Meet Me Halfway: Will Photoperiodic Responses of Interpopulation Hybrids of the Brown Marmorated Stink Bug Halyomorpha halys (Hemiptera: Heteroptera: Pentatomidae) Promote or Constrain Subsequent Invasions?

: Biological invasions often result from multiple invasion events. In the case of several subsequent invasive episodes, ‘newcomers’ are usually poorly adapted to local environmental conditions in contrast to the representatives of the already-established invasive population of the same species. Therefore, the mode of inheritance of life-history features determining the survival and performance of interpopulation hybrids is an important component of invasiveness. We investigated the mode of inheritance of the photoperiodic response in the brown marmorated stink bug, Halyomorpha halys , by crossing between the native population of South Korea and the invasive population of the Black Sea coast of Russia. The aim of this study was to predict the overwintering potential of the progeny from possible crosses between representatives of the native and invasive populations. The pre-adult development time and the incidence of winter adult diapause in the progeny of the interpopulation crosses were close to the average of the values recorded in ‘pure’ (unmixed) crosses. Female and male genotypes were equally important in the determination of these characters. Such a mode of inheritance is most likely determined by a polygenic control and would only partially promote subsequent invasion events from the native South Asian range of H. halys into Europe.


Introduction
Biological invasions are complex multi-step processes, starting with an intentional or accidental introduction; each invader has to somehow leave its native range, reach a new territory, and survive there. However, even the survival of the first unfavorable season (usually winter) does not complete the process of invasion. Very often, invasive episodes can repeat again and again, and the invasive individuals might arrive from the same or different regions. Thus, many (if not most) cases of invasions result from not a single invasion event, but multiple invasion events [1][2][3]. This process is generally referred to as multiple invasions. Although 'newcomers' are usually much less numerous than the The question about ability of H. halys hybrids to overwinter successfully has never been addressed. The aim of the present study is to understand the mode of inheritance of H. halys photoperiodic response, and thereby estimate the overwintering potential of the progeny from the possible crosses between native Asian and invasive Caucasian populations of this species. The earlier studies on H. halys demonstrated that day length influence not only the induction of adult diapause, but also the duration of pre-adult development [35,36]. Therefore, we investigated the mode of inheritance of both parameters.

Insects
The study was conducted with two laboratory populations of H. halys. The native population (hereafter referred to as the 'Andong population') originated from 30  It should be emphasized that South Korea is a part of the natural geographic range of H. halys, whereas in Sochi, this invader was recorded for the first time quite recently (in 2014) [26,27,[30][31][32][33][34][35][36].
Before the beginning of the experiment, both populations were reared in a laboratory at a temperature of 25-28 • C and a photoperiod of L:D 16:8 h (hereafter, the duration of the light and dark periods is given in hours). Under these optimal long-day conditions, diapause was never observed in H. halys. In the culture, insects were kept in ventilated transparent plastic containers (22 cm × 13 cm × 12 cm) and fed with peanuts, sunflower seeds, and broad-bean seedlings. In addition, nymphs were provided with wet cotton balls as a water source. During the experiments, all nymphs and adults were fed the same diet.

Parental Generation
To start the experiment, egg masses laid by females from the two laboratory populations were collected and subsequently kept at 25 • C and L:D 16:8 h. The hatched nymphs were reared under the same conditions. Upon emergence, the adults were randomly selected and paired up, according to the design of the experiment: four cross types, i.e., four combinations of males and females, from the two populations were tested. Hereafter, the cross types are indicated by the initials of the population origin of the parental female and male, respectively: two 'pure' crosses (within-population, unmixed, or pure lines) (AA and SS) and two hybrid (reciprocal, or mixed) crosses (AS and SA), where A stands for Andong and S stands for Sochi. When necessary, the emerged adults were kept individually isolated for up to 7 days waiting for the planned mate. These 'parental pairs' were kept in ventilated transparent plastic containers (height 12, diameter 10-12 cm) under the same optimal long-day conditions (25 • C and L:D 16:8 h) to ensure fast maturation and oviposition. Eggs laid by a female of each pair were collected 3 times a week, i.e., every 2 or 3 days.

Progeny Generation
The progeny that emerged from each egg mass (first 10-30 individuals) was reared separately. Nymphs were kept in the same size of ventilated transparent plastic containers under moderately diapause-inducing conditions: 25 • C and a near-threshold photoperiod of L:D 15:9 h. The progeny pre-adult development time (from the deposition of an egg mass to the emergence of adults) was recorded and the data for males and females were pooled, as our earlier study [35] demonstrated that there is no difference in the time of development between two sexes. The emerged progeny adults were collected every 2 or 3 days and kept in groups of up to 15 individuals per container under the same laboratory conditions. Our earlier studies demonstrated that, under these photoperiodic conditions, nearly all H. halys Diversity 2022, 14, 878 4 of 12 males and females from the Andong population matured, whereas most individuals from the Sochi population entered diapause [36].
Twenty days after emergence, all progeny adults were frozen and dissected. This age was also chosen based on our earlier studies [35,36] and constituted approximately 1.2 times the mean period from female emergence to the deposition of its' first egg mass under the optimal long-day conditions [39,40]. No less than 5 male and 5 female (usually more than 10 male and 10 female) progeny individuals of each parental pair were dissected. Data on the parental pairs with less than 5 dissected progeny individuals of each sex were discarded. Upon dissection, the reproductive state of males and females was evaluated by the criteria commonly used for H. halys and other pentatomids [15,35,36,[39][40][41][42][43][44] with a three-point scale used in our previous study [34]. Based on the state of their ovaries, females were separated into the following categories: (1) diapausing (no visible signs of development of ovarioles); (2) maturing or intermediate (initial stages of development of ovarioles); and (3) mature (ovaries fully developed, mature eggs or vitellogenic oocytes in ovarioles). Based on the state of their ectodermal sacs of accessory glands, males were separated into the following categories: (1) diapausing (ectodermal sacs were not filled with secretory fluids and shrunk); (2) maturing or intermediate (ectodermal sacs were only partly filled with secretory fluids); and (3) mature (ectodermal sacs were fully filled with secretory fluids).

Statistical Analysis
In total, the pre-adult development time of 2549 individuals was recorded (the progeny of 65 parental pairs, 13-20 pairs per each of the four cross types, 13-76 progeny individuals per parental pair), and then 1045 male and 1131 female progeny individuals were dissected. The data on the pre-adult development time were subjected to a two-stage statistical treatment: first, the mean time of the pre-adult development was calculated separately for each parental pair of each cross type; and second, these mean data were averaged again over all parental pairs of each cross type and used for further analysis. The proportions of the diapausing and mature progeny individuals were also calculated separately for each parental pair of each cross type. Thus, in order to avoid pseudoreplications, heterogeneity (the difference among parental pairs of the same cross type) was evaluated, using each progeny individual as an independent data point, whereas the differences between the cross types were estimated based on the means for parental pairs. For the final analysis, ANOVA, followed by the Tukey's HSD test, was used; nonparametric data (incidence of diapause) were ranked before this treatment. The proportions of diapausing and mature progeny were compared by chi-square test. All calculations were performed with SYSTAT 10.2. (Systat Software Inc., Richmond, CA, USA).

Pre-Adult Development Time
The mean pre-adult development time of individuals from the Sochi population (SS) was 9.0 days (i.e., more than 20%) longer than that of individuals from the Andong population (AA); the interpopulation hybrids had intermediate values ( Figure 1). The difference between the AS and SA hybrids was definitely not significant (p = 0.619), whereas the difference between the hybrids and within-population crosses was either statistically significant ( Figure 1) or only marginally not significant (p = 0.075 for the difference between the SA and AA crosses, and p = 0.126 for the difference between the AS and SS crosses).
The variation in the mean progeny pre-adult development time among parental pairs of the same cross type (heterogeneity) estimated by Tukey's HSD test was highly statistically significant (p < 0.001) for all cross types. When the means for the parental pairs were compared, the coefficients of variation calculated for pure crosses AA and SS, and for hybrids AS and SA, constituted 0.135, 0.082, 0.090, and 0.89, correspondingly (not shown), i.e., the variability of hybrids was not higher than that of pure crosses (see also in Figure 1). The variation in the mean progeny pre-adult development time among parental pairs of the same cross type (heterogeneity) estimated by Tukey's HSD test was highly statistically significant (p < 0.001) for all cross types. When the means for the parental pairs were compared, the coefficients of variation calculated for pure crosses AA and SS, and for hybrids AS and SA, constituted 0.135, 0.082, 0.090, and 0.89, correspondingly (not shown), i.e., the variability of hybrids was not higher than that of pure crosses (see also in Figure  1).

Diapause
A two-way ANOVA of the ranked data on the incidence of diapause in male and female progeny of all parental pairs (n = 130, that is 65 parental pairs × 2 sexes) showed that the proportion of diapausing progeny strongly depended on the cross type (F = 142.3, df = 3, p < 0.001): the difference between sexes (F = 18.1, df = 1, p < 0.001), as well as the interactions of these two factors (F = 7.5, df = 3, p < 0.001), were also significant. Tukey's HSD test, conducted separately for males and females, demonstrated that not only interpopulation difference, but also differences between the progeny of pure crosses and hybrids were highly significant (p < 0.001), whereas the difference between the two hybrid cross types was not significant (p = 0.224) for females, and only marginally significant (p = 0.042) for males.

Diapause
A two-way ANOVA of the ranked data on the incidence of diapause in male and female progeny of all parental pairs (n = 130, that is 65 parental pairs × 2 sexes) showed that the proportion of diapausing progeny strongly depended on the cross type (F = 142.3, df = 3, p < 0.001): the difference between sexes (F = 18.1, df = 1, p < 0.001), as well as the interactions of these two factors (F = 7.5, df = 3, p < 0.001), were also significant. Tukey's HSD test, conducted separately for males and females, demonstrated that not only interpopulation difference, but also differences between the progeny of pure crosses and hybrids were highly significant (p < 0.001), whereas the difference between the two hybrid cross types was not significant (p = 0.224) for females, and only marginally significant (p = 0.042) for males.
Indeed, most of the AA parental pairs produced only non-diapausing progeny, although up to 20% of the progeny of some pairs of this type entered diapause (Figure 2). Most of the parental pairs of the SS crosses, on the contrary, produced only diapausing progeny, although the progeny of some atypical pairs of this type included up to 40% of non-diapausing individuals. The progeny hybrids (AS and SA) demonstrated intermediate values compared to both pure crosses (AA and SS). Their variability was much higher, Diversity 2022, 14, 878 6 of 12 and the difference between the AS and SA hybrids, as noted above, was not statistically significant. It is noteworthy that in all graphs (Figure 2), the range of interpair variation in female progeny was wider than that in male progeny; this difference agrees well with the above shown differences in the significance of heterogeneity. Thus, although the mean diapause incidence was about 40% in the AS and SA hybrids, more than the half of the progeny of some parental pairs entered diapause, whereas in the progeny of some other pairs, diapause incidence was less than 20%. df = 12, p = 0.309).
Indeed, most of the AA parental pairs produced only non-diapausing progeny, although up to 20% of the progeny of some pairs of this type entered diapause (Figure 2). Most of the parental pairs of the SS crosses, on the contrary, produced only diapausing progeny, although the progeny of some atypical pairs of this type included up to 40% of non-diapausing individuals. The progeny hybrids (AS and SA) demonstrated intermediate values compared to both pure crosses (AA and SS). Their variability was much higher, and the difference between the AS and SA hybrids, as noted above, was not statistically significant. It is noteworthy that in all graphs (Figure 2), the range of interpair variation in female progeny was wider than that in male progeny; this difference agrees well with the above shown differences in the significance of heterogeneity. Thus, although the mean diapause incidence was about 40% in the AS and SA hybrids, more than the half of the progeny of some parental pairs entered diapause, whereas in the progeny of some other pairs, diapause incidence was less than 20%.

Adult Maturation
The non-diapausing fraction of the progeny included two categories of adults: mature and intermediate (maturing) (see the Section 2). The ratio between adults from these two categories also differed among cross types. A two-way ANOVA demonstrated that the proportion of mature males and females among non-diapause individuals (n = 110) depended on the cross type (F = 18.5, df = 3, p < 0.001) and sex (F = 28.7, df = 1, p < 0.001), whereas the interactions of these two factors were not significant (F = 1.0, df = 3, p = 0.405). A one-way ANOVA, followed by Tukey's HSD test, demonstrated that the proportions of mature males and females in both hybrid crosses (SA and AS) were highly significantly Diversity 2022, 14, 878 7 of 12 (p < 0.005), different from those in the AA pure crosses, and not significantly different from each other. Indeed, on the day of dissection, most of the non-diapausing males and females in the progeny of AA pairs were already mature, whereas about one-third of the progeny of the mixed pairs were still maturing (more precisely, they were in the intermediate state; Figure 3). The data on SS pure crosses are somewhat unclear because, as noted above, most of the parental pairs of this type produced only diapausing progeny. ture and intermediate (maturing) (see the Materials and Methods section). The ratio be-tween adults from these two categories also differed among cross types. A two-way ANOVA demonstrated that the proportion of mature males and females among non-diapause individuals (n = 110) depended on the cross type (F = 18.5, df = 3, p < 0.001) and sex (F = 28.7, df = 1, p < 0.001), whereas the interactions of these two factors were not significant (F = 1.0, df = 3, p = 0.405). A one-way ANOVA, followed by Tukey's HSD test, demonstrated that the proportions of mature males and females in both hybrid crosses (SA and AS) were highly significantly (p < 0.005), different from those in the AA pure crosses, and not significantly different from each other. Indeed, on the day of dissection, most of the non-diapausing males and females in the progeny of AA pairs were already mature, whereas about one-third of the progeny of the mixed pairs were still maturing (more precisely, they were in the intermediate state; Figure 3). The data on SS pure crosses are somewhat unclear because, as noted above, most of the parental pairs of this type produced only diapausing progeny.  The chi-square tests demonstrated that the heterogeneity of the percentage of mature individuals among the non-diapausing progeny was not statistically significant for both AS (n = 127, χ 2 = 17.9, df = 16, p = 0.332 and n = 187, χ 2 = 23.9, df = 16, p = 0.093 for male and female progeny, correspondingly) and SA (n = 100, χ 2 = 19.3, df = 19, p = 0.437 and n = 224, χ 2 = 15.0, df = 19, p = 0.720 for male and female progeny, correspondingly) hybrids, as well as for AA (n = 197, χ 2 = 9.8, df = 12, p = 0.634 and n = 218, χ 2 = 9.0, df = 12, p = 0.700 for male and female progeny, correspondingly) and for SS pure crosses (n = 6, χ 2 = 6.0, df = 3, p = 0.112 and n = 24, χ 2 = 0.9, df = 5, p = 0.966 for male and female progeny, correspondingly). However, the range of interpair variation (at least for the mixed pairs) was rather large (Figure 3). It is possible that the insignificance of heterogeneity was caused by low sample sizes (as only non-diapausing individuals were considered).

Discussion
Our experiments clearly demonstrate the following: (1) both pre-adult development time and incidence of adult diapause in interpopulation hybrids (AS and SA) of the brown marmorated stink bug are close to the average corresponding values of the two parental populations (AA and SS); and (2) in both parameters, the differences between the two reciprocal interpopulation crosses (AS and SA) are not significant. It should be emphasized that both the induction of adult diapause and the pre-adult development time are largely determined by the threshold of the photoperiodic response of the adult diapause induction: the proportion of diapausing individuals increases sharply when the day length decreases below the threshold level, whereas the pre-adult development time is longest when nymphs are reared under the pre-threshold day length [35,36]. Thus, the mode of inheritance of both studied parameters is determined by the mode of inheritance of the photoperiodic response.
We did not find any significant difference between reciprocal interpopulation hybrids (AS and SA). This means that male and female genotypes have an equal impact on both the duration of the pre-adult development and induction of adult diapause in H. halys. In certain other insect species (for example, in several lepidopterans), males play a leading role in the determination of diapause in progeny [53,[55][56][57][58][59][60], whereas in other cases, females have the same [61], or an even stronger [47][48][49], influence on progeny diapause than the males. Moreover, the relative importance of males and females might depend on the studied parameter of diapause: in the blow fly, Calliphora vicina R.-D., diapause induction is mainly determined by female genotype, whereas the duration of diapause depends on both parents [65]. In the Asian corn borer, O. furnacalis, diapause duration is also equally influenced by each parent, but the induction of diapause is mainly determined by male genes [56].
Regarding the main aim of our study, we conclude that whatever the mechanism of the observed mode of inheritance, it can only partially promote subsequent multiple invasions from the native range of the brown marmorated stink bug. If newcomers from Asia cross with the local conspecifics, the probability of the timely induction of winter adult diapause in the progeny will be half of that in the progeny of the local pairs from the established Diversity 2022, 14, 878 9 of 12 invasive population. However, it will be still much higher than could be expected in the case of an initial invasion, i.e., in the progeny of parents from the native (South Asian) populations of H. halys. At the same time, further potential newcomers from Europe are expected to be more pre-adapted to timely enter diapause than their conspecifics from Asia.