Characteristics of parthenogenesis in Cacopsylla ledi (Flor, 1861) were analyzed using cytological and molecular approaches. In all three populations studied from Finland, i.e. Turku, Kustavi and Siikajoki, males were present at a low frequency but were absent from a population from Vorkuta, Russia. In a follow-up study conducted in the Turku population during 2010–2016, the initial frequency of males was ca. 10 % and showed no intraseasonal variation, but then dramatically decreased down to approximately 1–2 % level in seasons 2015–2016. Male meiosis was chiasmate with some traces of chromosomal fragmentation and subsequent fusions. In most females, metaphase in mature eggs included 39 univalent chromosomes which indicated apomictic triploidy. Only a small fraction of females was diploid with 13 chiasmate bivalents. The frequency of diploid females approximately equaled that of males. COI barcode analyses showed that triploid females (N = 57) and diploids (7 females and 5 males) displayed different haplotypes, demonstrating that triploid females reproduced via obligate parthenogenesis. The rarity of diploids, along with the lack of males’ preference towards diploid females, suggested that most likely diploids were produced by rare triploid females which shared the same haplotype with the diploids (not found in the present analysis). Minimum haplotype diversity was detected in the Turku population, but it was much higher in Vorkuta with some indication for the mixed origin of the population. We suggest that functional diploids produced in a parthenogenetic population can give rise either to a new parthenogenetic lineage or even to a new bisexual species.

Cacopsylla ledi , Psylloidea , apomictic parthenogenesis, triploid females, diploid females, rare males, COI haplotypes

The great majority of psyllid species are characterized by bisexual reproduction. However, some members of, at least, in two genera, Cacopsylla Ossiannilsson, 1970 and Trioza Foerster, 1848 include all-female populations and are, therefore, suggested to be parthenogenetic. These are C. ledi Flor, 1961, C. rara (Tuthill, 1944), C. myrtilli W. Wagner, 1947, C. myrtilli ssp. canadensis Hodkinson, 1978, T. pletschi Tuthill, 1944 and T. abdominalis Flor, 1861 (Ossiannilsson 1972, Hodkinson 1976, 1978, Gegechkori 1985). Parthenogenesis of this kind is called thelytoky which is characterized by the presence of females that produce only daughters without fertilization. Thelytoky can be obligatory if only parthenogenetic populations are present throughout the whole distribution range of a particular species. Nevertheless, it can be facultative if both bisexual and parthenogenetic reproduction occurs, with cyclical parthenogenesis in aphids as a well-known example (Normark 2003, Vershinina and Kuznetsova 2016).

However, identifying facultative thelytokous taxa is usually difficult. In many animal taxa the existence of so called rare males was reported for some parthenogenetic lineages (Blackman 1972, Palmer and Norton 1990, Butlin et al. 1998, Martens 1998, Rispe et al. 1999, Simon et al. 1999, Delmotte et al. 2001, Snyder et al. 2006, Domes et al. 2007, Engelstadter et al. 2011, Maccari et al. 2013, Nokkala et al. 2013). These males can be either nonfunctional, i.e. incapable of producing haploid gametes (Taberley 1988, Nokkala et al. 2013) or functional with normal gamete production (reviewed by Maccari et al. 2013). Both kinds of males were found in geographically separate populations in the parthenogenetic psyllid Cacopsylla myrtilli (Nokkala et al. 2013). Females of this species are apomictic triploids (Nokkala et al. 2008). Cytological analysis has also discovered diploid females coexisting with the rare males especially at high altitude. These rare males, although being nonfunctional, copulate randomly with both parthenogenetic and “sexual” diploid females (Nokkala et al. 2015).

The characteristics of parthenogenesis in the Holarctic species C. ledi are poorly known. The species lives on wild rosemary Ledum palustre Linnaeus, 1753, and was previously recorded from Fennoscandia, the Baltic countries, Poland, Germany, Russia, Japan and Alaska (Ossiannilsson 1992). In addition to all-female populations, populations with males are known to exist, although quantitative frequencies of males are unknown. The males share the same chromosome number 2n = 25 (24 + X) with males of the genuine bisexual psyllid species (Kuznetsova et al. 1997).

In the present study, we planned to determine the frequency of males in four populations of C. ledi. In a follow-up study we planned to find out if the frequency of males would undergo any changes either during one reproductive season or between successive seasons. In addition, we planned to analyze chromosomes in mature eggs to determine the type of parthenogenesis and details of reproduction types of females in a particular population. Intrapopulational relationships between the females and males in a population were analyzed by using DNA barcode sequences.

Our findings show that details of parthenogenesis in C. ledi are similar to those found previously in C. myrtilli (Nokkala et al. 2015). Parthenogenetic females of C. ledi are triploid showing apomictic oogenesis in which normal meiosis is replaced by a modified mitosis. In addition to triploid females, there were also diploids showing normal chiasmate meiosis. Moreover, the presence of diploid females among obligate triploid parthenogenetic females was discovered for the first time in populations of C. myrtilli collected at various altitudes on the hill Rindhovda in southern Norway (Nokkala et al. 2015). Diploid females showing conventional meiosis were found at frequencies similar to those of rare males at three different altitudes. The highest percentage of both diploid females and males, 10%, was found in the high altitude (1000 m) population, above the tree line. Comparison of frequencies of diploids at different altitudes showed that environmental factors, like altitude, seem to significantly affect the production of diploids (Nokkala et al. 2015). The findings that rare males in those populations were nonfunctional, producing only diploid sperm (Nokkala et al. 2013), on one hand, and the fact that diploids showed the same COI haplotype as triploid females, on the other hand, collectively suggested that diploids resulted from reversions from triploidy and that the frequency of these reversions seemed to be influenced by environmental factors, e.g. altitude. In addition, reversions resulted in both males and diploid females with a similar frequency (Nokkala et al. 2015).

Haplotype diversity in the Turku population of C. ledi was extremely low, where the two recorded haplotypes differed by a single transition. However, much higher diversity was found in the Vorkuta population showing four different haplotypes among the small number of females studied. The close similarity of haplotypes Vorkuta 1 and Vorkuta 2 to that of Turku 1 suggests common ancestry. In contrast, Vorkuta 3 haplotype differed from all other haplotypes found in the population by three nucleotide changes, one of these being a transversion, indicating different origin of this haplotype.

In C. ledi triploid females and diploid females and males displayed different COI haplotypes. Our results prove that triploid females reproduce via obligate thelytoky. Those females, therefore, do not produce diploid females or males. To account for the occurrence of diploid individuals, it is tempting to speculate on the possibility of independent bisexual reproduction. Potentially, reproduction of this kind is possible, since diploid females display normal chiasmate meiosis and males despite some disturbances show virtually normal meiosis. Two observations, however, make bisexual reproduction unlikely. Firstly, the frequency of males in populations at Turku after the drop-down of diploids is very low and is the same magnitude or clearly below 10% like that of rare males in oribatid mites (Palmer and Norton 1990), in Artemia Leach, 1819 species (Maccari et al. 2013), or in C. myrtilli populations (Nokkala et al 2013). Secondly, those males do not discriminate against triploid females making bisexual reproduction highly unlikely. An alternative explanation implies that rare, still undiscovered triploid females which share the same haplotype with diploids are responsible for the production of diploid females and males via reversion from triploidy to diploidy. This means that the genotypes of triploid females would greatly affect their ability to produce diploids. In turn, this would also mean that C. ledi is an obligate parthenogenetic. However, additional studies are needed to confirm this suggestion.

Although in the short term it is difficult to see any advantage for the production of diploids in the long run they can provide further evolutionary opportunities for a parthenogenetic taxon. It is known that functional rare males can mate with closely related sexual females to give rise to a new parthenogenetic lineage. This type of parthenogenesis, known as contagious parthenogenesis (Simon et al. 2003), was analyzed in detail in Daphnia pulex Leydig, 1860 (Paland et al. 2005) and brine shrimps of the genus Artemia (Maccari et al. 2013). The origin of contagious parthenogenesis was also proven in the laboratory for D. pulex (Innes and Herbert 1988), for the aphid Myzus persicae (Sulzer, 1776) (Blackman 1972) and for Artemia spp. (Maccari et al. 2014). Apparently, new parthenogenetic lineages of contagious origin can also be produced by diploid females found in C. myrtilli (Nokkala et al. 2015) and in C. ledi (the present study), if these females are mated with males of a coexisting sexual species. The origin of contagious parthenogenesis could more easily occur via diploid females than through males as they are functional (or show normal meiosis) even in the case when males are nonfunctional (Nokkala et al. 2015). If both diploid females and functional rare males exist in a population, mating between diploids could occasionally result in diploid offspring, in spite of the fact that males do not discriminate against triploid females. Potentially, this could lead to the origin of a new bisexual species from a parthenogenetic lineage. This can also be the case in oribatid mites. Domes et al. (2007) considered phylogenetic relationships between certain taxa belonging to seven families of the oribatid mite group Desmonomata and concluded that a parthenogenetic lineage could evolve to a new parthenogenetic lineage or even to a new bisexual species. Apparently, the evolutionary potential of parthenogenetic lineages is high, complex and far more diverse than previously understood.

Parthenogenetic females of C. ledi are triploid with apomictic meiosis. Triploid females reproduce via obligate parthenogenesis. We also suggest that rare males and diploid females, if present, are produced by triploid females by reversions from triploidy to diploidy. This probably demonstrates how a parthenogenetic taxon can give rise either to a new parthenogenetic lineage or even to a new bisexual species.

Our sincere thanks go to Dr Gokhman for his valuable suggestion for improving the text. The financial support from the Russian Science Foundation (grant no. 14-14-00541) is gratefully acknowledged.