CLADOENDESIS OF EPHEMEROPTERA

A  B  C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z 

Autapomorphies and characters of unclear phylogenetic status.

Larval characters:

(1) Labial palp has 2^nd segment without large inner-apical projection, 3^rd segment widened and rounded. Muscle in 2^nd segment is usually retained (Kluge & Novikova 2011: Fig.39, 128, 152); in group sibiricum lost (ibid., Fig.101).

(2) Larva is adapted for inhabitance on stones in rapid current; it lost the primary siphlonuroid swimming specialization (which is retained in most other Baetofemorata) and has general appearance of "Acentrella-type" (Kluge & Novikova 2011: Fig.10–17, 18): body is short, with ventral side flattened; thorax is enlarged, with leg bases widely separated; abdomen is diminished. Head, retaining plesiomorphic hypognathous position (as in other Tetramerotarsata), at the same time is depressed dorso-ventrally, so mouthparts are shortened, and frons is somewhat projected above clypeus (Kluge & Novikova 2011: Fig.12, 16, 123–124). Legs are widely separated, long; femur has a row of long dense setae along outer margin; tibia has a similar setal row [see (3)]; tarsi are twisted so that claws are directed perpendicular to the general leg flatness. Each abdominal tergum is shortened thanks to reduction of its anterior part, so that medioanterior myo-sigilla are adjacent to anterior margin of tergum and do not form separate maculae (Kluge & Novikova 2011: Fig.110, 134) (on sterna medioanterior myo-sigilla retain form of oblique stripes separated from anterior margin — ibid., Fig.23, 99, 113). Cerci are long, their primary swimming setae are more or less reduced: either diminished , or vestigial, rarely completely lost (e.g., in fenestratum [Pseudocloeon]). Paracercus is reduced [see (5)]. Because of these modifications, larva lost ability for normal swimming: its legs, being widely separated, can not be stretched backward and pressed to the body; its abdomen, being too short, can not make undulate movements; its caudalii, consisting of long and nearly nude cerci, cannot serve as a swimming flipper. In natural condition larva does not swim, but crawls on stone surface. If put it into a glass with stagnant water and smooth walls, larva tries to swim: making non-effective dorso-ventral movements by abdomen, it swims up, than stops an slowly falls down passively, keeping its legs directed laterally, abdomen bent dorsally and cerci widely diverging.

In various species of Acentrella/fg1 these characters are expressed in various degrees. Degree of ventral flattening and shortening of abdomen can be characterized by ratio between width of 2^nd abdominal tergum (distance between tergalii bases) to its length, measured on dissected exuviae spread on slide: if larva is flattened ventrally and convex dorsally, its terga, being spread on slide, appear to be much wider than sterna (Kluge & Novikova 2011: Fig.10, 14). Larvae of Platybaetis lost the "Acentrella-type" habitus and got a unique specialization [see below, Platybaetis (2), (3)].

Besides Acentella/fg1, the same habitus of "Acentrella-type" independently appeared in some other, non-related taxa — e.g., Baetiella, Baetodes and Deceptiviosa (belonging to Baetungulata-non-Baetofemorata), Afroptiloides (belonging to Protopatellata) (see Index of characters [1a]). So this character, being an autapomorphy of Acentrella/fg1, is not enough to prove its holophyly.

(3) Larval tibia on outer-dorsal side bears a regular longitudinal row of setae; these setae can be long (Kluge & Novikova 2011: Fig.104, 138) or short (ibid.: Fig.67). This regular setal row on outer-dorsal side of tibia resembles the regular setal row on outer-dorsal side of femur, which occurs, besides Acentrella/fg1, in many other Baetofemorata. In Acentrella/fg1 setae of these femoral and tibial rows are often feathered, i.e. bear on their sides numerous very fine processes; in some species these processes are longer than half of distance between neighbouring setae (Fig.189), so that setae touch one another by these processes and thanks to this stay parallel one to another; in other species these processes are much shorter or absent. Presence of long setae is connected with general rheophilous adaptation [see (2)]; peculiarity of Acentrella/fg1 is that the tibial row of setae is always regular (unlike Baetiella and some others, whose tibiae bear irregularly situated long setae on the same place).
Besides Acentrella/fg1, the same regular tibial setal row is found in some non-related taxa — e.g., Baetodes/g2 (sine Prebaetodes), Deceptivosa, Echinobaetis (belonging to Baetungulata-non-Baetofemorata), Afroptiloides and Micksiops (belonging to Protopatellata) (see Index of characters [1.2.17]).

Some species of Acentrella/fg1, besides the regular row of setae along outer-dorsal side of tibia, have an additional regular row of setae along outer side of tibia; on hind legs setae of this additional row can be long, on middle and fore legs its setae can be shorter and poorly visible. Such additional row is present in two species belonging to the plesiomorphon Acentrella/fg2, and in some species of Liebebiella/g1; this allows to conclude, that species which have this additional row, do not form a holophylum, but are dispersed within Acentrella/fg1. Outside of Acentrella/fg1, this additional setal row is not found, so potential to its occurrence can be an autapomorphy of Acentrella/fg1.

(4) Larval abdominal terga lack wide scales [see Turbanoculata (12)] (see Index of characters [1.3.4]); it can bare only long membranous scales, whose bases are as small as bases of hairs.

(5) Larval paracercus is more or less reduced, never longer than 1/2 of cerci; in most species paracercus is vestigial, many times shorter than cerci, consists either of one, or of a few segments.

Among the species examined, only gnom [Pseudocloeon] has paracercus as long as 1/2 of cerci, with vestiges of primary swimming setae (Kluge & Novikova 2011: Fig.80); in other species primary swimming setae are absent, and paracercus consists either of several segments (ibid.: Fig.76), or of a single segment (ibid.: Fig.18). Various species differ by mode of larval/subimaginal transformation of paracercus: it can transform with or without wastes. In the first case only a part of tissues of larval paracercus give raise to subimaginal paracercus [which is always one-segmented — see Liberevenata (6)]; tissues of its distal part degenerate and their wastes are shed, reminding inside larval exuviae; such molt with wastes occurs if larval paracercus is either multisegmented (Kluge & Novikova 2011: Fig.75), or consists of one elongate segment (ibid.: Fig.178). In the second case, if larval paracercus consists of one short segment, it transforms to subimaginal paracercus completely, without wastes.

Judging by distribution of these characters among species of Acentrella/fg1, they are repeatable and reversible, i.e. in course of evolution larval paracercus was repeatedly shortened and repeatedly elongated; in some species (e.g., charadra [Acentrella]) these characters vary individually. So in Acentrella/fg1 characters connected with paracercus reduction can not be used for phylogenetic analysis (but the same characters are good apomorphies in some other mayfly groups).

Adult characters:

(6) In imago and subimago anteronotal protuberance is often conic and sharply projected upward; its shape is variable; usually in lateral view it has well expressed anterior-dorsal side, which is more or less straight and much longer than its posterior-dorsal side. Angle between the anterior-dorsal and the posterior-dorsal sides can be acute and well expressed, directed dorsally (Kluge & Novikova 2011: Fig.26, 90, 129, 145, 146), but sometimes this angle is obtuse, rounded and poorly expressed (ibid.: Fig.40, 130); in bispinosa [Acentrella], cylindroculata [Acentrella], mamasae [Platybaetis] and wallacei [Platybaetis] this character is poorly expressed, and anteronotal prtuberance is always low (ibid.: Fig.159, 184); in chantauensis [Baetis] this character is not expressed at all, and anteronotal protuberance is non-projected, shallow (ibid.: Fig.56).

Besides Acentrella/fg1, similarly projected anteronotal protuberance occurs in some other taxa. Such projected anteronotal protuberance of Heterocloeon was regarded to be a synapomorphy with Acentrella (Waltz 1994). The same character occurs in dubia [Cloe], which is recently attributed to Plauditus. Relationship between Acentrella/fg1, Heterocloeon and Plauditus is matter of discussion. The same projected anteronotal protuberance occurs in some evidently non-related taxa, belonging to Baetungulata-non-Baetofemorata (Deceptiviosa and Camelobaetidius) and to Protopatellata (an undescribed African species) (see Index of characters [2.2.5])

(7) Hind wing, if present of "Acentrella-type": very narrow, always lacks costal projection, has no more than 2 longitudinal veins (see Index of characters [2.2.60]); in female hind wing is smaller than in male (Kluge & Novikova 2011: Fig.29–31, 49–51). If hind wings are present, larva has narrow protoptera, whose size is different in male and female (ibid.: Fig.18–20). In many species hind wings are completely lost; in this case larva usually has small narrow vestiges of hind protoptera (ibid.: Fig.125-126), rarely lacks hind protoptera at all (in selected individuals of Platybaetis).

Here we are temporarily placing all species with hind wings to the group lapponica of the plesiomorphon Acentrella/fg2. Hind wings are absent in all species of the groups fenestratum, inexpectatum and sibiricum (also attributed here to the plesiomorphon Acentrella/fg2) and in all species of the holophylum Jubabaetis/g1.

Loss of hind wings and their protoptera repeatedly took place in many phylogenetic branches of Turbanoculata (see Index of characters [2.2.59], [2.2.62]).

(8) 2^nd segment of gonostylus usually (but not always) has inner-apical convexity (Kluge & Novikova 2011: Fig.1, 38, 57, 70, 77, 87, 96, 97, 109, 114, 132, 140, 149, 161, 162, 197, 205).

Character of larval/subimaginal transformation.

(9) In mature larva ready to molt to subimago, buds of subimaginal gonostyli are folded under larval cuticle by the "Acentrella-type" [see Baetovectata (3)]: 2^nd and 3^rd segments of gonostylus are directed medially so that the posteriormost point locates near joining of 2^nd and 3^rd segments (Kluge & Novikova 2011: Fig.25, 34, 43-44, 55, 65, 78, 82, 91, 106, 116-117, 131, 147, 157, 163, 174-176, 183); its early stage of development is shown in Fig.206. This type of folding is similar to the "Labiobaetis-type", where the posteriormost point locates on the 2^nd segment; it occurs in non-related taxa Labiobaetis/f1=Pseudopannota/g1, Indobaetis, Acerpenna and Echinobaetis, belonging to Baetungulata-non-Baetofemorata (see Index of characters [2.3.9]). In most other taxa of Turbanoculata the posteriormost point locates on the apex of last (3^rd) segment.

Variable charactere of Acentrella/fg1.

(10) Tarsi of imago and subimago have either maximum number of apical spines [see Tetramerotarsata (1)], or apical spine of the penultimate segment is lost. Maximum number of apical spines occurs in all examined species of the plesiomorphon Acentrella/fg2; probably, this condition is initial for Acentrella/fg1. In this case middle and hind tarsi of both sexes have 3 apical spines, i.e. 1^st+2^nd, 3^rd and 4^th tarsomeres are spine-bearing (Kluge & Novikova 2011: Fig.5); fore tarsus of female also has 3 apical spines, i.e. 2^nd, 3^rd and 4^th tarsomeres are spine-bearing (Kluge & Novikova 2011: Fig.4). In all examined species of Jubabaetis/g1 fore tarsus of female has only 2 apical spines, i.e. only 2^nd and 3^rd tarsomeres are spine-bearing. Among them, in both examined species of Platybaetis middle and hind tarsi of both sexes also have only 2 apical spines, i.e. only 1^st+2^nd and 3^rd tarsomeres are spine-bearing (Kluge & Novikova 2011: Fig.185). (See Index of characters [2.2.78.2], [2.2.83]). Probably, reduction of the distalmost apical spine took place in these taxa [see below, Jubabaetis/g1 (2) and Platybaetis (4)].

Plesiomorphies of Acentrella/fg1.

(11) The styligeral muscle is retained; it is always unpaired, usually narrow and parallel-sided (Grandi 1960:XXIII: Fig. XV; Grandi 1960:86-91: Fig.2; Kluge & Novikova 2011: Fig.3, 60, 132, 149, 161, 197; ).

(12) Eggs usually with regular net-like, rib-like or tuberculate sculpture.