. No body structure on borers has been detected that carries out this function. A borer generation typically develops before food runs out, allowing for dispersal to a new fruit. They build the gallery in one seed Þrst and then in the other, a strategy that facilitates allocation of food to their progeny. H. hampei seems to be a polyphagous mycetophage based on observation of its gut. Beaver (1989) stated that most ambrosial Coleoptera are primitively polyphagous, a reasonable hypothesis because the insectõs energy expense is lower. Although F. solani is found in borer tunnels (Morales-Ramos et al. 2000), we do not consider it the only ambrosial borer fungus because other fungi establish in the tunnel Þrst. F. solani was the last fungus to occur as a part of a fungal community succession inside the gallery. Ambrosial fungi grow as mycelia or yeast and are transported by the insect from an old to a new gallery (Beaver 1989). A recently constructed gallery is clean, free of feces and is where the eggs are placed. The bacteria and yeast observed during analysis of gallery and gut contents are most likely the result of constant borer feeding within the galleries. Ambrosial fungi are pleomorphic, changing from mold to the dense yeast that prevails throughout the tunnel (Batra and Batra 1967). Disinfection during the experiments permitted little fungal survival on the outside of the borer. Most fungi were found in the gut (11 species), although photographs (Fig. 2AÐD) indicate their presence on the exterior as well. The entomopathogenic fungus B. bassiana was not found in the gut, but it was in macerated individuals and ripe galleries. That this entomopathogen was found in 2% of macerated individuals may mean that it had succeeded in penetrating the exoskeleton. Although, a high percentage (74%) of live unwashed borers contained it at the end of the experiment, only 26% of B. bassiana strains found on borers managed to survive competition from interacting fungi. One of the fungi with the greatest survival ability was F. solani: at initial incubation, it was present in 58% of the petri dishes and in 82% when incubation ended. Using specimens collected in the Þelds of Chiapas, Mexico, and Allada, Benin (Dahomey), Rojas et al. (1999) isolated F. solani from 60% of H. hampei females fed an artiþcial diet for three generations. In our experiments, the highest percentage of F. solani was 58%, and this was obtained using live, unwashed borers. Macerated borer guts, in contrast, had very low percentages of this fungus (4Ð6%). F. solani is found as a parasite on a small percentage of unattacked ripe fruits, but the percentage rises on attacked fruits. Fungal presence in the endosperm is only possible when the fruit has been entered by the borer or is in an advanced state of decomposition. None of the fungi obtained in the experiments was found in healthy endosperm. Thus, interaction between H. hampei and fungi damages coffee fruit. The results obtained suggest interference between F. solani and B. bassiana. This does not necessarily indicate that H. hampei is growing F. solani, because P. echinulatum also limited B. bassiana growth. H. hampei may be using F. solani and P. echinulatum to protect itself from B. bassiana. ClariÞcation of how these three fungi interact would be of great value to research on biological control so that damage by this insect can be mitigated. Interactions between the plant, the borer, and the mycobiota are more complex than what has been reported in the literature. Our results indicate that the borer is not the only problem to be confronted by the coffee manager. An important additional effect is the introduction of diverse mycobiota with repercussions on the health of the coffee plant and potential consumers. Acknowledgments We thank Tiburcio Láez for taking scanning electron microphotographs and Marõ a Luisa Castillo, Yrma López, and Bruce Campbell for comments and suggestions on the manuscript. This research was funded by the Institute of Ecology, A.C. ( to G.C. and to A.B.) in Xalapa, Veracruz, México. References Cited Baker, P. S Some aspects of the behavior of the coffee berry borer in relation to its control in south Mexico (Coleoptera, Scolytidae). Folia Entomol. Mex. 61: 9Ð24. Barrera, J. F., P. S. Baker, J. E. Valenzuela, and A. Schwarz Introducción de dos especies de parasitoides africanos a México para el control biológico de la broca del café, Hypothenemus hampei (Ferrari) (Coleoptera: Scolytidae). Folia Entomol. Mex. 79: 245Ð247. Batra, S.W.T., and L. R. Batra The fungus gardens of insects. Sci. Am. 217: 112Ð120. Beaver, R. A Insect-fungus relationships in the bark and ambrosia beetles, pp. 121Ð143. In N. Wilding, N. M. Collins, P. M. Hammond, and J. F. Webber. [eds.], InsectÐ fungus interactions. Academic, London, England. Benavides, M. P., A. E. Bustillo, E. C. Montoya, R. Cárdenas, and C. G. Mejía Participación del control cultural, quõ mico y biológico en el manejo de la broca del café. Rev. Colombiana Entomol. 28: 161Ð165. Berryman, A. A Adaptive pathways in scolytid-fungus associations, pp. 145Ð156. In N. Wilding, N. M. Collins, P. M. Hammond, and J. F. Webber. [eds.], InsectÐfungus interactions. Academic. London, England. Beti, J. A., T. W. Phillips, and E. B. Smalley Effects of maize weevils (Coleoptera: Curculionidae) on production of aßatoxin B 1 by Aspergillus flavus in stored corn. J. Econ. Entomol. 88: 1776Ð1786. Booth, C The genus Fusarium. Commonwealth Mycological Institute Kew, Surrey, England. De la Rosa, W., J. L. Godínez, R. Alatorre, and J. Trujillo Biological activity of Þve strains of Metarhizium anisopliae upon the coffee berry borer Hypothenemus hampei (Col.: Scolytidae). Entomophaga 40: 403Ð412. De la Rosa., W., R. Alatorre, J. Trujillo, and J. F. Barrera. 1997a. Virulence of Beauveria bassiana (Deuteromycetes) strains against coffee berry borer (Coleoptera: Scolytidae). J. Econ. Entomol. 90: 1534Ð1538. De la Rosa, W., J. L. Godínez, R. Alatorre, and J. Trujillo. 1997b. Susceptibilidad del parasitoide Cephalonomia a diferentes cepas de Beauveria bassiana y Metarhizium anisopliae. Southwest. Entomol. 22: 233Ð242.

8 May 2004 CARRIÓN AND BONET: MYCOBIOTA ASSOCIATED WITH H. hampei 499 Domsch, K. H., W. Gams, and T.-H. Anderson Compendium of soil fungi. Academic, London, England. Ellis, M. B Dematiaceous Hyphomycetes. CAB, Wallingford, United Kingdom. Feng-Yan, B., and C. Quing-Tao Fusarium species on some insects from China. Acta Mycol. Sin. 10: 120Ð128. Frisvad, J. C Fungal species and their speciþc production of mycotoxins, pp. 239Ð349. In R. A. Samson, and E. S. van Reenen-Hoekstra [eds.], Introduction to foodborne fungi. Centraalbureau voor Schimmelcultures, 3rd ed. Institute of the Royal Netherlands Academy of Arts and Sciences, Amsterdam, The Netherlands. Haraprasad, N., S. R. Niranjana, H. S. Prakash, H. S. Shetty, and S. Wahab Beauveria bassiana: a potential mycopesticide for the efþcient control of coffee berry borer, Hypothenemus hampei (Ferrari) in India. Biocontrol Sci. Technol. 11: 251Ð260. Hawksworth, D. L., P. M. Kirk, B. C. Sutton, and D. N. Pegler Ainsworth & BisbyÕs dictionary of the fungi, 8th ed. CAB, Wallingford, United Kingdom. Le Pelley, R. H Las plagas del café. Ed. Labor. Barcelona, España. López-Vaamonde, C., and D. Moore Developing methods for testing host speciþcity of Phymastichus coffea LaSalle (Hym.: Tetrastichinae), a potential biological control agent of Hypothenemus hampei (Ferrari) (Col.: Scolytidae) in Colombia. Biocontrol Sci. Technol. 8: 397Ð 411. Mejía, J. W., and D. F. López Estudio de necesidades en manejo integrado de la broca del café en el departamento de Antoquia. Rev. Colombiana Entomol. 28: 167Ð 173. Morales-Ramos, J. A., M. G. Rojas, H. Sittertz-Bhatkar, and G. Saldaña Symbiotic relationship between Hypothenemus hampei (Coleoptera: Scolytidae) and Fusarium solani (Moniliales: Tuberculariaceae). Ann. Entomol. Soc. Am. 93: 541Ð547. Nagao, T., Y. Shimada, A. Sugihara, and Y. Tominaga Amino acid residues contributing to stabilization of Fusarium heterosporum lipase. J. Biosci. Bioeng. 89: 446Ð 450. Nagao, T., Y. Shimada, A. Sugihara, A. Murata, S. Komemushi, and Y. Tominaga Use of thermostable Fusarium heterosporum lipase for production of structured lipid containing oleic and palmitic acids in organic solvent-free system. J. Am. Oil Chem. Soc. 78: 167Ð172. Nagao, T., Y. Shimada, A. Sugihara, and Y. Tominaga Increase in the stability of Fusarium heterosporum lipase. J. Mol. Catal. B Enzymol. 17: 125Ð132. Northolt, M. D., and P.S.S. Soentoro Fungal growth on foodstuffs related to mycotoxin contamination, pp. 231Ð238. In R. A. Samson and E. S. van Reenen-Hoekstra [eds.], Introduction to food-borne fungi. Centraalbureau voor Schimmelcultures, 3rd ed. Institute of the Royal Netherlands Academy of Arts and Sciences, Amsterdam, The Netherlands. Pérez López, E. J., F. J. Posada Flórez, and M. T. González García Patogenicidad de un aislamiento de Fusarium sp. encontrado infectando la broca del café, Hypothenemus hampei. Rev. Colombiana Entomol. 22: 105Ð 111. Peterson, S. W., J. Pérez, F. E. Vega, and F. Infante Penicillium brocae, a new species associated with the coffee berry borer in Chiapas, Mexico. Mycologia 95: 141Ð147. Robledo, M., S. Marín, and A. Ramos Contaminación natural con micotoxinas en maõ z forrajero y granos de café verde en el Estado de Nayarit (México). Rev. Iberoamericana Micol. 18: 141Ð144. Rojas, M. G., J. A. Morales-Ramos, and T. C. Harrington Association between Hypothenemus hampei (Coleoptera: Scolytidae) and Fusarium solani (Moniliales: Tuberculariaceae). Ann. Entomol. Soc. Am. 92: 98Ð100. StatSoft, Inc Statistica for Windows, computer program manual. Tulsa, OK. Villacorta, A., and J. F. Barrera Techniques for mass rearing of the parasitoid Cephalonomia stephanoderis (Hymenoptera: Bethylidae) on Hypothenemus hampei (Coleoptera: Scolytidae) using an artiþcial diet. Vedalia 3: 45Ð48. Wellman, F. L Dictionary of tropical American crops and their diseases. The Scarecrow Press, Inc., Metuchen, NJ. Received 29 September 2003; accepted 23 December 2003.