EUROPEAN JOURNAL OF ENTOMOLOGY EUROPEAN JOURNAL OF ENTOMOLOGY

. The true bug Oxycarenus lavaterae has spread northwards through Europe from the Mediterranean at the beginning of the 21st century. Temperature and photoperiod in particular countries are important for the prediction of further spread. The timing and conditions for the termination of winter diapause in Bulgarian and Czech populations of O. lavaterae were determined. The progress in diapause development in the Bulgarian population was monitored by changes in the duration of the pre-oviposition period under long day (18L : 6D) conditions, which decreased gradually from October to June from 98 to 10 days at 20°C and 62 to 9 days at 25°C. There was no change in the photoperiodic response around the winter solstice in December/January, as is recorded for many other temperate insects. Instead, most overwintering females of O. lavaterae transferred to the diapause-promoting short-day photoperiod (12L : 12D) at 25°C did not start to oviposit even though they lived up to one year. The percentage of females that oviposited under short day conditions gradually increased from December to May, but only reached approximately 50% at most. In the Czech Republic, the number of individuals observed resting in overwintering aggregations on tree trunks gradually decreased up to early June. This bug seems remain in diapause until lime trees ﬂ ower and their seeds become available.


Spread of non-native species of Heteroptera
Between 2000 and 2020, several species of bugs, including agricultural pests, spread mostly northwards on several continents including Europe. The brown marmorated stink bug, Halyomorpha halys Stål (Heteroptera: Pentatomidae), became an invasive species in Europe, parts of Asia and America (Musolin et al., 2022). Locally it causes serious economic loss to crops and when overwintering is a public nuisance (Nixon et al., 2022). The southern green stink bug, Nezara viridula (L.) (Heteroptera: Pentatomidae), has spread into many parts of the world. Its reproductive winter diapause is controlled by photoperiod (Musolin, 2012). These species seem to have spread northwards due to climate change (Musolin, 2007;Chartois et al., 2021). In the north, they are subject to diff erent photoperiodic regimes, which are important for diapause induction and sometimes for its termination (Musolin et al., 2019). Another non-na-lime seed bug, O. lavaterae, which might limit its colonisation of geographical areas that diff er in terms of temperature and photoperiod, and the number of generations produced per year.

Laboratory experiments
The bugs used in the experiments were collected near Sofi a, Bulgaria (42°40´48˝N, 23°24´52˝E; 530 m a.s.l.), where the longest daylength is about 17 h at the summer solstice, and about 13 h including civil twilight at the autumn equinox. The photoperiods used in the experiments: short day (SD) = 12L : 12D and long day (LD) = 18L : 6D, resemble the photoperiodic conditions O. lavaterae experience in the fi eld and enable the results to be compared with those of another model insect, P. apterus, with a very similar life history (Hodek, 1971a(Hodek, , 1988.
In the fi rst experiment, the pre-oviposition period and longevity of diapausing bugs collected in the fi eld in November 1999 after transfer to the laboratory were recorded. The fi rst batch of bugs were not exposed to low temperatures, and they were directly placed in the laboratory at 25°C under both SD and LD conditions. The second batch of bugs was kept for 10 weeks at 10°C before transfer to 25°C and SD.
Bugs for the second experiment were collected from trees where they overwintered in early October 2005 and kept outdoors in several 500 ml glass jars covered with nylon net and provided with zigzag folded fi lter paper. Linden (Tilia cordata) seeds with their coats removed were strewn on the bottoms of the jars as a source of food; water in glass vials with cotton plugs were exchanged once a week. Random samples of the bugs were taken each month from mid-October 2005 through to mid-June 2006 and transferred to the laboratory and kept under a LD photoperiod and constant temperatures of 20 ± 2°C and 25 ± 2°C and under a SD photoperiod at 25°C ± 2°C. Light intensity, which can aff ect the duration of the photophase experienced (Saunders, 2022), was about 500 lx.
Pairs of bugs were kept separately in Petri dishes (8.5 × 1.2 cm) with zigzag folded fi lter paper. Each pair was provided with about 10 linden seeds, along with water in a glass vial. Presence of eggs and survival of males and females were checked each day, date of the fi rst oviposition and death of females were recorded, dead males were replaced. Percentage of egg laying females reactivated from diapause and duration of pre-oviposition period were recorded as they are widely used criteria for determining type of termination of diapause and its intensity, respectively (Tauber et al., 1986;Danks, 1987), the number of eggs was counted, and mean lifetime fecundity of the females that oviposit calculated.
The conditions for diapause induction were determined in the third experiment in which the fi rst generation of bugs were the off spring of those collected in March 2015 in the Czech Republic (see below). Larvae were divided between SD and LD conditions and reared to the adult stage. Fifteen pairs of bugs from each photoperiod were placed in Petri dishes and reared as above for 42 days. Egg laying was checked weekly.

Field observation
Aggregations of bugs in diapause on the trunks of linden trees (Tilia spp.) were observed in South Bohemia, Czech Republic (48°58´44˝N, 14°28´00˝E; 390 m a.s.l.) in several winters. In spring 2015, the sizes of the clusters of bugs (horizontal × vertical dimension) were repeatedly measured on three trees at about 2 m above the ground. regulated by environmental factors (mainly photoperiod, temperature and availability of food) (Musolin & Saulich, 2018).
Diapause termination is less studied in Heteroptera. To distinguish between the multiple physiological pathways for diapause completion, Hodek (1977Hodek ( , 2002 defi ned horotelic diapause development (natural slow process occurring in winter when photoperiod is short and temperature low) and tachytelic diapause development (reactivation = an accelerated process following exposure to photoperiod and temperature similar to that occurring in summer). The horotelic diapause development was studied in Pyrrhocoris apterus (L.) (Heteroptera: Pyrrhocoridae). Under a constant 20°C and short-day photoperiod, the intensity of diapause decreases to medium level. Full diapause termination is stimulated by exposure to cold temperatures (Košťál et al., 2008). A very short pre-oviposition period of only several days in Anthocoris nemorum (L.) (Heteroptera: Anthocoridae) transferred in January from outdoors to a laboratory indicates that it is no longer in diapause (Collyer, 1967;Herard & Chen, 1985; see also comments in Saulich & Musolin, 2009). Tachytelic termination of diapause in Riptortus clavatus (Thunberg, 1783) (Heteroptera: Alydidae) has critical photoperiod similar to diapause induction (between 13 and 14 h) (Numata & Hidaka, 1982). Diapause termination in Orius insidiosus (Say, 1832) (Heteroptera: Anthocoridae) is quickly terminated when photoperiod exceeds 12L : 12D, whereas Orius majusculus (Reuter, 1879) remains in diapause after two weeks under a long day 16L : 8D conditions (van den Meiracker, 1994).

Biology of Oxycarenus lavaterae
Since 1990, the lime seed bug O. lavaterae, which was originally distributed in the warm Palaearctic region including the Mediterranean (Pericart, 2001), has spread northwards. Large aggregations of O. lavaterae were reported for the fi rst time in Bulgaria in the winter of 1998/1999 on linden trees (Kalushkov, 2000). In the Czech Republic, it was fi rst recorded in autumn 2004(Kment et al., 2010. This species is considered a forestry pest (Velimirovic et al., 1992) or public nuisance (Wermelinger et al., 2005) in some countries.
Oxycarenus lavaterae develops on various Malvaceae species in the Western Mediterranean, while in Italy, central Europe and the Balkan Peninsula it occurs on linden trees (Tilia spp.) (Kalushkov & Nedvěd, 2010;Kment et al., 2010). In Bulgaria, the bugs gradually leave their hibernacula from March to April, start to feed and reproduce. Adults of the fi rst generation occur in mid-June, the second in late July and third in late September, and then overwinter in diapause (Kalushkov, 2000).
These bugs are easy to rear in the laboratory; the longevity at 20°C is 63 to 113 days and fecundity 250 to 390 eggs, depending on the time spent in outdoor storage prior to termination of diapause. Mating frequency is 36% and positively correlated with fecundity (Nedvěd et al., 2014).
The objective of this study was to determine the timing and prerequisites for termination of winter diapause in the

Statistical analysis
The duration of the pre-oviposition period was analysed using a one-way ANOVA for each combination of temperature and photoperiod. Values for each month of transfer were subsequently compared using the Fisher LSD post hoc comparison for LD conditions and unequal N HSD post hoc comparison for SD conditions. Longevity at SD was analysed using ANOVA and Fisher LSD test, because the numbers of individuals in each experiment were the same each month, whereas fecundity was analysed using ANOVA and unequal N HSD test, because the numbers of females that started to oviposit diff ered between months.
Diff erences in the duration of the pre-oviposition period at the two temperatures and under LD were analysed using a t-test for each month. Because there were nine parallel comparisons (= months), the level of signifi cance of the diff erences between the two temperatures in the same month were adjusted using the Bonferroni procedure (Table 1).

Diapause induction
Adults in the October samples used in the second experiment, which probably moulted from larvae in September, i.e. when the photoperiod was about 13.5L : 10.5D, did not reproduce. The third experiment testing conditions for diapause induction resulted in eight of 15 pairs laying eggs under LD compared to one of 15 under SD. That SD conditions are unsuitable for normal development was also indicated by the low lifetime fecundity (40-70 eggs) of the post-diapause parents in comparison with that recorded for suitable LD conditions (276 ± 150 eggs).

Eff ect of exposure to cold on diapause termination
In the fi rst experiment the 10-week exposure to cold (10°C) resulted in the activation of diapausing females. Two of 10 cold-treated females laid eggs after 20 and 31 days after transfer to SD, 25°C. Only four of the 44 females that were not exposed to cold laid eggs and only after a much longer pre-oviposition period (125-252 days) under SD at 25°C. Mean (± S.D.) longevity of the 40 SD females that did not oviposit was 155 ± 74 days (median 128, range 64-377 days), whereas that of the four that did oviposit lived for 170-293 days. For those kept under LD conditions, 27 of 42 (64%) females did oviposit, with a pre-oviposition period of 52 ± 15 days (median 54, range 25-82 days). Their mean longevity was 113 ± 66 days (median 93, range 57-378 days), whereas that of those that did not oviposit was shorter, 82 ± 28 days (median 71, range 57-144 days).

Termination of diapause under a long photoperiod
In the second, long term experiment, the percentage of females that oviposit was almost constant (about 90%) under LD photoperiod at both temperatures (20 and 25°C), and the duration of pre-oviposition period gradually decreased from October to June (Fig. 1, variability within groups is presented in Table 1). The pre-oviposition period diff ered at the two rearing temperatures from October to January. At the higher temperature (25°C), activation by the tachytelic diapause development was faster (pre-oviposition period was shorter) than at the lower temperature (20°C). Thus, the shortening of the pre-oviposition period   1). From that time onwards, the decrease in the duration of the pre-oviposition period continued at 25°C and followed the same straight line (Fig. 1), while the rate of decrease in the duration of the pre-oviposition period at 20°C was slower than before March and decreased at a similar rate (0.24 ± 0.05 days per day) and reached the same values as recorded at 25°C.

Termination of diapause under a short photoperiod
Most females transferred from outdoors to SD conditions in a laboratory from December to February remained in diapause and did not oviposit despite their long longevity ( Table 2). The percentage of females that oviposit increased substantially in March and continued to increase until May, when it was over 50%. There was an eventual decrease in June, connected with noticeable decrease in longevity.
Longevity of females under SD decreased substantially and almost linearly with the time of transfer. Mean decrease was 0.79 ± 0.04 days per day in outdoor conditions. The rate of decrease in the pre-oviposition time was 0.86 ± 0.04 days per day when kept outdoors. Duration of pre-oviposition period was similar to the mean longevity, which means that those females that started to oviposit did so about the time the others were dying at the highest rate. The longevity of females that oviposit was slightly longer than that of those that did not oviposit in each group, especially in March (Table 2).
Mean lifetime fecundity tended to decrease slightly with time of storage, but the diff erence was statistically insignifi cant due to high individual variability (Table 2). Fe-males activated from December to March oviposit three to four times before they die, while those activated from April to June oviposit once or twice (Table 2; F 6,71 = 5.67, P = 0.00007). The mean number of eggs per batch was 20 (Table 2) and the maximum 31 (female transferred to the laboratory in April).

Outdoor observations
Large overwintering clusters of adult O. lavaterae were observed in South Bohemia on trunks of linden trees in February 2011. These bugs dispersed in May and in October 2011, new overwintering clusters were observed. During the harsh winter of February 2012, these bugs most likely survived by basking in sun during daytime on the south and southwest sides of the trunks. After several cloudy days with severe frosts at night, virtually all bugs died. No bugs were observed during winter 2012/2013. However, in October 2013, there were large aggregations on the same trees. During the winter, they were on the south and southwest sides of the trunks and in March 2014, when temperatures increased to over 20°C, the bugs were on the eastern and in May (maximum temperatures over 30°C) on the northeast side of the trunks. Some of these bugs were observed in the crowns of the trees in May and others were still in the "overwintering" clusters in early June. The trees had fully developed leaves in May and fl owered in early June. In spring 2015, the size of the clusters of bugs observed on these trees gradually decreased, due to bugs moving to less sun-exposed places and into the crowns of the trees for feeding (Table 3). A small percentage of the bugs remained in diapause until early June when the lime trees fl owered.

Diapause strategy
This study revealed that O. lavaterae is sensitive to photoperiod throughout hibernation and in late spring. This strategy is reported in only a few species of Heteroptera, such as, Leptocorisa chinensis Dallas, 1852 (Heteroptera: Alydidae) (Tachibana & Watanabe, 2007). In this bug, the feeding of overwintered adults is a pre-requisite for egg maturation. The host plant (Setaria viridis; Poaceae) is available in mid-June. If they were not sensitive to photoperiod during the winter solstice and activity resumed in Table 2. Activation of Oxycarenus lavaterae females kept under a short-day photoperiod (12L : 12D) at 25°C. Longevity is measured in terms of number of days after transfer and Fecundity the lifetime fecundity of the females that oviposit (egg number, mean ± standard deviation); no signifi cant diff erences. N = 40 for each group. Lower case letters indicate means that diff er among months (ANOVA, post hoc comparisons: Longevity -Fisher LSD, other variables -unequal N HSD). Size of the clusters on particular dates (cm)

Month of transfer from outdoors to laboratory conditions
Feb 22 Mar 18 Apr 06 Apr 26 May 14 Jun 04 Jun 11 Tree 1 14 × 21 11 × 18 12 × 16 10 × 14 7 × 9 4 × 5 0 Tree 2 10 × 8 10 × 8 11 × 7 10 × 7 9 × 4 0 0 Tree 3 6 × 20 6 × 21 3 × 14 3 × 12 3 × 9 1 × 7 0 early spring the bugs would terminate dormancy long before their host plant became available. The loss of photoperiodic response is considered a criterion of diapause completion in most Heteroptera, e.g. the seed eaters Pyrrhocoris apterus (Hodek, 1971a(Hodek, , 1988 and Lygaeus equestris (Solbreck & Sillen-Tullberg, 1981). Pyrrhocoris apterus depends on fi nding the previous years' dry seeds of linden trees (Tilia spp.) in early spring. It seems that O. lavaterae, which also inhabits linden trees, does not rely on this food source, but waits for linden to fl ower in late spring. To synchronise with the blossoms, it remains sensitive to day length until late spring. Thus, there is a lack of the commonly used criterion of the end of diapause. It is tempting to conclude that diapause ends in March, when the slopes of the durations of the pre-oviposition periods recorded at two temperatures match ( Fig.  1). At that time O. lavaterae disperses from overwintering aggregations in Bulgaria. However, the pre-oviposition period would still be rather long (about 25 d) for post-diapause individuals. Field observations in the Czech Republic also indicate that these bugs can still be in diapause in May and June. Thus, it is likely that the diapause strategy of O. lavaterae is similar to the rare type described above for the alydid bug L. chinensis.

Decrease in the pre-oviposition period an indication of diapause development in O. lavaterae
Like most heteropterans inhabiting the temperate regions, adult O. lavaterae diapause overwinter. During late autumn, winter and spring, there is a gradual decrease in the duration of the pre-oviposition period measured after transfer to conditions suitable for development.
At 25°C and LD conditions, the decrease in the preoviposition period continued almost linearly from October until June, while at 20°C and LD conditions it underwent a change in the rate of decrease in March (Fig. 1). Under SD conditions, a large percentage of the females did not resume development in spring. The continuing decrease in the duration of pre-oviposition period in spring may either indicate further progress with diapause development or a shortening of post-diapause development. Because some of the overwintering bugs remained in clusters (presumably in diapause) till early June, while the others dispersed, there could be a mixture of these two processes.
The decrease in the pre-oviposition period during overwintering (from October to March) was much steeper at 20°C, leading in early March to values equal to those achieved at 25°C. At that time, the higher of the two laboratory temperatures did not produce faster activation (in combination with LD conditions) from dormancy. It is likely that the bug was no longer sensitive to the high temperature in March. Cold storage at 10°C shortened, even more, the pre-oviposition period of O. lavaterae.

Photoperiodic response and end of diapause
For the two seed-eating heteropterans, P. apterus (Hodek, 1971a(Hodek, , 1988 and L. equestris (Solbreck & Sillen-Tullberg, 1981) and many other adult insects that are in diapause when hibernating (Tauber et al., 1986;Danks, 1987;Sau-lich & Musolin, 2012), the insensitivity to photoperiod is recorded around the winter solstice. This is considered to be an indication of diapause completion/termination (i.e. end of diapause). At that time most females laid eggs after transfer from the fi eld to higher temperatures in a laboratory, both under LD and diapause promoting SD photoperiods. Outdoors they became quiescent due to low temperature or lack of food.
In the case of L. chinensis the long-lasting response is seen as an adaptation that keeps this bug inactive until its host plant becomes available. It is unlikely that this is the case for O. lavaterae as both its larvae and adults can be reared successfully on the dry seed of Tilia trees, thus, like P. apterus, it would appear to have food continually available outdoors on the ground under trees. It is more likely that fl ower buds or young seed are more suitable for egg maturation in females or development of the young instars of O. lavaterae.
Alternatively, O. lavaterae may retain an adaptation from the area of origin for feeding on herbaceous Malvaceae that are available later in spring (Pericart, 2001) and that linden (Tilia) is an alternative host plant in the absence of its preferred host.

Capacity for range expansion
Two congeneric species of the species studied here are pests: Oxycarenus hyalinipennis is an important pest of cotton worldwide (Slater & Baranowski, 1994), and Oxycarenus laetus is one of the most common and polyphagous pests in Southern Pakistan (Awan & Qureshi, 1996). Oxycarenus lavaterae has a similar dietary range (Kalushkov & Nedvěd, 2010) to O. hyalinipennis and is therefore a potential pest. Cold hardiness enables winter survival of part of the population in central Europe, but the reproductive potential is large enough to enable the populations to become abundant again in summer (Nedvěd & Kalushkov, in prep.). If the mild winters that occurred in central Europe over the last few years (Brázdil et al., 2023) were to occur in more northern countries, the expansion of the species is likely to continue.
The requirement of a long day photoperiod for the resumption of development recorded in this study seems to be adaptive in the Mediterranean, where this bug must wait for food plants to produce seeds. It did not prevent the spread of this species throughout temperate Europe, where linden seed is available year-round. Moreover, the continual sensitivity to photoperiod in spring is a rare phenomenon deserving further study. Even in northern Europe, these bugs can complete development. If their diapause is not terminated by photoperiod then they may be able to increase the number of generations completed per year even without climate warming (Yakamura & Kiritani, 1998).
This species is native to southern Spain, Near East and north Africa (Pericart, 2001). Investigations of the life history of O. lavaterae in its native range are needed. In addition, a search for this bug in northern Europe, helped by citizen science programs (Skuhrovec et al., 2021), should be organized.