Inheritance of diapause regulation in the multicoloured Asian ladybird Harmonia axyridis ( Coleoptera : Coccinellidae )

It is known that females from native populations of the multicoloured Asian ladybird, Harmonia axyridis, have a strong photoperiodic response (enter reproductive diapause under short photoperiods), whereas the proportion of diapausing females from invasive populations is less dependent on day length. The aim of the present study was to determine the mode of inheritance of these differences. The experiments were conducted with two laboratory populations of H. axyridis, the High Diapause (HD) population originated from Irkutsk (Southern Siberia, ca 52.3°N, 104.3°E) and the Low Diapause (LD) population originated from Sochi (North Caucasus, ca 43.6°N, 39.6°E). Reciprocal fi rst generation hybrids and reciprocal backcrosses were investigated. Under a strong diapause-inducing photoperiod (12 h) 100% of the females from the HD population and about 70% of those from the LD population entered diapause. First generation hybrids, as well as their backcrosses with individuals from the HD population, showed almost 100% diapause. Among the backcrosses with the LD population the percentage of diapausing females was widely variable (70–100%) but close to the average between the two populations. The comparison of reciprocal crosses did not reveal any signifi cant difference. We conclude that photoperiodic induction of diapause in the populations of H. axyridis studied is most probably determined by several genes (although one of them evidently plays a leading role with diapause being dominant over non-diapause) and that male and female genotypes are equally important in the determination of female reproductive diapause.

We studied reciprocal fi rst generation (F 1 ) hybrids and reciprocal backcrosses between F 1 and individuals from parental populations (Table 1).The hybridization experiments were performed in 2016-2017.Previously, both populations were reared under standard conditions: temperature of 20°C, day length of 18 h, relative air humidity of 75-85% and fed on the green peach aphid, Myzus persicae (Sulz.).In the experiments, larvae, pupae and adults of the parental individuals were reared under the same conditions.Their progeny (which were tested for the photoperiodic induction of diapause) were kept under a short (12 h) photoperiod, other conditions were the same.This day length was selected in order to estimate the general ability to enter diapause (rather than to compare photoperiodic thresholds) because, based on the results of our earlier study (Reznik et al., 2015a), a 12 h long day is perceived as "short" (diapause-inducing) by individuals from both native and invasive populations.Larvae of the parental and progeny generations were reared individually in plastic Petri dishes (60 × 12 mm) and fed on green peach aphids offered on seedlings of broad bean, Vicia faba L. Fresh food was provided to larvae and adults as needed (daily or every other day).Both parental individuals and their progeny were sexed immediately after emergence from pupae and were then kept in pairs on the same diet in Petri dishes of the same size.Parental individuals were kept alive until the required number of eggs was laid.Progeny adults were dissected 20 days after emergence to check for diapause; this period of time was selected based on the results of our earlier experiments (Reznik & Vaghina, 2011;Reznik et al., 2015a).It is well known that the most important sign of reproductive diapause in insects and, particularly, in ladybirds is inactive poorly developed ovaries (Tauber et al., 1986;Hodek, 2012).Therefore, during dissection of females, the state of the ovaries was described as "poorly developed" (follicles are not visible or their width does not exceed the width of a germarium) or "well developed" (follicles are well visible, non transparent).Then the percentage of diapausing (with poorly developed ovaries) females was calculated for the progeny of each parental pair.In addition, the state of the fat body was also recorded for each female as "poorly developed" (thin lobes interspersed around the internal organs) or "well de-standing of the biology of ladybirds (Hodek, 2012;Sloggett & Honěk, 2012).
Genetic aspects of the regulation of diapause are especially important when considering widely distributed and rapidly spreading invasive species that have to permanently adapt to different climatic conditions (Handley et al., 2011;Sloggett, 2012).One of them is the object of our study, the multicoloured Asian (or harlequin) ladybird, Harmonia axyridis (Pallas) originating from South-Eastern Asia, that has recently colonized numerous countries in America, Africa and Europe (Lombaert et al., 2010, Roy et al., 2016).Native and invasive populations of H. axyridis differ in various biological features (Sloggett, 2012).In particular, the females of Siberian and Korean native populations have a clear photoperiodic response (under short photoperiods, females enter reproductive diapause), whereas the percentage of females entering diapause in European and Caucasian invasive populations is less dependent on day length (Reznik et al., 2015a).These results indicate that there are genetic differences in the photoperiodic responses of individuals in native and invasive populations of H. axyridis the mode of inheritance of which is unknown.
Therefore, in the present work, we studied the inheritance of the diapause-controlling photoperiodic responses in H. axyridis.The results of this study are not only of theoretical but also of practical importance.Several authors note that hybridization between individuals from different native, invasive and mass-reared populations of H. axyridis has had an effect on the worldwide spread of this species (Facon et al., 2008(Facon et al., , 2011;;Lombaert et al., 2010;Handley et al., 2011;Turgeon et al., 2011;Seko et al., 2012).Thus, the mode of inheritance of diapause regulation could be a signifi cant factor in the further spread of the multicoulored Asian ladybird.On the other hand, the ability to avoid diapause and reproduce under short day conditions can be important for the mass rearing of H. axyridis and their use as biocontrol agents in greenhouses.

MATERIAL AND METHODS
The study was conducted with two laboratory populations of H. axyridis.The High Diapause (HD) native population originated from 420 adults collected 31.vii.2015 in the environs of Irkutsk, Southern Siberia, Russia (ca 52.3°N, 104.3°E).The Low Diapause (LD) invasive population originated from 670 adults collected 18.-28.vii.2015 in Sochi, Russian North Caucasus (ca 43.6°N, 39.6°E).It is known that Siberia is in the natural distribution range of H. axyridis, whereas in the North Caucasus the fi rst population of this species was found only several years ago (Belyakova & Reznik, 2013;Ukrainsky & Orlova-Bienkowskaja, 2014;Orlova-Bienkowskaja et al., 2015;Roy et al., 2016).
As noted above, our earlier study indicates that under short photoperiods (day length of 10-14 h) females from the native HD population entered reproductive diapause, whereas the percentage of females from the invasive LD population that entered diapause was lower and almost independent of photoperiodic conditions.The results indicate that photoperiodic responses of these populations are more strongly infl uenced by invasion history than latitudinal differences, because both the Siberian (Irkutsk, 52.3°N) and the Korean (Daegu, 35.9°N) native populations of H. axyridis show a clear photoperiodic response, whereas Table 1.The percentage of diapausing females in progeny of different crosses between individuals from High Diapause (HD) and Low Diapause (LD) populations of Harmonia axyridis (in the designation of crosses, the female is listed fi rst).This experiment was conducted at 20°C and 12L : 12D.The percentage of diapausing females in progeny (median and quartiles) is shown in the right column, values followed by the same letter are not signifi cantly different (p > 0.05 using the Tukey's HSD test of ranked data).veloped" (well-formed globules and interconnected lobes, the internal organs completely or nearly completely hidden).In total, the experiment included 106 parental pairs and 1068 dissected female progeny (Table 1).

Crosses
To reveal within-cross heterogeneity (to compare the numbers of diapausing and non-diapausing females in the progeny of different parental pairs within the same type of cross), the Fisher-Freeman-Halton exact test calculated using StatXact 10 was used.To compare different crosses, the percentages of diapausing females in the progeny of each parental pair were ranked and then Tukey's HSD test was used.To estimate the correlation between development of ovaries and fat body in individual females, Pearson chi-square test was used (these calculations were made using SYSTAT 10.2).In the text, table, and fi gure, untransformed results (percentages) are given.1, all females in the progeny of individuals from HD populations (HD × HD crosses) have poorly developed ovaries (i.e. were in diapause), whereas in the progeny of LD × LD pairs only about 70% of females entered diapause despite the strong diapause-inducing regime used in this experiment.Almost all of the progeny of fi rst generation hybrids, HD × LD and LD × HD (here and after, in designated crosses, the female is listed fi rst) entered diapause (Table 1, Fig. 1C) indicating that diapause is a completely (or almost completely) dominant character.The percentages of diapausing females in the progeny of backcrosses of F 1 hybrids with individuals from the HD population were not signifi cantly different from that in the progeny of HD × HD crosses (Table 1), which also indicates the dominance of diapause.

As indicated in Table
Backcrosses of F 1 hybrids with individuals from the LD population, in contrast, signifi cantly differed from HD × HD crosses (Table 1).Comparison of the distribution graphs showed that about half of the parental pairs of backcrosses with the LD populations (Fig. 1E) produced only diapausing progeny, like the HD × HD crosses (Fig. 1A), whereas in the other half of the pairs, the percentage of diapausing progeny varied similarly to that recorded for the LD × LD crosses (Fig. 1B).In combination, these data argue for monogenic inheritance with dominance of diapause induction.On the other hand, the percentage of diapausing progeny of different pairs of LD × LD crosses was very variable (Fig. 1B).Considering the standard experimental conditions, this variation indicates that some other genes had an effect and the LD population was relatively highly genetically heterogeneous.
The comparison of reciprocal crosses did not reveal any signifi cant difference (Table 1).In other words, male and female genotypes were equally important for the induction of reproductive diapause in females of the multicoloured Asian ladybird.
Finally, it should be noted that 98% of females with poorly developed ovaries (n = 988) and only 31% of those with well developed ovaries (n = 80) had well developed fat bodies (χ 2 = 504.6,p < 0.001).

DISCUSSION
As expected (Hodek, 2012), practically all of the females with poorly developed ovaries had a well developed fat body suggesting that they had entered reproductive diapause (which ensures successful overwintering) rather than merely delaying maturation.Most of the reproductively active females, on the contrary, had poorly developed fat bodies.Some individuals, however, at the time of dissection had both a well developed fat body and well visible follicles in their ovaries indicating, most probably, they had just matured after a short-term delay.Similar results are recorded in earlier studies (Reznik & Vaghina, 2006;Reznik et al., 2015b).
Different insects show various modes of inheritance of diapause regulation.Similar to that recorded in this study, diapause is dominant in the linden bug, Pyrrhocoris apterus L. (Dolezel et al., 2005), Drosophila littoralis Meigen (Lumme et al., 1975) and two Tetranychus species (Ignatowicz & Helle, 1986;Takafuji & Goka, 1999).In contrast, in certain other insects and mites diapause is a recessive character (Sakakibara & Kawakami, 1992;Suwa & Gotoh, 2006;Han & Denlinger, 2009;Chen et al., 2012;French et al., 2014).In the face fl y, Musca autumnalis De Geer (Kim et al., 1995) and fl esh fl y Sarcophaga similis Meade (Goto, 2009) diapause is only partially dominant.Finally, in some of the insects studied, the percentage of diapausing individuals in F 1 hybrids is close to the average between those of their parents (Kuang et al., 2011;Söderlind & Nylin, 2011;Chen et al., 2014;Paolucci et al., 2016).The degree of dominance of diapause can differ markedly even in closely related species.For example, both Tetranychus kanzawai Kishida and T. pueraricola Ehara et Gotoh show simple Mendelian inheritance of diapause, but in the fi rst species it is dominant (Takafuji & Goka, 1999), whereas in the second species it is a recessive character (Suwa & Gotoh, 2006).Moreover, crosses between individuals from different natural populations of Tetranychus urticae Koch show that diapause is a recessive character, whereas the experiments on the artifi cially selected non-diapause strain of the same species reveal the dominance of diapause (Kawakami et al., 2010).It should also be noted that the results of genetic analysis can depend on photothermal regimes and, possibly, on other details of experimental methods.For example, the fi rst studies conducted with the Asian corn borer, Ostrinia furnacalis Guenée indicate that diapause is incompletely dominant over non-diapause (Xia et al., 2012;Huang et al., 2013), but Fu et al. (2015) state that diapause is completely dominant and, fi nally, the experiments conducted by Xiao et al. (2015) show incomplete dominance of either diapause or non-diapause depending on the photoperiod.
We demonstrated that male and female genotypes have equal effects on the induction of reproductive diapause in H. axyridis.Earlier studies indicate that the male plays a more important role in the determination of diapause in its progeny, for example, in several species of Lepido ptera (Ikten et al., 2011;Chen et al., 2012;Xia et al., 2012;Huang et al., 2013;Fu et al., 2015;Xiao et al., 2015;Pruisscher et al., 2017).In some other insects in other orders, however, females have the same (Paolucci et al., 2016) or an even stronger (Kuang et al., 2011;Chen et al., 2014;French et al., 2014) infl uence than males on diapause in their progeny.In some cases, the relative importance of male and female parents depends on the parameter of diapause studied.For example, in the blow fl y, Calliphora vicina R.-D., the percentage of progeny entering diapause is mainly determined by the maternal genotype, whereas the duration of diapause depends on that of both parents (McWatters & Saunders, 1997).In the Asian corn borer Ostrinia furnacalis the duration of diapause is also equally infl uenced by each parent, but the induction of diapause is mainly determined by paternal genes (Xiao et al., 2015).It is also important to remember that the results of reciprocal crosses can be determined not only by the relative importance of maternal and paternal genotypes but also by epigenetic maternal effects (Denlinger, 2002;Saunders et al., 2002;Saunders, 2010;Meuti et al., 2015).
Our study revealed a high level of heterogeneity in the invasive population.High genetic diversity is commonly considered an important prerequisite for a biological invasion by the multicoloured Asian ladybird (Facon et al., 2008(Facon et al., , 2011;;Lombaert et al., 2010;Turgeon et al., 2011;Seko et al., 2012).Moreover, meta-analysis of the data on various invaders suggests that high levels of genotypic diversity in founder groups increase establishment success (Forsman, 2014), although there are exceptions to this rule (Handley et al., 2011).The European invasive population of H. axyridis quite recently invaded Russia and continues to disperse north-eastwards (Belyakova & Reznik, 2013;Ukrainsky & Orlova-Bienkowskaja, 2014;Roy et al., 2016).Our experiments, in particular, show that some families of the Caucasian population of H. axyridis are more prone to diapause and it can be expected that their progeny have more chances of spreading further north to European Russia.In this context, it should be noted that a similar study conducted under the same laboratory conditions with insects collected in the Sochi region 3 years earlier (Reznik et al., 2015a) report a much lower incidence of diapause.This difference suggests natural selection towards a high propensity to enter diapause.As regards further spread, it is known that while the European invasive population of H. axyridis continues to move eastwards, the native Siberian population is gradually dispersing westwards (Orlova-Bienkowskaja et al., 2015).Thus, it is conceivable that the two populations will meet.The outcome of this meeting, in particular, depends on the mode of inheritance of diapause regulation: both fi rst generation hybrids and backcrosses with the native population will show a strong photoperiodic response ensuring timely induction of diapause, which is crucial for survival during severe continental winters.
Summarizing, we conclude that photoperiodic induction of diapause in the populations of H. axyridis studied is most probably determined by several genes (although one of them evidently plays a leading role with diapause being dominant over non-diapause) and that male and female genotypes are equally important in the determination of female reproductive diapause.Under natural conditions, this mode of inheritance would benefi t the survival of hybrids between the native and invasive populations.As regards the mass rearing of non-diapausing strains, any occasional contamination could cause the induction of diapause in a signifi cant fraction of the progeny.

Fig. 1 .
Fig. 1.Percentages of parental pairs with different percentages of diapausing females in the progeny of different types of crosses between individuals from Low Diapause (LD) and High Diapause (HD) populations of Harmonia axyridis (in the designation of crosses, the female is listed fi rst).Statistical signifi cance of heterogeneity of parental pairs within the same type of crosses was calculated using the Fisher-Freeman-Halton exact test.