HAL Id: hal-00890763https://hal.archives-ouvertes.fr/hal-00890763

Submitted on 1 Jan 1989

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Microbiology of pollen and bee bread : taxonomy andenzymology of molds

M. Gilliam, D. B. Prest, B. J. Lorenz

To cite this version:M. Gilliam, D. B. Prest, B. J. Lorenz. Microbiology of pollen and bee bread : taxonomy and enzy-mology of molds. Apidologie, Springer Verlag, 1989, 20 (1), pp.53-68. �hal-00890763�

Original article

Microbiology of pollen and bee bread :taxonomy and enzymology of molds*

M. Gilliam, D. B. Prest B. J. Lorenz

US Department of Agriculture, Agricultural Research Service, Carl Hayden Bee Research Center,2000 E. Allen Road, Tucson, AZ 85719, USA

(received 26-5-1988, accepted 18-8-1988)

Summary — One-hundred and forty-eight molds were isolated from the following samples ofalmond, Prunus dulcis, pollen : floral pollen collected by hand; corbicular pollen from pollen trapsplaced on colonies of honey bees, Apis mellifera, in the almond orchard; and bee bread stored incomb cells for one, three, and six weeks. The majority of molds identified were Penicillia (32%),Mucorales (21%), and Aspergilli (17%). In general, the number of isolates decreased in pollen as itwas collected and stored by the bees. Each type of pollen sample appeared to differ in regard tomold flora and dominant species. Aureobasidium pullulans, Penicillium corylophilum, Penicilliumcrustosum, and Rhizopus nigricans were among the molds that may have been introduced by beesduring collection and storage of pollen. Mucor sp., the dominant mold in floral pollen, was not foundin corbicular pollen and bee bread. Tests for 19 enzymes revealed that most of the molds producedcaprylate esterase-lipase, leucine aminopeptidase, acid phosphatase, phosphoamidase, B-glucosi-dase, and Macetyl-B-glucosaminidase. Thus, enzymes involved in lipid, protein and carbohydratemetabolism were produced by pollen molds. Molds could also contribute organic acids, antibioticsand other metabolites.

pollen - bee bread - molds

Résumé — Microbiologie du pollen et du pain d’abeilles : taxonomie et enzymologie des moi-sissures. A l’aide de divers milieux microbiologiques possédant des pH différents, on a isolé 148moisissures des échantillons suivants de pollen d’amandier, Prunus dulcis : pollen de fleurs récoltéà la main; pollen en pelotes prélevé dans les trappes à pollen posées sur des colonies d’abeilles(Apis mellifica) dans un verger d’amandiers; et pain d’abeilles stocké dans les cellules des rayonsdurant une, 3 et 6 semaines. La majorité des moisissures identifiées sont des Penicillia (32%), desMucorales (21 %) et des Aspergillia (17%). C’est le pollen de fleurs qui est le plus riche en isolats,mais le plus pauvre en espèces. En général le nombre d’isolats diminue dans le pollen quand il estrécolté et stocké par les abeilles. Chaque type d’échantillon pollinique semble différer des autrespar la flore de moisissures et les espèces dominantes. Puisque les moisissures sont identifiéesd’après les besoins de croissance et la caractérisation microscopique et macroscopique des struc-tures morphologiques, les données biochimiques ne proviennent pas des tests taxonomiques. On a

* Mention of a trademark, proprietary product, or vendor does not constitute a guarantee or warrantyby the USDA and does not imply its approval to the exclusion of other products or vendors that mayalso be suitable.

donc analysé 19 enzymes chez 78 isolats, représentant 28 espèces, par le système API ZYM.Aucune moisissure ne produit de trypsine, de 8-glucuronidase, ni de a-mannosidase. La plupartdes moisissures produisent de la caprylate lipase-estérase, de la leucine aminopeptidase, de laphosphatase acide, de la phosphoamidase, de la 8-glucosidase et de la N-acétyle-l3-glucosamini-dase. Les moisissures du pollen produisent donc des enzymes impliqués dans le métabolisme desprotéines, des lipides et des glucides.

Ces résultats suggèrent que la flore de moisissures du pollen en pelotes et du pain d’abeillespeut résulter d’inoculations microbiennes par les abeilles et de modifications chimiques du pollendues aux substances ajoutées par les abeilles lors de la régurgitation du contenu du jabot et à lafermentation microbienne, qui permet à certaines espèces de survivre et à d’autres pas. Même si,dans nos échantillons, les moisissures étaient plus nombreuses que les espèces de Bacillus et leslevures, le pollen est rarement envahi par elles. Parce qu’elles sont susceptibles de fournir desenzymes, des acides organiques, des antibiotiques et d’autres métabolites, les moisissures méri-tent des études plus approfondies.

pollen - pain d abeilles - moisissures

Zusammenfassung — Mikrobiologie von Pollen und Bienenbrot : Taxonomie und Enzymolo-gie des Schimmels. Unter Verwendung verschiedener mikrobiologischer Medien mit unterschied-lichem pH-Wert wurden 148 Schimmelpilze von den folgenden Proben von Mandelpollen (Prunusdulcis) untersucht : Blütenpollen (von Hand gesammelt), Pollenhöschen (aus Pollenfallen an Bie-nenvölkern [Apis mellifera] im Mandelbaumgarten) und Bienenbrot, das 1, 3 und 6 Wochen in derWabe gespeichert war.

Die am häufigsten auftretenden Schimmelpilze waren Penicillia (32%), Mucorales (21 °l) undAspergilli (17%). Der Blütenpollen lieferte die meisten Isolate, aber die wenigsten Arten. Im all-gemeinen nahm die Anzahl der lsolate im Pollen mit dem Sammeln und Speichern durch die Bieneab. Jeder Pollentyp schien im Hinblick auf die Schimmelflora und die dominierenden Arten ver-schieden zu sein.

Da die Schimmelpilze aufgrund ihrer Wachstumserfordernisse sowie mikrosltopischer wiemakroskopischer Charakterisierung von morphologischen Strukturen identifiziert werden, erhältman von taxonomischen Untersuchungen keine biochemischen Daten. Daher wurden 78 Isolate,die 28 Arten repräsentieren, mit dem API ZYM-System auf 19 Enzyme analysiert. Kein Schimmel-pilze produzierte Trypsin, β-Glucuronidase oder a-Mannosidase. Die meisten Schimmelpilze produ-zierten Caprylat-Esterase-Lipase, Leucin-Aminopeptidase, saure Phosphatase, Phosphoamidase,β-Glucosidase und N-Acetyl-β-Glucosaminidase. Dies bedeutet, daß die untersuchten Pollen-schimmel Enzyme des Protein-, Fett- und Kohlehydratstoflwechsels produzieren.

Diese Ergebnisse deuten darauf hin, daß die Schimmelflora im Höselpollen und im Bienenbrotein Ergebnis folgender Einflüsse ist : mikrobielle Inokulation und chemische Veränderung des Pol-lens durch Zugabe von Honigmageninhalt durch die Biene; Drüsensekretion sowie mikrobielle Fer-menta6on, die manche Schimmelarten tolerieren, andere nicht. Obwohl Schimmelpilze in unserenProben weit zahlreicher waren als Bacillus spp. oder Hefen, wurde der Pollen selten von Schimmelüberwachsen. Die Schimmelpilze sollten als potentielle Spender von Enzymen, organischen Säu-ren, Antibiotika und anderer Metabolite intensiver untersucht werden.

Pollen - Dienenbrot - Schimmel

Introduction

Studies have shown for many years that

pollen and bee bread, that is pollen storedin comb cells of the hive, differ biochemi-cally, and extensive analyses have been

conducted on various floral and bee-col-lected (corbicular) pollens. The conver-sion of pollen to bee bread has often beenpostulated to be the result of microbial

action, principally a lactic acid fermenta-tion caused by bacteria and yeasts(Foote, 1957; Haydak, 1958). However,

the chemical and biochemical changesoccurring in pollen as it is collected andstored by honey bees, Apis mellifera, arenot clearly understood, and relatively littleis known about the microbiology of pollenand bee bread.

To better understand the nutrition of

honey bees, we studied the chemical, bio-chemical, and microbiological compositionof pollen from a single plant species be-fore, during, and after storage in combcells. Previous papers on the subject byresearchers at the Carl Hayden BeeResearch Center reviewed earlier workand reported results concerning yeasts(Gilliam, 1979a); Bacillus ssp. (Gilliam,1979b); fatty acids, sterols, vitamins, titra-table acidity, minerals (Loper et al., 1980);and protein content, amino acids, selectedenzymes, pH, and 10-hydroxy-!2-deca-noic acid (Standifer et al., 1980). In all thiswork, the same samples of almond, Pru-nus dulcis (Prunus communis), pollenwere utilized.

Even though molds are widely knownfor their abilities to degrade and to synthe-size numerous compounds including theproduction of many materials important tothe drug, food and chemical industries,they have received scant attention in api-cultural research concerned with pollenand bee bread. Early mycological re-

search recognized that certain molds arecommon saprophytes on and inside

honey bees and brood combs, but effortswere concentrated on dead bees; combs,particularly from dead colonies; and moldypollen (Betts, 1912; Burnside, 1927).Betts (1912) reported a species of Clado-sporium as well as Mucor erectus in corbi-cular pollen and Beftsia alvei, Eremascusfertilis, Gymnoascus setosus, Oosporafavorum, and Penicillium crustaceum in

pollen stored in combs. She noted that

honey appeared to be immune to attacksof molds. Burnside (1927) stated thatmost of the fungi collected by widespread

foraging of honey bees are probablyunable to become established within thebee or the hive. He found that Penicilliawere the most common molds within the

hive, Aspergilli occurred less frequently,and species of Mucor did not grow well onbrood combs.

Fungus-caused spoilage of provisionsand mortality of honey bees are rare

(Batra et aL, 1973). Recently Gilliam andVandenberg (1988) reviewed the literatureon fungi pathogenic or detrimental to

honey bees. Only Ascosphaera apiswhich causes chalkbrood disease in

honey bees is of economic importance.The pollen mold, Bettsia alvei, is not aserious problem since it does not growwell in cells that are filled with pollen andfinished with a layer of honey on top(Skou, 1972).

Burri (1947) stated that pollen is germ-free in blossoms that have not opened aswell as in opened blossoms if uncontami-nated by insect visitation or air currents.Neither of the two microbiological studiesof pollen and bee bread (Chevtchik, 1950;Pain and Maugenet, 1966) gave data onmolds, although Chevtchik (1950) men-tioned B. alvei as a possible consumer oflactic acid in bee bread.

Arizan et al. (1967) isolated Absidia

ramnosa, Aspergillus flavus, Aspergillusfumigatus, Aspergillus niger, Aspergillusterreus, Aspergillus versicolor, Mucor

mucedo, Penicillium clavigerum, Penicil-lium purpurogenum, Rhizopus nigricansand Trichothecium roseum from Indiancorn pollen collected by machine. Saingeret aL (1978) reported that Altemariaalternata was the most common isolate in

pollen from 3 herbaceous annual plants.Other molds isolated were Aspergillusflavus, Aspergillus luchuensis, Aspergil-lus nidulans, Aspergillus sulphureus, A.versicolor, Cladosporium oxysporum,Epicoccum purpurascens, Fusarium oxy-

sporum, Monilia fructigena, Monilia sito-phila, Monilia sp., Rhizopus sp., R. nigri-cans and Trichoderma viride. Molds isola-ted from bee bread by Egorova (1971)were A. flavus, A. versicolor, Mucor

alboalter, Penicillium granulatum, Penicil-lium solitum and Sporotrichum olivecum.

The present paper reports the resultsof the isolation and identification of moldsfrom almond floral pollen, corbicular pol-len, and bee bread stored in comb cells;analyses of enzymes produced by select-ed isolates; and comparison of speciesisolated with those previously reportedfrom honey bees in Arizona.

Materials and Methods

Details concerning bee colonies and collect-ions of pollen and bee bread are given by Gil-liam (1979a), Loper etal. (1980), and Standiferet al. (1980). The following samples of almondpollen were examined : fresh floral pollen col-lected by hand; corbicular pollen containing99.8% almond pollen from pollen traps placedon bee colonies in the almond orchard; andbee bread stored in cells for 1, 3, and 6 weeksin newly drawn combs of colonies of beesmaintained in a greenhouse. Each sample wasdivided into 4 sub-samples of approximately0.75 g each. Then each of the 4 sub-sampleswas homogenized by hand in 2.5 ml of sterile0.85% NaCi in a glass tissue grinder. Thehomogenates were plated (0.1 ml) in duplicateon acidified yeast extract-malt extract agarcontaining 1% glucose, pH 3.7—3.8 (Miller e tal., 1976); mycological agar with low pH (Difco,pH 4.8); nutrient agar (Difco) acidified with 0.1N HCI to pH 5.0; and eugon agar (Difco, pH7.0). One plate from each sub-sample wasincubated at 25°C and one at 37°C. All wereincubated under aerobic conditions excepteugon agar plates which were placed in 4%

C02. During a 2-week incubation period, plateswere examined periodically for mold growth.

When molds appeared, they were trans-ferred to plates of Czapek solution agar or maltextract agar (Difco) to allow time for sporulationand to test for purity. These plates were incu-bated at 25°C under aerobic conditions. Pure

cultures of isolates were lyophilized for preser-vation until tests were conducted. Molds weretested and identified according to Neergaard(1945), DeVries (1952), Cooke (1959), Ames(1961), Booth (1961), Morton and Smith (1963),Ellis (1965), Raper and Fennell (1965), Raperand Thom (1968), Zycha et al. (1969), Kendrickand Carmichael (1973), Samson (1974), vonArx (1975) and McGinnis et al. (1986).

Since molds, in contrast to bacteria and

yeasts, are identified on the basis of growthrequirements and microscopic and macroscop-ic characterization of morphological structures,biochemical data do not result from tests foridentification. Therefore, selected isolates weretested for 19 enzymes with the API ZYM sys-tem (Analytab Products) using the methods ofBridge and Hawksworth (1984). Suspensionsof some isolates such as Mucorssp. and Alter-naria tenuis were sonicated for 1—10 min to

separate fungal spores before inoculation intothe API ZYM strips. Also, malt extract agarand/or potato dextrose agar (Difco) were usedto prepare inocula of some cultures when

growth appeared to be less than optimal onCzapek solution agar.

Results

No attempt was made to determine whe-ther spores or mycelial elements wereisolated from pollen and bee bread. How-ever, molds were isolated on all 4 mediaused (Table I). Seventy-seven percent ofthe isolates were from media incubated at

25°C, and 23% were from media at 37°Cwhich is near the brood nest temperatureof 34°C (Dunham, 1929). However, theoptimum temperatures for most fungi arein the range of 20-30°C (Alexopoulos,1962). Isolations from floral pollenincreased with decreasing pH of themedia. In contrast, the highest percent ofisolations from corbicular pollen from thepollen trap was on acidified nutrient agarwith a pH of 5.0. Few isolations weremade from floral or corbicular pollen oneugon agar with a pH of 7.0, but thissituation changed with bee bread. The

percent of isolations on various mediawas similar in bee bread samples storedfor one week and for 3 weeks. Theseresults indicate that a variety of mediawith different chemical compositions andpH values incubated at both 25°C and37°C aerobically and under C02 shouldbe used for determining mycoflora of pol-len and bee bread.

One-hundred forty-eight molds wereisolated from pollen and bee bread, ofwhich 139 were identified (Table II). Over-all, the majority of molds identified werePenicillia (32%), Mucorales (21%), andAspergilli (17%). Floral pollen yieldedthe highest number of isolates but thesmallest number of species. In contrast,bee bread stored in comb cells for 3weeks had the fewest isolates but the

greatest number of different species. In

general, the number of isolates decreasedin pollen as it was collected and stored bybees.

The most frequent isolate was Mucorsp., associated exclusively with floral pol-len. All 19 isolates appeared similar, butspecies identification was not made. Theywere characterized by the production of

coenocytic mycelium, some sympodiallybranched sporangiophores containing

gemmae (chlamydospores) and yellowglobules, globose sporangia with finelyspinose walls, columella with a distinct

collarette, and elliptical smooth sporangio-spores. The other molds identified fromfloral pollen were found in at least oneadditional pollen source.

The second most frequently isolatedmold was Penicillium corylophilum. It wasassociated with corbicular pollen and allbee bread sources but not with floral pol-len. This was also the case for R. nigri-cans. Other species which first appearedin corbicular pollen and were then foundin bee bread were Aureobasidium pullu-lans, Cladosporium herbarum, Penicil-lium chrysogenum, and Penicillium crus-tosum. Aspergillus niger was a frequentisolate and was found in all types of pollensamples except bee bread stored for oneweek and was isolated most often frombee bread stored for 6 weeks. Another

frequently encountered mold was Penicil-lium cyclopium which was most abundantin bee bread stored for one week and wasfound in all sample types except corbicu-lar pollen. Cladosporium cladosporioideswas the only species isolated from all

sample types but was most prevalent infloral pollen. Molds isolated from more

than one source of bee bread but not flo-ral or corbicular pollen were Aspergillusamstelodami, A. flavus, and Paecilo-

myces varioti. Alternaria tenuis was foundin floral pollen and bee bread stored forone week, and Penicillium brevi-compac-tum was in floral and corbicular pollen butnot bee bread. Other than Chaetomiumelatum which was found in bee breadstored for 6 weeks, the remaining speciesthat were identified appeared once in onlyone type of pollen source other than floralpollen. Thus, each type of pollen sampleappeared to differ in regard to mold floraand dominant species since the predom-inant mold in almond floral pollen wasMucor sp.; in corbicular pollen, P. corylo-philum and P. crustosum were most com-mon; in bee bread stored in comb cells forone week, P. cyclopium and P. corylophi-lum were the most numerous isolates; inbee bread stored for 3 weeks, there wasno obvious dominant species; and in bee

bread stored for 6 weeks, A. niger wasmost common.

Eight unidentified molds in Table 11

were non-viable after lyophilization. Wewere also unable to assign the unidenti-fied Ascomycete to genus. It producedostiolate dark ascocarps with unbranchedterminal hairs. The asci were cylindricaland contained 4 ascospores. Ascosporeswere ellipsoidal to ovoid, smooth, non-apiculate, and yellow-brown to brown.Since it was difficult to determine by lightmicroscopy whether the ascospores weresingle or double-pored, scanning electronmicroscopy was used to reveal that themajority were single-pored, althoughdouble-pored ascospores were also pres-ent.

To determine enzymatic activity of themolds from pollen and bee bread,attempts were made to test at least onestrain of each species and more strains of

the species frequently isolated. However,Cladosporium sphaerospermum did not

survive lyophilization after identificationand could not be tested. Thus, 78 isolatesrepresenting 28 species were each anal-yzed for 19 enzymes. A total of 113 com-plete tests were conducted owing to repli-cation of tests, use of additional media for

preparing the inocula of strains which didnot grow well on Czapek solution agar,duplicate tests which were incubated aslong as 24 h, and tests comparing enzy-mology of spore and mycelial inocula ofselected strains. Aspergillus amsteloda-mi, Chaetomidium pilosum, Chaetomiumelatum, Thielavia sepedonium, Xylohy-pha bantiana, and the unidentified Asco-mycete failed to grow well enough onCzapek solution agar to yield sufficientinocula for API ZYM tests and were there-fore grown on potato dextrose agar andmalt extract agar and then tested for

enzymes. Other selected species were

also grown on various media, and theenzymology was compared to inoculafrom Czapek solution agar. Results oftests on inocula of the same strain pre-pared on different media gave the sameresults except that the concentrations ofone or two of the enzymes produced werein a few cases slightly higher when thegrowth medium was potato dextrose agarcompared to malt extract agar. These dif-ferences could be related to improvedgrowth of the molds and/or to the mediacomposition. Bridge and Hawksworth

(1984) also found some minor variationswith different media. We also noted a fewsimilar minor variations when the incuba-tion time was extended for 24 h for strainsof R. nigricans. Mycelial inocula producedthe same enzymes as spore suspensions,although in smaller quantities.

Results of enzymatic activities basedon identities of the isolates regardless ofthe pollen source are shown in Tables

III-VI. All 40 Penicillia tested producedleucine aminopeptidase, acid phosphata-se, phosphoamidase, and f3-glucosidase;none produced cystine aminopeptidase,trypsin, a-galactosidase, f3-glucuronidase,or a-mannosidase (Table 111). Most alsoproduced alkaline phosphatase, caprylateesterase-lipase, and N-acetyl-B-glucos-aminidase. Enzymes produced by Penicil-lia in the highest concentrations (> 20

nmol) were acid phosphatase by P. cory-lophilum, N acetyl-f3-glucosaminidase byP. cyclopium, and f3-glucosidase by P.crustosum.

All Aspergilli tested produced acid

phosphatase, phosphoamidase, B-gluco-sidase, and N acetyl-f3-glucosaminidase;none produced myristate lipase, trypsin,chymotrypsin, f3-glucuronidase, a-manno-sidase, or a-fucosidase (Table IV). Mostalso produced alkaline phosphatase, buty-rate esterase, caprylate esterase—lipase,

and leucine aminopeptidase. One of theA. flavus strains tested was var. columna-ris and gave the same reactions as strainsidentified as A. flavus. Enzymes producedin the highest concentrations (> 20 nmol)by Aspergilli were N-acetyl-B-glucosamini-dase and fi-glucosidase by A. niger, alka-line phosphatase by A. flavus, andalkaline and acid phosphatases and B-

glucosidase by A. versicolor.

All Murocales tested produced leucineaminopeptidase, and most producedacid phosphatase and phosphoamidase(Table V). Otherwise these molds pro-duced few enzymes. The only enzymesproduced in high concentrations (> 20

nanomoles) were phosphoamidase by allstrains of R. nigricans tested, acid phos-phatase by 3 of the 4 R. nigricans strains,and leucine aminopeptidase by Mucorracemosus.

Of the other molds tested, most pro-duced acid phosphatase and f3-glucosida-se (Table VI). Alkaline phosphastase,butyrate esterase, caprylate esterase—

lipase, leucine aminopeptidase, and

phosphoamidase were produced by52—67% of them. Enzymes produced inthe highest concentrations (> 20 nmol) byvarious molds are as follows : alkaline

phosphatase by Aur. pullulans from one-week bee bread; caprylate esterase—lip—ase by C. herbarum from corbicular pollenand by C. cladosporioides; leucine amino-peptidase by Aur. pullulans and Peyrone-lia sp.; acid phosphatase by Alt. tenuis,Arthrinium phaeospermum, Aur. pullu-lans, C. cladosporioides, C. herbarum,Peyronelia sp., Scytalidium sp., and X.bantiana; phosphoamidase by X. bantia-na; a-galactosidase by Aur. pullulans fromone-week bee bread and by C. cladospo-rioides from corbicular pollen; B-glucosi-

dase by Alt. tenuis and T. sepedoniumand by C. cladosporioides from 6-weekbee bread; and a-fucosidase by C. cla-

dosporioides from floral and corbicular

pollen.Enzymology of the 78 molds tested is

presented in Table Vil on the basis of thepollen sources of the isolates. No moldsproduced trypsin, f3-glucuronidase, or a-mannosidase. Few produced myristatelipase, and only one produced chymotryp-sin; these isolates were from corbicular

pollen. Valine aminopeptidase and cystineaminopeptidase were not produced byisolates from 3-week bee bread, nor wasthe latter enzyme produced by molds fromcorbicular or 6-week bee bread. Fewmolds from other pollen sources producedthese two peptidases. Of the glycosidases, a-glucosidase and a-fucosidast.were not associated with molds fromeither 3-week or 6-week bee bread and

were produced by few isolates from otherpollen sources. Both a- and B-galactosi-dases were produced by a higher percentof molds tested from floral pollen thanfrom other pollen sources. No molds fromcorbicular pollen produced B-galactosi-dase.

However, most molds from all pollensources produced caprylate esterase-lipase, leucine aminopeptidase, acid

phosphatase, phosphoamidase, B-gluco-sidase, and N acetyl-f3-glucosamidase. Ahigh percent of the isolates (> 50%) fromall sources gave positive reactions foralkaline phosphatase. This was also thecase with butyrate esterase except forthose molds from one-week bee bread.

Therefore, pollen molds producedenzymes involved in protein, lipid, and

carbohydrate metabolism.

Discussion

Seventy percent of the molds identifiedfrom almond pollen and bee bread wereAspergilli, Mucorales, and Penicillia. Wehave previously isolated from apiariansources in Arizona most of the species ofAspergilli and Penicillia found in pollenand bee bread as well as Aur. pullulans,C. claedosporioides, Mucorales, and

Peyronelia sp. (Gilliam and Prest, 1972,1977, 1987; Gilliam et al., 1974, 1977,1988). Frequent isolates in the presentstudy which are new records of moldsfrom apiarian sources in Arizona are

Alt. tenuis, Penicillium crustosum, andR. nigricans.

Molds which were not present in floralpollen but were frequent isolates from cor-bicular pollen and bee bread may havebeen introduced by the bees during col-lection and storage. The most obviousexamples are Aur. pullulans, P. corylophi-

lum, P. crustosum, and R. nigricans.Conversely, Mucor sp., the dominant moldin floral pollen, was eliminated in corbic-ular pollen and bee bread. Thus, as withyeasts (Gilliam, 1979a) and Bacillus ssp.(Gilliam, 1979b), the mold flora of corbic-ular pollen and bee bread may be theresult of microbial inoculations by beesand chemical changes in pollen resultingfrom additions by bees from regurgitationof honey sac contents and secretions ofglands as well as microbial fermentationwhich allow some species but not othersto survive. Even though molds were morenumerous than yeasts or Bacillus ssp. inour samples, pollen is rarely overgrown bymolds. Potential microbial spoilage of pol-len provisions may be controlled by anti-biotic substances produced by the normalmicroflora, bees, pollen, and/or honey.

Klungness and Peng (1983) examinedcorbicular pollen and bee bread with

scanning electron microscopy and foundno visible evidence of digestion or dam-age to pollen grain walls. They concludedthat microorganisms associated with beebread do not cause destruction of pollenintine or the cytoplasm, that substancesbees add to pollen during collection andstorage function as a preservative, andthat the regurgitation added to pollenprobably allows growth of some microor-ganisms and inhibits the growth of others.They observed that the few fungal sporesthat germinated produced hyphae lessthan 10 pm in length.

If microorganisms are responsible forfermentation and the accompanyingchemical changes of pollen stored in

comb cells by honey bees, the molds maybe a component of the required microbialcomplement. They could contribute anti-biotics, organic acids and enzymes, pro-ducts for which they are utilized industrial-ly. These compounds may limit the growthof deleterious microorganisms and provi-de enzymes for utilization of nutrients.

Indeed, we have found Aspergilli, Muco-rales, Penicillia, and other molds in beebread and guts of worker bees which in-hibit the growth of the chalkbrood fungus(Gilliam et aL, 1988).

Enzymology of our isolates revealedthat the major phosphatase was acid

phosphatase, although alkaline phospha-tase was produced by most isolates

except the Mucorales. Molds apparentlyare not able to participate in the initialbreakdown of long-chain fatty acids asevidenced by the lack of myristate lipase.However, they did produce butyrate es-rase and caprylate esterase—lipase

which act on shorter chain fatty acids.Leucine aminopeptidase was the majoraminopeptidase. Molds did not producetrypsin or chymotrypsin. Phosphoamida-se was produced by most isolates tested.

Results for glycosidases revealed thatMucorales were quite unreactive. How-

ever, most other molds produced B-gluco-sidase which hydrolyzes carbohydratessuch as cellobiose, salicin, amygdalin,and gentibiose. N Acetyl-f3-glucosamini-dase was also produced by most molds.This enzyme is involved in hydrolysis ofchitin. f3-Galactosidase which hydrolyzeslactose and a-glucosidase which hydro-lyzes sucrose, maltose, trehalose, andmelizitose were not produced by most ofthe molds.

In summary, molds as normal micro-flora in pollen and bee bread have recei-ved little attention. However, our resultsindicate that because they represented38% of the total number of microorga-nisms we isolated (Gilliam, unpublisheddata), produced a variety of enzymes, andare well known for production of seconda-ry metabolites such as antibiotics, pheno-lic compounds, terpenes, steroids, and

polysaccharides as well as enzymes, they

merit more intensive study. Even if mold

spores that germinate in corbicular pollenand bee bread produce short hyphae(Klungness and Peng, 1983), our resultswith selected isolates of various generaconfirmed those with Penicillia by Bridgeand Hawksworth (1984) that mycelia ino-cula produce th, same enzymes,although in reduced amounts, as sporesuspensions of the same strain. With thispublication and those on yeasts (Gilliam,1979a) and Bacillus ssp. (Gilliam, 1979b),we have now reported 77% of the totalisolates from our almond pollen and beebread samples. The remaining microorga-nisms will be described in a future publi-cation.

Acknowledgments

We thank Dr. L.N. Standifer of the Carl HaydenBee Research Center for the pollen samples,Dr. James T Sinski of the University of Arizonafor providing scanning electron microscopy andconsultation on the unidentified Ascomycete,and Drs. E.W. Herbert Jr., D.R. Jimenez, andJ.D. Vandenberg for reviewing the manuscript.

References

Alexopoulos C.J. (1962) Introductory Myco-logy. John Wiley and Sons, New YorkAmes L.M. (1961) A monograph of the Chaeto-miaceae. US Army Res. Div. Ser. No. 2Arizan D., Popa A., Mitroiu P., Toma C., SerbanM., Crisan 1. & Dobre V. (1967) Conservationdu pollen par des radiations ionisantes. Bull.Apicole 10, 43-50Batra L.R., Batra S.W.T. & Bohart G.E. (1973)The mycoflora of domesticated and wild bees(Apoidea). Mycopathol. Myco/. Appl. 49, 13-44Betts A.D. (1912) The fungi of the beehive. J.Econ. Biol. 7, 129-162Booth C. (1961) Studies of Pyrenomycetes. Vi.Thielavia, with notes on some allied genera.

Mycological Papers, No. 83, CommonwealthMycological Institute, Kew, EnglandBridge P.D. & Hawksworth D.L. (1984) The APIZYM enzyme testing system as an aid to therapid identification of Penicillium isolates.Microbiol. Sci. 1, 232-234

Burnside C.E. (1927) Saprophytic fungi asso-ciated with the honey bee. Mic;1. Acad. Sci. 8,59-86

Burri R. (1947) Die Beziehungen der Bakterienzum Lebenszyklus der Honigbiene. Schweiz.Bienen-Ztg. 70, 273-276Chevtchik V. (1950) Mikrobiologie pylov6hokvaseni. Publ. Fac. Sci. Univ. Masaryk 323,103-130

Cooke W.B. (1959) An ecological life history ofAureobasidium pullulans (deBary) Arnaud.

Mycopathologia 12, 1-45DeVries G.A. (1952) Contribution to the Know-lege of the Genus Cladosporium Link ex Fr.Centraalbureau voor Schimmelcultures, Baarn,The Netherlands

Dunham W.E. (1929) The relation of externaltemperature on the hive temperature during thesummer. J. Econ. Entomol. 22, 798-801

Egorova A.I. (1971) Preservative microflora instored pollen. Veterinariya 8, 40-41Ellis M.B. (1965) Dematiaceous Hyphomy-cetes. Vi. Mycological Papers, No. 103, Com-monwealth Mycological Institute, Kew, EnglandFoote H.L. (1957) Possible use of microorga-nisms in synthetic bee bread production. Am.Bee J. 97, 476-478

Gilliam M. (1979a) Microbiology of pollen andbee bread : the yeasts. Apidologie 10, 43-53Gilliam M. (1979b) Microbiology of pollen andbee bread : the genus Bacillus. Apidologie 10,269-274

Gilliam M. & Prest D.B. (1972) Fungi isolatedfrom the intestinal contents of foraging workerhoney bees, Apis mellifera. J. Invertebr.Pathol. 20, 101-103Gilliam M. & Prest D.B. (1977) The mycofloraof selected organs of queen honey bees, Apismellifera. J. Invertebr. Pathol. 29, 235-237Gilliam M. & Prest D.B. (1987) Microbiology offeces of the larval honey bee, Apis mellifera. J.lnvertebr. PathoG 49, 70-75

Gilliam M. & Vandenberg J.D. (1988) Fungi.In: Honey Bee Pests, Predators, and Diseases(R.A. Morse and R. Nowogrodzki eds.), ComellUniversity Press, Ithaca, New York (in press)

Gilliam M., Prest D.B. & Morton H.L. (1974)Fungi isolated from honey bees, Apis mellife-ra, fed 2, 4-D and antibiotics. J. Invertebr.Pathol. 24, 213-217 7Gilliam M., Taber S. III, Lorenz B.J. & PrestD.B. (1988) Factors affecting development ofchalkbrood disease in colonies of honey bees,Apis mellifera, fed pollen contaminated withAscosphaera apis. J. Invertebr. Pathol. 52,314-325

Gilliam M., Morton H.L., Prest D.B., Martin R.D.& Wickerham L.J. (1977) The mycoflora ofadult worker honey bees, Apis mellifera :effects of 2, 4, 5-T and caging of bee colonies.J. Invertebr. PathoL 30, 50-54

Haydak M.H. (1958) Pollen—pollen substi-tutes—bee bread. Am. Bee J. 98, 145-146Kendrick W.B. & Carmichael J.W. (1973)Hyphomycetes. In : The Fungi (G.C. Ainsworth,F.K. Sparrow and A.E. Sussman, eds.), Acade-mic Press, New York, Vol. IVA, pp. 323-509

Klungness L.M. & Peng Y. (1983) A scanningelectron microscopic study of pollen loads col-lected and stored by honeybees. J. Apic. Res.22, 264-271Loper G.M., Standifer L.N., Thompson M.J. &Gilliam M. (1980) Biochemistry and microbiolo-gy of bee-collected almond (Prunus dulcis) pol-len and bee bread. Apidologie 11, 63-73McGinnis M.R., Borelli D., Padhye A.A. & AjelloL. (1986) Reclassification of Cladosporiumbantianum in the genus Xylohypha. J. Clin.Microbiol. 23, 1148-1151Miller M.W., Phaff H.J., Miranda M., Heed W.B.& Starmer W.T. (1976) Torulopsis sonorensis, anew species of the genus Torulopsis. Int J.

Syst. Bacteriol. 26, 88-91Morton F.W. & Smith J. (1963) The generaScopulariopsis Banier, Microascus Zukal and

Doratomyces Corda. Mycological Papers, No.86, Commonwealth Mycological Institute, Kew,England

Neergaard P. (1945) Danish Species of Alter-naria and Stemphylium. Taxonomy, Parasitismand Economic Significance. Einar Munksgaard,CopenhagenPain J. & Maugenet J. (1966) Recherches bio-chimiques et physiologiques sur le pollenemmagasind par les abeilles. Ann. Abeille 9,209-236

Raper K.B. & Fennell D.I. (1965) The GenusAspergillus. Williams and Wilkins, Baltimore

Raper K.B. & Thom C. (1968) A Manual of thePenicillia. Hafner, New York

Sainger J.K., Garg A.P. & Sharma P.D. (1978)Mycoflora of some pollen grains. Acta Botan.Indica 6, 165-168Samson R.A. (1974) Paecilomyces and someallied Hyphomycetes. Studies in Mycology, No.6, Centraalbureau voor Schimmelcultures,Baarn, The Netherlands

Skou J.P. (1972) Ascosphaerales. Fiesia 10,1-24

Standifer L.N., McCaughey W.F., Dixon S.E.,Gilliam M. & Loper G.M. (1980) Biochemistryand microbiology of pollen collected by honeybees (Apis mellifera L.) from almond, Prunisdulcis. 11. Protein, amino acids and enzymes.Apidologie 11, 163-171 1von Arx J.A. (1975) On Thielavia and somesimilar genera of Ascomycetes. Studies in

Mycology, No. 8, Centraalbureau voor Schim-melcultures, Baarn, The Netherlands

Zycha H., Siepmann R. & Linnemann G. (1969)Mucorales. J. Cramer, Lehre, Germany