Effective Reduction in Natural Enemy Catches in Pheromone Traps Intended for Monitoring Orthotomicus erosus (Coleoptera, Curculionidae)

Abstract

Infestations have persisted following a sudden and intense outbreak of the bark beetle Orthotomicus erosus along the Croatian coast, necessitating a continuous battle against this pest. A recommended protective action is the utilization of pheromone traps for population surveillance. Previous monitoring efforts have recorded an exceptionally high capture rate of natural enemies using pheromone traps; these traps inadvertently prevented natural enemies from fulfilling their essential role in controlling bark beetle populations. To address and significantly diminish instances of this unintended capture, our study designed a modification to the Theysohn-type pheromone trap by integrating a metal mesh within the trapping container. An experimental setup was established in Marjan Forest Park, situated on a peninsula bordered by the sea on three sides and partly by the city of Split. For monitoring purposes, unmodified standard pheromone traps were deployed at the onset of a significant O. erosus outbreak in Croatia in 2018. Catch data from 2020 to 2022 show a marked decrease in the bark beetle population, indicating a shift toward a latent phase. In 2022, modified traps were integrated into the existing monitoring setup, consisting of 10 pairs, to evaluate whether modifications to the traps could significantly reduce the capture of the bark beetle’s natural enemies, specifically Temnoscheila caerulea, Thanasimus formicarius, and Aulonium ruficorne. The objective is to offer recommendations for forestry practices on employing pheromone traps with minimal disturbance to the ecological equilibrium. Our findings indicate that the modifications to the traps markedly decreased the capture of natural predators, particularly T. caerulea, which was the predominant predatory insect found in the traps. Simultaneously, captures of the target species, all bark beetles in the trap, were marginally reduced. This decrease in the capture rates of the target bark beetle species, O. erosus, is not considered problematic when pheromone traps are utilized primarily for monitoring purposes. The modifications to the traps significantly reduced the capture of common bark beetle predators, thereby facilitating a more balanced strategy in forest protection efforts.

1. Introduction

Bark beetles are considered one of the most damaging groups of insects to forests worldwide [1]. Out of this large group of insects, there are not many that cause huge losses in forestry. For the past few years, outbreaks of the Mediterranean pine engraver (MPE), Orthotomicus erosus Woll. (Coleoptera, Curculionidae, Scolytinae) on Aleppo pines (Pinus halepensis Mill.) have become a regular threat in Croatia [2]. Climate change, extreme drought, and secondary bark beetle attacks eventually led to enlarged bark beetle population levels, which then started attacking healthy trees. This increase in bark beetle populations in conifers is well known and documented in European forestry, most commonly for the spruce bark beetle (Ips typographus L.) [3,4,5,6,7]. Similar scenarios where MPEs exhibit aggressive behavior and act as primary pests have been documented in Dalmatia, the Southern Mediterranean part of Croatia [2]. Previously, concerning bark beetle attacks in this specific region, only very local outbreaks of Tomicus piniperda L. or T. destruens Woll. had been recorded [8,9]. The MPE is a reddish-brown beetle that is 2.7 to 3.5 mm long. It tunnels in the living part of the bark; its larvae are legless, white, and about 2.7 to 3.5 mm long; and their appearance does not change as they grow. The eggs are white, partly transparent, and about 1 mm long. The MPE is naturally distributed in Central Asia, the Middle East, Europe, and China. Although it is widespread throughout Europe, so far it has only caused damage in very warm Mediterranean areas. In France, Morocco, and Turkey, two generations have been identified per year, with three to four in South Africa and Tunisia, and three to five in Israel, where adults are active from March to October. Although previous research identified three [10] or four generations [11], more recent studies have shown that even five generations could be in the outbreak site in Marjan (Split, Croatia) [2]. A reason for this could be climate changes, primarily aridification, and a significant extension of the vegetation period in this region, with an unusually warm November and December (Pernek unpublished data). The MPE is also associated with xylophagous fungi, which probably play a role in the colonization of the host trees [12].

MPE attacks, as described in warmer Mediterranean regions [11,13,14,15,16], have never occurred before in Croatia [2]. The behavior of pest insects might change because of climate change, so mortality, reproduction, voltinism, and spatial distribution may be factors that favorably affect the pest insect [2,17,18]. Due to the extension of the growing season, the production of several generations above the usual number per year allowed for an exponential population growth and the development of an outbreak that destroyed 50% of the trees in the Marjan Forest Park after 4 years of heavy attack (Pernek unpublished). Control measures include the prompt cutting of infested trees and the use of bait logs, with pheromones used only for monitoring purposes. A mass trapping experiment was conducted in 2019, but it has not shown good results [18].

In general, pheromone traps of various designs are used for insect attraction, mostly for evaluating population densities, i.e., monitoring. Mass trapping is an exception [19,20,21]. Concerning the design and placement of pheromone traps, there is a large number of studies in the literature, especially for the most researched species of bark beetle: I. typographus [22,23,24,25]. In designing optimal lures and pheromone traps for the MPE as a target species, Erosowit^® (Witasek, Feldkirchen, Austria) has exhibited significantly higher catch rates compared to Pheroprax^® (BASF, Ludwigshafen, Germany) and it has also been significantly more selective, while single Theysohn traps (Theysohn, Salzgitter, Germany) have been shown to be an optimal solution [18]. Although traps are designed to target specific insects, such as other bark beetle species, many other species are also attracted. Mass trapping can unintentionally remove high numbers of predators that use bark beetle pheromones as kairomones [19] (Aukema et al., 2000), and such negative side effects have negative impacts [26]. For example, in the pheromone Pheroprax^®, components like ipsenol, ipsdienol, and (S)-cis-verbenol are responsible for attracting the antagonist Thanasimus dubius Fab. (Coleoptera, Cleridae) (Aukema et al., 2000). Pheromone traps have been used in Marjan to attract the MPE since 2018. Early on, it was apparent that many specimens of Temnoscheila caerulea Olivier (Coleoptera, Trogossitidae), an important predator of bark beetles, were present along with the target organism. This is concerning [19], because four or more generations of MPE have been detected each year [2,27], and a high capture rate of natural enemies should be avoided even if the traps are intended for monitoring or mass trapping. According to Martin et al. [28], the number of catches of T. caerulea could be reduced by modifying the pheromone traps.

Pheromone traps have been used for several years to capture MPEs. Throughout the years, the capture of non-target entomofauna, particularly predators, has been evident and identified as a major problem. The predatory species remain trapped and feed on the catches of other insects. Sometimes, in specific periods, the number of predators has been approximately equal to the number of MPEs, which is a major drawback of the pheromone trap. This was particularly evident when a large number of pheromone traps were used in an attempt to reduce the number of MPEs. Therefore, developing a way to reduce the number of non-target species has been necessary, especially concerning predatory and entomofauna captures. The use of protective nets in pheromone traps is not new, and their use has often proven effective for several bark beetle species. However, to our knowledge, there have been no such studies for MPEs. Now that the MPE has established itself in Croatia as a pest with the potential to build populations to levels that cause damage over large areas, this research should serve as a basis for future integrated forest protection. When used against the MPE, pheromone traps baited with an Erosowit^® lure catch large numbers of natural enemies; hence, the aims of this work were to (i) monitor the population dynamics of the MPE and (ii) improve trapping protocols in order to decrease unintentional captures.

2. Materials and Methods

2.1. Field Work

This study was carried out between 2020 and 2022 in the public forest (forest park) of Marjan, close to the city of Split, in a 200 ha Allepo pine stand. Marjan is a hill peninsula that was declared a forest park in 1976 and is located just a couple of hundred meters from Diocletian’s palace, which is in the center of Split. It is of great importance to the citizens of Split as a recreational area and historical site. The whole protected area of Marjan is ca. 300 ha in size, with forest making up approximately two-thirds of it. Throughout history, it was covered with holm oak and manna ash forest (Fraxino orni, Quercetum ilicis) that has gradually degraded. Today, it is mostly dominated by Aleppo pine. Geographically, Marjan is a peninsula, surrounded by the sea and connected to the mainland by a narrow strip on which the city of Split continues. Because the influence of the bark beetles’ arrival from the outside is almost negligible, this field experiment could be considered an open-air laboratory experiment.

Since 2017, MPE attacks have occurred regularly in the Marjan Forest Park [2]. Some parts of the forest have a greater proportion of deciduous trees, mostly holm oak (Quercus ilex L.), and these parts were avoided when installing the traps. Pheromone traps were set in early spring at the end of March 2020, 2021, and 2022, so the first sampling was always conducted on the last week of March. In total, 10 standard Theysohn traps were set each year, separated from each other by at least 100 m (Figure 1). In 2022, 10 modified traps were added as a pairing to the 10 standard traps (Figure 1). The modification was to avoid the capture of non-targeted insects. For this purpose, a special steel mesh with a 6 mm mesh screen was developed from scratch and installed on the collection container (Figure 2). The trap pairs were placed 20 m apart and at least 20 m from a healthy pine tree.

The catches in the pheromone traps were collected mainly from 7 to 14 days (9 days on average) depending on the weather conditions. The pheromone ampoules Erosowit^® were changed five times a year each time the liquid ran low, which was temperature-dependent. The catches were transferred into plastic trays, mixed with ethanol (70%), and labeled. Then, they were transported to the laboratory (Croatian Forest Research Institute, Jastrebarsko, Croatia) and examined under a microscope (Olympus BX2, Tokyo, Japan) after 2 weeks maximum.

2.2. Laboratory Work

A catch analysis was conducted on the pheromone traps in the laboratory for entomological research at the Croatian Forest Research Institute (Jastrebarsko, Croatia). The laboratory processing involved drying the insects at room temperature and sorting the species under a microscope. The insects were first sorted by taxonomic categories and dried to facilitate counting. Based on the taxonomic categories, the insects were identified using the available morphological keys [29]. The counting of the sorted catch was conducted manually. All larger insects, such as longhorn beetles, beetles with equally sized wings, and natural enemies, were separated and counted by hand in separate Petri dishes.

2.3. Statistical Analysis

A statistical analysis was conducted using R software version 4.2.2 [30]. To assess the statistical significance of the differences in the average number of beetles captured by the standard and modified pheromone traps, we employed the permutation paired t-test from the RVAide Memoire package version 0.9-81-2. Permutation methods were chosen due to their numerous benefits over conventional statistical tests. These methods excel in handling multiple comparisons and reducing the Type I errors that arise when numerous hypotheses are tested simultaneously. Unlike traditional tests, permutation tests do not presume any specific population distribution, making them versatile across various data types, including non-normal and non-parametric datasets. They depend on the permutations of the data rather than the actual data values, rendering them more resilient to outliers and other forms of data anomalies.

The effectiveness of the standard and modified traps was gauged by the proportion of captured species. The statistical significance of the differences between the proportions of beetle species captured in the standard and modified traps was assessed using the two-sample test for the equality of proportions, with a continuity correction.

3. Results

Between 2020 and 2022, catches were counted in all the traps to assess the population status of bark beetles. By adding modified traps in 2022, attention was focused on non-selective capture, i.e., the results obtained show to what extent, and to which predatory insects, the trap modification is beneficial. Ultimately, other targeted species of entomofauna were observed in the regular and modified pheromone traps (Table S1).

3.1. Catches of Target Organism MPE

MPEs caught in pheromone traps between 2020 and 2022 were compared with previous years [18]; the catches were constantly decreasing, indicating a population decline. Additionally, the catches between 2020 and 2022 show the collapse of the outbreak very clearly (Figure 3).

3.2. Efficiency of Modified Traps

In 2022, we recorded the number of beetles caught in standard and modified traps (

Table 1. Number of beetles caught in standard and modified traps during the entire 2022 season.

Species	Standard Traps				Modified Traps				Total
	Total	Max	Mean	St. Dev.	Total	Max	Mean	St. Dev.
Orthotomicus erosus	1645	472	164.5	137.4	1017	242	101.7	65.6	2662
Tomicus destruens	176	32	17.6	7.8	135	34	13.5	9.4	311
Hylurgus miklitzi	26,370	5622	2637.0	1543.9	21,051	4119	2105.1	1412.4	47,421
Other Scolytinae	152	67	15.2	19.3	366	157	36.6	46.6	518
Temnoscheila caerulea	874	188	87.4	48.1	486	125	48.6	32.5	1360
Thanasimus formicarius	60	23	6.0	6.8	22	7	2.2	2.0	82
Aulonium ruficorne	111	30	11.1	10.0	70	46	7.0	13.8	181
Monochamus galloprovincialis	30	14	3.0	4.1	1	1	0.1	0.3	31
Buprestidae	446	220	44.6	64.1	33	15	3.3	4.6	479
Other	229	51	22.9	16.7	104	18	10.4	5.0	333
All species	30,093	5970	3000.9	1666.3	23,285	4333	2328.5	1484.7	53,378

). We did not use the sample from trap pair #3 on 25 April 2022 for the evaluation, because the standard trap was damaged and the number of beetles could not be determined.

Although the pairs of standard and modified traps were placed close to each other and under similar conditions, the number of beetles captured was different (Figure 4). The average number of beetles captured in the standard traps was statistically greater than the average number of beetles captured in the modified traps (permutation paired t-test, p = 0.03).

The largest differences in the number of beetles captured in the standard and modified traps were found in trap pairs #7 and #10. In both cases, several times more beetles were caught in one type of trap than in the other during the summer (Figure 5 and Figure 6).

According to our observations, the efficiency of the modified traps was about 25% lower than that of the standard traps. We can see that fewer beetles of each species, except for Scolytinae, were caught in the modified traps than in the standard traps (Figure 7).

The average number of O. erosus captured in the standard traps was statistically higher than the average number captured in the modified traps (Table 1), as determined by the permutation paired t-test (p = 0.004). The substantial disparity in MPEs captured between the standard and modified traps is primarily attributed to traps #7 and #9 (Figure 8, Figure 9, Figure 10 and Figure 11). For both pairs of traps, the peak number of beetles captured occurred on the 237th day of the year (25 August 2022).

3.3. Effectiveness of Modified Traps

The main objective of this study was to develop and test modified traps that reduce the risk of trapping natural enemies of bark beetles. Therefore, in the second part of this study, we evaluated the effectiveness of the modified traps in separating bark beetles, their natural enemies, and other non-target species. We expressed the effectiveness of the traps by the proportions of target, non-target, and predatory species captured (

Table 2. Percentage of beetle species captured in standard and modified traps.

	Species	Percentage of Beetle Species in Traps (%)
		Standard	Modified	LL	UL	p
Targetspecies	Orthotomicus erosus	5.5	4.4	0.73	1.47	<10−8
	Tomicus destruens	0.6	0.6	−0.13	1.39	0.98
	Hylurgus miklitzi	87.6	90.4	−3.31	−2.24	<10−15
	Other Scolytinae	0.5	1.6	−1.25	−0.88	<10−15
	Total	94.2	96.9	−3.10	−2.39	<10−15
Predatory species	Temnoscheila caerulea	2.9	2.1	0.55	1.08	<10−8
	Thanasimus formicarius	0.2	0.1	0.04	0.17	<0.01
	Aulonium ruficorne	0.4	0.3	−0.03	1.70	0.20
	Total	3.5	2.5	0.70	1.28	<10−10
Non-target species	Monochamus galloprovincialis	0.1	0.0	0.05	0.14	<10−4
	Buprestidae	1.5	0.1	1.19	1.49	<10−15
	Other	0.8	0.4	0.18	0.45	<10−5
	Total	2.3	0.6	1.55	1.95	<10−15

L_L—the lower limit of the 95% confidence interval; U_L—the upper limit of the 95% confidence interval; p—p-value of the proportionality test.

). Consistent with the intent of the modified trap design, the proportion of beetle predators and non-target species captured decreased (Table 2). The proportion of predatory species captured in the modified traps decreased by 1% compared to standard traps. However, the overall efficiency of the modified traps was lower than that of the standard traps. This lower efficiency can be compensated for by placing additional modified traps. For the same anticipated total number of beetles captured, approximately 30% fewer predatory beetles will be captured in the modified traps than in the standard traps. The decrease in the proportion of non-target species was even more significant. For the same projected number of beetles captured, up to four times fewer non-target beetles are expected in the modified traps than in the standard traps. These decreases in the proportions of predator and non-target beetles in the modified traps were statistically significant.

4. Discussion

Although the catch numbers in different types of traps may vary significantly, they still provide information on the population status of bark beetles, indicating that the number of individuals caught is not as important [19]. Generally, pheromone traps do not have a curative function [19,31]. However, the aim is to catch a larger number of target insects, thereby excluding them from the population. However, natural enemies are extremely important in regulating bark beetle population dynamics [32,33]. They contribute to self-sustaining biological control over large areas and long timeframes [34,35]. The selectivity of the pheromone trap is therefore much more important, because by excluding predators from the population, their positive ecological impact is absent [36]. In the extreme case, the number of predators caught and removed from the population is so high that, considering their potential food consumption, we may cause more harm than good. The catch of predators may be so high that the number of bark beetles that would be destroyed in nature becomes significantly smaller than in traps [37]. Furthermore, the removal of predators captured through mass trapping may prolong bark beetle outbreaks [38].

The number of MPE catches in 2022 is negligible compared to catches from previous years, representing the bark beetle outbreak’s entry into latency. This is somewhat expected because of the high mortality rate of pine trees caused by the large-scale MPE attacks in previous years, where large numbers of larvae were found under bark. Due to competition within the species (intraspecific competition), the bark beetles do not have the space for normal development anymore [39]. This results in a reduction in their numbers and a decrease in brood productivity. This regulatory mechanism has been shown to be powerful in other species of bark beetles [40]. Additionally, their natural enemies have strengthened and created a strong barrier to the MPE population’s resurgence.

Previously used pheromones and traps were used in accordance with the literature [13] or based on research or experience gained in the first years of the outbreak [2]. In this study, the difference in the catches of target insects (MPEs) and non-target insects (predators) was analyzed to provide advice for future monitoring. By placing traps in the same location year after year, we can gain insight into population movements or trends that may help experts implement protection measures correctly and in a timely manner. This study compared pheromones and traps for the capture of MPEs, representing the first such research in Croatia, and the results should be valuable to forestry professionals in the field.

Our results show that the average number of O. erosus captured in the standard traps is statistically higher than that in the modified traps (Table 1), as determined by the permutation paired t-test (p = 0.004). The substantial disparity in MPEs captured between the standard and modified traps is primarily attributed to traps #7 and #9 (Figure 8, Figure 9, Figure 10 and Figure 11). For both pairs of traps, the peak number of beetles captured occurred on the 237th day of the year (25 August 2022).

The difference in the number of MPEs in the standard trap and modified trap #7 is less extreme. However, the trend in the number of MPEs in modified trap #7 does not match that in standard trap #7. This may be explained by the placement of modified trap #7.

Consistent with the intent of the modified trap design, the proportion of beetle predators and non-target species captured decreased. The 1% reduction in the proportion of predatory species in the standard traps indicates that, given the same total number of beetles caught, the modified traps captured about 30% fewer predatory beetles compared to the standard traps. The catch data from the last three years of research show a clear declining trend for the outbreak. The catches of both target and non-target entomofauna should be viewed in this context. Predator catches were much higher in the previous two years, which could be observed during the emptying of each trap; however, the data were not quantified. However, the predator–bark beetle ratio is likely even more unfavorable than our data suggest.

If we look at the non-target species as a whole, the decrease in proportion was even larger. For the same number of all beetles captured, up to four times fewer beetles of non-target species were captured in the modified traps than in the standard traps. The decreases in the proportion of beetles from predators and non-target species in the modified traps were statistically significant.

The natural enemy T. caerulea completely follows the population dynamics of the Mediterranean bark beetle. The data obtained are consistent with the literature [41], but they provide completely new insights into the number of generations and the first occurrences of these species during the year.

When reaching a high population density, MPEs can overwhelm the host defenses, thus causing tree mortality [42]. The MPE can be a serious pest in semi-arid regions, as confirmed by findings on living trees [27], because it is capable of colonizing living trees using the host exhaustion strategy, whereby employing aggregation pheromones attracts other individuals that launch large-scale attacks on the trees [17]. Concerning T. destruens, this species is attracted by terpenes emitted from wounds in trees attacked by pioneer bark beetles. This could be the reason why the number of catches of T. destruens in our study was relatively low compared to MPE or H. miklitzi.

Data on more effective pheromone traps will be valuable for forestry practice and open the possibility of additional innovations for reducing the capture of non-target organisms. Likewise, future research on modifications to pheromone traps should consider other species of harmful bark beetles.

5. Conclusions

In this study, the difference in catches of target insects (MPE) and non-target insects (predators) was analyzed to provide advice for future monitoring. The results showed that the average number of O. erosus captured in the standard traps is similar to the ones in the modified traps, while the number of captured beetles of non-target species were up to four times fewer in the modified traps than in the standard traps. The selectivity of the pheromone trap is therefore much more important, because by excluding predators from the population, their positive ecological impact is absent. This study represents the first such research in Croatia, and the results should be valuable to forestry professionals in the field, especially in protected areas where treatment options are limited. Furthermore, it should be kept in mind that the selectivity might change depending on the outbreak phase, and additional research is needed to provide more information about to what extent it is important to modify the pheromone trap.