Measures against pine weevil Hylobius abietis also reduce damage by Hylastes cunicularius and Hylastes brunneus

ABSTRACT Hylastes species are known to cause damage to conifers in plantations in northern Sweden, and in recent years an increase in seedling damage has been observed in southern Sweden. However, there are few studies on Hylastes spp and the damage it can cause, so there is a lack of knowledge regarding pest management. In order to investigate an eventual interaction between damage by Hylastes spp and the more well-known Hylobius abietis (L) we registered damage by these species. Unprotected spruce seedlings were compared with seedlings protected from Hylobius abietis by a mechanical coating or with an insecticide. The effect of mechanical site preparation (MSP) was studied, with half of the seedlings being planted in unprepared soil and the other half after MSP. Both seedling protection and MSP significantly reduced the level of damage caused by Hylastes spp. MSP reduced the proportion of affected and killed seedlings and reduced the level of damage at the root collar. Protecting the seedlings reduced the level of damage, and no difference was found between seedlings treated with an insecticide and those provided with a coating. Similar responses were observed with both containerized and plug plus seedlings. In conclusion, measures against Hylobius abietis seem to also prevent damage by Hylastes spp.


Introduction
Root-colonizing bark beetles (Coleoptera: Scolytidae) of the genus Hylastes are associated with coniferous forests, with 20 species known to occur in Europe.Some species of Hylastes are considered forest pests, causing mortality either by feeding on young coniferous seedlings, or transferring fungal pathogens (Witkovsky and Hansen 1985).Two species, Hylastes brunneus Erich, and Hylastes cunicularius Erich.occur across Sweden.They breed in newly dead wood, such as conifer stumps or logs with ground contact (Rudinsky and Zethner-Mǿller and 1967;Lindelöw et al.1993;Rahman et al. 2018), and feed on young seedlings.They appear to select similar habitats (Eidmann et al. 1977), although Tunset et al. (1993), in an experiment in which flying insects were captured in window traps baited with newly cut pieces of wood of Scots pine (Pinus sylvestris L.) and Norway spruce (Picea abies) (L.H. Karst), demonstrated that H. brunneus clearly preferred Scots pine, while H. cunicularius was exclusively attracted to spruce wood.It should be noted, however, that Palm (1931) occasionally found H. cunicularius in pine.
Adult H. cunicularius and H. brunneus (hereafter referred to as Hylastes) feed on the roots, root collars and stems of young conifers, which can result in high seedling mortality (Lindelöw et al. 1992;Piri et al. 2020).The adults' tunnel into the phloem and, under thin bark, can also tunnel into the xylem (Lindelöw 1992).The first sign of a severe attack is often seedlings losing their green color and becoming increasingly yellowish.Damage by Hylastes can therefore easily be mistaken as wilting caused by drought or other agents, and, because Hylastes feeding is often restricted to roots and root collars, the damage is hard to detect without destructively harvesting of the seedling (Rahman et al. 2018;Piri et al. 2020).In practical forestry, it is therefore likely that the degree of damage caused by Hylastes is underestimated.
Another important pest, the large pine weevil, Hylobius abietis, (L) (hereafter referred to as Hylobius) is a well-known, severe threat to young conifer seedlings.In contrast with Hylastes, extensive research has been carried out on Hylobius to ascertain effective methods for protecting seedlings from damage (Petersson et al. 2004;Nordlander et al. 2011;Lalík et al. 2020).Hylobius breeds in the stump roots of recently dead conifer trees, both Picea and Pinus spp.(Moore et al. 2004), and feeds mainly on the stem bark, damaging or killing the seedlings in the process (Leather et al. 1999;Örlander and Nilsson 1999).It feeds mainly at and above the root collar, and also slightly below the soil surface, but the most obvious sign of its presence is feeding scars higher on the stem.If Hylobius feeds extensively on the phloem and bark, stem girdling and death of the seedling can occur (Scott and King 1974;Eidmann and Klingström 1990).Seedling size, (mainly stem-base diameter), is related to mortality by Hylobius (Örlander and Nilsson 1999;Thorsén et al. 2001).Smaller seedling types, like containerized, are normally less able to tolerate pine weevil damage compared to bareroot or plug plus for example.However, if this also applies to damage of Hylastes is yet to be investigated.
The feeding habit of Hylastes differs from that of Hylobius mainly in that the former eats on the roots and root collar while the later eats on the seedlings stems.The gnawing is often deeper than that of Hylobius, and can penetrate as far as the sapwood.However, there is an area from the root collar to the lower part of the stem where any evidence of damage can be difficult to differentiate between the two species.
Severe Hylastes damage to Norway spruce plantations has been observed in north-central parts of Sweden, often several years after planting (Hellqvist 2010).For example, Lindelöw (1992) reported that death caused by Hylastes cunicularius seemed to occur mainly between the fourth and sixth growing seasons after cutting.The highest risk period for Hylobius damage to seedlings is during the first three growing seasons after harvest (Von Sydow 1997;Örlander and Nilsson 1999;Wallertz et al. 2016;Luoranen et al. 2017;Nordlander et al. 2017).Hylastes damage on the other hand can be diffuse over time and space, such that accumulated seedling mortality over several years may not be detected as clearly as for Hylobius during the first year after cutting (Lindelöw 1992).From studies in UK, Leahy et al. (2007) suggested that Hylastes ater Paykull damage is often incorrectly attributed to Hylobius, leading to overestimates of Hylobius damage and missing the contribution of H. ater to seedling deaths.H. ater is of European origin but is considered to be an introduced species in Chile and New Zeeland, it's main host being Pinus spp.
Treatment with insecticides has long been the most common way to protect seedlings against insect damage in Swedish forestry.Thorsén et al. (2001) noted that the proportion of severe seedling damage caused by both Hylobius and Hylastes decreased when insecticides were used.However, the use of insecticides was at the time of this experiment prohibited for certified forest owners, and the most common type of protection against Hylobius was and still is some kind of mechanical coating.As a result, the proportion of insecticide-treated seedlings was only 3% in 2021 (Nilsson et al. 2019).Several studies have shown that methods of mechanical site preparation (MSP) can also reduce damage caused by Hylobius (Lekander and Söderström 1969;Sutton 1993;Von Sydow 1997;Örlander and Nilsson 1999), especially if the planted seedlings are surrounded by pure mineral soil (Björklund et al. 2003;Petersson and Örlander 2003;Petersson et al. 2005).However, whether mechanical devices can also act as an obstacle for Hylastes has yet to be investigated.As Hylastes mostly feed on underground roots or perhaps close to the surface at the root collar, it may be that methods of protection designed to cope primarily with Hylobius damage will not be effective.Lindelöw (1992) argued that patches of scarified ground would probably only have a limited effect on Hylastes cunicularius because it is active over a longer period of time than other pest species, and the effect of MSP decreases as vegetation colonizes an area (Orlander et al. 1990).However, in southern Sweden, damage caused by Hylastes has also been observed during the first few years after cutting, when an effect of MSP could still be expected.To date, no study has been conducted to clarify a possible relationship between MSP and the level of damage caused by Hylastes.
Earlier studies have shown that fresh wounds on the host material enhance its attraction for both Hylobius (Tilles et al. 1986;Nordlander 1991) and Hylastes (Eidmann et al. 1991).It is also possible that previous damage caused by Hylobius feeding on conifer seedlings could increase the risk of Hylastes attack by releasing odors attractive to Hylastes and this eventual correlation could be something that should be taken into account.
This article presents the results of a multisite trial conducted in the south of Sweden, where untreated Norway spruce seedlings were compared with seedlings treated with either an insecticide (Merit Forest WG) or a mechanical coating used against Hylobius (Cambiguard or Ekowax).Half of the seedlings were planted after MSP, while the other half were planted in undisturbed humus.The aim of the study was to investigate if measures taken today to reduce Hylobius damage also prevent Hylastes damage.
The main hypotheses were: . insecticides provide effective protection against damage caused by Hylastes feeding, .mechanical coatings will not protect seedlings against, root-feeding, Hylastes species, .MSP will reduce attacks by Hylastes, .there is a correlation between Hylobius and Hylastes damage.

Experimental sites
The experiment was established at eight clear-felled areas, located in south-central Sweden, between latitudes 56°36´N and 57°10´N (Table 1).The sites were selected from clearcuts harvested in the last two years and representative of the area.Four sites were situated on fresh clear-cuts (harvested during the winter of 2016-2017), and the remaining four were on 1-year-old clear-cuts (harvested during the winter of 2015-2016).The previous stands had been dominated by either Norway spruce or Scots pine (Table 1).

Experimental design
A split-plot design was used, with three blocks at each site.
Each block was split into two plots consisting of six rows of either MSP or undisturbed ground.Within each MSP treatment, the combination of Norway spruce seedling type (containerized or plug plus) and seedling protection (unprotected, insecticide or mechanical) was randomly assigned to rows.In each row, ten seedlings were planted with an average spacing of two meters.Of these, five seedlings were later harvested and used in the study, i.e. the experimental unit.Thus, in total 2880 seedlings were planted, of which half, 1440, were harvested and analyzed.

Treatments
MSP was performed on each site in the early spring of 2017, and seedlings were planted in the middle of May of the same year.The MSP technique used was disc trenching, apart from at site 2, where mounding was used.Seedlings were either left untreated, or treated with an insecticide or coating, used as mechanical protection against Hylobius.The insecticide used was Merit Forest WG, with imidacloprid as the active ingredient.Before planting, the seedlings were dipped into a solution containing 1.40% by weight of the commercial product, supplied pulverulent, and mixed in water.The proportion of imidacloprid in the commercial product was 70% by weight.The following spring these seedlings were resprayed with the same dose.Cambiguard (Södra Skogsplantor AB) was used as mechanical protection on containerized seedlings, and Ekowax (Norsk wax) on plug plus seedlings.

Seedling stock
The containerized seedlings, were of Bredinge seed orchard origin, and had a height range of 20-40 cm.The plug plus seedlings, were of the same origin, with a height range of 25-50 cm.The containerized seedlings were grown in Hiko v93 trays for one year, whereas the plug plus seedlings had grown in a container for 10-12 weeks and thereafter were transplanted to an outdoor nursery bed and grown as bareroot seedling for another year (Dumroese et al. 2005).There they could grow more extensive and with larger root systems which makes it possible to grow larger diameters.All seedlings used in the trial were provided by Södra Skogsplantor AB, and the applications of Cambiguard and Ekowax were carried out at its nursery.

Measurements
Immediately after planting, the height of each seedling was measured.With the aim of describing the planting areas, the dominant soil type within a radius of 10 cm from each seedling was recorded as one of four classes: 0, undisturbed humus; 1, cultivated humus; 2, a mixture of mineral soil and humus; 3, pure mineral soil.The seedling height and length of the current year's leading shoot were recorded after the first and second growing seasons.
After the final assessment, half of the seedlings planted were harvested in order to estimate the degree of damage caused by Hylastes on stem bases and root systems.The seedlings were selected at random within the treatments, without regard to the vitality of the sampled seedlings, and analyzed in a laboratory environment.The debarked area caused by Hylastes feeding on stem bases and roots (>1 mm) was estimated to be the nearest 0.1 cm 2 .The attack from Hylastes differs from Hylobius (who is normally feeding higher up on the stem) in that it usually eats the bark on the seedling's roots and at the stem base.The gnawing damage often starts at the soil surface and then goes down to the roots.The gnawing is often deeper than that of Hylobius and can cut into the wood, a narrow passage is often visible in the bark.The severity of damage was scored according to the same scale used for Hylastes, Hylobius and other agents; no damage, damaged but alive or damaged and dead.

Statistical analyses
Only harvested seedlings were used in the calculations and statistical analyses.The two seedling types, containerized and plug plus, were analyzed together.Growth was calculated as the height at the time of harvest minus the height at planting.The response variables damaged by Hylastes, root area damaged, damaged by Hylobius, and growth, were analyzed with a mixed model using Proc Mixed if normally distributed or with Proc Glimmix if binomial (SAS 9.4, SAS Institute, Cary, NC, USA): where µ is the overall mean, a i is the random effect of the site (i = 1-8), b ij is the random effect of the block within the site (j = 1-3), γ k is the fixed effect of site preparation (k = 1-2), (bγ) ijk is the random effect of site preparation within site and block, δ l is the fixed effect of seedling type (l = 1-2),ζ is the fixed effect of seedling protection (m = 1-3) and ε ijkl is the experimental error.All the interactions between site preparation treatment, seedling type and seedling protection were also included in the model.Where significant treatment differences were detected, the treatments were separated by overall pair-wise comparisons using differences of least squares means, and Tukey-Kramer to adjust for multiple comparisons.For all tests, an α-value of 0.05 was used to indicate statistical significance.Residuals were checked for normality and constant variance using residual panels in SAS.In order to investigate any possible correlation between Hylobius and Hylastes damage, a correlation analysis was Site index = the highest the trees can reach within the reference age of 100 years, G = "gran" (Norway spruce) and T = "tall" (Scots pine).Volume m3 / ha-1 = standing volume.
carried out in Proc Corr in SAS using the number of damaged seedlings by each species and treatment.All seedlings were used in this analysis irrespective of treatment (n = 283 due to some missing values).Since the data was binomial a complementary test was performed in Proc FREQ to get a Chi 2value and a Phi-coefficient.

Results
The number of seedlings killed by Hylastes during the course of the experiment was relatively low (Table 2).For unprotected seedlings planted without MSP, 10% of containerized seedlings and 18% of plug plus seedlings died.For unprotected seedlings planted after MSP, the corresponding values were 7% for containerized seedlings and 8% for plug plus seedlings.For seedlings treated with either insecticide or a coating, the proportion of dead seedlings varied between 0-4%.Because of the relatively low numbers and many values around 0, the outcome of the statistical analyses was uncertain and no statistical differences were therefore reported.MSP and seedling protection as well as their interaction had a significant effect on seedlings damaged by Hylastes (Table 3).No significant effects of seedling type were found, indicating a similar response of MSP and protection regardless if the seedling planted was a containerized or plug plus seedling.
With MSP and seedling protection, the percentage of seedlings damaged by Hylastes decreased significantly (Figure 1).For unprotected seedlings, the attacks were reduced from almost 65% without MSP to around 40% with MSP.The numbers were in the same range for both containerized and plug plus seedlings.When seedling protection was applied, the positive effect of MSP was reduced.Overall, protected seedlings were less damaged than unprotected.No differences were found between the two types of protection, i.e. insecticide or mechanical coating.
MSP and seedling protection also significantly reduced the debarked area on the root collar of the seedlings (Table 3, Figure 2).Unprotected seedlings of both seedling types planted without site preparation had an average debarked area of 1.5 cm 2 in the root collar zone.With MSP, the area was reduced to around 0.7-0.8cm 2 .The average debarked area was in general lower for protected seedlings.The significant interaction between MSP and protection (Table 3) indicated that overall, seedlings planted without MSP but protected with coating were not statistically different from unprotected seedlings planted after MSP.
The more detailed investigation of soil type surrounding the seedling showed that there was a significant effect on the amount of seedlings damaged by Hylastes depending on the planting area (p < 0.0001) (Table 4).For seedlings planted in pure mineral soil, the proportion of damaged seedlings was significantly reduced compared with seedlings planted in areas with other soil types.There was also a significant interaction effect between soil type and seedling type (p = 0.0139).For containerized seedlings, there was a reduction in attacks when seedlings were planted in a mixture of humus and mineral soil compared with planting in pure humus or undisturbed soil.This was not seen for plug plus seedlings, showing that the overall effects of soil type on seedling attack were less evident in plug plus seedlings compared with containerized seedlings.
Seedling mortality and damage caused by Hylobius were also evaluated.The highest proportion of seedlings killed by Hylobius was found for unprotected containerized seedlings planted without MSP (21%).For plug plus seedlings under the same treatment, the proportion of killed seedlings was 9%.For other treatment combinations, the corresponding values varied between 0 and 6%.
Overall, a large proportion of the seedlings had been damaged, including killed, by Hylobius.Both MSP and seedling protection had a significant effect on seedlings damaged by Hylobius (p < 0.0001 for both) (Table 5).Without protection, 75% of the seedlings were damaged.For seedlings treated with insecticide or provided with a coating, the levels were lower, 60% and 42% respectively.No significant interaction effects were found between MSP, seedling type and protection.Different letters indicate significant differences between all treatment combinations.
We found a positive correlation between seedling damage by Hylastes and Hylobius.Seedlings damaged by Hylobius were more likely to also be damaged by Hylastes, with a correlation coefficient of 0.5726 (p < 0.0001), n = 283.The Chi 2value was 311, p < 0.0001 and the Phi-coefficient 1,05.
Containerized seedlings had an overall higher growth than plug plus seedlings, 12.2 cm versus 7.9 cm (p < 0.0001).They were slightly smaller at planting, thus indicating a higher growth rate.A significant interaction between MSP and seedling type (p = 0.0333) showed that containerized seedlings responded more to MSP: their growth was significantly lower without MSP, 11.4 cm, compared with MSP, 13.0 cm.No significant effect of MSP on growth was found for plug plus seedlings.Protection and seedling type by protection were also significant (p = 0.0001 for both).Insecticidetreated containerized seedlings achieved higher growth, 15.8 cm, than both untreated seedlings, 10.8 cm, and seedlings treated with a coating, 9.9 cm.The difference between seedling protections was greater for containerized seedlings.The growth of plug plus seedlings was slightly lower with a coating, 7.1 cm, compared with 8.5 cm for the insecticide treatment and 8.0 cm for the control.

Seedling protection
To date, damage by Hylastes has not been considered a major threat to conifer plantations, and the focus has been on preventing Hylobius damage.However, with more frequent observations of damage in practical forestry, an important question to address is whether there is a risk that damage by Hylastes will increase as prohibited insecticides are replaced by mechanical protection.
Chemical treatment with insecticides seems to have an effect on Hylastes damage (Thorsén et al. 2001, Eidmann et al. 1991, Lindelöw 1992), thus confirming the first hypothesis.Non-systemic pyrethroids, such as cypermethrin, have been widely used in Europe since the 1980s.Merit Forest WG, the insecticide used in this study, with the active ingredient imidacloprid, is an example of a broad-spectrum, systemic, synthetic neonicotinoid insecticide (Willoughby et al. 2020).The insecticide is supposed to spread throughout the whole seedling, and could therefore protect the roots to some extent.However, in the current study, there was no statistically significant  difference in protective effect between insecticide and coating protection, the proportion of seedlings killed by Hylastes varying between 0 and 4%.
To date, there have been very few studies on the effect of mechanical protection against Hylastes.Eidmann and Von Sydow (1989) observed that Hylastes was able to penetrate stockings used to protect against Hylobius and Merker and Sattler (1952) noticed that Hylastes could burrow under 5 cm deep barriers to be able to access spruce on the other side.However, since the mechanical protections used in this study appeared to be effective against damage caused by both Hylastes and Hylobius the second hypothesis was rejected.For seedlings provided with some kind of protection (insecticide or coating), the proportion of damage caused by Hylastes was more than halved compared with untreated ones.
Mechanical coatings are specifically designed to reduce Hylobius damage and are applied to the stem and root collar.One problem with applying mechanical coatings could be to achieve sufficient coverage of the root collar, an area where both Hylobius and Hylastes are known to feed.

Mechanical site preparation
MSP had a significant effect on Hylastes infestation, reducing the levels of damage, thus confirming our third hypothesis.The results clearly show that damage caused by Hylastes could be reduced when seedlings are planted in mineral soil.Also planting in a mixture of humus and mineral soil decreased damage but only for containerized seedlings, and to a lesser extent compared to planting in mineral soil.Several studies have shown that MSP can reduce damage caused by Hylobius (Sutton 1993;Örlander and Nilsson 1999;Sikström et al. 2020).However, MSP is a broad concept and can include a variety of planting environments during the period of establishment.We know that pure mineral soil surrounding seedlings after MSP reduces the level of Hylobius feeding on seedlings (Björklund et al. 2003;Petersson and Örlander 2003;Petersson et al. 2005), and Kindvall et al. (2000) found that Hylobius moves more directly and faster on pure mineral soil compared with humus, and is thus less likely to stop and feed on seedlings.A thorough literature search did not reveal any scientific analyses of the correlation between MSP and damage by Hylastes: most of the research was focused on Hylastes biology and population studies (Zethner-Mǿller and Rudinsky 1967;Lindelöw 1992;Rahman et al. 2018).However, Hylastes walks on the ground and locates the roots in the soil by scent (Eidmann et al. 1977), thus there is reason to believe that mineral soil probably could have a similar reducing effect on Hylastes damage as for Hylobius.

Hylobius-Hylastes relationship
The fourth hypothesis that there appears to be a correlation between previous Hylobius feeding and recent Hylastes attacks was confirmed.Hylastes are often accompanied by Hylobius (Munbo 1916), and many seedlings that are being damaged by Hylastes also have scars from Hylobius feeding.The reason why mechanical protection can reduce Hylastes damage of the roots and root collars requires clarification.The coatings could be a physical hinderance, but since they mostly are applied from the root collar or the substrate and up onto the stem (Lalík et al. 2020), we do not think that this is the main reason.
The application of the two coatings on the seedlings in this study was part of the regular work at the nursery and did not cover the roots, and possibly only small parts of the root collar, if any, and therefore should not have been a deterrent for root-feeding beetles such as Hylastes.Nordlander (1991) found that seedlings that were deliberately bark damaged were more easily found by Hylobius compared with undamaged ones, and Eidmann et al. (1993) found that damaged seedlings were more attractive to Hylastes than undamaged seedlings.Therefore, a more nuanced explanation is that feeding scars arising from Hylobius damage on seedlings will attract Hylastes as well as more Hylobius through the release of volatiles.Then if mechanical protection reduces that damage, the reduced level of volatiles will attract fewer subsequent insects.Mechanical protection will therefore work indirectly by reducing the attractiveness of the seedlings, rather than presenting a physical barrier.

Figure 1 .
Figure 1.Containerized and plug plus seedlings damaged by Hylastes (%) when planted without (No MSP) or with (MSP) mechanical site preparation.The seedlings were planted without any protection (unprotected) or treated with an insecticide or a coating.Different letters above the bars indicate significant differences between treatments.

Figure 2 .
Figure 2. Average area debarked by Hylastes (cm 2 ) in the root collar zone of containerized and plug plus seedlings planted either without (No MSP) or with (MSP) mechanical site preparation.The seedlings were planted without any protection (unprotected), or treated with an insecticide or a coating.Different letters above the bars indicate significant differences between treatments.

Table 1 .
Description of the experimental sites.

Table 3 .
P-values for the effects of mechanical site preparation (MSP), seedling type and protection, and their interactions, on seedlings damaged by Hylastes (%) and area of root collar damaged by Hylastes (cm 2 ).Bold numbers indicate statistical differences at α-level 0.05.DF = degrees of freedom.

Table 4 .
Effects of soil type in the planting area on attacks by Hylastes for the two seedling types.Different letters indicate significant differences between treatments.

Table 5 .
Proportion of seedlings damaged (including both damaged and killed seedlings) by Hylobius (%) with different treatment combinations.Different letters indicate significant differences between all treatment combinations.