Insects pollinating flowers are a familiar sight. But what happens when the "flower" is actually a fungus? Still "pollination", but now it is fungal spores. Read on to learn more about the fungi that mimic flowers...

Most of us are familiar with fungi in the form of mushrooms, some of which are brightly coloured and not likely to be mistaken for anything else. Certain fungi, however, are adept at disguising themselves – they are accomplished mimics, and there are multiple examples. Whilst animals employ mimicry mainly for protection, mimicry in fungi can serve different functions. It can be used to aid the dispersal of infectious spores, to facilitate gamete transfer in sexually reproducing fungi or to circumvent host defences.

Flower mimicry

Monilinia

The economically important blueberry can be seriously affected by the so-called mummy berry disease that is caused by a pathogenic ascomycete fungus (Monilinia vaccinii-corymbosi). This fungus first infects young shoots and leaves by ascospores, in other words sexual spores produced by meiosis that are ejected from the so-called asci. This infection then results in the formation of asexual spores called conidia, which infect blueberry flowers and cause them to produce mummified berries. It is during this second stage of infection that the most remarkable thing happens…

As the conidia are non-motile, they need to be transferred from the leaf or shoot onto open flowers. The fungus achieves this by an ingenious trick – it basically transforms the blueberry leaves into flowers. Not only do they reflect ultraviolet light at a wavelength similar to that reflected by the calyx of blueberry flowers, but they also produce a sweet odour. Thus, pollinators are attracted to the leaves and so transmit the conidia to the next flower they visit. The conidia will then germinate on the surface of the stigma, using up an exudate originally intended for pollen growth. Fungal hyphae enter the plant’s ovary through the style, so using exactly the same path that the pollen tubes would take. Generally, only very few highly specialised plant-pathogenic fungi infect their host through the stylar canal, presumably because pollen-pistil interactions are highly coordinated, involving specific signals. So do the fungal hyphae mimic the pollen tube? A recent study has suggested that this is indeed the case. Similar to pollen tubes, the conidial germ tubes of M. vaccinii-corymbosi showed highly directional growth, their tips adhering selectively to certain regions of the stylar canal. It is not yet known whether the pseudo-pollen tube co-opts the plant’s biochemical machinery or makes its own to ensure it is correctly guided (a case of such molecular mimicry has been proposed for two other fungal plant pathogens, Rhizoctonia solani and Phoma lingam). Hence the pathogenic fungus seems to have specialised on exploiting the blueberry’s reproductive infrastructure by imitating pollen growth anatomically and physiologically.

Smut fungi

Some basidiomycete smut fungi also employ floral mimicry to facilitate spore dispersal. The anther smut fungus (Microbotryum violaceum), which is an obligate parasite of many species of carnation (Caryophyllaceae), affects the plant’s actual flowers. They become sterile, with their anthers bearing the fungal spores rather than pollen. Pollinating insects transfer the spores onto healthy flowers, thus spreading infection. Spore transfer is enhanced by infected flowers blooming not only earlier but also for longer compared to healthy flowers. Curiously, a number of herbarium specimens (e.g. of Silene and Lychnis) had been misidentified as new species when in fact they were simply examples with infected anthers!

Rust fungi

Even more sophisticated are flower mimics termed pseudoflowers that not only fool insects, but also crab spiders (and botany students). They are induced by rust fungi, an important group of basidiomycetes. Best studied is Puccinia monoica that infects several species of Brassicaceae (e.g. Arabis spp.). This fungus dramatically alters the morphology of its host to facilitate gamete transfer among pseudoflowers. Its wind-borne spores infect the plant and prevent it from flowering. Instead, they induce the formation of elevated leaf rosettes that resemble the flowers of unrelated plants such as buttercups in size, shape, colour, smell and even nectar production! These pseudoflowers attract pollinating insects, which then carry the fungal gametes. Most rusts rely on insects for sexual reproduction, as they have outcrossing mating systems and gametes need to be transferred between different mating types. Interestingly, the pseudoflower scent, which consists of variable, mainly aromatic compounds, seems to go beyond simply mimicking or modifying the scent of blooming flowers and might serve to reinforce flower constancy among pollinators, thus increasing the reproductive chances of the fungus by reducing gamete loss. Pseudoflowers can have considerable effects on the reproductive success of nearby angiosperms, negative (by keeping pollinators for relatively long visits) as well as positive (by increasing overall pollinator visitation rate). Thus, the fungus not only affects its host plant, but also visiting insects and other flowering plants in the vicinity.

Other rust species are also known to induce pseudoflowers, e g. Puccinia arrhenatheri on the European barberry (Berberis vulgaris), P. punctiformis on the Canada thistle (Cirsium arvense) and Uromyces pisi on the cypress spurge (Euphorbia cyparissias).

Interestingly, mimicry can also work the other way round – fungus gnat flowers that are pollinated by flies mimic the fruiting bodies of mushrooms both morphologically and chemically. Examples are the extraordinary orchid Dracula chestertonii, which has a large lip resembling a mushroom cap and smells like champignons (producing typical mushroom scent components such as octan-3-ol), as well as several species in the families Araceae and Aristolochiaceae.

Carrion and faeces mimicry

Many flowers smell rather unpleasant, their scent imitating that of carrion or faeces so as to attract pollinating flies. But not only angiosperms (e.g. Rafflesia) have come up with this trick – some mosses and the so-called stinkhorn fungi (Phallaceae) use it to achieve spore dispersal. All these mimics emit volatiles that are also released by carrion or faeces, which many flies use as food sources and egg-laying sites. The deceptive stinkhorn fungi, however, do not seem to provide a brood place for larvae and do not offer much nutrition for the flies either (although they produce an edible secretion called gleba, which is consumed together with the spores that then germinate after passage through the insect’s digestive system). A gas chromatographic analysis of the scents of the octopus stinkhorn (Clathrus archeri) – thus named for its tentacle-like, foul-smelling fruiting body – and several fly-pollinated angiosperms provided evidence for convergent evolution in scent chemistry. The species investigated had similar scent profiles with compounds that are characteristic of carrion (e.g. oligosulphides) and faeces (e.g. phenol).

Termite egg mimicry

One of the most astonishing examples of an insect-fungus association is the mimicry of termite eggs by fungi. Termites are, of course, well known for their symbiosis with fungi that they farm for food. But this relationship is of a very different nature…

A deceptive fungus

In a termite colony, workers regularly collect the eggs the queen has laid and pile them together in a nursery chamber, where they are carefully looked after. Grooming, which is essential for egg survival, involves coating the eggs with antibiotic saliva, thus protecting them from desiccation and pathogenic infection. On close inspection, however, the egg piles of various termite species were found to contain a large number of so-called “termite balls” (sometimes, the number of termite balls even exceeds that of true eggs!). These small brown balls are fungal sclerotia, densely packed filaments that may germinate if conditions are favourable.

But how do they end up in a termite colony? Ingeniously, the fungus mimics termite eggs so closely that the workers cannot help but transport the balls into their nest and even groom them. As termite workers are virtually blind, this mimicry is not visual (in contrast to, for example, cuckoos, where the eggs often look remarkably like those of the host species). Instead, size and surface texture of the fungal balls seem to be of importance. Most termites were shown to care only for balls that have the same diameter as their eggs and are spherical and smooth. The smooth surface results from losing the outer shell-like layer that is typical of sclerotium-forming fungi. But such morphological camouflage is not enough – the balls also have to smell like termite eggs. It has recently been demonstrated that the egg recognition pheromone of termites mainly consists of two enzymes that are produced in the insects’ salivary glands as well as in their eggs and have been co-opted for chemical signalling – the antibacterial lysozyme and the cellulose-digesting b-glucosidase. While fungi do not produce lysozyme, b-glucosidase is, in fact, a common fungal enzyme (important not only for cellulose digestion in decay fungi, but also for pathogenicity in plant pathogens). Experiments confirmed that fungi in termite nests secrete b-glucosidase. The fact that both termites and fungi originally digested cellulose with the help of the same chemicals therefore provided the fungus with the opportunity to evolve chemical egg mimicry.

Symbiosis or parasitism?

The fungus clearly benefits from the association with termites, because it is transferred into a nearly competitor-free, protected environment. As a consequence of the loss of the outer shell, the sclerotia are prone to desiccation and infection, so grooming by the termite workers is necessary to ensure their survival. But what about the termites? An early experiment in the termite species Reticulitermes speratus provided evidence for enhanced survival of termite eggs in the presence of the fungus, thus supporting a mutualistic relationship. It was suggested that the fungus might produce antifungal and antibacterial compounds that keep putative pathogens in check. Later studies, however, concluded that this was most likely an artefact and found no difference in egg survival between termite colonies with and without fungal balls. As termites were not shown to consume termite balls, it is now generally assumed that the fungus is a parasite that exploits the brood care of its termite host – a “fungal cuckoo”, basically. How detrimental this exploitation actually is to the insect is not yet clear. The fungus does not germinate whilst in the egg pile but only afterwards. Old, shrivelled balls are removed from the pile and dumped into a corner of the nest, where they grow on termite excreta. Therefore, the fungus rarely kills termite eggs. However, the time and energy cost of caring for large numbers of fungal balls is probably substantial (at least in the short-term). By means of egg mimicry, the fungus has managed to gain access to termite nests and has therefore found an “evolutionary loophole” around the diverse defence strategies these and other eusocial insects have developed to prevent parasite invasion of their colonies. In a strikingly similar example, the “cuckoo butterfly” Phengaris alcon employs chemical mimicry to be picked up and cared for by workers of Myrmica ants.

The question remains whether the fungi are obligate or facultative parasites of the termites. Not much is known about the life cycle of termite egg-mimicking fungi, but recent phylogenetic analyses have indicated the existence of a free-living stage. Transmission between termite colonies seems to be horizontal, with fungi being brought intro the nest repeatedly by workers, and genetic analyses showed that termite colonies are often infected by multiple strains of fungi. This is in contrast to fungus-farming termites and ants that grow a monoculture of just one fungal strain, which in ants is maintained by vertical transmission of fungal cultivars from parent to offspring colonies and in termites probably by regular bottlenecking and other mechanisms. It has been argued that a mutualistic relationship should benefit from genetic uniformity, as competition between different fungal strains would lessen the benefit of the symbiosis, whereas parasitic fungi should gain from diversity as it increases parasitic virulence.

Evidence of convergence

Until recently, termite balls had only been found in species of the lower termite family Rhinotermitidae. Eight of these species belong to the temperate genus Reticulitermes, with four occurring in the United States (R. flavipes, R. virginicus, R. hageni and R. malletei) and four in Japan (R. speratus, R. kanmonensis, R. amamianus and R. miyatakei). The other species is the closely related Coptotermes formosanus, which probably originated from China, but is now widely distributed. In all these cases, the termite egg-mimicking fungus is an athelioid fungus (Basidiomycota, Agaricomycotina) of the genus Fibularhizoctonia (teleomorph: Athelia), and several phylogenetic analyses found no evidence of host race formation or geographic variation. It was therefore concluded that egg mimicry evolved only once in the atheloid group.

Intriguingly, however, another termite egg-mimicking fungus has recently been found in the higher termite family Termitidae. Trechispora sp. (Basidiomycota, Agaricomycotina) is a parasite of the subtropical termite Nasutitermes takasagoensis and phylogenetically distant from Fibularhizoctonia, thus suggesting two independent evolutionary origins of termite egg mimicry in the fungi. Both these fungi are sclerotia-forming decay fungi that are able to produce b-glucosidase, which probably served as a pre-adaptation for evolving termite balls. Interestingly, Reticulitermes and Nasutitermes show different size preferences – whilst the former only tend to termite balls that are similar in size to their own eggs, the latter favour termite balls that are slightly larger. It seems that both fungi have adapted the size of their termite balls to optimise acceptance.