Lygus rugulipennis on chrysanthemum: Supplemental prey effects and an evaluation of trap plants

The European tarnished plant bug, Lygus rugulipennis Poppius, is considered a major pest in chrysanthemum nurseries in The Netherlands. Adults puncture plant's apical meristem, after which the growing point splits and growth is inhibited. Flower buds and flowers can also be severely damaged. Both types of damage result in economic losses for growers. Despite the importance of this pest for chrysanthemum nurseries, there is only very limited information about L. rugulipennis development on chrysanthemum plants, Chrysanthemum × morifolium Ramat., and whether L. rugulipennis can be controlled using trap plants is not known. We investigated whether: (1) L. rugulipennis could develop from egg to adult on the vegetative and flowering stages of chrysanthemum; (2) their performance was enhanced when a supplemental prey source (Ephestia kuehniella Zeller eggs) or another common pest (the green peach aphid, Myzus persicae Sulzer) was present; and (3) there were alternative plant species more attractive than chrysanthemum for use as trap plants for local pest control or monitoring of L. rugulipennis. L. rugulipennis developed on both vegetative and flowering chrysanthemum stages without any additional food source. Nonetheless, when chrysanthemum was supplemented with E. kuehniella eggs, L. rugulipennis achieved the best performance in terms of the number of adults developed and faster developmental time. Interestingly, L. rugulipennis developed faster on chrysanthemum infested with the aphid M. persicae compared to non‐infested plants, however, there was no difference in the number of adults developed. In a trap plant experiment with 16 plant species in the vegetative stage, we found that white mustard, Sinapis alba L., was significantly more attractive than chrysanthemum to both adult and nymph L. rugulipennis. Further research is needed to evaluate the potential of S. alba as a trap plant for monitoring L. rugulipennis and how the presence of prey in the crop influences L. rugulipennis.

In Dutch chrysanthemum nurseries, L. rugulipennis is considered a pest insect that causes major crop losses (Hennekam et al., 2012).
Damage can easily be observed as plants infested with L. rugulipennis are up to 15 cm shorter than non-infested plants of the same age. This damage is caused by one or a few L. rugulipennis adults puncturing the plant's apical meristem, after which the growing point splits and its growth is inhibited. L. rugulipennis also damages flowers, resulting in spots on petals, irregular growth, and indentations, making flowers non-marketable products. Despite the importance of this pest for chrysanthemum nurseries, nothing is known about the ability of L. rugulipennis to develop its full life cycle on chrysanthemum plants, Chrysanthemum × morifolium Ramat. (Asteraceae) (Hennekam et al., 2012;Holopainen & Varis, 1991).
Although L. rugulipennis is considered as a pest, it is also an omnivore feeding on both plants and prey, like many other mirids (Coll & Guershon, 2002). Mirids range from phytozoophagy (prey-feeding herbivores) to zoophytophagy (plant-feeding carnivores) (Coll & Guershon, 2002). For example, the predator Nesidiocoris tenuis Reuter (Heteroptera: Miridae) is released as a predator in tomato crops, but is also considered as a pest when densities are getting high . Within the genus Lygus, the phytozoophagous western tarnished plant bug, Lygus hesperus Knight (Heteroptera: Miridae), is an important pest in cotton, but also a predator of whiteflies (Hagler & Naranjo, 2005). Prey feeding has also been reported for L. rugulipennis on other pests such as aphids (Holopainen & Varis, 1991;Salerno et al., 2007). However, to what extent the presence of supplemental prey affects the performance of L. rugulipennis on chrysanthemum plants is not known. Besides L. rugulipennis, other pests can be present in Dutch chrysanthemum nurseries, for example, the cotton aphid, Aphis gossypii Glover, and the green peach aphid, Myzus persicae Sulzer (both Hemiptera: Aphididae), are regularly found on this crop and can increase population size rapidly (Vehrs et al., 1992). They damage by sucking on plant parts and producing honeydew which encourages black sooty mould (Vehrs et al., 1992). Concurrent damage by Lygus pests and aphids can thus be common in chrysanthemum nurseries, yet is not known how the performance of L. rugulipennis is affected when aphids are present on the same plant.
As part of biological pest control, supplemental food, like Ephestia kuehniella Zeller (Lepidoptera: Pyralidae) eggs is often supplied to released natural enemies as part of a 'standing army' approach (Messelink et al., 2014;Pijnakker et al., 2020). This enables populations of biological control agents to be sustained and ready to eliminate pest insects early after crop colonization. However, if the pest is an omnivore, some caution is needed, because the provided food sources may promote pest growth (Leman & Messelink, 2015;van Rijn et al., 2002). Hence, pests might take advantage of biological control strategies aimed to favour carnivore insects. It is not known if L. rugulipennis benefits from supplemental food in chrysanthemum crops.
A promising alternative to the use of chemical control for L. rugulipennis is the implementation of 'trap crops'. A trap crop represents a plant that is more attractive for a pest than the main crop itself (Shelton & Badenes-Perez, 2006). Such attractive plants can be placed either between or around the main crop and act as a sink for the pest, thus facilitating supplemental management practices aimed at pest removal or elimination (Shelton & Badenes-Perez, 2006). This method is successfully used in various crops (reviewed by Hokkanen, 1991;Shelton & Badenes-Perez, 2006).
For Lygus spp., it has been shown that M. sativa can be successfully used as a trap crop (Sevacherian & Stern, 1974;Swezey et al., 2013).
In organic strawberry fields in California, L. hesperus preferred the interplanted trap crop M. sativa over organic strawberries (Swezey et al., 2007). Bugs present in the trap crop were then regularly removed by vacuuming-bugvac (Swezey et al., 2007). It was subsequently shown that L. hesperus did not disperse from M. sativa into strawberry plants (Swezey et al., 2013). Whether trap crops can also be implemented in chrysanthemum nurseries to control L. rugulipennis, as M. sativia works for L. hesperus, has not been investigated yet.
In this study, we conducted a series of experiments to investigate the performance and development of L. rugulipennis on chrysanthemum plants, C. × morifolium, and whether trap plants can be used for monitoring L. rugulipennis. Specifically, we investigated whether:  Per cage, five males and five females L. rugulipennis of 1-week-old, was released for 48 h and allowed to lay eggs. After 48 h all adults were removed and the treatments with supplemental prey were added. Three times a week, cages were checked for the presence of nymphs which were monitored until they reached the adult stage, after which they were removed. At the start of the experiment, all chrysanthemum plants had the same size (having serval branches and about 20 cm tall). When flower buds emerged, they were immediately removed. For each treatment, three replicates were performed. Cages were placed in a block design inside a greenhouse.
We scored the following fitness-related proxies: (1) the number of offspring developed until the adult stage; (2) the sex ratio of the developed adults, and (3) the developmental time from egg until the adult stage.

| Trap plant experiment
To screen for plant species more attractive for L. rugulipennis adults than chrysanthemum plants, a multiple-choice experiment with 16 different plant species (Table 1) was set up under greenhouse conditions. Plants were selected based on literature or on preliminary investigations. All 16 plant species were grown from seeds or young plant material ( Table 1). To ensure plants were free of pesticides, biological seeds were used. Lavender, Lavandula angustifolia Mil. 'Hidcote' (Lamiaceae) and common nettle, Urtica dioica L.
(Urticaceae) plants were bought/collected 2 months before the experiment started and placed per plant species in a quarantine cage and checked weekly on the presence of pest insects, which were removed when present. L. angustifolia was pruned after buying, so TA B L E 1 Plant species used for the trap plant experiment. Host plants for adults, nymphs, and oviposition as recorded by Holopainen and Varis (1991)     supplied by an automatic dripping system. The experiment was conducted at 20 ± 1C°, 70% RH.

| Development in the presence of supplemental food or aphids
The performance of L. rugulipennis on vegetative chrysanthemum plants was affected by the presence of additional food (Figure 2a).   (Figure 4). S. alba was the only plant species on which significantly more L. rugulipennis nymphs were found than on chrysanthemum (χ 2 = 11.977, df = 1, p < 0.001) (Figure 4). On S. tuberosum (χ 2 = 0.497, df = 1, p = 0.481) and U. dioica (χ 2 = 4.445, df = 1, p = 0.350) equally nymphs were found compared to chrysanthemum. On chrysanthemum significantly more nymphs were found than on 12 of the tested plant species (Figure 4).

| DISCUSS ION
In this study, we showed that L.  (Coll & Guershon, 2002). The pest L. rugulipennis is not a strict herbivore, but rather an omnivore, which seems to be beneficial for this pest's performance and may explain its successful survival in many cropping systems. Lygus rugulipennis prefers to feed on apical meristems and flower buds of plants (Varis, 1972), but it can also predate on other insects (Holopainen & Varis, 1991). We have seen L. rugulipennis nymphs and adults feeding on M. persicae, as well Varis (1972)  We also showed that L. rugulipennis achieved the best performance (i.e., developed 6.8 days faster compared to the control) when E. kuehniella eggs are supplemented as a food source. As E. kuehniella eggs represent high-quality additional food sources for supporting predators such as Orius laevigatus Fieber and Orius majusculus Reuter (both Hemiptera: Anthocoridae) in chrysanthemum nurseries (Messelink et al., 2019;Pijnakker et al., 2020), a side effect might be that L. rugulipennis can benefit from the addition of supplementary food sources. We suggest that caution is needed in chrysanthemum nurseries to maximize the performance of biological control agents while minimizing the benefits for pests.

F I G U R E 3
Percentage of Lygus rugulipennis adults recorded on 16 plant species during the trap plant experiment based on the sum of all observations. The asterisks above bars indicate significant differences between recorded L. rugulipennis adults on Chrysanthemum × morifolium 'Baltica' compared with the other plant species (p < 0.05 according to Chi-square test with Bonferroni correction). NS, no significant differences.
Our study indicated that S. alba in the vegetative stage is a potential candidate trap plant as part of an integrated approach to monitor and control L. rugulipennis in chrysanthemum nurseries.
S. alba can also be used as a trap crop for other pests (reviewed by The attractiveness of these varieties has to be tested between each other and between S. alba as it might be that some varieties are more attractive to L. rugulipennis. As we also used two varieties in our study, there was a difference of almost 3 days in development time between both varieties (see Figures 1b and 2b), however, we cannot compare

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data supporting the findings of this study are available from