- Email: [email protected]
Starvation period and age affect the response of female Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae) to odor and visual cues
ARTICLE IN PRESS
Journal of Insect Physiology 52 (2006) 729–736 www.elsevier.com/locate/jinsphys
Starvation period and age affect the response of female Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae) to odor and visual cues Melanie M. Davidson, Ruth C. Butler, David A.J. Teulon New Zealand Institute for Crop & Food Research Limited, Private Bag 4704, Christchurch, New Zealand Received 11 January 2006; received in revised form 19 March 2006; accepted 27 March 2006
Abstract The effects of starvation or age on the walking or flying response of female Frankliniella occidentalis to visual and/or odor cues in two types of olfactometer were examined in the laboratory. The response of walking thrips starved for 0, 1, 4, or 24 h to an odor cue (1 ml of 10% p-anisaldehyde) was examined in a Y-tube olfactometer. The take-off and landing response of thrips (unknown age) starved for 0, 1, 4, 24, 48 or 72 h, or of thrips of different ages (2–3 days or 10–13 days post-adult emergence) starved for 24 h, to a visual cue (98 cm2 yellow sticky trap) and/or an odor cue (0.5 or 1.0 ml p-anisaldehyde) was examined in a wind tunnel. More thrips walked up the odorladen arm in the Y-tube when starved for at least 4 h (76%) than satiated thrips (58.7%) or those starved for 1 h (62.7%, Po0.05). In the wind tunnel experiments the percentage of thrips to fly or land on the sticky trap increased between satiated thrips (7.3% to fly, 3.3% on trap) and those starved for 4 h (81.2% to fly, 29% on trap) and decreased between thrips starved for 48 (74.5% to fly, 23% on trap) and 72 h (56.5% to fly, 15.5% on trap, Po0.05). Fewer younger thrips (38.8%) landed on a sticky trap containing a yellow visual cue of, those that flew, than older thrips (70.4%, Po0.05), although a similar percentage of thrips flew regardless of age or type of cue present in the wind tunnel (average 44%, P40.05). r 2006 Elsevier Ltd. All rights reserved. Keywords: Frankliniella occidentalis; Host cues; Intrinsic factors; Walking; Flight
1. Introduction Western flower thrips, Frankliniella occidentalis (Pergande), is a pest on a variety of vegetable, fruit and ornamental crops (Lewis, 1997a). F. occidentalis can cause extensive damage to its host plant either directly through feeding or oviposition, or indirectly through the transmission of tospoviruses (Ullman et al., 1997). This pest is difficult to manage because of its small size (o2 mm long), cryptic habits, and resistance to a number of insecticides (Brødsgaard, 1989). Colored sticky traps have been used to monitor populations of F. occidentalis (Brødsgaard, 1993; Shipp, 1995; Shipp et al., 2000). However, colored sticky traps only catch flying thrips. Factors that increase the number Corresponding author. Tel.: +64 3 325 6400; fax: +64 3 325 2074.
E-mail address: [email protected] (M.M. Davidson). 0022-1910/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.jinsphys.2006.03.013
of thrips to fly could be used to improve monitoring, for example by allowing earlier detection of thrips’ presence, and could even increase their overall exposure to biological or chemical control strategies. Thrips’ take off, flight duration and landing may be influenced by several extrinsic and intrinsic factors. Likely extrinsic factors include temperature, light, wind and, to a lesser extent, relative humidity (Lewis, 1997b), as well as pheromones and kairomones. A number of studies have examined the influence that these factors have on flight (temperature, e.g. Shipp and Zhang 1999; Pearsall, 2002; light, e.g. O’Leary and Kirk, 2004; air flow, e.g. Teulon et al., 1999; pheromones, e.g. MacDonald et al., 2002; Hamilton, et al., 2005; and kairomones, e.g., Brødsgaard, 1990; Teulon et al., 1993a,b; 1999; Smits et al., 2000). Intrinsic factors that may influence thrips’ responses toward host cues, such as ovarian development, their mated state, or the senescences of host plants (Lewis,
ARTICLE IN PRESS 730
M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
1997b) have received less attention. However, the physiological state of insects could be expected to influence their behavioral responses toward external stimuli (Barton Browne, 1993). The time since feeding can be particularly important in altering an insect’s behavior toward host cues (Bernays and Chapman, 1994). For example, a higher number of starved than unstarved Colorado potato beetle (Leptinotarsa decemlineata (Say)) have been observed to fly from a summer population (MacQuarrie and Boiteau, 2003), and a greater percentage of starved female desert locust (Schistocerca gregaria (Forsa˚l)) take flight compared with those with full foreguts (Lambert, 1981). An adult female insect’s age could also affect its response to stimuli through physiological changes, such as ovarian development (Barton Browne, 1993). For example, Brevault and Quilici (1999) found that less than 5% of young (2 days post-adult emergence) female tomato fruit fly (Neoceratitis cyanescens (Bezzi)) landed on an orange sphere stimulus, compared with around 65% of older females (10 days postadult emergence). This study reports for the first time, the effect of starvation or age on the responses of female F. occidentalis to artificial host cues. The aims of this study were: (i) to compare the walking or flying response of thrips starved for different lengths of time to artificial host cues (visual and/or odor) in a Y-tube or low speed laminar flow wind tunnel; and (ii) to investigate the effect of age on the flying response of female F. occidentalis to artificial host cues in a wind tunnel. The odor used in the following experiments was p-anisaldehyde. An anisaldehyde compound was one of the first volatile compounds shown to attract higher numbers of thrips to traps baited with this compound compared to unbaited traps (Howlett, 1914; Morgan and Crumb, 1928). Greenhouse trials using p-anisaldehyde baited or unbaited traps demonstrated higher catches of F. occidentalis in the former (Teulon et al., 1993a,b). Previous experiments evaluating the response of F. occidentalis females to volatile compounds in a Y-tube olfactometer showed p-anisaldehyde elicted one of the strongest responses out of a range of compounds tested (Koschier et al., 2000). 2. Materials and Methods 2.1. Insects All F. occidentalis used in the following starvation experiments were from a colony maintained on potted flowering chrysanthemums, Dendranthema grandiflora (cv. Onyx time yellow). The colony was established in 2001 from wild thrips collected from commercial greenhouses (Auckland, New Zealand). The plants used for the colony were held in three temperature-controlled perspex cages (five plants per cage) with a temperature maintained at 25 1C under a 14:10 h light:dark cycle. A fresh uninfested plant was placed in each cage every 3 days and the oldest plant was removed at the same time.
Even-age female thrips were obtained using a green bean (Phaseus vulgaris L.) rearing method adapted from Loomans and Murai (1997). Briefly, this involved placing 200–300 adult males and females in a glass container (500 ml) that held five to six green beans and a 1.5 ml eppendorf tube with a hole in the lid containing approximately 0.2 g of ground bee-collected pollen. The bottom of the container was lined with a square of greenhouse plant matting (100 cm2) and filter paper (Whatman No. 1, 9 cm diameter). A 5 cm diameter hole was cut in the center of the metal screw lid, which was used to secure mesh (20 mm) over the opening of the container. After 3 days, the egg-laden bean pods were replaced with six fresh beans and the egg-laden beans were placed on top of filter paper and plant matting, along with bee-collected pollen in another 500 ml glass container. The glass containers were held in an incubator at a temperature of 2572 1C and a 14:10 h light:dark cycle. The beans were replaced every 5 days, and adults began to emerge after approximately 21 days. In order to obtain enough thrips for an experiment, adult females (along with males) were collected every 2 days for the older age group (10–13 days post-adult emergence) and held for 10 days in glass jars containing filter paper, fresh beans and bee-collected pollen, in the incubator. 2.2. Y-tube olfactometer experiment The response of walking female F. occidentalis starved for different lengths of time toward a volatile compound was evaluated in a glass Y-tube olfactometer following the method described by Koschier et al. (2000). The Y-tube has two branching arms at an angle of approximately 451 leading from a single tube, and each section is 60 mm long, with an internal diameter of 5 mm.The arms of the Y-tube were connected to glass Wheaton Micro Kits adapters that were in turn attached to 4 ml glass vials each containing a 1 cm2 piece of filter paper (Whatman No. 1). The Y-tube and Wheaton apparatus were placed in a grey box (to prevent external stimuli influencing thrips’ behavior), located in a darkened, air-conditioned room (22 1C with a 31 range). The Y-tube was positioned at an inclining angle of 251 and illuminated from above by a halogen lamp (780 lux, Beha digital Lux Hi Tester 93-1065L, Germany). One microlitre of the volatile compound, 10% p-anisaldehyde diluted in hexane (95%, BDH), was applied to filter paper in one vial, while 1 ml of hexane was applied to filter paper held in the second vial. Air was drawn through activated charcoal before entering the Wheaton apparatus and Y-tube using a suction pump (AR Harris Co. Ltd, New Zealand), producing an airflow of 5 cm/s through each arm and of 10 cm/s at the base of the Y-tube. Clean air was drawn through the Y-tube for 30 min before introducing the first thrips at the beginning of each bioassay. Connections between the activated charcoal, Wheaton apparatus, Y-tube and suction pump were with silicone tubing.
ARTICLE IN PRESS M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
Female thrips of unknown age were collected from the colony using an aspirator and held in a perspex ring cage (8 cm diameter, 5 cm high) (Loomans and Murai, 1997) with access to water only for 0, 1, 4 or 24 h prior to introducing them into the Y-tube olfactometer. An individual female thrips was released into the base of the Y-tube using a small aspirator. Most thrips walked up into the Y-tube within a few seconds, at which time the silicone air suction tubing was reconnected to the base of the glass Y-tube and timing of an individual was started. Thrips were observed as they walked up the stem of the Y-tube and entered either arm. When the thrips reached the far end of the respective arm their choice was recorded. On the rare occasion a thrips failed to make a choice after 3 min the thrips was removed and replaced with another thrips. After every five thrips that made a choice between the odor or clean air the Y-tube and Wheaton apparatus were rotated 1801 to avoid position effects. One replicate was completed after 25 thrips had chosen either arm, after which the Y-tube and Wheaton apparatus were thoroughly cleaned with acetone (99.5%, BDH) and allowed to dry before repeating the experiment. The experiment was replicated three times per starvation period to give a total of 75 choices per starvation period. 2.3. Wind tunnel olfactometer experiments A low-speed laminar air flow wind tunnel, described by Berry et al. (2005), was used to measure the flying response of thrips starved for different periods (0, 1, 4, 24, 48, or 72 h). The perspex observation chamber measured 0.58 wide by 0.55 m in height by 1.7 m in length and was illuminated by five 36W 84 cool lights alternating with five 36W 96 true light fluorescent tubes, producing an average of 10,000 lux (Beha, Digital Lux Hi Tester 93-1065L, Germany). During bioassays, room air, filtered through activated charcoal, was passed through the chamber in a laminar flow. The frame and floor of the wind tunnel were painted a charcoal gray to facilitate sighting of thrips. Sticky traps were constructed using two square glass plates (10 10 cm, 2 mm thick) with Tangle traps (Grand Rapids, Michigan) applied to the outward facing sides. A 98 cm2 yellow visual cue was obtained by removing the glue from commercially available sticky traps (Horriver, Koppert, the Netherlands). Visual cues were sandwiched between the glass plates. When no visual cue was required, the sticky trap was used without placing anything between the glass plates. In the wind tunnel, the sticky trap was hung from a clip attached to a retort stand, with the top edge of the sticky trap 20 cm from the roof of the wind tunnel, 0.5 m upwind of the thrips’ perspex release platform. The volatile odor (p-anisaldehyde, 0.5 or 1.0 ml) was applied to black filter paper (9 cm diameter, Schleicher & Schuell, Germany) and held within a glass petri dish (10 cm diameter) supported on a perspex platform 9 cm above the floor and 5 cm beneath the bottom edge of the sticky trap,
731
0.5 m upwind of the thrips’ release platform. The wind tunnel was wiped down between each odor treatment using 50% acetone (BDH) and clean air was flushed through for 20 min before starting the next treatment. The average airflow was 0.12 m/s (TA5 thermal solid probe anemometer; Airflow Developments Ltd, High Wycombe, England). For the starvation experiments, 25 female F. occidentalis (mixed ages) were collected from a laboratory colony with an aspirator and placed into a plastic petri dish lid (7 cm diameter, 0.5 cm height) covered with parafilm with a few drops of water on the parafilm, in turn covered with a smaller petri dish (3 cm diameter). Thrips were collected 5 min prior to release for satiated thrips (0 h) or collected 1, 4 or 24 h prior to release (Experiment A); or 0, 24, 48 or 72 h prior to release (Experiment B); or 0, 1, 24, or 72 h prior to release (Experiment C), and held in an incubator at 25 1C with a 14:10 h light:dark cycle. Thrips were released into the wind tunnel by placing the petri dish on a perspex release platform (9 cm tall) and removing the parafilm. One dish with 20 or 25 thrips was used for each treatment. The number of thrips that flew and of those that flew the number that landed onto the sticky trap after 5 min were recorded. For each experiment thrips were exposed to one of four treatments; no cue (transparent sticky trap only); visual cue; odor cue; or both visual and odor cues. In experiments A and B the odor cue was 0.5 ml of p-anisaldehyde, in experiment C it was 1 ml of panisaldehyde. All starvation experiments were replicated four times for the 16 treatments (four starvation periods and four cues) per experiment. For experiments A and B the replicates were run over 8 days, with half of the treatments run each day, and a whole set run in each pair of days. Within each day, there were two sessions (morning and afternoon) each using four treatments, and all of the treatments within a session had the same odor cue (i.e., all with or all without the odor), giving a split-plot structure. The four treatments within each session were chosen so that the starving time by visual cue interaction was partially confounded with session: within a session there was one treatment using each of the four starving times, and two with and two without a visual cue. For starvation experiment C, the day was divided into four sessions of four treatments. There were four replicates of the 16 treatments (four starvation and four cues), with a whole replicate run in a single day. Within a session, all treatments were with or without the odor cue, and there was one treatment using each of the four starvation periods, two with and two without a visual cue. The order of treatments within a given session was randomized, and treatments used within each session were randomized according to the design. For the even-age experiment, females for the younger age group (2–3 days post-adult emergence) were collected the day prior to a wind tunnel experiment using an aspirator and placed in groups of 20 into small petri dishes
ARTICLE IN PRESS 732
M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
(7 cm diameter, 0.5 cm height) covered with parafilm with a few drops of water on the parafilm, in turn covered with a smaller petri dish (3 cm diameter). In addition, the day before a wind tunnel experiment, older female thrips (10–13 days post-adult emergence) were placed in groups of 20 into small petri dishes. Both age groups were held in an incubator and starved overnight. The experiment was replicated twice (i.e. 50 thrips per age group for each cue) with a replicate run in 1 day. As for starvation experiment C, the treatments were arranged such that all four cues were used in each half day, with two runs of each age in each half day. The order of treatments was randomized, and treatments used within each session were randomized according to the design. An experiment was undertaken to determine if the method of collecting and storing the thrips affected their flying behavior in the wind tunnel. Eight batches of 25 thrips held in separate petri dishes using the method described above, were compared with more than 150 thrips held overnight in a single perspex ring cage, then collected in batches of 25 thrips and placed into separate petri dishes 5 min prior to release into the wind tunnel. These two groups were in turn compared to thrips collected from the colony 5 min prior to release into the wind tunnel (satiated thrips). The batches of 25 thrips were released into a wind tunnel containing one of the four treatments (no cue, visual cue, odor cue, both cues) described above. The order of treatments was randomized, and treatments used within each session were randomized according to the design. The experiment was replicated twice. The number that flew and number to land on a sticky trap after 5 min was recorded. In all of the experiments described above the temperature was between 21 and 25 1C and relative humidity was between 40 and 50% in the wind tunnel.
significantly greater than zero on the logit scale, significantly more than 50% of the thrips chose the treated arm. For all experiments, 95% confidence limits for the percentage of thrips responding were calculated on the transformed (logit) scale, and then back-transformed. All analyses were carried out with GenStat (Genstat Committee 2003), and a level of 5% was used throughout to determine significance. 3. Results 3.1. Y-tube olfactometer experiment More thrips chose the odor-laden arm of the Y-tube olfactometer than the clean air arm (Po0.05) regardless of starvation period (58–77%, Fig. 1). The percentage of thrips to choose the odor-laden arm increased significantly from 58.7% for satiated thrips to 77% for thrips starved for 24 h (Po0.05, Fig. 1). There was little difference between satiated thrips (0 h) choosing the odor arm and those starved for 1 h (62.7%). Likewise there was little difference between those starved for 4 (76%) or 24 h (Fig. 1). 3.2. Wind tunnel olfactometer experiments
2.4. Statistical analysis
In the experiment testing for the effect of collection and storage of thrips, odor did not affect the percentage of thrips to fly or land on a sticky trap (P40.05). Therefore, data are summarized over the treatments with and without odor. There was no difference between the two groups of thrips starved for 24 h and either held in separate petri dishes or a single perspex ring cage (P40.05) in relation to the mean percentage of thrips to fly in the presence or absence of a visual cue or percentage to land on a trap of those that flew when it contained a visual cue (Table 1). However, more thrips starved for 24 h flew or landed on a
Data from all experiments were analyzed with a binomial generalised linear model with a logit link (McCullagh and Nelder, 1989). The number of thrips to respond out of the total number of thrips exposed to a cue was analyzed. For the Y-tube experiments, the response was the number of thrips to choose the odor arm out of the total number of choices recorded (25 per replicate). For the wind tunnel experiments, the number flying out of total thrips introduced into the tunnel (20 or 25 thrips) and the number on the trap out of thrips that flew were analyzed separately. Comparisons between treatments were made as contrasts within the analysis of deviance done as part of the analysis (McCullagh and Nelder, 1989), using F-tests based on the residual deviance. For experiments carried out over several sessions or days, differences between dates and time of day were examined similarly. For the Y-tube experiment, tests for preference for the treatment over the control were made with t-tests of the parameters estimated on the transformed (logit) scale. On the logit scale, 50% has a value of zero, so if the parameter for a treatment is
Fig. 1. The mean percentage of female F. occidentalis starved for different lengths of time to choose the odor-laden arm of a Y-tube olfactometer. Error bars are 95% confidence limits.
ARTICLE IN PRESS M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
733
Table 1 The mean percentage (and 95% confidence limits) of female F. occidentalis to fly or land on a sticky trap of those that flew when not starved (0 h), or starved for 24 h from an experiment to evaluate the influence of collection and storage techniques used in the starvation experiments Treatment
No visual cue Visual cue
0h
24 h, thrips held in separate batches
24 h, thrips held in a single ring cage
% flew
% on trap
% flew
% on trap
% flew
% on trap
6 (2.20) 9 (3.23)
0 (0.4) 7 (3.16)
61 (45.75) 69 (53.82)
2 (0.10) 39 (28.51)
60 (44.74) 69 (53.82)
1 (0.9) 43 (32.55)
sticky trap containing a yellow visual cue than satiated thrips collected from the colony 5 min before release in the wind tunnel (Table 1). Where no visual cue was present, very few thrips that flew landed on the trap and the percentage landing on the trap was similar for all treatments (P40.05). In all of the experiments described below, there were no strong differences between days or sessions within days. In the starvation experiments, the percentage of thrips that flew from the total number released into the wind tunnel changed significantly with the amount of time that thrips were starved (Po0.001, Fig. 2). The percentage of thrips that flew increased between 0 and 4 h (Fig. 2A). A similar percentage flew when starved for between 4 and 24 h (Fig 2A). Fewer thrips flew after 72 h compared to those starved for 48 or 24 h (Fig. 2B, C, respectively). The percentage of thrips to land on the sticky trap out of those that flew increased with starvation period when the visual cue was present (Po0.001), but starvation had little effect on this percentage when there was no visual cue present (Fig. 2). In the absence of a visual cue, the presence of an odor cue applied at 0.5 ml had little effect on the proportion of thrips to fly or land on the sticky trap (P40.4, Fig 2A, B). However, fewer thrips flew in the presence of 1 ml of the odor cue (P ¼ 0.002, Fig. 2C). For the even-age experiment, the percentage of thrips to land on the trap of those that flew was significantly affected by the age of the thrips (Po0.01) and the presence of a visual cue (Po0.001, Table 2). No thrips landed on the sticky trap in the absence of a yellow visual cue irrespective of thrips age or odor cue. When the visual cue was present, fewer younger thrips (38.8%) than older thrips (70.4%) landed on the trap (Table 2). Odor had no effect on the percentage of thrips to land on the trap (P40.91). A similar percentage of thrips flew (overall mean 44%) irrespective of age or cue (P40.4 for all parameters and their interactions). 4. Discussion The results from the present study indicate that well-fed female F. occidentalis are less responsive to visual and/or olfactory cues in a Y-tube olfactometer or wind tunnel than thrips starved and held in petri dishes or perspex ring cages with access to water only for a minimum of 4 h, and that the age of female thrips appears to influence their response
to a visual cue in a wind tunnel. For walking thrips observed in a Y-tube olfactometer, fewer satiated thrips or those starved for 1 h chose the odor-laden arm compared with thrips starved for at least 4 h, suggesting that hungry thrips were more receptive to an odor cue (Fig. 1). For flying thrips observed in a wind-tunnel, starved thrips flew to a much greater degree than satiated thrips, irrespective of the presence of an odor and visual cue. The ability of thrips to take flight was somewhat reduced after not feeding for 72 h compared with thrips starved for 24 or 48 h (Fig. 2B, C). The age of a thrips affected its response to a visual cue, with fewer young thrips (2–3 days post-adult emergence) landing on a yellow cue than older thrips (10–13 days post-adult emergence). These results emphasize the importance of a thrips’ physiological state in interpreting its behavioral responses to external stimuli. The proportion of thrips of different ages, or of those that were starved for at least 4 h, to take flight was not affected by the presence of the yellow visual cue, although, not unexpectedly, it did influence where they landed. What was unexpected was the greater proportion of older females (70.4%) than younger females (38.8%) landing on the trap of those that flew when it contained a yellow cue. Brevault and Quilici (1999) observed fewer young female (2 days post-adult emergence) N. cyanescens landed on an orange sphere stimulus than older females (10 days post- adult emergence). They also found that the percentage of females to visit an orange sphere increased with egg load. In our study, the degree of ovarian development of females from either age group was not ascertained, nor were the females isolated from males, so their mated state was unknown. The presence of 0.5 ml of the odor cue did not affect the proportion of thrips taking flight or influence where they landed. However, 1.0 ml of the odor cue reduced the proportion of thrips taking flight, and this was most apparent for thrips starved for 24 h (with no visual cue) or 72 h (with and without a visual cue), thus our results demonstrate that thrips respond differently to two different concentrations of chemical. Teulon et al. (1999; Western flower thrips and anisaldehyde), and Berry et al. (2005; New Zealand flower thrips and ethyl nicotinate), have also shown reduced thrips take-off in the presence of a chemical odor. These previous studies and current results demonstrate that volatile compounds can arrest thrips’ flight. The principal treatment effect in the Y-tube experiment and mixed-age wind tunnel experiments is most likely
ARTICLE IN PRESS M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
734
Fig. 2. The mean percentage of female F. occidentalis to fly (of those released into the wind tunnel) or land (of those that flew) on a sticky trap (with or without a yellow visual cue) after being starved for (A) 0, 1, 4 or 24 h (B) 0, 24, 48 or 72 h, or (C) 0, 1, 24 or 72 h. The volume of odor cue used in A and B was 0.5 ml and, in C, 1 ml.The error bars are the smallest and largest 95% confidence limits for means in the figure.
Table 2 The mean percentage (and 95% confidence limits) of female F. occidentalis 2–3 d and 10–13 d post-adult emergence to fly or of those that flew the mean percentage to land on a sticky trap with or without the visual cue Treatment
No visual cue Visual cue
2–3 d
10–13 d
% flew
% on trap
% flew
% on trap
46.7 (34.5,59.3) 40.8 (29.2,53.6)
0.0 (0.0,6.4) 38.8 (25.6,53.9)
42.4 (30.5,55.2) 46.6 (34.2,59.4)
0.0 (0.0,7.1) 70.4 (56.1,81.6)
ARTICLE IN PRESS M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
starvation. However, the behavior of the thrips held in petri dishes or ring cages with access to only water may have also been affected by differences in relative humidity, lack of small crevices to hide in, restricted flight, and absence of olfactory, visual and mechanical stimuli from plants that thrips in the rearing cages would have been exposed to. We therefore would include a control treatment of thrips held in Petri dishes or ring cages with access to both food and water in future experiments. There were some differences between the wind tunnel experiments A, B, and C in the percentage of thrips to land on a sticky trap out of those that flew (Fig. 2). The highest percentage of thrips to land on the trap was recorded in experiment A (up to 87%) compared to B and C (53% and 69%, respectively). This may have been in part due to when the experiments were done. Experiment A was undertaken in mid summer, while experiments B and C were late winter and early spring respectively. It may have also been due to the age structure of the population in the laboratory colony. We have observed numbers of thrips decrease markedly over the winter months even in a controlled temperature room (personal observation), so as the numbers increase, the population is possibly dominated by younger thrips. Colored sticky traps are commonly used to monitor thrips activity for pest management (Brødsgaard, 1993; Shipp, 1995; Shipp et al., 2000). If the behavior of female thrips in a wind tunnel can be extrapolated to the field, then our results imply, at least for the color yellow, that sticky boards will mainly catch ‘hungry’ thrips and will potentially catch more older thrips than younger thrips, at least for the ages tested in the present study. Thus, yellow sticky boards will be useful in monitoring the arrival of older invasive thrips (which are presumably hungry) into a crop, but may not be useful in monitoring the development of a thrips population on a crop until the food resource is limited. For example, Rhainds and Shipp (2003) observed an increase in the proportion of F. occidentalis females caught on blue sticky traps with an increase in senescent chrysanthemum inflorescences. The challenge for those researching and developing improved thrips traps will be to find visual and chemical cues that recently fed thrips of all ages will respond to. To achieve an improved thrips trap, we require more studies on the influence of intrinsic and extrinisic factors that influence thrips’ host-finding behaviors. For example, several studies have shown that more F. occidentalis were caught on blue sticky traps of a specific hue than on yellow or white traps (Brødsgaard, 1989; Gillespie and Vernon, 1990; Roditakis et al., 2001), so how starved thrips or thrips of different ages will respond to visual cues of varying colors requires further investigation. Likewise, more information on the potential physiological differences between females of different ages (i.e. ovarian development, mated state) is needed before any conclusions can be drawn regarding their response to external stimuli, and investigations into the responses of thrips to a range of
735
chemical concentrations will further assist in the understanding of thrips’ host-finding behavior. The results from the current study concur with previous wind-tunnel experiments showing that visual cues provide a strong directional stimulus to thrips (Teulon et al., 1999; Smits et al., 2000). Nevertheless, there is evidence from field studies that in the absence of a visual cue, more thrips can be caught in black-colored baited water traps than nonbaited water traps (Teulon et al., 1999). Thrips’ response to odor appears to involve chemokinesis, and because of the conditions and experimental design used within the wind tunnel in the present study and in previous studies (Teulon et al., 1999; Smits et al., 2000; Berry et al., 2005), any chemically mediated response may be underestimated. The potential to exploit the behavior of thrips pest species toward visual cues has already been realised through the use of colored sticky traps for monitoring thrips populations. However, their use could be enhanced, and the use of olfactory cues in thrips management strategies could become a reality, with greater knowledge of the factors that influence thrips flight, such as ovarian development and mated state. Acknowledgements A postdoctoral fellowship from AGMARDT provided financial assistance for M. M. Davidson. The work was in part funded by the Foundation for Research, Science and Technology. References Barton Browne, L., 1993. Physiologically induced changes in resourceoriented behaviour. Annual Review of Entomology 38, 1–25. Bernays, E.A., Chapman, R.F., 1994. Host-plant Selection by Phytophagous Insects. Chapman & Hall, New York. Berry, N., Butler, R.C., Teulon, D.A.J., 2005. Responses of New Zealand flower thrips (Thrips obscuratus) to synthetic and natural stimuli (odour and colour) in a wind tunnel. New Zealand Journal of Crop and Horticultural Science 34, 121–129. Brevault, T., Quilici, S., 1999. Factors affecting behavioural responses to visual stimuli in the tomato fruit fly, Neoceratitis cyanescens. Physiological Entomology 24, 333–338. Brødsgaard, H.F., 1989. Frankliniella occidentalis (Thysanoptera: Thripidae)—a new pest in a Danish glasshouse. A review. Tidsskrift for Planteavl 93, 83–91. Brødsgaard, H.F., 1990. The effect of anisaldehyde as a scent attractant for Frankliniella occidentalis (Thysanoptera: Thripidae) and the response mechanism involved. SROP/WPRS Bulletin XIII, 36–38. Brødsgaard, H.F., 1993. Monitoring thrips in glasshouse pot plant crops by means of blue sticky traps. IOBC/WPRS Bulletin 16, 29–32. GenStat Committee., 2003. GenStat Release 7.1 Reference Manual, Parts 1-3, VSN International, Oxford. Gillespie, D.R., Vernon, R.S., 1990. Trap catch of western flower thrips (Thysanoptera: Thripidae) as affected by color and height of sticky traps in mature greenhouse cucumber crops. Journal of Economic Entomology 83, 971–975. Hamilton, J.G.C., Hall, D., Kirk, W.D.J., 2005. Identification of a maleproduced aggregation pheromone in the western flower thrips Frankliniella occidentalis. Journal of Chemical Ecology 31, 1369–1379. Howlett, F.M., 1914. A trap for thrips. Journal of Economic Biology IX, 21–23.
ARTICLE IN PRESS 736
M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
Koschier, E.H., De Kogel, W.J., Visser, J.H., 2000. Assessing the attractiveness of volatile plant compounds to Frankliniella occidentalis. Journal of Chemical Ecology 26, 2643–2655. Lambert, M., 1981. Observations on the laboratory take-off behavior of the desert locust, Schistocerca gregaria (Forska˚l) (Orthoptera: Acrididae). Bulletin of Entomological Research 71, 481–498. Lewis, T., 1997a. Thrips as Crop Pests. CAB International, Wallingford. Lewis, T., 1997b. Flight and dispersal. In: Lewis, T. (Ed.), Thrips as Crop Pests. CAB International, Wallingford, pp. 175–196. Loomans, A.J.M., Murai, T., 1997. Culturing thrips and parasitoids. In: Lewis, T. (Ed.), Thrips as Crop Pests. CAB International, Wallingford, pp. 477–504. MacDonald, K.M., Hamilton, J.G.C., Jacobson, R., Kirk, W.D.J., 2002. Effects of alarm pheromone on landing and take-off by adult western flower thrips. Entomologia Experimentalis et Applicata 103, 279–282. MacQuarrie, C., Boiteau, G., 2003. Effect of diet and feeding history on flight of Colorado potato beetle, Leptinotarsa decemlineata. Entomologia Experimentalis et Applicata 107, 207–213. McCullagh, P., Nelder, J.A., 1989. Generalized Linear Models. Chapman & Hall, London. Morgan, A.C., Crumb, S.E., 1928. Notes on the chemotropic responses of certain insects. Journal of Economic Entomology 21, 913–920. O’Leary, A.T., Kirk, W.D.J., 2004. The effect of wavelength on take-off in the western flower thrips. Acta Phytopathologica et Entomologica Hungarica 39, 123–129. Pearsall, I.A., 2002. Daily flight activity of the western flower thrips (Thysan., Thripidae) in nectarine orchards in British Columbia, Canada. Journal of Applied Entomology 126, 293–302. Rhainds, M., Shipp, L., 2003. Dispersal of adult western flower thrips (Thysanoptera: Thripidae) on chrysanthemum plants: impact of
feeding-induced senescence of inflorescences. Environmental Entomology 32, 1056–1065. Shipp, J.L., 1995. Monitoring of western flower thrips on glasshouse and vegetable crops. In: Parker, B.L., Skinner, M., Lewis, T. (Eds.), Thrips Biology and Management. Plenum Press, New York, pp. 547–555. Shipp, J.L., Zhang, Y., 1999. Using greenhouse microclimate to improve the efficacy of insecticide application for Frankliniella occidentalis (Thysanoptera: Thripidae). Journal of Economic Entomology 92, 201–206. Shipp, J.L., Wang, K., Binns, M.R., 2000. Economic injury levels for western flower thrips (Thysanoptera: Thripidae) on greenhouse cucumber. Journal of Economic Entomology 93, 1732–1740. Smits, P.H., Van Deventer, P., De Kogel, W.J., 2000. Western flower thrips: reactions to odors and colors. Proceedings of Experimental and Applied Entomology 11, 175–180. Teulon, D.A.J., Hollister, B., Cameron, E.A., 1993a. Behavioural responses of western flower thrips to anisaldehyde, and applications for trapping in greenhouses. IOBC/WPRS Bulletin 16, 177–180. Teulon, D.A.J., Penman, D.R., Ramakers, P.M.J., 1993b. Volatile chemicals for thrips (Thysanoptera: Thripidae) host-finding and applications for thrips pest management. Journal of Economic Entomology 86, 1405–1415. Teulon, D.A.J., Hollister, B., Butler, R.C., Cameron, E.A., 1999. Colour and odor responses of flying western flower thrips: wind tunnel and greenhouse experiments. Entomologia Experimentalis et Applicata 93, 9–19. Ullman, D.E., Sherwood, J.L., German, T.L., 1997. Thrips as vectors of plant pathogens. In: Lewis, T. (Ed.), Thrips as Crop Pests. CAB International, Wallingford, pp. 539–565.
Journal of Insect Physiology 52 (2006) 729–736 www.elsevier.com/locate/jinsphys
Starvation period and age affect the response of female Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae) to odor and visual cues Melanie M. Davidson, Ruth C. Butler, David A.J. Teulon New Zealand Institute for Crop & Food Research Limited, Private Bag 4704, Christchurch, New Zealand Received 11 January 2006; received in revised form 19 March 2006; accepted 27 March 2006
Abstract The effects of starvation or age on the walking or flying response of female Frankliniella occidentalis to visual and/or odor cues in two types of olfactometer were examined in the laboratory. The response of walking thrips starved for 0, 1, 4, or 24 h to an odor cue (1 ml of 10% p-anisaldehyde) was examined in a Y-tube olfactometer. The take-off and landing response of thrips (unknown age) starved for 0, 1, 4, 24, 48 or 72 h, or of thrips of different ages (2–3 days or 10–13 days post-adult emergence) starved for 24 h, to a visual cue (98 cm2 yellow sticky trap) and/or an odor cue (0.5 or 1.0 ml p-anisaldehyde) was examined in a wind tunnel. More thrips walked up the odorladen arm in the Y-tube when starved for at least 4 h (76%) than satiated thrips (58.7%) or those starved for 1 h (62.7%, Po0.05). In the wind tunnel experiments the percentage of thrips to fly or land on the sticky trap increased between satiated thrips (7.3% to fly, 3.3% on trap) and those starved for 4 h (81.2% to fly, 29% on trap) and decreased between thrips starved for 48 (74.5% to fly, 23% on trap) and 72 h (56.5% to fly, 15.5% on trap, Po0.05). Fewer younger thrips (38.8%) landed on a sticky trap containing a yellow visual cue of, those that flew, than older thrips (70.4%, Po0.05), although a similar percentage of thrips flew regardless of age or type of cue present in the wind tunnel (average 44%, P40.05). r 2006 Elsevier Ltd. All rights reserved. Keywords: Frankliniella occidentalis; Host cues; Intrinsic factors; Walking; Flight
1. Introduction Western flower thrips, Frankliniella occidentalis (Pergande), is a pest on a variety of vegetable, fruit and ornamental crops (Lewis, 1997a). F. occidentalis can cause extensive damage to its host plant either directly through feeding or oviposition, or indirectly through the transmission of tospoviruses (Ullman et al., 1997). This pest is difficult to manage because of its small size (o2 mm long), cryptic habits, and resistance to a number of insecticides (Brødsgaard, 1989). Colored sticky traps have been used to monitor populations of F. occidentalis (Brødsgaard, 1993; Shipp, 1995; Shipp et al., 2000). However, colored sticky traps only catch flying thrips. Factors that increase the number Corresponding author. Tel.: +64 3 325 6400; fax: +64 3 325 2074.
E-mail address: [email protected] (M.M. Davidson). 0022-1910/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.jinsphys.2006.03.013
of thrips to fly could be used to improve monitoring, for example by allowing earlier detection of thrips’ presence, and could even increase their overall exposure to biological or chemical control strategies. Thrips’ take off, flight duration and landing may be influenced by several extrinsic and intrinsic factors. Likely extrinsic factors include temperature, light, wind and, to a lesser extent, relative humidity (Lewis, 1997b), as well as pheromones and kairomones. A number of studies have examined the influence that these factors have on flight (temperature, e.g. Shipp and Zhang 1999; Pearsall, 2002; light, e.g. O’Leary and Kirk, 2004; air flow, e.g. Teulon et al., 1999; pheromones, e.g. MacDonald et al., 2002; Hamilton, et al., 2005; and kairomones, e.g., Brødsgaard, 1990; Teulon et al., 1993a,b; 1999; Smits et al., 2000). Intrinsic factors that may influence thrips’ responses toward host cues, such as ovarian development, their mated state, or the senescences of host plants (Lewis,
ARTICLE IN PRESS 730
M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
1997b) have received less attention. However, the physiological state of insects could be expected to influence their behavioral responses toward external stimuli (Barton Browne, 1993). The time since feeding can be particularly important in altering an insect’s behavior toward host cues (Bernays and Chapman, 1994). For example, a higher number of starved than unstarved Colorado potato beetle (Leptinotarsa decemlineata (Say)) have been observed to fly from a summer population (MacQuarrie and Boiteau, 2003), and a greater percentage of starved female desert locust (Schistocerca gregaria (Forsa˚l)) take flight compared with those with full foreguts (Lambert, 1981). An adult female insect’s age could also affect its response to stimuli through physiological changes, such as ovarian development (Barton Browne, 1993). For example, Brevault and Quilici (1999) found that less than 5% of young (2 days post-adult emergence) female tomato fruit fly (Neoceratitis cyanescens (Bezzi)) landed on an orange sphere stimulus, compared with around 65% of older females (10 days postadult emergence). This study reports for the first time, the effect of starvation or age on the responses of female F. occidentalis to artificial host cues. The aims of this study were: (i) to compare the walking or flying response of thrips starved for different lengths of time to artificial host cues (visual and/or odor) in a Y-tube or low speed laminar flow wind tunnel; and (ii) to investigate the effect of age on the flying response of female F. occidentalis to artificial host cues in a wind tunnel. The odor used in the following experiments was p-anisaldehyde. An anisaldehyde compound was one of the first volatile compounds shown to attract higher numbers of thrips to traps baited with this compound compared to unbaited traps (Howlett, 1914; Morgan and Crumb, 1928). Greenhouse trials using p-anisaldehyde baited or unbaited traps demonstrated higher catches of F. occidentalis in the former (Teulon et al., 1993a,b). Previous experiments evaluating the response of F. occidentalis females to volatile compounds in a Y-tube olfactometer showed p-anisaldehyde elicted one of the strongest responses out of a range of compounds tested (Koschier et al., 2000). 2. Materials and Methods 2.1. Insects All F. occidentalis used in the following starvation experiments were from a colony maintained on potted flowering chrysanthemums, Dendranthema grandiflora (cv. Onyx time yellow). The colony was established in 2001 from wild thrips collected from commercial greenhouses (Auckland, New Zealand). The plants used for the colony were held in three temperature-controlled perspex cages (five plants per cage) with a temperature maintained at 25 1C under a 14:10 h light:dark cycle. A fresh uninfested plant was placed in each cage every 3 days and the oldest plant was removed at the same time.
Even-age female thrips were obtained using a green bean (Phaseus vulgaris L.) rearing method adapted from Loomans and Murai (1997). Briefly, this involved placing 200–300 adult males and females in a glass container (500 ml) that held five to six green beans and a 1.5 ml eppendorf tube with a hole in the lid containing approximately 0.2 g of ground bee-collected pollen. The bottom of the container was lined with a square of greenhouse plant matting (100 cm2) and filter paper (Whatman No. 1, 9 cm diameter). A 5 cm diameter hole was cut in the center of the metal screw lid, which was used to secure mesh (20 mm) over the opening of the container. After 3 days, the egg-laden bean pods were replaced with six fresh beans and the egg-laden beans were placed on top of filter paper and plant matting, along with bee-collected pollen in another 500 ml glass container. The glass containers were held in an incubator at a temperature of 2572 1C and a 14:10 h light:dark cycle. The beans were replaced every 5 days, and adults began to emerge after approximately 21 days. In order to obtain enough thrips for an experiment, adult females (along with males) were collected every 2 days for the older age group (10–13 days post-adult emergence) and held for 10 days in glass jars containing filter paper, fresh beans and bee-collected pollen, in the incubator. 2.2. Y-tube olfactometer experiment The response of walking female F. occidentalis starved for different lengths of time toward a volatile compound was evaluated in a glass Y-tube olfactometer following the method described by Koschier et al. (2000). The Y-tube has two branching arms at an angle of approximately 451 leading from a single tube, and each section is 60 mm long, with an internal diameter of 5 mm.The arms of the Y-tube were connected to glass Wheaton Micro Kits adapters that were in turn attached to 4 ml glass vials each containing a 1 cm2 piece of filter paper (Whatman No. 1). The Y-tube and Wheaton apparatus were placed in a grey box (to prevent external stimuli influencing thrips’ behavior), located in a darkened, air-conditioned room (22 1C with a 31 range). The Y-tube was positioned at an inclining angle of 251 and illuminated from above by a halogen lamp (780 lux, Beha digital Lux Hi Tester 93-1065L, Germany). One microlitre of the volatile compound, 10% p-anisaldehyde diluted in hexane (95%, BDH), was applied to filter paper in one vial, while 1 ml of hexane was applied to filter paper held in the second vial. Air was drawn through activated charcoal before entering the Wheaton apparatus and Y-tube using a suction pump (AR Harris Co. Ltd, New Zealand), producing an airflow of 5 cm/s through each arm and of 10 cm/s at the base of the Y-tube. Clean air was drawn through the Y-tube for 30 min before introducing the first thrips at the beginning of each bioassay. Connections between the activated charcoal, Wheaton apparatus, Y-tube and suction pump were with silicone tubing.
ARTICLE IN PRESS M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
Female thrips of unknown age were collected from the colony using an aspirator and held in a perspex ring cage (8 cm diameter, 5 cm high) (Loomans and Murai, 1997) with access to water only for 0, 1, 4 or 24 h prior to introducing them into the Y-tube olfactometer. An individual female thrips was released into the base of the Y-tube using a small aspirator. Most thrips walked up into the Y-tube within a few seconds, at which time the silicone air suction tubing was reconnected to the base of the glass Y-tube and timing of an individual was started. Thrips were observed as they walked up the stem of the Y-tube and entered either arm. When the thrips reached the far end of the respective arm their choice was recorded. On the rare occasion a thrips failed to make a choice after 3 min the thrips was removed and replaced with another thrips. After every five thrips that made a choice between the odor or clean air the Y-tube and Wheaton apparatus were rotated 1801 to avoid position effects. One replicate was completed after 25 thrips had chosen either arm, after which the Y-tube and Wheaton apparatus were thoroughly cleaned with acetone (99.5%, BDH) and allowed to dry before repeating the experiment. The experiment was replicated three times per starvation period to give a total of 75 choices per starvation period. 2.3. Wind tunnel olfactometer experiments A low-speed laminar air flow wind tunnel, described by Berry et al. (2005), was used to measure the flying response of thrips starved for different periods (0, 1, 4, 24, 48, or 72 h). The perspex observation chamber measured 0.58 wide by 0.55 m in height by 1.7 m in length and was illuminated by five 36W 84 cool lights alternating with five 36W 96 true light fluorescent tubes, producing an average of 10,000 lux (Beha, Digital Lux Hi Tester 93-1065L, Germany). During bioassays, room air, filtered through activated charcoal, was passed through the chamber in a laminar flow. The frame and floor of the wind tunnel were painted a charcoal gray to facilitate sighting of thrips. Sticky traps were constructed using two square glass plates (10 10 cm, 2 mm thick) with Tangle traps (Grand Rapids, Michigan) applied to the outward facing sides. A 98 cm2 yellow visual cue was obtained by removing the glue from commercially available sticky traps (Horriver, Koppert, the Netherlands). Visual cues were sandwiched between the glass plates. When no visual cue was required, the sticky trap was used without placing anything between the glass plates. In the wind tunnel, the sticky trap was hung from a clip attached to a retort stand, with the top edge of the sticky trap 20 cm from the roof of the wind tunnel, 0.5 m upwind of the thrips’ perspex release platform. The volatile odor (p-anisaldehyde, 0.5 or 1.0 ml) was applied to black filter paper (9 cm diameter, Schleicher & Schuell, Germany) and held within a glass petri dish (10 cm diameter) supported on a perspex platform 9 cm above the floor and 5 cm beneath the bottom edge of the sticky trap,
731
0.5 m upwind of the thrips’ release platform. The wind tunnel was wiped down between each odor treatment using 50% acetone (BDH) and clean air was flushed through for 20 min before starting the next treatment. The average airflow was 0.12 m/s (TA5 thermal solid probe anemometer; Airflow Developments Ltd, High Wycombe, England). For the starvation experiments, 25 female F. occidentalis (mixed ages) were collected from a laboratory colony with an aspirator and placed into a plastic petri dish lid (7 cm diameter, 0.5 cm height) covered with parafilm with a few drops of water on the parafilm, in turn covered with a smaller petri dish (3 cm diameter). Thrips were collected 5 min prior to release for satiated thrips (0 h) or collected 1, 4 or 24 h prior to release (Experiment A); or 0, 24, 48 or 72 h prior to release (Experiment B); or 0, 1, 24, or 72 h prior to release (Experiment C), and held in an incubator at 25 1C with a 14:10 h light:dark cycle. Thrips were released into the wind tunnel by placing the petri dish on a perspex release platform (9 cm tall) and removing the parafilm. One dish with 20 or 25 thrips was used for each treatment. The number of thrips that flew and of those that flew the number that landed onto the sticky trap after 5 min were recorded. For each experiment thrips were exposed to one of four treatments; no cue (transparent sticky trap only); visual cue; odor cue; or both visual and odor cues. In experiments A and B the odor cue was 0.5 ml of p-anisaldehyde, in experiment C it was 1 ml of panisaldehyde. All starvation experiments were replicated four times for the 16 treatments (four starvation periods and four cues) per experiment. For experiments A and B the replicates were run over 8 days, with half of the treatments run each day, and a whole set run in each pair of days. Within each day, there were two sessions (morning and afternoon) each using four treatments, and all of the treatments within a session had the same odor cue (i.e., all with or all without the odor), giving a split-plot structure. The four treatments within each session were chosen so that the starving time by visual cue interaction was partially confounded with session: within a session there was one treatment using each of the four starving times, and two with and two without a visual cue. For starvation experiment C, the day was divided into four sessions of four treatments. There were four replicates of the 16 treatments (four starvation and four cues), with a whole replicate run in a single day. Within a session, all treatments were with or without the odor cue, and there was one treatment using each of the four starvation periods, two with and two without a visual cue. The order of treatments within a given session was randomized, and treatments used within each session were randomized according to the design. For the even-age experiment, females for the younger age group (2–3 days post-adult emergence) were collected the day prior to a wind tunnel experiment using an aspirator and placed in groups of 20 into small petri dishes
ARTICLE IN PRESS 732
M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
(7 cm diameter, 0.5 cm height) covered with parafilm with a few drops of water on the parafilm, in turn covered with a smaller petri dish (3 cm diameter). In addition, the day before a wind tunnel experiment, older female thrips (10–13 days post-adult emergence) were placed in groups of 20 into small petri dishes. Both age groups were held in an incubator and starved overnight. The experiment was replicated twice (i.e. 50 thrips per age group for each cue) with a replicate run in 1 day. As for starvation experiment C, the treatments were arranged such that all four cues were used in each half day, with two runs of each age in each half day. The order of treatments was randomized, and treatments used within each session were randomized according to the design. An experiment was undertaken to determine if the method of collecting and storing the thrips affected their flying behavior in the wind tunnel. Eight batches of 25 thrips held in separate petri dishes using the method described above, were compared with more than 150 thrips held overnight in a single perspex ring cage, then collected in batches of 25 thrips and placed into separate petri dishes 5 min prior to release into the wind tunnel. These two groups were in turn compared to thrips collected from the colony 5 min prior to release into the wind tunnel (satiated thrips). The batches of 25 thrips were released into a wind tunnel containing one of the four treatments (no cue, visual cue, odor cue, both cues) described above. The order of treatments was randomized, and treatments used within each session were randomized according to the design. The experiment was replicated twice. The number that flew and number to land on a sticky trap after 5 min was recorded. In all of the experiments described above the temperature was between 21 and 25 1C and relative humidity was between 40 and 50% in the wind tunnel.
significantly greater than zero on the logit scale, significantly more than 50% of the thrips chose the treated arm. For all experiments, 95% confidence limits for the percentage of thrips responding were calculated on the transformed (logit) scale, and then back-transformed. All analyses were carried out with GenStat (Genstat Committee 2003), and a level of 5% was used throughout to determine significance. 3. Results 3.1. Y-tube olfactometer experiment More thrips chose the odor-laden arm of the Y-tube olfactometer than the clean air arm (Po0.05) regardless of starvation period (58–77%, Fig. 1). The percentage of thrips to choose the odor-laden arm increased significantly from 58.7% for satiated thrips to 77% for thrips starved for 24 h (Po0.05, Fig. 1). There was little difference between satiated thrips (0 h) choosing the odor arm and those starved for 1 h (62.7%). Likewise there was little difference between those starved for 4 (76%) or 24 h (Fig. 1). 3.2. Wind tunnel olfactometer experiments
2.4. Statistical analysis
In the experiment testing for the effect of collection and storage of thrips, odor did not affect the percentage of thrips to fly or land on a sticky trap (P40.05). Therefore, data are summarized over the treatments with and without odor. There was no difference between the two groups of thrips starved for 24 h and either held in separate petri dishes or a single perspex ring cage (P40.05) in relation to the mean percentage of thrips to fly in the presence or absence of a visual cue or percentage to land on a trap of those that flew when it contained a visual cue (Table 1). However, more thrips starved for 24 h flew or landed on a
Data from all experiments were analyzed with a binomial generalised linear model with a logit link (McCullagh and Nelder, 1989). The number of thrips to respond out of the total number of thrips exposed to a cue was analyzed. For the Y-tube experiments, the response was the number of thrips to choose the odor arm out of the total number of choices recorded (25 per replicate). For the wind tunnel experiments, the number flying out of total thrips introduced into the tunnel (20 or 25 thrips) and the number on the trap out of thrips that flew were analyzed separately. Comparisons between treatments were made as contrasts within the analysis of deviance done as part of the analysis (McCullagh and Nelder, 1989), using F-tests based on the residual deviance. For experiments carried out over several sessions or days, differences between dates and time of day were examined similarly. For the Y-tube experiment, tests for preference for the treatment over the control were made with t-tests of the parameters estimated on the transformed (logit) scale. On the logit scale, 50% has a value of zero, so if the parameter for a treatment is
Fig. 1. The mean percentage of female F. occidentalis starved for different lengths of time to choose the odor-laden arm of a Y-tube olfactometer. Error bars are 95% confidence limits.
ARTICLE IN PRESS M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
733
Table 1 The mean percentage (and 95% confidence limits) of female F. occidentalis to fly or land on a sticky trap of those that flew when not starved (0 h), or starved for 24 h from an experiment to evaluate the influence of collection and storage techniques used in the starvation experiments Treatment
No visual cue Visual cue
0h
24 h, thrips held in separate batches
24 h, thrips held in a single ring cage
% flew
% on trap
% flew
% on trap
% flew
% on trap
6 (2.20) 9 (3.23)
0 (0.4) 7 (3.16)
61 (45.75) 69 (53.82)
2 (0.10) 39 (28.51)
60 (44.74) 69 (53.82)
1 (0.9) 43 (32.55)
sticky trap containing a yellow visual cue than satiated thrips collected from the colony 5 min before release in the wind tunnel (Table 1). Where no visual cue was present, very few thrips that flew landed on the trap and the percentage landing on the trap was similar for all treatments (P40.05). In all of the experiments described below, there were no strong differences between days or sessions within days. In the starvation experiments, the percentage of thrips that flew from the total number released into the wind tunnel changed significantly with the amount of time that thrips were starved (Po0.001, Fig. 2). The percentage of thrips that flew increased between 0 and 4 h (Fig. 2A). A similar percentage flew when starved for between 4 and 24 h (Fig 2A). Fewer thrips flew after 72 h compared to those starved for 48 or 24 h (Fig. 2B, C, respectively). The percentage of thrips to land on the sticky trap out of those that flew increased with starvation period when the visual cue was present (Po0.001), but starvation had little effect on this percentage when there was no visual cue present (Fig. 2). In the absence of a visual cue, the presence of an odor cue applied at 0.5 ml had little effect on the proportion of thrips to fly or land on the sticky trap (P40.4, Fig 2A, B). However, fewer thrips flew in the presence of 1 ml of the odor cue (P ¼ 0.002, Fig. 2C). For the even-age experiment, the percentage of thrips to land on the trap of those that flew was significantly affected by the age of the thrips (Po0.01) and the presence of a visual cue (Po0.001, Table 2). No thrips landed on the sticky trap in the absence of a yellow visual cue irrespective of thrips age or odor cue. When the visual cue was present, fewer younger thrips (38.8%) than older thrips (70.4%) landed on the trap (Table 2). Odor had no effect on the percentage of thrips to land on the trap (P40.91). A similar percentage of thrips flew (overall mean 44%) irrespective of age or cue (P40.4 for all parameters and their interactions). 4. Discussion The results from the present study indicate that well-fed female F. occidentalis are less responsive to visual and/or olfactory cues in a Y-tube olfactometer or wind tunnel than thrips starved and held in petri dishes or perspex ring cages with access to water only for a minimum of 4 h, and that the age of female thrips appears to influence their response
to a visual cue in a wind tunnel. For walking thrips observed in a Y-tube olfactometer, fewer satiated thrips or those starved for 1 h chose the odor-laden arm compared with thrips starved for at least 4 h, suggesting that hungry thrips were more receptive to an odor cue (Fig. 1). For flying thrips observed in a wind-tunnel, starved thrips flew to a much greater degree than satiated thrips, irrespective of the presence of an odor and visual cue. The ability of thrips to take flight was somewhat reduced after not feeding for 72 h compared with thrips starved for 24 or 48 h (Fig. 2B, C). The age of a thrips affected its response to a visual cue, with fewer young thrips (2–3 days post-adult emergence) landing on a yellow cue than older thrips (10–13 days post-adult emergence). These results emphasize the importance of a thrips’ physiological state in interpreting its behavioral responses to external stimuli. The proportion of thrips of different ages, or of those that were starved for at least 4 h, to take flight was not affected by the presence of the yellow visual cue, although, not unexpectedly, it did influence where they landed. What was unexpected was the greater proportion of older females (70.4%) than younger females (38.8%) landing on the trap of those that flew when it contained a yellow cue. Brevault and Quilici (1999) observed fewer young female (2 days post-adult emergence) N. cyanescens landed on an orange sphere stimulus than older females (10 days post- adult emergence). They also found that the percentage of females to visit an orange sphere increased with egg load. In our study, the degree of ovarian development of females from either age group was not ascertained, nor were the females isolated from males, so their mated state was unknown. The presence of 0.5 ml of the odor cue did not affect the proportion of thrips taking flight or influence where they landed. However, 1.0 ml of the odor cue reduced the proportion of thrips taking flight, and this was most apparent for thrips starved for 24 h (with no visual cue) or 72 h (with and without a visual cue), thus our results demonstrate that thrips respond differently to two different concentrations of chemical. Teulon et al. (1999; Western flower thrips and anisaldehyde), and Berry et al. (2005; New Zealand flower thrips and ethyl nicotinate), have also shown reduced thrips take-off in the presence of a chemical odor. These previous studies and current results demonstrate that volatile compounds can arrest thrips’ flight. The principal treatment effect in the Y-tube experiment and mixed-age wind tunnel experiments is most likely
ARTICLE IN PRESS M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
734
Fig. 2. The mean percentage of female F. occidentalis to fly (of those released into the wind tunnel) or land (of those that flew) on a sticky trap (with or without a yellow visual cue) after being starved for (A) 0, 1, 4 or 24 h (B) 0, 24, 48 or 72 h, or (C) 0, 1, 24 or 72 h. The volume of odor cue used in A and B was 0.5 ml and, in C, 1 ml.The error bars are the smallest and largest 95% confidence limits for means in the figure.
Table 2 The mean percentage (and 95% confidence limits) of female F. occidentalis 2–3 d and 10–13 d post-adult emergence to fly or of those that flew the mean percentage to land on a sticky trap with or without the visual cue Treatment
No visual cue Visual cue
2–3 d
10–13 d
% flew
% on trap
% flew
% on trap
46.7 (34.5,59.3) 40.8 (29.2,53.6)
0.0 (0.0,6.4) 38.8 (25.6,53.9)
42.4 (30.5,55.2) 46.6 (34.2,59.4)
0.0 (0.0,7.1) 70.4 (56.1,81.6)
ARTICLE IN PRESS M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
starvation. However, the behavior of the thrips held in petri dishes or ring cages with access to only water may have also been affected by differences in relative humidity, lack of small crevices to hide in, restricted flight, and absence of olfactory, visual and mechanical stimuli from plants that thrips in the rearing cages would have been exposed to. We therefore would include a control treatment of thrips held in Petri dishes or ring cages with access to both food and water in future experiments. There were some differences between the wind tunnel experiments A, B, and C in the percentage of thrips to land on a sticky trap out of those that flew (Fig. 2). The highest percentage of thrips to land on the trap was recorded in experiment A (up to 87%) compared to B and C (53% and 69%, respectively). This may have been in part due to when the experiments were done. Experiment A was undertaken in mid summer, while experiments B and C were late winter and early spring respectively. It may have also been due to the age structure of the population in the laboratory colony. We have observed numbers of thrips decrease markedly over the winter months even in a controlled temperature room (personal observation), so as the numbers increase, the population is possibly dominated by younger thrips. Colored sticky traps are commonly used to monitor thrips activity for pest management (Brødsgaard, 1993; Shipp, 1995; Shipp et al., 2000). If the behavior of female thrips in a wind tunnel can be extrapolated to the field, then our results imply, at least for the color yellow, that sticky boards will mainly catch ‘hungry’ thrips and will potentially catch more older thrips than younger thrips, at least for the ages tested in the present study. Thus, yellow sticky boards will be useful in monitoring the arrival of older invasive thrips (which are presumably hungry) into a crop, but may not be useful in monitoring the development of a thrips population on a crop until the food resource is limited. For example, Rhainds and Shipp (2003) observed an increase in the proportion of F. occidentalis females caught on blue sticky traps with an increase in senescent chrysanthemum inflorescences. The challenge for those researching and developing improved thrips traps will be to find visual and chemical cues that recently fed thrips of all ages will respond to. To achieve an improved thrips trap, we require more studies on the influence of intrinsic and extrinisic factors that influence thrips’ host-finding behaviors. For example, several studies have shown that more F. occidentalis were caught on blue sticky traps of a specific hue than on yellow or white traps (Brødsgaard, 1989; Gillespie and Vernon, 1990; Roditakis et al., 2001), so how starved thrips or thrips of different ages will respond to visual cues of varying colors requires further investigation. Likewise, more information on the potential physiological differences between females of different ages (i.e. ovarian development, mated state) is needed before any conclusions can be drawn regarding their response to external stimuli, and investigations into the responses of thrips to a range of
735
chemical concentrations will further assist in the understanding of thrips’ host-finding behavior. The results from the current study concur with previous wind-tunnel experiments showing that visual cues provide a strong directional stimulus to thrips (Teulon et al., 1999; Smits et al., 2000). Nevertheless, there is evidence from field studies that in the absence of a visual cue, more thrips can be caught in black-colored baited water traps than nonbaited water traps (Teulon et al., 1999). Thrips’ response to odor appears to involve chemokinesis, and because of the conditions and experimental design used within the wind tunnel in the present study and in previous studies (Teulon et al., 1999; Smits et al., 2000; Berry et al., 2005), any chemically mediated response may be underestimated. The potential to exploit the behavior of thrips pest species toward visual cues has already been realised through the use of colored sticky traps for monitoring thrips populations. However, their use could be enhanced, and the use of olfactory cues in thrips management strategies could become a reality, with greater knowledge of the factors that influence thrips flight, such as ovarian development and mated state. Acknowledgements A postdoctoral fellowship from AGMARDT provided financial assistance for M. M. Davidson. The work was in part funded by the Foundation for Research, Science and Technology. References Barton Browne, L., 1993. Physiologically induced changes in resourceoriented behaviour. Annual Review of Entomology 38, 1–25. Bernays, E.A., Chapman, R.F., 1994. Host-plant Selection by Phytophagous Insects. Chapman & Hall, New York. Berry, N., Butler, R.C., Teulon, D.A.J., 2005. Responses of New Zealand flower thrips (Thrips obscuratus) to synthetic and natural stimuli (odour and colour) in a wind tunnel. New Zealand Journal of Crop and Horticultural Science 34, 121–129. Brevault, T., Quilici, S., 1999. Factors affecting behavioural responses to visual stimuli in the tomato fruit fly, Neoceratitis cyanescens. Physiological Entomology 24, 333–338. Brødsgaard, H.F., 1989. Frankliniella occidentalis (Thysanoptera: Thripidae)—a new pest in a Danish glasshouse. A review. Tidsskrift for Planteavl 93, 83–91. Brødsgaard, H.F., 1990. The effect of anisaldehyde as a scent attractant for Frankliniella occidentalis (Thysanoptera: Thripidae) and the response mechanism involved. SROP/WPRS Bulletin XIII, 36–38. Brødsgaard, H.F., 1993. Monitoring thrips in glasshouse pot plant crops by means of blue sticky traps. IOBC/WPRS Bulletin 16, 29–32. GenStat Committee., 2003. GenStat Release 7.1 Reference Manual, Parts 1-3, VSN International, Oxford. Gillespie, D.R., Vernon, R.S., 1990. Trap catch of western flower thrips (Thysanoptera: Thripidae) as affected by color and height of sticky traps in mature greenhouse cucumber crops. Journal of Economic Entomology 83, 971–975. Hamilton, J.G.C., Hall, D., Kirk, W.D.J., 2005. Identification of a maleproduced aggregation pheromone in the western flower thrips Frankliniella occidentalis. Journal of Chemical Ecology 31, 1369–1379. Howlett, F.M., 1914. A trap for thrips. Journal of Economic Biology IX, 21–23.
ARTICLE IN PRESS 736
M.M. Davidson et al. / Journal of Insect Physiology 52 (2006) 729–736
Koschier, E.H., De Kogel, W.J., Visser, J.H., 2000. Assessing the attractiveness of volatile plant compounds to Frankliniella occidentalis. Journal of Chemical Ecology 26, 2643–2655. Lambert, M., 1981. Observations on the laboratory take-off behavior of the desert locust, Schistocerca gregaria (Forska˚l) (Orthoptera: Acrididae). Bulletin of Entomological Research 71, 481–498. Lewis, T., 1997a. Thrips as Crop Pests. CAB International, Wallingford. Lewis, T., 1997b. Flight and dispersal. In: Lewis, T. (Ed.), Thrips as Crop Pests. CAB International, Wallingford, pp. 175–196. Loomans, A.J.M., Murai, T., 1997. Culturing thrips and parasitoids. In: Lewis, T. (Ed.), Thrips as Crop Pests. CAB International, Wallingford, pp. 477–504. MacDonald, K.M., Hamilton, J.G.C., Jacobson, R., Kirk, W.D.J., 2002. Effects of alarm pheromone on landing and take-off by adult western flower thrips. Entomologia Experimentalis et Applicata 103, 279–282. MacQuarrie, C., Boiteau, G., 2003. Effect of diet and feeding history on flight of Colorado potato beetle, Leptinotarsa decemlineata. Entomologia Experimentalis et Applicata 107, 207–213. McCullagh, P., Nelder, J.A., 1989. Generalized Linear Models. Chapman & Hall, London. Morgan, A.C., Crumb, S.E., 1928. Notes on the chemotropic responses of certain insects. Journal of Economic Entomology 21, 913–920. O’Leary, A.T., Kirk, W.D.J., 2004. The effect of wavelength on take-off in the western flower thrips. Acta Phytopathologica et Entomologica Hungarica 39, 123–129. Pearsall, I.A., 2002. Daily flight activity of the western flower thrips (Thysan., Thripidae) in nectarine orchards in British Columbia, Canada. Journal of Applied Entomology 126, 293–302. Rhainds, M., Shipp, L., 2003. Dispersal of adult western flower thrips (Thysanoptera: Thripidae) on chrysanthemum plants: impact of
feeding-induced senescence of inflorescences. Environmental Entomology 32, 1056–1065. Shipp, J.L., 1995. Monitoring of western flower thrips on glasshouse and vegetable crops. In: Parker, B.L., Skinner, M., Lewis, T. (Eds.), Thrips Biology and Management. Plenum Press, New York, pp. 547–555. Shipp, J.L., Zhang, Y., 1999. Using greenhouse microclimate to improve the efficacy of insecticide application for Frankliniella occidentalis (Thysanoptera: Thripidae). Journal of Economic Entomology 92, 201–206. Shipp, J.L., Wang, K., Binns, M.R., 2000. Economic injury levels for western flower thrips (Thysanoptera: Thripidae) on greenhouse cucumber. Journal of Economic Entomology 93, 1732–1740. Smits, P.H., Van Deventer, P., De Kogel, W.J., 2000. Western flower thrips: reactions to odors and colors. Proceedings of Experimental and Applied Entomology 11, 175–180. Teulon, D.A.J., Hollister, B., Cameron, E.A., 1993a. Behavioural responses of western flower thrips to anisaldehyde, and applications for trapping in greenhouses. IOBC/WPRS Bulletin 16, 177–180. Teulon, D.A.J., Penman, D.R., Ramakers, P.M.J., 1993b. Volatile chemicals for thrips (Thysanoptera: Thripidae) host-finding and applications for thrips pest management. Journal of Economic Entomology 86, 1405–1415. Teulon, D.A.J., Hollister, B., Butler, R.C., Cameron, E.A., 1999. Colour and odor responses of flying western flower thrips: wind tunnel and greenhouse experiments. Entomologia Experimentalis et Applicata 93, 9–19. Ullman, D.E., Sherwood, J.L., German, T.L., 1997. Thrips as vectors of plant pathogens. In: Lewis, T. (Ed.), Thrips as Crop Pests. CAB International, Wallingford, pp. 539–565.