Chemical Ecology

Tessa R Grasswitz & Graeme R Jones. Encyclopedia of Life Sciences: Supplementary Set. Volume 21, Wiley, 2007.

Chemical ecology is the study of the structure, origin and function of naturally occurring chemicals that mediate intraspecific or interspecific interactions. It has developed into a mature science with diverse practical applications. At the heart of the discipline are modern analytical instrumentation, careful observational biology and good bioassay design.


Chemical ecology may be defined as the study of the structure, origin and function of naturally occurring chemicals that mediate interactions between individuals of the same species (i.e. intraspecific interactions) and/or between individuals of different species (interspecific interactions).

The chemicals concerned (termed ‘semiochemicals’ from the Greek semeion, a mark or signal) are distinct from hormones in that they are released into the external environment. They are further divided into pheromones (which mediate intraspecific interactions) and allelochemicals (which mediate interspecific interactions).

Pheromones are normally classified according to their function: sex pheromones (which mediate interactions between the sexes of a species), aggregation pheromones (which may attract both sexes of adults, and in some cases immatures), alarm pheromones, and so on. Some pheromones (‘releaser’ pheromones) elicit an immediate behavioural response, while others (‘primer’ pheromones) have a slower-acting ‘priming’ effect on the animal’s physiological state.

Allelochemicals are subdivided into kairomones, allomones and synomones. Kairomones are chemicals that in some way benefit the receiving organism but not the emitting organism, while allomones benefit the emitting organism but not the receiver. Synomones benefit both the emitting and receiving organisms. Thus kairomones include, for example, host- or prey-associated compounds that act as recognition ‘cues’ for parasites and predators, and allomones include defensive compounds of various types. The floral compounds emitted by plants that attract pollinating insects are examples of synomones (provided that the pollinator derives some benefit from the interaction in the form of nectar or pollen). Note that the classification of a particular compound depends on the ecological context in which it is being considered: thus, the same chemical can act both as a pheromone and as a kairomone, depending on the interaction concerned. Semiochemicals are often produced in – and are biologically active at – very low concentrations, a fact that has necessitated the development of specialized techniques for their collection, identification and testing.

The first semiochemical to be chemically characterized was the sex pheromone of the female silkworm moth, Bombyx mori, in 1959. This first breakthrough took 20 years and the extract of half a million female moths to yield enough material to allow the chemical composition of the pheromone to be determined. Once this was accomplished, the material was synthesized in the laboratory and the activity of the synthetic material was tested to confirm the initial identification. Since then, advances in chemical techniques have allowed us to work more easily with the minute quantities of material produced by insects and other organisms, and chemical signals have been found to play important roles in the ecology of a wide range of different organisms, including species of algae, crustacea, arachnids, fish, amphibians, reptiles and mammals. Most work, however, has been concerned with insect-related compounds, primarily because of the importance of insects as pests and the potential of semiochemicals to provide novel means of control.

The sizes and structures of semiochemicals reflect their function. Those that are produced by terrestrial organisms and that act at a distance are of necessity small, volatile molecules: insect sex pheromones, for example, are typically hydrocarbon derivatives with between 5 and 30 carbon atoms, and with molecular masses of less than 400. Semiochemicals that act at short range or on contact are larger compounds of low volatility. Pheromones evolved by aquatic organisms tend to be of relatively high molecular mass and are water-soluble; many are proteinaceous or steroidal in nature.

Experimental studies of chemical ecology begin with the hypothesis that a particular interaction is mediated by a chemical signal. The hypothesis is tested by means of a bioassay tailored to the biology of the organism(s) concerned. This is followed by an iterative sequence of collecting, testing, fractionating and identifying the compounds involved; ultimately, the putative semiochemical should be synthesized in the laboratory and its biological activity confirmed. At each stage of the process, it is critical to test the behavioural activity of the compound(s) by means of a relevant bioassay. The design of the assay is therefore of fundamental importance to the success of a project. However, in order to appreciate the diversity of approaches that may be adopted, it is necessary to review briefly the ecological functions and diversity of the main semiochemical groups.

Semiochemicals and Their Biological Roles


Chemical signals underlie much of insect behaviour and ecology. In social insects (including termites, ants and various species of bees and wasps), for example, semiochemicals mediate many aspects of colony functioning, including kin and colony recognition, reproduction and caste determination, alarm behaviour and recruitment. In nonsocial insects, mate location is generally mediated (wholly or in part) by pheromones, while other behaviours such as aggregation, oviposition, and, in semisocial spe-cies, alarm behaviour, have been shown to be affected by pheromonal cues in an increasing number of species. It is the insect sex pheromones, however, that have been most intensively researched.

Most insect sex pheromones are produced by the female and attract the male from a distance; less volatile compounds acting at closer range may also play a role in stimulating copulation. In some insect species, males have also been found to produce sex pheromones: for example, the so-called ‘aphrodisiac’ pheromones that promote sexual receptivity in the females of some Lepidoptera.

Insect sex pheromones are usually a blend of several different compounds, all of which must be present in the correct concentration and ratio for the pheromone blend to be behaviourally active. Slight differences in the ratios of the component compounds may occur in geographically distinct populations of the same species. Some components of the pheromone blend may be present in only trace amounts and, in many cases, the stereochemistry of specific components is also critically important as stereoisomers can differ widely in their biological activity. The sex pheromone produced by females of the oriental fruit moth (Grapholita molesta), for example, consists of a mixture of 92-95% (Z)-8-dodecenyl acetate (Z8-12Ac), 5-8% of (E)-8-dodecenyl acetate (E8-12Ac), and variable amounts of (Z)-8-dodecen-1-ol (Z8-12:OH), ranging from trace amounts to ~ 3% of the total mixture.

Most sex pheromones are synthesized de novo within the insect; in the few cases in which a dietary origin has been demonstrated for sex pheromone components, it has generally been in male-produced pheromones (e.g. the ‘aphrodisiac’ pheromones of some male butterflies and moths that utilize plant-derived pyrrolizidine alkaloids).

The large body of literature on insect pheromones is a reflection of their potential value in pest management. Amongst non-insect invertebrates, however, there is a growing body of work on the sex pheromones of the Arachnida (spiders, mites, and their allies) and also on the Crustacea; limited work has been done with other groups. Among the Arachnida, some spider sex pheromones have been shown to be airborne volatiles acting at a distance, whereas others are nonvolatile compounds deposited on the spider’s silk; in addition, adult male mites often exhibit a characteristic ‘guarding’ behaviour in response to low-volatility compounds present on the surface of female deutonymphs prior to their final moult. Mating will then occur as soon as moulting is complete. As with insects, the full male courtship sequence of arachnids is often dependent not only on the presence of the appropriate chemical cue(s) but also on visual and tactile cues provided by conspecific females. These factors should be taken into account when developing bioassays for chemical cues.



Kairomones, as compounds that are adaptively favourable only to the receiving organism, play two main roles in chemical ecology: as food-finding or food-recognition cues, and as predator-associated cues that elicit avoidance behaviour in potential prey. Relatively few studies have been conducted on the latter, but in the former case kairomonal compounds have been shown, for example, to mediate the attraction of various herbivorous insects to their host plants and the attraction of parasites and predators to their hosts and prey; in some cases, pheromones of the host or prey act as kairomones for their predators and parasites. For example, lekking males of the Mediterranean fruit fly (Ceratitis capitata) produce a pheromone that attracts not only conspecifics but also the predatory wasp Vespula germanica, so that the males suffer disproportionate mortality as a result. In a similar way, a spider that preys preferentially on the ant Iridomyrmex purpureus is attracted to the alarm pheromone of its prey.

Host or prey recognition stimuli may also act at short range or on contact, and these relatively nonvolatile compounds often form part of the host cuticle, frass, salivary or other secretions.


Synomones are allelochemicals that benefit both the emit-ting and receiving organisms. For example, healthy plants normally produce a characteristic range of volatile chemicals, the constituents of which (or their ratios) can change in distinctive ways when the plant is attacked by insect herbivores; in some cases, these altered chemical signals are exploited by predators and/or parasitoids of those herbivores, which use them as prey or host-location cues. In this way, the plants benefit from the resultant predation on the attacking herbivores, while the predators benefit from the increased foraging efficiency that plant cues can provide. In one such system, maize plants (Zea mays) attacked by larvae of the beet armyworm (Spodoptera exigua) respond by producing relatively large amounts of several volatile compounds (mainly terpenoids), which attract Micropletis croceipes, a parasitic wasp that attacks the armyworm larvae. In this case, the plant response has been shown to be induced by ‘volicitin’ (N-(17-hydroxylinolenoyl)-L-glutamine), a compound present in the oral secretions of S. exigua larvae.

Floral odours that attract pollinating insects also function as synomones: the plants benefit from pollen transfer and fertilization, while the insects benefit from the nectar and/or pollen on which they feed.


Allomones may be involved in either resource capture or defence. An example of the former, ‘aggressive’, use of allomones is provided by the Bolas spiders (Mastophora spp.) that synthesize the sex pheromone blends of their moth prey and use them as attractants to ‘lure’ their prey within reach of the specialized silken thread (bolas) that they use for prey-capture. The use of such ‘chemical mimicry’ is not confined to the animal kingdom: certain spider orchids (Ophrys spp.) that offer no food reward to potential pollinators produce mimics of the sex pheromones of their hymenopteran pollinators: pollen transfer is achieved when attracted males attempt to copulate with the flowers. In these cases, since there is no benefit to the receiving organism, the orchid odour functions as an allomone.

Defensive allomones have been found in a variety of animal taxa, including fish, amphibians and reptiles among the vertebrates, and a variety of invertebrates including centipedes, millipedes and whip scorpions as well as numerous species of insects. The compounds concerned represent a similarly wide variety of chemical groups, including aliphatic acids and aldehydes, alkaloids, peptides, phenols, quinones, terpenes, steroids and cyanogenic glycosides. As with sex pheromones, defensive allomones are often complex mixtures of chemicals, but they tend to be much less species-specific in their composition and biological activity. Indeed, their composition may vary between the sexes of the same species, or between adults andimmatures. The compounds may either be synthesized de novo or be derived from dietary sources, and in most cases they are effective against a variety of potential attackers.

Principles of Bioassay Design

The identification of chemical cues is critically dependent on good bioassay design. At each stage of the project, from extraction to synthesis, the putative chemical cue must be assessed by an appropriate bioassay – that is, an assay in which the ‘activity’ of the chemical is scored according to some readily quantifiable behavioural or physiological response of the test organism. Some idea of appropriate techniques can usually be gained from published studies with similar species, which can then be adapted to the biology and behaviour of the species under test, and to the aims of each particular project. Initial assays are likely to be either wholly laboratory-based or conducted under ‘semifield’ (field cage) conditions. Full-scale field trials may be undertaken eventually, but these are costly and are more likely to be affected by factors beyond the experimenter’s control.

An animal’s responsiveness to chemical cues can vary with both internal (physiological) factors such as age, sex, hunger level and mating status and external (environmental) factors such as temperature, light intensity, time of day and humidity. Good bioassay design should therefore be-gin with observations of the animal in the wild, so that some preliminary assessment of the importance of these factors can be made; the bioassay can then be designed accordingly. With some organisms, it is possible to maintain laboratory colonies for test purposes. While this makes it relatively easy to obtain sufficient test individuals, care must be taken to ensure that the genetic diversity of the colony is maintained over time, and that the colony does not adapt to captivity in ways that affect the behaviour under investigation. The ethical and legal aspects of maintaining animals in captivity should also be borne in mind.

In addition to trying to standardize the responding organism’s physiological state and appropriate environmental variables, the experimenter should eliminate the possibility of associative learning by using only ‘naive’ individuals (those that have never been exposed to the test stimuli prior to testing).

Other factors to be considered include the concentration of the test substance (which should be presented at realistic levels), and the way in which it is presented: in some cases, a chemical cue may only elicit a response if presented and perceived in the context of appropriate visual or tactile stimuli, or some approximation of them. For example, certain parasitoids that attack insect eggs respond best to host-associated chemical cues when they are applied to inert models of approximately the same size and shape as host eggs.

The bioassay should be designed around the hypothesized role of the chemical(s) and the way in which the test animal would normally respond to such cues. Thus, in the case of compounds thought to act at relatively long distances, the assay apparatus must allow appropriate behaviour to be expressed – for example, by the use of wind tunnels for flying insects, or flow-tank systems for aquatic organisms. In these cases, a decision must also be made regarding a suitable wind speed or water flow rate.

Some thought must be given at the outset to the way in which a positive response will be scored, and how the data are to be analysed. Is the animal to be given a choice between two or more stimuli, or are the test stimuli to be presented separately to different individuals? The statistical analyses and experimental design must be chosen accordingly.

Feeding assays (e.g. to test potential feeding stimulants or deterrents, or the physiological effect of certain compounds) pose additional problems. For example, in investigating the response of herbivores to leaf tissues, the response to putative feeding deterrents may depend on the type and concentration of feeding stimulants present; the levels of both types of chemicals may vary with the age and physiological state of the plant material, and may also change markedly if the tissue to be tested is cut from the growing plant rather than being presented as part of an intact plant. In some cases, it may be simplest to present isolated test compounds as part of a defined artificial diet (provided that one has been developed for the species under investigation). Again, it is important to appreciate that the animal’s response will be affected by its physiological state and prior experience. A decision will have to be made, for example, whether the test organisms are to be starved prior to the assay and, if so, for how long?

The ‘best’ bioassay for a particular purpose will be determined by the nature of the stimulus, its hypothesized biological role and the type of organism being tested. Although previously published studies offer useful ideas and guidelines, there is always room for innovation and creativity. Indeed, the assay chosen may have to be refined or modified as a study proceeds and as new questions arise.

Principles of Chemical Analysis

There are three principal steps towards elucidating the structure of semiochemicals – sampling, analysis and synthesis – each of which can pose a considerable challenge to the chemist. The choice of method for each step is dependent upon the chemical properties of the molecules involved. Organic molecules have a diversity of structures, differing in the number and type of atoms that they contain, their topology (e.g. chains or ring structures, or both), the functional groups within them (e.g. alkene, ether, ketone, aldehyde, ester, alcohol, amino, carboxylic acid, amide, etc.), the geometry of double bonds (E or Z), and the absolute stereochemistry of chiral centres. Semiochemicals are often so specific that only if all of these features are correctly identified will a synthetic compound elicit the expected behavioural response.

The most extensively studied compounds are insect pheromones, which, as stated earlier, are usually multicomponent mixtures present in nanogram or picogram amounts. Many of the techniques developed for insects have now been applied to the detection of pheromones produced by mammalian and aquatic organisms.

Sample Collection and Preparation

If the site of production of a pheromone can be narrowed down to a specific gland or area of an organism, it is often possible to dissect out the gland and extract the active compound(s) by immersing the tissue in a suitable solvent, such as hexane, dichloromethane or methanol. These extracts should then be bioassayed to ensure that activity remains in the sample before commencing chemical analysis. It is also sometimes possible to sample small glands directly using solventless injection techniques, in which the entire dissected gland is placed into the analytical instrument. Extracting polar compounds from substrates such as animal hair or plant fibres may require a more vigorous approach, such as hot solvent extraction.

In some species, pheromones are not stored in glands but are released into the environment directly after biosynthesis. Specialized techniques such as closed-loop stripping or dynamic headspace analysis have therefore been developed to collect pheromones (and other volatile semiochemicals) from the surrounding atmosphere. Such compounds can be trapped on adsorbent solid phases such as activated charcoal or Tenax then desorbed in the analytical instrument using either heat or a solvent.

Solid phase microextraction is a relatively new technique that is capable of adsorbing pheromones from both air and aqueous environments, and also by direct contact with surfaces on which pheromones have been coated. The great advantages of these adsorption techniques are that they are nondestructive, allow repeated sampling from the same individual organism(s) and can be used to monitor the diel rhythm of emission of airborne volatiles. Since these techniques are applicable to most airborne volatiles, they can be used, for example, to analyse floral odours, or to detect changes in plant volatile emission during herbivore attack.


Analysis can be considered a two-stage process of separation followed by detection. Separation is carried out using a chromatographic method, either gas chromatography (GC) for volatile molecules, or liquid chromatography (LC) (usually high-performance liquid chromatography (HPLC)) for nonvolatile, more polar molecules. The small quantities of pheromones that are normally available necessitate a detection method that is both extremely sensitive and universal (i.e. not structure-dependent), making mass spectrometry (MS) the method of choice. These separation and detection techniques are normally coupled together in ‘hyphenated’ techniques such as GC-MS and LC-MS.

In mass spectrometry, the molecule is ionized and the instrument measures the mass of the resulting ions. The resolution of these instruments can vary greatly, ranging from 0.1 Dalton (low resolution) through to high-resolution instruments that measure mass to six decimal places. The mass of the molecule is diagnostic for the molecular formula of the molecule, and further structural information can be gathered through fragmentation of the ions to give a characteristic mass spectrum.

GC-MS is a well-established technique in the analysis of insect pheromones, the majority of which are low-molecular weight, relatively nonpolar molecules. Gas chromatographs separate compounds on the basis of their size and polarity, with small, nonpolar molecules being eluted from the column first. For direct biological feedback on the separated compounds, gas chromatographs can be coupled to electroantennography detection (EAD) units, in which an insect antenna forms the detector. As the compounds elute from the column, they are blown over the antenna, which is connected to the gas chromatograph via electrodes. Active compounds in the mixture result in depolarization of the antenna, generating a potential difference across the electrodes that is amplified and recorded. It is also possible to record the responses of single olfactory cells on an antenna, thus assigning neurophysiological activity to individual pheromone components. It should be noted, however, that a compound that produces electrical activity in individual receptor cells will not necessarily elicit the full behavioural response in an intact organism.

Compounds that are less volatile and/or contain more polar groups (such as carbohydrate derivatives, steroids or peptides) are better suited to LC-MS techniques, though there are relatively few examples of the use of this technique. When dealing with nonvolatile plant semiochemicals, often milligram quantities can be extracted and purified using large-scale LC and HPLC methods. This allows alternative analytical techniques to be employed, such as infrared (IR) spectroscopy and nuclear magnetic resonance spectroscopy (NMR). NMR is particularly powerful as it can ‘observe’ both hydrogen and carbon nuclei within the molecule, enabling complete structural assignment. These techniques have been applied widely to the elucidation of many natural products from plants and marine organisms. Final confirmation of the correct structure of a semiochemical involves synthesizing the compound (or mixture) in the laboratory and then testing it to confirm that it elicits the same response as the natural substance in the bioassay. This final step has been greatly facilitated by the considerable advances made over the past 30 years in chemical synthesis, particularly with regard toasymmetric synthesis.

Applied Aspects of Chemical Ecology

As our knowledge of the ecological roles of chemical cueshas increased, so too has the practical application of that knowledge. Synthetic insect sex pheromones, for example, are now widely used as monitoring tools for the early detection of pest species. In this case, the pheromone blend is incorporated into a specialized trap that usually also includes an adhesive-coated surface that can be removed to check the number of insects caught in a given period of time. However, sex pheromones typically attract adult males, whereas the most damaging stage of most crop pests is usually the larva; in their simplest form, therefore, pheromone traps are used simply as ‘early warning’ devices, signalling the first influx of adults into the field. In more sophisticated systems, a threshold trap catch is used in conjunction with temperature data and computer-based pest development models to generate a predicted ‘best spray date’ for insecticide application.

Insect sex pheromones have also been used as a direct control method by disrupting pest mating behaviour. In this case, pheromone dispensers are distributed at relatively high density throughout the crop, with the aim of ‘saturating’ the atmosphere with synthetic pheromone. Against this background ‘noise’, wild males find it impossible to locate females and mating is prevented. Unmated females fail to produce fertile eggs, and the subsequent larval population is reduced as a result. The sex pheromone of the cotton pest Pectinophora gossypiella (pink bollworm) was one of the first to be developed for use in this way, and the technique has now been used successfully against a number of other lepidopteran pests.

Attempts at reducing insect populations through pheromone-based mass-trapping have been limited by the fact that sex pheromones attract only one sex (usually the males). Aggregation pheromones, which attract both sexes, are much more suitable for this approach, but are produced by relatively few insects. The best-known examples are those produced by various bark beetles, some of which are important forest pests. Mass-trapping has been used successfully for the Spruce bark beetle (Ips typographus) in Scandinavia, but such large-scale programmes are highly specialized, labour-intensive and expensive.

Both pheromones and kairomones have been used in so-called ‘attracticide’ strategies, whereby the chemical cue is used to attract and concentrate the pest in a discrete area to which pesticides are applied. House flies (Musca domestica), for example, have been controlled in poultry houses by combining ‘muscalure’ (the female sex pheromone) with insecticide-laced food baits.

There are other, more specialized uses for semiochemicals. In some bark beetles, for example, the pheromone blend changes once the beetles have reached a relatively high density, and new arrivals are repelled rather than attracted. Synthetic versions of these ‘anti-aggregation’ pheromones have been used for the short-term protection of stressed or damaged trees, which are most vulnerable to the beetles.

Alarm pheromones have been relatively little studied from an applied perspective, although early experiments with aphids demonstrated an enhanced uptake of insecticide when the aphid alarm pheromone ((E)-b-farnesene) was incorporated into the spray mixture. This was thought to be due to increased aphid movement in response to the pheromone, and hence greater exposure to the toxicant. Alarm pheromones may also provide a novel means of repelling undesirable, nonnative fish species from water-courses at risk of infestation; the Eurasian ruffe (Gymnocephalus cernuus), for example, is repelled by a substance released from the damaged skin of conspecifics, and it has been suggested that this could be used to control its outward spread from the St Louis River, into which it was accidentally introduced in the early 1980s.

Apart from the direct use of semiochemicals to modify pest behaviour, attempts have also been made to enhance the beneficial effect of natural enemies of insect pests by manipulating their responses to chemical cues. For example, both predatory lacewings (Chrysopidae) and ladybird beetles (Coccinellidae) have been successfully attracted by synthetic mixtures containing compounds found in aphid honeydew. Attempts have also been made to retain parasitic wasps in crops by treating plants with point sources of host-associated kairomones. However, there is a risk that this technique could result in reduced levels of parasitism if the wasps spend too much time examining these unproductive ‘false hosts’

A more promising method of manipulating the behaviour of parasitic wasps depends on their ability to learn specific chemical cues associated with the host or host habitat; indeed, they can be trained to respond to completely ‘unnatural’ odours (e.g. perfume blends), provided they have experienced them in association with their host. This remarkable learning ability can be used to ‘prime’ mass-reared wasps to respond preferentially to specific host/plant combinations, and is even being exploited in a novel US project aimed at using parasitic wasps as ‘bio-detectors’ for nerve gases and other battlefield toxicants.

Practical use is also being made of some mammalian pheromones, e.g. the male boar pheromone, which is used to increase the receptivity of females during artificial insemination.

Plant secondary chemicals have also been exploited in the control of insect pests. For example, seeds of the Neem tree (Azadirachta indica) contain azadirachtin, an insect antifeedant, and extracts of the seed have been used against a variety of insect pests.

In addition, the success of the various ‘polyculture’ techniques practised in traditional agricultural systems is based in part on the effect of plant volatiles on host-finding by insect herbivores. Specialist herbivores are repelled by nonhost plants but attracted by their host plants and related species. This effect has been exploited in the so-called ‘stimulo-deterrent’ or ‘push-pull’ strategy, whereby insects are both repelled from the crop by under-planting with deterrent plants and attracted away from the crop by more attractive plants grown around the perimeter.


Over the past 45 years, chemical ecology has developed into a mature science that is finding diverse applications. A deeper knowledge of these ‘signalling chemicals’ will lead us to a better understanding of our environment and more sustainable methods of pest control. While modern analytical instrumentation and new molecular biology techniques will both have their impact in the future, careful observational biology and good bioassay design will remain at the heart of the subject.