General trends of chromosomal evolution in Aphidococca (Insecta, Homoptera, Aphidinea + Coccinea)

Abstract Parallel trends of chromosomal evolution in Aphidococca are discussed, based on the catalogue of chromosomal numbers and genetic systems of scale insects by Gavrilov (2007) and the new catalogue for aphids provided in the present paper. To date chromosome numbers have been reported for 482 species of scale insects and for 1039 species of aphids, thus respectively comprising about 6% and 24% of the total number of species. Such characters as low modal numbers of chromosomes, heterochromatinization of part of chromosomes, production of only two sperm instead of four from each primary spermatocyte, physiological sex determination, "larval" meiosis, wide distribution of parthenogenesis and chromosomal races are considered as a result of homologous parallel changes of the initial genotype of Aphidococca ancestors. From a cytogenetic point of view, these characters separate Aphidococca from all other groups of Paraneoptera insects and in this sense can be considered as additional taxonomic characters. In contrast to available paleontological data the authors doubt that Coccinea with their very diverse (and partly primitive) genetic systems may have originated later then Aphidinea with their very specialised and unified genetic system.


Introduction
The name Aphidococca was recently introduced by Kluge (2010) for the taxon combining two closely related groups of Homoptera insects, aphids and scale insects. According to the paleontological data (see for example, Popov 2002, Shcherbakov 2007) scale insects (Coccinea) could originate from ancient aphids (Aphidinea) or aphid-like ancestors in the Triassic (Fig. 1). The close relationship of both groups is well supported by numerous morphological, anatomical, embryological, cytogenetic, physiological and other characters and, as it seems, is not disputed by any modern taxonomists. In the framework of cladistic taxonomy, aphids and scale insects are considered as sister groups (see for example, Wojciechowski 1992, Gullan andCook 2007 and others) originating from a common ancestor. However, various theoretical generalizations and attempts at analysis of any biological characters of aphids and scale insects are usually done separately for these groups. Below we shall try to analyze aphids and scale insects as a united group which can be exactly contrasted to other related groups of Paraneoptera insects with particular regard to their cytogenetics.
At present, about 5000 species of aphids and 8000 species of scale insects have been recorded from all over the world (Favret and Eades in on-line "Aphid species file" database: http://aphid.speciesfile.org, Ben-Dov et al. in on-line "ScaleNet" database: http://www.sel.barc.usda.gov/scalenet/scalenet.htm). There is no general agreement on the higher classification within both groups; the number of accepted families and their relationships are disputed in the papers of different modern authors. In general, the opposite tendencies (splitting vs. integration) of the families take place in scale insect and aphid modern systematics. Thus, some modern coccidologists (for example, Hodgson 2014) accept till 33 extant families of scale insects in contrast to the 15-19 "large" traditionally accepted families (Danzig 1986, Danzig andGavrilov-Zimin 2014), whereas the last taxonomic catalogue of aphids (Remaudière and Remaudière 1997) places all recent "true aphids" in the single family Aphididae, in contrast to the acceptance of 6-13 true aphid families by some other authors in addition to two families of "not true aphids", Adelgidae and Phylloxeridae (Börner 1952, Shaposhnikov 1964, Heie 1987, Heie and Wegierek 2009a. These opposite tendencies in the systematics of scale insects and aphids reflect, to our mind, the generally higher biological diversity of scale insects, which demonstrate more patterns of morphological, cytogenetic, physiological, and ecological specialization than aphids. Here, for further discussions we shall follow the system and nomenclature of Paraneoptera accepted recently in Gavrilov-Zimin and Danzig (2012) and Danzig and Gavrilov-Zimin (2014): Phylogenetic line Paraneoptera Martynov, 1923 (including 7 orders: Zoraptera, Copeognatha, Parasita, Thysanoptera, Homoptera, Coleorrhyncha, Heteroptera) Cohort Hemiptera Linnaeus, 1758 (= Condylognatha Börner, 1904, non (=Neococcoidea Borchsenius, 1950;= Neococcidea Bodenheimer, 1952) Within the scale insects we recognize 19 extant familes (Table 1). Within the aphids we follow the system of Shaposhnikov (1964Shaposhnikov ( , 1985 with minor changes (taking into account some conclusions of Heie and Wegierek 2009a, b) (see Table 3), and accept 15 recent families.
In the present paper we shall try to summarize data on chromosomal numbers, karyotypes and genetic systems of Aphidococca, mainly with regard to the evolutionary significance of these data, and try to demonstrate some previously neglected parallel tendences in the chromosomal evolution of aphids and scale insects. Two catalogues of chromosomal numbers and genetic systems are used as the basis for this discussion -the catalogue recently published by the first author (Gavrilov 2007) for scale insects, and a catalogue for  Shcherbakov and Popov 2002). Time periods: P 1 , P 2 Early (Lower) and Late (Upper) Permian T 1 , T 2 , T 3 Early, Middle and Late Triassic J 1 , J 2 , J 3 Early, Middle and Late Jurassic K 1 , K 2 Early and Late Cretaceous ₽ 1 Palaeocene ₽ 2 Eocene ₽ 3 Oligocene N 1 Miocene N 2 Pliocene R present time (Holocene). aphids, compiled in the present paper from numerous scattered publications on aphid cytogenetics, the main sources being the tables in Kuznetsova and Shaposhnikov (1973),  and the data from the monographs of Eastop (1994, 2006) as well as from the on-line compilation of these monographs  http://www.aphidsonworldsplants.info). We hope the combined catalogue will be useful for all collegues irrespective of any of our theoretical speculations.

Chromosome numbers
To date chromosome numbers have been reported for 482 species of scale insects belonging to 14 of the 19 known families and for 1039 species of aphids belonging to 14 families (all of those accepted here for the recent aphid fauna) (Tables 1-4), thus respectively comprising about 6% and 24% of the total number of coccid and aphid species. Thus, the greater knowledge of aphid karyotype diversity in contrast to that of scale insects is obvious at the species level as well as for the higher taxa in both these groups. The smallest chromosome number is the same for aphids and for scale insects, 2n=4, and known in species of the tribe Iceryini (Coccinea: Margarodidae), in the genus Apiomorpha Rübsaamen, 1894 (Coccinea: Eriococcidae) (Hughes-Schrader Table 2. Additions to the Gavrilov's (2007) catalogue of chromosome numbers and genetic systems of scale insects. (H -heterochromatinization of one haploid set of chromosomes without details of genetic system; P(o) -obligatory pathenogenesis).
Modal chromosomal numbers of Aphidococca as a whole, 2n=8, 10, 12, 18 are lower (with a small overlap) than in other Homoptera, and most other Paraneoptera groups that have been sufficiently studied to provide reliable data. Thus, comparable modal numbers are 2n=26 for Psyllinea (Maryańska-Nadachowska 2002), 2n=18, 20, 22, 26, 30 for Cicadinea Kirillova 1989, 1991), 2n= 14, 22, 26, 28,  34 for Heteroptera (Ueshima 1979, Kuznetsova et al. 2011, and 2n=18 for Copeognatha (Golub and Nokkala 2004). Aleyrodinea and Thysanoptera are too poorly studied for reliable comparison, but for both these groups there are no recorded chromosome numbers lower than 2n=20. What can be a reason for the comparatively low modal numbers of Aphidococca? It is well known that there is no direct correlation between chromosomal number and complexity of an organism. On the other hand, if we look for the most general character that Aphidococca share with another group with low chromosomal numbers, the Parasita, but not with other Paraneoptera groups, we shall see that the tendency for lower modal numbers within the Paraneoptera correlates with a tendency to larvalization of imaginal structures or neoteny, with reduction of the number of postembryonic stages to three-five in Aphidococca and Parasita, in comparison with the six developmental stages usually found in most Paraneoptera. The karyotype diversity within Aphidococca families can be characterized by a simple index of karyotypic variability (Kv) which is equal to the quantity of different diploid chromosome numbers in the taxon, divided by the number of cytogenetically studied species in this taxon. For example, in the family Diaspididae (Coccinea) six variants (2n= 6, 8, 10, 12, 16, 18) of the chromosomal number are known for 141 studied species. So, for Diaspididae, Kv is equal 6/141=0.04. Of course, Kv, based on the present available data may be changed when more species are studied, but it seems this change will not be very significant. Thus, if we calculate Kv for aphid families based on the old catalogue of Kuznetsova and Shaposhnikov (1973), we obtain values similar to those based on the present catalogue (Table 3), although the number of species studied has meanwhile increased 3-4 times. It is easy to see that Kv is smallest in the largest families of Aphidococca which include numerous poorly identified (recently diverged?) species: Aphididae (0.03), Diaspididae (0.04), Pseudococcidae (0.08). On the contrary, ancient families with a limited number of recent species show comparatively large Kv-s: Adelgidae (0.22), Phylloxeridae (0.45), Margarodidae s.l. (0.21). High Kv-s in some other families, for example, Eriococcidae (0.45) or Thelaxidae (0.66), are connected mainly with enormous variability of chromosomal number not in the family as a whole, but in one of the genera (Apiomorpha and Glyphina Koch, 1856 respectively).
In the higher (above family level) taxonomic groups of Paraneoptera the utility of Kv index is currently limited by the low percentage of studied species and by limited variation of chromosomal number itself, because there are thousands of species in these higher taxa, whereas chromosomal numbers higher than 2n=60 are very rare and higher than 2n=192 are unknown.
Moreover, most of the species in the young and large tribe Aphidini of the family Aphididae have 2n=8, and the same situation applies to the youngest and largest family of scale insects, Diaspididae, the overwhelming majority of species of which also have 2n=8. On the other hand, many genera of Aphidococca demonstrate significant or even extraordinary variation of chromosome number, and, moreover, several diploid numbers can be found in the same nominal species. The most impressive example of such variation is in the scale insect genus Apiomorpha with its 42 diploid numbers, ranging from 2n=4 to 2n≈192 in 47 studied species (Hughes-Schrader 1925, 1930, Cook 2000. A number of aphid genera, for example, Phylloxera Boyer de Fonscolombe, 1834, Glyphina Koch, 1856, Forda von Heyden, 1837, Tetraneura Hartig, 1841, Cinara Curtis, 1835, Lachnus Burmeister, 1835, Trama von Heyden, 1837, Amphorophora Buckton, 1876, Euceraphis Walker, 1870, Chaitophorus Koch, 1854 and others also demonstrate a great variability in diploid number both between and within nominal species (see Table 4).
Polyploidy is a very rare phenomenon in Aphidococca as in other Paraneoptera and probably does not play a significant role in the evolution of the group. For scale insects a polyploid (triploid) karyotype was reported for Physokermes hemicryphus (Dalman, 1826) from the family Coccidae (Nur 1979), but theoretically may be found to occur in some other species of soft scales, felt scales or mealybugs which have chromosome numbers three or four times those of species known to be diploid in the same genera. In aphids polyploid species are not known at all, but several cases of polyploidization in parthenogenetic populations have been reported (see discussion in . On the other hand, females usually have highly polyploid cells (Fig. 2d) in bacteriomes, peculiar organs which include intracellular symbiotic bacteria.
Accessory chromosomal elements have been found in several species of mealybugs (Pseudococcidae) (Nur et al. 1987, Gavrilov 2007 (Fig. 2g), in one species of the Margarodidae (Hughes -Schrader 1942), in two species of soft scales (Coccidae) (Gavrilov 2007) and in some armored scales (Diaspididae) (Brown 1960). Blackman ( , 1990 noted presumed B-chromosomes in numerous aphid species from different families, especially in anholocyclic populations, and these B-chromosomes are probably relicts of multiple X-chromosomes.

Evolution of genetic systems
In contrast to other Paraneoptera, all Aphidococca have spermatocyte and oocyte meiosis in larvae or in neotenic females (which are in fact equivalent to larvae as in scale insects) and demonstrate a multiplicity of very different and unique genetic systems, which are probably based on an original XX-X0 system, considered by Blackman (1995) as ancestral for all Paraneoptera insects (Fig. 3). In species possessing this system, the sex of the progeny is determined during spermatogenesis. Spermatozoa with X-chromosomes produce females and spermatozoa without X-chromosomes produce males. This usual type of XX-X0 spermatogenesis (similar to that of Copeognatha, for example) is known in some primitive scale insects (some Margarodidae, Ortheziidae, genus Puto Signoret, 1875 (Pseudococcidae)) (Hughes , 1944, Brown and Cleveland 1968 with only one peculiar character -spermatocytes fuse to form a quadrinucleate spermatid (Fig. 3). This fusion can be considered as a unique apomorphy of Coccinea. In some genera of Margarodidae, such as Aspidoproctus Newstead, 1901, Protortonia  Townsend, 1898, Llaveia Signoret, 1876, Llaveiela Morrison, 1927, Nautococus Vayssière, 1939 (all from the subfamily Monophlebinae) XX-X0 spermatogenesis is also complicated by the enclosure of meotic prophase I chromosomes in peculiar separate vesicles, instead of a single nuclear membrane. This phenomenon was discovered by F. Schrader and S. Hughes-Schrader and was comprehensively reviewed by Hughes-Schrader (1948). Moreover, it is interesting to note that in Protortonia (Coccinea: Margarodidae), in the second meiotic division, all chromosomes form a chain stretched between the two poles of the cell , which is similar to the well-known example of chain formation in plants of the genus Oenothera Linnaeus, 1753 (Onagraceae) and some other plants and animals (White 1973).
In most cases, species with the XX-X0 system have ony one pair of sex chromosomes in their karyotypes. For example, females of Porphyrophora polonica (Linnaeus, 1758) (Coccinea: Margarodidae) have 2n=12+XX and males have 2n=12+X. However examples of multiple sex chromosomes are also known. Thus, species of the family Adelgidae (Aphidinea) have up to four pairs of X chromosomes, and some species of the families Phylloxeridae, Eriosomatidae, Lachnidae and Drepanosiphidae (Aphidinea) have one-two pairs of sex chromosomes (see Table 4). In scale insects, only Matsucoccus gallicolus Morrison, 1939 (Margarodidae) has a multiple sex chromosome system with 6 pairs of X chromosomes (2n=28+12X in females and 2n=28+6X in males), which probably evolved as a result of fragmentation of an initial pair of X chromosomes (Hughes-Schrader 1948) and it seems the number of sex chromosomes in this species is the highest known in Insecta. Multiple sex chromosomes are also known in Cicadinea and Heteroptera and can be probably considered as a non-unique apomorpic character in different genera of proboscidian insects (Arthroidignatha). This character is not known in studied Copeognatha (Golub and Nokkala 2009), which is considered as an ancestor group for proboscidians. Hales (1989) reported a peculiar fusion of multiple X chromosomes with autosomes (X 1 +A, X 1 , X 2 +A, X 2 ) in somatic cells of Schoutedenia lutea (van der Goot, 1917) (Aphidinea, Greenideidae), that demonstrates a special genetic system unknown in other aphids and in Paraneoptera as a whole, but this phenomenon needs further investigation.
Hovewer, in the majority of studied scale insects and in all studied aphids sex determination is not brought about by stochastic combination of male and female chromosome sets during fertilization, because male and female gametes in most Aphidococca are cytogenetically identical and physiological sex determination takes place. Thus, in all studied Aphidinea, gametogenesis is of a unified type and based on an XX-X0 mechanism, but has unique features which are probably unknown in any other animals with XX-X0. One of the secondary spermatocytes (which includes autosomes only) is smaller in size and degenerates just after anaphase I. The second, larger spermatocyte gives origin to two sperms; both with one X-chromosome (see Manicardi et al. 2015 and our Fig. 2). Thus, aphid males give rise only to female-producing sperm, and sexual females also produce only female-producing oocytes, so that all sexuallyproduced progeny are female. On the other hand, parthenogenetic females can pro-duce embryos which are either XX or X0, using a special cytological mechanism in which the X-chromosome is lost in some of the oocytes (Orlando 1974, Blackman andHales 1986). Thus, sex of progeny is totally dependent on the physiology of the parthenogenetic female, which starts to produce sexuales under certain environmental conditions. This mode of gametogenesis is closely connected with cyclic parthenogenesis and is undoubtedly a unique apomorphy of Aphidinea. In general we suggest that the genetic system of aphids could be termed the Aphidoid system for the uniformity with the names of the genetic systems of scale insects (see below).
The majority of scale insects (almost whole superfamily Coccoidea) and aphids of the tribe Tramini (Lachnidae) demonstrate specific heterochromatinization of part of chromosomes in their diploid set. The species of scale insects with Lecanoid, Comstockioid, and Diaspidoid genetic systems feature obligate heterochromatinization of the paternal set of chromosomes in the males (Fig. 2e-f). Paternal genome heterochromatinization (PGH) is known in some groups of insects (see reviews of White 1973 andNormark 2003), but in each of these groups PGH has specific characters and forms unique genetic systems. The coccid species with systems Lecanoid, Comstockioid, and Diaspidoid can be purely sexual with identical male and female gametes, or demonstrate diploid arrhenotoky and deuterotoky in addition to heterochromatinization of the paternal set of chromosomes. In all these cases the sex of the progeny depends on rather enigmatic physiological processes occurring inside the female, as in the Aphidoid system. In the Lecanoid system, the heterochromatic chromosome set exists during all stages of the male life cycle. During meiosis in the male, the chromosomes do not pair and separate equationally during the first division. During the second division, two metaphase plates are formed, and the heterochromatic and euchromatic chromosomes then segregate to the opposite poles (Hughes-Schrader 1948, Nur 1980. As a result of meiosis, quadrinucleate spermatids are formed, but only the nuclei of maternal origin produce sperm (Fig. 3).
In the Comstockioid system, the heterochromatic set is partly (as separate chromosomes) eliminated during embryogenesis and different cells of the same tissue may differ in chromosome number. According to the number of eliminated chromosomes, several variants of the Comstockiella system are known: CL I -Comstockioid-Lecanoid intermediates, C varH -Comstockioid with one pair of paternal chromosomes, retained in different cysts, C C -complete Comstockioid. The course of spermatogenesis varies among the different taxa, depending on the number of non-eliminated heterochromatic chromosomes. If all these chromosomes are destroyed, the second division is absent (Brown 1965, Nur 1980. In the Diaspidoid system, the heterochromatic set has been completely lost, and adult males are haploid. Hence, spermatogenesis consists of a single equational division (Brown 1965, Nur 1980. In the aphid tribe Tramini (Lachnidae), almost all studied populations reproduce by thelytokous parthenogenesis and sex chromosomes have not been identified . Some of the chromosomes in the diploid set demonstrate heterochromatinization and even aggregation of heterochromatic elements in somatic cells until late prophase , thus resembling the Lecanoid-Comstockioid genetic system in scale insects. However, heterochomatic chromosomes in Tramini can vary significantly in number between populations and do not constitute a haploid set. These heterochromatic elements of Tramini are similar to B-chromosomes, and  suggest that they may be derived from ancestral redundant X chromosomes.
In the tribe Fordini (Aphidinea: Eriosomatidae), germ-line and somatic cells have radically different chromosome numbers . Unfortunately this very interesting phenomenon has not been additionally studied.
Hermaphroditism and Haplo-diploidy are known only in species of the tribe Iceryini (Coccinea: Margarodidae) (Hughes -Schrader 1948-Schrader , 1963. The hermaphrodites are diploid and similar to females in their morphology and mode of life. During embryogenesis the gonads of these insects do not undergo sexual differentiation. Later, in the crawlers, haploid nuclei appear in the gonads and form the central testicular part of a hermaphroditic gland. The haploid nuclei appear as a result of degeneration and elimination of one set of chromosomes. The peripheral ovarian part of the gland is diploid and formed a little later. Fertilization takes place either in the ovarian part or in the cavity of the ovo-testis. Fertilized eggs always develop into female-like hermaphrodites, which usually reproduce by self-fertilization. However, the hermaphrodites may also copulate with accidental haploid males, which sometimes develop from unfertilized eggs (Hughes-Schrader 1948). Haplo-diploidy is known in Iceryini scale insects only and is in fact a result of haploid arrhenotoky as in other insects with haploid males. Fertilized eggs produce diploid females and unfertilized eggs produce haploid males (Hughes-Schrader 1948).
To date, species with heteromorphic sex chromosomes (genetic system XX/XY, neo-XX/XY) have not been found among Aphidococca in contrast to larger groups of Paraneoptera: Cicadinea + Heteroptera, where these systems are very common and to Psyllinea + Copeognatha, where XX/XY (or neo-XX/XY) system is known in several species. On the other hand, in some species of scale insects, such as Newsteadia sp., Praelongorthezia praelonga (Douglas, 1891) (both from Ortheziidae), Lachnodius eucalypti (Maskell, 1892) (Eriococcidae), and Stictococcus sp. (Stictococcidae), both females and males have the same number of chromosomes, but without distinct sex chromosomes or peculiar heterochromatinization of the paternal set (as in the unique coccid systems Lecanoid, Comstockioid, and Diaspidoid). Thus, the Australian felt scale Lachnodius eucalypti, having 2n=18 in both females and males (Brown, 1967, Nur, 1980, is especially noteworthy. In other studied species of the genus Lachnodius Maskell, 1896 and in the family Eriococcidae as a whole, the Comstockioid system has been discovered, but in males of L. eucalypti heterochromatinization of the paternal set is absent. The 2n-2n system probably evolved in scale insects more than once and from different ancestral systems: from the system with heterochromatinization in L. eucalypti and Stictococcus sp. and from the XX-X0 system in Praelongorthezia praelonga (Nur 1980). Meiosis in L. eucalypti comprises one reductional division only (Brown and Chandra 1977), whereas in P. praelonga it comprises two divisions without an inverted meiotic sequence (Brown 1958).

Parthenogenesis
It seems that absolutely all aphid species and many scale insects can produce their progeny by parthenogenesis. In aphids the parthenogenesis can be cyclic (with alternation of thelytoky and deuterotoky -the apomorphic condition for Aphidinea) or anholocyclic (with continuous thelytoky). In scale insects no examples of cyclic parthenogenesis are known and parthenogenesis can be thelytokous, deuterotokous or arrhenotokous. On the other hand, there are probably a few obligatory thelytokous species of scale insects, such as Protopulvinaria pyriformis (Cockerell, 1894) and Eupulvinaria peregrina Borchsenius, 1953 (Gavrilov andTrapeznikova 2008), which never produce males in any population or geographical region. A great many species, often reported as thelytokous (see, for example, Nur 1990 for the review), in reality combine thelytokous reproduction with amphimixis, producing males amphimictically or parthenogenetically (diploid arrhenotoky and deuterotoky), and these males have, as usual for scale insects, paternal genome heterochromatinization. Some species variously manifest thelytokous and sexual lineages in different geographical regions or on different host plants (Nur 1990). Haploid arrenotoky (noted above for Icerini) is connected with haplo-diploidy and can be interpreted as facultative, rather than obligatory parthenogenesis.
Unfortunately it is impossible to say now exactly how many scale insects species are able to reproduce by parthenogenesis, and this ignorance hampers a detailed comparison of scale insects and aphids in this respect.

Conclusion
Finally we can underline the following parallel trends in the evolution of Aphidinea and Coccinea: 1) Low modal numbers of chromosomes. 2) Heterochromatinization of part of chromosomes.
We consider that at least some of these tendencies may be regarded as additional taxonomic characters, which support the erection of Aphidococca as a higher category differing radically from other Homoptera and more widely from all Paraneoptera groups.