Parasitoid Host Selection: Host and Host Food-Plant Cues

Walter Roachell

Colorado State University

Fort Collins, Colorado 80523

wroach@holly.colostate.edu

Successful parasitism by parasitoids of herbivorous insect hosts is preceded by 

several phases of host searching that lead females into the close vicinity of their 

potential host(13).а In each of these phases, the female often utilizes chemical 

stimuli to guide her in the search forа a suitable host.а Chemical stimuli emitted 

by members of the different trophic levels in the parasitoid's environment, such as 

the herbivorous host or the host's foodplant, are often used in the host-searching 

process(5).а Host derived stimuli are the most reliable indicators of the presence 

of hosts, but they may be hard to detect at long distances, whereas plant-derived 

stimuli are assumed to be more detectable(1).а 

The role of stimuli from different trophic levels in various phases of the host-

searching process, is hypothesized to be determined by the parasitoid's degree of 

specialization(1).а Specialists have been shown to exhibit behavioral responses to 

host-de rived cues, such as herbivore products and the herbivore itself, or to 

specific herbivore induced synomones.а For generalist, such fixed responses to 

specific stimuli do not seem to be functional or may be impossible due to more 

physiological constraints co mpared to specialist.а It has been hypothesized that 

the generalist is guided to their host by more general stimuli at first and 

subsequently learn to respond to more specific stimuli(11).а 

Due to the impressive variety of specialized lifestyles found in insect parasitoids, 

generalizations about their foraging behavior may be considered difficult to make.а 

Certain characteristics of the host's ecology can be important in the evolution ofа 

specific foraging mechanisms in parasitoids and in determining the usefulness of 

different sensory modalities in this foraging(12).а Host-derived cues provide the 

parasitoid with very specific information as to the identity of the prospective 

host, but usua lly these cues are in such low levels that they are hard to detect.а 

On the other hand, cues from the host's food plant have a much higher degree of 

detectability, but generally have a low reliability.а The purpose of this review 

article is to examine the foraging mechanisms of parasitoids, and determine how the 

reliability-detectability problem is solved.а <:s>

INTRODUCTION

Many natural enemies are known to discriminate between volatile chemicals emitted by 

uninfested and herbivore-infested plants.а Chemical stimuli emanating from the 

plant-host complex can originate from the herbivore, the plant, or from interactions 

bet ween the herbivore and the plant.а In spite of the lifestyle diversities, insect 

parasitoids have certain aspects of foraging in common.а Long distance location by 

parasitoids seems to be essentially guided by information from the food of the host, 

whileа at shorter distances information from the host itself becomes increasingly 

important.а Parasitoids encounter a great variety of stimuli that they may use in 

host location.а The appropriateness and usability of this information depends on two 

factors:а 1) its reliability in indicating host presence, accessibility and 

suitability and 2) the degree to which stimuli can be detected(12).

INFORMATION FROM THE HOST

Stimuli derived directly from the host are generally the most reliable source of 

information because they can inform the parasitoid of the presence, identity, 

availability and suitability of the host.а Although this type of stimuli has a high 

reliability, it is often limited by low detectability.а Low detectability of host 

derived information has two inherent constraints that limit its use as stimuli for 

host location.а In terms of mass, hosts are small components of a complex 

environment and if they produce any information at all, it will be small in amount.а 

Secondly, there should be constant selection on the hosts for inconspicuousness as a 

way to avoid parasitization and predation(12).а 

As stated above, one of the constraints of host derived cues is the relative small 

amounts of volatiles produced by the host.а Host derived cues are difficult to 

detect at long distances and thus become more important the closer the parasitoid 

gets to the host.а There are however several types of host cues that are used by 

some parasitiods to locate hosts from long distances.а While communicating 

interaspecifically through pheromones, herbivores are often much more conspicuous to 

their natural enemies th at exploit these pheromones as kairomones in long distance 

herbivore location(14).а A well known example is that of predators and parasitoids 

of bark beetles that respond to volatile aggregation pheromones of their victims.а 

Similarly, parasitoids exploit the sex pheromones of their hosts as kairomones(2).ааа 

Various pieces of direct and indirect evidence indicate that within a habitatа 

Trichogramma spp. are able to distinguish between host-infested and uninfested 

areas.а In this phase, foraging behavior seems to be mediated by odors originating 

from adult female host insects(11).а Studies have been done that deal with the means 

byа which Trichogramma detect host communities within a habitat.а This research has 

provided evidence that suggests that the sex pheromone of the host may play a 

significant role in this respect.а 

The eggs of the cotton bullworm, Heliothis zea, were found to be heavily parasitized 

by naturally occurringа Trichogramma spp. in cotton plots treated with synthetic sex 

pheromone of H. zea than in control plots(2).а These findings aroused the interest 

of several researchers as to the behavioral mechanism underlying the observed 

effect.а Increased parasitization in treated plots could be due to attraction of 

parasitoids from a d istance.а Noldus & van Lenteren studied this problem in a 

comparable host-parasitoid system:а the cabbage armyworm,а Mamestra brassicae and T. 

evanescens.ааа 

The experiment consisted of a bioassay using calling virgin moths as the order 

source.а The basic set-up consisted of a four-armed airflow olfactometer.а Three of 

the virgin moths were placed in a glass cylinder, which was then connected to one 

arm ofа the olfactometer.а A second glass cylinder containing 3-5 male moths was 

attached to the exhaust air stream, and their behavior was used as a check for the 

actual presence of sex pheromone in the air stream.а A parasitoid was released into 

the olfactometer chamber only after calling activity of at least two of the male 

moths had been observed. The parasitoid was observed for 10 minutes(4).а 

During the 10 minute observations, the parasitoids made a significantly higher 

average number of visits to the flow field containing the odor of the moths than to 

the opposite control field.а The insects stayed significantly longer in the flow 

field co ntaining the odor of the calling moths compared to the control fields(4).а 

These results demonstrate that T. pretiosum can show a behavioral response to the 

odor emitted by callingа H. zea moths.а The fact that a reaction of the wasps was 

elicited only by female moths during calling activity and concurrent with male moth 

responses, strongly suggests that the parasitoids were indeed responding to the sex 

pheromone of the moth.а The reaction of T. pretiosum to its host's sex pheromone may 

be illustrative of many species of egg parasitoids(4).ааа 

The research seems to confirm that the sex pheromone of H. Zea serves as a kairomone 

for T. pretiosum.а Noldus et al surmises that it indeed seems adaptive for an egg 

parasitoid to use host sex pheromone as an indicator for the probable presence of 

host eggs, because as far as known no other long distance c ues more directly 

connected to the eggs themselves are available(4).аа 

Apart from the use of host pheromones, herbivore odors are difficult to detect at 

long distances.а As expected, herbivore derived cues become more important the 

closer the foraging parasitoid comes to their host.

INFORMATION FROM THE FOOD OF THE HOSTаа 

Stimuli from the food of the host or prey are usually more readily available, 

because of the food's relatively large biomass, but are less reliable predictors of 

host or prey presence and suitability. The amount of herbivory on the plant strong 

ly affects the usefulness of the volatile cues from the plant.а If the infestation 

is very high, information from the plant is very reliable, and can even replace the 

more difficult-to-detect host information(12).а 

Plants show some astonishing adaptations that promote the presence of predators and 

parasitoids.а A good example of this is the relationship between acacias and ants.а 

Specialized structures on the acacia provide the ants with food and shelter.а The an 

ts seem offer the plants protection against herbivory.а In light of the astonishing 

adaptations plants exhibit, it should not be surprising that some have proposed that 

plants may actively guide the predators and parasitoids to the herbivore(6).а Recent 

stu dies have revealed that plants that are under attack by herbivores initiate the 

release of chemical "signals" that parasitoids use to locate their victims.а 

Herbivore-induced plant chemical responses are usually considered direct defenses 

against the herbivorous attackers.а Herbivore-injured plants release relatively 

large amounts of volatiles that are attractive to parasitoids.а Since the plant and 

the pa rasitoid both benefit from this interaction it has been suggested that the 

plant responses serve to recruit the parasitoids.а Turlings and Tumlinson argue that 

induced production of plant chemicals evolved first as a direct defense against 

herbivores, and that the attractive function evolved secondarily(8).а Three aspects 

ofа plant signal seem most relevant to the ongoing discussion on whether or not 

herbivore damaged plants actively lure natural enemies.а (a) The signal should be 

clear enough to the insectsа so that it can be perceived and distinguished from 

background noise.а (b) The signal has to be specific enough to reliably indicate the 

presence of a suitable host or prey.а (c) The signal will have to be emitted during 

the period of time that the natural enemies forage(6).

SIGNAL CLARITY

Turlings and Tumlinson studied the host searching behavior of the parasitoidа 

Cotesia marginiventris, and discovered that plants emit a strong odor when under 

attack by caterpillars.а Rather than responding to odor cues coming directly from 

their hosts, females ofа C. marginiventris are strongly attracted to volatiles 

emitted by the caterpillar damaged plants(9).а Research focusing on the 

identification of the emitted chemicals revealed a clear difference between corn 

seedlings that were under recent herbivoreа attack and seedling that had been fed on 

for more that six hoursа 

Fresh feeding damage results in a significant release of (Z)-3-hexenal, (E)-2-

hexenal, (Z)-2-hexenol, (E)-2-hexenol and (Z)-3-hexenyl acetate.а These volatiles, 

also known as "green leaf volatiles" are the only compounds detected at this stage.а 

Afterа several hours, however, emission of large amounts of terpenoids were 

observed. (figure 1)ааа 

It was found that this emission of specific chemicals could not be induced by 

mechanical damage alone.а Plants that had been fed upon by caterpillars for 2 hours 

released far more of these terpenoids on the following day than plants that had been 

dama ged with razor blades during the same 2 hours.а When caterpillar regurgitate 

was applied to sites that had been freshly damaged with a razor blade, the plants 

released terpenoids in amounts equal to those released by caterpillar damaged 

plants(6).ааа 

Signals from herbivore damaged plants are further enhanced by the fact that the 

chemical emissions are not limited to the damaged sites.а For corn seedlings, all of 

the induced compounds are released throughout injured plants.а Unharmed leaves of 

dama ged plants show an increase in the release of terpenoids(8).а The chemicals 

emitted by the plants upon herbivory should be easily detected by insects and thus 

could serve as very clear signals to parasitoids and predators.а The amount of 

volatiles producedа per hour by the plant is much greater than the amounts normally 

seen in insect pheromone communication.а Only a few nanograms of sex pheromone per 

hour can be detected a compared to several micrograms of a particular substance 

emitted by one corn seedling(8).а 

Turlings and his group conclude that the chemicals emitted by the plants in response 

to herbivory are easy to detect by insects and are clearly distinguishable from the 

extremely low levels of odors emitted by unharmed plants.а Moreover, the blend of 

induced terpenoids emitted by herbivore damaged plants is quite different from the 

odors emitted by unharmed or mechanically damaged plants.а Unlike the "green leaf 

volatiles" and constitutive terpenoids that may be released by mechanical damage, 

the inducedа substances represent a signal that is dependably associated with the 

presence of herbivores.а The signal is enhanced by the systematic nature of the 

plant response and makes the whole plant stand out as an odorous beacon that should 

be easily distinguishable from the surrounding plants.а In short, herbivore injured 

plants provide foraging natural enemies with a very clear odorous signal that 

indicates the presence of potential hosts or prey.(6).

SPECIFICITY OF SIGNAL

For the signals to be useful to the parasitoid, they should indicate the presence of 

a suitable host and be distinguishable from odors that are associated with non 

hosts.а Some of the chemicals disseminated from the infested plants are merely due 

to mechanical disruption of plant cells, and thus are not specific for herbivore 

damage(12).а Figure 2 illustrates the enormous variety in volatile blends that 

several different plant species release when damaged by herbivores.а In this case 

all of the depicted blends are the result of feeding damage inflicted by the beet 

armyworm Spodoptera exigua(7).

Until recently, evidence for herbivore specific cues has been limited.а Turlings et 

al. can now substantiate the variability of plant and host cues with their work onа 

Cotesia marginiventris.а C. marginiventris is a generalist that attacks the larvae 

of many Lepidoptera.а Research done by Turlings et al. on the foraging behavior ofа 

C. marginiventris has shown some ability to learn to distinguish between odors from 

different herbivores feeding on the same variety of plant. When an herbivore feeds 

on different structures of the same plant the odors emitted vary.а The increase inа 

response to host related odors after experience is greatest to the odors emitted by 

the plant-host complex that the female experienced.а The learning process that must 

be involved is triggered by a brief contact with host by-products and does not 

require a ctual contact with the host.а It has been shown that when females are 

given a longer experience period, including ovipositions, they did not appear to 

respond differently than females that had a much shorter experience without 

ovipositions.а The effect of the experience on the preference for host related odors 

was shown to last at least for several hours, and is likely to be an important 

factor determining host-searching behavior of C. marginiventris in the field(10).

TIMING OF SIGNAL

It seems that signals would be most effective if emitted as soon as, or at least 

shortly after a herbivore starts damaging a plant.а Moreover, the volatiles would be 

most effective if they are emitted during the time of day when natural enemies are 

most likely to forage(6).а Work has been done to confirm whether or not the timing 

of volatile release is in tune with the parasitoids needs.а 

Several experiments were designed to monitor odor emissions by corn plants over a 

long period of time.а A method was used that allowed non-destructive collection of 

volatiles from growing plants.а Volatiles were sampled at periodic intervals 

throughout one or more photoperiods.а Volatiles were monitored at 2 hour intervals 

for up to 3 days.а Figure 3A represents the emission of several compounds that are 

representative of the total blend emitted by the corn seedlings.а Figure A 

graphically shows that the LOX products are released almost instantaneously in 

response to damage, but rapidly decrease after damage ceases.а Figure 3B represents 

the emission of terpenoids after the corn seedling was damaged and treated with 

regurgitate.

The result was a delayed but dramatic increase in terpenoid emissions during daytime 

hours that was still detected on the third day after initial damage(6).а 

Although there is some delay, the plants seem to respond quickly enough for 

parasitoids and predators to effectively exploit the volatile signals.а The plants 

give off the strongest signals during the photoperiod when natural enemies tend to 

forage.а I nstead of the plants timing their responses to the active periods of 

parasitoids, it is more likely that one of the reasons natural enemies forage mostly 

during the day is because there are more visual and chemical cues available to them.а 

It is possible that the plants diurnal metabolism allows for volatile emissions only 

during the day.а On the other hand, the biology of night blooming plants suggests 

that there is no constraint on when volatiles could be emitted(6).

CONCLUSION

As herbivores have adapted to be inconspicuous to their natural enemies, it can be 

expected that they would emit very little in terms of odors that parasitoids could 

use to locate them.а Plants on the other hand benefit from the attraction of 

parasitoi ds, and may have developed means such as chemical signals to reveal the 

presence of herbivores.а It is unlikely that herbivore-damaged plants initiate the 

production of chemicals solely to attract parasitoids and predators.а The signaling 

role probably evolved secondarily from plant responses that produce toxins and 

deterrents against herbivores and antibiotics against pathogens(6).

It is clear that parasitoids most often utilize the cues from the host plant when 

foraging for potential hosts.а Insects overcome the constraints of the unreliable 

nature of the host plant cues by associative learning, and the ability to 

distinguish between herbivore damage plants and mechanically damaged plants.а The 

parasitoids are able to distinguish between "green leaf volatiles" and the blend of 

induced terpenoids emitted by herbivore damage.

The nature of the signals is obviously under the control of the plants.а Turlings et 

al. suspects that if the plant-produced chemicals are targeting the herbivores, then 

possible variations in chemicals that the plants produce may mainly result from 

different adversaries.а Turlings proposes that perhaps the plants do not need to 

release highly specific signals that differ when different herbivores attack it.а 

The plants so not necessarily suffer when natural enemies make "mistakes" and are 

attracted to the signals that are induced by herbivores they cannot attack.а As long 

as the "right" natural enemies are attracted as well, the signal has served its 

purpose for the plant.а It seems therefore up to the parasitoid to deal with some of 

the unreliable aspects of the signals(6).а In order to supplement the unreliable 

nature of the plant volatiles, the parasitoids often use other signals such as close 

range chemical contact, visual and vibrational cues in conjunction with herbivore 

induced volatiles from the host plant.ааааааааааааааааааааааааааааааааааааааааааа 

REFERENCES

1. Jacqueline BF, Greevliet, Vet LEM, Dicke M (1994)а Volatiles from damaged plants 

as major cues in long range host searching by the specialist parasitoidа Cotesia 

rubecula.а Ento Exp Appl 73:289-297.

2. Lewis WJ, Nordlund RC, Gueldner PEA, Tumlinson JH (1982)а Kairomones and their 

use for management of entomophagous insects.а J Chem Ecol 8:1323-1331.а 

3. Misell RF, Frazier JL, Nebeker TE (1984)а Response of the clerid predatorа 

Thanasimus dubius to bark beetle pheromones and tree volatiles in a wind tunnel.а J 

Chem Ecol 10:177-187.

4. Noldus LPJJ (1988)а Response of the egg parasitoid Trichogramma petiosum to the 

sex pheromone of its hostа Heliothis zea.а Ento Exp Appl 48:293-300.

5. Price PW, Bouton CE, Gross P, McPheron BA, Thompson JN, Weis AE (1980)а 

Interactionsа among three trophic levels:а influence of plants on interactions 

between insect herbivores and natural enemies.а Ann Rev Ecol Systа 11:41-65.

6. Turlings TCJ, Loughrin JH, McCall PJ, Ursula SR, Lewis WJ, Tumlinson JH (1995)а 

Howаа caterpillar-damaged plants protect themselves by attracting parasitic wasps.а 

Proc Natl Acad Sci USA 92:4169-4174.

7. TurlingsTCJ, Wackers FL, Vet LEM, Lewis WJ, Tumlinson JH (1993)а in Insect 

learning: Ecological and Evolutionary Perspectives, eds. Lewis AC, Papaj DR.а 

Chapman & Hall,аа New york.а pp. 51-78

8. Tulings TCJ, Tumlinson JH (1991)а Do Parasitoids use herbivore-induced plant 

chemical defenses to locate hosts?а Fl Ento 74:42-49.

9. Turlings TCJ, Tumlinson JH (1990)а Exploitation of herbivore-induced plant odors 

by hostаа seeking parasitic wasps.а Science 250:1251-1253.

10. Turlings TCJ, Scheepmaker JWA, Vet LEM, Tumlinson JH, Lewis WJ (1990)а How 

contact foraging experiences affect preferences for host-related odors in the larval 

parasitoidаааа Cotesia marginiventris.а J Chem Ecol 16:1577-1589.

11. Vet LEM, Dicke M (1992)а Ecology of infochemical use by natural enemies in a 

tritrophic context.а Annu Rev Ent 37:141-172.

12.а Vet LEM, Dicke M , Wackers FL (1991)а How to hunt for hiding hosts:а the 

reliability detectability problem in foraging parasitoids.а Neth J Zoo 37:141-172.

13.а Vinson SB (1975)а Biochemical coevolution between parasitoids and their hosts.аа 

Evolutionary Strategies of Parasitic Insects and mites.а Price PW, ed.а pp. 14-48.а 

Plenumа Press, New York.

14.а Weseloh RM (1981)а Host location by parasitoids.а Semiochemicals:а their role 

in pest management.а Nordlund DA, Jones RL, Lewis WJ, ed. pp.а 79-95.а John Wiley & 

Sons, Inc. New York.