Optimization of Traps Used in the Management of Monochamus galloprovincialis (Coleoptera: Cerambycidae), the Insect-Vector of Pinewood Nematode, to Reduce By-Catches of Non-Target Insects

Abstract

A possible tactic to survey and control Pine Wilt Disease is the use of semiochemical-baited traps to capture the insect-vector, the pine sawyer Monochamus galloprovincialis (Olivier) (Coleoptera: Cerambycidae). The most common chemical lure used is the Galloprotect Pack, which includes the aggregation pheromone ([2-undecyloxy] ethanol), a host monoterpene (α-pinene), and bark-beetle pheromones (ipsenol and 2-methyl-3-buten-1-ol). This lure also attracts non-target species, including bark beetles (Coleoptera: Curculionidae: Scolytinae) that use ipsenol (Ips sexdentatus (Boerner)) and 2-methyl-3-buten-1-ol (Orthotomicus erosus (Wollaston)) as pheromones, but also large numbers of their natural enemies, Temnoscheila caerulea (Olivier) (Coleoptera: Trogossitidae), Aulonium ruficorne (Olivier) (Coleoptera: Colydiidae), and Thanasimus formicarius (L.) (Coleoptera: Cleridae), and other saproxylic insects (Coleoptera: Cerambycidae). These catches cause a decrease in biodiversity of the forest insect communities, and the removal of predatory insects may favour bark beetle outbreaks. Thus, our project objective was to test trap modifications to try to reduce catches of non-target insects. Modifying the multifunnel trap’s collection cup by placing a 0.5 cm mesh in the drainage hole allowed the escape of all predator beetles (Cleridae, Trogossitidae, Colydiidae, and Histeridae) in 2020, and retained only two Trogossitidae in 2021, against 249 specimens caught in the non-modified collection cup. This simple modification thus allowed the escape of almost all predators, while maintaining the traps’ efficiency at catching the target species, M. galloprovincialis.

1. Introduction

Pine wilt disease (PWD) is a complex system that involves the infection of susceptible host trees by the pinewood nematode (PWN) Bursaphelenchus xylophilus (Steiner & Buhrer 1934; Nickle, 1970), through the activity of its insect vectors, the pine sawyer from the Monochamus genus. PWN is native to the North American east coast, but has now invaded many parts of the world, including Japan (1905), China (1982), Korea (1988) [1,2,3,4], and more recently, Europe. Since its detection in Portugal in 1999 [5], it has dispersed to Spain [6,7] and Madeira [8].

As a quarantine organism in Europe, strict measures have been implemented to prevent the progression of PWN to new areas. Globally, in almost all countries with PWN infestations, detection efforts are carried out annually, both for symptomatic trees [1,2,9,10] and for its insect-vectors during their flight periods [1,2,9,10,11,12,13,14,15,16,17], and all the material collected (wood samples and captured insects) is sent to a specialized laboratories for PWN screening.

Aerial or ground application of chemical insecticides has been used to control Monochamus adults in PWD-affected Asian countries [15]. However, in addition to the cost of insecticide applications over large areas, such applications can cause serious side effects, such as environmental contamination and toxicity to non-target organisms [14,18].

The strategy to control PWD caused by PWN worldwide focuses on infested host trees, with winter surveys of conifer forests for the detection and removal of infested trees, initially implemented in Japan [1] and followed in China [2,9], Korea [10], and Portugal [11]. A complementary measure targets the insect-vector populations by trapping during their flight periods, to decrease vector populations and slow the disease spread [14,17].

For the capture of Monochamus spp. in Europe, standard protocols use Teflon-coated black Lindgren multifunnel or black cross-vane traps [19], spaced 250–300 m apart [20] and lured with Galloprotect Pack (SEDQ, Barcelona, Spain), which includes the host volatile α-pinene, two bark beetle pheromones which act as synergists (ipsenol and 2-methyl-3-buten-1-ol) and the insects’ aggregation pheromone ([2-undeciloxyl] ethanol), known as monochamol [21]. Although developed for M. galloprovincialis (Olivier), this combination proved to be an effective lure for other Monochamus species in Europe [22] and in the PWN native range of eastern North America [23,24,25].

The Monochamus species evolved to be attracted by bark beetle pheromones [26,27,28,29,30,31,32,33], because they need weakened or recently dead hosts for egg-laying. If Monochamus spp. arrive while a tree is still under bark beetle attack, not only are the tree tissues more nutritious, but the larvae can enhance their nourishment through intraguild predation on bark beetle larvae [27,34].

However, when a trap is placed to capture Monochamus spp., using lures that include host tree volatiles and bark beetles’ pheromones, many other non-target Coleoptera are caught. Non-target species usually found in these traps include other large wood borers (Cerambycidae and Buprestidae) and smaller Cleridae, Colydiidae, Histeridae, and Trogossitidae, which include beneficial predators of xylophagous and bark beetles [23,35,36]. Catches of non-target species have been reported while attempting to control a number of forest pest insects [37,38,39]. These unintended catches of non-target species can have considerable negative impacts, reducing biodiversity and populations of beneficial predatory insects. These difficult issues have led some to question the wisdom of using baited traps to assess the spread of PWD and other diseases in several European countries [40,41,42,43]. Attempts have been made to reduce bycatch by modifying conventional traps, but with very limited results [39,44].

In this context, our objectives were to carry out a quantitative assessment of the negative effects of the M. galloprovincialis trapping methodology on non-target insects and to test modifications to the insect collection containers to allow smaller insects to escape. Here, we report that a simple modification to the collection cup can eliminate by-catch of most nontarget insects.

2. Materials and Methods

2.1. Assay Location and Time of Experiment

Field experiments were carried out over two consecutive years (24 July–21 October 2020; 5 July–27 September 2021) in a pure 40 years old maritime pine (Pinus pinaster) stand in Central Portugal (Seia municipality) (40°16′22″ N 7°42′12″ W), infested by PWN and with confirmed presence of insect vectors.

2.2. Experimental Design

The traps used in the assays were variations of black 12-element multi-funnel traps (WITASEK PflanzenSchutz GmbH, Vienna, Austria). All traps were baited with Galloprotect 2D Plus (SEDQ, Barcelona, Spain), and placed 50 m apart, hung from live branches in the lower canopy of dominated pines, at 6 an average height, in a Latin-square experimental design (3 different types × 3 repetitions). The distance between traps overlapped the effective areas of neighbouring traps (100–150 m radius) [20] in order to maximize the probability of catching beetles. Traps were serviced and rotated weekly (indicated by arrows in Figure 1) to minimize the location effect on captures. Lures were renewed after 6 wk, according to the manufacturer’s recommendation, and the experiment was terminated after 12 wk, after each trap type was exposed in every position, on four different occasions (Figure 1).

Traps used in 2020 trials were the model commercially available in 2012 (V), the 2019 commercial model (C) (Figure 2), and the same 2019 model with a modified collection cup (M), where the bottom of the cup was removed and replaced by a 0.5 cm plastic mesh (Figure 3). The number of funnels and consequently the size of the 2012 and 2019 trap versions were similar, although there were some minor differences in the funnel connections and the collection cup design, which was larger and shorter in the 2012 model.

In 2021 assay, three versions of the 2019 model were tested with the same Latin-square experimental design (Figure 1): the 2019 commercial model (C), the modified collection cup tested in 2020 (M) and a prototype collection cup made with 0.5 cm metallic net in the drainage hole (V), prepared by the manufacturer, WITASEK PflanzenSchutz GmbH, (Vienna, Austria) (Figure 3).

In both years, insects captured weekly in each trap were identified to species level, using morphological taxonomic keys [45,46,47], and counted.

2.3. Data Analysis

The results of each experiment were statistically analysed to determine significant differences between the evaluated parameters (number of specimens caught in the traps of Cerambycidae, Scolytinae, Trogossitidae, Colydiidae, Cleridae, and Histeridae). Since the number of different insect species caught per trap and per date of collection was frequently zero and total values were quite small, a square root transformation was applied to the raw data [48]. One-way analysis of variance (ANOVA) was used to compare mean values of captured beetles per trap and collection, followed by post-hoc least significant difference (LSD) tests to establish homogeneous groups with a confidence level of 95% (α = 0.05), using the Statistica software package, version 12 [49].

3. Results

In 2020, a total of 67 specimens of the target M. galloprovincialis were caught, with a female-biased sex ratio (3.8:1). No statistically significant differences between the number of specimens caught among the tested trap types were found, F_{(2, 105)} = 0.7932; p = 0.4551. The multifunnel 2012 model and the 2019 model with a modified cup caught 27 and 25 specimens, respectively, compared to the 2019 model (12 beetles). The results obtained by the modified cup revealed that replacement of the cup bottom with plastic mesh did not allow the target species to escape, thus not affecting the main purpose of this trapping procedure.

In this assay, several other cerambycid beetles were caught, in lower numbers, and were equally distributed among the trap types. These included 21 Arhopalus syriacus (Reitter), 9 A. ferus (Mulsant), 9 Spondylis buprestoides (Linnaeus), and 2 Acanthocinus griseus (Fabricius), which were found in the 2012 trap (Figure 4).

Of the 2385 scolytids caught, most were Orthotomicus erosus (Wollaston) (92.1%) and Ips sexdentatus (Boerner) (5.5%), with few remaining specimens being Hylurgus ligniperda (Fabricius) (32 beetles) and Tomicus sp. (10 beetles) (Figure 5A,B). As expected, no scolytids were found inside the modified collection cup, because all of the above bark beetle adults are less than 3 mm in length, and so could easily escape through the plastic mesh, resulting in highly significant statistical differences when compared with the other trap models, F_{(2, 105)} = 16.0165; p < 0.0001 (Figure 5C). The 2012 model was more efficient at catching these nontarget species, capturing more than twice as many (1596 specimens) as the 2019 trap (789 specimens) (raw data is available in the supplemental material file).

During the 2020 assay, 133 predatory beetles were caught in the traps, including 99 Temnoscheila caerulea (Olivier) (Trogossitidae), 28 Aulonium ruficorne (Olivier) (Colydiidae), 5 Thanasimus formicarius (L.) (Cleridae), and 1 Histeridae specimen (Figure 6).

The results obtained for the predatory beetles were analogous to those described for scolytids, with no specimens recovered from the traps with the modified cup, although T. caerulea is a considerably bigger beetle than the scolytids and other predatory species (Figure 6). Statistically significant differences were observed between trap types, F_{(2, 105)}= 7.4906; p = 0.0009, with the modified collection cup being different than the two other trap types (Figure 7).

In the 2021 assays, fewer M. galloprovincialis were caught (43 adult beetles), with no statistical differences in numbers collected in the different trap types, F_{(2, 105)} = 1.5591; p = 0.2152. Eight specimens were found in the traps with the modified cup (the same used in 2020), while in the 2019 trap model and the traps with the prototype cup (with a net in the drainage hole), 18 and 17 beetles were caught, respectively.

Regarding the other cerambycid species trapped in 2021, higher numbers of A. syriacus, A. ferus, and S. buprestoides were captured in the 2019 trap model, thus revealing these saproxylic beetles were also able to escape from modified cups (Figure 8). No specimens of A. griseus were found.

In the 2021 assay, the collection cup with mesh on the bottom or in the drain hole (Modified and Prototype) allowed the escape of all scolytids, as in 2020, resulting in highly significant differences when compared to the commercial 2019 model, F_{(2, 105)} = 15.0404; p < 0.0001. The number of scolytids caught in 2021 was considerably less (44 specimens) than in the 2020 assay (7 O. erosus, 29 I. sexdentatus, 7 H. ligniperda, and 1 Tomicus spp.) (Figure 9).

Finally, during the 2021 assay, 249 predators were caught, with the Trogossitidae family (T. caerulea) contributing 88% (219 specimens), followed by Colydiidae (23 specimens), 6 Cleridae (T. formicarius), and 1 Histeridae specimen.

Almost all specimens were caught in the commercial 2019 model cup, with the exception of two T. caerulea found in the prototype cup. Consequently, there were highly significant differences, F_{(2, 105)} = 30.0779; p < 0.0001, between the number of predator specimens caught in the different trap types (Figure 10).

4. Discussion and Conclusions

One strategy for the management of PWD-affected pines involves trapping its insect vectors, cerambycids in the genus Monochamus. Typically, pine sawyers are considered secondary species, needing other abiotic agents (e.g., severe drought or forest fires) or biotic agents (e.g., bark beetles) to weaken host trees so that they can be successfully colonized. The co-evolution of these species to respond to host volatiles and bark beetles’ pheromones allows them to efficiently locate suitable host trees still under attack, and this selectivity is likely chemically mediated [50,51,52].

Traps used for pine sawyers have included water pan traps, cross-vane traps with a black silhouette [53,54], black multi-funnel traps [19,55], or transparent cross-vane traps [17,19]. Trials helped demonstrate that either a 12-element black multi-funnel or a black cross-vane trap performed similarly in capturing M. galloprovincialis [19].

North American researchers determined that M. titillator (F.) was attracted to a blend of bark beetle pheromones and turpentine tree host volatiles [56,57]. Analogous attraction was later confirmed with other Monochamus species, such as M. clamator (leConte), M. scutellatus (Say) [27,51], and the European M. galloprovincialis [17,29].

However, lures consisting of a combination of host tree volatiles and bark beetle pheromones, not surprisingly, also attract significant numbers of non-target insects, including useful saproxylic species, and beneficial natural enemies of pest species [44,58,59,60,61,62].

In forests, a plethora of organisms contribute to nutrient cycling, canopy thinning, gap dynamics, soil structure, hydrology, and successional pathways [63]. Useful predatory species can exploit the pheromones of their prey, or other indirect chemical cues such as the volatiles emitted by trees under attack, to locate their prey [40,41,42]. Bakke and Kvamme [64] suggested that predators may also be attracted to volatiles emitted by captured beetles inside a trap.

The main insect predator species caught in our study, including T. caerulea (Trogossitidae), A. ruficorne (Colydiidae), and T. formicarius (Cleridae), had previously been collected in traps baited with pheromone blends of I. sexdentatus [41,60,65,66,67,68] and are considered important bark beetle natural enemies, contributing to their population control by preying on the last instar larvae and callow adults under the bark [62,68,69].

Catching these predators constitutes a problem for bark beetle management and forest insect diversity and equilibrium. Trapping control techniques must strive to be sustainable, impacting primarily or preferably exclusively the target species. Significant catches of nark beetle predators in commercial trap designs can contribute to undesirable effects on pest populations [70]. For example, it was estimated that a reduction of predator densities could lead, within 4–20 bark-beetle generations, to a doubling of the pest population, prolonging bark beetle outbreaks [71]. The use of commercial multifunnel traps to control the PWN insect-vector has this unwanted effect on the bark beetles’ populations that already benefit from PWN-infested trees as breeding substrate, since all surveyed pines were colonized by bark beetles, O. erosus, at least [72].

The results of our study demonstrate the low specificity of the Galloprotect Pack lure, as well as its possible negative effects against beneficial insects when used with the commercially available traps. However, simple modifications to the collection cups essentially eliminated catches of all non-target species, including other saproxylic cerambycids (Arhopalus spp. and S. buprestoides), bark beetles, and predatory species.

However, the fact that the modified cup does not retain bark beetles is apparently not desirable for general pine forest management, because I. sexdentatus and O. erosus are pests in the Mediterranean area that cause considerable mortality and must be controlled. The inclusion of Monochamus pheromone in the Galloprotect Pack lure has been demonstrated to have a strong repellent effect on bark beetles, drastically reducing their catches [23]. Therefore, for the integrated protection of pine forests, it will probably be necessary to simultaneously deploy traps with specific lures for bark beetles.

Finally, considering that the average size of most Monochamus species is similar, the modified collection cup can be effective worldwide in retaining mainly the target Monochamus and allowing the escape of smaller non-target species. However, testing of the modified collection cup in new areas with new species would probably be prudent before deploying the modified trap in operational programs.