A Review of the Possible Influences of Semiochemical Cues on the Multiple Trophic Levels of the Phytophagous Mite, Tetranychus urticae, and its Phytoseiid Predators

Abstract  
The concept that plants defend themselves by releasing volatile chemicals when 
under attack by herbivorous caterpillars has been explored and demonstrated.  
These volatiles subsequently attract Hymenopteran natural enemies of the pest to 
the benefit of the plant.  Studies involving anthocorid predators and psylla-
infested pear trees demonstrated the predator's attraction to these types of 
volatiles as well.  A similar interaction was seen in the response of predatory 
thrips towards bean plants damaged by Tetranychus urticae in both laboratory and 
field studies, showing a preference for those damaged by the pest as opposed to 
those just manually damaged. It appears that the predacious Phytoseiid  mites 
commonly used for biological control of T. urticae also make use of  chemical 
cues in locating their prey.  The role of these interactions on predator 
foraging efficiency is explored as part of the larger picture of predator-prey 
population dynamics. The need to understand behavioral mechanisms underlying 
simple and free distribution population dynamics models is asserted.  The 
effects of using selected vs. unselected lines of predacious mites in studies 
has been explored as well,  looking at the possible implications for improving 
efficacy of phytoseiid predators with genetic manipulation. On a lower trophic 
level than that of predators seeking prey, Tertanychus urticae seems to use  
volatile cues to determine the presence of  heterospecific ,( Frankliniella 
occidentalis), or conspecific herbivores on a potential food source and will 
respond accordingly.  There are several possible scenarios explored as to why 
this response may occur, taking into consideration what the beneficial, or 
possibly detrimental, results of the choice made might be.  It also appears that 
the predacious Phytoseiid mites will, in certain studies, avoid prey patches 
with conspecific predators present.  This is thought to occur due to the release 
of an odour by the predator, which then elicits production of another odour, by 
the prey.  It has been suggested that these odours, and not just physical 
inspection of a prey patch, are what determines the predator's acceptance or 
rejection of a prey patch, determining dispersal or aggregation, and that this 
may be carried out from a distance. Finally, leaf age plays a roll in chemicals 
emitted and subsequent attractiveness to predators,  as well as simple physical 
features of certain host plants as they relate to shelter and ability of  mites 
to disperse.

Introduction
The interaction between host plants , herbivores, and the predators that seek 
them out has been explored in a variety of  complexes.  The use of plant 
released volatiles under attack by caterpillars, by their natural enemies, to 
locate them over distances has been demonstrated (Tumlinson, Lewis, Vet 1993).  
Aggregative responses by anthocoreds to caged Psylla-infested trees led to the 
study of a similar kind of relationship in that predator-prey complex.  Results 
demonstrated that there are apparent differences between volatiles in uninfested 
vs. infested leaves and, anthocoreds respond to compounds isolated from infested 
leaves. (Scuraeanu, Drukker, Bruin, Posthumus, Sabelis 1997).  In both the 
laboratory, as well as in the field, studies involving the predatory thrips, 
Scolothrips takahashii  clearly demonstrated its use of the herbivore-induced 
plant volatiles in foraging .  These demonstrate that predators and parasites 
are using tools beyond simple physical searching in finding prey effectively.  
Phytoseiid mites, used for biological control of Tetranychus urticae, appear to 
be no exception.  T. urticae  seems to make use of volitles itself  in 
determining acceptability of a host.  The predacious mites not only use them to 
find  prey but seem to use them to avoid patches already occupied by 
conspecifics.  There are other potential influencing factors including the age 
of the host plant, species of predacious mite, and possibly the genetic 
predisposition of a specific mite, as well as simple physical features of a host 
plant that can determine the interactions of this predator-prey complex.
Herbivores Use of Plant Volatiles
It appears that not only do the potential natural enemies of initial herbivore 
pests take advantage of  plant released volatiles, but that secondary herbivores 
may be using these cues as well. This can be manifested as either an avoidance 
response or an attraction.  It has  been suggested  that a secondary herbivore 
would react to volatiles produced by a plant under attack due to potential 
changes, either positive or negative, in the acceptability of the host plant . 
(Pallini, Janssen, Sabelis 1996)   In some cases, presence of  another herbivore 
may be positive due to the plant's reduced defensive ability brought on by the 
initial attack .  In others, it may be negative due to induced resistance and 
potential competition.  It may be simply that these volatiles make it easier for 
the secondary herbivore to find a potential host.  T. urticae females have been 
shown to respond to plant odours (Dicke 1986) but a recent study demonstrated 
their avoidance of cucumber plants infested with the western flower thrips,  
Frankliniella occidentralis (Pergande), a heterospecific herbivore.  While these 
two are commonly present in the same crops, intraguild predation on mite eggs by 
the thrips is known to occur.  Therefore it would be to the advantage of the 
mite to avoid such a host if possible.  This hypothesis was further tested and 
results in both olfactometer bioassays, as well as in greenhouse studies, were 
able to demonstrate a significant avoidance response  by T. uticae to the 
presence of F. occidentralis.  While the results of  studies regarding response 
to infestation by conspesifics were less dramatic, they did show some tendencies 
toward plants that were already infested with T. uticae. (Pallini, Janssen, 
Sabelis 1996) A previous study (Dicke 1986) had contrasting results that my be 
partially attributed to the difference in spatial scale of the experiments. It 
is suggested by the authors of the second study that what was initially shown 
was really equivalent to movement within a plant. What the mites may actually be 
doing is using the odours to localize the plants and then, secondarily, be using 
them to search for empty spots within the plant and avoid those already 
inhabited. (Pallini, Janssen, Sabelis 1996).    There are other differences in 
technique that may explain some of the difference in response by T. urticae to 
conspecifically infested plants in these two studies . One of the most important 
may be simply the fact that spider-mite infested bean and cucumber plants do not 
necessarily produce the same volatile signals.  Bean plants infested with T. 
urticae  produce a terpinoid linalool and it is thought to cause dispersal of T. 
urticae. (Dicke et al. 1993) Cucumber plants infested with the same herbivore 
have not been found to produce this terpinoid.  (Takabayashe et al. 1994)   A 
final observation as to why spider-mites might seek out a host already infested 
with conspecifics is that they could be looking to take advantage of the 
established protective webbing. (Pallini, Janssen, Sabelis 1996).  
Selection of Phytoseiid lines for Attraction
Predacious mites must search for their prey on two levels. Prey-infested patches 
need to be located and then prey within a patch must be found.  Spider mites 
tend to occur in clusters due to the fact that offspring don't  move great 
distances from were they were born. It is suggested that the volatiles produced 
by spider mite infested plants may serve to keep the predators within an area 
while they move around and search for the prey. (Margolis, Sabelis, Boyer 1997).  
Variation in response by P. persimilis in some studies was possibly due to 
either differences in the physiological state of the predator or condition of 
the plant.  However, it might also be partially due  to a genetic variation of 
the predator itself.  In order to explore this possibility further, studies 
using selected lines based on their initial responses to spider mite infested 
beans were conducted. After one generation of selection  the line with positive 
response was showing a significant difference from the base population in 
response to the infested bean plants.  

Food depravation was also a factor and, as would be expected, in all cases 
starved mites of all three lines were more responsive to herbivore induced 
odors.  In the presence of high stimulus, the mites from the positive line 
showed greater response overall, regardless of feeding history
Differences in rate of oviposition, feeding, and prey density did not appear to 
affect residence time of individual mites. However the line selection did, with 
the positive lines showing shorter residence times that the base populations.  A 
possible explanation for this increased movement was suggested as a response to 
increased  sensitivity to smaller volatile concentrations .  (Margolies, 
Sabelis, Boyer 1997). The other possible conclusion  that was considered is that 
this increased sensitivity is advantageous in that it could result in increased 
efficiency in prey finding by positively selected lines.  The success of a 
predator in terms of it's evolutionary selection as well as it's success as a 
biocontrol agent,  is of course closely related to it's searching ability and 
efficiency.  The results taken together seem to indicate that there is a genetic 
component to the big picture as well as just condition of the plant or predator 
at a given time.
Avoidance by Predators of  Prey Patches with Conspecifics
Simple models of foraging often assume that the foragers are aware of how many 
competitors are present, presumably by inspection or by learning from previous 
experience.  Some of the earlier and more rigid behavioral  submodels assumed 
that predators chose patches regardless of  the presence of conspecifics and 
that predators stayed in one prey patch for an entire lifetime. These have been 
somewhat relaxed.(Janssen, Bruin, Jacobs, Schraag, Sabelis 1997)  Decreases in 
the rate of reproduction, as well as in feeding rate and number of prey caught 
have been shown in higher density predator patches. (Yao & Chant 1990) It is 
reasonable to assume that females who avoid high density patches would benefit.   
Past studies have demonstrated repeatedly  the fact that residence time 
decreases as conspecific numbers increase. ( reviewed by Sabelis 1985)  A recent 
study (Janssen, Bruin, Jacobs, Schraag , Sabelis 1997) was able to demonstrate 
another possible source of odours that produce this response. It appears that it 
is not actually the odour of conspecific predators or of just the damaged prey , 
both of which on their own elicit an attractant response.  Differences in 
response according to position of the predators in the odour source seem to 
indicate it is a more complicated series of events that leads to avoidance by 
conspecifics.  When the predators in the source were upwind of the prey patch, 
the expected avoidance response was seen.  However, when they were down wind 
this was not observed.  Furthermore, when adult prey were  removed, so was the 
negative response to that prey patch.  In fact patches with conspecifics, and no 
prey, are attractive. (Janssen, Bruin, Jacobs, Schraag, Sabelis 1997)   
  The tentative conclusion is that it is an odour released by the prey, in 
response to odours from the predators upwind, that causes the effect on 
conspecifics downwind. This is possibly an alarm pheromone produced by the 
spider mite warning other conspecifics of the predator's presence.  It was not 
necessary for them to be being fed upon to elicit the response therefore it is 
not strictly damage to prey that alerts conspecifics.  (Janssen, Bruin, Jacobs, 
Schraag, Sabelis 1997).  The implications are that  these predators don't 
actually need to visit the site to determine the presence or absence of 
competitors but a decision can be made from a remote location.  What affect 
these cues have on the large spatial scale of distribution  is not totally 
clear.(Janssen et al. 1997).  An important point that is brought forth by the 
authors is that this basic behavioral mechanism would greatly reduce the time 
and energy required to determine what prey patches would be desirable.  In 
addition, the inclusion of this type of mechanism would have significant effects 
on predictions made by population models and therefore should be further 
investigated. (Janssen, Bruin, Jacobs, Schraag, Sabelis 1997) 
Generalists vs. Specialist Phytoseiid mites: Influence on Aggregation Patterns
A study involving three commonly used Phytoseiid mites for biological control of  
T. urticae showed varying responses to distant patches and was related by the 
authors to range in specificity of  polyphagy of the predators and density 
differences of prey patches.  T. occidentalis, with a narrower range of 
polyphagy, was responsive to prey patches but not necessarily with varying rates 
according to density. A. andersoni,  a generalist,   appeared to search randomly 
without regard to density either.  Interestingly enough,  the highly 
specialized, oliogophagous , P. persimilis was responsive to high density prey 
patches and was the only one of the three to demonstrate the ability to find 
small prey patches from a short distance. It was asserted that degree of 
aggregation increased with degree of specialization (Zhang, Sanderson 1997)  As 
was suggested in the studies above conducted by Janssen et al. 1997, the 
response in these studies indicated that the P. Persimilis were  also responding 
to volatile cues from upwind. However, the authors felt that the evidence 
presented in this study did not support the theory that these predators are 
actively orienting to more distant patches with higher density, i.e.: non-random 
patch-entry due to these cues alone.  It is asserted, however, that different 
underlying behavioral processes possibly affect patterns of aggregation.  
(Waage, 1979; Lessells, 1985; Morris and Karevia, 1991). Further studies 
conducted in the greenhouse seemed to support the concept of variability in 
aggregation due to different degrees of  specialization within various spatial 
scales.  The more highly specialized P. persimilis aggregated strongly on lower 
spatial levels and was more random in higher spatial levels.  The less 
specialized A. andersoni  showed and aggregation pattern that was exactly 
opposite.  These results give a characterization  of the different patterns of 
aggregation and support the idea of behavioral mechanisms and their potential 
effects as well as showing that  different degrees of polyphagy may play a role 
in determining patterns. (Zhang, Sanderson 1997).
Effects of Leaf Age on Attractiveness of Volatiles to Phytoseiid mites
In the earlier mentioned study regarding aggregation responses by anthocorid 
predators to Psylla-infested trees,  it was observed that there existed a  
variation in the blend of volitiles both qualitatively and quantitatively. It 
was suspected that this could be at least in part attributed to differences in 
leaf age. (Scutareanu, Drukker, Bruin, Posthumus, and Sabelis 1997) In addition 
earlier studies by Takabayashi et al.(1994) were cited as having found similar 
differences due to age in cucumber leaves.   In the study by Takabayashi  et. al 
published in 1993, bioassays were conducted to determine if cucumber leaf age 
affected the response of the predacious mites.  In addition, starved vs. fed 
mites and  mites reared on bean plants vs. those on cucumbers where looked at to 
rule out other  possibly influencing factors.  In testing the different 
combinations, it was determined the only factor affecting preference by the 
phytoceiid mites was leaf age. Both  the starved mites and the well fed mites 
preferred young infested leaves to uninfested ones.  However when both were 
offered old leaves,  there was no preference for old infested leaves over 
uninfested old leaves.  While this supports the fairly well established 
assertion that there is an herbivore-induced volatile involved in attracting 
predators to Phytophagous mites, it demonstrates another influence due to 
variation in the host plant.  When an equal combination of old and new infested 
leaves was combined, the preferential response by P. persimilis raised on Lima 
beans for young infested leaves seemed to be negated.  It is interesting to note 
that those raised on cucumber where not deterred by the mixture and still showed 
a positive response. Previous studies (Dicke et al., 1990b) had mixed leaves of 
different ages as the possible effects of leaf age were unsuspected at that 
time, and so had different results.  A number of compounds were isolated and 
found to be present only among the volatiles isolated from infested leaves and 
not from ones that had only been artificially damaged  or not damaged at all.  
It was interesting to note that one of these was methyl salicylate found to 
cause positive responses in the Psylla-pear complex mentioned above. It was not 
a quantitative difference in compounds that seemed to account for differences in 
response to old vs. young leaves, but a qualitative one. There is an apparent 
exchange in the amount of two of the oximes, (3-methylbutanal O-methyloxime 
32.0%, and an unknown oxime) isolated , as well as the nitriles ( 2-
methylbutanenitrile, 2-methylpropanenitrile, and 3-methylbutanenitrile).  These 
nitriles and oximes were ,at the time of this study, only known to have been 
found in the plant  leaf volatiles associated with apple leaves infested by T. 
urticae (Takabayashi et al. 1993).   The two main chemicals induced by 
herbivores also show changes according to age of leaves tested,  ((E) -(-ocimene 
and(3E)-4,*-dimethyl-1,3,7-nonatriene.) The first is produced less in the 
younger infested leaves and the second relatively more in younger infested 
leaves.(Takabayashi et. al 1994).  The main volatiles being emitted by undamaged 
cucumber leaves were identified and shown to be distinctly different from the 
main compounds mentioned above.     As is illustrated with the variance in 
results shown above, understanding the effect of something as simple as leaf age 
can have a significant bearing on results of studies.  
Physical Aspects of the Host, and Prey Abundance Affects on Phytoseiid mites 
 Finally, by backing up and taking a much simpler view, or at least a non-
semiochemical view,  the simple effects of plant structure due to phylogeny of 
the host have been demonstrated.  Coming down to this lower trophic level, a 
study of host influences on the dynamics of these populations reveals an element 
that can not be dismissed.  The host plant can provide shelter, alternate food 
sources, and in some cases hinder movement and dispersal of both Phytophagous 
mites and their prey. (Karban, English-Loeb, Walker, Thaler 1995) In looking at 
these potential effects, several species of grapes were studied and variables 
such as density of prostrate leaf hairs and  density of erect hairs along main 
veins and axils were rated.  Also included in analysis was the presence or 
absence of  domatia which are hypothesized to serve as shelters for phytoseiids 
(O'Dowd and Willson 1989)  Results of this particular study did not indicate 
high numbers of T. pacificus McGregor present and there was not much variation 
between grape species.  More predacious mites were found than Phytophagous 
species and no difference due to gender of the plant was indicated.  It was the 
leaf surface characteristics that seemed to explain a large part of the 
variation.  Positive associations with greater density of vein hairs,  leaf axil 
bristles, as well as with the presence of domatia  were indicated.  It  In this 
particular complex, results indicate that it is these key host plant features 
and not presence or absence of prey species that determined aggregation of  the 
predacious mite T. caudiglans.  Reduction in desiccation of eggs due to 
increased shelter may be one  reason that there are higher numbers of T. 
caudiglans found on leaves with these characteristics.  It can not, however, be 
extrapolated that the leaves with higher numbers of the predators have lower 
numbers of Phytophagous mites.  

Summary

In reviewing the multiple trophic levels that can be involved in the predator- 
prey complexes, it quickly becomes obvious that the possible interactions are 
endless.  Even when focusing on a more specific example such as that of the 
highly specific P. persimilus  and T. urtica, host plant differences and  
seasonal effects become evident and have the potential to alter results if not 
accounted for.  It would seem reasonable to assume when applying models to make 
predictions about population dynamics, consideration for as many possible, and 
some as yet unrecognized, influences would be a wise approach. 

References

Dicke, M., Vandermass, K.-J., Tadabayashi, J., and Vet, L.E.M.(1990b) Learning 
affects response to volatile allelochemicals by predatory mites. Proc.Exp. Appl. 
Entomolgy N>E>V Amsterdam 1:31-36. 

Dicke, M.(1986) Volatile  spider mite pheromone and host plant kairomone, 
involved in spaced out gregariousness in the spider Tetranychus urticae, Physiol 
Entomology 11:251-262

Dicke, M., Bruin,J., Sabelis, M.W.(1993) Herbivore-induced plant volitiles 
mediate plant-carnivore, plant-herbivore and plant-plant interactions; Talking 
plants revisited. In: Schultz, J. Raskn I (eds) Plant signals in interactions 
with other plants, American Society of Plant Physiologists, Rockville, Maryland, 
U.S.A. pp182-196.

Dicke, M., (1994) Local and Systemic production of volatile herbivore-induced 
terpenoids: their role in plant-carnivore mutualsim. J Plant Physiol 143:465-
472.

Janssen,A., Bruin, J., Jacos, G., Schraag, R., and Sabelis,M. (1997). Predators 
use volatiles to avoid prey patches with conspecifics. Journal of Animal 
Ecology.66, 223-232.

Karban, R., English-Loeb, G., Walder, A., Thaler, J.(1995) Abundance of 
phytosiied mites on Vitis species: effects on leaf hairs, domatia, prey 
abundance and plant phylogeny. Experimental and Applied Acarology, 19,189-197. 

O'Dowd, D.J., Willson, M.F., (1989) Leaf domatia and mites on Australian plants: 
ecological and evolutionary implications. Biol. J. Linn. Soc. 37,99191-236.

Palline, A., Janssen, A., Sabelis, M. (1997) Odour-mediated responses of 
phytophagous mites to conspecific and heterospecific competitors. Oecologia 
110:179-185.

Scutareanu, P., Drukker, B., Bruin,J., Posthumus, M., Sabelis, M.,(1997) 
Volatiles from Psylla-infested pear trees and their possible involvemnet in 
attraction of anthocorid predators. J.of Chem. Ecol., Vol. 23, No. 10.

Shimoda, T., Takabayashi, J., Ashihara, W., Takafuju, A.(1997) Response of 
predatory insect Scolothrips takahashii toward herbivore-induced plant volatiles 
under laboratory and field conditions. J. of Chem. Ecology vol.23, No 8, 1997

Takabayyashi er al.(1993)Herbivore-induced synomones of cucumber plants. J.of 
Chem. Ecology, vol. 20, N.2, 1994.

Trichilo, P.J., Leigh, T.f. (1986) Predation on spider mite eggs by the western 
flower thrips, Frankliniella occidentalis(thysnoptera: thripidae) an opportunist 
in a cotton agroecosystem. Environ. Entomology 15:821-825.

Tumlinson, J.H., Lewis, J.W., Ber.L.E.M.(1993) How Parasitic wasps find their 
Hosts. Scientifice American, March 1993,pp.100-106.

Yao, D.S., and Chant, D. A. (1990) Changes in body weight of two species of 
predatory mites(Acari: Phytoseiidae) as a consequence of feeding on an 
interactive system. Experimental and Applied Acarology, 8, 195-220.

Zhang, Z., Sanderson, J.P.(1997) Patterns, mechanisms and spatial scale of 
aggregation in generalist and specialist predatory mites(Acar: Phytoseiidae) 
Exper. and Applied Acarology 21,393-404(1997).