Thermal sum requirements for development and flight initiation of new‐generation spruce bark beetles based on seasonal change in cuticular colour of trapped beetles

The spruce bark beetle Ips typographus is the most important pest on Norway spruce in Europe. To improve predictions of bark beetle phenology in a warmer climate, correct thermal sums representing the development time of the new generation are needed. A standardized method for classifying adults into five different colours was used for describing the seasonal change in colour of beetles in breeding substrate and in weekly trap catches from regions located in southern, central and northern Sweden in 2015–2020. Repeated sampling of I. typographus from breeding substrate demonstrated that the adults get gradually darker, from light brown when newly moulted to dark brown or black in the following summer. In spring, four to five colours co‐occurred among trapped beetles. Over time, the proportions of darker individuals increased until the two lightest colours were absent except for 2 cases out of 21 trapping locations/years. Thereafter, the individuals of the two lightest colours started to occur again, indicating that they belonged to the new generation. The average thermal sum from start of flight of parental generation in spring until onset of new‐generation flight in summer was higher for southern Sweden [lower developmental threshold (LDT) 5°C = 744 degree‐days (dd); LDT 8.3°C = 467 dd] than for northern Sweden (LDT 5°C = 668 dd; LDT 8.3°C = 418 dd). New‐generation flight occurred in every year and region, but generally constituted only a small proportion of total seasonal flight activity.


INTRODUCTION
The spruce bark beetle Ips typographus (L.) is the most important insect pest on Norway spruce Picea abies (L.) Karst in Europe, killing large volumes of mature trees. Tree mortality caused by I. typographus exceeded 150 million m 3 from 1950 to 2000 in Europe, and damages have increased substantially in recent years (Grégoire et al., 2015;Hlásny et al., 2021;Jönsson et al., 2012;Schelhaas et al., 2003;Seidl et al., 2011). Due to climate change, damage levels are predicted to increase even more in the future (Seidl et al., 2014). One factor that may contribute to increased damages is an increase in number of generations per year (voltinism) in a warmer climate. Models, based on different climate scenarios and on thermal sums required for I. typographus development, have been used for predicting the increase in voltinism in different parts of Europe in the future Jönsson et al., 2012). It is important that the thermal sums used in such models are based on data for the specific areas included in the models.
Lower developmental threshold (LDT) temperatures limiting I. typographus development have been determined by rearing beetles at different constant temperatures in the laboratory (Annila, 1969;Wermelinger & Seifert, 1998). Most commonly, LDTs of 5 and 8.3 C are assumed for modelling I. typographus generation development.
Similarly, accumulated effective thermal sums (accumulated daily mean temperatures above LDT) required for complete development from egg to mature adult were determined either in laboratory (Annila, 1969;Wermelinger & Seifert, 1998) or in field studies (Baier et al., 2007;Berec et al., 2013;Harding & Ravn, 1985;Ogris et al., 2019;Öhrn et al., 2014). In the field studies, thermal sums were calculated from colonization of trap trees to start of emergence of new-generation beetles. Models predicting development rates and generation development, and thus when emergence of the new generation can be expected, are based on known thermal requirements and information about air temperature, topography and solar radiation for a given locality (Baier et al., 2007). A major drawback of studies using trap trees is that, unless beetles are dissected and examined for the presence of mature eggs, it is not known whether emerging newgeneration beetles are in a reproductive state or in reproductive diapause (i.e. just emerging from breeding material for hibernation in the litter). No dissections were conducted in the field studies mentioned above. Another issue is the difficulty to apply the results from local studies including a few trap trees to landscape level. That is because it is generally not known to what extent breeding substrates are utilized (or available) at different altitudes and sun exposures which will influence developmental rates. In addition, there may also be local I. typographus adaptations to differences in regional climatic conditions. An alternative approach to the trap tree studies mentioned above is to investigate weekly trap catches of I. typographus from monitoring programs for the first date of occurrence of new-generation beetles in the summer. Based on the information on start of flight in spring, and temperature data from climate stations, thermal sums required for completion of development and start of flight of new-generation beetles can then be determined. This approach has several advantages: (1) The fact that beetles were attracted by a pheromone bait is a strong indication that they were not in reproductive diapause. (2) The trapped beetles originate from many different localities and types of breeding substrates in the surrounding landscape and thus give a good estimate for when, and at which temperature sum, the flight of the landscape-wide population of new-generation beetles is initiated.
(3) Weekly monitoring of I. typographus is conducted in many regions and thus offers easy access to trapping materials.
In northern Europe, bivoltine populations of I. typographus are not predominant. Thus, from a population dynamic perspective, it is important to determine to what extent new-generation beetles reproduce before the winter. Climate chamber experiments, including Swedish I. typographus populations, demonstrated that at shorter day lengths, an increasing proportion of the new-generation beetles entered reproductive diapause and that some of the beetles from northern regions showed an obligatory diapause (Schebeck et al., 2022;M. Schroeder & Dalin, 2017). However, it is very likely that laboratory results regarding induction of diapause do not correspond with field observations under fluctuating temperatures and natural light conditions. Thus, it is difficult to translate laboratory results to certain proportions of reproductive new-generation beetles under field conditions. Analyses of trapping materials may offer a possibility to determine the proportion of total I. typographus seasonal flight activity that is constituted by newgeneration flight. This proportion can then be used as a proxy for the influence of new-generation beetles on the population dynamics.
Two morphological traits have been used to distinguish between parent and new-generation beetles of I. typographus: density of bristles on the pronotum and elytra (Harding & Ravn, 1985) and body colour (Öhrn et al., 2014). Harding and Ravn (1985) separated beetles emerging from trap logs into parent beetles and new-generation beetles by the lower density of bristles on the pronotum and elytra of the former.
However, the practicality of this method for classifying beetles from trap catches is questionable. Friction between the beetles in the trap, and the possibility that some new-generation adults may already have bored into trees before being trapped, can damage the bristles. Thus, differentiation by colour appears to be a more useful method for classifying the age of beetles collected in pheromone traps.
When beetles develop from pupae to adults, the cuticle is first soft and pale. As the proteins in the cuticle become sclerotized and melanized, the exoskeleton hardens and darkens (Moret & Moreau, 2012;Noh et al., 2016;Thompson et al., 2002). Merker and Wild (1954) reported that yellow/light brown to black beetles were present both in early spring (before flight period) and later in summer in colonized trees.
Knowledge of the long-term change in cuticle colour of I. typographus is limited. Thus, a detailed understanding of colour change over the season is required in order to use colours as indicator of new-generation flight activity in the summer. So far, no study on I. typographus or any other bark beetle species has recorded the transitions in colour from the newly-moulted adults until colonization of new trees for a particular cohort of beetles. In addition, no earlier study has used a standardized method to classify I. typographus based on colour.
The aims of this study were: (1) to develop a standardized method to classify I. typographus individuals based on their cuticular colour; (2) to describe the seasonal colour change of I. typographus in breeding material and from trap catches in Sweden, and based on this information to determine the date of flight initiation of the new generation; (3) to determine the thermal sum required from start of flight of parent beetles in spring until onset of new-generation flight in summer along a climatic gradient; and (4) to estimate the proportion of total seasonal trap catch constituted by new-generation beetles as a proxy for their potential influence on population dynamics.

MATERIALS AND METHODS
In the study, we (1)

Colour classification
The cuticular colour of I. typographus was classified by the same person with a stereo microscope at Â10 magnification. Five different colours were selected from the Natural Colour System ® © (NCS). The NCS provides a straight-forward method to evaluate colours by plain eyesight. The system is based on the six psychological primaries: white, black and the elementary colours green (G), yellow (Y), red (R) and blue (B). Nonelementary colours are described as a relationship between two of the elementary colours. Thus, a red-brown hue will be more reddish than yellowish, for example, 70% red, which would result in the colour Y70R. In addition, lightness is expressed in percentages as the degree of blackness. Chromaticness is shown in percentages as a degree of saturation, where 0 is monochrome. For instance, the individual nuance of a red-brown bark beetle with the hue Y70R can have 60% blackness and 20% chromaticness. This results in the NCS colour notation S 6020-Y70R, where S denotes NCS 1950 standards.
Both blackness and reddishness were used in the classification of beetle development. The following five colours were chosen and referred to as: light brown (S 6020-Y30R), intermediate brown (S 8010-Y30R), red brown (S 8010-Y70R), dark brown (S 8502-R) and black (S 9000-N). When classified, the beetles were placed individually on a neutral-coloured background (grey: S 6500-N) next to a colour palette with the five colours and the closest match was determined for both ventral side of thorax and elytra ( Figure 1).

Seasonal colour change of I. typographus in breeding material
The development, and colour change of new-generation beetles, was recorded in two wind-felled spruces in central Sweden in 2019. One From 20 May 2020 to 6 August 2020, beetles were collected from attacked wind-felled and standing trees to record the colour change of parent beetles over time. In Kronåsen, beetles were collected four times from a wind-felled tree. In Fiby, beetles were collected four times from two wind-felled trees from 1 June 2020 to 24 June 2020. The first attacks on these trees were recorded on F I G U R E 1 Colour palette used in the classification of Ips typographus. From left to right: light brown (S 6020-Y30R), intermediate brown (S 8010-Y30R), red brown (S 8010-Y70R), dark brown (S 8502-R) and black (S 9000-N). The beetles were placed on the palette and the closest match was determined for both ventral side of thorax and elytra under 10 times magnification 20 May 2020. In addition, beetles were collected from one standing tree under attack on 26 May 2020 and from five standing trees, attacked later in the season, from 29 June 2020 to 8 August 2020.
The beetles were stored in the freezer before the colour of each individual was classified with the method described above. From each region and year, we determined the colour of 100 beetles per weekly catch if the trapped numbers allowed. If available, we analysed 50 beetles from 2 different trapping sites (out of the 5 available sites in each region). We chose the localities with the highest catch numbers per week with particular focus on the early and late season when catches were generally lower. In case the samples of the two chosen localities in each region contained less than 100 beetles, we included beetles from the three remaining trapping sites.

Start
In addition, in 2020 we included beetles trapped in seven other localities in Sweden: Örsundsbro, Svenljunga, Misterhult, Nordmaling, Ljungsbro, Karlskrona and Åmål ( Figure 2). In these localities, only one trapping location with a group of three traps was used for monitoring.
Otherwise, the trapping and the analysis was conducted in the same way as for the three regions described above. Thus, we classified 100 beetles per week for each locality if the catch number allowed. Thermal sum required from start of flight in spring until onset of new-generation flight in summer Thermal sums were calculated from start of I. typographus flight in the spring until the samples constituted at least 5% and 10% newgeneration beetles for 5 years of trappings from Tönnersjöheden (2015,2017,2018,2019,2020), and 4 years for Vindeln (2015,2016,2018,2019) and Siljansfors (2015Siljansfors ( , 2018Siljansfors ( , 2019Siljansfors ( , 2020. Vindeln 2020 was not included in the thermal sum analysis due to a 3-week period with zero catches prior to the start of flight of new generation. Based on previous studies, we decided to use the two most commonly applied LDTs of 5 C (Annila, 1969;Harding & Ravn, 1985;Öhrn et al., 2014) and 8.3 C (Baier et al., 2007;Wermelinger & Seifert, 1998). The thermal sum was calculated in degree-days (dd) as the accumulated sum of the daily mean air temperatures subtracted by the LDT. Starting date of flight in spring was defined as the first day of at least 16.5 C maximum air temperature in the first week with an accumulated catch sum of at least 100 beetles (including catches from prior weeks). The traps were set up prior to the expected start F I G U R E 2 Trapping locations in Sweden for Ips typographus individuals that were classified by colour. Main localities represent the three regions in which several years of trapping were conducted, while beetles were trapped in the extra localities in 2020 only. The attacked wind-felled and standing trees from which beetles were colour classified were situated about 30 km from the trapping location Örsundsbro of the spring flight based on the weather forecast and snow conditions in the north (Figures S1-S3). Only in Vindeln 2018, the maximum air temperature exceeded 16.5 C during the 4 days prior to the baiting date (maximum air temperature varied between 19.4 and 23.1 C). Hence, we decided to include the four warm days prior to baiting in the calculation of the thermal sum for this year in Vindeln.
For Tönnersjöheden, Siljansfors and Vindeln, we used temperature data from climate stations at the research stations, which also were the closest climate stations to the trapping localities. Mean daily temperatures were calculated from 1440 readings per day (1 per minute). The distance of the trap sites to the climate stations varied from 3 to 43 km (Table 1). We could not acquire our own temperature measurements T A B L E 1 Coordinates, distances to climate stations, total number of Ips typographus caught and the estimated percentage of this catch constituted by new-generation beetles in the three main regions Tönnersjöheden, Siljansfors and Vindeln, and in the seven additional localities included in 2020

Region
Year

Colour classification
When comparing the colour of ventral side of thorax with the colour of elytra of 24,797 checked beetles from the three main regions, 61.1% had the same colour, 37.1% differed one step, 1.8% two steps and 0.04% three steps on our five-step colour classification. The proportions of beetles among the five colours were somewhat more even for ventral side of thorax than for elytra ( Figure 3). The elytra colour of some of the beetles was not homogenous making classification difficult. Thus, we decided to use data from the thorax in the analyses.

Seasonal colour change of I. typographus in breeding material
The wind-felled trees in Kronåsen and Fiby were colonized by . From 20 May 2020 to 17 June 2020, the beetles were collected from a newly-colonized wind-felled tree in Kronåsen. In Fiby Urskog the beetles were collected from a wind-felled and standing tree currently under attack in 20 May 2020 and 26 May 2020, respectively, a wind-felled tree from 01 June 2020 to 24 June 2020, and a group of five standing trees from 29 June 2020 to 06 August 2020. Each date at the X-axis represents one inspection. The arrow indicates the first date with Ips typographus emergence holes in 2019.
Observe that the X-axis is "broken" for the winter and spring inspections. The number of beetles checked each date is given above the bars  (Table S1).  Table S2.
In 2020, when seven extra localities were included (but Vindeln excluded), there was a clear pattern with somewhat higher thermal sums in the southern locations compared with the two northernmost locations (Nordmaling and Siljansfors) and the northwestern location Åmål (Figure 8a,b and Table S3).
The mean number of days AE SE between the start of flight in spring until ≥5% of the catch consisted of new-generation beetles was 89 AE 6.08 in Tönnersjöheden (mean date = 20 July/21 July), 80 AE 7.12 in Siljansfors (mean date = 31 July/1 August) and 79 AE 5.12 in Vindeln (mean date = 31 July/1 August) ( Table S2).

Proportion of total seasonal trap catch constituted by new-generation beetles
When estimating the proportion of total seasonal flight activity con-     et al., 2000). The underlying mechanism explaining the long-term gradual darkening of I. typographus and its potential adaptive value is not clear. Melanization has been connected to increased immunity and to increased body temperature when exposed to solar radiation in earlier insect studies (Fedorka et al., 2013;Krams et al., 2016).
Average thermal sums required from initiation of I. typographus flight in spring until start of the new-generation flight in summer were 10% higher at the 5% level and 20% higher at the 10% level for the southernmost region (Tönnersjöheden) than for the northernmost region (Vindeln). One factor that could explain the lower thermal sum required in the north compared with in the south is that I. typographus may be more constrained to sun-exposed breeding material, and thus exposed to higher temperatures during development than recorded by climate stations, in the north. Two empirical studies suggest that this could be the case: in southern Sweden wind-felled trees in both sun-exposed and shaded conditions were colonized even though sun-exposed were preferred (Göthlin et al., 2000), while in the north only up-rooted trees in sun-exposed conditions were colonized (Schroeder & Lindelöw, 2003). There is only one previous study about thermal sum requirements for development of a new generation of I. typographus in Sweden F I G U R E 9 Proportion of total Ips typographus seasonal trap catch being made up by new-generation beetles plotted against the date when the new-generation flight started (defined as the first trap emptying when new generation constituted at least 5% of total catch) in Tönnersjöheden (south), Siljansfors (central) and Vindeln (north). The numbers above symbols give the remaining seasonal thermal sum at lower developmental threshold of 8.3 C T A B L E 2 Summary of this and earlier studies on the thermal sums required from flight/colonization by parent beetles in spring to emergence/flight of new-generation Ips typographus. North Zealand, Denmark 573 (sunexposed) 685 (shaded) • Field study • Pheromone-baited stem sections that after colonization were hung in emergence traps, beetles classified as new generation based on colour and bristles, 1980-1981, one sun-exposed and one shaded locality each year • Period: start colonization until first emergence • Air temperature, thermal sum based on hourly mean temperature Harding and Ravn (1985) Southern  ) under 6 different temperatures, teneral new-generation adults provided with fresh bark for maturation feeding, determination of developmental threshold (8.3 C) • Period: (1) development into adults, construction of maternal gallery until development of new-generation adults; (2) maturation feeding, from boring into bark piece until emergence from bark piece (and capable of reproduction) • Air temperature Wermelinger and Seifert (1998) (Continues) (Öhrn et al., 2014). The study was based on the timing of emergence of new generation from stem sections in shaded conditions with known colonization dates ( Table 2). The average thermal sum was 40% lower than our results for Tönnersjöheden in southern Sweden.
In contrast, a study using the same methods as Öhrn et al. (2014) but conducted in Denmark, showed results similar to our study for shaded conditions (Harding & Ravn, 1985, Table 2). We have no explanation for the discrepancy in results between our and the Danish study on one hand and the study by Öhrn et al. (2014) on the other hand. Studies conducted in Austria, Switzerland and Slovenia all reported slightly higher thermals sums, 18% to 23%, compared with our results for Tönnersjöheden in southern Sweden (Table 2). This is expected as day length determines how large proportion of new-generation I. typographus that will be reproductive, respectively, be in reproductive diapause (Doležal & Sehnal, 2007;Schebeck et al., 2022;Schroeder & Dalin, 2017). In addition, the proportions of new-generation flight activity decreased with latitude in agreement with Schroeder and Dalin (2017) demonstrating that the proportion of reproductive new-generation I. typographus decreased with latitude for a given day length.
In Sweden, there are no observations of larvae and pupae of I. typographus surviving winter and in central and southern Europe mortality is high (Dworschak et al., 2014;Faccoli, 2002). Thus, it is very important that the remaining seasonal thermal sum is high enough for the offspring of the new-generation adults to reach adult stage before winter. According to Baier et al. (2007), 60% of the total thermal sum required for complete development of I. typographus (including maturation feeding of adults) is needed for development into adults. In the present study, the average remaining thermal sum for the southernmost region   (Schroeder, 2003).
From a management perspective, our study demonstrates that Thus, it is important to provide forest owners from these regions with predictions about when attacks by the new generation can be expected.
In monitoring programs, it may be tempting to assess beetle colour directly in the field. However, this is not an easy task to conduct with accuracy without good light conditions, magnification and ref-

SUPPORTING INFORMATION
Additional supporting information may be found in the online version of the article at the publisher's website.