Feral Honey Bees in the Sonoran : Propolis Sources other than Poplars (Populus spp.) Eckhard Wollenweber3 and Stephen L. Buchmannb a Institut für Botanik der Technischen Hochschule, Schnittspahnstraße 3, D-64287 Darmstadt, Germany b Carl Hayden Bee Research Center, USDA - ARS, 2000 East Allan Road, Tucson, A Z 65719, U. S. A. Z. Naturforsch. 52c, 530-535 (1997); received April 25, 1997 Apis mellifera, Feral Bees, , Propolis, Source, Ambrosia, Encelia, Resinous Exudates In Central Europe and other temperate regions the lipophilic bee hive product propolis normally originates from the bud exudate of poplar trees that is collected by the bees. Based on bee observations, various other sources have been discussed in particular for tropical regions, but in only few cases the origin has been proved by analytical methods. We have analyzed propolis samples from managed honey-bees as well as from feral bee colonies in the Sonoran Desert. Propolis collected in hives out of flight reach of poplars contained flavonoid aglycones and other phenolics that point to specific as the source of propolis in this area, namely Ambrosia deltoidea and Encelia farinosa.

Introduction based on bee observations only and not on the Propolis or bee-glue is a resinous or sometimes chemical analysis of propolis. wax-like bee hive product that has been used by The best indicator for the origin of European man since ancient times for its pharmaceutical propolis is the pattern of flavonoid aglycones pre­ properties. It is still used as a remedy in folk medi­ sent in this lipophilic material (Hausen et al., 1987; cine (Ghisalberti, 1979), as a constituent of “bio­ Wollenweber et al., 1990). Normally the flavonoid cosmetics” and for numerous further purposes pattern (“fingerprint”) of propolis can hardly or (Vanhaelen and Vanhaelen-Fastre, 1979; Hausen not at all be discerned from that of poplar bud et al., 1987; Wollenweber et al., 1990). Bees use exudate (cf. Wollenweber, 1975), the latter being this material to seal hive walls and its entrance, to its source. During a plant collection trip in Ari­ strengthen the border of the combs, to embalm zona, the senior author wondered where bees liv­ dead invaders. It is now well known and it has ing in the Sonoran Desert might get their propolis been confirmed by many studies that at least in material. Which plant source would they use in Central Europe the bees almost exclusively collect this area? Are they desperately searching for the this material from the bud exudate of poplar trees, scarce cottonwood trees, or have they adopted to in particular of black poplar, Populus nigra. Ban­ collect lipophilic material from other plant kova et al. (1992) have shown that also in Mongo­ sources? Analysis of propolis flavonoids should lia poplar buds (Populus suaveolens) are the allow us to answer this question. source of propolis material. Deviant vegetable The first propolis samples from the Sonoran propolis sources have been reported from Russia, Desert area that we studied came from bees hives South Africa, Australia, Hawaii etc. Bees are even in the vicinity of Tucson/, where the scat­ said to collect materials like asphalt, mineral oil or tered vegetation is dominated by xeromorphic paint as a substitute in case of need (König 1985 and cacti, whereas cottonwood (Populus) is a, 1985 b). Many of these statements, however, are very scarce. A study of the flavonoid aglycones of these samples revealed that even in that region the bees must have access to poplar trees, as the propolis flavonoid patterns were typical for Popu­ Reprint requests to E. Wollenweber. lus bud exudates (Wollenweber and Egger, 1971; Telefax: (49) 6151/166878 Wollenweber, 1975; Wollenweber et al., 1987a).

0939-5075/97/0700-0530 $ 06.00 © 1997 Verlag der Zeitschrift für Naturforschung. All rights reserved. D

Dieses Werk wurde im Jahr 2013 vom Verlag Zeitschrift für Naturforschung This work has been digitalized and published in 2013 by Verlag Zeitschrift in Zusammenarbeit mit der Max-Planck-Gesellschaft zur Förderung der für Naturforschung in cooperation with the Max Planck Society for the Wissenschaften e.V. digitalisiert und unter folgender Lizenz veröffentlicht: Advancement of Science under a Creative Commons Attribution Creative Commons Namensnennung 4.0 Lizenz. 4.0 International License. E. Wollenweber and S. L. Buchmann • Feral Honey Bees in the Sonoran Desert 531

Fremont cottonwood, Populus fremontii S. Wats, same race/subspecies). The samples were collected occurs, indeed, naturally in the Arizona before the Arizona Africanized honey bee in­ (Benson and Darrow, 1981). The “fingerprint” of vasion, so do not represent propolis from that these propolis samples matches with bud exudate race. of P. fremontii, although this poplar is extremely A) Propolis from managed colony within easy rare in the Tucson area where it grows only spo­ flight range (about 0.5 mile) of Rillito River and radically in watercourse beds (Wollenweber et al., some poplar trees. The colony has been in the 1990). Only one flavonoid was detected (xantho- same spot for about 6 years and was full of propo­ microl, 5,4'-dihydroxy-6,7,8-trimethoxy flavone) lis. Coll. H. G. Spangler (Carl Hayden Bee Re­ that cannot be derived from Populus, but might search Center, Tucson, Arizona), June 10, 1988. originate from the leaf exudate of an asteraceous B) Propolis from the Carl Hayden Bee Re­ desert such as an Ambrosia species (Wollen- search Center. Mixed managed colonies within weber et al., 1987b; 1995) or a Baccharis species easy flight range of the Rillito River poplar trees. (Wollenweber et al., 1986; 1989). Finally, propolis Coll. S. Buchmann, December 16, 1990. samples were obtained from colonies of feral bees C) Propolis and melted wax from small entrance found in rock outcrops beyond the flight range of around mouth of Colony No. DS-2. Colony dead any poplar trees. The insects were observed col­ when sampled. Dripping Springs, Organ Pipe Cac­ lecting material from the leaf surfaces of Astera­ tus National Monument, Ajo, Arizona. Coll. S. ceae belonging to the genera Ambrosia, Baccharis, Buchmann, February 2, 1989. Encelia and Gutierrezia. One would also expect D) Dark propolis floor sample. Feral colony that the bees use the abundant leaf resin of the in rock wall of wash at the Arthropod Discovery most common shurb in that area, the creosotebush Center, Sonoran Arthropod Studies, Institute, Larrea divaricata (Moc. and Sesse) Cov. The resin Tucson Mountains, Tucson Mountain Park. Live of the elephant tree, Bursera microphylla A. Gray, colony above. Coll. S. Buchmann, August 23, is another possible source that one might think of. 1990. Analysis of these feral bees’ propolis samples was, E) Propolis from feral colony OWL-1. Colony therefore, of high interest. In the following we re­ on ledge behind at Owl Head Buttles, port the results of a study on this subject, done Pima Co., Arizona.. Coll. John Edwards, March with propolis from the hives of cultivated bees as 12, 1990. well as from the nests of feral bees out of flight F) Sample from feral colony TORT-1, Honey range from any poplars. Bee Canyon, Rancho Vistoso, Pima Co., Arizona. Life colony. Coll. John Edwards, March 12, 1990. G) Scrapings from very large debris midden. Or­ Material and Methods gan Pipe Cactus National Monument, Ajo, Ari­ Aerial parts of Ambrosia deltoidea were col­ zona. Copper Mountain Col. no. 1. Colony long lected west of Tucson city limits. Aerial parts of dead. Coll. S. Buchmann, 1989. Encelia farinosa were collected in the foothills of H) Propolis sample from “Artificial Swarm Mount Lemon near Tucson/AZ (G. Yatskievych, Scale Colony Experiment.” Arthropod Discovery May 1990). Air-dried material of both plants was Center, Tucson Mountains, Pima Co. Sonoran Ar­ briefly rinsed with acetone to dissolve the lipo­ thropod Studies, Institute, Tucson, AZ. Large philic exudates. Identification of flavonoid agly­ amounts of Ambrosia deltoidea nearby, probably cones in Ambrosia deltoidea has been reported no Populus trees within flight range. (Propolis is previously (Wollenweber et a l, 1987 b). For the only from March - November of 1990). Coll. S. present study samples of the acetone-wash were Buchmann, November 8, 1990. used for direct comparisons with propolis samples. Propolis samples J - N came from feral honey The following collection data refer to the prop­ bee colonies living within a Schmidt/Thoenes olis samples now studied. They were collected “swarm trap” located within Organ Pipe Cactus from the hives of cultivated bees, i.e. managed National Monument, Ajo, Arizona. They were all European honey bees (Apis mellifera ligustica) collected by Steven C. Thoenes. Specific locations and from the nests of feral bees (wild bees of the are as follows: 532 E. Wollenweber and S. L. Buchmann • Feral Honey Bees in the Sonoran Desert

J) Colony 1.5 miles from Senita Basin. February tribution revealed that the flavonoid pattern cor­ 14, 1991. responds to that found in the exudate of Ambrosia K) Mesquite bosque approx. 1 mile from Quito- deltoidea. The characteristic exudate flavonoids of baquito Springs. February 14, 1991. this plant are the major products, 3'-desmethoxy- L) Quitobaquito Springs area. February 14, sudachitin (5,7,4'-trihydroxy-6,8-dimethoxy fla- 1991. vone) and xanthomicrol (5,4'-dihydroxy-6,7,8-tri- M) Burro Springs near Quitobaquito Springs methoxy flavone), and the minor components, area. February 14, 1991. N) Dripping Springs. No scutellarein-6-methyl ether (hispidulin), luteolin- cottonwood trees in area. January 15, 1992. 7-methyl ether, 6-methoxy luteolin (nepetin), sid- The phenolic portions of ground propolis sam­ eritiflavone (5,3',4'-trihydroxy-6,7,8-trimethoxy ples were extracted with acetone and with warm flavone), kaempferol, quercetin, rhamnetin (quer- ethanol, respectively. Flavonoid aglycones were cetin-7-me), rhamnazin (quercetin-7,3'-diMe) and identified by direct TLC comparison with markers, ombuin (quercetin-7,4'-diMe). The presence of based on the senior author’s experience with the two coumarins, scopoletin and ayapin, is also typi­ analysis of flavonoid aglycones in general, and cal for the “fingerprint” ofAmbrosia deltoidea on with poplar bud exudates and with propolis in par­ TLC, where they appear as light blue fluorescent ticular (Wollenweber, 1975; Wollenweber et al., spots (Wollenweber et al., 1987b). Thus, their pres­ 1987a; Wollenweber et al., 1990). Propolis samples ence in these propolis samples confirms this plant and plant exudates were also compared directly source. Triangular leaf bur sage, Ambrosia del­ by TLC. Thin layer chromatograms were run on toidea (Torrey) Payne, is, in fact, omnipresent in polyamide DC-11 with solvents A) petrol 100- most Sonoran Desert sites. The presence of A m ­ 140° - toluene - methylethylketone - methanol brosia exudate constituents in propolis was further 12:6:1:1 (v/v), B) toluene - petrol 100-140° - confirmed by GC/MS analysis of several of our methylethylketone - methanol 12:6:2:1 (v/v), and samples. Apart from the flavonoid aglycones it re­ C) toluene - methylethylketone - methanol vealed identical major peaks in both, propolis and 12:5:3 (v/v), sometimes also on silica with tolu­ Ambrosia exudate, which do not correspond to ene - MeCOEt 9:1 (v/v) and with toluene - di- flavonoids. The relevant products were not iden­ oxane - HOAc 18:5:1 (v/v). Chromatograms were tified, but nonetheless this result has additional viewed under UV366 before and after spraying value as evidence. with “Naturstoffreagenz A” (Nat. A). Authentic Sample K originates from a feral European Apis flavonoid aglycones and coumarins for compari­ mellifera colony in a swarm trap. This sample ex­ son purposes were available in E. W.’s lab. hibits flavonoids derived from Populus bud exu­ date as well as flavonoids derived from Ambrosia deltoidea leaf exudate. Obviously the bees have Results and Discussion collected both materials. Propolis samples A and B, collected from man­ Samples L and M, also from feral bee colonies, aged Apis mellifera colonies, exhibit the character­ exhibit neither the flavonoid pattern of Populus, istic flavonoid patterns of poplar bud exudates, al­ nor the flavonoid pattern of Ambrosia deltoidea, though it is not completely identical with that of although both sources are present within the bees Populus fremontii. The hives are within flight flight range. Curiously enough, we here found range of Fremont cottonwood. some highly lipophilic products that are charac­ The collection sites of samples C - J are all out teristic of another asteraceous plant, namely the of flight range of any cottonwood tree. These sam­ brittle bush Encelia farinosa A. Gray. This species ples show different wax contents, different colors, occurs on rocky or gravelly slopes and mesas in and somewhat different solublility. In sample G the Arizona deserts and adjacent areas (Benson (“colony long dead”) the flavonoid content is very and Darrow, 1981). Chromenes (benzopyrans) and low. In each case the flavonoid pattern does not benzofurans are common inEncelia species like in match that of poplar bud exudate. After isolation many other (Proksch and Rodriguez, of two major flavonoids and identification of some 1982). These compounds are non-polar products minor flavonoids, a data base search for their dis­ that fluoresce on irradiation with long wave UV- E. Wollenweber and S. L. Buchmann • Feral Honey Bees in the Sonoran Desert 533

light (Proksch et al., 1984). They might, therefore, As a matter of fact, Clusia minor and Clusia major well correspond to the non-polar fluorescent spots (Guttiferae) were found to produce the same pol­ observed on TLC in Encelia farinosa leaf wash as yprenylated benzophenones in their flowers and well as in propolis samples L and M. According to bees have been observed collecting this resin. For Proksch et al. (1984) the accumulation of chro- comparison purposes the same authors analyzed menes and benzofurans is strictly correlated to the propolis from temperate regions: Spain, Bulgaria, presence of resin ducts. With the help of fluores­ Canada, Arizona, New Zealand. In each case the cence microscopy they have proved that in Encelia typical flavonoids found in the bud exudate of Po­ farinosa these compounds are stored exclusively in pulus sect. Aigairos were abundant. On the other the resin ducts and the surrounding cells. It ap­ hand, only a few of the tropical samples showed pears unlikely, however, that bees pierce the stems trace amounts of flavonoids which in one of them of brittle bush to suck the resin, so we checked the could be identified to be 6-methoxylated flavones. localization once more. When undamaged aerial This type of flavones is often found in exudate on parts are rinsed with acetone, the solution actually the leaf surfaces of Asteraceae (Wollenweber and exhibits exactly those non-polar spots on TLC that Valant-Vetschera, 1995) and of Labiatae (Tomäs- we observed in propolis samples L and M. It is Barberän and Wollenweber, 1990). Hence we as­ assumed, therefore, that at least a small amount of sume that such plants might be the source in a few these products is also excreted by and deposited tropical propolis samples. on leaves and stems and this material is used by Bonvehf and Coll (1994) analyzed 15 samples of bees as a source of propolis. With regard to the propolis originating from China, from Brazil and function of propolis, the insecticidal and anti­ from Uruguay. They identified 24 phenolic com­ microbial effects of these phenolics (Proksch et al., pounds among which benzoic acid and benz- 1983; Isman and Proksch, 1985) might be a wel­ aldehyde derivatives as well as flavonoids were come property. abundant. Several of the phenolic compounds Sample N exhibits not only the flavonoid identified by these authors might be collected pattern of Ambrosia deltoidea. In addition, it from poplar bud exudates (chrysin and tecto- shows the typical Encelia farinosa products. Obvi­ chrysin, galangin, pinocembrin and pinostrobin, ously these bees have collected their propolis vanillin). The preponderance of apigenin and api- material from both Asteraceae species. genin-4'-methyl ether, however, indicates marked In view of the observations mentioned in the differences between propolis from Central Europe Introduction we assume that other propolis sam­ and propolis from China and South Ameria and ples might contain flavonoid aglycones that are points to different sources. The origin of glycosides characteristic for the leaf exudates of Baccharis such as rutin, hesperetin and naringin has not been and of Gutierrezia species among other plant clarified; the same is true e. g. for gallic acid. sources. By contrast, it is unlikely that Larrea is A recent study on antimicrobial compounds in used as a source for propolis because considering Brazilian propolis (Agaet al., 1994) reports three its ubiquity in the area, its phenolic constituents typical compounds: 3,5-diprenyl-4-hydroxycin- should already have been found in the samples an­ namic acid, 3-prenyl-4-dihydrocinnamoloxycin- alyzed so far. namic acid and 2,2,-dimethyl-6-carboxyethenyl- Tomäs-Barberänet al. (1993) studied a total of 2H-l-benzopyran. Again the origin of these com­ 38 propolis samples from tropical Venezuela by pounds, i.e. the plant source is not known. It is HPLC analysis. 29 samples were produced by im­ assumed, though, that they might come from ported Apis mellifera, the others by five indige­ Baccharis and Flourensia species that have been nous species of stingless bees. Most of these tropi­ shown previously to produce the same prenylated cal propolis samples showed a rather uniform cinnamic acid derivatives (cf. Bankova et al., 1995). phenolic profile that is characterized by the pres­ Bankova et al. (1995) also report on the antibac­ ence of polyprenylated benzophenones, while terial activity of Brazilian propolis. Major constitu­ some samples were devoid of any phenolic com­ ents that they identified comprise diprenyl aceto- pound. The uniform phenolic pattern pointed to a phenone, 2Z,6£-farnesol, a sesquiterpene alcohol, single plant species or plant genus as the source. dihydrocinnamic acid and p-coumaric acid. Only 534 E. Wollenweber and S. L. Buchmann • Feral Honey Bees in the Sonoran Desert traces of an unidentified flavone and of an uniden­ ence of several compounds in both materials. The tified flavanone were detected. It is obvious that present results with cottonwood and triangular the propolis samples studied here were not col­ leaf bur sage are even more convincing as they are lected from poplar buds. Bankova et al. (1995) in based on completely identical sets of flavonoids their Brazilian propolis samples did not find the (“fingerprint”) in both,Populus fremontii and polyprenylated benzophenones that Tomäs-Bar- Ambrosia deltoidea exudates and propolis. berän et al. (1993) had found in material from Venezuela nor did they encounter prenylated cin­ Acknowledgements namic acid derivatives as they have been reported by Aga et al. (1994) for propolis from Brazil. Two The authors are grateful to Mr Steven J. Prchal of the samples studied, however, contained con­ of the Sonoran Arthropod Studies Institute in Tuc- siderable amounts of prenyl acetophenone and di­ son/AZ for permission to place honey bee colo­ prenyl acetophenone, respectivly. This might point nies on their property and to collect propolis sam­ to Flourensia heterolepis, a plant producing preny­ ples. Thanks are due to Dr. George Yatskievych lated hydroxyacetophenones (Bankova et al., (Missouri Botanical Garden, St. Louis, MO) who 1995). collected Ambrosia deltoidea and Encelia farinosa In summary it can be stated t