Reticulate Evolution in the Appalachian Aspleniums Warren H

Reticulate Evolution in the Appalachian Aspleniums

Warren H. Wagner, Jr.

Evolution, Vol. 8, No. 2. (Jun., 1954), pp. 103-118.

Stable URL: http://links.jstor.org/sici?sici=0014-3820%28195406%298%3A2%3C103%3AREITAA%3E2.0.CO%3B2-Q

Evolution is currently published by Society for the Study of Evolution.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/ssevol.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Fri Feb 8 10:53:51 2008 RETICULATE EVOLUTION IN THE APPALACHIAN ASPLENIUMS

Received, October 20, 1953

point species and comparisolis of them with their various intermediates suggest To he able to postulate the course of that within the complex totalling 11 de- evolution in a group of organisnls on the scribed entities, the real extremes are only basis of indirect evidence is one of the three in number. These three species are goals of phylogenetic research, since in A. z?contc~nunt,A, plntyneuron, and A. Inany groups direct evidence is difficult l-lti:ophyllt~71~.A11 the remaining taxa. or impossible to obtain. Occasionally a comn~onlytreated as species or hybrids- worker may be fortunate enough to con- including the familiar A. pilznntifidzwl and firm his indirect and comparative tech- A. b~ndleyi-lie someri~herebetween these niques by direct evidence, such as the dis- three extremes in their morphology. covery of a postulated ancestral type, or ' Anlong students of the Filici~leaeit has the production of a postulated form by been rather conventio~lal to consider as hybridization experiments. In the pres- hybrids only those intertnediate forms ent study of the evolution of the Appala- ~vl~ichoccur as single, sterile plants with chian Aspleniums, a complex group of obvious parents growing nearby. But small ferns of the eastern United States, with the increasing recognition of allopoly- the methods used have been indirect ones, ploidy as a factor in species forination in morphological, anatomical, and cytological. plants, sterility per se is 110 longer the sole But in future years it is to he expected signpost of hybrid origin Indeed, the that experimental proof of the conclusions best-known fern hybrid. Asjlenitl~webe- of this study will be forthcoming. The noides (A. plntynez~ronx A rhizophyl- indirect steps used as evidence of reticu- lt~i~z),though usually sterile, occurs as a late evolution in the Appalachian Asple- fertile for111 in one large population in niums are thus subject to valiclatlo~l by Alabama. It has not, however, been ex- direct tests. anlined cytologically until the present 1tTherry(1925, 1936) pointed out that study (IVagner, 1953), nor have any the Appalachian Aspleniutns "form a ser- others of the Appalachian Aspleniu~ms ies showing intermediates between cer- been so investigated. The aim of the pres- tain long-recognized species." His basic end-point species were five in number: ent work has been not only to clarify the Asp2eniz~nz pinnatifidztnz (lobed spleen- cytological picture of the group, but also wort), A. nzo~ttnnzt~lz(mountain spleen- to use any other available illdirect ap- proaches to its evolutionary history. Thus IT ort), A. bl-ndleyi (cliff spleen~vort). A. fllnt~we?~l-on(ebony or brownsteln spleen- certain lnorphological and anatomical fea- wort), and A. hi,-oplzyllzt~~z(the ~valk- tures were found valuable in interpreting ing-fern, usually treated as Cn~~tptosortts the intermediate for~ns. The evidence rlzi,-oplzyllz~sbut for sake of sinlplicity re- now itldicates that n complex of eight cle- tained in Asplenit~tnhere). Detailed stud- scribed entities, A. ebenoides, A. pinna- ies of the morphology of these five end- tifidtiiqc, A. t~*?~dellii,'4. ke?ztz~clziense,A. gvnvesii, A. brndleyi, A. brndleyi x A. 1 This study \\as aided by a Faculty Research Grant from the Horace H. Raclihatll School of platyneui,on, and A. brc~dleyix A. ?%on- Graduate Studies. fnnlcnc, has arisen as a result of hybridiza-

~VOI~~TIOX8: 103-118. June, 1954. 103 104 WARREN H. WAGNER, JR.

tion between three ancient and original blades, (2) dark-brown color of the leaf species. axis present only at the base of the flat- The three basic species are entirely tened petiole, (3) "glossy" upper leaf distinctive, whereas the remaining taxa surface, resulting from the markedly elon- constitute a difficult group, as shown by gate, practically straight-walled form of diverse herbarium identifications and the epidermal cells, and (4) occurrence taxonomic interpretations. Aspleniuvn strictly in acid-rock crevices. The closest montanum (fig. 1, RI) shows notably the relative of A. nzontanuvn appears to be the foIlowing peculiarities : (1) triangular, wide-ranging A. adiantuwz-nigruvn of the long-petiolate, 2-3-times dissected leaf western United States, Africa, Eurasia,

FIG.1. Concept of relationships in the Appalachian Aspleniums. R. Asplenizhm vhizophyllum; P. A. platyneuron; M. A. monta~tum;PR. A. ebenoides; RM. A. pilz~tatifidzrm;PM. A. bradleyi (frond on left, Frederick Co., Va., Gilbert 250; frond on right, Madison Co., Mo., Russell) ; RMM. A. trtfdellii; RMPM. A. gravesii; PRM. A. ke,ltuckiet~se. RETICULATE EVOLUTIOIV 10.5

TABLEI

Species 2?2 11 Asplerziuin nzontanuriz Harford Co., Md. Pilre Co., Ohio

-4splenium platynez~ron Montgomery Co., Md. Harford Co., Md. Hoclring Co., Ohio Licking Co., Ohio Hardy Co., U'. Va.

.lspleniu~~rrhizophyllz~,iz Ross Co., Ohio - 36 Fairfield Co., Ohio 72 - Shellandoah Co., Virginia 72 36 Alonroe Co., Illdialla - 36

.4spleniuin ebc?zoides Alontgomery Co., Md. 72 (72 univalents) Hale Co., Alabama (Groff) 144 72 Hale Co., Alabama (Logue) 144 72

Aspleniunz pinnati$dz~nz Shenandoah Co., \la. Hardy Co. W. Va. Yorlr Co., Pa. Licking Co., Ohio hlonroe Co., Indiana

Aspleniunz brndleyi Harford Co., Md. Pilce Co.. Ohio

Asplcnz~tnztrudellzi Yorli Co., Pa. 108 (ca. 36 univalents, ca. 36 bivalents) Hardy Co., \V. Va. 108 (ca. 36 univalents, ca. 36 bivalents) Pilre Co., Ohio 108 (ca. 36 univalents, ca. 36 bivalents) (Voucher specimens deposited in the Department of Botany University of Michigan) and Hawaii, a species which also possesses derillal cells oblong, undulate-walled, and the "glossy" upper epidermal cells. A. (4) occurreilce on roclis and soils of a 1l~onfnn1c1lzis confined to the eastern variety of pH reactions. It appears to be United States, occurring from northern illost closely related to A. t~iclzoutanesand Georgia and Tennessee ill the Appalachian A. rpsiliens, and its hybrids with the Region to western R/Iassachusetts. A. foriner have been twice recorded (A.vir- platyneztron (fig. 1, P), the second of the giniculn Mason). A. plntynezr~onhas a basic species, has the follo\ving distinc- very wide range, extending from Texas tions : (1) linear- to oblanceolate-elliptic, and Florida in the south to IYisconsin once-pinnate, short-petiolate leaf blades, and I

istics : (1) triailgular-atteiluate, simple and young leaflets placed in saturated leaf-blades, with extremely long, rooting aqueous solution of paradichlorobenzene tips, (2) leaf axis green except for the for 3 hours to shrink the chromosomes, base of the petiole, (3) upper epidermal then placed ill fixative. Meiotic studies cells as in A. platyneuron, but the venation were made from young sori, fixed directly. pattern anastomosing, and (4) occurrence The fixative used was 4 parts chloroforn~; confined almost esclusively to well-shaded, 3 parts ethyl alcohol; and 1 part glacial moss-covered tops and sides of rocks and acetic acid. The specimens were squashed boulders, the pH reaction of the rock sub- after 24 hours in the solution, and stained stratum predominantly circumneutral. ill aceto-orcein. Its closest relative is A. sibi~icuwzof northeastern Asia. A. rlzizoplz)lllum, like A. platyfzeztron, is a rather commoil fern and has a broad range, extending from Ala- The first of the intermediates to be bama and Georgia north to Minnesota discussed is Asplenium ebenoides. Prob- and Quebec. The i~lorphological char- ably no other fern of the New World acteristics of these three basic species are has attracted so much attentioil as this intricately blended aillong the members hybrid. Wherever its parents, A. platy- of the Appalachian Asplenium complex. neuron and A. ~hizophyllu~~z,grow in close The chroinosome numbers (table 1) of proximity, this intermediate appears with the basic species cannot be used to distin- gratifying regularity, but usually as a guish them since they all have the same solitary plant, and with sufficient rarity to make its discovery a challenge to the -2n = 72, with 36 pairs at meiosis. The number 36 seems to be characteristic of plant collector. Although its hybrid ori- the entire family Aspleniaceae, as was gin was strongly suspected, the finding of found in European species of Asplenium, the large and obviously self-reproducing Ceferach, and Plqillitis by R~lantoil(1950) populatioil of this fern at Havana Glen, and Hawaiian species of Diellin by Wag- Hale Co., Alabama, some eighty years ner (1952). ago, cast doubt on its true nature. Never- theless, the careful experiment of Slossoi~ METI-IODS (1902) produced plants by hybridization Observations upon the different de- of the putative parents which were mor- scribed taxa were made in the field and phologically like A. ebenoides, thus firmly in the herbarium. Some of the entities establishing the hypothesis of its hybrid are, however, exceedingly rare, and field origin. Until the present, however, the observation of these was impossible, al- cptological behavior of the hybrid re- though localities where they had been col- mained unknown, and the capacity of the lected in the past n-ere ~isited. Dried Alabama population to reproduce re- leaflets were cleared in sodium hydroxide mained unexplained. Two possible hy- solutions, stained with tannic-acid and potheses-hereditary obligate apogatny, iron chloride, and mounted on micro- and allopolyploidp-were considered as scope slides in diaphane in order to possible explanations of the fertility of this examine venation patterns and other ana- usually sterile plant. Obligate apogamy, tomical features. Leaf-outline drawings it should be noted, is already known in were made on a tracing table, and epi- two species of Aspleniuuz, A. wtonanthes dermal-cell drawings with a Rausch and and A. resiliens, both of which, though Lomb microprojector. For the cptologi- chiefly subtropical, occur in the eastern cal study, living plants n-ere gro\vn under United States. greenhouse conditions by Mr. TTTalter F. The cptological studies of two living Kleinschmidt at the University of Michi- plants of the usual, sterile form of A. gan Botanical Gardens. Somatic chromo- ebe~toidesfrom Rlontgo~neryCo., Mary- some counts were obtained from crosiers land, reveal that there is complete non- FIG.2. Leaves showing extreme irregularity. A-G. Aspleniunz ebenoides (Asplenium rhizophgllzruz X A. plnt>neliron) : A. Jefferson Co., W. Va., Wagner G. Rawlbtgs 2016; B. Montgomery Co., Md., Pnlvzer 1899; C. Shenandoah Co., Va., Wagner 241; D. Hale Co., Ala., Maxolz G. Pollard; E. Hale Co., Ala. (cultivated, Groff) ; F. Loudoun Co., Va., Gilbert 237; G. Montgotnery Co., Md., WagPter. H. Asplenzu~nine.l-pectatzint (A. rlzizophyllu+nX A. mta- muraria), Adatns Co., Ohio, Bmun. I-M. Asple?ziu+n pitznatifidn+n (A. rhizophyllthm X A. montani~nz): I. Patrick Co., Va., Helle~;J. Jefferson Co., W. Va., Palvzei; I(. Jackson Co., Ill., French; L. Madison Co., Mo., Pi+zkerton:M. Warren Co., Ky., Sadie Price. 108 WARREN H. WAGNER, JR. pairing of chromosomes in meiosis. At homa and Georgia to New Jersey, is rather meiotic metaphase there are visible 72 rare, becoming abundant only locally on univalents, 36 from each of the parents steep rock cliffs. However, in recent times (fig. 7, C). By contrast, the investiga- the possibility of hybrid origin of this fern tions of fertile Alabama specimens of this has received no attention. As early as taxon from two collections reveal that ISSO, D. C. Eaton, who described it as the allopolyploid hypothesis explains the a species, wrote the following: situation : dividing somatic cells show 144 "A. bradleyi varies a good deal in the shape chromosomes (fig. 7, D,), and during the of the fronds and in the degree of incision of first meiotic metaphase 72 normal-appear- the pinnae, the narrower and less divided forms ing chromosome pairs are evident (fig. 7, having some resemblance to A. platynettvott and D) The spores produced are normal, the larger forms looking more like A. mon- but like the stomata, are larger than those tanum, or the European A. lanceolatum. If there could be a hybrid between A. platyneuron of the diploid relatives. and A. montanum, it would be very much like The only certain record of the normal, our plant." sexual life cycle in A. ebenoides is this occurrence in Hale Co., Alabama. Whether Figure 1, PM, shows two of the different vegetative reproduction explains the mul- leaf forms of this taxon to illustrate its tiple plants sometimes found in other variability. With present-day knowledge localities is not known. A. ebenoides may of hybridization in the ferns, and with the produce young plants not only at the tips evidence now at hand, its describer would of the leaf blades, but also at the tips of not have expressed as much doubt as to the pinnae; as many as 8 or 10 young its possible hybrid nature. Morphologi- plants may be produced on a single luxuri- cally it is indeed intermediate between the ant frond. The large fronds of this plant tw; parents suggested by him :The rachis are usually conspicuously irregular in the of A. platyneuron is dark brown, and lobulation of the blade (fig. 2, ,4-G), giv- nearly round in cross-section (fig. 3, P), ing them a characteristic bizarre appear- with two delicate ridges running along the ance. top side, the ridges separated from each It is possible that the origin of the fertile other by a distance of approximately one- form of A. ebenoides by allopolyploidy oc- half the diameter of the rachis. In con- curred in recent times and only in one trast, A. montanum has a green rachis locality, and that this explains why its which is flattened-triangular in cross- distribution is so narrow compared to the section, the two adaxial ridges appearing broader distributions of the two following as conspicuous, rounded flanges, separated intermediate types. by a wide, shallow groove containing a medial bulge. The midrib of A. bradleyi (fig. 3, PM) is intermediate, being brown only in the lower half, and but slightly "Bradley's spleenwort" or "cliff spleen- flattened; the two ridges are separated wort," though wide-ranging from Okla- from each other a distance approximately

FIG.3. Diagrams from tracings of midrib cross-sections: P. A. platyneuron; PM. A. bradley; M. A. ~nontantint. RETICULATE EVOLUTIOX 109

FIG. 4. Epidermal cells. rl-C. Upper epidermis: A. Asplenir~irz pltrty~le.iii-oii: A,. Duval Co., Fla., Chz~rclaill;.k2. Bullock Co., Ga.! HaQcv. C. A. bvndleyi: B,. Dade Co., Mo., Steyer- mark 40273; R,. Eto~vallCo., Alabama, Eqgei-t. C. A. nzoiltn~i~~rir,*llleghany Co., N. C., Corvcll 10817. 13, E. Lower epidermis : D. A.pl~rty~~et~vou,Duval Co., Fla., Clzrlrchill; E. A. bvndleyi, Dade Co., Mo., Sfeyerr~zark40273. 110 WARREN H. WAGNER, JR. equal to the diameter of the rachis, and 1902), to be discussed below. The larger the upper groove formed between them fronds in any population may be espected shows only a slightly elevated, central to have elongated pinnae at the base of bulge. Such other A. platyneuron fea- the blade (forma elongatuwz Morton). A. tures as more or less truncate basal mar- pinnatifidum also resembles A. ebenoides gins of the leaflets and reduced lower in combining the peculiar features of A. pinnae are combined with the elongated rhi,-oplzylluvz with those of the more typi- tendency of the upper epidermal cells cal Aspleniums. Concerning the genus (fig. 4, cf. A, R, and C), the stalked and Ca.tazptosorusas conventionally recognized divided basal pinnae, and the restriction and typified by A, rhizophyllztwz, Copeland to acid-rock habitats characterizing A. (1947, p. 170) wrote: "Camptosorus is IIZOIZ~~IZZLI~." evidently derived from Asplenium. A. No individuals of A. bradleyi have yet pinnutifidurn Nutt. presents a more defi- heen recorded which are diploid and nite place of origin in the parent genus." sterile, although they should be sought Thus Copeland considers A. pinnatifidum for in habitats where the supposed parents (which has also been treated as Campto- occur side-by-side in appropriate con- so~z~spinnutifidus Wood) as an Aspleniuwz ditions. The chromosome number of A. prototype of Camptosorus. Perhaps the bradleyi populations which have been niost suggestive resenlblance of the two studied is in accord with the allopolyploid ' taxa is the attenuated, simple tip of the hypothesis, being 144, with normal pair- leaf which, though usually shorter in A. ing at meiosis (fig. 7, F). Hybrids of A. pinnatifidztm is much like that of A. rhizo- bradleyi have been recorded with both A. plzyllzt?~~.However, production of young platyneuro~zand A. ~2zonta~zunz.Probably plants on the leaf tip in A. pintlatifidu~nis A. bradleyi originated by allopolyploidy decidedly rare and I have found only a in the past at least once, and perhaps sev- few examples. The other major distinc- eral times, but it now has a rather broad tion of A. rlzi,-oplzyllztm from more typi- range, extending in particular well to the cal Aspleniums, that of anastomosing west of the westernmost known limit of veins, is much more common than bud- one of its putative parents, A. nzontanzm. production in A. pinnatifiduf~z,but the vein areoles form casually and infre- cluently. The lilcelihood is far greater that the This intermediate is a familiar but A. r7zi,-opIzyllu.taz-lilre features of A. pin- rather uncommon fern which occurs from natifidztm, rather than indicating com- Alabama and Georgia northward to New munity of origin, arose abruptly through Jersey, Indiana, and Oklahoma. Its re- hybridization of A. q~tontanztwwith A. semblance to A. ebenoides is sufficiently ~lzi,-ophyllzi~n.The first suggestion in the close so that the two species are frequently literature of this possibility seems to be confused. Eaton wrote (1579) that it that by the writer (1950). The chromo- "bears considerable resemblance to A. some numbers of the five populations thus ebenoides but has a green, herbaceous far investigated accord with this hypoth- midrib or rachis, a sinuous-margined pro- esis, having the 4n number, and four longation, thicker texture, and is very additional populations examined showed rarely, if indeed ever, proliferous." Lilte the corresponding large-sized stomata. A. ebenoides, A. pinnatifidunz frequently Not only does A. pinnntifidwz show tend- attracts attention by the irregular appear- encies toward the features of anastomos- ance of its leaf-forms (e.g., Copeland ing veins, attenuate leaf-tip, and leaf-tip reproduction of A. rhixophyllum, but its 2 Reports of A. bradleyi on "litnestone" have proved to be errotleous. Cf. Wherry, Amer. lamina is thick-textured as in A. man- Fern Jour., 21 : 111, 1931. tatlztwZ; moreover there is a definite tend- RETICULATE EVOLUTION 111 ency toward elongation of the upper epi- tively in Pennsylvania, West Virginia, derillal cells, and the basal pinnae or lobes, and Ohio, and the findings not only bear though simple, have the same outline as in on its own interpretation but on that of A. lnontanufr4. The habitat of A. pinnati- A. fiinnati,iidt(?~zas well. Although its fiduwz is usually nlediacid or subacid soil describer considered it "possibly in on rocli cliffs as in A. wzontanu~~z,although part the result of hybridization between Craw (1932) found 3 occurrences in A. pinnntifidum and A. wzontanu~" Indiana to have neutral soil-reaction, and (Wherry, 1925), its most recent treat- the writer has seen it in close associati011 ment (Morton, in Gleason, 1952) is as a with Pellaea glabella, a predominantly variety of A. pi?znatifidztwz. In 1932, how- calcareous soil plant, at Blacli Hand, ever, Dr. Paul Kestner of Lausanne, 0hi0.~ Switzerland, found that spores of all One distinctive feature of A. pinnnti- specinlens of A. trttdellii he received from fidztm which points especially strongly to Georgia and Tennessee were sterile. In its hybrid origin is the irregularity of the spite of the sterility of its spores-which blade outline, seen primarily in larger may also be recognized under the micro- leaves. The two other hybrids between scope by their irregularity and abortion- A. rhizoplzyll?~~g~and more typical As- this plant, remarkably, is often relatively pleniums (A. plclynetwon, fig. 2, A-G; common where it occurs. But apparently A. ruta-mztraria, discovered by Braun, . it is never found in the absence of the 1939, fig. 2, H) liliewise display this two presumed parents, A. pin~mtifiduwz condition. Rather than a symmetrical and A. Izzolltalzztm. system of pinnae or lobes as is character- Present cytological evidence illakes it istic of most ferns, the blades of these probable that A. trudellii is a hybrid, but hybrids show conspicuous irregularities its morpholog~7,though intermediate, usu- in pattern, short lobes next to long ones ally seems son~ewhatcloser to A. pilaplati- on the same side of the blade, and differ- fidzclzt than to A. ~zzontanztza. The chromo- ent patterns of lobulation on opposite some number in all the plants from the ~icles.~ three populations studied is 108, i.e.. 3n. Also at meiosis there are approximately 36 pairs and 36 univalents, as ~vouldbe Oddly enough, the widespread plant expected in a baclicross of an allopolyploid lcnown as A, trudellii, which has been hybrid of A. ~zzonfanuwtand A, rhizophyl- interpreted both as a variant of A. pinplati- 1u.tzz with one of its parents. In 18 well- fidum and as its hybrid with A. montanum, spread sporocytes the total units of an! lacks the peculiarity of irregular leaf- kind estimated averaged 73.2 (70-95). blades, a feature which is sometimes help- with 36.3 bivalents (33-42) and 36.9 ful in distinguishing it from A. pinnati- univalents (33-39). The genetic con- fidzt~zt. A, trudellii has been investigated stitution of A. frz~delliimay thus be now cytologically from three localities, respec- considered to he two genomes of A. man- tn?zzr~gzand one genome of A, rhizophyl- 3 Dr. E. T. Wherry has since informed me 11~1yt. that the same species-association occurs at In spite of its production in consider- Cumberland Falls, I

DI D 2

-400m mu. + FIG.5. Upper epidermal cells. A. A. ~noiztan?lm:A,. Macon Co, N. C., Correii 6679, A. Sevier Co., Tenn., Tryon 33. B. A. trztdellii: Bl Etowah Co., Alabama, Eggert: B?. Kentucky, Rule. C. A. fiin~zatifiduwz:C,. Carter Co., Mo., Steyermark 11875; C2. Union Co., Ill., Hubricht BZZ87. D. A, rlzi~opityllz~wz:Dl Smith Co., Va., Svrall; D,. Shenandoah Co., Va., .41-t- 900. RETICULATE EVOLUTION 113

\vide range suggests ancient origin at least above. The stomata of one pintla or frag- once by doubling of chromosomes in an ment of all these plants under cliscussioil original hybrid. Plant collectors should may vary in length as iuuch as 11 to 18 certainly seek in the field individuals of inicrons, and the averages of different .4. pinnatijidlt??~\vhich are diploid and collections of one species as much as 5 sterile, to be expected where A. rlziao,blz~ll- to 10 microns. Nevertheless it has usu- luin and A. Iizontanzrm occur in close ally proven possible by averaging 30 meas- proximity. urements of stomatal lengths of a pintla from a single collectioil to clistiilguish the polyploid levels. The fertile tetraploids have coilspicuously large stomata. Iiidi- vidual collections of lcilown or theoretical Aspleniuwz kentuckiense, discovered and triploids may have values ~vhichoverlap described first by ;\/lcCoy (1936), repre- sonie of the diploids and some of the tetra- sents the theoretical central point niorpho- ~loids.as shown in the chart below of logically in the entire Appalachian Asple- averages of 30 stoinatal lengths in nii- ilium complex. Although McCoy did not crons of each of 31 collections : discuss its hybrid origin, Wherry (1936) first suggested that it represented an in- Diploids termediate between A. pinnntifidz~~nand (range of averages of 11 collections, 39-45). A. plafyneuron. (The other plant, A. A. vlzkophg~llum(3 collections) . .4l (40-41) A. ebe~zoides (2 collections) . . . . .40 (3940) sfofleri,nientioned by Wherry in this con- A. platylzezlvou (3 collections) . . .42 (40-43) nection, is now believed to involve an A. +~zontanuriz(3 collections) . . . . .42 (4&45) additional species outside the coniplex.) Triploids Besides the morphology of A. 17entucki- (range of averages of 9 collections, 42-51) cnse, which is intermediate between the A. bradleyi X nzonta~zt~~lt(2 collections) two supposed parents, additioilal observa- 44 (4245) tions indicate that it is the cross of A. A. trudellii (4 collections) ...... 46 (42-51) platynezwon and A. pinnatifidzdm: (1) A. lzentzlckiense (3 collections) . .45 (43-48) three plants were collected by Mr. Floyd Tetraploids Bartley in Hay Hollow, Pike Co., Ohio, (range of averages of 11 collections, 46-58). A. gvavesii (4 collections)...... 49 (46-53) in the immediate vicinity of the two puta- A. ebenoides (1 collection)...... 54 tive parents; (2) in southern Illinois, A. pin~zatifidz~nz(3 collections) . . .55 (50-58) where A. lzentuckiense was found by Earl A. bvadleyi (3 collections) ...... 56 (52-58) (specimen in Chicago i\luseum), only A. Asplenizm gravesii, described by Maxon platynez~ronand A. pinnatifidz~wz of pos- (1918) as a hybrid of A. pinlzatifidt~mand sible parents are present, neither A. brad- A. bradleyi, has been found in a limited lcjli nor A. ~nonta~z~ritzbeing lcnown from number of localities in Georgia, Alabama, that state. Specimens believed to repre- Ohio, and Pennsylvania. "Judging A. sent A. plnfyneuron x A. pznntrfiJid~i7jiare now known from scattered points in Ken- gravesii on both gross and minute char- tucky, Ohio, Illinois, and Arkansas, and acters, there can be little doubt of its they all possess abortive spores. Al- hybrid nature, niaking all allowance for though no living plants have been avail- the unusually high variability of the sup- able for study, the stomatal sizes of pre- posed parents, of one or the other of served material indicate the 3n condition. which it might at first be thought an ex- Stomata1 size seems especially useful trenle state" (Maxon, loc. cit.). A. in the determination of dried niaterial of gravesii niay be distinguished from A. the rarer ferns in this complex, the make~ztl~ckie~zseby having less brown on the terial being relaxed in sodium hydroxide leaf-axis, pinna shape, thicker texture of solutions, cleared, and treated as described the blade, somewhat Illore elongate upper 114 WARREN H. WAGNER, JR.

I 1 I .-40o0mu. - FIG.6. Epidermal cells : A. Aspleltili~z cbenoides, upper epidermis: A,. Hale Co., Alabama, !!-faxo!z & Poilard (412) ; A?. Montgomery Co., Md., Wngrlc~. B. A. keiztl~ckie~ue,S. Illinois, Earl: B,. Upper surface; B?.Lower surface. C. A. graoesii, Trenton, Ga, Graves: C,. Upper; C,. Lower surface. epidermal cells (cf. fig. 6, B and C), and These two iilterinediate ferns, A. grave- the usually some\vhat larger stomata1 sizes sii and A. ke?ttz~ckiense,combine morpho- suggesting the 411 state. Hori~ever,the logically distinctions of all three of the two ferns are similar ~uorphologically,and basic diploid species, in respect to such young plants especially may prove to be features as the extent of brown on the leaf- difficult to identify on gross characteris- axis, the degree of attenuation of the leaf- tics. Both are exceedingly rare, and no tip, pinna shape, etc. But A. herztllckiense living material has been seen by the au- possesses unusual theoretical interest in thor or is expected in the near future. this problem since it presumably repre- RETICULATE EVOLUTION 115

FIG. 7. Chromosomes of Appalachian Aspleniums. A. Aspleniztnz platynettro?t: A,. Mont- gomery Co., Md., 2r1 = 72; A?. Hardy Co., W. Va., 36 pairs. B. Aspleitilt+n rhiz~ophyllztuz: B,. Shenandoah Co., Va., 212 = 72; Monroe Co., Indiana, 36 pairs. C. Aspleniz~nz ebenoides (sterile form), Montgomery Co., Md. : C,. 2n = 72; C, and C,. 72 univalents at meiotic metaphase. D Asplelzit~wz ebciioides (fei tile form), Hale Co., Alabama : Dl. 212 = 144; D,. 72 pairs. E. Asplenitr~iz ~~zolitantriii,Harford Co., Md.: E,. 212=72; E,. 36 pairs. F. Asfileni~~+nblpadleyi: F,.Pike Co., Ohio, 212 = 144; FA.Harford Co., &Id., 72 pairs 116 WARREN H. WAGNER, JR. bents the precise n~orphological central virginica) ancl others ~vhich occur as point bet\?-een the "poles" of the triangle normal species. Only recently, however, of basic, cliploicl species (fig. 1, PRM). has allopolyploicly in species of the Fili- Thus it should theoretically be possible cineae been subjected to investigation: to produce triploid A. kentltckiense in all Manton (1950) has shown that seven of the following experimental crossings : European ferns with ilorillal life-cycles, including Diyopteris filix-?was and Poly- sticlzz~~~~trclrleatc~?)~, are either suspected or demonstrated allopolyploids. >lore recently (Manton and Wallter, 1953) it has been sho~vnthat the American Dryopteris clintoninnn 1s a hexaploicl ancl "is perhaps Stehhins (1950) has discussed the role the ainpl~idiploicl hybrid between normal of polyploidy'in plant evolution, and cites D. cristtrtn [a tetraploid species] and D. a number of established or suspected al- goldinnn [a diploid] ." lopolyploids anlong flowering-plants, in- Perhaps the illost interesting aspect of cluding Gnleopsis tctrirhit (G. pc~bescens- the Appalachian Aspleniuin comples is speciosa) and Iris versicolor (I. setosn- that the polyploid tava combine features

740 mu. ---

FIG.8. Chromosomes of Appalachian Aspleniurns. A. .~S~/C~I~I~III~ii~il(~tifi(l~iiii: -4,. Shenan- doah Co., Va., 211 = 144; A2. Monroe Co., Indiana, 72 pairs. B. .4. ti.iitlcIlii: B,. Yosk CO., Pa., 212 = 108; Bz. Yorlc Co., Pa., approximately 36 pairs and 36 univalents at meiotic 11letapliase; B:, B4. Hardy Co., 14'. Va., approximately 36 pairs and 36 univalents. RETICULATE EVOLUTIO?J 117

of diploids of widely different morphology of the origin of such a form should be and relationship. The suspected allo- recognized by field-workers. polyploid, "Scolopendriuwz hybl.idz~711," of Europe (Manton, 1950) likewise com- bines features of very unlike putative parents, Asplenium (Scolopend9,iu11z) lzelni- The Appalachian Aspleniunls comprise onitis and A. (Ceteraclz) cetel,aclz. The 11 described taxa, of which 3 represent polyploids of the Appalachian Asplenium i~~orphologicalextremes and the remainder coinplex appear to be "typical or genoinic intermediates. A. ebenoides is a usually allopolyploids" (Stebbins, op. cit., p. 226) sterile hybrid of A. platyneuron and A. resulting from hybridization of distantly ~~lzixoplzyllu~~zwith 72 univalents at meiotic related species in which the chromosoines inetaphase, although one population from are so different that nor~nalpairing would Alabama is a fertile allopolyploid, for~lling be itnpossible in the diploid. The poly- 72 normal bivalents at meiosis. A. brad- ploid intermediates between the presum- lcyi is apparently the allopolyploid hybrid ably basic diploid species in this coinplex of A. wtontanuwz and A. plat3ine1i~o?t;it are clear-cut intermediates, and there is back-crosses with both parents. A. pill- evidently no tendency toward intro,ores- 91ntifidz~11tlikewise is evidently an allopoly- sion between parent species since gene ploid hybrid, ancl its morphology, the ir- interchange through the diploid hybrids regularity of the leaves, and the pairing is extremely unlikely. behavior of its putative baclccross (A. Two of the diploid species of this con- f~udellii)with one of the parents indicates plex, Aspleniti~z rrlzizophyllu~~zand A. that its parentage is A. wzonfnnzmt x A. plat3ineu~on, are much inore abundant ~l?izophyllzt~a.A. Kent~icl~ienseancl A. than their apparent allopolyploid deriva- grazlesii are both evidently trihybricls, the tives, in contrast to many of the previously former 3n, the latter 4n, and '4.Izwztz~cizi- reported allopolyploids atnong plant spe- ense morphologically represents the theo- cies. One of the allopolyploids discussed retical central point of the whole cotnplex. here, namely A. ebenoides, is very limited A hypothesis of reticulate evolution is in range, while the two others, A. b?,adle)li thus presented for the Appalachian Asple- and A. pinnatifiduw?-although generally niurns which postulates that three original less common than A, nzontanu~a where cliploid species, A. ~~?ontanu~z,A. platy- their ranges overlap with it-have ex- netiron, and A. rhizoplzylluwz, have given tended their ranges farther ~vest~varclthan rise to 8additional taxa through hybridiza- this presun~ablyparental, diploid species, tion. but not farther northward. Of the theoretically possible sterile hybrids in this complex, five still remain to I an1 indebted to Dr. G. Ledyard Steb- be discovered, but their arising under bins and to Dr. Edgar T. Wherry for natural field conditions seeins extremely critical reading of the manuscript, and to uillikely because of parental differences Mr. Walter F. Kleinschiniclt for gron-ing in ranges and habitat preferences. Even the living plants. I wish also to thank what is perhaps the most expected of the those persons who so kindly supplied liv- still unknown hybrids, A. ~hiaophyllu~ax ing materials for study, including Mr. A. pinnatifidui?a, may never be found Floyd Bartley, Mr. Donald F.M. Brown, under natural conditions because these AIr. Neal W. Gilbert. Miss Mary E. Groff, species so rarely occur close together. Of Dr. Charles B. Heiser, Dr. Everett G. the four theoretically possible fertile poly- Logue, Dr. Erich Steiner, Mr. Harry ploid hybrids, only one-hexaploicl A. Truclell, and Dr. ST7herry, and those kenfuckiense-has not yet been fou~lclin herbaria which sent dried materials, in- the natural state, although the possibility cluding the University of Pennsylvania, 618 WARREN H. WAGNER, JR.

U. S. National i\Iuseutn, Chicago Mu- MANTON,I.. AND S. WALKER. 1953. Cytology seum, University of Illinois, and Missouri of the Dvyopteris spiiz~~losacomplex in East- eril North America. Nature, 171 : 1116- 13otailical Garden. Mr. Hubert M.Vogel- 1118. man helped me with the drawings. MAXON,WILLIAM R. 1918. A ne\v hybrid Aspleiziltln. Amer. Fern Journ.. 8 : 1-3. McCoy, TI-IOMASN. 1936. A new .Isple?ziunz from Iientucky. Amer. Fern Journ., 26: ALSTON,A. H. G. 1940. Notes on the supposed 104-106. hybrids in the genus Aspleniutn found in SLOSSON,MARGARET. 1902. The origin of Britain. Linn. Soc. London Proc., Session Asplelti~wt ebelzoides. Torrey Bot. Club 152, 1939-40 (Part 2) : 132-144. Bull., 29: 487-495. BRAUN,E. LUCY. 1939. A new fern hybrid- STEBBINS,G. L., JR. 1050. Variation and Aspleniunz cryptolepsis X Camptosorlts rhizo- evolution in plants. Kew York. plzyllzrs. Amer. Fern Journ., 29 : 133-135. WAGNER,W. H., JR. 1950. Cytotaxonomic COPELAND,E. B. 1902. Two fern monstrosities. analysis of evolution in Pteridophyta (re- Bot. Gazette, 34: 143-144. view). Evolution, 5 : 177-181. -. 1947. Genera Filicum. Waltham, Mass. -. 1952. The fern genus Diellia. Univ. CRAW,JOE E. 1932. Hydrogen-ion reaction of Calif. Publ. Bot., 26 : (no. 1) : 1-212. native Indiana fern soils. Butler Univ. Bot. -. 1953. A cytological study of the Ap- Studies, 2 : 151-158. palachian spleenworts. Amer. Fern Jot~rn., EATON,D. C. 1879-80. Ferns of the United 43 : 109-114. States. I and 11. Salem, Mass. WHERRY,E. T. 1925. The Appalachian Asple- GLEASON,H. A. 1952. The new Britton and . niums. Amer. Fern Journ., 15 : 47-54. Brown 'Illustrated Flora of the northeastern -. 1927. Notes on Asplcizizi~c trzldellii. United States and adjacent Canada. Vol. I. Amer. Fern Journ., 17: 135-138. Lancaster, Pa. WHERRY,E. T., AND WILLIAMD. GRAY. 1936. MANTON,I. 1950. Prol>lems of cytology and Variants of some Appalachian Aspleniums. evolution in tile Pteridophl ta. Cambridge. Amer. Fern Journ., 26: 77-86.