The JapaneseSocietyJapanese Society for Systematics

May, 1978 Acta Phytotax. Geobot. 33

Venuloid Idioblasts in Pteris and Their

Systematic Implications

W. H. WAGNER, Jr.*

w. H. v fft- : i / Je s vvKoeelkse. ng$mHa

In many ways anatomical characters have been relied upon more in pteridophytes

than in spermatophytes in working out taxonomic relationships, especially during the early development of classificatien at the levels of genus and family. Nevertheless, we are still engaged in seeking new characters that can be exploited in solving controversial

problems of relationships and this research continues. In this article I wish to discuss

"false the peculiar veins'' that are found in the brake , Pteris. Such structures have

"spicular ``recurrent [Cfalse "Scheinnerven," been referred to as idioblasts," veins," veins,"

"sclerites," "venzaloid and other terms. As a name fbr this phenomenon I propose idio- blasts." This would cover all structures of various fbrms and homologies which produce

the illusion of tiny veinlets scattered in the interveinal, submarginal, or sinus regiens of

the lamina. They are not attached to the true veins. Ordinarily the venuloid idio-

blasts can be seen with a high-powered hand lens or dissecting microscope with no

difllculty, appearing as narrow Iines or streaks. In dried herbarium specimens they are often slightly raised. Such veinlet-like structures are widelv . scattered in the various groups of ferns, but are most familiar in such families and subfamilies as Marattiaceae

in Marattiales and Hymenophylloideae and in the . In some genera (e.g., 7Vichomanes, Angioptere's) they have been used fbr purposes of species discrimination. The presence of venuloid idioblasts has also been used as a source of evidence for relationships at the higher levels of genus and family. As part of a general study ef venuloid idioblasts in leaves, I wish to describe in this paper some of my observations in the genus Pteris. This is especially appropriate on this occasien because Professor Motozi TAGAwA probably dealt more than any other

individual with the species of Pten-s that display these structures. Indeed he described

several of these taxa, I wish here not only to expand the existing descriptions of the

venuloid idioblasts of Pteris, but also to discuss some of their possible systematic im- plications.

* Department of Botany, University of Michigan, Ann Arbor, MI 48le9, U. S. A. Research supported by NSF DEB75-03550. Taxonemic Applications ofFoliar Characters. I wish to acknowledge Florence S, NVAGNER for her constant assistance in this prQject.

NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics

34 Acta Phytotax. Geobot. Vol. XXIX, Nos. 1-5

Generally speaking thc presence of venuloid idiob]asts in the brake ferns has gone unnoticed by the majority ef authors. BowER (1928) did not mention them in Pteris, nor did CopELAND or OGuRA In connection with Pterz's WALL･, (1947) (1972). .arewilleana "between HoLTTuM (1954) noticed that the veins on both surfaces are irregular short ralsed lines, commonly two rows between each pair of veins." In his synopsis of the genus Pteris in Japan, Ryukyu, and Taiwan, Wang-Cheung SmEH (1966) used what he termed

[[false-veins ``Subsect. in mesophyll" te key out 2. Cadieri." He also adopted the

"many ``few contrast false-veins" vs. faIse-veins'i as the primary division in his key to the seven taxa in this section, name]y P. -ptanzatensis (TAGAwA) TAGAwA, P. angustipinna

muttijitla KuRATA, P. WALL., TAGAwA, P. tyukruensisTAGAwA, P. PoiR.,P. kidoi .arevilteana and P. cadieri CHRisT, A modification of this treatment was also used by SHiEH (1975)

in the Flora of Taiw, an. , Description. The fbllowing descriptions of the venuloid idioblasts of Pterzs are based primarily on the species that represent the two extreme types of frond structure, namely P. multifdd with linear, entire major segments, and P. grevilleana with pectinate or deeply lobed major segments. The specimens used are in the University of Michigan

Herbarium*. Frond scgments were cleared in sodium hydroxide aqueeus solution, and

stained with tannic acid and ferric chloride. It became evident early in this study that the idiohlasts actually do not occur in the mesophyll, as has been reported; instead they are scattered in the upPer and lower epi-

dcrmis, They are, however, somewhat overlapped by adjacent epidermal cells, as shown

in Figure 1. The cross-sectional diameters of the epidermal cells and the idioblasts are

sma}ler by one-halfto one-fifth than the true mesophyll cells. There are usually two to four layers of inesophyll parenchyma cells visible in section, and they are rounded and lack projections.

'

Fig. 1. Cross-sectien of a leaf segrnent of Pteris gretiUeana showing position of venuloid idioblasts in the upper and lower epidermises,

* Volcher specimEns'for anatomical preparations. (In University ofMichigan Herbarium unless otherwise indicated). Pterisgrevilteana-CmNA: Tonkin, Balansa 1978; western Tonkin, Bonin May 1B87. BoRNEo: Sarawak, Penrisson Mt., Brooks 14 in May 191O. P. cadien'-AssAM: Maknon Forest, Mann s.n., CopELA)m Herb. 13777; CHiNA: Kwantung, Tsing Wan Shan ["P. Iauii CHiNG sp. nov.'i], Lau 21S4; Hainan, Chim Fung Mt., Lau 5469; Fokien Prov., M55, CopELAND Herb. 7575; FoRMesA, Chilung, Simoaawa 558. P. multijida-A?AN: Hondo: YamamotQ in Settsu, 71)gasi 709; KoREA: Cheju-Do, C7itLng in-eno 1713; CHiNA: Kwantung, Fan Shiu Shan, Lau 2758; Man Chi Shan, 71sang 264I4; Shantung, Huang Shan, UNiTED STATEs oF AMERicA: AIabama, Mobile, Dttkes in 1904 Fern Sec. Herb.). (]eiao2894; ' (Amer. P. mnttijidu × semipinnata?-OKiNAwA: Nuha-oyama, Amano 6738. P, wdouensis: RyuKyu, Amami-isl.: Naze-city, Slerizawa 11765.

NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics

May, 1978 Acta Phytotax. Geobot. 35 ' ' -t -t ttttt- - t

The idioblasts are narrow, approximately 20-25 microns in diameter, and the

secondary walls are extremely thick, the lumen being only 6-9 microns wide. Their lengths vary from O.! to 1.2 (rarely more) mm., and most of them are between O.5 and

1.0 mm. Thc longest idioblasts tend to be those located approximately midway between

the costa and the margin of the segment, and the shortest are fbund near the cQsta, and especially along the margin. With few exceptions, they taper at both ends to pointed

Fig, 2. View ofcleared segment of P. gmoilleana under Nomarski interference contrast lighting showing the similar refraction of true veins and false veins. ' tips. Fertile segments show the idioblasts running along the adaxial surface and the

rolled margin above the sorus, but not in the false indusium itsel£ The idioblasts have

a slightly scalloped appearance as seen under the dissecting rnicroscope and in clearings

duq to the greater er lesser overlap of the numerous adjacent epidermal cells, The

refractive characteristics of the false veins under Nomarski interference contrast illumi-

nation are similar to those ef the true veins (Fig. 2). Usualiy the idioblasts are nearly straight or only slightly curved. Strongly curved

or angular ones are most often fbund close tQ the costa in the angle of vein departure.

Also, along the segment margins many of the idioblasts curve upward toward the apex

of the segment. Although there are occasional individuals with more or Iess knobbed

tips, only rarely do the idioblasts branch. On the abaxial surfaces one sometimes finds

examplcs of forking where the tip comes into contact with the auxiliary cells of the

stomates.

Pteris grevilteana (Fig. 2) displays the most striking development of spicular cells, and these are immediately evident to the eye, even with a low powered hand lens. They are very numerous and crowded, and a typical interveinal area including both laminar surfaces totals roughly between 20 and 40 idioblasts (Fig. 2). [I]hey occur throughout the interveinal areas and there is no special tendency to cluster along the true veins.

Numerous examples ofjoined pairs or triplets of idioblasts can be found. They are

NII-Electronic Library Service The JapaneseSocletyJapanese Society forforPlantSystematlcs Plant Systematics

36 Acta Phytotax. Geobot. Vol. XXIX, Nos. 1-5

Fig. 3. Distribution of venuloid idioblasts in Pteris multti7de showing their tendeney to d'u'hte'r drotu'iui the truc vcins.

joined in various ways. Although they may sometimes be contiguous for nearly or quite their fu11 length, they are commonly attached near tlteir tips. Carefu1 examination

often is needed to estab]ish that what appears to be a single unusually long idioblast is

in reality two or three connected along their tips. Pteris multijda (Fig. 3, 4) differs in having only relatively few idioblasts, usually O-5, rarely more, per interveinal area. These tcnd to cluster along the true veins, and this may make it possible to overlook them in a casual examination. Joined idiob]asts are rare in P. muttifda, except along the midrib, and curved examples are also not so commen as in I'. grewilleana. However, short individuals, as in the latter, do tend to occur mainly along the costa and along the margin. P. ryulpuensis (Fig. 5) has the fewest idioblasts of any J of the taxa I studied. wM-v#i/･,/////T,g/.i･.･ix,'g,f' ;･" " ,e/g･/itwh,i,s,ny-i,,/g'i"zee･g.ma,l,/sew.. . ..gts/a･-#'ggli//r,\･isii:.'/･"'ps ' ci ww th: ' ' ' -' ca･srm.t,. -g' - de -i,//S,/s // .k.if/T,,/::gif\ nc . gs/i･g i/i'#./g:x:-g･eelk/4'},g･i sigS･,Y･･ist/j?{,//t-/r.lg' ss･/:ii//Y.l･g

'ik' grf,IillS･/l/r,//'//1,g･i･',,rr･il",l:11ii-l':111?l･,i/11/l,lti･l･Li,l'tlSl'･li#,I,//lilllli-,i･,-i,ii,.,E,11e..'ijll'l.,$',//tsi.l//f-,fl/Tit"Tl,g,kll,lw" mekg:msznys-&im' -g･-i -t/tde'-'i,･: ,k g'l'i s/k ti･ i･ i- iLe,l ;, L, l. ;, :S' le-.;e i/ i/t" :, k'･ - ' i,･:k' 'X"'be",･X- '' ' ' ･= :ee-･!mitmp.{.AtMeein'#'tk -: ' -ali. ･ ･･,,wa,,lktge･il.l.･ -,iillli,li,/in :･pa'.l. me. li,lfli,,lk/l/tt.i tmeg ii,\i･ii//V/I,/ktk .g/tT,:liilli x 'ww'' ffk..'.asgem,scwnm, . ..,.M.,..:.'-w.. twdi.ee.m. k M..'::eenm:.'.'ul-epmembww:m',

Fig. 4. Detail of venuloid idioblasts in Pteris multijitla showing their relation to the veins and to the sorus.

Plants of the general description ofP. cadien-, i.e., those with more or less dimorphous fronds that show irregular combinations of the characters of P. grevilleana and P. multijida,

NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics

May, 1978 Acta Phytotax. Geobot. 37

Fig. 5. Pteris rvttig,uensis shewing interveinal area with very sparse venuleids.

appear to have intermediate expressions ofthe idioblasts (Fig. 6). They occur in numbers ranging from 10-30 in the interveinal areas. However some specimens rescmbling P. cadieri (e.g., Amano 6738 from Okinawa, which appears to be a hybrid of P. semipinnata L. with P. multijicla) have fewer idioblasts per interveinal area. Discussion. It is intriguing to speculate on the origin of the venuloid idioblasts of Pteris. They are apparentiy very rare in the genus as a whole, being known in only

a few taxa of the western Pacific and southeastern Asia, One hypothesis might be

that these structures arose entirely de noeve from ordinary epidermal cells. Another

hypothesis, and the one which I would favor, is that they had their erigins in the more

or less elongate epidermal cells above and below the true veins. These cells tend to have

somewhat straighter walls and to have greater length than the majority of adjacent

epidermal cells in the interveinal regions. According to this hypothesis these epidermal cells located over the veins provided some sort of mechanical strengthening and thcy became exaggerated to the present form of short sclerenchymatous fibers. Subsequent evolution would call for the progressive diflUsion of such cells into the epidermises of the

interveinal regions until they had an essentia]ly uniform distribution.

Both extremes described above are fbund: In P. multtiitla we see the conditien ef few idioblasts, these tending to cluster closely along the true veins, In P. grevitteana, in contrast, we see numerous idioblasts, and these are thoroughly spread throughout the interveinal areas. If this hypothesis is true, then, P. multifcla displays the primitive condition, and P. grevilleana thc most specialized. Other taxa, such as P. cadieri, are intermediate.

I do not doubt that P. grevilleana and P. muttoficla are related, in spite of the seemingly radical differences in their frond structure. The Sino-Japanese area has been cited by

WALKER (1962) as one ef the centers of speciation in the genus Pteris. We are familiar with several cases ofhybridization between species with linear, unlobed pinnae and with

NII-Electronic Library Service The JapaneseSocietyJapanese Society forforPlant Plant Systematics

38 Acta Phytotax. Geobot. Vol. XXIX, Nos. 1-5

Fig. 6, Ptert's cadieri interveinal areas showing numbers of venuloids intermediate 1..L...." n n".-:11.n-n nrA ....T"C.TA VCLYVt;t;tt l . gTeule-aunLt tLllu ift"e['usuu.

deeply lobed pinnae. In Hawaii occur obvious intermediates betwe'en P. cretica L. and

P. irregutaris KAuLF. knewn as P. × hillebranth CopEL. which not only appear to be hybrids ' but which have become established as an apogamous species (WAGNER, !974). In the case ofP. multiaurita AG. and P. euacfriaun'ta RETz., T. WALKER (1958) was able to demon- strate an almost complete hybrid series between them. In their hybrid swarm sexual

reproduction prevails and there is no less of ability for the chromosomes to pair. The plant known as P. otaria BEDD. is an especially common hybrid form. The striking point abeut these and other known or suspected hybrids in Pteris that involve sharp differ- ences in leafblade structure is that the compromise between parental extremes is strongly ' irregu}ar (c£ WAGNER, 1962; WALKER, 1962, fig. 2; WAGNER, 1974, Fig. 6). I would like to suggest that P, cadieri is analogeus to P, × hillebrandii and P. × otaria in that it arose by hybridization and that it displays irregular frond structure. SHiEH

"fronds (1975) describes P. cadieri as having dimorphous, sterjle frends.,. with one or two pairs of lateral pinnae, pinnae pectinate, pinnatifid, or irregularly producing seg- ments, in the simplest form the terminal pinna simple and linear." The development of venuloid idieblasts in P. cadien' is intermediate between that of P. grevilleana and P. of L6vE, L6vE, and multijida (Fig. 6). According to the cytotaxonomical compilation Picm SERMoLu (1977) both P. grevilleana and P. multijida are tetraploids with 2n=116.

==`87' and P. as MiTm (1975) also reports P. cadieri as being n apogamous, rpulpuensis n=58. Further work is desirable befbre we can evaluate the significance of these ob-

servatlons. I suspect that hybridization has taken place between P. greviUeana and P. multifda (and) or other species with linear pinnae to produce ofthe description ofP. cadieri. Several different species may be involved, thus making a complex picture. To NAKAiKE's (197e) list of hybrid Pteris in Japan wc may have to add others not heretofore suspected of having had hybrid origin, not only P. cadien' but P, kidoi, P. rvordyuensis, and possibly others.

NII-Electronic Library Service The JapaneseSocietyJapanese Society for Plant Systematics

May, 1978 Acta Phytotax. Geobot. 39 .. . .-. t tt tttt .-......

The mqjor significance of the unusual anatomical character discussed in this paper involves the systematic relationships ef the genus Pteris-a ]ong undecided question. For many years it was a tradition to associate the brakes with the worldwide fern genus, Pteridium, mainly because of the gross appearance of the cutting of the fronds and the marginal coenosori, However, in recent years there has been a tendency to question this, and indeed to separate the genus Pteris and its allies wholly from Pteridium, placing the former with the adiantoid-cheilanthoid ferns, and the latter with the denn- staedtioid-cyatheoid ferns. One major impetus for this change came from the discovery

by cytologists that the Pteris complex is characterized by the base chromosome number

of x=29, iike typical adiantoid-cheilanthoid ferns with x=29 or 30, whereas Pteridizam

"The has a base number ofx=26. MicKEL has written (1973) that genus Pteris presents a problem that is not easy to solve. I would place it with the Adiantaceae on the basis

of its reflexed marginal false indusium, chromosome number of n=29, and commonly pedate architecture. On the other hand it does have in some species polycyclic stelcs

as in Dennstaedeia, and epipetiolar branches, but the branches are entirely abaxial rather

than Iateral on the petiole.'' I submit that the presence of epidermal idioblasts of the type described here, albeit rare in the genus as a whole, may have a systematic bearing on the classification of Pteris at the familial level. It is well known that similar structures are found in the genus Adiantum, although again in a minority of the species. Also, those epiphytic ferns pre-

sumably derived from ancestral adiantoicl-cheilanthoid ferns, the Vittarioideae, have

similar idioblasts in the majority of genera and species. In my survey of venuloid idio- blasts in fems in general I have found no other group beside the adiantoids, pteroids,

and vittarioids where venuloids of this description occur. All other venuloid idioblasts in ferns have strikingly difllerent structure, and appear to be examples of convergent rather th'an parallel evolution. Those of the filmy ferns, Hymenophylloideae, for example, apparently involve changes and detachments of whole

veins, and those of the giant ferns, Marattiaceae, modifications of the true mesophyll.

Both are multicellular false veins, and neither can be homologized with the isolated

epidermal idioblasts described here, On the other hand, the false veinlets ofthe adiantoid

and vittarioid ferns are so much alike that they are certainly homologous. This is not

to say that they are necessarily directly derived from one another. They more probably represent parallel evolution in three separate lines. Thus it is the predisposition to

the evolution of such structures that betrays systematic aMnity.

Literature Cited

BowER, F. O. 1928. TlieFerns (Filicales),' Vol. III. The Leptosperangiate Ferns. Cambridge U, Press. P. vili, 306. CopELAND, E. B. 1947, Genera Filicurn. Chronica Botanica, Waltham, Mass. P. 247. HoLTTuM, R. E. 1954. Ferns ofMalaya, Flera ofMalaya, Vel. II. Governrnent Printing OMce, Singapore. P. iv., 622, appendix.

NII-Electronic Library Service The Japanese SocietySooiety for Plant SystematicsSystematios

XXIX ． − 40 Acta Phytotax． Gcobot ． Vo1 ． ， No5 1 5

and R ． E ． G ． PIcHI SERMoLLI ．1977， Cytotaxonomical Atlas of the Pteridophyta． LOvE ， Askell， Doris L6vE ，

Cra er Hirschberg P xv 五i 98 Straussand ， ．． β ． ． The classification and of the Dennstaedtiaceae ・Bot ・ ・ MIcKEL ， J．T ．1973 phylogenetic position Jour

1．innaean Soc ．67 suppL l： 135−144， ， Bu1L NipPon Dental Gollege M π ul ， Kunio ．1975． chromosome numbers of Japanescμ eridophytes ． ，

4 − General Edu ぐ ation ， No ． ： 221 27L of ま ferns in ． Printing center colleges of Agric・and NAKAIKE ， Toshiyuki ．1970．Index hybr d Japan ， Engineering， Univ． of Tokyo ， P ，62． OGuRA ， Yudzuru ．1972． Comparative Anatomy of Vegetative Organs Qf the Pteridophytes． Revised ed ． 置 ． 2．Encyclopcd 三a of Plant Anatomy ． Gebruder Borntraegcr． p ． vi ，501

− 飴 rn ’σ埋5 and Talwan ・ Bot ・Mag ・ sHIEH，wang chcung ．1966． A synopsis ofthe genus P inJapan， Ryukyu ， − Tokyo 79： 283 293．

− ． voL 1．Epoch Publ ． ． In Li．且 ui lin et aL Flora of Taiwan ， S 田 EH Wang Cheung 1975． ， ， ， ， ， − Co ．，Taipei， Taiwan ． P．280 301 ． ．Phytomorphology 12 ； wAGNER ， w ． H ．， Jr．1962． Irregular morphological develoP皿 ent in hybrid ferns n 弓− t A ∩ OJ Iuu ． ，61 ： WAGNER W 且 ．1974． Structure of spores in relation to fern ． Ann ． MissouriBot ．Gard ， ， Jr phylogeny 332−353 ．

且 some species of L ． Evolution 12 ： 82−92． WALKER ， Trevor G ．1957． ybridization in PtertS − WA 正 KER Trevor G ．1962． Cytology hl the f壱rn P ’爾 ∫ LEvolution ユ6 ：27 43 ． ， and evolution genus

て き た し 新 しい 形 質 摘 要 シ ダ 植物 の 分 類 形質 と し て い ろ い ろ の 形 態 学 的形質 が 採 用 さ れ ，

つ い こ で ノ モ ト ソ ウ の 偽 脈 に い て 論 じ る が ・ こ を 求 め る 努 力 も続 け られ て る。 の 論 文 は ， イ 属 て い る の で 以 外 の こ の 種 の の 種 の 構 造 は こ れ ま で 様 々 の 呼 び 名 で 呼 ば れ て お り， 混 乱 し ， 脈 構

型 造 を ひ っ くる め て 「脈 状異 型 細 胞 venuloid idioblasts」 と呼 ぶ こ とを 提 唱 し た い 。脈 状異 細

シ シ ン シ の い の ビ ン イ コ ケ シ ノ ブ亜 科 ， ラ 亜 胞 は ダ 植物 い ろ ろ 群 に み られ る が ， リ ュ ウ タ 科 ，

科 に 普 通 の もの で あ る 。

モ の の よ さ れ て い る よ う に 葉 肉 の 組 織 の う ち に イ ノ ト ソ ウ 属 も で は ， 脈 状異型 細胞 は く記 載 は く 径 20〜25 ミ ク ロ ン あ る の で は な くて ， 上 下 の 表 皮組 織 内 に 散在 して い る 。異 型 細 胞 幅狭 ，

〜 〜 0． 1． の で る 。 通 常 端 は で 壁 は 厚 い 。長 さ は 0．1 1．2mm で ， 多 くの も の は 5 Omm 範 囲 あ 両

の で っ た り枝 分 か れ した りす る は 特 狭 くな り， 先 端 は 尖 る 。 表 面 観 で は 異 型 細 胞 は 直線 的 ， 曲

殊 な 場 合 な どだ け で あ る 。

シ さ る イ ノ モ ト ソ ウ で は異型 細 胞 ア ガ タ シ ダ に は こ の 種 の 異型 細 胞 が 多 く， 容 易 に 観 察 れ 。

の の ュ キ ュ イ ノ モ ト ソ ウ で は ， が 少 な く， 脈 近 くに 集 まる 傾 向 が あ る で 見落 しや す い 。 リ ウ ウ

バ マ こ べ っ カ ワ ア ク サ シ ダ で は の 形 質 筆 者 が 調 た も の の う ち で ， 異 型 細胞 は 最 も少 な か た 。 リ に も大 き な変 異 が 認 め られ た 。 の そ の に つ イ ノ モ ト ソ ウ 属 の 脈状異型 細胞は ア ジ ァ 東 部 の 種 に だ けみ られ る も で あ り， 起源

つ て き の で は な い か と 考 い て もい く か の 仮設 が立 て られ る が ， 筆 者 は表皮 細 胞か ら変 形 し た も

い こ の の つ い て よ な 型 胞 を もつ ホ ウ ラ イ シ ダ属 や シ シ ラ ン え た 。 ま た ， 属 類縁 に ， 冏 じ う 異 細

シ ・イ ノ モ ト ソ ウ ・シ シ ラ ン 群 以 外 の も の の 脈 亜 科 と の 類 似 に も注 目 した い 。 ホ ゥ ラ イ ダ 群 群

3 の つ は 互 に 来 し 合 状 異 型 細 胞 は構 造 上異 っ た もの で ， 相 似 の 形 質 で あ る 。 上 記 群 も 形 質 相 由

こ の に し て こ れ ら 3 群 っ た と い う よ り は 同 じ よ う に形 成 さ れ て きた も の だ ろ うが ， 形 質 を 指 標

の 関係 を 考 え る こ と は 強 ち無 理 と は い え な い だ ろ う 。

一 NII-ElectronicN 工 工 Eleotronio Library Service