A cladistic analysis of relationships at the generic level was carried out for the Eriophyoidea. The analysis was based on a total of thirty-five characters drawn from the examination of investigated specimens and original descriptions of species in 17 genera examined. These characters were polarized by comparison with the generalized Tydeidae, which was chosen as the outgroup. Three equally parsimonious trees were produced using branch and bound procedure of PAUP after successive character weighting. Three distinct clades were revealed in the strict consensus tree: clade A (Pentasetacus, (Novophytoptus, Mackiella)), clade B (Phytoptus) and clade C (Trisetacus, (Nalepella, ((((Sierraphytoptus, Ashieldophyes), Phyllocoptes), ((Aberoptus, Cecidophyes), (Nothopoda, Eriophyes))), (((Diptacus, Diptilomiopus), Rhinophytoptus), Rhyncaphytoptus)))). The Phytoptidae is paraphyletic, whereas the Eriophyidae ( Sierraphytoptinus) and the Diptilomiopidae are both monophyletic. The current classification systems within the Eriophyoidea are appraised, with suggestions on further study on the phylogeny of the superfamily, and on the revision of the classification systems to reflect the natural relationships at the familial and generic levels.
Introduction
The cladistic analysis of the Eriophyoidea (Acari: Prostigmata) is discussed as part of a research project on the systematics and phylogeny of the genera of the superfamily Eriophyoidea.
The important role of eriophyoid mites in damaging plants, transmitting plant diseases, and controlling weeds has led to extensive research on various aspects of this group, including taxonomy, morphology, physiology, cytogenetics and ecology (Nault et al. 1967; Jeppson et al. 1975; Easterbrook 1978, 1979; Andrés 1982; Keifer et al. 1982; Helle & Wysoki 1983; Krantz & Ehrensing 1990; Kuang et al. 1992; Amrine & Stasny 1994; Kuang et al. 1995;). The major problem has been the lack of well-corroborated hypotheses of relationships in the Eriophyoidea.
The concept of the Eriophyoidea or Tetrapodili as a natural group within the suborder Pro stigmata is virtually uncontested. The superfamily is well defined by a number of synapomorphies: two pairs of legs with few setae; feathered or rayed empodia; absence of stigmata; peritremes, or tracheae; transverse genital opening (Krantz 1978). In contrast, considerable disagreement exists regarding the relationships within the Eriophyoidea. Four main classification systems of eriophyoid mites have been proposed. Shevtchenko (1974) suggested three superfamilies within the Tetrapodili: Phytoptoidea, including the families Phytoptidae and Sierraphytoptidae; Trisetoidea, including Trisetidae and Nalepellidae; Eriophyoidea, including Eriophyidae, Phyllocoptidae and Rhyncaphytoptidae. This system has been used by some acarologists in the former Soviet Union and other eastern European countries (Kuang 1986). Boczek et al. (1989) divided the Eriophyoidea into 6 families, namely the Ashieldophyidae, Pentasetacidae, Nalepellidae, Phytoptidae, Eriophyidae and Diptilomiopidae. The most widely used system was devised by Newkirk & Keifer (1975), who divided the Eriophyoidea into three families: Nalepellidae, Eriophyidae and Rhyncaphytoptidae. Recently Amrine & Stasny (1994) divided the Eriophyoidea into three families: Phytoptidae, Eriophyidae and Diptilomiopidae; their much more detailed system is quite similar to Newkirk & Keifer's in terms of the taxonomic structure. All the classification systems were established largely by character differentiation; their primary purpose is to provide identification guides, not to reflect the evolutionary history. A few analyses of relationships between different families in the Eriophyoidea (Huang & Huang 1990; Kuang et al. 1992; Sukhareva 1994; Kuang et al. 1995) have failed to provide strong support for any of the existing classification systems or to reinforce the notion that the existing classifications do not reflect natural relationships.
In view of these problems, the goal of the present study is to report the results of a cladistic analysis of relationships at the generic level in the Eriophyoidea, and to appraise the current classification system at the familial level.
Selection of taxa
The Eriophyoidea consist of ca. 240 described genera (Amrine & Stasny 1994). Limited by the computer power to find the most parsimonious cladogram in a reasonable amount of time and our own research budget and time, we selected 16 typical genera; six are type genera of the six tribes in the Phytoptidae (Pentasetacus Schliesske, Trisetacus Keifer and Nalepella Keifer; Novophytoptus Roivainen, Phytoptus Dujardin, Mackiella Keifer); six are type genera of the six subfamilies of the Eriophyidae (Aberoptus Keifer, Nothopoda Keifer, Ashieldophyes Mohanasundaram, Cecidophyes Nalepa, Eriophyes von Siebold, Phyllocoptes Nalepa); two genera each from the two subfamilies of the Diptilomiopidae (Diptacus Keifer, Diptilomiopus Nalepa, Rhinophytoptus Liro and Rhyncaphytoptus Keifer). The generalized Tydeidae was chosen as the outgroup, because the Tydeoidea was recently shown to be a sister group of the Eriophyoidea (Norton et al. 1993).
Character analyses
This section presents a survey of the characters used in the cladistic analysis of the Eriophyoidea. Terminology and notation follow those of Grandjean (1939, 1947), which were applied to the Eriophyoidea by Lindquist (in press, not seen by us; cited by Flechtmann 1995), and Newkirk and Keifer (1975) in parentheses. Some characters are illustrated in Figs. 1–7. The character states recognized are briefly outlined, accompanied by their respective codes, after which a more detailed discussion follows. The data are summarized in the character state matrix of Table 1. Amrine's classification system (Keifer's classification system), which is familiar to many acarologists, is used in the following character discussion to facilitate understanding.
Setae vi (median anterior setae) (Fig. 1): (0) present; (1) absent.
Vi are a pair of setae in the outgroup Tydeidae, but there is only a single seta on the median anterior part of prodorsal shield in some eriophyoid mites. It only occurs in some genera of the Phytoptidae. It is present in Pentasetacus, Trisetacus and Nalepella, but is absent in Novophytoptus, Phytoptus, Mackiella, Aberoptus, Nothopoda, Ashieldophyes, Cecidophyes, Eriophyes, Phyllocoptes, Diptacus, Diptilomiopus, Rhinophytoptus, and Rhyncaphytoptus.
Setae ve (anterior pair of prodorsal shield setae) (Fig. 1): (0) present; (1) absent.
Ve are the second pair of setae on the prodorsal shield. They occur in the Phytoptidae (except Boczekella Farkas, Trisetacus, Nalepella, Phantacrus Keifer and Setoptus Keifer), but are absent in the families Eriophyidae and Diptilomiopidae.
Dorsal tubercles: (0) absent; (1) present.
Dorsal tubercles are located on the posterior part of the prodorsal shield. Their presence, absence, location (on rear margin, ahead of rear margin, at lateral margins), orientation (longitudinal, transverse or oblique) and shape (cylindrical, on tapered flange) are often used to differentiate the subfamilies, tribes, genera and species. The dorsal tubercles are absent in the Tydeidae but present in most genera of the superfamily Eriophyoidea, except Ashieldophyes and Cecidophyes.
Setae sc (prodorsal setae) (Fig. 1): (0) present; (1) absent.
Sc setae are located on the dorsal tubercles in the genera of the Eriophyoidea; exceptions are: Asetacus Keifer, Asetadiptacus Carmona, Diptilomiopus, Neodialox Mohanasundaram, Neodiptilomiopus Mohanasundaram, Rhynacus Keifer, Vilaia Chandrapatya & Boczek and Vimola Chandrapatya & Boczek. In genera that do not have dorsal tubercles such as those in the tribe Cecidophyini Keifer, sc are absent. Sc setae are present on the prodorsal shield in the Tydeidae. The direction of sc is often used to differentiate the subfamilies, tribes, genera and species in the Eriophyoidea. The length of sc is used by some acarologists to separate different species.
As for sc in the genus Ashieldophyes, Mohanasundaram (1984) stated that the genus is characterised by the absence of dorsal shield and setae. Recent careful examination of the specimens of this genus by Dr. Amrine (personal communication) reveals that prodorsal setae sc are present, and they are located on the outer margins of the shield, about 1/3 back from the front margin.
Naso (anterior lobe of prodorsal shield): (0) absent; (1) present, (unordered).
Naso shows a variety of forms in the Eriophyoidea. Many genera in the family of Phytoptidae do not have this character, while most Phyllocoptinae mites of the Eriophyidae have the lobe. Among the genera studied, the naso is present in Pentasetacus, Nalepella, Sierraphytoptus, Mackiella, Cecidophyes, Phyllocoptes, Diptacus, Diptilomiopus, Rhinophytoptus and Rhyncaphytoptus, but absent in Trisetacus, Novophytoptus, Phytoptus, Aberoptus, Nothopoda, Ashieldophyes and Eriophyes.
Spine(s) on the naso: (0) absent; (1) present.
Not all the nasos have spines. The spines are present in nasos of Acadicrus Keifer, Acarhynchus Keifer, Aculus Keifer, Dicruvasates Abou-Awad & El-Borolossy, Neoepitrimerus Kuang & Li, Neoshevtchenko Kuang & Zhuo, Sierraphytoptus and Tetraspinus Boczek. Among the genera studied, only Sierraphytoptus shows a spine on the naso.
Location of setae sc: (0) ahead of the rear margin; (1) at the rear margin.
Sc setae are located ahead of rear margin in the genera of Pentasetacus, Trisetacus, Nalepella, Phytoptus, Sierraphytoptus, Mackiella, Nothopoda, Ashieldophyes Eriophyes, Phyllocoptes, Diptacus, Diptilomiopus, Rhinophytoptus and Rhyncaphytoptus; and at the rear margin in the genera Novophytoptus and Aberoptus. They are absent in Cecidophyes. There are often two pairs of sc setae in the Tydeidae and other prostigmatid families. Setae sc of Ashieldophyes are located on the outer margins of the shield, about 1/3 back from the front margin (Amrine, personal communication).
Direction of setae sc: (0) forward; (1) backward; (2) upward or inward.
The prodorsal setae sc in Pentasetacus, Trisetacus, Nalepella, Phytoptus, Mackiella, Diptacus, Rhinophytoptus and Rhyncaphytoptus are directed forward; and those in Novophytoptus, Aberoptus and Nothopoda are directed backward. Setae sc in Ashieldophyes are directed upward and laterally (Amrine, personal communication). Usually setae sc directing forward or backward are long, whereas those directing upward or inward in Sierraphytoptus and Phyllocoptes are short.
The different understanding of the location and direction of setae sc has caused turmoil in the taxonomy of the Eriophyoidea. In their publication of “Revision of types of Eriophyes and Phytoptus”, Newkirk and Keifer (1971) removed species from Aceria and reassigned them to the genus Eriophyes; the genus Aceria discontinued. Mites previously assigned to Phytoptus were assigned to the new genus, Phytocoptella (Newkirk & Keifer 1971). Their work was based on careful library research and the discovery of older species names, having priority, which resulted in the designation of types and thus the redefinition of generic concepts (Amrine & Stasny 1994). Shevtchenko (1974b), Lindquist et al. (1977), and Lindquist and Manson (1987) objected to the designation of the new types for Eriophyes and Phytoptus because of long established usage, and began petitions to the International Commission of Zoological Nomenclature for the restoration of the types for Aceria, Eriophyes and Phytoptus. In March 1989, the Commission published Opinion 5721 which restored the former uses of the disputed genera and set the types for each genus as follows: 1. Eriophyes von Siebold, type species Phytoptus pyri Pagenstecher; 2. Phytoptus Dujardin, type species Phytoptus avellanae Nalepa; 3. Aceria Keifer, type species Eriophyes tulipae Keifer; and 4. Colomerus Newkirk & Keifer, type species Eriophyes gardeniella Keifer. As a direct result of this confusion, many acarologists since 1971 have placed eriophyoid mites into the wrong genus (especially the confusion of Eriophyes and Aceria) and sometimes, in the wrong family (Amrine & Stasny 1994). It is hoped that researchers working on eriophyoid mites will now be able to correctly assign species due to the Opinion 5721 of International Commission of Zoological Nomenclature (1989), and Catalog of the Eriophyoidea (Acarina: Prostigmata) of the world by Amrine and Stasny (1994).
Body shape: (0) worm-like; (1) fusiform (spindle-shaped).
The body shape varies among the Eriophyoidea. Pentasetacus, Trisetacus, Novophytoptus, Phytoptus, Mackiella, Nothopoda, Ashieldophyes and Eriophyes have the worm-like body; Nalepella, Sierraphytoptus, Aberoptus, Cecidophyes, Phyllocoptes, Diptacus, Diptilomiopus, Rhinophytoptus and Rhyncaphytoptus have the spindle-shaped body. Many genera in the Phytoptidae and a few genera in the Eriophyidae have the worm-like body, whereas many genera in the Eriophyidae and all genera in the Diptilomiopidae except the genus Stenarhynchus Mohanasundaram have the spindle-shaped body.
Cheliceral curvature: (0) evenly curved; (1) abruptly curved.
The chelicerae in the families Phytoptidae and Eriophyidae are evenly down-curved, whereas those in the family Diptilomiopidae abruptly bent down near base. Gnathosoma in the Diptilomiopid mites are large in comparison to body, attenuate, and contain long cheliceral stylet. Mites of this family are often called big-rostrum mites. They are leaf vagrants only, rarely causing much damage to their hosts (Jeppson et al. 1975). Among the genera examined, the chelicerae curve abruptly in Diptacus, Diptilomiopus, Rhinophytoptus and Rhyncaphytoptus.
Location of genital area: (0) not appressed to coxae II; (1) appressed to coxae II.
All the genera in the subfamily Cecidophyinae of the Eriophyidae have a genital area that is appressed to the coxal area. In all the other subfamilies, the genital area is usually not appressed to the coxal area and there are several ventral annuli between the coxal area and the genital area. The number of ventral annuli between the genital area and the coxae has been used by a few acarologists to describe species. The genital area in Aberoptus, Ashieldophyes and Cecidophyes are appressed to the coxal area; whereas the genital area in the other genera are not appressed to the coxal area. The genital area in the Tydeidae is far away from the second coxae.
Setae c1 (subdorsal setae) (Fig. 7): (0) present; (1) absent.
Setae c1 are present only in 10 genera of the Phytoptidae (Acathrix Keifer, Anchiphytoptus Keifer, Austracus Keifer, Boczekella, Fragariocoptes Roivainen, Pentasetacus, Phytoptus, Prothrix Keifer, Sierraphytoptus and Trisetacus) and absent in most other genera of the Eriophyoidea. The presence, ring position, length, spacing etc. of c1 are used to describe and differentiate the subfamilies, genera, tribes and species.
Setae d (first ventral setae) (Fig. 7): (0) present; (1) absent.
With the exception of Ashieldophyes, all genera examined have setae d. In almost 240 reported genera of the Eriophyoidea, setae d are absent in the following 15 genera: Acamina Keifer, Ashieldophyes, Cecidodectus Nalepa, Diptilorhynchus Mondal & Chakrabarti, Hemiscolocenus Mohanasundaram, Hornophyes Mohanasundaram, Knorella Keifer, Neocupacarus Das & Chakrabarti, Neodicrothrix Mohanasundaram, Neomesalox Mohanasundaram, Proneotegonotus Mohanasundaram, Retracus Keifer, Scolocenus Keifer, Schizacea Keifer and Tumescoptes Keifer. The presence, ring position, length and spacing of setae d are often used to describe and differentiate genera and species.
Setae e (second ventral setae) (Fig. 7): (0) present; (1) absent.
With the exception of the genus Ashieldophyes, all genera examined have setae e which are usually the shortest of the three pairs of ventral setae. Among 240 described genera of the Eriophyoidea, setae e are absent in the following 19 genera, namely, Acamina Keifer, Amerineus Flechtmann, Asetilobus Manson, Ashieldophyes, Cecidodectus Nalepa, Dicrothrix Keifer, Hornophyes Mohanasundaram, Knorella Keifer, Neocupacarus Das & Chakrabarti, Neodicrothrix Mohanasundaram, Paracolopodacus Kuang & Huang, Paraphytoptella Keifer, Phyllocoptacus Mohanasundaram, Proneotegonotus Mohanasundaram, Prophyllocoptes Mohanasundaram, Ramaculus Manson, Schizacea Keifer, Surapoda Boczek & Chandrapatya and Tumescoptes Keifer. The presence, ring position, length and spacing of setae e are often used to describe and differentiate genera and species.
Setae lb on coxae I (first pair of forecoxal setae) (Fig. 2): (0) present; (1) absent.
Setae lb are absent in the forecoxae of Nothopoda and Diptilomiopus, and present in those of other genera examined. Mohanasundaram (1984) stated that Ashieldophyes is characterised by the absence of dorsal shield and setae, first and second ventral setae, and first and second coxal setae. After careful examination of the specimen, Dr. Amrine (personal communication) found that all three coxal setae are present. Usually there are two pairs of setae on the forecoxae and one pair on the hindcoxae. Variation occurs in the forecoxal setae, but not in the hindcoxal setae.
Seta bv on femur I (forefemoral seta) (Fig. 3): (0) present; (1) absent.
A single seta bv is usually located near mid-ventral 1/2 on the forefemur. The presence, length and position are used to describe and differentiate genera and species. It is present in Pentasetacus, Trisetacus, Nalepella, Phytoptus, Sierraphytoptus, Mackiella, Aberoptus, Nothopoda, Ashieldophyes, Cecidophyes, Eriophyes, Phyllocoptes, Rhinophytoptus and Rhyncaphytoptus, but absent in Novophytoptus, Diptacus and Diptilomiopus.
Seta l″ on genu I (foregenual seta) (Fig. 3): (0) present; (1) absent.
With the exception of the genus Diptilomiopus whose seta ľ is absent, all the other genera examined have seta l″. Seta l″ is always dorsal or laterodorsal, and directed upward. The presence and length of seta l″ are used to differentiate genera and species.
Seta l′ on tibia I (foretibial seta) (Fig. 3): (0) present; (1) absent.
Seta l′ is usually located at basal 1/3 of tibia I and is a very important character for differentiating tribes and genera. The presence, length and position on tibia are often used to describe genera and species. Among the genera examined, seta 1″ is absent on tibia I of Aberoptus, Nothopoda and Diptilomiopus, but is present in those of other genera.
Seta bv on femur II (hind femoral seta) (Fig. 4): (0) present; (1) absent.
The hind femoral seta bv is located at mid-ventral 1/2 on the hindfemur. The presence, length and position are used to describe and differentiate genera and species. It is present in Pentasetacus, Trisetacus, Nalepella, Phytoptus, Sierraphytoptus, Mackiella, Aberoptus, Nothopoda, Ashieldophyes, Cecidophyes, Eriophyes, Phyllocoptes, Rhinophytoptus and Rhyncaphytoptus, but absent in Novophytoptus, Diptacus and Diptilomiopus.
Seta l″ on genu II (hind genual seta) (Fig. 4): (0) present; (1) absent.
With the exception of the genus Diptilomiopus, all the other genera examined have seta l″ on genu II. The presence or absence of seta l″ on genu II is often used to differentiate tribes and genera in the Eriophyoidea.
Tibia: (0) normal; (1) reduced or fused.
In the subfamilies Aberoptinae and Nothopodinae, tibiae are reduced or completely fused with tarsi; foretibiae never bear setae. All the other eriophyoid mites have normal tibiae which are always distinct from tarsi, fortibial seta nearly always present except in a few genera. Among the genera studied, tibiae in the genera of Aberoptus and Nothopoda are reduced or fused with tarsi, whereas tibiae in other genera do not show this reduction.
Solenidion φ on tibia I (solenidion on foretibia): (0) present; (1) absent.
Solenidion φ is a lateral foretibial spur. The foretibiae in the Phytoptidae usually bear φ, while those of the Eriophyidae and Diptilomiopidae rarely do. Solenidion φ is present in Pentasetacus, Trisetacus, Nalepella, Phytoptus and Mackiella, but absent in Novophytoptus, Sierraphytoptus, Aberoptus, Nothopoda, Ashieldophyes, Cecidophyes, Eriophyes, Phyllocoptes, Diptacus, Diptilomiopus, Rhinophytoptus and Rhyncaphytoptus.
Ridge(s) or trough(s) on the hysterosoma: (0) absent; (1) present.
A slight longitudinal subdorsal furrow is present on the hysterosoma of Diptacus. A central ridge is present on hysterosoma in Diptilomiopus. The ridge is not stable, and is not very obvious in the genus Rhyncaphytoptus; in other words, ridge appear in some species and does not appear in other species of Rhyncaphytoptus. Ridge(s) or trough(s) are absent in other genera studied.
Empodium (Fig. 6): (0) simple (normal); (1) not normal (divided, palm-shaped etc.).
The length, position, number of rays, configuration (divided or undivided), and shape of empodium are important for distinguishing subfamilies, tribes, genera and species. Empodia in Pentasetacus, Diptacus and Diptilomiopus are divided, those in Aberoptus are reduced to a bristle; and those in other genera are normal (undivided). Empodia in the Tydeidae are undivided.
Setae c2 (lateral setae) (Fig. 7): (0) present; (1) absent.
With the exception of Diptilomiopus, all the other genera examined have setae c2. In the Eriophyoidea, setae c2 are absent in only 15 genera (Acarhis Keifer, Africus Smith Meyer & Ueckermann, Amrineus Flechtmann, Dacundiopus Manson, Diptilorhynacus Mondal, Ghosh & Chakrabarti, Diptilomiopus, Lambella Manson, Levonga Manson, Pseudodiptacus Chakrabarti, Ghosh & Das, Rhynacus Keifer, Thacra Keifer, Vilaia Chandrapatya & Boczek and Vimola Boczek). It is interesting that 13 of the above 15 genera belong to the family Diptiliomiopidae; only two genera (Amerineus and Thacra) belong to the Eriophyidae. The presence, ring position, length and spacing of setae c2 are used to describe genera and species.
Ridges of the female genital coverflap (Fig. 2): (0) absent; (1) one longitudinal row; (2) two longitudinal rows or transverse lines.
Scoring or sculpture of female genital coverflap is a very important character which is used to differentiate subfamilies, tribes, genera and species. In the Cecidophyinae, the female genitalia, which noticeably project from hysterosoma venter, are appressed to coxae; longitudinal ridges of female genital coverflap typically are in 2 ranks. In other subfamilies in the Eriophyoidea, genitalia lie more on level with hysterosoma venter, and are not appressed to coxae; female genital coverflap are variably sculptured, and ridges are rarely in 2 ranks. Usually female genital coverflap in the Phytoptidae are without ridges except in Fragariocoptes Roivainen.
Among the genera studied, there is no sculpture on the female genital coverflap of Pentasetacus, Trisetacus, Nalepella, Novophytoptus, Phytoptus, Sierraphytoptus, Mackiella, Nothopoda, Ashieldophyes, Diptilomiopus and Rhinophytoptus. There are 2 ranks of ridges on the female genital coverflap of Cecidophyes. On the female genital coverflap of Eriophyes, Phyllocoptes and Rhyncaphytoptus, sculpture varies. Some species have smooth coverflap, while. others have one longitudinal row ridge.
Spatulate or shovel-shaped projections on legs: (0) absent; (1) present.
In the subfamily Aberoptinae of Eriophyidae, spatulate or shovel-shaped projections are present, either on rostral termen, or on tarsi; legs lacking spatulate appendages are very stout; segments are shortened or combined. With the exception of the genus Aberoptus, all the other genera studied and the Tydeidae do not have spatulate or shovel-shaped projections on legs. The location of spatulate or shovel-shaped projections is used to differentiate genera and species within the Aberoptinae.
Lateral opisthosomal differentiation: (0) absent; (1) differentiate into dorsal tergites and ventral sternites.
The bodies of mites in Pentasetacus, Trisetacus, Novophytoptus, Phytoptus, Nothopoda, Ashieldophyes and Eriophyes are elongate and wormlike, whereas those in Nalepella, Sierraphytoptus, Mackiella, Aberoptus, Cecidophyes, Phyllocoptes, Diptacus, Diptilomiopus, Rhinophytoptus, Rhyncaphytoptus are more or less fusiform and are differentiated into dorsal tergites and ventral sternites.
Dorsal annuli: (0) not extended laterally; (1) extended laterally or with indentations.
With the exception of Sierraphytoptus whose dorsal annuli project a little laterally, all the other genera examined have normal dorsal annuli which do not extend laterally. The extension of dorsal annuli in some eriophyoid mites is seen as a peculiar character, and is used to describe and differentiate genera.
Sternal line: (0) absent; (1) present.
Sternal line is absent in Aberoptus, Nothopoda, Ashieldophyes, Cecidophyes and the Tydeidae, but is present in the genera Pentasetacus, Novophytoptus, Phytoptus, Sierraphytoptus, Mackiella, Phyllocoptes, Diptacus, Rhinophytoptus and Rhyncaphytoptus. It is unstable in Trisetacus, Nalepella, Eriophyes, Diptilomiopus among which some species have sternal line whereas others do not. The presence, development and shape of sternal line are important for differentiating and describing genera and species.
Setae h1 (accessory setae) (Fig. 7): (0) present; (1) absent.
Setae h1 are present in the genera Pentasetacus, Trisetacus, Nalepella, Novophytoptus, Phytoptus, Sierraphytoptus, Mackiella, Ashieldophyes and the family Tydeidae; and absent in the genera Aberoptus, Nothopoda and Cecidophyes. They are present in some species and absent in other species in the genera Eriophyes, Phyllocoptes, Diptacus, Diptilomiopus, Rhinophytoptus and Rhyncaphytoptus. The length, presence and spacing of setae h1 are used to decribe genera and species.
Spermathecal tubes: (0) long; (1) short.
Spermathecal tubes are usually long in the Phytoptidae and short in the Eriophyidae and Diptilomiopidae. They are long in Pentasetacus, Trisetacus, Nalepella, Novophytoptus, Phytoptus, Sierraphytoptus and Mackiella, but short in Aberoptus, Nothopoda, Ashieldophyes, Cecidophyes, Eriophyes, Phylllocoptes, Diptacus, Diptilomiopus, Rhinophytoptus and Rhyncaphytoptus. The orientation, length, position of spermathecal tubes and spermathecae are often used to describe species. Due to the techniques of specimen preparation and the limitation of microscope, the spermathecal tubes may not be discernible and therefore not given in the descriptions of some species.
Length of setae sc: (0) very long; (1) long; (2) short; (3) absent.
Length of sc is a relative concept. Very long means that the seta is longer than the distance between 2 tubercles; long means the seta is longer than half the distance; short means the seta is shorter than half the distance. Sc setae are very long in Pentasetacus, Trisetacus, Nalepella, Novophytoptus, and Phytoptus, long in Aberoptus, Eriophyes, Phyllocoptes and Rhyncaphytoptus, and short in Sierraphytoptus, Nothopoda, Ashieldophyes, Diptacus and Rhinophytoptus. They are absent in Cecidophyes and Diptilomiopus.
Microtubercles on dorsal annuli: (0) absent; (1) present.
Microtubercles are absent on the dorsal annuli of Sierraphytoptus, Mackiella and Ashieldophyes, but present on dorsal annuli of Pentasetacus, Trisetacus, Novophytoptus, Phytoptus, Aberoptus, Nothopoda, Cecidophyes and Eriophyes. They vary among the species in Nalepella, Phyllocoptes, Diptacus, Diptilomiopus, Rhinophytoptus and Rhyncaphytoptus. Taking Diptacus as an example, there are microtubercles on dorsal annuli in D. flocculentus Keifer, D. ragarifoliae Keifer, D. gigantorubra Xin & Dong, D. prunorum (Keifer), D. pseudocerasis Kuang & Hong, D. rubra Keifer and D. swensoni Keifer; no microtubercles on dorsal annuli in D. aceris Kuang & Hong, D. calicoryli (Keifer), D. castaneae Kuang & Feng, D. gigantorhynchus (Nalepa), D. guangxiensis Kuang & Hong, D. liquidambaris Kuang & Feng, D. maackiae Kaung & Feng, D. sacramentae (Keifer) and D. ulmi Kuang & Huang. When microtubercles are present on dorsal annuli, there are variation in the form (triangle, bead, elongate, blunt, rounded etc.), location (a little ahead of ring margin, on ring margin, reaching ring margin etc.) and density.
Location comparison between setae la and 2a (Fig, 2): (0) la ahead of 2a; (1) la same line as 2a; (3) la behind 2a.
The location of coxal setae (lb, la and 2a) is quite stable, although few acarologists use them in describing the eriophyoid mites. Keifer (1938) was the first to compare coxal setal locations, and Smith Meyer always compares coxal seta locations in her descriptions (e.g. Smith Meyer 1981, 1989; Smith Meyer & Ueckermann 1995). In the genera Trisetacus, Nalepella, Phytoptus, Sierraphytoptus, Nothopoda, Ashieldophyes, Cecidophyes, Eriophyes, Phyllocoptes, Diptacus, Diptilomiopus, Rhinophytoptus, Rhyncaphytoptus, setae la are ahead of setae 2a; in the genera Pentasetacus, Mackiella and Aberotptus, setae la are almost on the same line as setae 2a; and in the genus Novophytoptus, setae la are behind of setae 2a.
FIGURES. 1–7.
A sample of characters and notations in the Eriophyoidea. 1, prodorsal shield and associated setae of Pentasetacus araucariae Schliesske; 2, coxal area of Aceria tulipae (Keifer); 3, foreleg of A. tulipae; 4, hindleg of A. tulipae; 5, simple empodium; 6, divided empodium; 7, lateral view of Phytoptus avellanae Nalepa.
TABLE 1.
Character state matrix for the outgroup and the examined eriophyoid genera
Analysis and results
Characters of each genus were determined by examining specimens and original descriptions representing as many species as possible. The character state matrix given in Table 1 was analyzed using the parsimony program PAUP 3.0 (Swafford, 1991) on a Macintosh Centris 650 Computer. In Table 1, unknown or unstable character states are coded ‘?’. All characters are ordered except the character 5 (the prodorsal shield lobe). The family Tydeidae of Tydeoidea, which is the sister group of the Eriophyoidea according to Norton et al. (1993), was selected as the outgroup.
The branch and bound procedure produced 33 equally parsimonious trees (length = 79; consistency index = 0.506; retention index = 0.606). According to Farris (1988), successive weighting may have the advantage of providing a means of basing groups on more reliable characters without making prior decisions on. weighting. Successive weighting was used to identify if there were more characters in the present character set supporting monophyletic groups. Branch and bound procedure produced 3 equally parsimonious trees after the first and second reweighting. Consensus procedure produced 1 strict consensus tree (Fig. 8).
FIGURE 8.
A cladogram showing the hypothetical relationships among selected genera of the Eriophyoidea. It is the strict concensus tree of three equally parsimonious trees. 8a = sc directing forward; 8b = sc directing upward or inward; 33a = sc long; 33b = sc short. See text for details.
The monophyly of Novophytoptus + Mackiella is supported by two synapomorphies: the absence of seta vi (character 1) and setae c1 (character 12). The monophyly of Novophytoptus + Mackiella + Pentasetacus is supported by one synapomorphy: coxal setae la not ahead of coxal setae 2a.
Sierraphytoptus + Ashieldophyes forms a monophyletic group, based on two synapomorphies: short setae sc (character 33) on prodorsal shield and presence of microtubercles on dorsal annuli (character 34). The group (Phyllocoptes, (Sierraphytoptus, Ashieldophyes)) is united by upward or inward setae sc on the prodorsal shield (character 8). The monophyly of Aberoptus + Cecidophyes is supported by two synapomorphies: female genital area close to coxae (character 11) and presence of ridges on female genital coverflap (character 26). The monophyly of Nothopoda and Eriophyes is based on three synapomorphies: the absence of naso (character 5), worm-like body (character 9) and dorsoventrally subequal opisthosoma (character 28). The sister-group relationship of (Aberoptus, Cecidophyes) and (Nothopoda, Eriophyes) is supported by the sternal line (character 30) and setae h1 (character 31).
Diptacus and Diptilomiopus form a monophyletic group, based on four synapmorphies: the absence of setae bv on femur I (character 16), the absence of setae bv on femur II (character 19), the presence of ridges on hysterosoma (character 23) and divided empodium (character 24). The group ((Diptacus, Diptilomiopus), Rhinophytoptus) forms a monophyletic group, based on the short sc on the prodorsal shield. The Diptilomiopidae, represented by (((Diptacus, Diptilomiopus), Rhinophytoptus), Rhyncaphytoptus), forms a monphyletic group, based on the abruptly curved chelicerae (character 10).
All the genera in the Diptilomiopidae + Eriophyidae, and the genus Sierraphytoptinus from Phytoptidae form a monophyletic group, supported by four synapmorphies: the absence of setae vi (character 1), the absence of solenidion on foreleg (character 22), short spermathecal tubes (character 32) and reduced length of setae sc (character 33). The sister group of this large group is Nalepella; this is based on two synapmorphies: the absence of setae c1 (character 12) and differentiated opisthosoma (character 28) (with reversal in Nothopoda and Eriophyes). The monophyly of above genera and Trisetacus is supported by two synapmophies: the absence of setae ve (character 2) and the fusiform body (character 9) (with reversal in Nothopoda and Eriophyes).
Discussion
Studies on the higher level relationships in the Eriophyoidea are few (e.g. Huang & Huang 1990; Kuang et al. 1992). Both phenetic and cladistic approaches have been used.
Huang and Huang (1991) used both phenetic and cladistic methods to investigate the phylogenetic relationships among 15 eriophyoid species which belong to 11 subfamilies. Their results show that phenetic analysis based on morphometric measurements failed to reveal natural groups and the phenograms are inconsistent with classical taxonomy. In their cladistic analysis of the same 15 species, Huang and Huang (1990) did not select an outgroup; character polarization was based on the assumption that the Diptilomiopidae have plesiomorphic states of characters. They used the program “Phylip” to construct the cladogram and reported a tree with length 32. They did not mention what procedure they used and whether the reported tree was the shortest one. We did an analysis of their data (in their Table 5) using PAUP; the branch and bound procedure produced 62 shortest trees of equal length 28. Clearly, their tree of 32 steps is not the shortest. Nevertheless, Huang and Huang (1991) concluded that both Eriophyidae and “Rhyncaphytoptidae” (i.e. Diptilomiopidae) are monophyletic, whereas “Nalepellidae” (i.e. Phytoptidae) is paraphyletic. They did not explain why the five sections of the Phyllocoptinae scattered in different clades within the Eriophyidae.
Kuang et al. (1992) used polyacrylamide gel electrophoresis to examine patterns of esterase isozymes of five eriophyoid mites: Trisetacus juniperus (Nal.), Boczekella pseudolaris Kuang & Shen, Tegolophus fontanesiae Kuang & Hong, Aculus ligustri Keifer and Rhyncaphytoptus lonicera Kuang & Zhuo. Cluster analyses of the isozyme patterns and 18 morphological characters using Euclidean distance and coefficient of similarity showed that the Diptilomiopidae is more similar to the Eriophyidae than to the Phytoptidae. A similar study on karotypes and 18 morphological characters of 10 eriophyoid mite species showed that the Eriophyidae and Diptilomiopidae are more similar to each other than either to the Phytoptidae (Kuang et al. 1995).
Our cladistic analysis showed that the Eriophyidae and Diptilomiopidae are closer to each other than to the paraphyletic Phytoptidae (Fig. 1). The Phytoptidae have more plesiomorphous characters than the other two families. The genus Pentasetacus has the most plesiomorphous characters in the Eriophyoidea.
A comparison of the current classification systems of the Eriophyoidea is presented in Table 2. The use of numbers of dorsal setae by current classifications for the diagnosis of the Phytoptidae emphasize its differences from the other families; there is a lack of synapomorphies for the Phytoptidae. Huang & Huang (1990) also showed that four subfamilies in Nalepellidae form a paraphyletic group. We feel that it is probably more reasonable to divide the Phytoptidae into several monophyletic families: e.g. Pentasetacidae, Novophytoptidae, Phytoptidae, Trisetacidae, Nalepellidae and Sierraphytoptidae based on the present study. However, before the new system is formally proposed, more detailed research is needed because many more species of the Eriophyoidea remain undescribed and other characters (e.g. DNA data) have not been studied. It seems to us that the system erected by Boczek et al. in 1989 is a step toward dividing the paraphyletic Phytoptidae, although their idea on the Ashieldophyidae is incorrect due to the wrong description by Mohanasundaram in which Ashieldophyes was diagnosed as not having dorsal shield and setae and coxae setae (Mohanasundaram 1984). In fact, these characters are present (Amrine, persona] communication).
TABLE 2.
Comparison of classifications of the eriophyoid mites
Although the genera of Eriophyidae + Sierraphytoptus from Phytoptidae form a monophyletic group, based on the direction of dorsal setae sc which is backward or upward or inward, there is a lack of apomorphous characters which can be used to define all the genera in the Eriophyidae. The relationships within the family are well defined (Fig. 1). The monophyly of the Diptilomiopidae is supported by a single synapomorpy: abruptly curved chelicerae on the characteristic big gnathosoma. The sister group relationship between the Eriophyidae and Diptilomiopidae are strongly supported (Fig. 1).
The cladistic study on the Eriophyoidea presented here, although tentative and incomplete, identifies the phylogenetic relationships among the families, appraises the current classification systems, and underscores the need for increased study of the families Phytoptidae and Eriophyidae which are poorly defined. The relatively low consistency index in the cladogram indicates a relatively high degree of homoplasy in the morphological characters used in the current classification of the Eriophyoidea.
Most studies on the Eriophyoidea have emphasized finding the differences among taxa and describing new taxa. It is high time that more attention should be paid to critical analyses of characters (morphological, molecular, etc) and more efforts be devoted to finding shared derived characters that define monophyletic groups in the Eriophyoidea.
Acknowledgements
This research project was completed when the senior author was a Darwin Fellow at the International Institute of Entomology, London. We thank the Darwin Initiative for the Survival of Species, Department of Environment of the United Kingdom for financial support.
We thank Prof. V. K. Brown, Ms. A. Walker, Dr. J. LaSalle, Dr. G. Watson, Dr. D. Agassiz, Mr, G. Sandlant and Mr G. Kibby of International Institute of Entomology for their help and assistance during the study.
We thank Profs. Jin-Hua Ding, Xia-Nian Cheng, Hua-Guo Dai and Yin-Chang Wang, Ms. Jian-Yu Chu, Mr. Hai-Kou Wang, Ms. Ye-Hua Yang and Mr. Yong-Yin Wang of Nanjing Agricultural University, China for their help and assistance.
We thank Dr. M. K.P. Smith Meyer and Dr. E. A. Ueckermann of Plant Protection Research Institute, Pretoria, South Africa for the loan of the specimens; Professor J. W. Amrine, Jr., Division of Plant and Soil Sciences, College of Agriculture and Forestry, West Virginia University, U.S.A. for allowing the senior author to examine specimens in his collections; Dr. R. Smiley of Systematic Entomology Laboratory, USDA for arranging for the senior author to examine Keifer's collection; Dr. Kun-Wei Huang of National Museum of Science, Taichung, Taiwan and Dr. H. Dastych of Zoologisches Institut und Zoologisches Museum, Universitat Hamburg, Germany for the loan of the specimens in their care; Prof. C.H.W. Flechtmann of University of Sao Paulo, Brazil, Professor Tsan Huang of National Chung Hsing University, Taiwan, Prof. M. Mohanasundaram of Tamil Nadu Agricultural University, India and Dr. V. V. Ramamurthy of Division of Entomology, Indian Agricultural Research Institute, India for sending reprints. We are especially grateful to Prof. J. W. Amrine, Jr. who provided a lot of useful information and valuable suggestions.
Space and facilities during this study were kindly provided by the Department of Entomology, The Natural History Museum, London.
References
Notes
[1] Note added in proof: The paper by E. E. Lindquist on external anatomy and notation of structures in eriophyoid mites, which was cited in this paper but was not seen by us when this paper was prepared, was published in June 1996. It is added below.
Lindquist, E.E. (1996) External anatomy and notation of structures. In: Lindquist, E.E., Sabelis, M.W. & Bruin, J. (eds.) Eriophyoid Mites — Their Biology, Natural Enemies and Control. Amsterdam: Elsevier, pp. 3–31.
