Chemical Mimicry in Pollination

Elisa J. Bernklau

Entomology Dept.

Colorado State University

Fort Collins, CO 80523

bernklau@lamar.colostate.edu

Introduction

The use of chemical signals may be the dominant form of communication in the insect world.а Many insects have evolved highly complex and specific chemical signals with 

which to communicate within their own species.а It is not surprising that other 

organisms have evolved the ability to exploit these communication systems in order 

to fulfill their own needs.а One method of exploitation involves mimicking the 

chemical signals used by insects.а In a system of chemical mimicry, a single 

compound or a mixture of compounds is produced by an organism to elicit a specific 

behavioral response by an organism of a different species.а 

Chemical mimicry, which is used by vertebrates, invertebrates, plants and fungi, can 

be divided into unique categories based on the outcome of the relationship.а These 

different classes include aggressive mimicry, reproductive mimicry, dispersal 

mimicry, group mimicry, and predator mimicry (Wiens 1978).а One of the most 

intriguing systems is the use of chemical mimicry by plants in order to attract 

insect pollinators.а 

Most pollination systems have evolved as mutualistic relationships in which both 

organisms are rewarded:а the insect obtaining pollen, nectar, waxes or scents from 

the flower and the plant achieving reproduction through the transfer of its pollen 

by the insect.а Chemical mimicry has evolved in insect/flower systems where no 

reward is received by the insect.а The plant, which is the sole benefactor, lures 

the insect to its flower by producing specific scents.а These odors may mimic insect 

pheromones, food sources, broodsites or prey odors.

Pollination mimicry has been studied in some detail and the literature contains 

several excellent reviews of the subject.а A 1978 review entitled Mimicry in Plantsа 

(Wiens 1978) includes a section on pollination mimicry in which Wiens briefly 

describes some different categories of chemical mimicry and gives a few examples of 

each.а A chapter entitled 'Chemical Mimicry' (Stowe 1988), in the book Chemical 

Mediation of Coevolution, contains a review of all known chemical mimicry systems.а 

In the section on pollination mimicry, Stowe cites numerous examples of pheromone 

and brood-site mimicry.а He summarizes the similarities in compounds (identified to 

date) produced by flowers and their insect pollinators and provides some possible 

explanations for differences that have been found between insect and plant-produced 

compounds.а He also suggests possible evolutionary patterns for various classes of 

chemical mimicry.а In a 1990 article, Borg-Karlson (1990) provides a comprehensive 

review of the relationships between orchids of the genus Ophrys and their insect 

pollinators.а In this review, Borg-Karlson describes the detailed chemical analyses 

and field studies conducted by herself and her associates.а She also discusses 

previous field work and chemical identifications carried out by other experts in the 

field (Bergstrom, Paulus, Kullenberg).а A 1994 review by Dettner (1994) provides 

examples of particular compounds that are produced by plants in different 

pollination systems and the specific insects that are effected by them.

Plant Mimicry of Insect Pheromones

Some form of mimicry is used by approximately 50% of orchid species to attract 

insects of specific taxa for pollination.а Visual mimicry of female insects by the 

flowers was believed to be the main factor in these relationships until 1961 when 

Kullenberg first suggested that chemical mimicry may also play a role (Kullenberg 

1961).а Researchers have since demonstrated this to be the case and the chemistry of 

many pollination mimicry systems is now known.а According to Borg-Karlson (1990), 

the most important studies on the Ophrys orchid/insect pollination systems are those 

compiled by Kullenberg (1973), Kullenberg et al. (1984), Warncke and Kullenberg 

(1984) and Paulus and Gack (1986).

In these systems, male insects are attracted by odors produced by the flowers 

(specifically by the labellum).а These odors excite the insects sexually and cause 

them to approach and investigate the flowers more closely.а In some instances, the 

males will go so far as to attempt to copulate with the flower.а This behavior, 

termed 'pseudocopulation', was first observed by Correvon and Pauyanne (Pouyanne 

1917; Correvon and Pouyanne 1923).а During the insect/flower interaction, pollinaria 

are inadvertently attached to the head or abdomen of the male insect who then 

pollinates the next flower that he visits (Borg-Karlson 1990).

The odors produced by these orchids are blends of several compounds that closely 

resemble different types of insect pheromones.а These include:а odors produced by 

eclosing females, sex attractants produced by virgin females, scents produced by 

mated females, male courtship pheromones, and male scent markers (Stowe 1988).а A 

flower species or variety may produce one specific type of odor or a combination of 

the different pheromones.

Visual and Tactile Cues

In addition to the chemical attractants, many of the orchid species use visual and 

tactile cues to dupe their pollinators.а The combination of visual, tactile and 

olfactory cues may mimic a female insect so well that the male is not able to 

distinguish the difference.а The pheromone odor cues serve to attract patrolling 

insects to the flowers from a distance.а Many of these compounds also stimulate the 

males sexually.а The visual stimuli then serve as close-range attractants that, in 

some cases, cause the sexually-stimulated males to actually alight on the flower.а 

Borg-Karlson (1986) noted that in some cases Andrena bees will not alight on the 

odor source without the proper visual cues.а However, the visual cues are not 

important for all hymenoptera species.а Tactile cues such as the firmness and 

smoothness of the labellum help to guide the male into a position that insures his 

contact with the pollinaria.

Examples of Pheromone Mimicry

All but one example of pheromone mimicry in plants occur in the family Orchidaceae.а 

At least 30 species of one genus (Ophrys) produce compounds in common with their 

insect pollinators.а While most of the insects involved are bees and wasps, a few 

plants depend upon beetles for pollination (Stowe 1988).а 

Because this phenomenon is so widespread in the orchid family, it has been 

extensively studied and many of the specific insect/orchid relationships have been 

defined.а For example, orchids of the genus Ophrys are pollinated by males of the 

hymenoptera families Andrenidae, Anthophoridae, Colletidae, Megachilidae, Sphecidae 

and Scoliidae and by two species of beetle (Elateridae and Scarabaeidae) (Borg-

Karlson 1990).а All Cryptostylis flowers attract Lissopimpla excelsa wasps 

(Ichneumonidae).а Wasps of the subfamily Thynninae are attracted to related orchids 

(Chiloglottis, Drakaea, Caladenia, Arthrochilus, Paracaleana, and Specula) (Stowe 

1988).а Calochilus attracts scoliid wasps, Caleana attracts pergid sawflies,Ophrys 

diurisа and O. galilaea attract male halictid bees (Halictus marginatus) (Borg-

Karlson et al. 1985), and Leporella fimbriata attracts male alate ants (Myrmecia 

sp.) (Stowe 1988).а Guiera senegalensis (Combretaceae), which is the only non-orchid 

plant known to mimic insect sex pheromones (Kullenberg 1961) attracts male sphecid 

wasps (Tachysphex).а A very complete listing of specific orchids and their 

pollinators was compiled by Kullenberg (1973; 1976; 1985).

Chemical Analysis

A review by Borg-Karlson (Borg-Karlson 1990) details the chemical analysis of Orchid 

volatiles and field studies of the insect-Ophrys relationships conducted by herself 

and her associates and by other researchers (Bergstrom, Paulus, Kullenberg).а 

Chemical analyses resulted in the identification of the main compounds produced by 

flowers in the Ophrysа genus as well as by Andrena, Eucera and Colletes bees.а In 

addition, these studies revealed similarities between the insects and their 

associated flowers and helped to determine the specific compounds responsible for 

the attraction in some of these systems.

A combination of techniques was used for extracting flower and insect specimens to 

insure the collection of compounds of both low and high volatility.а These methods 

included making sorption extracts using Porapak Q or Tenax GC, low temperature 

trapping with liquid nitrogen or carbon dioxide in ethanol, and extracting the 

flower and insect parts in hexane or methanol (Borg-Karlson 1990).а GC-MS was used 

to identify particular compounds. 

Bioassays and Field Studies

 After identifying the key components of the volatile blends, the effects of 

collected volatile samples as well as synthetic mixtures of those blends on the 

behavior of the insects were tested.а Field studies were conducted in the natural 

mating areas of the insects using a revised version of a bioassay that was 

originally developedа by Kullenberg (1961).а In the bioassays, they applied the test 

compounds to insect "dummies" which were constructed of black nylon velvet attached 

to an insect pin.а Physiological responses of the insects to key compounds were 

tested in the laboratory using an electroantennagram technique.

Compounds Isolated from Insects and Flowers

Using the techniques described above, Borg-Karlson detected approximately 100 

volatile compounds in the flower extracts (Borg-Karlson et al. 1985).а Of these, 

aliphatics and terpenoids were determined to be the main components.а The aliphatic 

compounds, including saturated hydrocarbons, aldehydes (octanal and nonanal), 1- and 

2-alcohols and esters, were always present in the largest quantities. аTerpenoids 

(geraniol, citronellol and E,E-farnesol) were always present as well, but the 

amounts varied greatly (Borg-Karlson 1990).а 

Compounds isolated from the mandibular glands of Andrena bees include aliphatics (1- 

and 2-alcohols, 2-ketones and esters) and terpenoids (geraniol, citronellol and 

farnesol).а Borg-Karlson separated the Andrena species into four distinct groups 

based upon the mixtures of these volatiles.а Oxygenated open chain mono- and 

sesquiterpenes (geranial and neral) and their alcohol derivatives were found to be 

the main constituents in the first group.а A second group of insects shares a 

mixture of geranial, neral, their alcohols, aliphatic alcohols and aldehydes.а A 

third group was found to produce mainly aliphatic alcohols, esters, and 2-ketones, 

along with smaller amounts of monoterpenes (geranial) and spiroacetals.а The fourth 

group of bees included only one species (Andrena squalida), which produces only a 

series of esters.а 

Compounds isolated from the Dufour's gland of Andrena bees include terpene esters 

and aliphatic esters (Borg-Karlson 1990).а 

Colletes bees contain linalool, geranial and neral in their mandibular gland 

secretions.а Cephalic secretions of the Eucera bees have only a few volatile 

compounds and these included alcohols, terpenoids and linalool (Borg-Karlson 1990).

Similarities Between Floral and Insect Volatiles

Many similarities were found in the chemicals produced by the flowers and by their 

pollinators.а In fact, several Ophrys sp. release compounds identical to those found 

in the mandibular and/or Dufour's glands of their pollinators.а Similarities are 

seen between Ophrys lutea and Ophrys fusciа and several species of Andrena bees, 

Ophrys arachnitiformes-araneiferae and bees of both the Andrena and Colletes genera, 

Ophrys insectifera and two digger wasps (Argogorytes mystaceus and A. fargei ) and 

Ophrys scolopax heldreichiiа and males of the bee species Tetralonia cressaа (Borg-

Karlson 1990).а Volatile compounds common to Andrena bees and Ophrys flowers 

include1-and 2-alcohols, 2-ketones, acetates, esters, geraniol, geranial, 

citronellol, citronellal, linalool, E,E-farnesol and farnesol esters.а Colletes bees 

and their flower hosts both produce geranial neral and linalool.а Eucera bees share 

geranial and linalool, benzaldehyde, 4-methyl phenol, 1,4-benzoquinone, methyl 

esters, ethyl esters and acetates with their flowers (Borg-Karlson 1990).а 

Behavioral Responses of Pollinators

Some of these compounds were found to be responsible for the attraction of male bees 

and wasps in the field tests.а The most important components in the Andrena-Fuscii-

Luteae and Arachnitiformes-Araneiferae attractions were concluded to be geraniol, 

E,E-farnesol and aliphatic 1-alcohols (Borg-Karlson 1990).а Andrena bees and digger 

wasps (Argogorytes spp. and Campsoscolia ciliata) showed the most vigorous responses 

to these compounds.а Borg-Karlson notes that in field tests these insects began to 

expose their genitalia before they even encountered the model dummies.а Males of 

Andrena squalida responded to odors of Ophrys splendida by eliciting a digging 

behavior that is common to sexually-stimulated bees.

In related studies it was found that some coleopteran pollinators are attracted to 

the alcohol compounds produced by some orchid species (Dettner and Liepert 1994).

Pollinator Specificity 

Taxonomy of Ophrys is currently based upon morphological characteristics.а Borg-

Karlson notes that this system is often confusing because consistent differences 

between species-subspecies, varieties, forms and hybrids can be very difficult to 

determine by morphology.а Kullenberg and Bergstrom (Kullenberg and Bergstrom 1976) 

have proposed that a taxonomic division based on pollination, insect attraction and 

chemical analysis of insect and flower volatiles might be more exact because the 

flower selection of particular insects to particular flower scents is fairly 

selective. 

The exact degree of selectivity in these mimicry pollination systems is not yet 

clear (Borg-Karlson 1990).а The 'one-pollinator-one-Ophrys taxon' hypothesis is 

supported by field studies conducted so far.а Paulus and Gack (Paulus and Gack 1986) 

observed that four different forms of Ophrys fusca were pollinated by four different 

genera of hymenoptera.а In addition, Kullenberg proposed that Andrena bees can be 

divided into three categories based on their attraction to O. fusca and O. lutea 

(Kullenberg et al. 1985).а Some degree of specificity is also supported by the 

chemical similarities shown by Borg-Karlson's analysis.а However, other studies show 

that a variety of insects that are all of a particular 'physiological type' are 

attracted to Ophrysа flowers that are of a similar 'biochemical type' (Borg-Karlson 

1990).а For example, in Europe and Africa Andrena flavipesа pollinates several taxa 

of Ophrys fusci-luteaeа (Warncke and Kullenberg 1984).а 

Borg-Karlson (1990) points out that the odors emitted by Ophrysа flowers often 

include mixtures of compounds that are sure to attract more than one species of 

insect (at least within a particular genus).а The key components of the insect 

attraction that were identified by Borg-Karlson can be found in most Ophrysа 

species, but the amounts and proportions of these compounds vary greatly.а 

Quantitative differences may be partly responsible for the variation in attraction 

of different pollinators to Ophrys flowers.а Visual and tactile cues may also play 

an important part in the final selection of flowers by specific insects (Borg-

Karlson 1990).

Variation in Plant and Insect Odors

The large number of compounds involved in odor production as well as the variations 

that occur among individual plants make it difficult to determine the exact mixture 

produced by any one flower taxon.а Even though the chemical similarities have been 

determined and the behavioral effects of certain compounds have been demonstrated, 

it is still difficult to know exactly what is going on in the natural setting.а It 

is possible that the congruence between some insect and flower odors is actually 

much closer in nature than can be shown in the laboratory.а As pointed out by Stowe 

(1988), some compounds that are extracted and identified from the insects and/or 

flowers may never actually be released by the organism.а In addition, the mixtures 

of compounds that are released (especially by the flowers) may vary at different 

times of the day.а Important signal compounds that are produced in tiny amounts may 

be missed or overlooked during chemical identification.а Finally, variations in the 

volatile mixtures between individual plants may not be discerned when flowers are 

pooled into a single mix for extraction.а 

On the other hand, differences between the plant and insect odors in the natural 

setting may be greater than has been shown in the laboratory.а Extra compounds found 

in the plant odor blends that are not produced by the insects may greatly effect the 

overall odor and subsequent behavior of the insects.а Also, plant volatile blends 

may not include some compounds that are generated in the insect glands.а In 

addition, the ratios of individual compounds may vary between the insect and the 

plant.а Finally, the mixtures produced by the plants may change greatly throughout a 

24 hour period.а 

Broodsite or Food Mimicry

Some flowers mimic the odors of dung and/or carrion to attract insects (mostly 

beetles and flies) for pollination.а These systems are widespread in the plant 

kingdom and have been extensively reported on.а This type of relationship is found 

in ten plant families (Annonaceae, Araceae, Aristolochiaceae, Asclepiadaceae, 

Burmanniaceae, Hydnoraceae, Orchidaceae, Rafflesiaceae, Sterculiaceae, and 

Taccaceae) (Wiens 1978).

The odors produced by these flowers smell like decaying protein or feces and are 

very unpleasant to humans.а However, the scents are highly attractive to 

coprophilous beetles and flies that oviposit or feed at carrion or dung.а 

While the primary cues that the insects respond to are olfactory, secondary visual 

cues are produced by some plants.а For example, several species of Araceae have 

enlarged bracts, mottled purple petals and attenuated appendages.а With a 

combination of these features, the flower bears a striking resemblance to a mass of 

foraging flies (Wiens 1978).

Chemistry of Dung/Carrion Mimicry

Chemicals that are responsible for the carrion-like odors include ammonia, 

alkylamines, cadaverine and putrescine.а More fecal-like odors are also produced by 

skatole and indole (Dettner and Liepert 1994).а These odors are enhanced in some 

flowers (of the Araceae) by the production of heat and carbon dioxide.а Heat 

increases the production of volatile compounds and carbon dioxide, which is produced 

as a by-product of the heat production may increase the length of time that an 

insect remains at the flower (Dafni 1984).

PollinatorTraps

In addition to mimicking attractive odors, many dung or carrion flowers have evolved 

elaborate traps that increase the likelihood of pollination.а In general, flies and 

beetles are not active pollinators and must be induced into having adequate contact 

with the reproductive structures of the flower (Wiens 1978).а To insure pollination, 

these plants mimic food or brood-site odors to attract the insects to the flowers 

and lure them inside.а Once inside, the pollinators are trapped in the flower for an 

extended period of time.а In most systems the insects are eventually released by the 

flower or allowed to escape, but in a few cases the insects are permanently trapped 

or even killed (Stowe 1988).

Mimicry of Decaying Fruit

In a similar mimicry system, flowers produce odors that resemble those of either 

fresh or decaying fruit.а The main pollinators in these cases are beetles and small 

flies.а Methylesters may play a role in this attraction as they are present in many 

fruit odors as well as in many flower scents (Stowe 1988), (Gottsberger 1977).а 

Prey Mimicry

Wiens (1978) reported on an unusual variation of reproductive mimicry in which 

plants attract parasite or predator pollinators by mimicking their prey (hosts).а In 

some cases, the insect is rewarded with nectar and in other cases there is no 

reward.а Occasionally, the insect is duped into laying its eggs in the flower.а 

Because the flower is not a true broodsite, the eggs either do not hatch or the 

hatchlings die for lack of proper food (Stowe 1988).а In these mimicry systems, the 

flower is sometimes pollinated when a parasite tries to sting what they think is 

their prey on the flower.а In other cases, the flowers lure their pollinators inside 

and use elaborate traps to contain the insects long enough to insure pollination.а 

Examples

Three genera of orchid (Ada, Brassia and Encyclia) are known to use this 

"pseudoparasitism" method of attracting pollinators.а In addition, two unrelated 

genera of orchids (Epipactis and Paphiopedilum) may mimic the compounds produced by 

aphids in order to attract female syrphid flies that normally lay their eggs near 

aphid infestations.а According to Stowe (1988), visual cues and nectar are also 

important in attracting the flies to the flowers.а Because the presence of aphid 

odors stimulates oviposition by the flies, it is likely that the plants are indeed 

mimicking the chemistry of aphids even though these examples have not been verified.а 

One species of orchid (Oberonia thwaitesii) that is pollinated by ants may also 

mimic aphid odors (Faegri and van der Pijl 1979). 

Cost/Benefit of Pollination Mimicry

One question that is often debated in the literature is whether or not pollinators 

are impaired by their participation in these systems where no reward is received for 

their pollination efforts.а In the systems involving mimicry of food or broodsite 

odors, the cost to the insects varies.а In some cases, a duped insect only loses 

some time while searching for food that is not there or while he is trapped inside a 

flower for an extended period.а The cost is much higher, however, when a female is 

fooled into laying her eggs on a false broodsite.а 

The costs and/or benefits to the pollinators in systems where plants mimic female 

insects are more difficult to determine.а Kullenberg (1985) and Burns-Balogh (1985) 

believe that the non-rewarding flowers may actually benefit their pollinators.а For 

example, the flowers serve to concentrate male insects in a suitable habitat where 

females are likely to be found.а In addition, the males who are sexually stimulated 

after their encounters with flowers may spend more time looking for females to mate 

with.а 

On the other hand, Stowe (1988) argues that the non-rewarding mimics are harmful to 

their pollinators. Because the male insects lose time that would otherwise be spent 

seeking females, they ultimately achieve lower mating success.а He argues that there 

is much less profit in time spent visiting the flowers than in time spent in 

searching for females.а He believes that males who can discriminate between flowers 

and female insects will have the selective advantage.

However, these plants depend upon insects for pollination and it would not be to 

their ultimate advantage to out-compete female insects for the attention of the 

males.а As pointed out by Weins (1978) some of the plants in these systems produce 

flowers after male pollinators have become active but before the females emerge.а In 

this way, they are able to acheive pollination and still avoid competition with 

females.

References Cited

Borg-Karlson, A.K. 1990. Chemical and ethological studies of pollination in the 

genus (Ophrys) (Orchidaceae). Phytochemistry 29: 1359-87.

Borg-Karlson, A.K., Berstrom, G., and Groth, I. 1985. Chemical basis for the 

relationship between Ophrys orchids and their pollinators II.а Volatile compounds of 

Orphrys lutea and O. fusca as insect mimetic attractants/excitants. Chem. Scr. 25: 

283-94.

Borg-Karlson, A.K., and Tengo, J. 1986. Odor mimetism?а Key substances in Ophrys 

lutea-Andrena pollination relationship (Orchidaceae: Andrenidae). J. Chem. Ecol. 12: 

1927-41.

Burns-Balogh, P., Borg-Karlson, A.K., and Kullenberg, B. 1985. Evolution of the 

monandraceous Orchidaceae VI.а Evolution and pollination mechanisms in the subfamily 

Orchidoidea. Can. Orchid J. 3: 29-57.

Correvon, H., and Pouyanne, M. 1923. Ibid. 24: 372.

Dafni, A. 1984. Mimicry and deception in pollination. Annu. Rev. Ecol. Syst. 15: 

259-78.

Dettner, K., and Liepert, C. 1994. Chemical mimicry and camouflage. Annu. Rev. 

Entomol. 39: 129-54.

Faegri, K., and van der Pijl, L. 1979. The Principles of Pollination Ecology. 102-

106.

Gottsberger, G. 1977. Some aspects of beetle pollination in the evolution of 

flowering plants. Plant Syst. Evol. Suppl. 1: 211-226.

Kullenberg, B. 1961. Studies in Ophrys pollination. Zool. Bidr. Uppsala 34: 1-340.

Kullenberg, B., and Bergstrom, G. 1973. The pollination of Ophrys orchids. In 

Chemistry in Botanical Classification. Nobel Symp., Stockholm, Nobel Foundation.

Kullenberg, B., and Bergstrom, G. 1976. Hymenoptera Aculeata males as pollinators of 

Ophrys orchids. Zool Scripta 5: 13-23.

Kullenberg, B., Borg-Karlson, A.K., and Kullenberg, A.L. 1985. Field studies on the 

behavior of the Eucera nigrilabris male in the odour flow from the flower labellum 

extract of Ophrys tenthredinifera. Nova Acta Reg. Soc. Sci. Ups. Ser. V:C 3: 79-110.

Kullenberg, B., Buel, H., and Tkalcu, B. 1984. ├bersicht von Beobachtungen ÷ber 

Besuche von Eucera- und Tetralonia M┼nnchen auf Ophrys - Bl÷ten (Orchidaceae). Nova 

Acta Reg. Soc. Sci. Ups. Ser. V:C 3: 

Paulus, H.F., and Gack, C. 1986. Neue Befunde zur Pseudokopulation und 

Best┼uberspezifit┼t in der Orchideengattung Ophrys - Untersuchungen in Kreta, 

S÷ditalien und Israel. Die Orchidee (Sonderheft) 39: 48-86.

Pouyanne, M. 1917. La f▌condation des Ophrys par les insectes. Bull. Soc. Hist. nat. 

Afr. N. 8: 6-7.

Stowe, M.K. 1988. Chemical Mimicry. In Chemical Mediation of Coevolution. K. C. 

Spencer, eds.а Academic Press, Inc., London.

Warncke, K., and Kullenberg, B. 1984. ├bersicht von Beobachtungen ÷ber Besuche von 

Andrena- undа Colletes cunicaularius - M┼nnchen auf Ophrys-Bluten (Orchidaceae). 

Nova Acta Teg. Soc. Sci. Ups. Ser. V:C 3: 

Wiens, D. 1978. Mimicry in plants. Evol. Biol. 11: 365-403.ааааааааааааааааааааааааааааааа