THE TAXONOMIC POSITION OF THE PSILOTALES IN THE LIGHT OF OUR KNOWLEDGE OF DEVONIAN PLANT LIFE* F. P. JONKER Laboratory of Palaeobotany and Palynology, Utrecht, The Netherlands ABSTRACT exclusively Devonian Psilophytales and of the latter the Rhyniaceae are regarded The order Psilotales, containing the two recent genera Psilatum and Tnzesiptel'is only, is usually usually as the closest relatives. Among classified by taxonomists and palaeo botanists those who take this st?nd I mention G. M. in the phylum Psiiophyta. This concept is, Smith (1955), Magdefra.u in Strasburger in spite of the synangia, based on the primitive (1971) and Lcmoigne (1968c). The concept general appearance reminding one of that of the of the last mentioned author will be dis• Devonian Rhyniaceae. Numerous attempts have been made to derive both genera, including their cussed further on. Darrah (1960) states synangia, from either the Rhvniaceae or from other that the status of the putative primitive Psilophyta. These attempts' sometimes led to the nature of the plant body in the Psilotaceae acceptance of a series of missing Jinks but this concept is, however, not supported bv palaeo· is controversial although there is a strong botanical data. tendency to accept the family as a pel'sis• In the opinion of the present reporter the order tent remnant of the Psilopsida. Andrews has kept a primitive general appearance of its (1961) states that Psilatum has been reaarded Devonian ancestors but in its synangia and in its monolete spores it is more advanced. by some botanists as a very simpl~ land In its anatomy, its microphyllous leaves, and in its plant, possibly a very ancient type that gametophyte perhaps, it shows more affiinity to the has managed to survive for several hundreds phylum Lycopodiophyta. The present reporter of millions of years. Chadefaud (1950) clssiflCs the two genera as a separate order in the isosporous Lycopodiophyta. In his opinion the even regarded them as ancestral to the order is to be derived from the Protolepido• Bryophytes. Others view its simple organi• dendraceae. zation as a result of degeneration from a Available palaeobotanical data in the Devonian more complex ancestor. Thus the former are in support of this concept. The present author has to admit, however, that between the Devonian viewpoint regards it as primitive and the and recent times no palaeobotanical or palyno• latter as reduced. Nothing is known, how• logical data are known to support his concept. ever, of fossil records and there are no related modern plants which offer much aid in INTRODUCTION settling the problem. There are, however, quite a number of objections, even scruples, to an attribution without more to tne Psilophyta to which and twoTmesiptcrisrecent gennaBernh. PsilatumhaVE duringSw. THEa long time already, ?,ttracted the phylum further only Devonian fossils would interests of plant taxonomists and phylo• belong. Of these objections I now mention genists. Eames ('1936) consideled them the already one, namely thE' IE'aves, Psilatum most primitive vascular plants. Tne pri• has very small, spinelike, microphyllous mitive geneml appearance of, especially, leaves without a midvein. The leaves of Psilatum reminds of the Middle-Devonian Tmesipteris are considerably larger and Rhyniaceae: tIle genera Rhynia Kidst. et provided with a mid vein. For this reason Lang and H arneaphytan Bargh. et Darrah, Lotsy (1909) regarded the Psilatum leaves by the green stiff, at first sight almost as reduced, connected with its xerophytic leafless, strictly dichotonous aeri,,>.!shoots and semisaprophytic life. arising from a rootless creeping rhizome. The upper parts of the shoots of both The two genJra, together forming the Psilatum and Tmesipteris are provided, family Psilotaceae "md the order Psilot?.1es, moreover, with bifurcate leaves of the have accordingly been classified by most same size, supporting a very short stalk authors in the Psilophyta, together with the which bea.rs a trilocular synangium in *Contributed to the Paheobotanical Conference, Birbal Sahni Institute of Palaeobotany Silver Jubilee, December 1971. 33 34 THE PALAEOBOTANIST Psilotum and a bilocular synangium in fundamentally verticillate in organization T11lesipteris. This led, as no fossils are and this might point to a relationship with available, Professor Lam (1948) to his well• Equisetophyta, for instance Sphenophyllales. known ingenious, theoretical-phylogenetic, As neither Psilotum nor Tmesipteris are but highly hypothetical derivation from articulate or show a verticillate organiza• the Rhyniaceae with their terminal (stachyo• tion in their aerial shoots, I reject this sporous) sporangia. concept, the more as other conformities In this way he defended and tried to with Sphenophyllales are scarcely evident. save the attribution to the Psilophyta In later papers Salmi (1923b; 1925) without taking into account other features brought forward anatomical evidence to which contradict such an attribution as bring the Psilotales nearer to the Devonian expounded below. Needless to say that genus Asteroxylon which, in his opinion, Lam's concept lacks every palaeobotanical supports the old view that the Psilotales basis. Lam, consequently, rejects t11eterm are related to the Lycopods. sporophylls for the bifurcate leaf-like organs In 1930 Campbell defended an attributiOn supporting the synangia. Other features to the Lycopodiophyta and the same concept which might be objections to an attribution was held by Zimmermann (1959). TIle to Psilophyta we find, besides in the le2.ves possible relationship to Lycopodiophyta was and the sporangia, in the anatomy and, discussed by a number of authors without perhaps, in the gametophytes. Next to t2.king a definite standpoint with respect the general appear2.nce the 2.bsence of roots to that. In this paper I am defending an is actually the only fe2.ture that is in favour attribution, as a separate order, to the iso• of an attribution to the Psilophytes. sporous Lycopodiophyta and in the following For these reasons a number of authors I will discuss my concept in comparison defend the attribution to a separate phylum with arguments against a relationship to of Pteridophytes or, if they prefer to con• Psilophyta- Rhyniaceae. sider the Pteridophytes a phylum alto• gether, to a separate classis. Such an is.o• ANATOMY lated taxonomic position h<l.sbeen defended by a number of authon, who, however, do Psilophytes always have a haplostele in not take account of fossil records. Palaeo• their rootstocks and aerial shoots. Primi• botanists who defend an isolated position tive Lycopodiophytes have, in their aerial are, for instance, Gothan and Weyland snoots, an actinostele of which in more (1964). A different position has been taken advanced representatives the xylem either up by Lotsy (1909) who, on the basis of the may divide into vertical plates (plectostele) presence of a vascular bundle in the sporan• as in certain Lycopodium species, or become gium stalk, decided to a relationship with tubular and filled with a central medula. the Sphenophyllales, a view shared more Those siphonostelic conditions occur, for or less by Scott. instance, in a number of Lepidodmdrales. Sahni (1923a) regarded the sporangio• The complicate anatomy of Selaginellales phore of the Psilotaceae, on the basis of and Isoetales is left out of consideration observed abnormalities, as consisting of here. The roots often show in the above axes bearing alternate whorls of sterile mentioned actinostelic, plectostelic, and si• lobes and sporangia. This concept was phonostelic groups a haplostelic anatomy. shared by Bierhorst (1956) who interpreted The aerial shoots of Psilotum are actinos• the fertile appendage (i.e. the stalked synan• telic in principle. The hollow xylem body gium with its subtending bifid or trifid encloses initially a parenchymatous medulla bract) of the Psilot2.ceae as bearing a verticil which in the older parts becomes scleren• of sporangia (two or three remaining) fused chymatous. According to Bierhorst (1954b) to a central axis, and a verticil of sterile only rhizomes greater than 1 mm in dia• appendages (two or more remaining). In meter usually possess a complete stele. P. nudum and P. complanatum the vascular Tmesipteris also has a siphonostelic vascular trace which enters the base of the sporangia• cylinder with a central parenchymatous bearing portion of the fertile appendage, pith. These facts point into the direction branches into three parts. Bierhorst, conse• of the Psilotales showing a rather primitive quently, re-introduced the hypothesis th2.t Lycopoc1iophytic anatomy whiCI1,110wever, f61rtile appendages in the Psilotaceae are presents a more advanced stage than the JONKER- THE TAXONOl\IIC POSITION OF THE PSILOTALES ' 35 • most primitive Lycopodiophytic anatomy A still more recent view was started by which is characterized by a massive acti• Lemoigne (1968a, b; 19690.) who found nostelic protostcle without pith. There are archegonia of the Pteridophyte type sunken no points of contact to tne llaplostele of in semiglobular tisme bodies on the stems the Rhyniaceae. of Rhynia gwynne-vaughanii and never on any part of Rhynia major. GAMETOPHYTE He also published still vague indications for the presence of antheridia in small The gametophyte of Tmesipteris ha.s been warts on the aerial shoots of R. gwynne• studied extensively by Lawson (1917), by vaughanii (1969b). As he also states that Holloway (1917, 1921), and by Bierhorst the terminal sporangia are only known in (1953,