Correlations Among AC Electronic Monitoring Waveforms ... - BioOne

PEST MANAGEMENT AND SAMPLING

Correlations Among AC Electronic Monitoring Waveforms, Body Postures, and Stylet Penetration Behaviors of Lygus hesperus (Hemiptera: Miridae) ANDREW R. CLINE1

AND

ELAINE A. BACKUS2

Environ. Entomol. 31(3): 538Ð549 (2002)

ABSTRACT A behavioral ethogram for feeding and nonfeeding behaviors of Lygus hesperus Knight, the western tarnished plant bug, was developed using third instars on host and nonhost plants, as well as an artiÞcial diet. Individual insects were simultaneously recorded with an AC electronic monitoring system and a time-lapse video cassette recorder. The electronic monitor provided information on the various stylet penetration behaviors exhibited, whereas the video was used in evaluating nonfeeding and surface exploration behaviors on plants and stylet activities in the diet. From these recordings, seven distinct feeding waveforms and over 20 nonfeeding behaviors were described. ArtiÞcial diet correlations were performed to verify ingestion and possible salivation activities during other waveforms. Some behaviors were found only to occur on particular substrates. These included the release of a salivary bubble following a period of feeding on the artiÞcial diet and a continual labial dabbing behavior on mature cotton plants. An association was observed between ingestion and speciÞc positions of the labium and antennae during the construction of an ethogram. A subsequent experiment was conducted that conÞrmed this relationship on mature and immature cotton. KEY WORDS Lygus hesperus, behavioral ethogram, probing behavior, herbivory, electropenetration graph, plant bugs

MANY MIRIDS, ESPECIALLY those in the genus Lygus, are agricultural pests of many economically important crops including cotton Gossypium hirsutum L. (Sorenson 1936, Baker et al. 1946, Allen and Goede 1963, Broersma and Luckman 1970). Due to recent successes in managing other injurious cotton insects (e.g., the Boll Weevil Eradication Project and Bt cotton for heliothine lepidopterans), Lygus bugs have become more important pests (Tingey et al. 1975). The two plant bugs reported to be most damaging to upland cotton in the United States are Lygus lineolaris (Palisot de Beauvois) and Lygus hesperus Knight. The former is restricted to the eastern two-thirds of the United States and Canada, whereas the latter is predominantly a western pest. Both species are polyphagous; L. hesperus is known to feed from plants belonging to more than 35 families (Schwartz and Foottit 1998). All phytophagous hemipterans, including mirids, use piercing-sucking mouthparts to imbibe plant ßuids. The severity of plant bug injury to cotton is due to the preferred feeding locations of these insects as well as the cellular damage they impose. Both Lygus pest species prefer to feed from the main stem of mature plants, on fruiting bodies, and on developing terminals (Mauney and Henneberry 1979, Snodgrass 1998), typ1

E-mail: [email protected]. Department of Entomology, University of Missouri-Columbia, 1Ð 87 Agriculture Building, Columbia, MO 65211. 2

ically on the Þfth through seventh nodes of the plant (Wilson et al. 1984). Nymphs have a preference for developing squares, whereas adults prefer the vegetative structures (Snodgrass 1998). Lygus bugs impose signiÞcant damage apparently through both mechanical and chemical alterations caused by the combined action of their stylets and saliva. Strong (1970) proposed that Lygus injury to cotton involves both the mechanical breakdown of vegetative and reproductive tissues, and the disruption of hormone ßow from cells producing them. The injury caused from these species results in both lowered yield (i.e., via abscission of fruiting bodies) and reduced lint quality (i.e., mechanical puncturing of developing buds causing discoloration) (Parencia 1978). This damage to cotton alone results in annual losses greater than $75 million dollars (Anonymous 1999). Although there is some knowledge of the physiological responses of plants to feeding by L. hesperus (Strong 1970, Tingey and Pillemer 1977), much less is known of the precise feeding behaviors employed by these bugs. The most reliable tool used to elucidate such behavior is electronic monitoring of insect feeding (McLean and Kinsey 1964), also known as AC electropenetration graph (EPG) (Walker 2000). The system is based on the physical principle that as ßuids of varying electrical conductivity (e.g., saliva, plant cellular matter) pass through the insectÕs mouthparts, a measurable change in resistance is created. These

0046-225X/02/0538Ð0549$02.00/0 䉷 2002 Entomological Society of America

June 2002

CLINE AND BACKUS: STYLET PENETRATION BEHAVIORS OF L. hesperus

changes in resistance can then be attributed to different behavioral processes such as salivation and ingestion (McLean and Kinsey 1967, 1968). An EPG yields electrical waveforms that can be correlated with stylet movements and salivation events within plant tissues (e.g., Kabrick and Backus 1990). In the 37 yr since its development, the EPG has mostly been used to study homopterans. This work has emphasized vascular cell-ingesters such as aphids (Tarn and Adams 1982, Cole et al. 1993, Prado and Tjallingii 1994, Harrewijn and Kayser 1997, and Ramirez and Niemeyer 1999) and other sheath-feeding leafhoppers (Chang 1978, Port 1978, Raman et al. 1979, Khan and Saxena 1988, and Wayadande and Backus 1989). The only lacerate-and-ßush feeders studied extensively to date are Empoasca spp. leafhoppers (Backus and Hunter 1989, Hunter and Backus 1989, Kabrick and Backus 1990, Calderon and Backus 1992). The heteropteran work has focused primarily on members of the Suborder Pentatomamorpha, Coreidae [Anasa tristis (DeGeer)] (Bonjour et al. 1991, Cook and Neal 1999). The objectives of this study were to provide the Þrst documentation of Lygus (in fact, any mirid) feeding waveforms, to perform artiÞcial diet correlations to assess stylet activities and other biological meanings of each waveform, such as ingestion, laceration or salivation, and to determine whether any speciÞc body postures were correlated with particular stylet activities identiÞable by waveforms. This information will provide important background information for further studies on the quantiÞcation of L. hesperus feeding. Materials and Methods Plant Rearing. ÔCoker 312Õ (a susceptible cultivar of cotton) was seeded, Þve per pot, into 15-cm plastic pots containing a mixture of Pro Mix and 12.0 g Osmocote (Scotts-Sierra Horticultural Products, Marysville, OH) 14-14-14 fertilizer. Plants were grown in growth chambers under 32:26⬚C and a photoperiod of 16:8 (L:D) h and watered ad libitum every day between 0900 and 1100 hours (Purseglove 1968). When the Þrst true leaves appeared, two seedlings were randomly chosen and removed from the pots to reduce the water stress and competition for light among the remaining plants. Plants were reared for six more weeks, or until development of squares. One of the three remaining plants in a pot was chosen for each monitoring session; the other two plants were discarded. Choice of test plant depended on three criteria: (1) overall health of the plant, (2) the number of squares present, and (3) the location of the squares on the sympodial branches, i.e., to facilitate minimal movement of the plant during video observations. Insects were placed on one side of the test square in mature cotton or on the adaxial leaf surface on cotton seedlings to allow maximum viewing of the insect. Both okra (Hibiscus esculentus L.) and caulißower (Brassica oleracea L.) were grown under the same protocol as cotton. The experimental procedures on okra and caulißower were performed on the adaxial

539

leaf surfaces of seedlings as in the cotton seedlings. Upon completion of a monitoring session, the plant was discarded. Insect Rearing. A production colony of L. hesperus was maintained in a windowless, quarantine rearing room under ⬇27⬚C, 70% RH, and a photoperiod of 16:8 (L:D) h. Eggs were originally acquired from Biotactics (Riverside, CA) (APHIS import permit no. 33219) in 1997, but in 1998 the colony became reproductively self-sufÞcient. Insects were housed in plastic rearing containers with an organza mesh cloth placed between the lid and container. The insects were fed a complex commercial diet contained within a plastic/ paraÞlm packet. Commercial diet packets were supplied weekly by Biotactics; diet composition was based on those Þrst described by Debolt (1982) and Patana (1982). A packet was placed paraÞlm-side down on the organza mesh top through which the insects pierced the artiÞcial diet with both their mouthparts and ovipositors. The diet packets served as both a feeding substrate and as an ovipositional site. Diet packs and organza mesh were changed every 1Ð2 d for adults, and every 3Ð5 d for nymphs. After 2 d, the adults typically oviposited the maximum complement of eggs into the diet packets. Each packet containing eggs was then placed into a new container with four to Þve crumpled paper towels. New nymphs were typically produced within 6 Ð 8 d. Approximately 100 nymphs were produced from each egg pack. To reduce fungal contamination, old diet packets were discarded from the containers when secondinstar nymphs were visible. Third-instar nymphs (distinguishable by a black spot on the dorsal abdomen and reduced wing pads) were used in all experiments. Simple Artificial Diet Preparation. A 100-ml aliquot of distilled water was autoclaved for 20 min. While the water was continuously stirring on a hot plate in a laminar ßow hood, 12.00 g of sucrose was added and allowed to solubilize (⬇2Ð3 min). After the solution cleared, 8.00 g of SeaPrep (FMC BioProducts, Rockland, ME) agarose was slowly added to the stirring solution to make a 12% (wt:vol) sucrose and 8% (wt: vol) agarose mixture. Particles of green Chinese stick ink were obtained by sanding a small portion of stick ink with Þne-grade sandpaper. Ink powder was suspended in distilled water, and a dilute slurry of the smallest particles was pipetted from the surface after several minutes of particle settling for a Þnal concentration of 5% (wt:vol). Equal volumes of the ink and sucrose/agarose solutions were then combined in a container on the stirring hot plate. The Þnal artiÞcial diet contained 6% sucrose, 4% agarose, and 2.5% ink particles (wt:vol). While still warm, the diet was poured onto an eight-compartment tissue culture chamber slide (Habibi et al. 1993). A slit was cut into one side of the plastic gasket for insertion of a copper wire, which provided substrate voltage to the diet from the electronic monitor. Approximately 3Ð5 ml of diet was used on each slide to a depth of 1Ð2 mm. Before the diet cooled, paraÞlm was stretched over the top. The slides were kept in a refrigerator at ⬇10⬚C until needed. Before use, the paraÞlm of each slide was

540

ENVIRONMENTAL ENTOMOLOGY

brushed with a 1% sucrose solution to stimulate stylet penetration by the nymphs. EPG/Electronic Monitoring. L. hesperus nymphs were electronically monitored using one channel of a four-channel version of the “Missouri Monitor” (Backus and Bennett 1992; type 2.2 E.A.B., unpublished data). With the aid of a stereomicrosope, each insect was attached to a 12.7-␮m gold wire (Sigmund Cohn, Mt. Vernon, NY) with silver print paint (n-butyl alcohol solvent, Ladd Research Industries, Burlington, VT) at the dorsal junction between the abdomen and thorax. The tether was 4 Ð5 mm long, allowing the insect considerable range of movement. After tethering, each naive, diet-reared insect was acclimated on a plant for 1 h, then starved by dangling for 1 h. On all plant material, the electrical signal applied to the substrate was 75 mV at 500 Hz. All artiÞcial diet work was performed with a 225 mV signal at 500 Hz, due to the less conductive nature of the nonproteinaceous substrate. All recordings were made within a copper wire mesh Faraday cage to reduce the amount of ambient electronic noise entering the system. The substrate voltages for each plant and artiÞcial diet were determined through a trial-and-error process during which the nymphs were exposed to various voltages and frequencies. Observations were made to determine which voltage yielded the most probing (both frequency and duration) with the highest resolution waveforms. These voltages also had to prevent irregular tarsal movements or prevent insects from walking to the tip of the plant surface (indicating the surface was too “hot”). The threshold voltage was determined to be around 125 mV, because at any higher voltage the insect did not probe. Windaq (Dataq Instruments, Akron, OH) analogto-digital analysis hardware and software was used for real-time display of waveform output. The Windaq settings for all experiments were the same so that standardization and correlation of waveform appearances could be possible among varying substrates. The compression was held constant at a value of 5, the seconds per division at 4, and the sample rate at 25 per second. Videomicrography. As each insect was being electronically monitored for probing behaviors, a simultaneous video recording was made of the external movements of the insect using a Javelin Chromochip II (model JE3462HR) high-resolution color video camera (Javelin Systems, Torrance, CA) attached to a Wild M-5 Apo stereomicroscope (Wild Instruments, Basel, Switzerland). The camera/microscope apparatus was connected to a Panasonic AG-6740 time lapse video cassette recorder (VCR) connected to a Panasonic Series 1381 color video monitor (Will Electronics, St. Louis, MO). Individual insects were recorded on a 6-h setting to maximize the number of frames (12) per second the VCR would record. Further analysis was done on a frame-by-frame basis. Most video recordings were at 12⫻ magniÞcation; however, when the insect was motionless, 25⫻ magniÞcation was used. All artiÞcial diet recordings were made at 25 and 50⫻ magniÞcations to provide better resolution of the

Vol. 31, no. 3

insectÕs stylets; this higher magniÞcation was possible due to the transparent nature of the diet. Ethogram Compilation. Fourteen third instars were individually electronically monitored and simultaneously videotaped on one of three different host plants: cotton (mature and immature plants), okra, and caulißower. Cotton and okra belong to the family Malvaceae (one of the preferred host families for Lygus bugs), and caulißower to Brassicaceae; caulißower served as a nonhost plant comparison. All insects were given access to the adaxial leaf surface of each plant, except on mature cotton, on which they were placed on developing squares. Immature cotton plants were deÞned as those having only the Þrst true leaves (⬇2-wk-old plants). Mature cotton was used at the 6-wk stage of development when squares were evident. Caulißower and okra were both used at 4 wk of development, and were considered immature due to the absence of fruiting bodies. Each insect was recorded for between 30 min and 4 h, from 1000 to 1400 hours. A total of ⬇9.4 h of recordings was made. Recordings were terminated when the insect had not probed for a minimum of 10 min. The insect and plant were then discarded. Two insects were recorded on caulißower, four on okra, four on immature cotton, and four on mature cotton. One or two insects were recorded per day. Observations were used only to assemble ethograms; behaviors were quantiÞed by total waveform duration but were not statistically tested due to the small sample size of insects for each plant type. Artificial Diet Correlations of Waveforms. Twelve third instars were electronically monitored and simultaneously videotaped while feeding on 12 different artiÞcial diets. Stylets were observed at 50⫻ under the videomicrography set-up for 12Ð72 min, being sure the stylet tips were in focus at all times. Notes were made on which waveforms corresponded to various stylet movements, as well as whether individual ink particles were ßowing (1) into or (2) out of the insectÕs stylets (representing 1 ] likely ingestion or 2] salivation or ßuid egestion, respectively) or were moved by the stylet tips. One or two insects were recorded per day, between 1000 and 1400 hours each day. Ingestion/Body Posture Correlation. Observations during the experiment for ethogram compilation suggested that there was a correlation between a stereotypical body posture and either C-type (primarily ingestion) waveforms (see waveform descriptions, below). Therefore, a separate experiment using mature and immature insects was designed to test this hypothesis. Eight insects were electronically monitored and videotaped for varying periods (1Ð2 h), based on their relative activity and performance of C-type ingestion waveform (see Results for description of waveforms), on both immature and mature cotton. No insect was used on a previously fed-upon plant, and each insect was recorded only once on one individual plant. As ingestion behaviors were exhibited by an insect, the simultaneous video recording was analyzed to determine characteristic body postures occurring during these ingestion events. Body postures were

June 2002 Table 1.


541

Video ethogram of Lygus hesperus on cotton and other host plants

Symbol

Behavior

␥ (C1 or C2) ␤ (B) ␣ (A) ␦ (D) ␧ (E) ⍀1 (Z) ⍀2 (Y) ⍀3 ⍀4 ⍀5R, L ⍀6 ⍀7L, R ⍀8L, R ⍀9 ⍀10 ⍀11 ⍀12 ⍀13 ⍀14 ⍀15

Ingestion behaviorÑmouthparts remain in a Þxed position; labial angles remain constant Drilling behaviorÑstylets are inserted into and out of the substrate; labial angles change Test probeÑa rather short duration penetration event with only one insertion Excretory dropletÑtip of abdomen touches substrate with coincident release of excreta Salivary bubbleÑas stylets are withdrawn from artiÞcial diet, a saliva droplet is left on substrate WalkingÑany directional movement in which the insect progresses towards another area Standing/RestingÑno movement in any direction Body reorientationÑslight movement that results in insect changing body orientation Substrate antennationÑantennal segments are brushed against substrate surface Antennae grooming with foreleg tarsiÑforeleg tarsi brush the third and fourth antennal segments Foreleg tarsal groomingÑtarsi are rubbed together Left or right foreleg/Midleg groomingÑtarsi from both legs on same side are rubbed together Left or right midleg/Hindleg groomingÑtarsi from both legs on same side are rubbed together Hindleg tarsal groomingÑhindleg tarsi are rubbed together or against abdomen Labial grooming with foreleg tarsiÑtarsi are swept over the terminal three labial segments Labial dabbing of substrateÑlabium is extended under head and brießy touches substrate Labial reverberations against abdomenÑlabium continuously touches the venter of the abdomen Tarsal twitchingÑany tarsi which is continuously in motion independent of the other tarsi Labial draggingÑlabium is situated beneath abdomen but touches the substrat Insect out of Þeld of view or adjustedÑinsect has moved out of video Þeld and/or is adjusted with brush

Letters in parentheses indicate the EPG ethogram waveforms associated with videotaped behaviors.

measured using a protractor and several camera angles to determine the positions of the labium and antennae. The experiment produced a total video/EPG duration of ⬇6.5 h containing 41 C-type waveform events. This sample size was similar to a previous electronic monitoring study (Ecale and Backus 1994) describing a postural correlation for Empoasca fabae (Harris), and was considered sufÞcient for the current study. All ingestion events used in this study had corresponding video recordings in which the labium was in clear view. Measurement and Statistical Analysis. Waveforms were displayed postacquisition by Windaq Playback software so that waveform types could be determined and durations measured, then input by hand or via a modiÞed data input program (Van Giessen and Jackson 1998) into Excel. VCR timing was converted to event durations for each observed behavior and input manually into Excel. Mean durations between electronic monitoring and video recordings for the ingestion correlation experiment were tested using a paired t-test via SAS (SAS Institute 1985) to determine if a signiÞcant difference existed. Results Video Ethogram. The L. hesperus video ethogram comprised 23 distinct behaviors (Table 1), including those for both feeding and nonfeeding. The majority of videotaped behaviors exempliÞed nonfeeding activities that involved grooming or cleaning. This phenomenon illustrated the use of a multitude of sensory systems on the labium, antennae, and tarsi to distinguish the acceptability of potential food sources. Video recordings allowed direct visualization of all movements produced by the nymphs on the leaf or square surface, the one aspect that was lacking from electronic monitoring recordings. The electronic monitoring ethogram, however, provided detailed in-

formation about the feeding behaviors occurring internally within the plant, often not visible with the video. In total, nine behaviors (waveforms) were derived through electronic monitoring (Table 2). Seven of the nine behaviors were either directly (A, B, C1, C2, E, and F) or indirectly (D) involved with feeding. The other two waveforms (Z and Y) were associated with either resting or active body movements between feeding intervals. The electronic monitoring data also showed preliminary trends in how L. hesperus alters its feeding parameters based on the type of substrate that was being fed upon. The feeding activity on caulißower indicated that, with only one stylet penetration event, this plant is a nonpreferred food source. The other plants were readily fed upon, each with ⬎100 probing events, and all probes contained ingestion behaviors. Results from the ethogram compilation experiment and the artiÞcial diet correlation experiment were combined to develop the biological meanings, or definitions, of each waveform. ArtiÞcial diet studies allowed visualization of the ßuid ßow dynamics around the inserted stylets, and added additional information necessary for understanding the biological meanings of each waveform. The following sections describe each of the ethogram behaviors. EPG Ethogram: Probing Behaviors. Probing (synonymous with stylet penetration, Backus 2000) behaviors include all activities performed while the insectÕs stylets are inserted within the plant. L. hesperus probing behaviors represented ⬇33% of the total access time the nymphs had on plants. Waveform A. The most frequently occurring feeding waveform, expressed on all plants as well as on simple artiÞcial diet, was the A waveform (termed ÔCÕ waveform in Cline 2000). It lasted 2Ð7 s in a single, short-duration or “test probe.” Waveform A was actually a multicomponent waveform whose subpatterns

542 Table 2.


Vol. 31, no. 3

EPG waveform ethogram of Lygus hesperus on cotton and other host plants Total waveform duration for each windaq behavior

Symbol (video code) C1 (␥) C2 (␥) B (␤) A (␣) D (␦) E (␧)a F (z)b Z2 (⍀2) Z1 (⍀1, 3Ð15)

Windaq waveform/Behavior

Cotton (mature) Cotton (immature)

Flat-line: ingestion Wavy-line: ingestion Jagged-line: drilling/laceration Singular or dual spike: test probe High amplitude squared-off: excretory droplet production Prolonged high amplitude spike: salivary bubble Series of high amplitude short duration spikes: repeted labial insertion Low amplitude ßat baseline: resting/standing Low amplitude variable baseline: any non-resting behavior

Okra

Caulifower

424.8 (1) 923.6 (8) 866.4 (35) 600.3 (146) 1.6 (1)

184.5 (3) 504.4 (5) 5541.4 (164) 997.4 (226) 4.1 (2)

1,800.6 (4) 456.6 (5) 1,040.6 (44) 775.5 (239) 3.7 (2)

0 0 0 1.6 (1) 0

0 0

0 0

0 0

0 0

3,602.8 (227) 2,303.9 (57)

7,497.9 (405) 1,975.8 (21)

2,321.6 (297) 415.9 (16) 934.5 (26) 556 (16)

Greek letters in parentheses represent the video ethogram codes. Data presented are the total number of seconds spent by the entire cohort of insects on that plant (total waveform duration). Numbers in parentheses represent the number of times the waveform occurred (i.e., waveform events). a Occurred only in artiÞcial diet experiments. b Occurred in one insect during subsequent experiment on mature cotton.

occurred in a stereotypical order (Fig. 1). The Þrst subpattern (1) was a single or double spike of about 1Ð2 s in duration, of high relative amplitude. This was followed by a lower-amplitude, ßatter waveform lasting 1Ð 4 s (2), sometimes including one to several small spikes. The last subpattern (3) was a pull-out spike of varying amplitude lasting ⬇1 s. Videotapes on both plants and simple artiÞcial diet showed a highly stereotypical series of stylet activities associated with this waveform. Within a single probe there was one thrust of the stylet bundle (i.e., one “channel cut”), almost always with curvature to the left. To accomplish this, the full stylet bundle was inserted through the paraÞlm and a short distance into the diet. The right maxillary stylet (and perhaps also the right mandibular stylet although resolution was not sufÞcient to determine) then advanced ahead of the left stylet(s), and due to the inherent curvature of the stylets, the extension curves to the left. This exposed a trough made up of the interior of the right food canal; the stylet was held motionless for a second or two before being withdrawn back to the position of the left stylet(s), whereupon the entire stylet bundle was withdrawn. In the semiviscous, artiÞcial diet, ink particles Þrst moved away from and then toward the stylet tips. This supported the logical interpretation of the waveform that subpattern 1 represented salivation of high conductivity (i.e., either large volume rapidly, or highly conductive chemical composition) during the initial thrust of the stylet(s), then subpattern 2 represented a brief interlude of ßuid uptake, followed by a short, rapid burst of saliva concomitant with rapid stylet withdrawal (also possible stylet friction) during subpattern 3 (the pull-out spike). During this process, the insectÕs body remained motionless except for the movement of the stylets, labium and lowering of the head and antennae. Waveform B. The second most common type of waveform was B. In contrast to A, it was the most variable, least stereotypical of the waveforms. It was composed of multiple increases and decreases in am-

plitude, yielding a very jagged waveform (Fig. 2). Relative amplitude could be quite variable from low (⬇40 Ð50% of maximum) to very high (75Ð 85% of maximum), including both extreme dips in amplitude nearly to the baseline, as well as very high pull-out spikes if a probe was ended after a waveform event of B (Fig. 2). Some probes were composed entirely of B; others contained either C1 or C2 (described below) in addition to B. Observations on simple artiÞcial diet showed that the stylets were rapidly and repeatedly protracted into and partially retracted from the substrate. The stylets always began this series of thrusts by curving to the left as in waveform A, then alternately cutting more channels in and out, gradually continuing in a counterclockwise manner until the farthest rightward extension had been reached, whereupon the stylets were removed. This activity was very similar to laceration movements made by stylets of E. fabae (Hunter and Backus 1989). However, unlike in E. fabae, movement was not strictly planar; stylet thrusts occurred radially in any direction away from the stylet insertion point. These thrusts moved ink particles in various directions, with most (75Ð 80%) of the particles being forced away from the stylets. This may have resulted from salivation or mechanical pushing of the particles, and likely both. As in waveform A, the stylets often extended past one another. The body posture of continuous labial bending (making multiple and varied angular inßections and deßections) as well as head bobbing were also correlated with B. Typical stylet insertion depth was shallow to moderate, not unexpectedly given the uniformity of diet composition. However, on plants, stylet depth could be very deep, resulting in occasional double-bends of the labium to allow the farthest extension of the stylet tips. We thus termed this type of behavior “drilling/laceration.” Waveform C1. The least common waveform was C1 (previously termed ÔA1Õ in Cline 2000) or low-amplitude (primarily) ingestion waveform (Fig. 3). C1 was a relatively ßat waveform interspersed with low-am-

June 2002


Fig. 1. Typical Ôtest probeÕ or ÔAÕ waveform. Two distinguishable spikes characterize this A waveform: an insertion spike (1) and a pullout spike (3), as well as an intervening period of low amplitude ÔtastingÕ of plant ßuids (2). These waveforms are also typically very short in duration ranging from 1.5 to 7 s. This waveform has been magniÞed to illustrate the subpatterns within it. Vertical bars represent time units (each interval is ⬇1. 7 s), and all waveforms are measured from left to right.

plitude spikes at a position about halfway between baseline and maximum peak amplitude. C1 was observed not only in the artiÞcial diet but also in all plants in the study except caulißower. Although rare, when it occurred, C1 was performed in long durations (10.8 Ð1,249.7 s); the longest being exhibited on okra. On simple artiÞcial diet, the stylets were held motionless during waveform C1, with their tips offset from one another. Stylets also appeared to be held at

543

Fig. 2. Representative electronic waveforms of the drilling/laceration behavior, waveform B. Waveform B is denoted by a high amplitude multi-spiked appearance. Vertical bars represent time units (each interval is ⬇1. 7 s), and all waveforms are measured from left to right.

a stereotypical depth (described further below). Ink particles moved consistently toward the stylet tips, and simultaneously the insectÕs abdomen rocked steadily up and down. Similar labial bending, stylet extension, and abdominal rocking also occurred on plants during this waveform occurrence, suggesting that the likely ingestion observed in artiÞcial diets also occurred in plants. Waveform C2. A more common waveform was C2 (Fig. 4), which also contained ingestion. Unlike C1, C2 had periodic, wave-like moderately high-amplitude interspersed spikes, suggesting the appropriate name of wavy-line ingestion. C2 occurred more frequently than C1, and was often contained within multiple

544


Vol. 31, no. 3

Fig. 3. A characteristic ingestion waveform is visible on the right two-thirds of the diagram (labeled C1). Note the moderate level in relative amplitude and short-spike appearance. This is a more horizontally compressed view than the other waveform, to show the full length of two test probes (labeled A) and one multi-waveform probe (labeled B-A-B). Vertical bars represent ⬇1. 5 s and waveforms are read from left to right. Vertical bars represent time units (each interval is ⬇1. 7 s), and all waveforms are measured from left to right.

events and numerous probes by the same insect. Observations on plants and artiÞcial diets were the same as for C1. It is logical that the increased conductivity represented by wave-like spikes corresponds to brief spurts of watery salivation interspersed among primarily ingestion events. However, there was no indication of particle movement (either away from or toward stylets) when this wavy portion of C2 was observed in artiÞcial diet. It is likely that particles could not be moved by such small spurts of saliva because they were embedded in a viscous diet. Other ink-particle correlation studies (e.g., McLean and Kinsey 1965, Hunter and Backus 1989) used liquid diet, allowing freer movement for the particles. However, L. hesperus nymphs would not maintain stylet contact with liquid diet in preliminary experiments. Waveform E. The salivary bubble, or E waveform, was recorded only on the artiÞcial diet, and resembled the high amplitude squared-off appearance of the D waveform (described below) (Fig. 5). Waveform E is Fig. 5. Waveform representing the formation and release of an excretory droplet onto the substrate. The excretory waveform ÔDÕ is designated by a high amplitude singular spike with a plateau-like top. This waveform has been reduced to illustrate the high amplitude nature of the waveform, which exceeds the sensitivity range of the monitor as exempliÞed by the ßattened top. Vertical bars represent time units (each interval is ⬇1. 7 s), and all waveforms are measured from left to right.

Fig. 4. The low amplitude, wavy ingestion waveform C2. Vertical bars represent time units (each interval is ⬇1. 7 s), and all waveforms are measured from left to right.

probably an excessive form of the pullout spike, a high-amplitude peak often associated with the termination of probing events. The higher-than-usual amplitude probably reßects the larger voltage applied to the artiÞcial diet. This waveform may also be an artifact of the relatively low viscosity of the diet. Due to its occurrence in artiÞcial diets only, we excluded the E waveform from any further analysis. Waveform F. This waveform resembles waveform B but with more frequent excursions in amplitude cor-

June 2002


545

Fig. 7. Both of the nonprobing waveforms, including walking and resting waveforms, Z1 and Z2, respectively. The walking waveform appears as a variable baseline with low amplitude spikes and bumps. The resting waveform has a ßat baseline appearance. Vertical bars represent time units (each interval is ⬇1. 7 s), and all waveforms are measured from left to right. The relative amplitude of Z1 and Z2 have been increased for labeling purposes.

Fig. 6. Waveform F resembles waveform B, however, it usually has much higher relative amplitude, connoting the multiple reinsertion of the stylets into the substrate. Vertical bars represent time units (each interval is ⬇1. 7 s), and all waveforms are measured from left to right.

responding to the repetitive insertion of the stylets into the substrate (Fig. 6). Typically, the labium was held directly beneath the insect, with small deßections of the labium indicative of stylet insertion. The stylets were rapidly reinserted at only very shallow depths. Although F was an infrequently observed waveform, its duration was similar to some of the longer-duration B waveforms. Interestingly, this behavior was only exhibited on cotton squares. EPG Ethogram: Nonprobing Behaviors. Nonprobing activities (especially Z1 and Z2, see below) comprised the bulk of activities expressed by L. hesperus nymphs, a result also found by other Lygus researchers (HatÞeld et al. 1982). Nonprobing activities included all behaviors during which there was no penetration of the plant by the insectÕs mouthparts. Waveform D. This waveform corresponded to the formation and release of an excretory droplet onto the feeding substrate. Waveform D does not occur with

other piercing-sucking insects, such as aphids and leafhoppers, which propel their excreta away from the plant, resulting in no electrical detection of the behavior. L. hesperus, however, formed an excretory droplet and subsequently dragged its abdomen across the substrate until the droplet detached from the anus. The release of the droplet took no longer than 3 s, and resulted in a very high amplitude squared-off waveform (Fig. 5) due to direct contact with the abdomen. Waveform D was observed several times on plants but only once on artiÞcial diet; no particle movement was observed in the diet because the release of an excretory droplet occurred on the surface only. Waveform Z1. Z1 waveform (Fig. 7) was a type of baseline waveform that corresponded to any active movements exhibited by an insect. These activities included the ⍀1, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, and 15 behaviors (Table 1) in the video record on both plants and diet. The waveform resembled a very low amplitude wavy to jagged appearance. Waveform Z2. This waveform was also a type of baseline (Fig. 7), and is associated with resting behaviors wherein the insect exhibits no external movements. It is correlated with either ⍀2 (standing/ resting) or ⍀15 behaviors (insect out of focus or being adjusted, Table 1) on the video record. Its appearance was a straight-line baseline of low amplitude. Waveform Z2 was also seen on the artiÞcial diets; however, no ßuid ßow was seen because these are only surface activities. Nonprobing time was spent primarily (⬎66%) in nonlocomotor (Z2) activity when the insect simply stayed at rest in one area. Because this behavior expended more than two-thirds of the insectsÕ time on a plant, it was therefore an essential component of the speciesÕ behavioral repertoire. During the Z2 behavior, the nymph held the antennae in a stereotypical manner at particular angles with respect to the rest of the body. These antennal angles were very similar to the antennal angles seen during ingestion periods (see below) and resting. Ingestion/Body Posture Correlation. A stereotypical body posture, primarily composed of labial and antennal orientations, was 100% correlated with every C-type waveform ingestion event observed. Stereo-

546


Vol. 31, no. 3

Fig. 8. The representative body postures associated with ingestion events. (A) Lateral view of Lygus nymph head, showing three distinct labial angles (␪L1, ␪L2, and ␪L3) and one antennal angle (␪A1). The labial angles are present during all ingestion events, regardless of leaf location, whereas the antennal angle occurs only during interveinal ingestion. (B) Dorsal view of Lygus nymph head illustrating the second antennal angle (␪A2) which also occurs only during interveinal ingestion.

typical angles were formed between the cervix and Þrst labial segment, as well as the second and third, and third and fourth segments of the labium. Antennal orientations coincided with angles between the antennal scape and a sagittal and transverse plane that originated at the antennal socket. When viewed laterally, the labium produced three distinctive angles during most (⬎99%) probing events on both immature and mature cotton (Fig. 8). The Þrst angle was formed between the ventral surface of the cervix and the Þrst labial segment (␪L1). The second angle occurred between the second and third labial segments (␪L2). The third occurred between the third and fourth labial segments (␪L3). At the onset of ingestion waveforms and throughout their duration, the second labial angle (␪L2) remained Þxed at 70 ⫾ 2⬚. The Þrst and third labial angles were somewhat variable, although ␪L1 typically approximated a right angle, and ␪L3 was always obtuse. During interveinal ingestion on cotton leaves, the antennae were also held in a stereotypical manner; in a slightly elevated orientation, 10 Ð15⬚ above a horizontal midline that transects the antennal socket. There was also a forward rotation ⬇40⬚ anterior from a sagittal plane that transects the antennal socket (Fig. 8). The antennae were positioned in this manner during ingestion episodes in which the insect was not directly on a primary leaf vein. This antennal orientation was also observed in resting behaviors. The durations of the 41 ingestion events (C1 or C2) were highly variable, lasting from several seconds to over 10 min. This variability did not seem to correspond to the age of the plant or one particular insect, because most insects employed both long and short ingestion behaviors. A paired t-test detected no signiÞcant difference between the durations recorded on the video equipment compared with those from the electronic monitoring system (t-calc. ⫽ 1.384, t-crit. ⫽ 2.021, df ⫽ 40; Table 3).

Discussion Waveform Interpretations. The behaviors represented by the A waveform appear to function in part as a mechanism for determining the quality and palatability of the food source. A large expulsion of watery saliva to solubilize plant constituents was followed by brief uptake of ßuids into the anterior portion of the foregut (the precibarium [Backus 1988] or epipharyngeal organ) for gustatory discrimination by chemosensilla located there. This interpretation was strongly supported by the Þnding that waveform A was the only feeding behavior expressed on the nonhost plant caulißower, and after its completion no further feeding was pursued. Test probes as a means of gustatory discrimination are a behavior well-known from the feeding of aphids and other sheath-feeding homopterans, though they have never been documented for heteropterans. It was especially surprising to see “test probes” performed in large numbers by L. hesperus, because they tended to occur infrequently and just preceding long ingestion probes in other insects. On plants, A-containing test probes were performed in bouts of many probes in rapid succession. But, unlike the short-duration probes of Empoasca spp., they were not done in the same vicinity. Between two contiguous probes, the insect walked rapidly to a new site on the leaf or square, often covering many millimeters in only a second. Thus, test probes were performed all over the leaf or square. We speculate Table 3. Calculations of mean, standard error of the mean (SEM), t-calc., and P value for the cohort of 41 ingestion periods Variable

n

Mean

SEM

EPG VID Difference Combined

41 41 41

70.32 69.42 0.898

14.92 14.84 0.649

t-crit.

t-calc.

P value

2.021

1.369

17.47

June 2002


that there may be an additional function of these short-duration probes, i.e., to “soften-up” the plant tissues, or perhaps somehow overcome some type of plant defense, by repeatedly puncturing and salivating into the plant. Waveform B is analogous to the Ia (multiple-cell laceration) waveform of Empoasca spp. leafhoppers (e.g., Calderon and Backus 1992), because it encompasses several types of stylet activities simultaneously. It likely represents intracellular penetration of numerous cells, salivation, and active stylet thrusting, but no (or at most very little) ingestion. Its primary function in feeding appears to be to mechanically and chemically solubilize plant tissues, as indicated by both variable waveform amplitudes (stylet activity and/or salivation) and video recordings of head movements. This is probably associated with preparation of the area for future ingestion events (A.R.C., unpublished data). C1-like waveforms have been seen in virtually all other insects studied with electronic monitoring experiments (e.g., aphids, Tjallingii 1978; leafhoppers, Hunter and Backus 1989; and heteropterans, Bonjour et al. 1991), and have consistently been correlated with ingestion. The relatively ßat-line appearance supported that little salivation was occurring. The regular, very short spike was likely to represent rhythmic pumping of the cibarium, indicating active ingestion of ßuids from areas of low hydrostatic pressure within the plant. Although this was the longest duration L. hesperus waveform, it was quite short compared with ingestion durations for most other hemipterans (it can continue for many hours in sheath-feeding homopterans). The comparatively short sucking suggested that ingestion was occurring from nonconducting cells in the plant (i.e., not phloem sieve elements or xylem tracheary elements), although this cannot be stated conclusively at this time. Brief ingestion could be occurring from conducting cells that became subsequently blocked by callose and P protein, or ingestion could be of previously solubilized material. Both cases occur with the similar Ic waveform in Empoasca spp. The C waveform has also been found in leafhoppers, both sheath-feeders (for which it has been termed ÔRÕ, see review discussion in Wayadande 1994), and lacerate-and-ßush feeders (the Ib waveform of Empoasca spp., e.g., Calderon and Backus 1992). Nothing exactly like it has been seen in aphids. Ib by Empoasca spp. occurs primarily in mesophyll and parenchyma cells in interveinal regions of succulent leaf material, and is correlated with cell emptying in those areas (Njihia 1996). It was surprising to see C2 in viscous diet, because Ib had never been documented on viscous or liquid artiÞcial diet (Hunter and Backus 1989). Although we have no histological tissue correlations for L. hesperus in this study, C2Õs strong resemblance to Ib suggests that it too is a mesophyll cell puncturing and ingestion behavior, with brief spurts of salivation for occasional solubilization of cellular material. Waveform F, which consisted of a high amplitude, highly variable waveform, was observed only with nymphs feeding on cotton squares. The behavior may

547

therefore correspond to changes in internal tissue or vasculature that occur in plant reproductive structures as they develop from vegetative structures. This behavioral plasticity, which is dependent on the plantÕs developmental stage, indicates the ability of L. hesperus nymphs to ascertain the type of structure from which they are feeding and adjust their feeding strategy accordingly. The destructive nature of the F behavior, as seen by the multiple insertions of the stylets, may be a potential method for causing square shed. If so, plant breeders may be able to develop cultivars that eliminate this type of behavior, as shown through EPG screening. As seen by other Lygus researchers, these insects spent the majority of plant access time in nonprobing behaviors (HatÞeld et al. 1983). This time at rest putatively serves many functions, including reduction of metabolic activity and subsequent energy expenditure, anabolic activity during which the insect was assimilating ingested foods, and predator/parasitoid avoidance or detection, as well as other functions. The resting time was typically spent with the antennae raised above the antennal transverse plane, providing evidence that this was deÞnitely a period of detecting volatiles (whether to recongnize conspeciÞcs or potential predators) via antennal sensilla just above the plant surface boundary layer. The very large surface area-to-volume ratio of these nymphal insects also may result in relative inactivity as a means of conserving water reserves. EPG/Body Posture Correlation. The importance of combining EPG and videomicrography was demonstrated in both the ethogram and postural correlation studies. EPG can only be performed with a limited sample size of insects, and also creates a no-choice feeding situation due to the restrictions on movement imposed by the wire tether. However, EPG can provide deÞnitive correlations of ingestion with certain body postures, as shown by the paired t-test. The present correlation of video recordings with waveforms will allow future studies to use rapid, visual observation of an insectÕs body posture to predict occurrence of ingestion. Thus, it will be possible to perform experiments with large sample sizes and multiple feeding substrates. The stereotypical antennal orientation during interveinal ingestion may be an adaptive posture to detect volatiles at the plant surface boundary layer. Ingestion is an important indicator of a temporal feeding commitment by L. hesperus to a particular host plant. Ingestion on cotton can extend from 10 s to ⬎10 min. The time taken to consummate ingestion behaviors is therefore a vulnerable period for the insect. Stylet pullout and labial retraction under the abdomen take 1Ð2 s (A.R.C., unpublished data), sufÞcient time for attack by predators or parasitoids. By placing those antennal segments that contain chemosensilla in a position to detect natural enemies, the nymph would be able to readily respond accordingly. The potential predators of L. hesperus not only include entomophagous groups such as nabids, but also their conspeciÞcs,

548


which are known to be cannibalistic (A.R.C., unpublished data). The postural correlation of the labium and ingestion events has several broad-ranging implications. Due to the Þxed second labial angle, L. hesperus appear to prefer a particular ingestion depth on cotton. Such a preferred depth is unlikely to always coincide with the position of a preferred plant tissue (e.g., phloem, mesophyll), providing yet more evidence of their lacerate-and-ßush strategy of feeding, ingesting from any solubilized tissue available. Histological work is one of the next steps planned to elucidate how the feeding behaviors exhibited by these insects causes the morphological and physiological changes in damaged tissues. This preferential feeding site may be associated with avoidance of deterious plant chemicals or physical barriers. Also, the feeding at one particular depth may represent phylogenetic inertia. If preferred hosts in the evolutionary past of this family contained plant tissues of differential nutrition or palatability, selection pressures to alter this depth may not be strong enough to change this pattern of behavior. The compilation of ethograms, although sometimes thought to be a method used only by early ethologists, remains important in answering both basic and applied questions about an insectÕs behavioral repertoire. Use of videomicrography and EPG allow detailed description of feeding and nonfeeding behaviors exhibited by piercing-sucking insects. Here we provide the Þrst electronic monitoring/EPG ethogram of an important pest species. The information given herein has implications not only for the control of damaging feeding behaviors of these insects (e.g., screening for resistant cultivars) but also for explaining the plasticity and frequency of feeding activities that occur within a single phytophagous insect species. Our work can facilitate further experimentation on both applied and basic questions, such as (1) how do different cotton cultivars alter the types of feeding expressed by L. hesperus nymphs, (2) do different host plants have particular feeding strategies or tactics associated with them, (3) how does the behavioral repertoire of L. hesperus change over its lifespan, and (4) how does the feeding ethogram of L. hesperus compare with its congeneric relative L. lineolaris? Acknowledgments We thank the following people for help in preparation of the manuscript: JoAnna Cline for her drawings of the Lygus heads. Floyd Shockley for help with the tables and Þgures, and Mark Linit and Bruce Barrett for providing comments on earlier versions of the manuscript. This article was partially funded by Monsanto. This research was, in part, supported by the Missouri Agricultural Experiment Station.

References Cited Allen, W. W., and S. E. Goede. 1963. The relationship of lygus bugs and thrips to fruit deformity in strawberries. J. Econ. Entomol. 56: 823Ð825. Anonymous. 1999. Project underway to battle cotton pest in Mississippi Delta. Entomol. Soc. Am. Newsl. 22: 1 and 3.

Vol. 31, no. 3

Backus, E. A. 2000. Our own Jabberwocky: clarifying the terminology of piercing-sucking behaviors of homopterans, pp. 1Ð13. In G. P. Walker and E. A. Backus (eds.), Proceedings, Principles and Applications of Electronic Monitoring and Other Techniques in the Study of Homopteran Feeding Behavior. Thomas Say Publications in Entomology. Entomological Society of America, Lanham, MD. Backus, E. A., and W. H. Bennett. 1992. New AC electronic insect feeding monitor for Þne-structure analysis of waveforms. Ann. Entomol. Soc. Am. 85: 437Ð 444. Backus, E. A. 1988. Sensory systems and behaviours which mediate hemipteran plant feeding: A taxonomic overview. J. Insect Physiol. 34: 151Ð165. Backus, E. A., and W. B. Hunter. 1989. Comparison of feeding behavior of the potato leafhopper Empoasca fabae (Homoptera: Cicadellidae) on alfalfa and broad bean leaves. Environ. Entomol. 18: 473Ð 480. Baker, K. F., N. C. Snyder, and A. H. Holland. 1946. Lygus bug injury to lima beans in California. Phytopathology 36: 493Ð503. Bonjour, E. L., W. S. Fargo, J. A. Webster, P. E. Richardson, and G. H. Brusewitz. 1991. Probing behavior comparisons of squash bugs (Heteroptera: Coreidae) on cucurbit hosts. Environ. Entomol. 20: 143Ð149. Broersma, D. B., and W. H. Luckman. 1970. Effects of tarnished plant bug feeding on soybean. J. 69: 127Ð129. Calderon, J. D., and E. A. Backus. 1992. Comparison of the probing behaviors of Empoasca fabae and E. kraemeri (Homoptera: Cicadellidae) on resistant and susceptible cultivars of common beans. J. Econ. Entomol. 85: 88 Ð99. Chang, V.C.S. 1978. Feeding activities of the sugarcane leafhopper: identiÞcation of electronically recorded waveforms. Ann. Entomol. Soc. Am. 71: 31Ð36. Cline, A. R. 2000. Stylet penetration behaviors of the western tarnished plant bug (Hemiptera: Miridae) on cotton. M.S. thesis, University of Missouri, Columbia. Cole, R. A., W. Riggall, and A. Morgan. 1993. Electronically monitored feeding behavior of the lettuce root aphid (Pemphigus bursarius) on resistant and susceptible lettuce varieties. Entomol. Exp. Appl. 68: 179 Ð185. Cook, A. C., and J. J. Neal. 1999. Feeding behavior of larvae of Anasa tristis (Heteroptera: Coreidae) on pumpkin and cucumber. Environ. Entomol. 28: 173Ð177. Debolt, J. W. 1982. Meridic diet for rearing successive generations of Lygus hesperus. Ann. Entomol. Soc. Am. 75: 119 Ð122. Ecale, C. L., and E. A. Backus. 1994. A characteristic body posture of the potato leafhopper is correlated with probing. J. Entomol. Sci. 29: 490 Ð 495. Habibi, J., E. A. Backus, and T. H. Czapla. 1993. Plant lectins affect survival of the potato leafhopper (Homoptera: Cicadellidae). J. Econ. Entomol. 86: 945Ð951. Harrewijn, P., and H. Kayser. 1997. Pymetrozine, a fastacting and selective inhibitor of aphid feeding. In-situ studies with electronic monitoring of feeding behavior. Pestic. Sci. 49: 130 Ð140. Hatfield, L. D., J. L. Frazier, and J. Ferreira. 1982. Gustatory discrimination of sugars, amino acids and selected allelochemicals by the tarnished plant bug Lygus lineolaris. Physiol. Entomol. 7: 15Ð24. Hatfield, L. D., J. Ferreira, and J. L. Frazier. 1983. Host selection and feeding behavior by the tarnished plant bug, Lygus lineolaris (Hemiptera: Miridae). Ann. Entomol. Soc. Am. 76: 688 Ð 691. Hunter, W. B., and E. A. Backus. 1989. Mesophyll-feeding by the potato leafhopper, Empoasca fabae (Homoptera: Cicadellidae): results from electronic monitoring and

June 2002


thin-layer chromatography. Environ. Entomol. 18: 465Ð 472. Kabrick, L. R., and E. A. Backus. 1990. Salivary deposits and plant damage associated with speciÞc probing behaviors of the potato leafhopper, Empoasca fabae, on alfalfa stems. Entomol. Exp. Appl. 56: 287Ð304. Khan, Z. R., and R. C. Saxena. 1988. Probing behavior of three biotypes of Nilaparvata lugens (Homoptera: Delphacidae) on different resistant and susceptible rice varieties. J. Econ. Entomol. 81: 1338 Ð1345. Mauney, J. R., and T. J. Henneberry. 1979. IdentiÞcation of damage symptoms and patterns of feeding of plant bugs in cotton (Gossypium hirsutum). J. Econ. Entomol. 72: 496 Ð501. McLean, D. L., and M. G. Kinsey. 1964. A technique for electronically recording aphid feeding and salivation. Nature (Lond.) 202: 1358 Ð1359. McLean, D. L., and M. G. Kinsey. 1965. IdentiÞcation of electrically recorded curve patterns associated with aphid salivation and ingestion. Nature (Lond.) 205: 1130 Ð 1131. McLean, D. L., and M. G. Kinsey. 1967. Probing behavior of the pea aphid, Acyrthosiphon pisum. I. DeÞnitive correlation of electronically recorded waveforms with aphid probing activities. Ann. Entomol. Soc. Am. 60: 400 Ð 405. McLean, D. L., and M. G. Kinsey. 1968. Probing behavior of the pea aphid, Acyrthosiphon pisum. II. Comparisons of salivation and ingestion in host and non-host plant leaves. Ann. Entomol. Soc. Am. 61: 730 Ð739. Njihia, S. 1996. Cellular responses of Phaseolus beans to speciÞc probing behaviors of Empoasca kraemeri (Ross & Moore) (Homoptera: Cicadellidae). M.S. thesis, University of Missouri, Columbia. Parencia, C. R., Jr. 1978. One hundred twenty years of research on cotton insects in the United States. USDA-ARS Agric. Handb. 515. Patana, R. 1982. Disposable diet packet for feeding and oviposition of Lygus hesperus (Hemiptera: Heteroptera: Miridae). J. Econ. Entomol. 75: 668 Ð 669. Port, G. R. 1978. Studies on feeding biology of Auchenorrhynca (Homoptera). Ph.D. dissertation, University of London. Prado, E., and W. F. Tjallingii. 1994. Aphid activities during sieve element punctures. Entomol. Exp. Appl. 72: 157Ð 165. Purseglove, J. W. 1968. Tropical crops: Dicotyledons. Longman, London. Raman, K. V., W. M. Tingey, and P. Gregory. 1979. Potato glycoalkaloids: effect on survival and feeding behavior of the potato leafhopper. J. Econ. Entomol. 72: 337Ð341. Ramirez, C. C., and H. M. Niemeyer. 1999. Salivation into sieve elements in relation to plant chemistry: The case of

549

the aphid Sitobion fragariae and the wheat, Triticum aestivum. Entomol. Exp. Appl. 91: 111Ð114. SAS Institute. 1985. SAS userÕs guide: statistics, 1985 ed. SAS Institute, Cary, NC. Schwartz, M. D., and R. G. Foottit. 1998. Revision of the nearctic species of the genus Lygus Hahn, with a review of the palearctic species (Heteroptera: Miridae). Mem. Entomol. Int. 10: 1Ð 428. Snodgrass, G. L. 1998. Distribution of the tarnished plant bug (Heteroptera: Miridae) within cotton plants. Environ. Entomol. 27: 1089 Ð1093. Sorenson, C. J. 1936. Lygus bugs in relation to occurrence of shriveled alfalfa seed. J. Econ. Entomol. 29: 454 Ð 457. Strong, F. E. 1970. Physiology of injury caused by Lygus hesperus. J. Econ. Entomol. 63: 808 Ð 814. Tarn, T. R., and J. B. Adams. 1982. Aphid probing and feeding, electronic monitoring, and plant breeding. In K. F. Harris and K. Maramorosch (eds.), Pathogens, vectors, and plant diseases. Academic, New York. Tingey, W. M., and E. A. Pillemer. 1977. Lygus bugs: crop resistance and physiological nature of feeding injury. Bull. Entomol. Soc. Am. 23: 277Ð287. Tingey, W. M., T. F. Leigh, and A. H. Hyer. 1975. Glandless cotton: susceptibility to Lygus hesperus Knight. Crop Sci. 15: 251Ð253. Tjallingii, W. F. 1978. Electronic recording of penetration behavior by aphids. Entomol. Exp. Appl. 24: 721Ð730. Van Giessen, W. A. and D. M. Jackson. 1998. Rapid analysis of electronically monitored homopteran feeding behavior. Ann. Entomol. Soc. Am. 91: 145Ð154. Walker, G. P. 2000. A beginners guide to electronic monitoring of homopteran probing behavior. In G. P. Walker and E. A. Backus (eds.), Principles and applications of electronic monitroing and other techniques in the study of Homopteran feeding behavior. Thomas Say Publications. Entomological Society of America, Lanham, MD. Wayadande, A. C. 1994. Electronic monitoring of leafhoppers and planthoppers: feeding behavior and application in host-plant resistance studies. In M. M. Ellsbury, E. A. Backus, and D. E. Ullman (eds.), History, development, and application of AC electronic insect feeding monitors. Thomas Say Publications. Entomological Society of America, Lanham, MD. Wayadande, A. C., and E. A. Backus. 1989. Feeding behavior of the potato leafhopper (Homoptera: Cicadellidae) on chlordimeform- and phosmet-treated alfalfa. J. Econ. Entomol. 82: 766 Ð772. Wilson, L. T., T. F. Leigh, D. Gonzalez, and C. Foristiere. 1984. Distribution of Lygus hesperus (Miridae: Hemiptera) on cotton. J. Econ. Entomol 77: 1313Ð1319. Received for publication 25 May 2001; accepted 27 December 2001.