Microbiology of the Larval Provisions of the Stingless Bee, Trigona hypogea, an Obligate Necrophage

Abstract

Five species of the genus Bacillus were the only microbes found in the larval provisions of Trigona hypogea, a bee for which dead animal tissue has replaced pollen as the sole protein source. Bacillus pumilus was the most frequent isolate followed by B. megaterium, B. subtilis, B. circulans, and B. licheniformis. These bacteria were metabolically active and produced many enzymes including proteases, amylases, and esterases. They may have a fundamental role in the metabolic conversion, fermentation, and preservation of the food of perennial colonial insects that rely on stored food in tropical environments. THE MELIPONINES ARE A PANTROPICAL GROUP of eusocial, stingless bees native to tropical forest. They live in perennial colonies and have a single, physogastric, flightless queen. Nests of stingless bees are usually subterranean or aboreal and are occasionally associated with ant or termite nests. The entrance is either a small opening just large enough to allow one bee at a time to pass or a more elongated tube-like entrance that may have a flared opening. Although colonies generally consist of fewer than several thousand adults, foraging bees visit a wide variety of resources, and large amounts of food are stored in the nests (Schwarz 1948, Wille 1983, Roubik 1983). Plant resins, other exudates, secreted wax, and mud used for nest construction, combined with the extreme selectivity of nesting sites, make the nest waterproof and highly resistant to attack from predators. Microbes associated with bees may be parasites, commensals, or mutualistic. Some can infect bees, and spoilage microbes can destroy pollen stores (Batra et a/. 1973). Also, the presence of microorganisms in the nests of perennial colonial insects that rely on stored food in humid, tropical environments may be fundamentally important in the metabollc conversion or preservation of the food. Stingless bees continuously forage for nectar and pollen in tropical forests, and colony food stores are kept in separate but grouped dusters of storage pots. The pots are built of secreted wax mixed with plant resins, the latter possibly serving as a biocidal agent (Michener 1974, Roubik 1983). Thus, these bees resemble some groups of bumble bees (Bombus spp.) with regard to food storage behavior but, unlike Bombus, construct or remove large numb rs of storage pots within the nest and do not rely on the availability of vacated brood cells for modification to food storage containers. Trigona workers provision cells with pollen and honey mixed with glandular secretions prior to oviposition by the queen. Unlike Apis honey bees, stingless bees use mass provisioning, and each larva is provided with the total amount of food required for adult development before the cell is sealed. Resources collected by foraging Trigona workers are much more diverse than those collected by Apis workers and include pollen, nectar, fungal spores, exudates from homopteran insects, plant resins containing terpenes and other sap exudates, mud, vertebrate feces, and carrion from vertebrate carcasses. True obligate necrophagy was first reported by Roubik (1982) for Trigona (Trigona) hypogea Silvestri. Dead animal tissue has replaced pollen as the sole protein source for these unique bees. Trigona hypogea bees lay down odor trails between the nest site and the animal carcass, which can be located within a matter of hours, and utilize massive recruitment to send hundreds of foragers to the site. Bees on vertebrate carcasses were observed to actively attack and drive off flies and wasps that had begun to feed. Thus, dipteran oviposition and larval competition for the carcass could not occur. Roubik (1982) has observed these bees feeding on the carcasses of toads, frogs, snakes, lizards, birds, fish, and monkeys. Trigona hypogea bees collect nectar and produce and store copious amounts of honey within large colonies (Roubik 1983). Unlike all other known Trigona species, they never collect pollen from flowers. The nests lack stored pollen in either the storage pots or in the provisions 28 BIOTROPICA 17(1): 28-31 1985 I Mention of a proprietary product or company name does not constitute an endorsement of this product by the U.S. Department of Agriculture. 2 Received 2 December 1983, accepted 24 February 1984. This content downloaded from 157.55.39.231 on Wed, 05 Oct 2016 04:23:37 UTC All use subject to http://about.jstor.org/terms in larval cells. Morphological adaptations of T. hypogea indude reduced corbiculae on the hind legs and strong mandibles with five enlarged teeth (Roubik 1982). Bees can be observed to bite off, masticate, and consume large pieces of musde and other tissue from the carcasses. They do not carry externally unmodified pieces of animal flesh back to the nest. The surface of the animal carcass often appears to glisten, and this is apparently due to salivary and other secretions produced by the feeding worker bees. We (Roubik and Buchmann unpublished) hypothesize that the younger worker bees further process the pre-digested material in their massive hypopharyngeal glands by a mechanism analogous to that of nurse bees of Apis mellifera which consume pollen and produce brood food or royal jelly via the hypopharyngeal glands. These younger Trigona bees then transfer the brown glandular secretion to large cerumen storage pots which are sealed when filled. When a section of brood comb has many empty larval cells, workers remove glandular material from the larger pots to provision individual brood cells. The queen then oviposits on the surface of the provisions, and the cells are capped. The brownish liquid secretion obtained from larval cell provisions and storage pots is highly viscous and has a pH of 3.0-4.0 which is very similar to the harsh, acidic Apis royal jelly. Other analyses (Buchmann, Schmidt, Roubik, and Law unpublished) have revealed a high protein concentration (-20%), large numbers of amino acids and hydrocarbons, and lipid and carbohydrate levels very similar to those found in royal jelly of Apis. Considering the rapidity with which dead animal flesh putrefies in the tropics due to microbial activity, the metabolic conversion of this rich protein foodstuff is the only viable strategy for long-term food storage by these bees. Many nests have been examined to date, and no evidence of rotten carrion has ever been found within a larval cell or large storage pot. Because of the distinctive fermentation odor, larval provisions of T. hypogea were examined for bacteria, yeasts, and molds to characterize microbes that might play a role in metabolic conversion or preservation of food. MATERIALS AND METHODS On February 18, 1982, brood provisions of T. hypogea were collected from a nest in Panama. The nest was opened by removing a side of a log, the batumen was carefully removed, the newest comb area and sealed storage pots were located and exposed, and the larval cells and intact storage pots were rapidly dissected indoors. The cells were opened with sterile forceps, and the contents of each cell were removed with a separate sterile capillary pipet. The pipets were sealed with Critoseal, placed in sterile test tubes, frozen immediately, and kept frozen during transport and until analyzed in Tucson. Two 20-30 ,ul samples of pooled larval provisions were inoculated onto duplicate plates of blood agar, nutrient agar (Difco), TYG agar (40 g Difco tryptic soy agar, 3 g yeast extract, 10 g glucose, and 1 liter distilled water), Czapek solution agar (Difco), and YM-1 agar (Wickerham 1951) to isolate bacteria, molds, and yeasts. TYG agar was developed by one of us (B.J.L.) and is an excellent growth medium for bacteria associated with bees and plants. To test for the presence of anaerobes, duplicate tubes of thioglycollate medium without indicator (Difco) were inoculated. One plate or tube of each medium was incubated at 250 and the other at 3 7C under aerobic conditions. During a 2-week incubation period, plates and tubes were periodically examined for microbial growth, and any colonies that developed were restreaked onto plates of TYG agar to test for pure cultures. These plates were incubated at 37?C. Cell suspensions from each microbial colony were stained by the Gram method, and all were found to belong to the genus Bacillus. This was confirmed by microscopic examination of the provisions. In this manner, the size, shape, and location of the spores within the sporangia were noted, and the morphology of the vegetative cells was determined. The organisms were maintained on TYG agar slants and were tested and identified according to Gordon et al. (1973) except that motility was determined in motility test medium (BBL) rather than microscopically. Additional information on enzymatic activity of each of the isolates was obtained by testing for 19 enzymes using the API ZYM system (Analytab Products) according to the manufacturer's directions.