Journal of Insect Behavior

, Volume 26, Issue 1, pp 109–119

Phototaxis, Host Cues, and Host-Plant Finding in a Monophagous Weevil, Rhinoncomimus latipes


  • Jeffrey R. Smith
    • Department of Entomology and Wildlife EcologyUniversity of Delaware
    • Department of Entomology and Wildlife EcologyUniversity of Delaware

DOI: 10.1007/s10905-012-9343-7

Cite this article as:
Smith, J.R. & Hough-Goldstein, J. J Insect Behav (2013) 26: 109. doi:10.1007/s10905-012-9343-7


Rhinoncomimus latipes is a monophagous weevil used as a biological control agent for Persicaria perfoliata in the eastern United States. Density of adult R. latipes and resulting feeding damage has been shown to be higher in the sun than in the shade. This study aimed to determine whether phototaxis, sensitivity to enhanced host cues from healthier sun-grown plants, or a combination is driving this behavior by the weevil. A series of greenhouse choice tests between various combinations of plant and light conditions showed that R. latipes is positively phototactic, responsive to host cues, and preferentially attracted to sun-grown plants over shade-grown plants. From our experiments, we hypothesize two phases of dispersal and host finding in R. latipes. The initial stage is controlled primarily by phototaxis, whereas the later stage is controlled jointly by host cues and light conditions.


Host-plant findingphototaxishost plant cuesCurculionidaeRhinoncomimus latipesPersicaria perfoliata


A major goal in the study of plant-herbivore interactions is to determine the mechanisms governing insect host-plant finding and selection. It is generally accepted that host-plant selection is a catenary process consisting of a sequence of behavioral phases or “reaction chains” (Tinbergin 1951; Atkins 1980; Schoonhoven et al. 2005). For example, Finch and Collier (2000) suggested that host-plant selection in pest insects of cruciferous plants involves three stages, the first governed by volatile host plant chemicals, the second by visual stimuli, and the final stage by chemicals on or in the leaf surface. The specific cues used in host finding are highly variable among insect families, and even among species within a family (Schoonhoven et al. 2005). Our goal here was to examine the factors influencing the host finding behavior of a monophagous weevil, Rhinoncomimus latipes Korotyaev (Coleoptera: Curculionidae).

Rhinoncomimus latipes is a small black weevil, approximately 2 mm long, that feeds and reproduces exclusively on Persicaria perfoliata (L.) H. Gross (Colpetzer et al. 2004; Korotyaev 2006; Frye et al. 2010). In 2004, R. latipes was introduced in the Mid-Atlantic States as a biological control agent for P. perfoliata (Hough-Goldstein et al. 2008, 2012; Lake et al. 2011). Persicaria perfoliata, or mile-a-minute weed, is an annual, invasive vine introduced from Asia to the United States in the 1930s (Moul 1948). The plant often grows over and outcompetes native vegetation and is now found in the eastern United States from Massachusetts to North Carolina and westward to Ohio and West Virginia (Oliver 1996; Kumar and DiTommaso 2005; EDDMapS 2012).

Field observations and experiments have shown that density of adult R. latipes and resulting feeding damage to mile-a-minute weed is higher in full sun compared to shaded areas (Hough-Goldstein and LaCoss 2012). Mile-a-minute grows best in full sun, but will tolerate some shade (Mountain 1989; Hough-Goldstein 2008; Hough-Goldstein et al. 2008). Thus, the observed preference of weevils for full-sun areas may be due to an attraction to high light conditions, higher quality sun-grown host plants, or both. Although they demonstrated that weevil density and feeding damage was higher in full sun areas, Hough-Goldstein and LaCoss (2012) did not segregate light conditions and resultant host plant quality to determine their importance in influencing host-plant finding.

Orientation using the sun, through phototaxis or by maintaining straight-line movement at a fixed angle to the sun (the light-compass reaction), is common in insect dispersal flights (Johnson 1963; Atkins 1980; Jermy et al. 1988). Positive phototaxis has been shown for other curculionids (Hollingsworth et al. 1964; Cross et al. 1976; Meyer 1976; Kjaer-Pedersen 1992). However, in these species, positive phototaxis plays a secondary role in host finding. For example, while the boll weevil (Anthonomus grandis grandis [Boheman]) is attracted to light, migration by boll weevils was found to be determined primarily by attraction to plant odors (Mitchell and Taft 1966). Positive phototaxis is also important in host finding by the cabbage seedpod weevil (Ceutorhynchus assimilis [Paykull]), but it plays a secondary role to chemical cues (Kjaer-Pedersen 1992). Similarly, positive phototaxis has been shown in the alfalfa weevil (Hypera postica [Gyllenhal]) in laboratory settings, but time of day was found to be the driving force behind this weevil’s dispersal activity (Prokopy and Gyrisco 1965; Meyer 1976).

Enhanced host plant cues also may contribute to increased R. latipes activity on mile-a-minute grown in full sun conditions. Growth in full sun has been shown to increase biomass of mile-a-minute plants (Hough-Goldstein 2008). The plant vigor hypothesis predicts that larger, more vigorously growing plants will be more attractive to insect herbivores than slower growing shaded plants (Price 1991; Cornelissen et al. 2008). Attraction to host plant cues, both chemical and visual, has been shown for many curculionids. For example, attraction to host plant volatiles has been shown for boll weevils (McKibben et al. 1977), pepper weevils, Anthonomus eugenii Cano (Addesso and McAuslane 2009), and cranberry weevils, Anthonomus musculus Say (Szendrei et al. 2009). Björklund et al. (2005) found that walking pine weevils (Hylobius abietis L.) responded to both odor and visual host stimuli, and traps with both types of stimuli caught more weevils than either stimulus alone. Reeves and Lorch (2011) reported that aquatic milfoil weevils, Euhrychiopsis lecontei Dietz, may use chemical cues if plants are not initially seen, but use vision at closer distances to precisely locate their host plants.

In order to more fully understand the mechanisms for dispersal and host-plant finding in R. latipes, greenhouse trials were conducted to test weevil preference for combinations of light conditions and host plant quality. Our objectives were to determine if the weevil is positively phototactic; if sun-grown plants are more attractive to weevils than shade-grown plants; and if both are true, what is their relative importance in influencing weevil host finding behavior? Two sets of experiments were designed, one to test weevil behavior when they lacked a suitable host plant, and the other to determine whether weevils would abandon one host plant and light condition combination for a different set of conditions.

Materials and Methods

Eight greenhouse trials were conducted during the summer of 2011. Each trial had five replicates, each consisting of two fine mesh, white, tent-like cages measuring 61 × 61 × 61 cm (BugDorm 3, Bioquip Products, Rancho Dominguez, CA). The two cages were connected by a tube constructed from metal window screening and thin metal wire, supporting overlapping cage sleeves that created a passage approximately 18 cm in diameter and 20 cm long. Four beige ceramic tiles were used to create a floor for each cage. For shade treatments, greenhouse blackout cloth was cut into panels and sewn into fitted covers that blocked the front, rear, top, and side panels of the cages.

For each trial, 100 weevils were collected from P. perfoliata plants at White Clay Creek State Park, Newark, DE the morning of the experiment. The weevils were separated into either five petri dishes with 20 weevils each (for trials where weevils were to be released in the central tube) or ten petri dishes with ten weevils each (for trials where weevils were to be released in each of the two linked cages). Weevils were not sexed, but were allocated at random to the replicates and treatments.

The potted P. perfoliata plants used for the trials were grown from seed in a greenhouse. The seeds were collected from White Clay Creek State Park, Newark, DE on 23 October, 2010, and kept at room temperature for 3 weeks. On 12 November, 2010, the mile-a-minute seeds were combined with moist peat moss in a self-sealing plastic bag and put in a refrigerator at approximately 4 °C where they remained until they were planted. Seeds were planted 3 to 5 weeks prior to the start of each trial. Seeds were planted in Redi-Earth potting soil (Grace-Sierra, Milpitas, CA) in 7.5 cm square pots, and placed in a mist room. When plants were approximately one and a half weeks old, the seedlings were transplanted to 15 cm diameter round pots with Pro-Mix BX soil (Premier Tech Horticulture, Quakertown, PA) and moved to a greenhouse room. At this time plants designated as shade-grown plants were placed on a metal greenhouse bench under a frame draped with blackout cloth with an open north face and wire mesh bottom to allow air ventilation and indirect sunlight. Shade-grown plants were grown under these conditions for between 2 and 4 weeks. All other plants were grown in full sun. Each individual trial used plants of the same age. Plants were watered daily at the base of the plants, immediately after weevils were counted.

The trials conducted and their conditions are summarized in Table 1. Trials were conducted in two separate rooms of the greenhouse (assigned randomly) to allow two trials to be run simultaneously. Each pair of cages was oriented in an east–west direction and the assignment of Side A and Side B conditions was randomized prior to each trial; however each set of conditions had either two west and three east or three west and two east cages to minimize any possible differences between the east and west cage. Trials began at either 12 noon or 2 PM, when the petri dishes were placed and left open either in the central tube (to test weevil choice in the absence of a suitable host) or at the base of the plants or the bottom of the cage in each of the two cages (to test whether weevils would abandon one set of conditions for another).
Table 1

Rhinoncomimus latipes greenhouse choice trails conducted with two linked cages (side A and B), summer 2011

Dates of trial

Side A conditions

Side B conditions

R. latipes release site

Corresponding figure

7/11 to 7/15


Shade, Plant

Connecting Tube


7/18 to 7/22


No plant

Connecting Tube


7/11 to 7/15


Shade-grown plant

Connecting Tube


8/8 to 8/12

No plant

Shade, Plant

Connecting Tube


8/1 to 8/5


Shade, Plant

In cages


7/25 to 7/29


Shade, Shade-grown plant

In cages


7/18 to 7/22

Shade-grown plant

Shade, Plant

In cages


8/15 to 8/19

No plant

Shade, Plant

In cages


The number of weevils in Side A and Side B was recorded 15 min, 30 min, 60 min, 90 min, and 120 min following release, and then daily for the next 4 days at approximately the same time as they were released. Weevils were not counted when they were still in the petri dish or in the central tube. Weevils on the cage structure and the plant were counted separately, but combined for data analysis. For shaded cages, the fitted cover was removed when weevils were counted, and then replaced. During the first trial, air temperatures in the paired sun and shade cages were measured daily using a mercury thermometer placed in the cages during the time that weevil counts were conducted.

After the completion of trials comparing response to sun-grown plants and shade-grown plants, plants were cut at soil level and immediately weighed to obtain fresh biomass. Plants were then placed in paper bags and put in a drying oven at 60 °C for 72 h after which they were weighed again to obtain dry biomass.

The total numbers of weevils in each of the two connected cages at each time interval were compared using two-tailed paired t-tests. Fresh biomass and dry biomass of sun-grown and shade-grown plants, and temperatures in sun and shade cages, also were compared using two-tailed t-tests.


In the first set of experiments (Fig. 1) 20 weevils were released in the central tube. When given a choice between a cage in the sun and a shaded cage, each with a plant of equal quality, significantly more weevils were found in the cage in the sun at all observed time intervals (P ≤ 0.004, Fig. 1a). When allowed to choose between two cages both in full sun, one with a plant and one without, weevils preferentially selected the cage with the plant (P ≤ 0.043, Fig. 1b); differences were significant beginning on the second day following release and for the remainder of the week. Similarly, when presented with two cages both in full sun, one with a sun-grown plant and the other with a shade-grown plant, significantly more weevils selected the cage with the sun-grown plant from 60 min throughout the end of the trial (P ≤ 0.020, Fig. 1c). Finally, when given a choice between a cage in the sun with no plant and a shaded cage with a plant, weevils selected the sun cage with no plant at all observed time intervals (P ≤ 0.043, Fig. 1d).
Fig. 1

Mean (± SEM) number of weevils in each of two linked cages when 20 weevils were released in the connecting tube between (a) a cage with a plant and a shaded cage with a plant; (b) two cages both in full sun, one with a plant and one without; (c) two cages both in full sun, one with a sun-grown plant and one with a shade-grown plant; and (d) a cage with no plant and a shaded cage with a plant. N = 5; *, significantly different at P ≤ 0.05, two-tailed paired t-test

In the second set of experiments (Fig. 2) ten weevils were released in each cage, at the base of the plant or on the bottom of the cage if no plant was present. When weevils were released at the base of two host plants of equal quality, one in sun and one in shade, significantly more weevils were observed in the cage in the sun at time intervals from 60 min through day 5 (P ≤ 0.020, Fig. 2a). From day 2 until the end of the trial, an average of more than ten weevils was found in the sun cages, indicating that weevils had moved from the shaded cage to the sun cage (Fig. 2a). Likewise, when released at the base of a sun-grown plant in the sun and a shade-grown plant in a shaded cage, weevils moved towards the cage in the sun with the sun-grown plant, and were found in significantly higher numbers in that cage from 90 min on, with the exception of day 4 (P ≤ 0.049, Fig. 2b). On average, there were more than ten weevils in the sun cage at 120 min, day 2, and day 3, again indicating that weevils had moved from the shaded cage through the central tube to the cage in the sun (Fig. 2b).
Fig. 2

Mean (± SEM) number of weevils in each of two linked cages when 10 weevils were released in each of two cages, (a) a cage with a plant and a shaded cage with a plant; (b) a cage in sun with a sun-grown plant and a shaded cage with a shade-grown plant; (c) a cage with a shade-grown plant and a cage with a sun-grown plant in the shade; and (d) a cage with no plant and a shaded cage with a plant. N = 5; *, significantly different at P ≤ 0.05, two-tailed paired t-test

When weevils were released at the base of a shade-grown plant in a sun cage paired with a sun-grown plant in a shaded cage, weevil numbers were relatively equal in the two cages from day 2 on (Fig. 2c). During the initial observation period, elevated counts in the sun cage primarily reflected weevils that became active and left the petri dishes more quickly than weevils in the shaded cage. Throughout the week, no more than ten weevils were found in either of the cages (Fig. 2c). Similarly, when weevils were released at the bottom of a cage in the sun with no plant and at the base of a sun-grown plant in the shade on the other side, weevil numbers were initially higher in the sun as weevils left the petri dish, but were very similar on days 2–5, and never exceeded ten weevils in either cage (Fig. 2d).

Plants grown in full sun had significantly higher biomass, both fresh and dry, than shade-grown plants following 2, 3, and 4 weeks of growth under the different conditions (Table 2). Temperatures in the sun and shade cages, measured for the first trial, showed no differences at any point in the week (Table 3).
Table 2

Average fresh (A) and dry (B) plant biomass (mean ± SEM) for trials using sun and shade grown plants


Time grown under different conditions


2 weeks

3 weeks

4 weeks

A. Fresh biomass (g)



27.47 ± 2.75

37.09 ± 2.02

55.00 ± 3.13


1.11 ± 0.23

0.42 ± 0.11

0.40 ± 0.12









B. Dry biomass (g)



4.10 ± 0.43

4.10 ± 0.26

13.79 ± 0.75


0.12 ± 0.04

0.07 ± 0.02

0.04 ± 0.01









Table 3

Daily average temperature (mean ± SEM) in sun and shade cages during the first trial

Temperature (°C)

1 d

2 d

3 d

4 d

5 d


33.4 ± 0.19

33.0 ± 0.16

31.4 ± 0.19

30.7 ± 0.12

30.2 ± 0.12


33.8 ± 0.12

32.9 ± 0.19

31.7 ± 0.20

30.5 ± 0.22

30.5 ± 0.19














Based on these experiments, R. latipes is highly phototactic and responsive to host plant cues. When the weevils had no host plant (in this study, placed in the central tube between two linked cages) they were strongly attracted to the sun cage versus the shaded cage, even when the sun cage lacked a plant and the shade cage had a high quality sun-grown plant. This shows a strong positive phototaxis, which overrides response to host cues when the weevil initially lacks a host. However, when both sides were in sun, weevils placed in the central tube were attracted to the side with a plant, and were differentially attracted to a high quality sun plant rather than a low quality shade-grown plant. Therefore, when light conditions are equal, weevils that lack a host plant are attracted to host cues, and are more attracted to high quality than low quality plants. These experiments simulated a scenario in nature where the weevil lacks a suitable host plant, which could occur through excessive herbivory or environmental stresses on the plant. The experiments showed that in these situations, dispersal and host finding is primarily controlled by positive phototaxis and secondarily by response to host plant cues.

The greater response by R. latipes to light than to plant cues when the weevils were not on the host plant is similar to the results of Kjaer-Pedersen (1992), who found that cabbage seedpod weevils responded to host odors by moving upwind (odor-conditioned anemotaxis) but when host plants were not present and wind speeds were low the weevils moved toward the sun. He interpreted this as an orientation response ensuring that weevils seeking host plants would not overfly an already visited area. Jermy et al. (1988) similarly suggested that movement toward the sun, or in a straight line using light-compass orientation, can be an adaptive trait enabling insects to quickly leave a habitat without host plants and reach another habitat. In our system, R. latipes that are faced with either no mile-a-minute or heavily damaged mile-a-minute probably disperse using the sun for orientation in search of new host plants, rather than relying on host plant cues to find nearby, and most likely also poor quality plants.

Weevils in the presence of a host plant (in this study, placed at the base of plants on either side of a tube connecting two linked cages) actively moved off the plant they were placed on in the shade and sought out the plant in the sun if the plants on both sides were of equal quality or if the plant in the shade was inferior. This suggests that positive phototaxis also is important in the later stages of host-plant finding and selection. However, if the plant in the shade was of better quality than the plant in the sun, weevils did not redistribute themselves into the sun cage, with weevil numbers remaining below ten in both cages. These experiments indicate that the later stages of host-plant finding and host-plant selection are controlled jointly by light conditions and host plant cues. The balance between these two factors is likely due to a combination of the energetic and reproductive risks and rewards for leaving one plant in search of another. Factors contributing to this balance include: fitness of the current plant, plant cues from nearby plants, light conditions, and the likelihood of finding a new host plant.

The distance from a plant at which host cues, whether chemical or visual, begin to override alternative behavioral controls is not well studied, but recent work has begun to explore this question. Insect olfactory neurons are highly responsive to blends of specific host plant volatiles (Bruce and Pickett 2011), but the distance at which insects respond to these volatiles is difficult to measure (Jermy et al. 1988; Schoonhoven et al. 2005; Finch and Collier 2012). Reeves and Lorch (2011) estimated the visual active space of the milfoil weevil at 17.5 cm. In our experiments where weevils were placed in the central tube between two cages, the weevils were approximately 35 cm from a plant placed in a cage, and probably responded primarily to host plant volatiles; however, no effort was made to mask the plants from view, so there may have been a visual response as well.

The concept that there are distinct phases in the host-plant finding and selection process, each controlled by different factors, is generally accepted by researchers in the field (Schoonhoven et al. 2005). From our experiments, we conclude that there are two distinct phases of host-plant finding in R. latipes. The first stage appears to be predominately controlled by attraction to light, but is secondarily influenced by plant cues. The later stages are influenced jointly by plant cues and response to light conditions. Our final experiment, depicted in Fig. 2d, shows both stages of host-plant finding simultaneously. The weevils released at the bottom of a sun cage that lacked a plant remained in the initial stage of host finding and therefore remained in the sun cage despite the host cues given off by the plant in the shaded cage. Conversely, weevils on the plant in the shade exhibited the later stages of host finding and remained on the high quality plant, controlled more by strong host cues than the desire to move toward the sun. We suspect that the weevils in the experiment shown in Fig. 2d remained in the cages where they were placed, rather than moving back and forth between cages. If weevils were moving between the two cages, the trends from day 2 to day 5 of Figs. 1d and 2d would look similar. Instead, weevils initially placed in the connecting tube moved into the sun cage and stayed there, with very few weevils moving into the shaded cage to find the plant, whereas weevils placed in the shaded cage with the plant remained at high numbers in that cage throughout the week.

Our results are consistent with the findings of Hough-Goldstein and LaCoss (2012), who showed that weevils in the field preferentially fed and oviposited on mile-a-minute in the sun, compared to artificially shaded plants. Their field experiment was similar to the trials depicted here in Fig. 2a and b, where weevils actively moved off plants in the shade and sought out plants in the sun. They attributed this both to selection for oviposition on the larger, more vigorous plants in the sun, consistent with the plant vigor hypothesis of Price (1991), as well as an attraction to sunny areas, where generally warmer temperatures would allow for more rapid growth and development of the weevil. Because there was no difference in the temperature within the sun and shade cages in the present study, temperature preference is unlikely to be the primary factor driving the behavior observed here, but may be part of the evolutionary reason for it.

Further studies are needed to fully understand host finding behavior in this weevil. In our study we did not separate visual and chemical cues, and do not know their relative importance in influencing weevil behavior in each stage of host finding. Further studies can refine the catenary process laid out in this paper, by modifying or further dividing our proposed phases, as well as by more clearly defining the boundaries between the phases. A better understanding of host finding in this weevil will provide strong empirical evidence that can be used in the construct of theories that attempt to generalize host finding to higher taxa of insects. In addition, understanding plant finding behavior in this insect may prove important in predicting efficacy of R. latipes as a biological control agent in different host habitats.


We thank White Clay Creek State Park for allowing us to collect weevils and seeds from their land. Additionally, we thank Kaity Handley and Connie Gatlin for helping to set up and monitor experiments as well as Bill Bartz and Rodney Dempsey for their expertise and help with designing greenhouse protocols. Funding was provided by the USDA Forest Service.

Copyright information

© Springer Science+Business Media, LLC 2012