Behavioral Ecology Vol. 11 No. 6: 606-613
© 2000 International Society for Behavioral Ecology
Attraction of a predator to chemical information related to nonprey: when can it be adaptive?
Laboratory of Entomology, Wageningen University, P.O Box 8031, NL-6700 EH, Wageningen, The Netherlands
Address correspondence to M. Dicke. E-mail: Marcel.Dicke{at}users.ento.wag-ur.nl .
Received 17 September 1999; revised 27 January 2000; accepted 23 February 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Information specificity can be important to animals in making optimal decisions. However, it is not always necessary to use every level of specificity. We analyzed the response of the predatory mite Phytoseiulus persimilis to plant-produced information related to a nonprey herbivore. This predator is a specialist feeding on spider mites in the genus Tetranychus. Caterpillars of Spodoptera exigua cannot serve as prey. Plants respond to an infestation by herbivores with the emission of volatiles that attract carnivorous enemies of the herbivores. Conspecific plants infested with different herbivore species can emit blends that are qualitatively identical, while differing in the ratios of blend components. However, different plant species emit volatile blends that differ qualitatively. We demonstrated that the predator P. persimilis is attracted to volatiles from bean plants infested with S. exigua caterpillars, but that this attraction is affected by predator starvation and host-plant experience. One-hour and 24-h starved predators were made to represent predators that just lost a prey patch versus predators that have totally lost a prey patch. Predators reared on spider mites on bean were attracted to bean plants infested with caterpillars when starved for 1 h but not when starved for 24 h. Both predator groups were attracted to bean plants infested with prey (i.e., spider mites). One-hour starved predators can use the odor to relocate the rewarding prey patch they just lost contact with, and using a general olfactory representation of the blend is sufficient for relocation. In contrast, for 24-h starved predators, the perception of a plant's odor blend is unlikely to represent the prey patch lost, and discriminating between an odor blend representing prey or nonprey will avoid investing time in finding a nonprey herbivore. In contrast, predators that had been reared on spider mites on cucumber and thus had experienced a qualitatively different odor blend were not attracted to volatiles from caterpillar-infested bean plants. They were attracted to spider mite-infested bean plants, irrespective of starvation level. To cucumber-experienced predators, the perception of bean plant odor cannot represent the prey patch lost, but only a new prey patch. Being discriminative and only responding to prey-infested plants is adaptive in this situation. Our results are discussed in the context of optimal information processing.
Key words: Acari, experience, herbivore-induced plant volatiles, learning, Lepidoptera, phenotypic variation, Phytoseiulus persimilis, prey selection, Spodoptera exigua, starvation, Tetranychus urticae.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The responses of animals to changing environmental conditions are a central issue in ecology. These responses include morphological and behavioral responses, and they may be genetically fixed and/or phenotypically plastic. The degree of phenotypic plasticity can have considerable consequences for population dynamics and evolution of animals (Doebeli, 1995
;
Papaj, 1994;
Rollo, 1994). Prey selection
is an important factor affecting reproductive success of predators. Foraging
decisions during prey selection affect not only the rate of prey encounter,
but also the quality of the prey encountered, and thus foraging decisions
affect reproductive output. Therefore, these decisions are under strong
natural selection. This is especially true for small-sized predators that have
little reserves to survive periods of prey scarcity and for whom predation
rate and food quality are strongly correlated with rate of reproduction
(Sabelis, 1992).
Optimal foraging theory assumes that animals are omniscient and thus make
adaptive decisions (Stephens and Krebs,
1986). This assumption has often been criticized as being
unrealistic, but evidence in its favor is rapidly increasing (e.g.,
Janssen et al., 1997;
Papaj and Lewis, 1993; Vet,
1996,
1999). Rewarding and
unrewarding experiences influence the informational state of animals and
result in phenotypic variation in foraging decisions
(Papaj et al., 1994;
Vet et al., 1998). For
instance, previous experiences can inform animals about the availability of
resources and variation therein (Geervliet
et al., 1998a; Jaenike and
Papaj, 1992; Papaj and Lewis,
1993; Vet, 1996,
1999) or about the level of
competition to be expected (Visser et al.,
1990).
Herbivore-induced plant volatiles attract carnivorous arthropods
Chemical stimuli are an important source of information for arthropods and
may convey information about the presence and quality of resources,
competitors, and enemies (Carde and Bell,
1995; Roitberg and Isman,
1992). Predatory arthropods that forage for herbivores exploit
volatiles related to herbivores. These volatiles do not originate from their
prey itself, but from the prey's food plant, which produces the cues in
response to herbivore damage (Dicke,
1999; Dicke and Sabelis,
1992; Turlings et al.,
1993). This information emitted by plants can be specific for the
herbivore that damages the plant, and thus the signal may be a reliable
indicator of herbivore presence and identity
(Dicke and Vet, 1999;
Vet and Dicke, 1992). Odor
blends emitted by conspecific plants that are infested by different herbivore
species show qualitatively similar blends that differ in the contribution of
individual components (DeMoraes et al.,
1998; Dicke and Vet,
1999; Takabayashi et al.,
1991; Turlings et al.,
1993). Although these differences seem to be small, carnivorous
arthropods can (learn to) discriminate between them
(Du et al., 1998;
Geervliet et al., 1998b;
Powell et al., 1998;
Turlings et al., 1993;
Vet et al., 1995). Odor blends
emitted by heterospecific plants that are infested by the same herbivore
species differ much more, both qualitatively and quantitatively, and
carnivores can discriminate between these blends
(Dicke et al., 1990b;
Geervliet et al., 1998a;
Takabayashi and Dicke, 1996;
Turlings et al., 1993).
Because the chemical information provided by plants is variable, phenotypic
plasticity in the behavioral responses of carnivorous arthropods is expected
and has been amply demonstrated (Dicke and
Vet, 1999; Dicke et al.,
1998; Geervliet et al.,
1998a,b;
Turlings et al., 1993;
Vet et al., 1995).
The best studied form of phenotypic plasticity in these behavioral
responses is associative learning, where foraging success is the unconditioned
stimulus and the plant volatiles constitute the conditioned stimulus
(Turlings et al., 1993;
Vet et al., 1995). A recent
study showed that associative learning in a parasitic wasp may result in the
discrimination of odor blends that are only marginally different
(Vet et al., 1998). Moreover,
associative learning may also result in generalization (sensu
Smith, 1993), where stimuli
that resemble the experienced conditioned stimulus elicit a response when the
conditioned stimulus is not offered (Vet
et al., 1998). By retaining information from both successful and
unsuccessful foraging experiences, carnivorous arthropods are able to make
precise foraging decisions adapted to changing environmental conditions
(Papaj et al., 1994;
Vet et al., 1998).
Most research on phenotypic plasticity in the behavioral response of
carnivorous arthropods to plant odors has been done on parasitoids
(Turlings et al., 1993;
Vet et al., 1995). After
locating and parasitizing a host, these parasitic wasps continue foraging for
new hosts or new host patches. The chemical stimuli perceived during
successful oviposition affect subsequent foraging decisions. An oviposition in
a single host or a 20-s exposure to host products can be enough to affect
foraging decisions (Turlings et al.,
1993; Vet et al.,
1995).
Processing variable plant information by arthropod predators
Foraging behavior of arthropod predators is different from that of
parasitoids in several respects. Predation of a prey by a predator usually
takes much more time than the parasitization of a host by a parasitoid.
Predators usually remain longer in a prey patch, and this is especially true
for predators that prey on species that form aggregations. Thus, predatory
arthropods are faced with different temporal dynamics during foraging and in
the number of foraging decisions made per unit time. This is reflected in the
duration of an experience that is needed to change foraging decisions.
Foraging decisions of the predatory mite Phytoseiulus persimilis are
not affected by experiences that last for fewer than 3 days, but an exposure
of 6-7 days to herbivore-induced plant odors in the presence of prey leads to
significant effects (Dicke et al.,
1990a; Krips et al.,
1999).
In addition to being attracted to herbivore-induced plant volatiles,
predators that stay in a prey patch for a long time may also be arrested by
herbivore-induced plant volatiles. For example, the predatory mite P.
persimilis responds to steep gradients of the herbivore-induced plant
odors at the edge of a prey patch by returning into the patch, where it will
continue to feed on the prey that induce the plant odors
(Sabelis et al., 1984). When
predators use the odors to stay in the prey patch of which the quality is
known, much less detailed information on odor composition is needed than for
locating a new prey patch.
Information processing hypothesis and tests
It is of great importance to animals to limit the input of irrelevant
information, and this can be done by using the general characteristics of an
odor blend, rather than its specific composition
(Bernays, 1996;
Bernays and Wcislo, 1994;
Milinski, 1990). We therefore
hypothesized that after prolonged exposure to the induced volatiles in a prey
patch, predators will respond to a general "image" of the complex
mixture rather than to the specific mixture with all its details. If this is
true, we expect that after prolonged exposure to the odor blend by being in a
prey patch, predators will respond similarly to odor blends that differ in
minor respects. Upon termination of the exposure, we expect the predators to
lose the generalized response with time. Thus, we hypothesized that predators
that have just left the prey patch (and thus the odor source) will be
attracted to odor blends that differ slightly from the one to which they have
been exposed, while predators that have left the prey patch for a prolonged
period of time will either not be attracted or will be attracted to a lesser
degree.
We tested these hypotheses with the predatory mite P. persimilis.
This specialist predator is well known to respond to plant odors induced by
their prey, the two-spotted spider mite, Tetranychus urticae (Dicke
et al.,
1990b,c;
Sabelis and Dicke, 1985). To
our surprise, the predators were also strongly attracted to volatiles emitted
by lima bean plants that were infested with caterpillars of Spodoptera
exigua, which cannot serve as food for the mites
(Shimoda and Dicke, 1999). The
two-spotted spider mite and the armyworm S. exigua both are extremely
polyphagous herbivores that share many host plant species
(Brown and Dewhurst, 1975
Hill, 1983;
Van de Vrie et al., 1972),
have a similar global distribution (CAB,
1972; Van de Vrie et al.,
1972), and can occur on the same plant species (e.g.,
Karban, 1986;
Stout et al., 1998).
The experience of the predators in the study by Shimoda and Dicke
(1999) (i.e., satiated
individuals that were reared on T. urticae on lima bean) might
explain the unexpected result of attraction of the predators to volatiles
induced by nonprey. These predators may have generalized the odor blend of
lima bean leaves infested with prey mites and also respond to the similar
blend of lima bean leaves infested with nonprey caterpillars. We investigated
this possibility by comparing the response of P. persimilis to
volatiles from lima bean leaves that are infested with either T.
urticae or S. exigua. We studied the response of predators that
had been reared on plants infested with T. urticae and that had been
removed from the food source, which is also the odor source, for a short
duration (1 h) or for a long duration (24 h). Predators that had been removed
from the food source for only 1 h represent predators that accidentally
wandered out of the prey patch and are still in its neighborhood
(Sabelis and Dicke, 1985). In
contrast, predators that had been removed from the food source for 24 h
represent predators that had lost the prey patch (e.g., because of long-range
dispersal) and that are unlikely to encounter the same prey patch again
(Sabelis and Dicke, 1985).
To distinguish between the effects of a loss of experience and the effects
of starvation, we investigated predators with different rearing histories
(i.e., predators reared on lima bean plants or cucumber plants infested with
two-spotted spider mites). The odor blends emitted by lima bean and cucumber
plants infested with T. urticae differ in chemical composition, both
qualitatively and quantitatively (Dicke et al.,
1990b,c;
Takabayashi et al.,
1994a).
In this study we investigated the behavioral response to odors from lima bean leaves infested by either prey mites or nonprey caterpillars. According to the hypothesis, predators that have been reared on lima bean and have left the prey patch during a short time (1 h) respond to a general "image" of the odor. Therefore, they are expected to have a more similar response to lima bean plants infested with either T. urticae or S. exigua than predators that have left the prey patch for a longer period of time (24 h). Predators that have been reared on spider mites on cucumber have had no experience with odor of infested lima bean plants. Therefore, they have no possibility to generalize the odor induced by spider mites on lima bean leaves. Consequently, we expected that among predators with cucumber experience, those that have left the prey patch for 1 or 24 h would have similar responses to odors from lima bean plants infested by spider mites or caterpillars; both are expected to respond to spider mite-infested bean plants and neither to caterpillar-infested bean plants.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Organisms
Lima bean plants (Phaseolus lunatus cv. Sieva) or cucumber plants (Cucumis sativus, cv. Lange groene giganten) were grown in sterilized soil in a greenhouse (20-30°C, 50-70% relative humidity, and a photoperiod of at least 16 h). Two-spotted spider mites (Tetranychus urticae Koch; Acarina: Tetranychidae) were mass-reared on lima bean and cucumber plants. Two- to three-week-old lima bean plants or 4- to 6-week-old cucumber plants were infested with spider mites and grown in a greenhouse under similar conditions as those used for normally rearing the plants. Beet armyworms, Spodoptera exigua (Hbn.) (Lepidoptera: Noctuidae) were reared on a semisynthetic diet (Smits, 1987
) in a climate cabinet (26 ± 1°C, 70-80% relative
humidity, and 16 h:8 h light:dark). Eggs were obtained from the culture and
introduced onto a lima bean leaf in a Petri dish in a climate room (23
± 1°C, 50-70% relative humidity, and 16 h:8 h light:dark) for about
7 days to obtain larvae with suitable ages (most eggs had hatched after 4
days). The predatory mites (Phytoseiulus persimilis Athias-Henriot; Acarina: Phytoseiidae) were mass-reared on lima bean leaves infested with two-spotted spider mites in a greenhouse (20-30°C, 50-70% relative humidity, and photoperiod of at least 16 h). Standardized experimental predators were obtained as follows. We obtained predator eggs by placing five fertilized females from the mass-rearing on an infested lima bean leaf or a cucumber leaf in each of a number of Petri dishes (9.2 cm diam) in a climate room (23 ± 1°C, 50-70% relative humidity, and 16 h:8 h light:dark). After 2 days, the adult females were removed, and their offspring were reared in the Petri dish on the spider mites on a leaf of the same plant species for 7 or 8 days. Adult females (8-10 days old; i.e., 2-4 days since the final moult) of the two populations (i.e., reared on spider mites on bean or on cucumber leaves) were used in this study.
Before each experiment, 10-20 predators of either population were introduced into a Petri dish containing a small piece of moist filter paper for 60 ± 30 min or for 24 ± 4 h in a climate room (23 ± 1°C, 50-70% relative humidity, and 16 h:8 h light:dark) to obtain predators with different levels of starvation.
Thus, we used four experimental predator groups in the experiments: (1) reared on lima bean leaves and starved for 1 h, (2) reared on lima bean leaves and starved for 24 h, (3) reared on cucumber leaves and starved for 1 h, and (4) reared on cucumber leaves and starved for 24 h.
Odor sources
We presented three different types of odor source to predators from each of
the four experimental groups. Here, we only used lima bean plants that were
12-18 days old. The first odor source was spider mite-infested plants. Freshly
cut, uninfested lima bean plants with two primary leaves each were put
individually with the stem in a glass vial (1.8 cm in diam and 7 cm in height)
with water. Twenty adult spider mite females were introduced on each leaf. The
infested plants were incubated in a climate room (23 ± 1°C, 50-70%
relative humidity, and 16 h:8 h light:dark) for 3 days. Just before the
experiments, the cut edge of the stem of each plant was wrapped in moist
cotton wool. We used five infested plants (200 spider mite females on 10
leaves in total) with moist cotton wool as odor source.
The second source was caterpillar-infested plants. Freshly cut, uninfested lima bean plants with two primary leaves each were put individually with the stem in a glass vial (1.8 cm in diam and 7 cm in height) with water. Two first-instar larvae of S. exigua within 24 h after hatching were confined to each leaf with a clip cage (2 cm in diam and 1.5 cm in height, upper and lower surface were covered with parafilm and fine mesh, respectively). Subsequently the plants were incubated in a climate room (23 ± 1°C, 50-70% relative humidity, and 16 h:8 h light:dark) for 3 days. Once every day, we changed the position of each clip cage. The last position change was made just before the plants were used in the bioassay. Five infested plants (20 caterpillars in 10 cages on 10 leaves in total) with moist cotton wool were used as odor source. Other conditions and procedures were the same as described above.
Finally, we used clean plants as the odor source. Uninfested lima bean plants were cut and placed with the stem in water in a vial for 3 days in a climate room (23 ± 1°C, 50-70% relative humdity, and 16 h:8 h light:dark). Five uninfested lima bean plants (10 leaves in total) with moist cotton wool were used as odor source.
Bioassay
A closed-system Y-tube olfactometer
(Takabayashi and Dicke, 1992)
was used to determine the predators' response to different odor sources. The
airflow through each olfactometer arm was 4 l/min, which was checked with flow
meters. For details of the olfactometer set-up, see Takabayashi and Dicke
(1992). We conducted
experiments assessing behavioral response to (1) spider-mite infested plants
versus clean plants, (2) caterpillar-infested plants versus clean plants, and
(3) spider-mite infested plants versus caterpillar-infested plants. This was
done for four predator groups: predators that had been reared on spider mites
on lima bean leaves and were starved for either 1 h and 24 h before the
experiment, predators reared on spider mites on cucumber leaves and were
starved for either 1 h and 24 h before the experiment.
Predators were individually introduced into the olfactometer at the
starting point on an iron wire, which was positioned in the center of the
glass tube. The behavior of a predator on the wire was observed for a maximum
of 5 min. The observation was terminated when the predator reached the far end
of one of the arms. Predators that did not reach the end of either arm within
5 min (classified as "no choice," an event that occurred in only
about 5% of all predators tested) were excluded from the statistical analysis.
In every experiment, we alternately tested 1-h starved and 24-h starved
predators that had been reared on the same plant species within an experiment,
so that they were exposed to the same odor sources. The connections of odor
source containers (2-1 conical flasks) to the olfactometer arms were exchanged
after every 10 predators (i.e., five predators of each of the two predator
groups) to control for any unforeseen asymmetry in the setup. In every
experiment, we tested more than 90 satiated or starved predators on 4-6
experimental days. For each experiment on each day new odor sources were used.
The experiments were conducted at 22 ± 2°C. The data for the 1-h
starved and 24-h starved predators in an experiment were subjected to a
chisquare test. The null hypothesis was that the predators had a 50:50
distribution over the two odor sources. In addition, the responses of the two
predator groups (1 h versus 24 h starvation) in an experiment were subjected
to a contingency table test (Sokal and
Rohlf, 1981) with a null hypothesis of no difference in response
among the two predator groups with different starvation level.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Predators reared on spider mites on lima bean plants
Predators that had been starved for 1 h and for 24 h were significantly attracted to volatiles from spider-mite—infested lima bean plants (
2 = 10.6, p =.001 and 2 = 9.9,
p =.002) (Figure 1). In contrast, when offered volatiles from caterpillar-infested bean plants
versus those from uninfested bean plants, only 1-h starved predators were
significantly attracted, while the distribution of 24-h starved predators was
not significantly different from 50:50 (responses of the two predator groups
differed significantly; contingency table with Yates correction,
2 = 8.3, p =.004). When spider-mite—infested
plants were offered versus caterpillar-infested plants, more predators chose
spider-mite—infested plants. This was only statistically significant for
24-h starved predators (2 = 1.2, p =.27 for 1-h
starved and 2 = 5.2, p =.022 for 24-h starved
predators). The responses of the two groups with different levels of
starvation were not significantly different (contingency table with Yates
correction, 2 = 0.44, p =.51); when the data for the
two experimental predator groups were pooled, there was a significant
preference for spider-mite—infested plants over caterpillar-infested
plants (2 = 5.8, p =.016). The data in
Figure 1 show that 24-h starved
P. persimilis responded differently to volatiles from lima bean
plants infested with spider mites than to odors from bean plants infested with
caterpillars. Such a differential response was not recorded for 1-h starved
P. persimilis.
|
Predators reared on spider mites on cucumber plants
For predators that had been reared on cucumber plants, the responses to
volatiles from lima bean plants infested with spider mites and caterpillars
were similar for the two groups with different levels of starvation
(Figure 2). The predators
preferred volatiles from spider-mite—infested lima bean plants over
those from uninfested lima bean plants (2 = 5.4, p
=.020 for 1-h starved and 2 = 5.1, p =.023 for 24-h
starved predators). In contrast, the predators did not prefer volatiles from
caterpillar-infested plants over those from uninfested plants
(2 = 0.9, p =.35 for 1-h starved and
2 = 1.1, p =.30 for 24-h starved predators). As a
consequence, they clearly preferred the volatiles from
spider-mite—infested lima bean plants over volatiles from
caterpillar-infested lima bean plants (2 = 7.1, p
=.008 for 1-h starved and 2 = 6.7, p =.010 for 24-h
starved predators).
|
In conclusion, in contrast to predators that had been reared on spider mites on lima bean, the response of predators reared on spider mites on cucumber plants was not affected by starvation level. The predators discriminated among volatiles from lima bean infested with S. exigua caterpillars and from lima bean plants infested with spider mites, regardless of duration of starvation.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Phenotypically plastic behavior enables animals to adjust their foraging decisions to prevailing conditions and experiences. When local circumstances quickly change, adaptive foraging decisions can be made, but when local circumstances are relatively constant, decisions based on a fixed "rule of thumb" may be more cost effective. Animals are confronted with an overwhelming amount of information, and interpreting all details of the information is costly (Bernays and Wcislo, 1994
; Milinski,
1990). Therefore, discrimination among similar types of
information is adaptive only when this has a net benefit. Odor blends emitted
by plants infested by different herbivore species are similar: they usually
consist of the same components, albeit in different ratios (e.g.,
DeMoraes et al., 1998;
Takabayashi et al., 1991;
Turlings et al., 1993, see
Dicke and Vet, 1999, for
review). Our data demonstrate that discrimination by the predatory mite P.
persimilis among such similar odor blends depends on previous
experiences.
Starvation of P. persimilis does not affect the attraction of the
predators to lima bean leaves infested with their preferred prey, T.
urticae. In contrast, the response to lima bean leaves infested with the
nonprey caterpillars is affected by starvation and by the plant species on
which the predators had fed before the experiment. Predators reared on lima
bean leaves infested with T. urticae were attracted to volatiles from
caterpillar-infested bean leaves when they had been removed from the rearing
plant for 1 h, whereas predators that had been removed from the rearing plant
for 24 h were not attracted by caterpillar-infested bean leaves. Cucumber
leaves infested with T. urticae emit a volatile blend that is
qualitatively different from the emitted by lima bean leaves infested with
T. urticae (Dicke et al.,
1990b,c;
Takabayashi et al., 1994a).
The behavior of predators exposed to this dissimilar odor blend is different:
these predators were attracted to prey (spider mites)-infested bean leaves but
not to nonprey (caterpillar)-infested bean leaves. Starvation did not affect
these responses.
The different behavioral responses of 1-h and 24-h starved predators are not caused by the different levels of starvation per se, but by losing familiarity with the rewarding odor source. This is clear from the differences between predators that have experienced the same odor source as the one presented in the olfactometer (infested lima bean leaves) and predators that have experienced a different odor source (infested cucumber leaves) from the one presented in the olfactometer. Starvation affects the response to volatiles from caterpillar-infested lima bean plants in predators reared on infested lima bean, but not in predators reared on infested cucumber.
The responses of predators to nonprey-related bean plant odors versus odors from uninfested bean plants are also reflected in their choices in the two-choice experiment where plant odors related to nonprey are offered versus plant odors related to prey. Of the four predator groups (two rearing histories x two starvation levels), only 1-h starved predators reared on bean did not discriminate between the two odor sources, whereas predators from the other three groups preferred prey-related odors over nonprey-related odors. The difference between the two predator groups reared on bean is not statistically significant. This may to some extent be a result of the statistical test being not very discriminative. We have recently done the same experiment with the two predator groups reared on spider mites on bean, but with a different spider mite density in the prey-related odor source, and this yielded similar results (Shimoda and Dicke, unpublished data). This further supports the data presented here.
In this study we used an olfactometer to study the behavioral responses of
the predators to prey-related plant volatiles. Behavioral choices recorded in
this bioassay provide reliable information on the predator's behavior under
field conditions, as has been clear from studies by Janssen
(1999) and Zemek and Nachman
(1999).
The data presented support our hypothesis. Predators are attracted to odors
from nonprey-infested plants when they have recently been feeding on
prey-infested leaves of the same plant species. However, after a longer period
away from the prey-infested leaves, the attraction to odors from plants
infested with nonprey disappears. In the real world this represents the case
where predators that forage in a prey patch and accidentally leave the patch
respond to the odors from the prey patch to return to their food source
(Sabelis et al., 1984). This
response to a recently lost prey patch does not require discrimination between
volatile blends that marginally differ, such as blends emitted by plants
infested with different herbivore species
(DeMoraes et al., 1998;
Takabayashi et al., 1991;
Turlings et al., 1993). After
all, the most likely herbivore patch to be found is the one just lost. Thus,
responding to a general olfactory representation of the blend is not likely to
result in a nonprey patch being found. However, when the prey patch has been
lost for a long time, the perception of an odor source is not likely to
represent the one just lost. Thus, an evaluation needs to be made of whether
the odor represents a patch with suitable or unsuitable prey, and hence
discrimination between different blends is functional.
Predators that have recently left an odor-emitting prey patch do not
respond to all details of the odor blend. This seems adaptive because the
central nervous system has a limited capacity to analyze and retain detailed
information, as has been demonstrated for a wide range of animals as well as
for humans (Bernays and Wcislo,
1994; Milinski,
1990). The time devoted to analyzing the details of the chemical
information can alternatively be spent on analyzing information needed, for
example, to locate prey individuals or to avoid falling victim to enemies
(Milinski, 1990). The recorded
behavior fits in the neural hypothesis of the limitation of information input
(Bernays and Wcislo, 1994).
The predatory mite P. persimilis is a specialist on spider mites
in the genus Tetranychus that can feed on several hundred species of
host plants (Van de Vrie et al.,
1972). The predators temporarily specialize on the chemical
information of a prey-infested plant after experience and thus discriminate
between information from different plant species
(Dicke et al., 1990a;
Krips et al., 1999;
Takabayashi et al., 1994b).
This relates to information that is qualitatively different
(Dicke and Vet, 1999;
Turlings et al., 1993). In
addition, with experience on one plant species, the predators also responded
to a general "image" of the complex blend and did not discriminate
among blends from one plant species infested with different herbivore species.
Such blends are qualitatively similar, but differ in the contribution of
individual components (Dicke and Vet,
1999; Turlings et al.,
1993). The nondiscriminative response disappeared within 24 h
after the odor patch was lost. Discriminating between blends that are
qualitatively similar is more difficult than discriminating between ones that
are qualitatively different (Vet et al.,
1998), and thus such discrimination should only be done when it
has a clear benefit. Responding to all the details of a complex blend is
obviously not required when the response to the odor blend is used to stay in
the prey patch. In contrast, during distant location of prey patches, the
odors are used to discriminate between odor patches that differ in
profitability, so as to avoid allocating energy to locating nonprofitable prey
patches. In this latter situation, there is a clear benefit for discrimination
among similar odor blends that represent herbivore patches that significantly
differ in quality and consequently in prospects for reproductive success.
Studies of parasitoids, bees, and bumblebees have shown that insects can
learn to discriminate among rewarding and unrewarding foraging sites
(Dukas and Waser, 1994;
Geervliet et al., 1998; Papaj et al.,
1994; Smith, 1993;
Vet et al., 1995). With
increasing experience, including, for example, a combination of rewarding and
unrewarding experiences, parasitoids, bees, and bumblebees can increase the
ability to discriminate between highly similar foraging sites. This does not
seem to be the case for the predators tested here. However, there are some
important differences in the temporal characteristics of foraging behavior of
predators such as predatory mites on the one hand and parasitoids and bees on
the other. Parasitoids and bees make more foraging decisions per unit time
than predators that forage within a prey patch and that can remain in a prey
patch for all their lives or even for several generations, depending on the
size of the patch. During this prolonged stay in the prey patch, the
arrestment based on responses to steep odor gradients at the edge of the patch
(Sabelis et al., 1984) are
important. It is exactly during this arrestment behavior that a response to
all details of the volatile blend is not important. Recording an attraction to
odors emitted by plants infested by nonprey does not always represent an
inability of the carnivores to discriminate, but can be a functional behavior
in which the animals do not discriminate. The value of resources for an animal
is not constant but is affected by the specific condition of the animal, and
thus resource value may vary over time. Phenotypic plasticity allows animals
to modulate their behavioral decisions as affected by the conditions
experienced. A consequence of responding to a general representation of an
odor blend is that errors will be made because the animal is less
discriminative than it could be. However, if significant benefits are obtained
by not analyzing all the details of the informational input, responding to a
general representation can be clearly adaptive. We have demonstrated that
whether all details of an odor blend affect the response of an animal can be
dependent on the animal's experiences. Therefore, rather than concluding on
the adaptiveness of an animal's behavior from isolated studies of its
response, an analysis of the variation in its response to environmental
variation is important. This study is a building block to the development of
contextual chemical ecology, which is an evolutionary approach to chemical
ecology that aims at studying the adaptiveness of variation in animal behavior
(Robertson et al., 1995;
Vet, 1999).
|
ACKNOWLEDGEMENTS |
|---|
We thank Herman Dijkman for rearing plants, mites, and caterpillars, N. G. Derksen-Koppers for providing eggs of Spodoptera exigua, Gerrit Gort for statistical advice, and Junji Takabayashi for useful comments during this study. The manuscript benefited from constructive comments by Paul Grostal, Jennifer Thaler, Louise Vet, and three anonymous reviewers. M.D. was supported in part by the Uyttenboogaart-Eliasen Foundation.
|
FOOTNOTES |
|---|
T. Shimoda is now at the Laboratory of Ecological Information, Graduate School of Agriculture, Kyoto University, Kyotio 606-8502, Japan.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bernays EA, 1996. Selective attention and host-plant specialization. Entomol Exp Appl 80: 125-131.
Bernays EA, Wcislo WT, 1994. Sensory capabilities, information processing, and resource specialization. Q Rev Biol 69: 187-204.
Brown ES, Dewhurst CF, 1975. The genus Spodoptera (Lepidoptera: Noctuidae) in Africa and the Near East. Bull Entomol. Res 65: 221-262.[Web of Science]
CAB, 1972. Distribution maps of pests. Pest: Spodoptera exigua (Hb.). Series A. Map no. 302. London: Commonwealth Agricultural Bureaux.
Carde RT, Bell WJ (eds), 1995. Chemical ecology of insects 2. New York: Chapman and Hall.
DeMoraes CM, Lewis WJ, Paré PW, Alborn HT, Tumlinson JH, 1998. Herbivore-infested plants selectively attract parasitoids. Nature 393: 570-573.
Dicke M, 1999. Evolution of induced indirect defense of plants. In: The ecology and evolution of inducible defenses (Tollrian R, Harvell CD, eds). Princeton, New Jersey: Princeton University Press; 62-88.
Dicke M, Mass K J van der, Takabayashi J, Vet LEM, 1990a. Learning affects response to volatile allelochemicals by predatory mites. Proc Exp Appl Entomol 1: 31-36.
Dicke M, Sabelis MW, 1992. Costs and benefits of chemical information conveyance: proximate and ultimate factors. In: Insect chemical ecology: an evolutionary approach (Roitberg RD, Isman MB, eds). New York: Chapman and Hall; 122-155.
Dicke M, Sabelis MW, Takabayashi J, Bruin J, Posthumus MA, 1990b. Plant strategies of manipulating predator-prey interactions through allelochemicals: prospects for application in pest control. J Chem Ecol 16: 3091-3118.
Dicke M, Takabayashi J, Posthumus MA, Schütte C, Krips OE, 1998. Plant-phytoseiid interactions mediated by herbivore-induced plant volatiles: variation in production of cues and in responses of predatory mites. Exp Appl Acarol 22: 311-333.
Dicke M, van Beek TA, Posthumus M A, Ben Dom N, van Bokhoven H, de Groot AE, 1990c. Isolation and identification of a volatile kairomone that affects acarine predator-prey interaction: involvement of host plant in its production. J Chem Ecol 16: 381-396.[Web of Science]
Dicke M, Vet LEM, 1999. Plant-carnivore interactions: evolutionary and ecological consequences for plant, herbivore and carnivore. In: Herbivores: between plants and predators (Olff H, Brown VK, Drent RH, eds). Oxford: Blackwell; 483-520.
Doebeli M, 1995. Phenotypic variation, sexual reproduction and evolutionary population dynamics. J Evol Biol 8: 173-194.
Du Y, Poppy GM, Powell W, Pickett JA, Tremblay E, 1998. Strategies involved in the location of hosts by the parasitoid Aphidius ervi Haliday (Hymenoptera: Braconidae: Aphidiinae). Biol Control 11: 104-112.
Dukas R, Waser NM, 1994. Categorization of food types enhances foraging performances of bumblebees. Anim Behav 48: 1001-1006.
Geervliet JBF, Ariëns S, Dicke M, Vet LEM, 1998a. Long-distance assessment of patch profitability through volatile infochemicals by the parasitoids Cotesia glomerata and C. rubecula (Hymenoptera: Braconidae). Biol Control 11: 113-121.
Geervliet JBF, Vreugdenhil AI, Vet LEM, Dicke M, 1998b. Learning to discriminate between infochemicals from different plant-host complexes by the parasitoids Cotesia glomerata and C. rubecula (Hymenoptera: Braconidae). Entomol Exp Appl 86: 241-252.
Hill DS, 1983. Agricultural insect pests of the tropics and their control. Cambridge: Cambridge University Press.
Jaenike J, Papaj DR, 1992. Behavioral plasticity and patterns of host use by insects. In: Insect chemical ecology, an evolutionary approach (Roitberg BD, Isman MB, eds). New York: Chapman and Hall; 245-264.
Janssen A, 1999. Plants with spider-mite prey attract more predatory mites than clean plants under greenhouse conditions. Entomol Exp Appl 90: 191-198.
Janssen A, Bruin J, Jacobs G, Schraag R, Sabelis MW, 1997. Predators use volatiles to avoid prey patches with conspecifics. J Anim Ecol 66: 223-232.
Karban, R, 1986. Induced immunelike resistance to spider mites in cotton. Cal Agric 40: 11-15.
Krips OE, Willems PEL, Gols R, Posthumus MA, Dicke M, 1999. The response of Phytoseiulus persimilis to spider-mite induced volatiles from gerbera: influence of starvation and experience. J Chem Ecol 25: 2623-2641.
Milinski M, 1990. Information overload and food selection. In: Behavioural mechanisms of food selection (Hughes RN, ed). Berlin: Springer Verlag; 721-736.
Papaj DR, 1994. Optimizing learning and its effect on evolutionary change in behavior. In: Behavioral mechanisms in evolutionary ecology (Real LA, ed). Chicago: University of Chicago Press; 133-153.
Papaj DR, Lewis AC (eds), 1993. Insect learning. Ecological and evolutionary perspectives. New York: Chapman and Hall.
Papaj DR, Snellen H, Swaans K, Vet LEM, 1994. Unrewarding experiences and their effect on foraging in the parasitic wasp Leptopilina heterotoma (Hymenoptera: Eucoilidae). J Insect Behav 7: 465-481.
Powell W, Pennacchio F, Poppy, GM, Tremblay E, 1998. Strategies involved in the location of hosts by the parasitoid Aphidius ervi Haliday (Hymenoptera: Braconidae: Aphidiinae). Biol Control 11: 104-112.
Robertson IC, Roitberg BD, Williamson I, Senger SE, 1995. Contextual chemical ecology: an evolutionary approach to the chemical ecology of insects. Am Entomol 41 (4): 237-239.
Roitberg BD, Isman MB (eds), 1992. Insect chemical ecology: an evolutionary approach. New York: Chapman and Hall.
Rollo CD, 1994. Phenotypes: their epigenetics, ecology and evolution. London: Chapman and Hall.
Sabelis MW, 1992. Predatory arthropods. In: Natural enemies: the population biology of predators, parasites and diseases (Crawley MJ, ed). Oxford: Blackwell; 225-264.
Sabelis MW, Dicke M, 1985. Long-range dispersal and searching behavior. In: Spider mites. Their biology, natural enemies and control (Helle W, Sabelis MW, eds). World Crop Pests, vol. 1b. Amsterdam: Elsevier; 141-160.
Sabelis MW, Vermaat JE, Groeneveld A, 1984. Arrestment responses of the predatory mite, Phytoseiulus persimilis, to steep odour gradients of a kairomone. Physiol Entomol 9: 437-446.
Shimoda T, Dicke M, 1999. Plant volatiles induced by non-prey caterpillars affect olfactory preference of predatory mites for those induced by their prey spider mites. J Chem Ecol 25: 1585-1595.
Smith BH, 1993. Merging mechanism and adaptation: an ethological approach to learning and generalization. In: Insect learning: ecological and evolutionary perspectives (Papaj DR, Lewis AC, eds). New York: Chapman and Hall; 126-158.
Smits PH, 1987. Nuclear polyhedrosis virus as biological control agent of Spodoptera exigua. (PhD dissertation). Wageningen, The Netherlands: Wageningen Agricultural University.
Sokal RR, Rohlf FJ, 1981. Biometry. New York: Freeman.
Stephens DW, Krebs JR, 1986. Foraging theory. Princeton New Jersey: Princeton University Press.
Stout MJ, Workman KV, Bostock RM, Duffey SS, 1998. Specificity of induced resistance in the tomato, Lycopersicon esculentum. Oecologia 113: 74-81.
Takabayashi J, Dicke M, 1992. Response of predatory mites with different rearing histories to volatiles of uninfested plants. Entomol Exp Appl 1: 109-113.
Takabayashi J, Dicke M, 1996. Plant-carnivore mutualism through herbivore-induced carnivore attractants. Trends Plant Science 1: 109-113.
Takabayashi J, Dicke M, Posthumus, MA, 1991. Variation in composition of predator-attracting allelochemicals emitted by herbivoreinfested plants: relative influence of plant and herbivore. Chemoecology 2: 1-6.
Takabayashi J, Dicke M, Takahashi S, Posthumus MA, van Beek TA, 1994a. Leaf age affects composition of herbivore-induced synomones and attraction of predatory mites. J Chem Ecol 20: 373-386.
Takabayashi J, Dicke M, Takahashi S, Posthumus MA, van Beek TA, 1994b. Volatile herbivore-induced terpenoids in plant-mite interactions: variation caused by biotic and abiotic factors. J Chem Ecol 20: 1329-1354.
Turlings TCJ, Wäckers FL, Vet LEM, Lewis WJ, Tumlinson JH, 1993. Learning of host-finding cues by hymenopterous parasitoids. In: Insect learning (Papaj DR, Lewis AC, eds). New York: Chapman and Hall; 51-78.
Van de Vrie M, McMurtry JA, Huffaker CB, 1972. Ecology of tetranychid mites and their natural enemies: A review. III. Biology, ecology and pest status and host-plant relations of Tetranychids. Hilgardia 41 (13): 343-432.
Vet LEM, 1996. Parasitoid foraging: the importance of variation in individual behaviour for population dynamics. In: Frontiers of population ecology (Floyd RB, Sheppard AW, eds). Melbourne: CSIRO Publishing; 245-256.
Vet LEM, 1999. From chemical to population ecology: infochemical use in an evolutionary context. J Chem Ecol 24: 31-49.
Vet LEM, Dicke M, 1992. Ecology of infochemical use by natural enemies in a tritrophic context. Annu Rev Entomol 37: 141-172.[Web of Science]
Vet LEM, de Jong AG, Franchi E, Papaj DR, 1998. The effect of complete versus incomplete information on odour discrimination in a parasitic wasp. Anim Behav 55: 1271-1279.[Web of Science][Medline]
Vet LEM, Lewis WJ, Carde RT, 1995. Parasitoid foraging and learning. In: Chemical ecology of insects 2 (Carde RT, Bell WJ, eds). New York: Chapman and Hall; 65-101.
Visser ME, van Alphen JJM, Nell HW, 1990. Adaptive superparasitism and patch time allocation in solitary parasitoids: the influence of the number of parasitoids depleting a patch. Behaviour 114: 21-36.
Zemek R, Nachman G, 1999. Interactions in a tritrophic acarine predator-prey metapopulation system: prey location and distance moved by Phytoseiulus persimilis (Acari: Phytoseiidae). Exp Appl Acarol 23: 21-40.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Ament, M. R. Kant, M. W. Sabelis, M. A. Haring, and R. C. Schuurink Jasmonic Acid Is a Key Regulator of Spider Mite-Induced Volatile Terpenoid and Methyl Salicylate Emission in Tomato Plant Physiology, August 1, 2004; 135(4): 2025 - 2037. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||