F. M. Carpenter.
The Lowe Permian Insects of Kansas. Part 11. The Orders Protorthoptera and Orthoptera.
Psyche 73:46-88, 1966.

Full text (searchable PDF, 8580K)
Durable link: http://psyche.entclub.org/73/73-046.html

The following unprocessed text is extracted from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.

THE LOWER PERMIAN INSECTS OF KANSAS. PART 11. THE ORDERS PROTORTHOPTERA
AND ORTHOPTERA1
BY F. M. CARPENTER
Harvard University
The two preceding papers in this series dealt with representatives of seven closely related families of the Order Protorthoptera occurring in the Elmo limestone.
The present paper treats additional families of more diverse relationships within that order and also covers several families of the Order Orthoptera.
The problems involved in the systematics of the Palaeozoic orthop- teroids are intrinsically very great, mainly as a result of our frag- mentary knowledge of most species but also as a result of the variability of the venation within species. It was my belief more than twenty years ago (1943, pp. 76-77) that the classification of the Palaeozoic orthopteroids as suggested by Handlirsch first and by Martynov later was not realistic in the light of our knowledge at that time.
Since then many additional orthopteroids have been de- scribed, mostly from the Lower and Upper Permian strata of the Soviet Union. These new fossils have added greatly to our knowl- - -
edge of the early history of the orthopteroid complex. Through the courtesy of Dr. B. B. Rohdendorf, Arthropod Section, Palaeontolog- ical Institute. Academy of Sciences, in Moscow, I had the opportunity in 1961 of studying both the undescribed and described material in the collection of the Institute; and of discussing with Dr. Sharov, Dr. Martynova, Dr. Bekker-Migdisova and other staff members of the Institute various oroblems of insect evolution. I would be remiss if I did not acknowledge at this time my gratitude to the entire staff of the Institute for their kindness and help during and subsequent to my stay.
During the past decade I have been able to study additional orthop- teroids collected at the Elmo locality and especially in the Midco beds in Oklahoma. Two additional trips to, the Institute de Palionto- logic in Paris have enabled me to make further examination of the Commentry fossils, which I still consider (in spite of the remarkable fossil insects from Tchekarda in the Soviet Union) the foundation on which our understanding of Palaeozoic insects rests. 'This research has been supported in part by a National Science Founda- tion Grant, No. GB 2038.
Part 10 of this series was published in the Proc. Amer. Acad. Arts Sci., 78 ~185-219, 1950.
Pnche 73:46-88 (1066). hup //psyche enlclub orgi73173-IM6 html



================================================================================

19661 Carpenter - Protorthoptera and Orthopiera 47 In the present state of our knowledge, the classification of the Palaeozoic orthopteroids is necessarily based on the venation of the fore wings, the hind wings and body structures being very little known at best and entirely unknown in by far the majority of species. Since considerable difference of interpretation exists in even the re- cent literature on the orthopteroid venation, I consider it necessary to present here my own views on the homologies of wing veins in these particular insects and indeed in insects in general. I find that few students of insects have any understanding of the problems of vein homology or of the current status of the subject. The following account is intended to present the background and the nature of my own views used throughout this paper and the subsequent parts in the series.
Although some preliminary attempt was made by Hagen (1870) to hokologize the wing veins of insects, Red tenbacher ( I 886) was the first to make a significant contribution to the subject. He pro- posed the recognition of six main veins, which he termed the costa, subcosta, radius, media, cubitus and anal. In reaching his conclusions, he considered the general correlation of the positions of the veins, as well as a primitive alternation of topography, i.e., convexity and concavity. The Redtenbacher System of nomenclature was followed by Comstock and is actually the one which has been in general use, although it is commonly referred to as the Comstock-Needham Sys- Comstock's first publications on wing veins appeared in 1892. In 1895, J. G. Needham, then a graduate student under Comstock, began a new approach to the study of wing vein homology and the ontogenetic development of wings and their veins. Results of these studies were first published in a series of articles under joint author- ship of Comstock and Needham in 1898 and 1899. An extensive series of papers, mainly by Needham, appeared in subsequent years and in 1918 Comstock brought together in book form a compilation of what had been done in his and Needham's laboratories. They con- dueled that the various patterns of wing venation in insects had been derived from a common ancestral type and that the veins of different orders could be homologized. The Redtenbacher System of nomen- clature was used by them, although no significance was attached to the convexity or concavity of the veins. As noted above, the innovation brought into their venational studies was the ontogenetic method. Noting that in such primitive insects Tomstock himself pointed out (1918, p. 11) that this nomenclature should be recognized as the Redtenbacher System, not the Cornstock-Needham System.




================================================================================

48 Psyche [March
as the Plecoptera only tracheae could be seen in the developing wings, and that the pattern of venation of the adult wing agreed closely with the pattern of tracheation in the wing pad, they concluded that tracheae determined the positions of the veins (Comstock, 1918, p. 12). They also concluded that the ontogenetic history of the tracheal pattern recapitulated the phylogenetic history of the vena- tion in the group of insects concerned. Applying these principles to the Odonata and Ephemeroptera, for example, they reached the un- expected conclusion that a branch of the radius vein had crossed over a branch of the media in the course of the evolution of these groups; the trachea appeared to cross over in the wing pad and this, in their view, meant that the vein had done likewise in previous geologic time. Objections to the tracheation theory of vein deter- mination and especially to the recapitulative conclusions were raised by several students of fossil insects and insect evolution (e.g., Till- yard, Martynov, Carpenter, Fraser ) in the period from 1923-1935. In 1935, Needhain reiterated his stand on the ontogenetic-phylogenetic relationship of tracheae and veins ; and in I 95 I, he published a more detailed discussion in defense of this thesis, especially as it related to the Odonata; although a substantial part of his paper was an at- tempt to ridicule in a personal manner all individuals who had dis- agreed with him.;
As Needham himself indicated ( 1935, p. 129) there had not been undertaken up to that time a thoroughgoing investigation of the de- velopment of nymphal wings of any species, at least with respect to the development of tracheae and veins. Shortly after, however, such an investigation was made by Holdsworth ( I 940, 1941 ) , this con- sisting of a histological study of the development of wing pads, tracheae and veins, starting with the earliest beginnings of the wing buds. The plecopteran, Pteronarcys, was chosen because Comstock and Needham considered the stone-flies as demonstrating most clearly the tracheal determination of veins. Holdsworth's results were strikingly clear: the tracheae did not enter the main area of the wing pads until the blood spaces or lacunae between the blocks of epidermal cells had already established the positions of the veins. The tracheae, as they grew longer, simply entered the lacunae which had already been blocked out, following the lines of least resistance. Variation in the tracheal branching was obvious and usually several Qne can only regret that this final paper on this subject by Needham was so vindictive. It contributed nothing to science and detracted from Needham's image as a scientist. It also earned a black mark for the Amer- ican Entomological Society for publishing it.



================================================================================

19661 Carpenter - Protorthoptera and Orthopiera 49 lacunae received no tracheae. Eventually, the epidermal cells lining [he lacunae, including those without tracheae, secreted the cuticular materials which finally formed the veins. The obvious conclusion from this investigation was that the tracheae did not determine the positions of the veins. What Comstock and Needham had observed was the entrance of the tracheae into the wing pad, followed by vein formation, which ultimately closely resembled the tracheal pattern. What they did not see was that the blood lacunae, along which the veins would form, were already blocked out, before the development and extension of the tracheae.
Holds-worth's conclusions have been corroborated by the investiga- tions of Henke ( 1953) and of Leston ( I 962) on the inter-relation- ships of veins and tracheae, demonstrating that the lacunae in wing pads aze the precursors of veins, the tracheae merely occupying the available lacunae. Smart ( 1961) has shown that the cutting of the main tracheae in the wing pad of Periplaneta resulted in degeneration of tracheal branches and in retracheation but with an abnormal pat- tern, which, however, had no effect on the normal venational pattern. His conclusion was that the pattern of tracheation of the nymphal or the pupal wing could not be taken as fundamental in determining the homologies of the veins.* As the situation now stands, the Comstock- Needham method ot determining homologies of veins, which domi- nated investigations of wings for the first half of the present century, must be regarded as a side issue which actually led nowhere. How- ever, it must also be emphasized that many of the conclusions reached by Comstock and Needham, not involving their ontogenetic method, are perfectly valid.
Another approach to the problem of homologies was introduced by Lameere in 1923, as 2. result of his extended and important studies on the Carboniferous insects of Cornmentry, France. Impressed by the regularity of the convexities and concavities, he concluded that there were originally two media veins and two cubitus veins, one of each being convex ( + ) and the other concave ( - ) ; these he termed the media anterior (MA), media posterior (MP), cubitus anterior (CUA) and cubitus posterior (CUP). He believed that some insects had both convex and concave elements, while others had various combinations of one or the other. Support? for his con- " have given this detailed summary of the Comstock-Needham method of determining wing homologies because their conclusions, based on this tech- nique, have become firmly implanted in American entomological literature and in current texts. See, for example, the 1963 edition of Borror and DeLong's "An Introduction to the Study of Insects."



================================================================================

50 Psyche [March
clusions has come from the study of Palaeozoic insects and more primitive groups of living insects with the result that the Lameere view has been generally accepted as a working hypothesis by stu- dents of fossil insects and insect evolution. In this connection, one should recall that Redtenbacher in his original account of vein homologies used the alternation of convexities as part of the evidence for his system of homologies. Unfortunately, the convexity or con- cavity of several veins has been lost in most orders of insects. The subcosta ( - ) , radius ( + ) , radial sector ( - ) , anterior cubitus ( 4- ) ,md posterior cubitus ( - ) tend to retain their topography in mem- branous wings, although virtually all veins in thick tegmina or elytra appear to have lost their topographic positions. The anterior media ( + )
and posterior media ( -) have generally come to lie flat in the wing membrane, except in the palaeopterous orders, where they are distinctly different. As a working hypothesis, I am assuming the presence of both of these veins in the early neopterous stock; there is some evidence from the pattern of these two veins in closely related taxa that they have been retained even in the endopterygote line (see, for example, Carpenter, 1940, Adams, 1958). Histological investigations on the development of convex and concave veins are still needed. Holdsworth included some histological observations in his work previously cited, but his studies were limited to one species. Mayfly wings, treated with caustic potash, separate into their orig- inal membranes, all the convex veins being on the dorsal membrane and all the concave veins on the ventral membrane (Speith, 1932 ; Holdsworth, 1941). Holdsworth noted that, although there was not this sharp difference in Pteronarcys, most of the cuticular material of the convex veins appeared to be formed in the dorsal epidermal layer and most of that of the concave veins in the lower epidermal layer.
It is not improbable that this is generally the case. As noted above, veins have tended to lose the topographic characteristics in tegmina or elytra; and it is possible that a previously concave vein might eventually acquire a convex position secondarily if the tegmen became membranous. However, I regarded the latter occurrence as probably a rare event and consider convexities or concavities of veins as due to the original condition, unless strong evidence exists to the contrary.
In my own work on insect evolution, therefore, I use the follow- ing terminology for wing veins: costa ( + ), subcosta (- ), radius ( + ), radial sector (- ), anterior media ( + ), posterior media ( -) anterior cubitus ( + ), posterior cubitus ( -), and anals ( +, -, or flat). The term postcubitus was suggested by Snodgrass



================================================================================

19661 Carpenter - Protorthoptera and Orthoptera 51 for the first anal of Comstock and Needham; however, I see no reason to make this change especially since the new name would al- most certainly be confused with Lameere's posterior cubitus men- tioned above.
Order Protorthoptera
As noted above, the Palaeozoic orthopteroids present unusual prob- lems in classification. The Blattodea, although part of this phylo- genetic complex, are not included in the present discussion, since they are usually regarded as comprising a distinct order. The Manteodea and Phasmatodea are as yet unknown in Palaeozoic strata. We are therefore concerned in this discussion with the living order Orthop- tera (i.e., Saltatoria) and with a bewildering variety of orthopteroid fossils, some of which appear to be close to the Orthoptera, but others which are suggestive of the Blattodea, Manteodea, Phasmatodea, Plecoptera, or combinations of two 01- more of these groups. Un- fortunately, our knowledge of about four-fifths of these species is restricted to the fore wings or even to only a part of the fore wings. Handlirsch (1906) recognized two main extinct orders in the complex, the Protorthoptera and Protoblattoidea, but found it neces- sary to recognize a third category, "Protorthoptera vel Protoblat- toidea" for the species which he could not clearly assign to one or the other. As more Palaeozoic insects became known, a gradual diminution of the distinctions between the Protorthoptera and Pro- toblattoidea resulted and the number of genera in the "Protorthoptera vel Protoblattoidea" category became nearly as great as the number in the Protoblattoidea itself. In 1937, Martynov suggested the sep- aration of the several non-saltatorial families into a distinct order, Paraplecoptera, leaving in the Protoi-thoptera only the saltatorial forms. More recently, this proposal has been amplified and somewhat altered by Sharov, who has suggested additional differences between the Protorthoptera, Protoblattoidea, and Paraplecoptera. This in- volves the transfer of a few species (Oedischiidae) with well de- veloped jumping hind legs into the true Orthoptera, restricting the Protorthoptera to one family, having an incipient saltatorial modifica- tion of the legs, with the bulk of the Palaeozoic orthopteroid families going into the Paraplecoptera and Protoblattoidea. Before considering Sharov's proposed classification, I wish to dis- cuss certain aspects of the venation of the fore wing of these Orthop- tei-oids, at least those features which involve differences in interpretation. Sc, RI, Rs, CUP and the anals present no difficulties in their homologies, but the media (and to some extent CuA) is a



================================================================================

52 Psyche
[March
different matter. In the orthopteroids, as noted above, the media does not show the clear division into a convex anterior branch and a concave posterior one. It is often deeply forked and the posterior branch may be strongly concave or only slightly concave or even neutral (flat), but I think it can be safely said that there is no wthopteroid known in which the anterior branch of the media is convex. We have no way of knowing, therefore, whether in such cases the entire media consists only of MP (with a flattened anterior branch) or of LIA and MP, with a flattened MA. The only positive criterion by which we can identify a vein in the orthopteroids as homologous with MA of the Palaeoptera is by its convexity -which none have. I think there is enough evidence, however, to justify the probable determination of the anterior branch of M as MA in some families of orthopteroids, but the determination is only a working hyp~thesis.~
Another area of controversy is the relationship between CuA and M. In the majority of the orthopteroids there is some type of con- nection between M and CuA if only a short cross-vein. In others (as Stereopteridae, figures 10-13 of the present paper), CuA curves upwards and fuses with part of M before diverging off as an in- dependent vein. It should be noted that there is marked individual variation in the nature and amount of this coalescence. In others. such as the Blattinopsidae, there is a strongly convex stem of M (see figures 7 and 8 of this paper) which become abruptly flat or concave after the divergence of a short, convex, posterior branch. I think it probable here that the anterior branch of CuA is fused with M from the very base until the point of divergence. A somewhat similar situation appears to occur in the Oedischiidae and related families (these being treated here as true Orthoptera), but I believe rhe homologies are different (see figure 15). The stem of M, in- stead of being markedly convex, is flat or even concave. The short vein which diverges towards CuA is rather weak in the Oedischiidae, althoug,h it may be stronger in other, related families. In this case, 'In my own desciiptive accounts of the Paleozoic orthopteroids I use the designation MA and MP if the posterior branch is definitely concave and the anterior branch flat; if the posterior branch is flat like the anterior one I use the designation M for the entire system; if all branches of the media are concave, I use the designation MP for all. EXPLANATION OF PLATE 4
Lmatofihora typa Sellards.
Photograph of specimen No. 3539, Museum
of Comparative Zoology, showing prothoracic lobes, with hair covering and reticulated pattern, Original.




================================================================================




================================================================================

54 Psyche [March
I consider the divergent vein as a modified cross vein, which in many cf the orthopteroids appears in diverse forms (e.g., Strephocladidae, figures I and 2). It is my opinion, therefore, that the connections between CuA and M are of a diverse nature in the orthopteroids and that these connections have arisen independently many times. Regarding Sharov's proposed classification of the Palaeozoic orthop- teroids, I have previously ( 1954) adopted Zeuner's suggestion (also accepted by Sharov), that the Oedischiidae are true Orthoptera; Sharov has with good reason made a similar inclusion of a few related families (of which the Permelcanidae, figure 18, is a rep- resentative). He then proposed restricting the Protorthoptera to the single family Sthenaropodidae, defining (1960, p. 295) the order as including those orthopteroids with "dorso-ventral flattening of the body, cursorial hind legs, lacking the two rows of spines on the hind margin of the tibia, by the small precostal area lacking the numerous veinlets and by the absence of an undifferentiated concave MA2." This definition I find much too narrow for an order; it might well fit a family - a small one - but certainly not an order. The re- mainder of the oithopteroids which I have previously included in the Pi-otorthoptera, Sharov proposes to divide into the Protoblattodea and the Paraplecoptera. The former order he would restrict to those species having wide coriaceous fore wings, the absence of a clearly defined division of the media stem into two main branches, MA and ^\IF, by large coxae and by general resemblance to Blattodea. In this case, Sharov's characterization seems to be much too broad and generalized. Certainly the coriaceous nature of the fore wings varies greatly within orders (e.g., Orthopteraj ; in some the fore wings are truly men~branous but in others they are definite tegmina or even dytra.
So far as the division of the media into MA and MP is concerned, I question that this is clearly divided in any of the orthop- teroids; as noted above, there is no orthopteroid that has a convex, and therefore, definite, MA. The coxae are known in very few of the species that Sharov would place in the Protoblattodea and, once again, I cannot see this as an ordinal characteristic. The Paraplecop- tera are distinguished by Sharov by the presence of membranous, elongated fore wings, by the clearly defined division of the median into MA and MP and by the general resemblance of the insects to the Plecoptera. On examining the genera which Sharov includes in the Paraplecoptera, as described and figured in the Osnovy (1962 j, I find many families (e.g., Spanioderidae, Probnidae, Strephocladidae, etc.) in which the fore wings are distinctly coriaceous and as rela-



================================================================================

19661 Carpenter - Protorthoptera and Orthoptera 55 tively broad and oval as those of the previous order. The condition of MA and MP has already been commented upon. I can see no justification in Sharov's account for the recognition of the Protorthoptera, Protoblattodea and Paraplecoptera as separate orders, and I propose to place all of these without subgrouping in the order Protorthoptera. Admittedly, the Protorthoptera as thus con- stituted would bs almost: certainly polyphyletic. But it seems to me that Sharov's classification would recognize two polyphyletic orders, (Pi-otoblattodea and Paraplecoptera) with the order Protorthoptera itsel-f so narrowly defined as to include only one family. In all probability, the Palaeozoic Orthopteroids were not evolving just in che direction of the living orders Blattodea, Plecoptera, and Orthop- tera but, as a result of radial evolution, in many directions. Certainly this is what one would expect from the geological record of other groups of animals. The setting up of the three orders Protoblattodea, Paraplecoptera and Protorthoptera would seem to me to conceal what were almost certainly the real evolutionary lines of these insects. Hence, I prefer to group these orthopteroids into one large complex t h e Protorthoptera- until we have enough evidence to indicate what the several lines of evolution have been. I do not believe that we have that now.
I am convinced that Sharov is correct in maintaining that the Lem- matophoridae are not sufficiently different from the Liomopteridae, etc., to justify separation in a distinct order, Protoperlaria. Certainly, as Sharov points out, both fore and hind wings of the Lemmato- phoridae and related families can be distinguished from those of other Protorthoptera only with the greatest difficulty. I cannot agree wtih Sharov! however, in his claim that the paranotal lobes in the Lemmat~pho~idae were continuous and formed a pronotal shield as in Liomopteridae, instead of being independent lobes, as Tillyard and I had described them. Sharov states that his study of the pub- lished photographs in till yard'^ ( I 928) and Carpenter's ( I 935 ) papers shows that the lobes unite in front and behind. Although
photographs are extremely useful in the study of fossils, they are no substitute for the actual specimens.
Till yard's drawing and mine
were based on different specimens and were made several years apart. I have re-examined the material in both the Harvard and Yale col- lections since the publication of Sharov's paper and I cannot agree with the interpretation which he has made from the published photo- graphs. Photographs of the thoracic region of two specimens of Lemmatophora typa Sellards are included here (plates 4 and 5). The first of these shows a specimen which is not quite in a symmetrical



================================================================================




================================================================================

position; it shows especially clearly the form of the individual lobes. The second specimen, which is the one originally figured by Tillyard, shows the thorax in a more symmetrical position. When Tillyard's original photograph was made. plant fragments and other organic debris covered much of the thorax, obscuring the form of the para- notal lobes posteriorly. Subsequently, as shown in the photograph on plate 5, this debris was removed, presumably by Tillyard himself. The paranotal lobes are reddish-brown in color, like the true wings; the plant fragments and the debris are black, so that the two are more distinctive in the actual fossil than is apparent in a black-and- white photograph. In any event, I do agree that these paranota are not sufficient to justify the separation of the Lemmatophoridae from the Protorthoptera.
In the preceding papers in this series, eight families of Protorthop- tera were considered: Lemmatophoridae, Probnidae, Liomopteridae, Chelopteridae, Stereopteridae, Demopteridae, Phenopteridae and Protembiidae. In the present paper three additional families are covered, the Strephocladidae, Blattinopsidae, and Tococladidae, and the Stereopteridae are discussed further, in the light of new material. Family Strephocladidae Martynov
Strephocladidae Martynov, 1938, p. 100.
Fore wing: coriaceous ; precostal area absent; Sc well developed, extending to mid-wins; or beyond, with several to many forked branches; Rs arising before mid-wing; RI extending well towards iipex, with several oblique branches leading to margin beyond Sc; Rs very well developed, with several to many long branches, usually without forks except for the branches in the apical part of the wing; M forked before origin of Rs, the anterior branch often touching Rs briefly or connected to it by a short, stout cross vein; M with sev- eral long branches, usually simple, independent of R basally, often touching CuA briefly or connected to it by a stout cross vein or pos- sibly by an anastornosed branch; Cu independent of M basally; CUP arising near base ; CuA directed longitudinally, giving rise to several