Arrestment Responses of Eretmocerus Species and

Journal of lnsect Behavior, Vol. 5, No. 4, 1992 Arrestment Responses of Eretmocerus Species and Encarsia deserti (Hymenoptera: Aphelinidae)to Bemisia tabaci Honeydew Orit Shimron, 1 Abraham Hefetz, 1'2 and Dan Gerling 1 Accepted July 30, 1991; revised November 21, 1991 The responses of Eretmocerus sp. and Encarsia deserti (Gerling & Rivnay), parasitoids of Bemisia tabaci (Gennadius) to host secretions were investigated. Upon contact with honeydew excreted from the host, the parasitoids' walking speed decreased, whereas the rate of angular turning increased (the walking speed of Eretmocems sp. and Encarsia deserti decreased from 2.0 to O. 8 and from O. 8 to 0.3 ram~s, respectively, and the respective angular turning rates increased from 123.3 to 489.6 and from 267.9 to 979.2 deg/cm). Moreover, the wasps generally turned back toward the secretion as soon as they lost contact with it. This induced arrestment responses resulted in increased searching time by the parasitoids. Laboratory bioassays were conducted in which paper disks in petri dishes containing different amounts of honeydew, equivalent to the secretions of 1, 5, or 50 larvae, were offered to female wasps. The wasps responded by arrestment even to honeydew produced by a single larva. While the response of Eretmocerus sp. was quantity dependent, being more intense at higher amounts of honeydew, that of E. deserti was of equal intensity at all honeydew quantities assayed. KEY WORDS: arrestment; contact chemicals; kairomone; Bemisia tabaci; Eretmocerus sp.; Encarsia deserti; host location; honeydew. INTRODUCTION The process of host location by parasitoids involves the use of different stimuli, commonly including contact chemicals. The reaction of parasitoids when con- Department of Zoology, Tel-Aviv University, Tel-Aviv, Ramat-Aviv 69978, Israel. 2 To whom correspondence should be addressed. 517 0892-7553/92/0700-0517506.50/0 9 1992 Plenum Publishing Corporation

518 Shimron, Hefetz, and Gerling tacting host chemicals can be defined as arrestment, provided that they congregate as a result of an undirected kinetic reaction, either by slowing or stopping their locomotion (orthokinesis) or by changing the rate of turning (klinotaxis) (Kennedy, 1978; Waage, 1978; van Alphen and Vet, 1986). The overall effect of the arrestment responses is that the parasitoids spend more time searching in areas that are contaminated with host secretions, resulting eventually in host contact and parasitization. The chemicals which induce arrestment responses may originate from different host-associated sources such as host frass in Orgilus lepidus (Muesebeck) (Hendry et al., 1973), Microplitis croceipes (Cresson) (Lewis and Jones, 1971), and Rhyssapersuasoria (L.) (Spradberry, 1970), moth scales in Trichogramma evanescens (Westwood) (Lewis et al., 1972), or host salivary gland secretion in Cardiochiles nigriceps (Viereck) (Vinson and Lewis, 1965). Host honeydew was found to be the origin of the arrestment kairomones for Microterys flavus (Howard) (Vinson et al., 1978), Aphidius nigriceps (Ashmead) (Bouchard and Cloutier, 1984), and Encarsia formosa (Gahan) (van Vianen and van de Viere, 1988). Eretmocerus sp. and Encarsia deserti (Gerling & Rivnay) are two species of parasitic wasps which parasitize immature instars of the sweetpotato whitefly, Bemisia tabaci (Gennadius). [They were introduced into Israel in order to assist the native species Eretmocerus mundus (Mercet) and Encarsia lutea (Masi) in controlling the whitefly population that has become a major pest of cotton in the last 15 years (Geding et al., 1980).] In this study we characterize the arrestment responses of Eretmocerus sp. and E. deserti toward secretions of their host B. tabaci. MATERIALS AND METHODS Bemisia tabaci (Gennadius) were reared on young cotton plants in an incubator at 28~ and 60% relative humidity, under a 10:14 (L:D) photoperiod. Second- and third-instar host larvae were subjected to parasitization by Eretmocerus sp. or E. deserti in a separate incubator under the same conditions. Female parasitoids taken for experimentation were 24-48 h old and were kept for 24 h prior to each test in a vial containing honey solution. Since the females had emerged in an environment containing hosts and their products, they were considered experienced. Each female was tested only once. Behavioral characterization of the parasitoids' response to contact chemicals was carried out in a petri dish lined with green paper (simulating a green leaf), in the center of which a single green paper disk (1.5-cm diameter), treated with the assay material (honeydew or blank control), was placed. In each test, one wasp was released at the edge of the petri dish and its movement was videotaped for 5 min. The petri dish was held upside down in order to mimic the underside of a leaf. These videotapes were then used for the analysis of the

Arrestment of Parasitic Wasps 519 wasps' behavior. The route taken by each wasp was traced directly from the television screen onto a transparency, from which the travel distances and the degree of turning were later measured. Two parameters were used in the analysis of movement: (a) average speed of walking (mm/s) and (b) rate of angular turning expressed as the sum of turning angles per total distance traveled (deg/ cm). Within each test the movement of the wasps before reaching the honeydewcovered paper disk was compared to their movement while on the disk, whereas between tests the behavior of the wasps before reaching the treated disk was compared to their behavior in a petri dish with a clean disk. Choice experiments were also conducted in petri dishes. Four equal, green paper disks (1.5-cm diameter), one of which was impregnated with a test material, were glued to the green lining. A petri dish containing four untreated disks served as a control. In each experiment, one female wasp was introduced into the petri dish and was observed for 20 min (from the moment she started walking). The time spent by the wasp on each of the paper disks was recorded. For the qualitative assays, honeydew was collected on paper disks from more than 100 B. tabaci larvae for 2-3 days. The disks were used later in the bioassay. To study the effects of increasing amounts of honeydew on the intensity of the arrestment responses, the test disks were placed under different numbers of B. tabaci larvae (1, 5, or 50 larvae) for 24 h. The time spent by each wasp on each of the paper disks in the petri dish and the number of returns to the control vs test disks after leaving them while searching were used to assess the magnitude of the response. For the qualitative and quantitative tests, significance was determined using nonparametric tests, Mann-Whitney or Kruskal-Wallis, as needed. The differences between numbers of returns to the control vs test disks were tested using a chi-square test of independence. RESULTS General observations of the behavior of the wasps in the vicinity of whitefly honeydew indicated that the latter induced an arrestment response. In the absence of stimuli from contact chemicals, i.e., in the control experiments, both Eretmocerus sp. and E. deserti usually traveled over most of the surface of the dish, walking relatively fast (average speed of 1.7 and 0.7 mm/s, respectively) with few turnings (average angle of 139.3 and 199.1 deg/cm, respectively) (Table I and Fig. 1). A similar searching pattern and speed of walking were also observed in petri dishes containing disks with honeydew, before contact with the contaminated area had been made. Upon contact with the secretion, walking speed decreased drastically, averaging 0.8 mm/s for Eretmocerus sp. and 0.3 mm/s for E. deserti (Table I), whereas the rate of angular turning increased to 489.6 and 979.9 deg/cm, respectively (Table I). The wasps generally turned back toward the secretion when they lost contact with it, resulting in an intensified

520 Shimron, Hefetz, and Gerling Table I. Walking Speed (mm/s) and Turning Rates (deg/cm) of Eretmocerus sp. and Encarsia deser~i With and Without the Presence of Bemisia tabaci Honeydew ~ Before contact with the In the contaminated Species Control contaminated area area Average turning rates (mean + SE), n = 6 Eretmocerus sp. 139.3 + 13.7 a 123.3 _ 12.5 a 489.6 + 70.6 b Encarsia deserti 199.1 + 25.4 a 267.9 + 65.3 a 979.9 _ 14.5 b Average walking speed (mean + SE), n = 6 Eretmocerus sp. 1.7 -t- 0.3 a 2.0 _ 0.4 a 0.8 + 0.2 b Encarsia deserti 0.7 + 0.1 a 0.8 + 0.1 a 0.3 +_ 0.05 b *Data within rows that are followed by a different letter are significantly different (P < 0.05) by Mann-Whitney U test. Test search that concentrated in the area in which whitefly secretion existed (Table I1). The duration of the arrestment response was examined in the choice experiment. Both species responded positively to contact with paper disks impregnated with B. tabaci honeydew by spending more time on them than on the control disks (Table III). Eretmocerus sp. averaged 343.5 + 44.6 s on the treated disks and 44.1 _ 7.8 s on the control disks; Encarsia spent 329.3 + 65.5 s on the treated disks and 50.3 + 9.1 s on the control disks. Both species responded even to the relatively small amounts of honeydew excreted by 1 or 5 larvae for 24 h (Table III). However, the magnitude of their arrestment response to different honeydew amounts differed. The response of Eretmocerus sp. to the equivalent of secretion by 50 larvae was about threefold higher than the response to either 1 or 5 larvae (Table III). In contrast, the response of E. deserti was already high when exposed to secretions of one larva and did not change significantly when exposed to higher doses of honeydew. DISCUSSION Reproductive success as a component of fitness in parasitoids is dependent, among other factors, upon the number of suitable hosts found. Searching efficiency for hosts, therefore, is likely to be under strong natural selection. As a consequence, many species have evolved mechanisms that enable them to detect and orient toward their hosts from a distance. Among the means utilized by these parasitoids for orientation to their host is the perception of their chemical products. Although kairomones are not the sole factors involved in host location,

A2 ~- Y Start Fig. 1. Examples of typical walking patterns of Eretmocerus sp. (A1 and A2) and Encarsia deserti (B1 and B2) in petri dishes containing (1) an untreated paper disk vs (2) a paper disk impregnated with B. tabaci honeydew.

B 1 B2 Start Fig. 1. Continued

Arrestment of Parasitic Wasps 523 Table II. Cumulative Number of Returns (to the Same Disk that Was Last Visited) Made by Eretmocerus sp. and Encarsia deserti to the Contaminated Disks vs the Control Disks in a Four-Choice Experiment Returns Species Control (avg.) ~ Test n pb Eretmocerus sp. 62 174 11 < 0.001 Encarsia desert 45.5 108 13 < 0.001 "The control values are averages of the three clean disks. hthe differences between number of returns to the control disks and number of returns to the test disks were tested using a chi-square test of independence. Table IlL Average Time Spent by Eretmocerus sp. and Encarsia desert on Paper Disks Smeared with Whitefly Honeydew, or on Clean Control Disks, in a Petri Dish During 20 rain of Bioassay Treatment of the test disk" n Control Test (s SE) (s SE)/' P" Eretmocerus sp. Pool 15 343.5 + 44.6 44.1 + 7.8 <0.001 Blank 9 45.1 + 13.9 51.1 + 17.2 ns 1 larva 14 137.9 + 27.5 70.3 + 7.24 ns 5 larvae 18 159.5 + 24.1 61.1 5.42 <0.001 50 larvae 15 454.1 46.1 34.0 + 5.8 <0.001 Encarsia desert Pool 7 329.3 65.5 50.3 9.1 <0.005 Blank 11 80.1 13.5 95.1 19.2 ns 1 larva 12 394.1 53.6 71.4 10.0 <0.001 5 larvae 12 363.4 36.0 93.0 15.2 <0.001 50 larvae 12 454.6 + 45.1 72.7 14.4 <0.001 Treatments of the paper disks were as follows: pool--honeydew secreted by more than 50 B. tabaci larvae for 2-3 days; 1, 5, and 50 larvae--secretions from 1, 5, and 50 B. tabaci larvae, respectively, placed above the test paper disk for 24 h; blank--all four disks in the petri dish clean. ~'Control--average time spent on the three clean disks in the same petri dish that contained the test disk. 'Mann-Whitney U test was used. many studies have demonstrated that they are important cues for host location by parasitoids (Weseloh, 1981, and references therein). Our results with Eretmocerus sp. and E. desert demonstrated a kairomonal effect of host honeydew. This was expressed as an arrestment response involving the intensification of searching behavior in areas which contain hosts rather than a directed movement toward the stimuli source. Several other parasitoid species were reported to respond to the honeydew of their hosts by intensified search. Aphid nigriceps

524 Shimron, Hefetz, and Gerling (Ashmead) spent more time searching on potato plants previously infested with aphids than on fresh plants and exhibited arrestment when crossing a honeydewtreated area (Bouchard and Cloutier 1984). Microterysflavus (Howard) responded to honeydew of its host Coccus hesperidum (L.) with stimulated search (Vinson et al., 1978), and Encarsiaformosa (Gahan) was shown to stay longer on leaves that were artificially smeared with whitefly (Trialeurodes vaporariorum) honeydew as long as they contacted the secretions (van Vianen and van de Viere, 1988). Lewis (1976) suggested that it may be possible to increase the efficiency of parasitoids through manipulation of their host location behavior. Indeed, it has been shown in other parasitoids that contact chemicals which were artificially applied caused increased levels of parasitization in the field by localizing the parasitoid foraging in selected areas (Lewis et al., 1972; Gross et al., 1975). Similarly, it might be possible, through artificial application of honeydew or its components, to cause Eretmocerus females to search more intensively in patches that otherwise would be left unparasitized. The application of such a program would be feasible only after further study of the parasitoid's behavior in the field and its response to different methods of kairomone application. The "all-or-none" response of Encarsia deserti as opposed to the "quantity-dependent" response of Eretmocerus sp. probably reflects the capability of the latter to spend more time in areas that contain higher host densities, where they have a better chance of finding a suitable host for parasitization. E. deserti, on the other hand, presumably spends equal time in all patches irrespective of the number of B. tabaci larvae present. In this respect, Eretmocerus sp. seems more efficient than E. deserti, as it has a density-dependent response to hostderived kairomone. This was also shown for Asobara tabida (Nees) (Galis and van Alphen, 1981), Nemeritis canescens (Gravenhorst) (Waage, 1978), and Aphidius nigriceps (Noldus et al., 1988). In general, kairomones elicit host location by parasitoids through two mechanisms: long-range chemoreception, which enables location of a host environment, and short-range chemoreception, which focuses search over small areas and which depends on contact with a substrate on which a chemical had been placed. This search eventually results in host contact (Weseloh, 1981). Although it is difficult to differentiate between these two kinds of perceptions, it is clear that in the cases of E. deserti and Eretmocerus sp., close-range chemoreception is utilized. The reaction to the stimuli (i.e., honeydew) was seen only after direct contact with the kairomone. In such cases, the need for temporal coincidence between the host and its product is clear: otherwise, the parasitoids will be arrested in areas devoid of hosts, keeping them from searching other potential host areas. Arrestment by honeydew, which is usually indicative of host presence, furnishes such temporal coincidence. This increases the efficiency of host location, placing the parasitoid in a better position for reproductive success. It

Arrestment of Parasitic Wasps 525 should be taken into consideration, however, that host location is only one of the factors that affect the parasitoid efficiency. In order to assess the overall efficiency of the parasitization process, it is necessary to determine the influence of other factors, such as patch time allocation and clutch size, among others. ACKNOWLEDGMENTS We wish to thank Dr. Gerrit Pak for analyzing some of the walking patterns. Parts of this study were supported by grants from the "Ben-Gurion Foundation" (O. Shimron) and the Israeli Cotton Council (A. Hefetz). REFERENCES Bouchard, Y., and Cloutier, C. (1984). Honeydew as a source of host-searching kairomones for the aphid parasitoid Aphidius nigriceps (Hymenoptera: Aphidiidae). Can. J. Zool. 62: 1513-1520. Galls, F., and van Alphen, J. J. M. (1981). Patch time allocation and search intensity of Asobara tabida Nees (Braconidae), a larval parasitoid of Drosophila. Neth. J. Zool. 31:596-611. Geding, D., Motro, U., and Horowitz, R. (1980). Dynamics of Bemisia tabaci Gennadius (Homoptera: Aleyrodidae) attacking cotton in the costal plain of Israel. Bull. Entomol. Res. 70: 213-219. Gross, H. R., Lewis, W. J., and Jones, R. L. (1975). Kairomones and their use in the management of entomophagous insects, iii. Stimulation of Trichogramma achaeae, T. pretiosum and Microplitis croceipes with host seeking stimuli at time of release to improve their efficiency. J. Chem. Ecol. 1: 431-438. Hendry, L. B., Greany, P. D., and Gill, R. J. (1973). Kairomone mediated host-finding behavior in the parasitic wasp, Orgilus lepidus. EntomoL Exp. Appl. 16: 471-477. Kennedy, J. S. (1978). The concepts of olfactory "arrestment" and "attraction." Physiol. Entomol. 3: 91-98. Lewis, W. J., and Jones, R. L. (1971). Substance that stimulates host-seeking by Microplitis croceipes (Hymenoptera: Braconidae), a parasite of Heliothis species. Ann. Entomol. Soc. Am. 64: 471-473. Lewis, W. J., Jones, R. L., and Sparks, A. N. (1972). A host seeking stimulant for the egg parasite Trichogramma evanescens, its source and a demonstration of its laboratory and field activity. Ann. Entomol. Soc. Am. 65: 1087-1089. Lewis, W. J., Jones, R. L., Gross, H. R., and Nordlund, D. A. (1976). The role of kairomones and other behavioral chemicals in host finding by parasitic insects. Behav. Biol. 16: 267-289. Noldus, L. P. J. J., Lewis, W. J., Tumlinson, J. H., and van Lenteren, J. C. (1988). Olfactometer and wind tunnel experiments on the role of sex pheromones of noctuid moths in the foraging behavior of Trichogramma spp. In Voegele, J., Waage, J. K., and van Lenteren, J. C. (Eds), Trichogramma and Other Egg Parasites, Les Colloques de L'INRA No. 43, Paris, pp. 223-238. Spradberry, J. P. (1970). Host finding by Rhyssa persuasoria (L.) an ichneumonid parasite of siricid woodwasp. Anim. Behav. 18: 103-114. van Alphen, J. J. M., and Vet, L. E. M. (1986). An evolutionary approach to host finding and selection. In Waage, J. K., and Greathead, D. J. (eds.), Insect Parasitoids, Academic Press, London, pp. 23-61. van Vianen, A., and van de Viere, M. (1988). Honeydew of the greenhouse whitefly, Trialeurodes vaporariorum (Westwood), as a contact kairomone for its parasite Encarsia formosa Gahan. Med. Fac. Landbouww. Rijksuniv. Gent. 53: 949-954. Vinson, S. B., and Lewis, W. J. (1965) A method of host selection by Cardiochiles nigriceps. J. Econ. Entomol. 58: 869-871.

526 Shimron, Hefetz, and Gerling Vinson, S. B., Harlan, D. P., and Hart, W. G. (1978). Response of the parasitoid Microterysflavus to the brown soft scale and its honeydew. Environ. Entomol. 7: 874-878. Waage, J. K. (1978). Arrestment responses of the parasitoid, Nemeritis canescens, to a contact chemical produced by its host, Plodia interpunctella. Physiol. Entomol. 3: 135-146. Weseloh, R. M. (1981). Host location by parasitoids. In Nordlund, D. A., Jones, R. L., and Lewis, W. J. (eds.), Semiochemicals: Their Role in Pest Control, Wiley, New York, pp. 79-95.