Chromosome Studies of Brazilian Ferns

Chromosome numbers, habitat and general geographic distributions are reported for a sample of ferns from the vicinity of Manaus. These numbers are compared with those reported for species in other regions. The report of n=ca. 154 for Schiza.ea incurvata is the first record of the genus in the American tropics and the others are first counts of these species in continental South America. In Lindsaea n=42 and n=ca. 84, are somewhat lower and appear to represent a line distinct from the species of the paleotropics with n=47. In Trichomanes n=32, 64 and 128 appear to be levels of polyploidy based on X= 8, while n=72 is possibly based on X = 9 and suggest possible diverse lines in the genus. Pityrogramma calomelanos with n=ll6 is considered as an octoploid based on X= 29 . Ou r sample cov~ring about ten percent of the pteridophytes in the area we collected, yielded numbers ranging between n=32 and n = 154, and generally appears to represent high polyploid levels. The American tropics has one of the world's richest fern floras encompassing some 4000 species, but cytologically it is poorly known. Cytological surveys of pteridophytes were initiated with studies centering on the British species by Manton (1950) and were soon extended to many regions as lndia, Ceylon, Japan, West Africa, New Zealand and eastern North America. The most comprehensive study of American species is the survey of Jamaican ferns by Walker (1966) w hich encampasses 270 of some 540 species listed for the island. This is a major col'ltribution to the subject and the most useful reference for our work. There are essentially no cytological investigations of t he ferns of continental South America. This point is clearly made in a reAlice F. Tryon (1) Hortensia P. Bautista (2) l zonete da Silva Araújo (3) cent paper by Britton (1974) which summarizes tho work in fern cytology and indicates they are rather well sampled except for those of South /\merica and mainland China. The opportunity to include a section on cytology in the Class of Pteridophyta, taught during September and October 197 4, as part o f the Curso de Botânica Tropical do Instituto Nacional de Pesquisas da Amazônia was taken to begin work with Brazílian students. We hope that this sample of our accomplishments over a relatively brief period will motivate further studies in the Amazonian area and in other parts of South America. MATERIALS AND METHODS Leaves with young sporangia were fixed in Farmer's solution consisting of 1 part glacial acetic acid to 3 parts absolute alcohol. Some material fixed in 96% alcohol softened, for example, see Fig. 10 of Trichomanes pilosum, and the chromosomes are not wellspread. The si ides were prepared by removing sporang ia from the leaves into a drop of fixative. A drop o f aceto-carm i ne was added after the solution evaporated; the slide was then heated over an alcohol lamp, and squashed. Preparations were made permanent by inverting t he slides on glass rods in 95% alcohol, untíl the cover glass dropped, and mounting in Diaphane. Meiotic f igures obtained in eleven species are illustrated and discussed in this paper. Collections were made at four localities in the vicin ity of Manausat the Instituto Nacional ( 1} ~ Gray Herbarium, Harvard University, Cambridge, Massachusetts 02138. ( 2) INPA, Museu Paraense Emílio Goeldi, Belém. Bolsista do CNPq. ( 3} Instituto Nacional de Pesquisas da Amazônia, Manaus. Bolsista do CNPq. ACTA AMAZONICA 5(1):35-43. 1975 -35 de Pesquisas da Amazônia (INPA), at the Reserve Florestal Ducke (km 26), on the Manaus to ltacoatiara road, at the Estação Experimental de Silvicultura Tropical (km 60), and at km 130 on the Manaus to Caracaraí road. Permanent slides and herbarium vouchêrsare deposited at the Gray Herbariurn, Harvard University. Species, voucher Schizaea Meiotic number General Distribution S. incurvata Schkur INPA, Conant 1103 Manaus. ca. 154 This is one of the two species of Schizaea frequent on the open, white sandy campina areas that are interspersed arnong the forests, near Manaus. This and S. pennula Sw. form l<:1rge colonies in open sunny places or often arnong shrubs at the edge of the forest. The sterile leaves of both species are filiform, characteristic of a group of species in Schizaea, but the fertile ones are quite distinct. The fertile segments of S. incurvata are attached laterally to a central rachis and are often conduplicate as in Fig. 1. Those of S. pennula are clustered at the petiole apex in a tufted arrangement. This report of ca. 154 in Schizaea incurvata (Fig. 2). the first cytological record of the genus for the American tropics, is relatively high for the pteridophytes. There are some pt·oblerns in working with such high numbers of chromosomes but there are also cornpensations in these plants. They have large sporangia with many more megasporocytes than in Guianas, Amazonian , Brazil, Upper Orinoco and Gran Sabana, Venezuela, Vaupés, Colombia. spec ies of t he Polypodiaceae in which there usually are only 16 spore mother cells per spomngium. In this material of Schizaea the chwmosomes were easily spread, and meiosis appears to be synchronized with numerous nuclei in different phases of diakinesis within a single sporangium. fhe chromosome number in Schizaea incurvata is clearly polyploid although the levei is difficu lt to establish. The lowest number reported for the genus is n= 77 in S. asperula Wakefield, from New Zealand. The same number is reported in S. dichotoma (L.) Smith, from Ceylon while another specimen of the species, from New Zealand is reported as n= 540. This large number of chromosomes approaches that in Ophioglossum reticulatum L., from India which is illustrated by Ninan (1958) with n= 630, the largest nurnber of chromosomes reported for any living organ ism. lt is interesting to note that the leaf form in Schizaea is a simple one as it is in Ophioglossum.

cent paper by Britton (1974) which summarizes tho work in fern cytology and indicates they are rather well sampled except for those of South /\merica and mainland China.The opportunity to include a section on cytology in the Class of Pteridophyta, taught during September and October 197 4, as part o f the Curso de Botânica Tropical do Instituto Nacional de Pesquisas da Amazônia was taken to begin work with Brazílian students.We hope that this sample of our accomplishments over a relatively brief peri od will motivate further studies in the Amazonian area and in other parts of South America.

MATERIALS AND METHODS
Leaves with young sporangia were fixed in Farmer's solution consisting of 1 part glacial acetic acid to 3 parts absolute alcohol.Some mat erial fixed in 96% alcohol softened, for example, see Fig. 10 of Trichomanes pilosum, and the chromosomes are not wellspread.The si ides were prepared by removing sporang ia from the l eaves into a drop of fixative.A drop o f aceto-carm i ne was added after the solution evaporated; the slide was then heated over an alcohol lamp, and squashed.Preparations were made permanent by inverting t he slides on glass rods in 95% alcohol, untíl the cover glass dropped, and mounting in Diaphane.
Meiotic f igures obtained in eleven species are illustrated and discussed in this paper.Collections were made at four localities in the vicin ity of Manaus-at the Instituto Nacional ( 1} ~ Gray Herbarium, Harvard University, Cambridge, Massachusetts 02138.
( 2) -INPA, Museu Paraense Emílio Goeldi, Belém.Bolsista do CNPq.( 3} -Instituto Nacional de Pesquisas da Amazônia, Manaus.Bolsista do CNPq. de Pesquisas da Amazônia (INPA), at the Reserve Florestal Ducke (km 26), on the Manaus to ltacoatiara road, at the Estação Experimental de Silvicultura Tropical (km 60), and at km 130 on the Manaus to Caracaraí road.Permanent slides and herbarium vouchêrs-are deposited at the Gray Herbariurn, Harvard University.This report of ca.154 in Schizaea incurvata (Fig. 2). the first cytological record of the genus for the American tropics, is relatively high for the pteridophytes.There ar e some pt•oblerns in working with such high numbers of chromosomes but there are also cornpensations in these plants.They have large sporangia with many more megasporocytes than in Guianas, Amazonian , Brazil, Upper Orinoco and Gran Sabana, Venezuela, Vaupés, Colombia.

Schizaea
spec ies of t he Polypodiaceae in which there usua lly are only 16 spore mother cells per spomngium.In this material of Schizaea the chwmosomes were easily spread, and meiosis appears to be synchronized with numerous nuclei in different phases of diakinesis within a single sporangium.fhe chromosome number in Schizaea incurvata is clearly polyploid although the levei is difficu lt to establish.The lowest number reported for the genus is n= 77 in S. asperula Wakefield, from New Zealand.The same number is reported in S. dichotoma (L.) Smith, from Ceylon while another specimen of the species, from New Zealand is reported as n = 540.This large number of chromosomes approaches that in Ophioglossum reticulatum L., from India which is illustrated by Ninan (1958) with n= 630, the largest nurnber of chromosomes reported f or any living organ ism.lt is interesting to note that the leaf form in Schizaea is a simple one as it is i n Ophioglossum.

Some of the extreme morphological and ecological diversity among the species of
Trichomanes is shown in the sample of species cytologically studied.A li of them grow in generally moist, shaded situations but occu-PY quite different habitats.Trichomanes arbuscu/a forms dense colonies near the edge of tr.e water, along stream banks .The leaves are dimorphic with the fertile ones long petio• late (Fig. 3).Leaves of T. cristatum are monomorphic, caespitose and with a deeply pec-Cl1romosome studies ... tinate Iam i na.Only a few plants were found in dense moss at the base of a large tree.Trichomanes e/egans has exceptionally large leaves for the genus with the lamina finely dissected and more than 30 em long (Fig .5).We found only a si ngle plant along a stream bank with the leaves overhanging the water.Plants of T. pi/osum were abundant on the face of shaded sandstone rocks .The caespitose leaves (Fig. 9) are covered with dense indument.Trichomanes tanaicum is scandent and quite a rare species although with a rather wide distribution in northern South America.We found only a single plant (Fig. 7) on an upright, dead tree trunk.
The species of Trichomanes are excellent cytological subjects for many have-relative:y few, and remarkably large chromosomes as in Fig. 6 of T. elegans and Fig. 8 of T. tanaicum.The bivalents in T. arbuscula (Fig. 4) are so-mewhat smaller but are in an earlier, diplotene stage prior to diakinesis.This species has ca.128 bivalents, the highest number known for the genus which has also been reported in T. crispum L. from Jamaica, by Walker (1966).He also has a record of n=64 in T. arbuscu/a from Trinidad, which is considered as possibly 16 X.Thus our collection from Brazil with n= 128 may be 32X.The chromosome 1\ Fíg .5 -Voucher collectíon of Trichomanes elegans Rích .with híghly divíded monomorphic leaves .Fig. 6 -Chromosomes of Trichomanes elegans at diakine~ís n=32, X 1000.Fig. 1 -Volicher collection of Trichomanes tanaicum Hook.with scandent rhizomes.The numerous elongate leaves with slender, marginal sori.Fíg. 8 -Chromo?omes of T. tanaicum at diakinesis n = 32, X lOOO.number of T. pilosum between n=62-68 (Fig. 1 O) is not clear beca use of severa!superimposed bivalents, but certainly represents a relatively high po lyploid levei.
The species of Trichomanes are also good cytological subjects because of the gradate sorus.Meiosis can usually be found on the receptacle between the basal point where sporangia are initiated and the apex where they are dehiscing.Our record of n=32 in both T. elegans and T. tanaicum, which are morphologically very distinct, indicates that number may be established in different evolutionary I ines.Consideration of the chromosome numbers along with other aspects of the plants suggests that n=72 in T. cristatum (Fig. 11) may be constituted from a line based on 9 rather than 8 as in the other species.
Trichomanes elegans is a very distinctive species and has been treated in subgenus , .. .,_.. . . . ... Pachychaetum, in its own section Davalliopsis, in the treatment of the genera of the Hymenophyllaceae by Morton (1968).In that work, T. tanaicum and the other species that have been studied here are placed in different sections of the subgenus Achomanes.
Chromosome numbers have been reported in about 77 species of the Hymenophyl laceae and there is a considerable range of numbers between 11 and 128.Derivation of the numbers in the Hymenophyllaceae has been dis-cussed in considerable detail by Walker (1966) and he postulates 6, 7, 8, and 9 as base numbers.In this system our records of 32, 64 and 128 would be based on 8, whi le 72 would be on 9.The scheme of Walker seems more plausible than one proposed by Vessey & Barlow (1963) in which chromosomes are lost and different sets are combined to arrive at the correct sum.In that scheme n= 32 would be based on X= 17 wHh the loss o f one and doubling to 32.Plants are often abundant in open, sunny sites especially in disturbed places along roadsides and on cleared hillsides.There were large colonies at km 130 where we collected ou r specimen at the edge ot the forest.Pityrogramma belongs to a complex of genera that have chromosome numbers based on X= 29, including Anogramma, Eriosorus, and Jamesonia, although one of the most distinctive species, Pityrogramma triangularis (Kau lf.) Maxon, in Ca lifornia ha,~ a diploid race with n==30 and a tetrapolid race with n= 60.Severa!nuclei in our preparations of Pityrogramma ca/omelanos are clearly n= 116 as Fig , 12.The Wide ranging from Mexico south to Bolívia and Argentina, Antilles and Floridêõ species is reported as n= ca.120, from Jamaica , by Walker (1966).That same number is given for a specimen from northern lnd ia, and there are also records of n=116 for plants from southern lndia and Formosa.The species is regarded as wholly American, in the taxonomic treatment of the genus by R. Tryon ( 1962), but it i s noted that it h as been i ntroduced in many are as of the paleotrop:cs.lt is a frequent species in the American tropics.Further cytological work to determine whether both numbers do i ndeed occur in different popu lations and especially whether lower diploid populations occur in the American trop•cs, will be of much interest.

ca. 84
Mostly Amazonian, also in M ato Grosso and Peru

Tryon et alli
Lindsaea is a frequent genus in the moist forests around Manaus and two or three species may occupy the same si te.The plants are terrestrial , commonly on stream ban!<s and in open places in the forest.Lindsaea lancea var.falcata occurs in ali of the localities we collected and there are many scattered plants in open places in the forest.The leaves have compact pinnae with a large terminal segment (Fig. 13) .Lindsaea schomburgkii is a rarer species growing in open, more exposed places with longer leaves and somewhat coriaceous pinnae.Lindsaea diva- ricata and L. quadrangularis have similar more complex leaves with smaller segments.L. divaricata (Fig. 15) occurs in somewhat wetter sites.
There are few cytolog ical records for American species.The chromosomes are difficult to study because of their relatively small size and the tendency for bivalents to adhere.Sporangia at meiosis are quite small as compared to other genera .A report of Lindsaea arcuata Kze. with n= 84 was made by Mickel et ai. (1965) based on a collection from Oaxaca, Mexico.Lindsaea portoricensis Desv.was reported from Jamaica as n= ca.88, by Walker (1966) and was considered to be tetraploid.On the basis of this earlier work on the genus.it appears that our material of L. lancea var.falcata (Fig. 14) and L. divaricata (Fig. 16) represent diploids and L. schomburgkii (Fig. 18) is tetraploid.The cell shown here of L. quadrangularis (Fig. 17) with 84-88, largely univalents and some multivalents, at arrow, is one of three cells examined with mostly unpaired chromosomes.The univalent and multiva lent associations in this material indicate an abnormal meiosis.
In the treatment of American Lindsaea by Kramer ( 1957).L. /ancea v ar.ta/cata and L. schomburgkli are placed in subsection Terminales and L. arcuata, L. divaricata and L. quadrangularis in subsection Decrescentes.There appears to be no cytological differentiation between these subsections since species with n= 42 occur in each of them.
There are many cytological records for paleotropical species of Lindsaea and meiotic Chromosome studies ... chromosome numbers range between 34 and 155.The most frequent, n= 47, is reported in nine species from widely distant areas as L. vieillardii Mett.from New Caledonia , L. chienii Ching, from Japan, L. concinna J. Sm., from Australia, and in L. parallelogramma v A. v R. from Malaya.However, there are other numbers also reported for species of the paleotropics as n=34 in L. linearis Sw. from New Zealand and n= 82 in L. caudata Hook.from Ceylon which show that changes in chromosome numbers have occurred other than polyploid doubling of the complement.The record of n= 42 in L. trichomanoides Dryand., by Brownlie (1956) suggests a possible common base number among species of both reg ions.
However, the gross morphology of L. trichomanoides, especially the leaves is so different from the specimens that we have stud ied that it appears a connection between them is probably a distant one.

DISCUSSION
The relatively small sample of chromosome numbers reported for the Amazonian region does not allow broad comparisons with other geographic areas as has been dane for England where the whole fern flora has been examined.The number of pteridophyta in Amazonia is estimated at about 300 species and about 100 occur in the are a we collected around Manaus; our cytological sample represents about ten percent of these .Some preliminary comments can be made relative t o other records especially for Lindsaea and Trichomanes.The report of L. por toricensis with n=ca.88, from Jamaica was considered a tetraploid, thus our record of n= 42 is probably diploid and n= ca.84 tetraploid.Severa!Lindsaea species with n=47 in the paleotropics suggest that there are distinct evolutionary I ines in these two regions.
Cytologically Trichomanes is one of the best known genera and there are reports of the following meiotic numbers: 21 , 22, 26, 32, 33, 34, 35, 36, 56, 64, 68, 72, 108, 128.Most are at higher leveis than our reports.lf 8 is considered a base number for the group, n= 32 would represent an octoploid and our highest number n= 128 would be 32X.Pityrogramma ca/ome/anos with n = 116 represents an octoploid based on X = 29.The highest number among the species we report, n=ca.154, in Schizaea incurvata, is intermedtate in the series of numbers reported for species of the paleotropics and one of North America with 77, 94, 96, 103, 270, 325 (± 30).350-370, and 540.A base number for Schizaea was proposed as X= 77.lf this is accepted, our specimens would represent a relatively low levei of ploidy, in spite of the high number of b:valents.
The following tab le summarizes our reports on Amazonian fern cytology.

Species
of the two species of Schi zaea frequent on the open, white sandy campina areas that are interspersed arnong the forests, near Manaus.This and S. pennula Sw. form l<: 1rge colonies in open sunny places or often arnong shrubs at the edge of the forest.The sterile leaves of both species are filiform, characteristic of a group of species in Schizaea, but the fertile ones are quite distinct.The fertile segments of S. incurvata are attached laterally to a central rachis and are often conduplicate as in Fig. 1.Those of S. pennula are clustered at the petiole apex in a tufted arrangement.