DANIEL McALPINE MEMORIAL LECTURE 1999

Bees and fungi, with special reference to certain plant pathogens

  D.E. Shaw

C/– Department of Primary Industries, Meiers Road, Indooroopilly, Queensland 4068, Australia

Daniel McAlpine

 This lecture commemorates the life and work of Daniel McAlpine and his contribution to plant ­pathology. McAlpine, a Scotsman, matriculated at London University, studied botany and geology at the Royal School of Mines, South Kensington, as well as biology under Professor Thomas Huxley, and was later appointed Professor of Natural ­History in Edinburgh. He migrated to Australia in 1884 and after teaching at Ormond College was ­appointed Vegetable Pathologist with the Department of Agriculture in Victoria in 1890. He investigated the plant disease situation in Australia and extended this knowledge to scientists and primary producers in various publications on plant disease in many crops and other plants. His work during 26 years has rightly earned him the title of ‘Father of Plant Pathology’ in this country (Carne 1927–28; Fish 1970; 1976; White 1981).

Introduction

It is estimated that the world population will be 10 billion in 2050 AD (Rothschild 1998). The world’s population depends on (in the first or other degrees) the vegetation of the planet for food, fodder, fibres, fuel, timber, paper, pharmaceuticals, oxygen, water etc. Two of the many components contributing to the wellbeing of the earth’s vegetation are (1) the work of the plant pathologists, and (2) successful pollination.

The biotic pollinators include bees, wasps, flies, butterflies, moths, beetles and other insects, birds, marsupials, monkeys, lizards etc. with bees, ­especially honey bees, being predominant in many parts of the world (Proctor et al. 1996).

 Early records of honey hunters, beekeeping, and biotic pollination

The earliest record of man’s involvement with bees is a pictorial one of honey hunters on the wall of a cave in Spain ca. 6000 BC, and the first record of beekeeping is also a pictorial one from Egypt ca. 2400 BC (Crane 1983).

Although a carving from the palace of an Assyrian king ca. 860 BC shows hand pollination of date palm (Buchmann and Nabhan 1996), and although Aristotle noted floral constancy by bees in the 4th century BC (Proctor et al. 1996), there was no knowledge of the role of insects in pollination until a few hundred years ago. About that time in the UK, Miller (1691–1771) worked with tulips and bees, and Dobbs (about 1750) confirmed his work and noted floral constancy. In Germany, Kölreuter (1733–1806) carried out experiments on insect pollination, examined many nectars, and concluded that nectar was the source of honey (Proctor et al. 1996).

Main field activities of honey bees

The four main field activities of honey bees and their relatives are the collection of nectar, pollen, propolis and water. Instead of nectar, bees in some areas collect insect honeydew. Nectar (and insect honeydew) is high in sugars; pollen in protein, ­vitamins, lipids and minerals; propolis (used for in-hive repair) in resin, wax and essential oils. Water is needed for the dilution of larval food and for cooling and humidifying the interior hive environment. Nectar and water are transported to the hive internally in the honey stomach, whereas pollen and propolis are carried externally in pollen baskets (corbiculae) on the outside of the back legs.

Bees and fungi

Fungi are associated with bees in various ways, ­including the following:

1. As atmospheric mycospora which may be ­deposited on flowers and on bees.

2. In nectar and honey. In nectar yeasts predominate; in honey most spores do not grow (due to the high osmotic pressure etc.) but, if there, they may persist when honey is diluted.

3. On and in pollen. For example, pollen of ­lucerne may carry the fungus Verticillium alboatrum when incidentally transported by the alfalfa leafcutting bee Megachile rotundta (see Table 1).

4. In the hive and on provisions. Many species of fungi have been recorded.

5. On and in bees, and as pathogens of bees. The fungus Ascosphaera apis (the cause of chalkbrood) and other species of Ascosphaera have been examined molecularly by Anderson et al. (1998). Four new species in this genus have also been described, and three other isolates including A. apis, not previously known in Australia (Anderson and Gibson 1998).

6. In substitution and supplementary food. Brewers yeast is one of the important constituents of these foods.

7. In incidental transport during normal foraging.

8. In the collection of fungal spores in lieu of ­pollen.

Only the two last items will be discussed in the following sections.

Incidental transport

Plant pathogenic propagules may be incidentally transported by bees of various genera and species during their normal foraging as well as by many other insects. The foraging may involve the collection of nectar, or insect honeydew (secreted especially by aphids and scale insects), or fungal honeydew as in species of Claviceps, and the sweet pycnidial secretion as in some rust fungi. Only those records involving bees and plant pathogens are listed in Table 1, together with references. Only two of the items are mentioned below.

Various studies in the United States of America (USA) and Canada have tested fungi, mainly Gliocladium roseum, as possible control agents of fungal pathogens by bees, including Sclerotinia sclerotiorum causing stem rot of canola (Brassica napus) by honey bees (Apis mellifera) (Table 1). Tests carried out in Canada against Botrytis cinerea causing grey mould of strawberry and raspberry flowers used both honey bees (A. mellifera) and bumble bees (Bombus impatiens). Bee-vectored inoculum effectively controlled B. cinerea in petals, stamens and flowers of strawberry and in flowers of raspberry. However, raspberry fruits were not ­adequately protected, presumably because B. cinerea is able to infect drupelets directly as well as by ­invasion of flowers (Table 1).

In the USA, when the fungus Puccinia monoica infects its aecial host (species of Brassicaceae, ­especially Arabis) it causes leaves to be yellow and clumped, visually similar to the flowers of the buttercup (Ranunculus sp.), and hence called ‘pseudoflowers’. The pseudoflowers exude a sweet solution which attracts hallictic and andrenid bees as well as other insects. The insects help in transferring pycniospores from the pycnia of infected plants to other pycnia. The pseudoflowers also produce a distinctive floral fragrance composed primarily of aromatic alcohols, aldehydes and esters, whereas infected plants produce only green leaf volatiles, mainly terpinoides, ubiquitous from vegetative ­tissue. The yellow pseudopetals also reflect UV light, so that the pseudoflowers give gustatory, olfactory and visual cues to bees (Table 1).

Canada thistle (Cirsium arvense) plants infected with the systemic sexual state of Puccinia punctiformis also produce fragrant volatiles but ­although this is stated (Connick and French 1991) to facilitate cross-fertilisation by attracting insects, these have not apparently included bees.

The collection of spores in lieu of pollen

The collection of spores by honey bees was first recorded in 1875, and since then the records have all involved bees in the family Apidae, mainly the European honey bee (Apis mellifera). The fungi collected in lieu of pollen have included various plant pathogens causing rust and powdery mildew diseases, and species of Neurospora (non-pathogenic to man, other animals and plants), as well as a few other fungi. The Neurospora has been mainly recorded on filter mud, the precipitated impurities obtained during cane sugar extraction. The records, although infinitesimal in number compared with pollen collection , have now occurred on all continents (except Antarctica) and in the East and West Indies. The records are not expanded here but are listed with references in Table 2, with common names as in the original papers. Some aspects of the collections are briefly discussed in the following sections.

Duration of collecting  

Data on this aspect are not available for some records, but have been recorded for others including the following. Bees collected spores of powdery mildew (Oidium farinosum) from apple trees during 14 days in Germany (Kraus 1920). In Israel, large numbers of bees gathered the spores of poplar rust (Melampsora larici-populina) in October–November 1940. In 1941 the rust ­appeared in the latter half of August and spores were collected until the beginning September, when the bees disappeared. At the beginning of October the bees reappeared at the start of a second rust attack (Minz 1942).

In January and February 1954, 1955, 1957 and 1958, microscopic examination of corbicular loads in the Mahebleshwar area of India showed that they consisted uniformly of spores of Zaghouania oleae which infects Olea dioica. None of the loads ­collected in January–February 1956 showed the spores, perhaps due to the flowering of Nilgirianthus heyeanus v. neesii, which blooms only once in three years (Deodikar et al. 1958).

In the Netherlands, bees collected spores of Melampsora larici-populina in August–October 1986 (Moraal 1988), while in Mexico spores of Cronartium conigenum were collected from rusted pine cones in June–July (Trujillo Flores and Peña Garcia 1989). Corbicular loads sampled on 24, 28 and 29 September 1992 in the United Kingdom contained spores of poplar rust (M. larici-populina) (Kirk 1994; 1996, personal communication).

Conidia of Neurospora were present and being collected by bees during a visit by Drs G.R. and A.M. Stirling to a heap of filter mud ca. 110 km north of Brisbane (Table 2). During the first visit they estimated that there were 20 bees per m². The bees were still foraging on the same heap during a second visit 10 days later (although there was ­reduced coverage of fungus due to rain) when they estimated 40 bees per m². Colonies of Neurospora occurring at sites in a coastal strip over 1000 km long in eastern Australia (Shaw 1990a) have been checked only intermittently during the last two decades, but bees foraging on this resource (Neurospora) have been recorded 25 times during a period of 22 years (Shaw and Robertson 1980; Shaw 1990b; 1993; 1998 and herein).

The continuous spore collection at some of the sites listed in Table 2 indicates the apparent acceptance by the housekeeping bees of the propagules and the continued recruitment of foragers to these resources.

Reasons for the ‘in lieu’ collecting  

Some ­authors attributed the collection of fungal spores to a dearth of pollen at the time. In New Zealand, Bennie (1942) stated that the reason bees collected spores of Puccinia graminis (avenae) from a chaff cutter which had cut rusted oat straw the previous day was because the severe winter had killed all the gorse and broom blossom with a resultant shortage of pollen in early spring. In India, aeciospores of Zaghouania oleae were collected in those years when a local tree (Nilgirianthus heyneanus var. neesii) was not flowering, as it blooms only once in three years (Deodikar et al. 1958). During 1963 in Jamaica, honey bees were seen to visit allspice (Pimenta dioica) at a time when trees bore no flowers but which were infected with Puccinia psidii. Loads in the corbiculae consisted of pure samples of urediniospores (Chapman 1964). In Florida, bees were recorded collecting spores of Puccinia oxalidis on Oxalis spp. at a time when frost had killed perhaps 95% of the local blossoms (Wolfenbarger 1997). In South Africa, Melampsora spores growing on the weed Euphorbia geniculata were collected during periods of general pollen scarcity (Johannsmeier 1981). Collections of spores of Cronartium conigenum from Pinus spp. in Mexico were particularly abundant from mid-June to mid-July, a period of few flowers, but diminished later with the occurrence of a greater variety and quantity of pollen (Trujillo Flores and Peña Garcia 1989). In New South Wales, honey bees collected conidia of Oidium sp. on rose leaves in a garden during a two years’ drought (Ray 1981 personal communication).

On the other hand, various authors recorded the collection of fungal spores in lieu of pollen at times when flowers were present in the vicinity. For ­example, bees collected spores of Caeoma nitens rust of wild dewberry in Texas because it furnished an easy and abundant supply, although some plants of various species were flowering in the vicinity (Lang 1901). In Germany, bees were noted at the time of apple blossoming collecting either the ­conidia of powdery mildew (caused by Oidium farinosum) or apple pollen (Kraus 1920). In Sarawak, the Giant Asian honey bee (Apis dorsata) twice collected urediniospores of tropical maize rust (caused by Puccinia polysora) although ample pollen was available on the same plants, and which pollen, in fact, was being collected by the stingless bees (Trigona spp.) (Turner 1974). Spores of Melampsora larici-populina on poplars were collected by honey bees in Western Australia, although urediniospores of plum rust (caused by Tranzschelia pruni-spinosa) were also present but were ignored (Dell 1977). Bees collected spores of willow rust (caused by Melampsora sp.) in the USA and although some bees were visiting sage (Salvia sp.) in the same area, they were very few compared with the activity in the willows, i.e., the bees clearly preferred foraging the willow rust to nearby sage which provides both pollen and nectar (Williams and Tomlinson 1985). Bees in South Africa collected spores of Melampsora ricini on the castor oil plant, ignoring flowers in the vicinity (Wingfield et al. 1989).

While there are now records of bees collecting Neurospora conidia, these have been from large blooms of the fungus which have provided ample spores. For example, in Bolivia, the bees gathered a load of spores possibly in a third or less of the time commonly utilised when collecting pollen from flowers, as in an area of a few cm² they encountered sufficient quantities to complete their load (Kempff Mercado 1955). This was also the experience in eastern Australia, where bees collected from large areas of Neurospora, but not from small colonies, or those with limited sporulation under drought conditions, where the relatively few spores available were apparently not a sufficient reward for the energy that would be expended in foraging (Shaw 1998).

Therefore, fungal spores have been collected at times of pollen dearth, but also at other times when presumably the reward (in spores) has been sufficient to repay the energy expended in the collection.

Fungal spores are, of course, available throughout the entire period of a bee’s daytime foraging, whereas pollen of some flowers is only available at certain times of the day. One further point should be mentioned, viz., that if bees collect fungal spores in lieu of pollen they will not be fulfilling their role as pollinators. However, as some records occurred in times of pollen dearth there may in fact have been reduced pollination, even if the bees had not been collecting fungal spores at the time.

 Size of pollen and fungal spores collected by bees

Although pollen grains measure from 5 to over 200 µm in diameter (Faegris and Iversen 1989), and even >300 µm (Roubik 1989), the size of grains usually collected by bees is in the smaller range of up to 100 µm with mean 34 µm (Roberts and Vallespir 1978) with grain contents lipid-rich rather than starchy (Baker and Baker 1983). The size of the fungal spores collected by bees (rust ­urediniospores ca. 20–40 × 10–25 µm; powdery mildew conidia ca. 20–30 × 10–20 µm, and ­Neurospora ­conidia ca. 5–15 µm) falls within the range of grains usually collected by honey bees.

Type of bee  

All the records in Table 2 refer to the European honey bee (Apis mellifera) except for two records of the Giant Asian honey bee (A. dorsata) (both Turner 1974) and two records of Trigona spp. (Trigona cf. branneri) (Burr et al. 1996) and T. amalthea (Roubik 1989). The preponderance of records involving A. mellifera may merely reflect the attention given to this species in beekeeping milieus around the world.

Spore forms collected  

Most of the rust fungi collected were urediniospores, with only a few records of aeciospores or aeciospore-like propagules (Zabriskie 1875; Deodikar et al. 1958); ustilospores of Ustilago (Betts 1912) and of Ustilaginales (Maurizio 1975); and conidia of Oidium spp. and Neurospora. All these spores are dry deciduous spores with ease of detachment and are normally wind-blown (anemophilous). However, there are now two records of species of stingless bees (Trigona) foraging on the fruiting bodies of a stinkhorn (Staheliomyces cinctus, Phallaceae) and the gleba of an unidentified fungus in Ecuador and Panama, respectively, when the bees removed parts of the thalli before packing the pieces into their corbiculae.

Colour  

Honey bee are usually highly sensitive to long-wave ultra violet (LW UV) and insensitive to red, with a visible spectrum of 300–ca. 650 nm, against the human range of 400–800 nm (von Frisch 1950; Seeley 1995). Massed Neurospora conidia (which microscopically appear individually colourless) reflect LW UV (Shaw and Robertson 1980), as do powdery mildew conidia (Shaw 1990b). Rust fungi, however, whose urediniospores and aeciospores are individually pigmented either in the cell wall and/or in the cytoplasm, probably have non-reflectance (absorbance) as uredinia on leaves and massed detached urediniospores of Melampsora larici-populina, Puccinia maydis, P. oxalidis and P. paullula did not reflect LW UV (Shaw 1990b). The nearer the colour of the massed rust spores to yellow, however, the more likely they are to fall within the range of the bees’ visible spectrum.

Fungal constancy  

Microscopical checks of corbicular samples of many pathogens and Neurospora spp. listed in Table 2 showed nearly 100% homogeneity of contents, i.e., there was collecting constancy for fungal spores, just as there is floral constancy for pollen collection from profitable sources. Exceptions were those reported by Kraus (1920) where in a few cases bees collected apple pollen and then changed to mildew spores.

Destination of fungal loads  

Corbicular loads of pollen are stored in hive cells around the brood. Fungal spores collected by bees have also been ­recorded in a few cases in hive cells. These records include those of Zaghouania oleae whose spores were detected in broodcomb in India (Deodikar et al. 1958). In another case in Jamaica (Chapman 1964) the comb contained a high proportion of urediniospores of Puccinia psidii collected from allspice. In Western Australia, spores of Melampsora larici-populina were located in pollen stores in a hive (Dell 1977) and spores of M. euphorbiae were found stored in considerable quantity in comb in ­Pakistan (Ahmad and Ahmad 1986). Spores of Cronartium conigenum collected from Pinus were found in the hive cells in Mexico (Trujillo Flores and Peña Garcia 1989). Of course, fungal loads ­recovered in pollen traps at hive entrances are not available in samples from hive storage cells.

Nutritive requirements of honey bees  

This ­aspect was summarised in Shaw (1990b) as follows. Pollens with less than 20% crude protein cannot satisfy colony requirements for optimum production (Kleinschmidt and Kondos 1976), whereas bees consuming more than 23% protein were able to ­successfully rear brood (Herbert and Shimanuki 1978). Fungal yeasts, of course, have been used in pollen substitute and supplementary diets for many years. Ten amino acids (arginine, histidine, leucine, isoleucine, lysine, methionine, phenyalalanine, threonine, tryptophan and valine) are considered ­essential for growth of bees and three others (glycine, proline and serine) which are not essential for growth exert a stimulating effect on suboptimal growth levels (De Groot 1953).

Chemical composition of fungal spores  

The general composition of fungal cells was comprehensively reviewed by Birkenshaw (1965), Lilly (1965) and Wolf (1982). As this was summarised by Shaw (1990b) it is not repeated here. However, protein records for spores of Melampsora occidentalis were 9.1% (McGregor 1978 personal communication), 25.9% for Puccinia graminis tritici (Shu et al. 1954) and 25.8% for Neurospora sitophila (Owens et al. 1958). One major and two minor carotenes have been reported for P. gr. tritici (Irvine et al. 1954) and eight for N. crassa (Goldie and Subden 1973).

The effect on the brood  

Very little information is available on the effect of ingestion of fungal spores on the brood. No deleterious effects were noticed after collection of the rust spores Caeoma nitens (Zabriskie 1875) in the USA. In Switzerland, the eating of rust spores (given as ‘Uredineae sp.’) collected by honey bees, increased the length of life of the bees to some extent, and had only a slight effect on the hypopharyngeal glands and fat body (Maurizio 1950). There were no adverse effects on the bees as a result of collecting rust spores of Zaghouania oleae during pollen dearths in India (Deodikar et al. 1958). On the other hand, the storing of spores of Melampsora euphorbiae in the comb in considerable quantity at three sites in ­Pakistan resulted in fairly high mortality of brood and bees (Ahmad and Ahmad 1986), although no details were given. In a survival study of honey bees using various pollens in the USA, Schmidt et al. (1987) included a collection of rust spores ascribed to ‘Uromyces euphorbiae?’ (a synonym of U. proeminens) of unknown plant source, which (with three pollens among several) induced decreased life span of bees. In later tests (Schmidt et al. 1989) honey bees did not avoid ‘a rust’ (unidentified) and some pollens, but did avoid pollen from three other plants.

In the pollen storage cells of the hive, pollen (and fungal spores if any) would be mixed, so brood would receive a varied diet. Further experiments are required on the effect of feeding identified fungal spores to bees as a sole diet and in mixtures with identified pollens.

Plant pathological risk  

Several workers considered that rust spores collected in lieu of pollen might cause dissemination of the pathogen. Lang (1901), for example, thought this might be the case for the rust (caused by Caeoma nitens) of wild dewberry in the USA, and Turner (1974) wondered whether the Giant Asian honey bee (Apis dorsata) could assist in the spread of tropical maize rust (caused by Puccinia polysora). Other workers, however, considered this unlikely. Kraus (1920) was convinced that bees gathering conidia of the apple mildew fungus (Oidium farinosum) never visited healthy leaves, so that there was little if any chance of ­infection. Moraal (1988) also considered it unlikely that the collection of poplar rust spores in the Netherlands had much effect on the development and propagation of the disease.

As was pointed out previously (Shaw 1990b) and is reiterated here, spores packed into the corbiculae in lieu of pollen are moistened with liquid from the honey stomach, and this, and the physical compaction in the corbiculae and later in the hive cells, may exclude them from dissemination. Spores may also adhere loosely to the bee’s body parts and hairs (as do pollen grains (Free and Williams 1972) and as illustrated by Leach (1940)), and as found with tropical maize rust spores (Turner 1974) and conidia of Neurospora sitophila (Shaw (1993), ­although auto-grooming of such grains is part of the pollen-collecting process in honey bees. There is therefore the possibility that there may be in-hive transfer of any spores remaining on the body after auto- or during allo-grooming, or in-hive contact. No efficient transfer of avocado pollen, however, occurred within the hive in tests carried out recently (Ish-Am and Eisikowitch 1998). Also, even if spores were transferred during in-hive contact, or if launched into air currents if blown from the bee’s body, they would still need to be viable at the time, and land on a susceptible host and find suitable conditions for germination.

Also, the bees will probably return again and again to a profitable source, viz., the pathogen on already infected hosts. This is a different state of affairs from the incidental transport of pathogenic propagules during normal foraging, when bees (and other insects) visit uninfected as well as infected plants. Therefore, while this aspect – the plant pathological risk involved in the collection of spores in lieu of pollen – should not be overlooked, there would seem little chance of transmission, even given optimum conditions for infection.

Future records  

More observers are required to determine whether bees are making further collections, and if so, what fungi from what hosts or substrates are involved. It would seem desirable for any future records to include information on pathological, mycological, apicultural and environmental aspects, such as listed previously (Shaw 1990b). The amount of fungal spores collected is infinitesimal when compared with the amount of pollen ­collected, but it has occurred, and no doubt will ­continue to occur, in some cases at least in crisis situations during periods of pollen dearth (Shaw 1990b).

The chemical composition of the spores, especially the protein levels of at least some of them, would seem to remove these fungi from the category of worthless substances, as in Faegri and Pijl (1979), and place them rather in the category of nutrients. Neurospora, for example, is non-pathogenic to man, other animals and plants, and is already used as a human food in Indonesia and Sarawak.

Recent work on ancient bees

Bees and wasps have long been thought to have evolved in the early Cretaceous (146 to 65 mya or Ma) with flowering plants. The origin of flowering plants was itself regarded by Darwin as an ‘abominable mystery’. And what is the position 150 years later regarding the origin of flowering plants? The most recent ordinal classification (APG 1998) stated (in modern terms) that the position of the root of the flowering plant phylogeny remains elusive. The work published less than a year ago on what is claimed to be the earliest flowering plant, Archaefructus liaoningensis, gives its date at 140 mya or earlier (Sun et al. 1998).

Nests and groups of cocoons (ichnofossils) ­attributed to ancient bees have now been found in strata dated at 220 mya in the Petrified Forest ­National Park, USA. This work extends the range of bees and their nesting behaviour (reflecting ­social interaction) by 140–155 million years to 220 mya, and indicates that bees and wasps almost certainly predate flowering plants by at least 100 million years (Bown et al. 1997); Hasiotis 1997; 1998 (1999); Hasiotis et al. 1995; 1996a; 1996b). The ichnofossil cell linings provide biochemical evidence of a phylogenetic link to modern bees in the Colletidae and Anthophoridae (Kay et al. 1997). This work also raises the question as to the nutrients of the ancient bees (if they predated flowering plants) and Hasiotis et al. (1999b) suggested that they ­probably foraged on carrion, gymnosperm and cycad pollen, resins, plant juices and fungi.

It is not known whether fungi were present at the time, as their fossil record does not extend back that far. Rust fungi in Hyalopsora, Milesina, and Urediniopsis are considered by some rust specialists to be the most primitive of the rusts – (although Hart (1988), after a phylogenetic study, placed the suprastomatal tropical rusts as the most primitive) – and these three genera are heteroecious rusts, i.e., they have host alternation between ferns and firs. It is known that present day bees will collect rust spores from present day gymnosperms, such as honey bees have done in Mexico with spores of Cronartium conigenum on pine (Table 2), but whether the ancient bees did so, and whether there was host alternation at that time, is not known. ­Although of quite a different magnitude to the origin of the flowering plants, I think that the origin of host alternation in the rust fungi is another ‘abominable mystery’. This origin may have happened millions of years ago, but the fact of alternation of hosts is with us today, and I suggest that this should now be investigated with molecular tools. The study should not only include the disparity of the host pairs on the telial and aecial plants, but the inter-woven aspects as to why haploid basidiospores borne on the telial host cannot infect that host, and why the binucleate aeciospores borne on the aecial host cannot infect that host, but can infect the telial host.

The body fossil insect records include a Meliponine bee in New Jersey amber dated at 80 mya and a stingless bee with resin in the corbiculae at 20 mya (Roubik 1989; Poinar 1993; Ross 1998). What would be highly desirable would be the finding in authentic amber of an ancient bee with fungal spores in the corbiculae. If ever such is found, it would give important data on the age of the fungus and on bee behaviour.

Acknowledgements

Slide illustrations shown during the lecture were those of the author (some previously published in the Transactions of the British Mycological Society and The Mycologist) or reproduced from publications of the following: The Archaeology of Beekeeping, and The Pollen Loads of the Honey Bee: International Bee Research Association, Cardiff, CF10 3DT, UK; The Hive and the Honey Bee: Dadant & Sons, Hamilton, Ill., USA; Guide to Bees and Honey: Cassell, Wellington House, Strand, London, UK; Nature’s Use of Colour in Plants and Flowers: Eurobook, Wallingford, Oxon OX10 0XU, UK; Amber: The Natural Time Capsule: The Natural History Museum, London, UK. The original slide of powdery mildew on pumpkin is held by the ­Department of Primary Industries, Indooroopilly, and that of Neurospora sp. on a heap of filter mud by Drs G.R. and A.M. Stirling. The use of facilities at Indooroopilly is gratefully acknowledged.

References

Table 2 Records of fungal spores collected in lieu of pollen by honey bees (Apis mellifera), two records by A. dorsata and two records by species of Trigona