An Environmentally Friendly Method to Protect Box Trees (Buxus spp.) from Attacks by the Invasive Moth Cydalima perspectalis

Abstract

The non-native invasive box-tree moth Cydalima perspectalis causes severe damage to ornamental box trees (Buxus spp.) and natural boxwood stands. So far, no promising natural enemy of C. perspectalis has been discovered in Europe. Many garden owners would like to protect their box trees from C. perspectalis without the use of insecticides, which also harm other arthropod species. In a controlled experiment under natural conditions, we tested whether box trees covered with a net during the flight period of C. perspectalis are as well protected against the moth as trees treated with a bioinsecticide. After 1 year, the box trees covered with a net during the moths’ flight activity (monitored by light traps) showed no damage by larvae (average loss of leaves 0%), as did box trees regularly treated with a bioinsecticide (control group 1). In contrast, box trees with no protection (untreated box trees; control group 2) lost 97.7% of their leaves due to larval feeding. In a second experiment, we investigated whether defoliated box trees can recover when covered with a net during the flight period of the moth. Protected by the net, the emerging new leaves were not attacked by C. perspectalis. After 1.5 years, the trees had 24% of their original foliage again. Our study demonstrates that temporarily covering box trees with a net protects them against damage by C. perspectalis in an effective way.

1. Introduction

The box-tree moth Cydalima perspectalis (Walker 1859), native to eastern Asia, was first recorded in Germany and the Netherlands in 2007 and was most likely introduced through the nursey trade [1,2]. Since then, the non-native moth has spread across Europe, causing devastating damage to ornamental box trees (Buxus spp.) in private gardens, cemeteries, and urban and historical parks [3,4] and also affecting natural boxwood stands (Buxus sempervirens L.) in southern and western Europe [5,6,7,8]. More recently, C. perspectalis has been recorded in North America (first reported in Toronto, Canada, in 2018 [9]) and in Africa (first detected in Constantine, Algeria, in 2018 [10]). Nowadays, the box-tree moth is considered an invasive species in several European countries and is on the way to becoming a cosmopolitan invasive species [4]. Ornamental box trees are highly valued by many garden owners and can grow old (>100 years [4]). As evergreen trees with dense foliage, box trees are suitable for topiary, which, in turn, increases their value [4].

The box-tree moth is a multivoltine species that can have up to three generations per year in northwestern Switzerland [3]. Female box-tree moths lay their eggs on Buxus spp. leaves. The larvae feed on the leaves and, at high population densities, also on the bark of box trees, affecting different varieties of horticultural box trees in a similar way [11]. Winter diapause occurs in the larval stage, and moths of the first generation emerge in late May and June [3]. In the summer, the larvae can defoliate a box tree within a few weeks. So far, knowledge about the natural enemies of C. perspectalis in its area of origin is extremely poor, and no promising natural enemy has been discovered in Europe [4]. Predation by birds is low, probably due to the high levels of toxic alkaloids sequestered by the larvae [12], and is restricted to the summer generation. Wasps have occasionally been observed feeding on C. perspectalis larvae. The only parasitoids detected in Europe that parasitize C. perspectalis are polyphagous species [13]. Therefore, garden owners and urban park managers often rely on chemical control with contact or systemic insecticides, which are effective but may harm the moth’s natural enemies and other arthropod species using the box trees for shelter, such as arachnids and other insects [4,14]. Biopesticides based on the bacterium Bacillus thuringiensis would be the preferred option for ornamental box trees due to their relatively minor environmental impact [4]. However, there is an increasing number of garden owners who no longer want to use pesticides to maintain their gardens.

We aimed to examine how horticultural box trees can be protected from attacks by C. perspectalis without the use of pesticides. In particular, we tested whether fine-mesh nets, such as those used to protect fruit trees, berry bushes, and smaller columnar trees from pests, such as cherry fruit flies and apple and plum moths, also protect against attacks by box-tree moths in a controlled field experiment.

Garden owners frequently remove defoliated and severely damaged box trees. However, most of these trees are not dead. The box trees will sprout again next spring and, if there are no further feeding attacks by C. perspectalis larvae, will regenerate completely within 4–6 years [15]. In a second experiment, we investigated the extent to which defoliated box trees recover if they are covered with a fine-mesh net during the flight period of C. perspectalis.

2. Materials and Methods

2.1. Experimental Box Trees

We obtained 18 box trees (Buxus sempervirens) from the Merian Gardens in Münchenstein, near Basel, in 2020. These box trees had been planted as a hedge around 1968. In previous years, the box trees had been treated with Delfin^® (Altermatt Biocontrol Suisse, Grossdietwil, Switzerland). Due to a re-design of the Merian Gardens, the box trees had to be removed on 16 October 2020. On the same day, we planted each box tree in a plastic pot (58 cm in diameter, 48 cm high) filled with standard garden soil. The planted box trees were transported to the Institute for Conservation Biology at the University of Basel and held there in the courtyard until April 2021. During this time, the trees were regularly irrigated and covered with nets to protect them from attacks by the box-tree moth. Thus, the 18 box trees were of the same age and origin and were exposed to the same environmental conditions for a long time before the experiment.

2.2. Protection Experiment

To examine whether a fine-mesh net that is stretched over a box tree protects the tree from damage by C. perspectalis, we randomly assigned six trees each to three groups. In one group, the box trees were covered with a fine network (Lutrasil Pro 19X^®, an agricultural fleece hereafter referred to as a net) that was sewn in the shape of a suitable cover (Figure 1). Lutrasil Pro 19X^® is permeable to light, air, and water; weighs 19 g/m²; and protects against insects (Soft Nonwovens, Quaregna Cerreto, Italy). The cover was closed with a rubber band or laundry clovers on the pot (Figure 1). In another group (control group 1), the box trees were treated with Delfin^® (Altermatt Biocontrol Suisse, Grossdietwil, Switzerland), a biological insecticide that was regularly sprayed on the box tree leaves. Delfin^® contains Bacillus thuringiensis ssp. kurstaki, a bacterium that produces a toxin (endotoxin) [16]. This toxin causes a specific fatal intestinal disease in the larvae after they consume the leaf material [16]. Delfin^® was prepared following the manufacturer’s (Altermatt Biocontrol Suisse, Grossdietwil, Switzerland) instructions. We sprayed the entire foliage of the box trees with Delfin^® every 3–4 weeks (preferably after heavy rainfall). The box trees assigned to the third group (control group 2) served as controls and were not treated at all.

The height of the experimental box trees averaged 79.1 ± 1.2 cm (mean ± SE, range 70–88 cm, n = 18) and did not differ between the three experimental groups (ANOVA, F_2,15 = 0.183, p = 0.83).

Since the density of box-tree moths in the settlement area varies spatially according to the frequency of untreated box trees in private gardens, we carried out the experiment at three different localities. On 28 and 29 April 2021, we transported the box trees to three gardens in the suburbs of Basel, Switzerland. Six box trees (two from each experimental group) were placed in the backyard of the Hörnli cemetery (hereafter Hörnli; 47.5614 N, 7.6414 E; elevation 274 m a.s.l.), six box trees in a private garden in Muttenz (hereafter Muttenz; 47.5222 N, 7.6492 E; elevation 309 m), and six box trees in a private garden in Reinach (hereafter Reinach; 47.5060 N, 7.5983 E; elevation 295 m; Figure 1). The experimental box trees were placed in lawns (Figure 1). The three localities were in residential areas with gardens and were 4.4 to 7.0 km apart. In neighboring gardens, there were numerous ornamental box trees at distances of 20–40 m. Therefore, our experimental box trees were often visited by C. perspectalis moths to lay their eggs.

The experiment was conducted from 29 April 2021 to 28 April 2022. At the start of the experiment, we recorded the percentage of undamaged leaves on all box trees (in each case 100%). Box trees of one group were covered with a fine-mesh net during the flight activity period of C. perspectalis from 29 April to 3 November 2021. To assess the extent of the damage caused by feeding C. perspectalis larvae over time, we estimated the percentage of undamaged leaves to the nearest 5% on all box trees on 4 August, 9 September, and 3 November 2021 and 28 April 2022 (trees with intact foliage = 100%; Table S1). This standardized method has been used for several years to record damage to box trees in natural box-tree forests. The estimate was always carried out by the same person (B.B.). The extent of grazing damage was expressed as 100% minus the percentage of undamaged leaves. During warm periods, the box trees were watered.

2.3. Seasonal Flight Activity of C. perspectalis

To monitor the timing of emergence and the abundance of C. perspectalis moths, we used two light traps placed in Hörnli and Reinach. Light traps give much more reliable results than pheromone traps [17] because due to multiple introductions of C. perspectalis from different locations [18], the moths use different pheromones [17]. The light traps operated daily from May to mid-November in 2021–2022 in Hörnli and 2021–2023 in Reinach. The trapping was stopped in late fall when we no longer found any C. perspectalis moths in the traps for at least 2 weeks. Using a timer, the light was switched on at the beginning of dusk and switched off at the end of dawn. The modified Heath light traps consisted of a box (30 cm × 30 cm × 30 cm) with an attached funnel. At an angle of 120°, three transparent slats were attached at the light source (8 W, with increased UV in the range of 250–300 nm). The slats routed approaching moths into the funnel and the box. In the box, Mikorex^® (HGZ, Heusenstamm, Germany) was used as a killing agent. We emptied the traps on a weekly basis and counted the number of C. perspectalis individuals.

2.4. Recovery of Defoliated Box Trees

Defoliated box trees usually sprout again and recover in the following years, provided the emerging leaves are not eaten by C. perspectalis larvae. In a second experiment, we examined whether defoliated and severely damaged box trees recover when protected from further attacks by C. perspectalis larvae by a net over 1.5 years. In this experiment, five defoliated and heavily damaged box trees of the former control group 2 in the first experiment were temporarily covered with a net (as described earlier) from 28 April to 16 November 2022 and again from 17 May to 15 November 2023. We assessed the recovery of these box trees by estimating the number of new leaves, expressed as a percentage of the leaves of a healthy tree, on 22 June, 7 September, and 16 November 2022 and on 17 May, 5 July, and 15 November 2023 (Table S2). The seasonal flight activity of C. perspectalis was recorded in 2022 and 2023, as described earlier.

2.5. Data Analyses

We used the estimate of the percentage of undamaged leaves (accuracy 5%) to calculate the extent of grazing damage by C. perspectalis larvae (100% minus the percentage of undamaged leaves) in each box tree. All statistical analyses were performed in R [19]. Repeated analysis of variance was used to examine the effects of study site (Hörnli, Muttenz, Reinach), treatment nested within study site (experimental group: temporarily covered with a net; control group 1: Delfin^® application; control group 2: untreated trees) and time (damage to leaves recorded on 4 August, 9 September and 3 November 2021 and on 28 April 2022) and the interactions between study site and time and between treatment and time on the extent of grazing damage (% damaged leaves) in box trees. Before the repeated analysis of variance was conducted, the percentages of damaged leaves were subjected to arc-sinus square-root transformation. We checked whether the residuals were normally distributed and had equal variances at different times (sphericity).

3. Results

3.1. Protection Experiment

Box-tree moths were captured with light traps from 9 June to 6 October 2021, with a break from 14 July to 18 August (Figure 2). The first period corresponded to the summer generation, while the flight activities of the individuals of the second and third generations partially overlapped in the second period (Figure 2). More box-tree moths were caught in Reinach (59 in total) than in Hörnli (34 in total), but the flight activities occurred at both trap sites at the same time (Figure 2).

At the end of the experiment, on 28 April 2022, the box trees covered with a net during the moths’ flight activity showed no damage by C. perspectalis larvae (damaged leaves: 0%; Figure 3A). Similarly, the box trees treated with Delfin^® had no feeding damage, with the exception of one tree, which showed minor damage (to 2% of the leaves). This group had an average loss of 0.3% of leaves (Figure 3B). In contrast, box trees with no protection (control group 2) lost 97.7% of the leaves due to feeding by C. perspectalis by the end of the experiment (Figure 3C). The three groups of box trees were treated differently and, therefore, differed in the extent of damage: the temporarily used net protected the trees well from the moth (

Table 1. Summary of repeated analysis of variance examining the effects of study site, treatment (=experimental group), and time on the extent of grazing damage (% damaged leaves) in box trees. Sites: Hörnli, Muttenz, and Reinach. Treatment: group 1, temporarily covered with a net; control group 1, Delfin^®; control group 2, untreated. Time: damage to leaves recorded on 4 August, 9 September, and 3 November 2021 and on 28 April 2022. Significant p-values (p < 0.05) are in bold.

Factor	d.f.	F	p
Site	2,9	3.16	0.094
Treatment [site]	6,9	21.61	<0.0001
Time	3,27	31.03	<0.001
Site × time	6,27	5.27	0.001
Treatment [site] × time	18,27	10.75	<0.0001

). Interestingly, the larvae’s main feeding activities did not take place in the three gardens at the same time. In Hörnli, most of the leaves were eaten early in spring 2022 (by C. perspectalis that overwintered on the trees), while in the other two gardens, the greatest damage occurred between 4 August and 9 September 2021 (second generation of C. perspectalis). This resulted in significant site × time and treatment × time interactions (Table 1).

3.2. Recovery of Defoliated Box Trees

Box-tree moths were captured with light traps from 25 May to 2 November 2022, with a break from 29 June to 27 July (Figure 2), and from 7 June to 1 November 2023, with no apparent break in July (Figure 2). During these periods, the box trees were covered with nets.

At the beginning of the study, on 28 April 2022, the severely damaged box trees only had 2.8% leaves left (range 0–5%, n = 5; Figure 4A). In the following weeks, the proportion of newly sprouted leaves steadily increased (Figure 5). Protected by the net, the new leaves were not attacked by C. perspectalis. After 1.5 years, the trees had 24% of their original foliage again (range 15–35%; Figure 4B and Figure 5).

4. Discussion

The non-native invasive C. perspectalis causes enormous damage to box trees in parks and gardens. Natural enemies cannot control the invasive moth in Europe [4]. So far, chemical control with contact or systemic insecticides and the use of biological insecticides are the only ways to preserve ornamental box trees [4]. Increasingly more garden owners want to care for their ornamental box trees in an environmentally friendly way. In this context, it is important to note that biological insecticides based on Bacillus thuringiensis can also have indirect negative effects on other organisms, such as birds that eat caterpillars or invertebrates that eat dead caterpillars [20]. The toxin produced by B. thuringiensis can also accumulate in the soil, affecting soil invertebrates, fungi, and other bacteria [20,21]. In our study, we presented a novel method that effectively protects box trees from being damage by C. perspectalis larvae without the use of toxic substances.

Horticultural box trees are of great cultural and historical importance in parks, cemeteries, and private gardens [14]. Buxus spp. are one of the oldest and most commonly cultivated ornamental shrubs, grown for their diverse forms and evergreen foliage [22]. Many garden owners have cared for their box trees for decades and are, therefore, prepared to make greater efforts to protect them against C. perspectalis. What is less known is that numerous species live exclusively on Buxus spp. Mitchell et al. [14] reported that 43 fungi, 3 algae, and 18 invertebrate species have been recorded only on Buxus spp. in Europe and the Caucasus, suggesting that these species are obligate on box trees and are most at risk if these trees disappear. Temporarily covering the box trees with a net could have little disadvantage for any associated species.

Exclusion nets are increasingly being used as a sustainable alternative to chemical control to protect a variety of agricultural crops from insect pests [23,24]. In our study, we demonstrated that this form of protection can also be successfully used on a valuable horticultural tree. Females of C. perspectalis cannot lay their eggs on the leaves if the box trees are covered with a net. Various insect nets, such as those available in garden shops, can be used for this purpose. However, the mesh size should be small (<0.2 mm). Our method is cost effective. We produced the covers ourselves, with material costs being less than EUR 1 per cover. Nets made of both biodegradable material and synthetic material, like plastic, can be used. Synthetic nets can be used for several years but are less environmentally friendly to produce and dispose of.

The timing of covering is most important. The trees should be covered during the active flight periods of the moths. In Basel, as in other warmer regions of Europe, C. perspectalis has three generations per year [3]. Moths of the first generation occur from May to July and those of the second and third generations from August to the end of October. Our light trap data from 2021–2023 confirm this seasonal pattern. The observed temporal occurrence of C. perspectalis in the 3 years also largely corresponded with the flight activity recorded in Basel in preceding years (2009–2020 [25]). In most cases, continuous monitoring of C. perspectalis moths is not possible. We, therefore, generally recommend using protective nets for 5–6 months.

For box trees that are exposed to extreme sunlight, it should also be ensured that the net is not directly on the leaves. Spacers, e.g., in the form of empty PET bottles that are inserted upside down into the tree, can be useful for this purpose. The method described here is best suited for solitary box trees. Boxwood hedges are much more difficult to wrap with a net so that no C. perspectalis moths can penetrate.

The success of protection with a net that we described is comparable to the success of repeated use of a bioinsecticide (box trees of group 2 in our experiment), but the time required for the garden owner is significantly less. For the optimal effect of a bioinsecticide, the box trees should be sprayed every 3 to 6 weeks and even earlier after heavy rainfall. Because the tiny young larvae hiding between the leaves and their feeding marks can easily be overlooked [11,26], spraying is often carried out too late. These disadvantages do not exist when protecting the trees with a net.

Heavily damaged and completely defoliated box trees are often removed by garden owners and park managers. Most of these trees are still alive, provided their bark has not been removed by grazing larvae. The trees will produce new leaves again in the following years. However, the emerging leaves must be protected from further damage. We showed that with temporary use of a protective net, completely defoliated box trees can slowly recover. Field data show that completely defoliated natural box tree stands in northwestern Switzerland recover within 4–6 years, provided no further C. perspectalis larvae attack the trees [15]. This suggests that garden owners should be patient with their box trees.

5. Conclusions

Many garden owners want to manage their gardens in an environmentally friendly way and no longer use insecticides. The results of our study showed that box trees can be protected from attacks of C. perspectalis by temporarily covering them with a fine-mesh net. This cost-effective method also allows defoliated trees to regenerate and may have minor effects on other invertebrates.