Experimental and Applied Acarology 27: 27–37, 2002.
© 2002 Kluwer Academic Publishers. Printed in the Netherlands.
Plant feeding by a predatory mite inhabiting
S. MAGALHÃES1,? and F.M. BAKKER2
1Section Population Biology, University of Amsterdam, Kruislaan 320, 1098 SM Amsterdam,
2 MITOX, Kruislaan 320, 1098 SM Amsterdam, The Netherlands
(Received 23 May 2002; accepted 20 September 2002)
Abstract. Plant feeding by arthropod predators may strongly affect the dynamics of bi-
and tri-trophic interactions. We tested whether a predatory mite, Typhlodromalus aripo, feeds
upon its host plant, cassava. This predator species is an effective biological control agent of
Monoychellus tanajoa (the cassava green mite or CGM) a herbivorous mite speci?c to cassava.
We developed a technique to detect plant feeding, based on the use of a systemic insecticide.
We found that T. aripo feeds upon plant-borne material, while other predatory mite species,
Neoseiulus idaeus and Phytoseiulus persimilis, do not. Subsequently, we measured survival of
juveniles and adult females of T. aripo and N. idaeus, both cassava-inhabiting predator species,
on cassava leaf discs. Survival of T. aripo was higher than that of N. idaeus. Thus, T. aripo was
able to withstand longer periods of prey scarcity. Because CGM populations ?uctuate yearly
and are heterogeneously distributed within plants, plant feeding may facilitate the persistence
of populations of T. aripo in cassava ?elds and its control of CGM outbreaks.
Key words: plant feeding, alternative food sources, omnivory, Typhlodromalus aripo, cassava
Many predatory arthropods utilize plant-borne material, such as pollen, nec-
tar, exudates or leaf tissue, in addition to their primary diet (Price et al.,
1980; Overmeer, 1985; Coll and Guershon, 2002). Plant feeding by preda-
tors may alter the dynamics of predator–prey interactions, for example by
allowing persistence of predator populations in absence of prey (Bakker and
Klein, 1992), by increasing predator numbers or by decreasing their preda-
tion rate (Eubanks and Denno, 1999; Lalonde et al., 1999; Van Rijn and
Tanighoshi, 1999; Van Rijn et al., 2002). Moreover, attraction and arrestment
?Author for correspondence (Fax: +31-20-5257754; E-mail: firstname.lastname@example.org)
of predators by means of plant-provided alternative food sources may affect
plant ?tness (Bentley, 1977; Heil et al., 2001; Van Rijn et al., 2002).
We tested the occurrence of plant feeding in Typhlodromalus aripo
DeLeon, a predatory mite species inhabiting cassava. Cassava is a subsistence
crop in several regions in the tropics. It is attacked by a few herbivores, among
which is the cassava green mite (CGM), Mononychellus tanajoa Bondar. In
absence of predators, CGM populations may cause severe damage to cassava
plants (Yaninek et al., 1989). T. aripo is the most effective predator against
CGM. It was ?rst reported by Bakker and Klein (1992) from the apices of
cassava. In plants with T. aripo, CGM is generally absent from the apices
(Bakker, 1993), but it occurs lower in the plant. Since T. aripo migrates to
the lower strata at night (Onzo et al., 2002), it could feed upon CGM during
these migrations. Moreover, CGM migrates to the upper strata during the day,
both in presence and absence of leaf-dwelling predators (Magalhães et al.,
2002). Thus, T. aripo could feed on CGM that migrates to the apex. However,
electrophoresis data show that a large proportion of T. aripo collected in the
?eld do not contain CGM in their gut (Bakker, 1993). The upper strata of
cassava plants harbour few prey species. Apart from CGM, some species
of thrips and of white?ies are found, but T. aripo feeds poorly on these
prey (Bakker, 1993). Therefore, we hypothesize that T. aripo feeds on plant-
borne material from cassava. Testing this hypothesis is the aim of the present
Materials and Methods
Experiments were done in Cruz das Almas (Brazil) and in Amsterdam (The
Netherlands). We present the cultures in the two locations separately and the
experimental procedure together.
Cassava (variety: Cigana Preta) was grown in a screenhouse at the
Empresa Brasileira de Pesquisa em Agropecuaria (EMBRAPA). Plants were
planted as stakes (circa 20 cm) in 2.5-l plastic pots (surface: 20 cm2), and
watered every other day. CGM was reared on young potted plants in another
screenhouse. Clean plants were infested by placing CGM-infested leaves at
the base of one or more petioles. Every month, cultures were supplemented
with CGM individuals collected from several ?elds around the station.
Typhlodromalus aripo and Neoseiulus idaeus Denmark and Muma were
reared in an acclimatized room at 24 ± 2?C, and 70–90% RH. Each rearing
unit consisted of a PVC arena (20 × 20 cm) on top of a sponge, surrounded
by wet tissue and placed in the middle of a 30 × 30 × 10 cm plastic tray with
water on the bottom. This set-up prevented escapes and enhanced humid-
ity (Mégevand et al., 1993). Predatory mites were fed three times a week
by introducing two CGM-infested leaves in the rearing trays. Every month,
predator cultures were supplemented with specimen collected in the ?eld.
Cassava (variety: CMC 40) was grown in a greenhouse at 25?C,
70% RH and 16L/8D photoperiod at the University of Amsterdam. Plants
were planted as stakes (circa 20 cm) in 1-l pots (surface: 15 cm2) with soil
and a 28N, 14K, 14P fertilizer. CGM was reared on intact plants in a sepa-
rate greenhouse compartment. Clean plants were infested by placing CGM-
infested leaves at the base of one or more leaf petioles. Typhlodromalus aripo
(from Benin, West-Africa) and Phytoseiulus persimilis Athias-Henriot (from
Sicily, Italy) were reared in a climate room, under the same conditions as
those in the greenhouse compartments. The rearing procedure for T. aripo
was the same as the one followed in Brazil. Every three months, cultures
were renewed with specimen that had been collected in the ?eld. P. persimilis
was reared on bean leaves infested with Tetranychus urticae Koch placed on
top of inverted pots. Each rearing unit consisted of two pots placed on a tray
with water and covered with a 40 × 40 × 60 cm Plexiglas box. Infested leaves
were added thrice a week.
To detect plant feeding in T. aripo, we developed a technique based on
the use of a systemic insecticide. Replicates differed in a few technical de-
tails (cf. Table 1), but the basic principle was maintained across all
Table 1. Experimental details
24 ± 2
25 ± 5
Relative Humidity (%)
Substrate for leaf-discs
For each replicate, we selected 4 three-week-old plants from the same
variety (Table 1). We placed a systemic insecticide (Temik 10G®, active
ingredient: aldicarb) on the pots of two of them at rates of 1–5 g/l (Table 1)
and left the other two pots untreated. We watered the two sets of plants every
other day during 15 days, enough time for the insecticide to be translocated
from roots to leaves (Ridgway et al., 1967; Chamberlin et al., 1992), and
not enough for it to be lost (Arienzo et al., 1991; Chamberlin et al., 1992).
Then, we cut leaf discs (diam. 2 cm) from the two types of plants and placed
them on a substrate that would keep them fresh (Table 1). We placed one
T. aripo per leaf disc and assessed mortality 24 h later. Under the condition
that bio-availability of the insecticide was restricted to inside the leaf tissue,
any increase in mortality in insecticide leaf discs compared to control leaf
discs would be the result of intake of pesticide-contaminated plant materi-
al by active feeding. To test if the only possible route of pesticide uptake
was through feeding, additional treatments were included in the experimental
To test for contamination of the leaf surface, bio-assays with two addi-
tional species of predatory mites were included as a negative control. The
species used, N. idaeus and P. persimilis, are relatively specialized on one
prey type (McMurtry and Croft, 1997), thus expected not to feed on plant
material. Mortality of these species on insecticide leaf discs invalidates the
test design because it might indicate that the bio-availability of insecticide
was not restricted to the leaf tissue (it may also result from these species
feeding on the plant, but this explanation cannot be distinguished from the
Alternatively, a negative result (i.e., no mortality) in N. idaeus or P. per-
similis versus a positive result with the test species T. aripo could also be
caused by differences in susceptibility to the pesticide. To test the suscep-
tibility of these species, we performed slide-dip tests (Busvine, 1971): mites
were glued by their backs to a double-sided sticky tape on a microscope slide,
then dipped for 15 s in solutions with different concentrations of the active
substance. Twenty-four hours later, mortality (i.e., the percentage of mites
not moving their legs when dabbed with a brush) was assessed, and compared
with that of mites on a control slide that had been dipped in water (control).
Mortality of the individuals tested could also result from contamination
of the substrate supporting the leaf discs. To test if this was the case, we
removed insecticide leaf discs after the test and placed control leaf discs
on the same substrate. On these leaf discs, we placed individuals from the
species that showed high mortality on insecticide leaf discs. One day later,
we assessed mortality. Occurrence of mortality would indicate contamination
of the substrate. Finally, to test whether our method was adequate to induce
mortality in leaf-feeding mites, we assessed mortality of CGM on insecticide
versus control leaf discs.
To test whether plant feeding affected the survival of predatory mites that
inhabit cassava (T. aripo and N. idaeus), we measured survival of juveniles
and females of these species on cassava leaf discs without prey. Egg cohorts
of T. aripo and N. idaeus were produced by well-fed females placed on CGM-
infested cassava lea?ets for 24 h. Subsequently, eggs were placed individually
on cassava leaf discs (diam. 2 cm). As soon as the eggs hatched, leaf discs
were replaced every day until the individuals died. Individuals that escaped
the set-up were used as censored data in the analysis. Sample sizes for total
data: N. idaeus – 60, T. aripo – 41; uncensored data: N. idaeus – 22, T. aripo –
19. To measure survival of adult females without prey, we did cohorts of both
species until the deutonymph stage, then transferred each female individually
to a leaf disc with conspeci?c males and CGM. As soon as the female had
oviposited its ?rst egg, she was moved to a clean leaf disc and followed the
same treatment as the juveniles. Sample sizes for total data: N. idaeus –
52, T. aripo – 70; uncensored data: N. idaeus – 31, T. aripo – 40. Survival
curves were constructed using the Kaplan–Meier method. Differences
between curves were analysed with the Gehan’s Wilcoxon test (Hosmer and
Nearly all T. aripo died 24 h after being placed on insecticide leaf discs
(Table 2). No mortality was observed on control discs. The positive reference
treatment (CGM) showed a similar mortality pattern. On leaf discs placed on
the same substrate where insecticide leaf discs had been, no mortality was
Table 2. Mortality (mean ± S.E) of CGM, T. aripo, N. idaeus and P. persimilis on
leaf discs from plants with and without systemic insecticide. ‘Substrate control’
refers to mortality assessed on leaf discs placed on the same substrate where the
insecticide leaf discs had been placed
(N = 70)
(N = 70)
(N = 10)
(N = 60)
0.97 ± 0.013
0.14 ± 0.039
0.02 ± 0.017
Table 3. Slide-dip test in (a) Brazil and (b) Amsterdam. Indicated
are the fractions of mites that died upon exposure to each concen-
tration of the insecticide. N per insecticide concentration = 10,
except for 0.001 g/l and control in (b), where N = 20 initially, to
compensate for a few individuals that escaped from the sticky tape
observed. This indicated that the insecticide was not available from the sub-
strate. No N. idaeus and only few P. persimilis died when placed on insecti-
cide leaf discs, and none died on control leaf discs. The slide-dip tests (Table
3) yielded only small differences among the phytoseiid species at very low
insecticide concentrations. Thus, relative to the level of pesticide expected in
the plant (Andrawes et al., 1971, 1973), similar mortality levels on insect-
Figure 1. Longevity of T. aripo (black dots) and of N. idaeus (white dots) on cassava leaf
discs: (a) juveniles, (b) adult females.
Figure 1. (Continued.)
icide leaf discs could be expected among the species used. Together, these
results show that T. aripo feeds on the plant while N. idaeus and P. persimilis
When placed on clean cassava leaf discs, T. aripo juveniles survived longer
than N. idaeus juveniles (p = 0.004, Figure 1a). Similarly, survival of
T. aripo adult females was higher than that of N. idaeus adult females
(p < 0.001, Figure 1b).
Our results show that T. aripo feeds upon cassava leaves, whereas P. per-
similis and N. idaeus do not. Similar results were reported by Porres et al.
(1975), where one out of four predatory mite species tested appeared to feed
on green plant tissue, and by Soares et al. (1996).
Among the most widely used methods to detect plant feeding by arthro-
pods are direct observations (Al-Wahaibi and Walker, 2000; Messchendorp
et al., 2000; Pedrosa-Macedo, 2000; Montserrat, 2001) and measuring feed-
ing damage (Meyer, 1993; Agrawal et al., 1999; Greenberg et al., 2000).
However, these methods are not universally applicable, since direct obser-
vations are only possible with relatively big arthropods and plant feeding
does not necessarily entail visible damage (Montserrat, 2001). Many stud-
ies compare life histories of arthropods on plant tissue and on an arti?cial
substrate as a test for plant feeding (e.g., Ruberson et al., 1986; Eubanks
and Denno, 1999; Gillespie and McGregor, 2000). While this method clearly
shows the effect of plants on the life history of predatory arthropods, it does
not necessarily demonstrate that this effect is due to feeding. In fact, many
characteristics of the leaf substrate (microclimate, leaf topography) may af-
fect the life history of predatory arthropods (Kareiva and Sahakian, 1990;
van Lenteren et al., 1995). Therefore, differences in life-history traits of
arthropods placed on leaf versus plastic substrates may be interpreted in sev-
eral ways. In some studies, plant tissues have been labelled, by either radiola-
belling (Porres et al., 1975; Ostrom et al., 1997; Armer et al., 1998) or by
staining (Armer et al., 1998). The analogous use of a systemic insecticide to
detect plant feeding simpli?es this methodology.
Using this new method allowed us to show that T. aripo takes up nutrients
from cassava leaves, but the speci?city of this predator–plant association is
as yet unclear. It seems unlikely that T. aripo has taken up nutrients from
cassava exudates because N. idaeus, known to feed upon cassava exudate
(Tanigoshi et al., 1993; Mégevand and Tanigoshi, 1995), did not die in our
test. Moreover, Temik is not available from the leaf surface (Ingram et al.,
1997). Thus, T. aripo has probably fed upon plant-borne material and not
upon microorganisms from the phyllosphere. Since cassava leaves contain cy-
anide (Haque and Bradbury, 2002), T. aripo must have developed adaptations
to overcome this plant defence. This suggests that the interaction predator–
plant is speci?c, but this hypothesis needs further testing. Moreover, T. aripo
occurs on the upper strata of cassava plants. Due to the apical dominance of
cassava, nutrient density in the apices is expected to be high. Thus, the within-
plant distribution of T. aripo may be related to the within-plant availability of
Typhlodromalus aripo survived longer than N. idaeus on cassava leaf discs.
Probably, the nutrients that T. aripo took from cassava leaves contributed to
this higher survival. However, differences in survival may also be related to
differences in metabolic rates between species. In any case, both juveniles
and adults of T. aripo are able to survive long periods without prey. CGM
populations are heterogeneously distributed in space and time (Noronha and
Silva, 1998). Moreover, predatory mites are passive dispersers, thus prone to
land on plants without prey. Since the dispersing stage of predatory mites is
the young female (Kennedy and Smitley, 1985), this stage (and the eggs it
carries) is likely to undergo periods of prey scarcity. Feeding on the plant
may allow for the persistence of T. aripo populations in absence of prey and
contribute to the success of T. aripo as a biological control agent of CGM.
We are grateful to Aloyseia Noronha and Edmilson Santos Silva for help at
the EMBRAPA station. IITA is thanked for shipments of T. aripo, Bas Pels
for providing P. persimilis and Rhone-Poulenc/Aventis for providing Temik.
We thank Marta Montserrat for statistical help and Maria Nomikou, Jan Bruin
and three anonymous referees for comments. SM was funded by the TMR
research program, FB by MITOX.
Agrawal, A.A., Kobayashi, C. and Thaler, J.S. 1999. In?uence of prey availability and induced
plant resistance on omnivory by western ?ower thrips. Ecology 80: 518–523.
Al-Wahaibi, A.K. and Walker, G.P. 2000. Searching and oviposition behaviour of a mymarid
egg parasitoid, Anagrus nigriventris, on ?ve host plant species of its leafhopper host,
Circulifer tenellus. Entomol. Exp. Appl. 96: 9–25.
Andrawes, N.R., Bagley, W.P. and Herret, R.A. 1971. Metabolism of 2-methyl-2-
(methylthio)propionaldehyde O-(methylcarbamoyl)oxime (temik aldicarb pesticide) in
potato plants. J. Agr. Food Chem. 19: 731–737.
Andrawes, N.R., Romine, R.R. and Bagley, W.P. 1973. Metabolism and residues of temik
aldicarb pesticides in cotton foliage and seed under ?eld conditions. J. Agr. Food Chem.
Arienzo, M., Basile, G., Palmieiri, G., Lucisano, A. and Amorena, M. 1991. Assorbimento
e traslocazione dell’ Aldicarb in alcune specie vegetali coltivate. Agrochimica XXXV:
Armer, C.A., Wiedenman, R.N. and Bush, D.R. 1998. Plant feeding site selection on soybean
by the facultatively phytophagous predator Orius insidiosus. Entomol. Exp. Appl. 86:
Bakker, F.M. 1993. Selecting phytoseiid predators for biological control, with emphasis on the
signi?cance of tri-trophic interactions. PhD Thesis, University of Amsterdam.
Bakker, F.M. and Klein, M.E. 1992. Transtrophic interactions in cassava. Exp. Appl. Acarol.
Bentley, B.L. 1977. Extra?oral nectaries and pugnacious predators. Annu. Rev. Ecol. Syst. 8:
Busvine, J.R. 1971. A critical review of the techniques for testing insecticides. Commonwealth
Agricultural Bureaux, London.
Chamberlin, J.R., Todd, J.W., Farrow, J.M. and Mullinix, B.G. 1992. Aldicarb
residue persistence in leaf terminals of ‘Florunner’ peanut. J. Econ. Entomol. 85:
Coll, M. and Guershon, M. 2002. Omnivory in terrestrial arthropods: mixing plant and prey
diets. Annu. Rev. Entomol. 47: 267–297.
Eubanks, M.D. and Denno, R.F. 1999. The ecological consequences of variation in plants and
prey for an omnivorous insect. Ecology 80: 1253–1266.
Gillespie, D.R. and McGregor, R.R. 2000. The functions of plant feeding in the omnivo-
rous predator Dicyphus hesperus: water places limits on predation. Ecol. Entomol. 25:
Greenberg, R., Bichier, P., Angon, A.C., MacVean, C., Perez, R. and Cano, E. 2000. The
impact of avian insectivory on arthropods and leaf damage in some Guatemalan coffee
plantations. Ecology 81: 1750–1755.
Haque, M.R. and Bradbury, J.H. 2002. Total cyanide determination of plants and foods using
the picrate and acid hydrolysis methods. Food Chem. 77: 107–114.
Heil, M., Koch, T., Hilpert, A., Fiala, B., Boland, W. and Linsenmair, K.E. 2001. Extra-
?oral nectar production of the ant-associated plant, Macaranga tanarius, is an induced,
indirect, defensive response elicited by jasmonic acid. Proc. Natl. Acad. Sci. USA 98:
Hosmer, P.W. and Lemeshow, S. 1998. Applied Survival Analysis. John Wiley and Sons,
Ingram, J.C., Groenewold, G.S., Appelhans, A.D., Delmore, J.E., Olson, J.E. and Miller, D.L.
1997. Direct surface analysis of pesticides on soil, leaves, grass and stainless steel by static
secondary ion mass spectrometry. Environ. Sic. Technol. 31: 402–408.
Kareiva, P. and Sahakian, R. 1990. Tritrophic effects of a simple architectural mutation in pea
plants. Nature 345: 433–434.
Kennedy, G.G. and Smitley, D.R. 1985. Dispersal. In: Spider Mites – Their Biology, Nat-
ural Enemies and Control, W. Helle and M.W. Sabelis (eds), pp. 233–242. Elsevier,
Lalonde, R.G., McGregor, R.R., Gillespie, D.R. and Roitberg, B.D. 1999. Plant feeding by
arthropod predators contributes to the stability of predator–prey population dynamics.
Oikos 87: 603–608.
Magalhães, S., Janssen, A., Hanna, R. and Sabelis, M.W. 2002. Flexible antipredator be-
haviour in herbivorous mites through vertical migration in a plant. Oecologia 132:
McMurtry, J.A. and Croft, B.A. 1997. Life-styles of phytoseiid mites and their role in
biological control. Annu. Rev. Entomol. 42: 291–321.
Mégevand, B. and Tanigoshi, L.K. 1995. Effects of prey deprivation on life table attri-
butes of Neoseiulus idaeus Denmark and Muma (Acari: Phytoseiidae). Biol Contr 5:
Mégevand, B., Klay, A., Gnanvossou, D. and Paraiso, G. 1993. Maintenance and mass
rearing of phytoseiid predators of the cassava green mite. Exp. Appl. Acarol. 17:
Messchendorp, L., Gols, G.J.Z. and van Loon, J.J.A. 2000. Behavioural observations of
Pieris brassicae larvae indicate multiple mechanisms of action of analogous drimane
antifeedants. Entomol. Exp. Appl. 95: 217–227.
Meyer, G.A. 1993. A comparison of the impacts of leaf-feeding and sap-feeding insects on
growth and allocation of goldenrod. Ecology 74: 1101–1116.
Montserrat, M. 2001. Estudi de la relació depredador-presa en alguns heteròpters presents en
el cultiu de cogombre. Phd Thesis, University of Lleida.
Noronha, A.C. and Silva, E.S. 1998. Flutuação populacional de Typhlodromalus aripo
(De Leon) (Acari: Phytoseiidae). EMBRAPA 64: 1–2.
Onzo, A., Hanna, R., Zannou, I., Sabelis, M.W. and Yaninek, S.J. 2002. Dynamics of refuge
use: diurnal, vertical migration by predatory and herbivorous mites within cassava plants.
Oikos (in press).
Ostrom, P.H., Colunga-Garcia, M. and Gage, S.H. 1997. Establishing patterns of energy ?ow
for insect predators using stable isotope ratios: ?eld and laboratory evidence. Oecologia
Overmeer, W.P.J. 1985. Alternative prey and other food sources. In: Spider Mites – Their
Biology, Natural Enemies and Control, W. Helle and M.W. Sabelis (eds), pp. 131–139.
Pedrosa-Macedo, J.H. 2000. Biology and behavior of the strawberry guava saw?y, Hap-
lostegus epimelas Konow 1901 (Hymenoptera: Pergidae), in southern Brazil. Proc.
Entomol. Soc. Wash. 102: 129–134.