Karyology of some Brazilian species of Alismataceae

Abstract

The family Alismataceae (sensu Cronquist) comprises 11 genera and 75 species of aquatic and semiaquatic herbs. In Brazil, only the two most speciose genera, Echinodorus Rich. ex Engelm. and Sagittaria L., are present. The chromosome number 2n = 22 is common to all species. We present karyotypes for eight species, together with their asymmetry index (TF%) and total chromatin length (TCL). All karyotypes have high karyotypic asymmetry and TCL, due to their large and mostly acrocentric chromosomes. The chromosomal evolution of the Alismatidae group is discussed.

aquatics, centric fission, chromosomal evolution, chromosomes, cytotaxonomy, Echinodorus, Sagittaria

INTRODUCTION

The family Alismataceae in the order Alismatales (Cronquist, 1988) comprises aquatic and semiaquatic, rhizomatous herbs, with aerial, floating or submerged leaves, growing in muddy and water-logged substrates. Members of the family can be recognized by their milky sap, basal placentation and fruits, which are mostly achenes (Haynes & Holm-Nielsen, 1994). It consists of 11 genera and about 75 species native to tropical, subtropical and subtemperate regions worldwide. In Brazil only the two most speciose genera, Echinodorus and Sagittaria, are present.

Dahlgren & Clifford (1981) established the order Alismatales with just one family, Alismataceae, that included species previously assigned to Limnocharitaceae. Later, Dahlgren, Clifford & Yeo (1985) included in the order five different families: Aponogetonaceae, Butomaceae, Hydrocharitaceae, Limnocharitaceae and Alismataceae. Cronquist (1988) separated Hydrocharitaceae as the only member of Hydrocharitales, and transferred Aponogetonaceae to the order Najadales (comprising Aponogetonaceae, Scheuzeriaceae, Juncaginaceae, Potamogetonaceae, Ruppiaceae, Najadaceae, Zannichelliaceae, Posidonaceae, Cymodoceaceae and Zosteraceae). The three remaining families (Alismataceae, Limnocharitaceae and Butomaceae) were kept in the Alismatales. Judd et al. (1999), through a phylogenetic approach using mainly the rbcL sequence, recognized two major clades in the Alismatales. The first comprises Alismataceae (including Limnocharitaceae), Hydrocharitaceae and Butomaceae; the second the Potamogetonaceae, Ruppiaceae, Zosteraceae, Posidoniaceae, Zannichelliaceae, Cymodocaceae and Najadaceae. The Araceae appears as a subclade and was considered sister to the remaining families of the order. As in Dahlgren & Clifford (1981), Limnocharitaceae was included in Alismataceae. Analysing the evolution of hydrophily in Alismatidae, Les, Cleland & Waycott (1997) obtained a consensus tree with the rbcL gene that grouped Alismataceae and Limnocharitaceae in the same clade, strongly supported by bootstrap values.

According to Judd et al. (1999), the inclusion of Limnocharitaceae within Alismataceae is also supported by some shared morphological features, such as presence of laticifers, hypogynous flowers, three to many carpels, one to many ovules, curved embryos, six to many stamens and biforaminate to multiforaminate pollen grains. Main differences between the two families, given by Tomlinson (1982), are the number of whorls of the carpels (two to many in Alismataceae and usually one in Limnocharitaceae), placentation (basal in Alismataceae and parietal in Limnocharitaceae) and fruits (indehiscent nutlets in Alismataceae and dehiscent follicles in Limnocharitaceae).

Members of Alismataceae, sensuCronquist (1988), have been studied cytologically by some authors (Harada, 1956; Sharma & Chatterjee, 1967; Lepper, 1982; Mehra & Pandita, 1984; Uchiyama, 1989), but just a few species, especially from the neotropics, are chromosomally documented. Haploid chromosome numbers of n = 5–13 have been reported, with 7–11 being the most common (Haynes & Holm-Nielsen, 1994). In both Echinodorus and Sagittaria, the number 2n = 22 is the most frequent (Harada, 1956; Sharma & Chatterjee, 1967; Mehra & Pandita, 1984).

The aquatic plants present many problems for traditional systematists. The loss or reduction of structures considered adaptative for terrestrial life, such as lignified tissue and cuticle, are commonplace and some characteristics, such as floating leaves, dissected submerged leaves, aerenchymatous tissue, etc., have arisen repeatedly among unrelated groups of aquatic angiosperms (Sculthorpe, 1967 and Les, 1991, cited in Les & Haynes, 1995). Furthermore, many morphological features exhibit extensive phenotypic plasticity that is influenced environmentally rather than genetically (Wooten, 1986; Scremin-Dias & Barros, 2001) and several leaf forms may occur on the same plant (Judd et al., 1999). All these elements complicate understanding of the homologies among hydrophytes, because the characters states have a high probability of being convergent or nongenetically based (Les & Haynes, 1995).

In order to trace the evolutionary karyotypic trends in the group as a whole, and to incorporate cytogenetic data for Alismataceae, we analysed karyotypes of a number of species of Echinodorus and Sagittaria and compared them with previous data available for the Alismatales sensuCronquist (1988).

MATERIAL AND METHODS

All species were collected in the states of São Paulo (SP) or Mato Grosso do Sul (MS), Brazil, and are presented in Table 1. Mitotic studies were made from root tips collected in the wild, pretreated in a mixture of saturated solution of paradichlorobenzene (PDB) and cycloheximide 0.009% (2 : 1) for 5 h at 16–18°C and fixed in 1 : 3 acetic-alcohol for 24 h. The tips were washed in distilled water, hydrolysed in 5 N HCl for 10 min at room temperature and squashed in a drop of 45% acetic acid. Slides were stained in 2% Giemsa (Guerra, 1983) and mounted permanently with Entellan. For calculating the size of the chromosomes at least 10 cells with good chromosome morphology, similar contraction and spreading ability were measured, and the mean length (L) was calculated for each pair. Satellites were considered only when observed in at least three cells of the same species. Nomenclature for centromeric position was that of Guerra (1986). The total chromatin length (TCL) and karyotypic symmetry (TF% rate), based on the short arm (S) and on chromosome length (L) – TF% = 100ΣS.ΣL⁻¹ (Huziwara, 1962), were calculated together with the CI (centromeric index) for a better comparison between karyotypes. Due to the small number of cells obtained, it was not possible to obtain all karyotypic data for Echinodorus bolivianus (Table 1, Fig. 3).

Table 1.

Data obtained for Alismataceae (Echinodorus and Sagittaria). Abbreviations: Pop, populations analysed; Nos, voucher numbers; 2n, somatic chromosome number; KF, karyotypic formula (m = metacentric, sm = submetacentric, a = acrocentric); CPSC, chromosome pair showing secondary constriction; RMCL, range in mean chromosome length (µm); TF%, symmetry index; TCL, total chromatin length (µm)

Species	Pop	Nos	2n	KF	CPSC	RMCL	TF%	TCL
E. aschersonianus Graebn.	III	J.Y.C. 1011	22*	2m + 20a	XI	2.8–7.5	16.1	87.7
E. bolivianus (Rusby) Holm-Niels.	VI	J.Y.C. 1101	22	2m + 20a	–	–	–	–
E. grandiflorus (Cham. & Schltdl.) Micheli	II, IV, V	J.Y.C. 1073, 091, 0103	22	2m + 20a	VI, XI	2.3–6.7	16.1	81.1
E. longipetalus Micheli	II, IV	J.Y.C. 1071, 1072	22*	4m + 18a	IV	3.7–9.1	20.5	121.1
E. macrophyllus (Kunth) Micheli	I		22*	2m + 20a	VIII, XI	2.9–6.7	12.9	86.6
E. pubescens (Mart.) Seub.	V	J.Y.C. 0102	22*	2m + 20a	XI	2.9–8.6	14.5	106.4
E. tennellus (Mart.) Buchenau	III	J.Y.C. 1075	22*	2m + 20a	–	2.9–8.5	15.2	100.8
S. montevidensis Cham. & Schltdl.	I		22	2m + 20a	XI	3.2–7.9	15.6	100.3
S. rhombifolia Cham.	II, IV, V	J.Y.C. 1012, 093,;1074, 0101	22*	2m + 2sm + 18a	IX	2.4–7.9	18.7	91.3

Populations: I Campinas, SP – 22°54′2020″S, 47 °3′39″ W; II Casa Branca, SP – 21°46′26″, 47 °05′11″; III Itapetininga, SP – 23°35′30″, 48°3′11″; IV Itirapina, SP – 22°15′10″, 47°49′22″; V Pereira Barreto, SP – 20°38′18″, 51°6′ 33″; VI Bonito, MS – 21°20′ 28″, 56°33′ 34″.

*

= first documentation.

−= not observed.

Figures 2–10.

Mitotic metaphases of Echinodorus and Sagittaria, all 2n = 22. Scale bars = 10 µm. Fig. 2. Echinodorus aschersonianus. Fig. 3. E. bolivianus. Fig. 4. E. grandiflorus. Fig. 5. E. longipetalus. Fig. 6. E. macrophyllus. Fig. 7. E. pubescens. Fig. 8. E. tennellus. Fig. 9. Sagittaria montevidensis.Fig. 10. S. rhombifolia. Arrows point to satellite position.

RESULTS

All species studied have the same chromosome number, 2n = 22 (Figs 2–10), with some karyotypic differences between species and no major differences between populations of the same species (Table 1). All show large chromosomes, with chromosome length varying gradually between species, but with each species strongly bimodal (Fig. 1). Chromosome size varies from 2.3 to 9.1 µm (Table 1), with Echinodorus grandiflorus presenting the smallest TCL (81.1 µm) and E. longipetalus the largest (121.1 µm).

Figure 1.

Ideograms of Alismataceae species, indicating satellite position when present.

Most of the chromosomes are acrocentric with exception of the largest pair, which is always metacentric. In E. longipetalus and S. rhombifolia, not only is the largest pair metacentric, but also the smallest, which is metacentric/submetacentric (Figs 1, 5, 10). This large number of acrocentric chromosomes is reflected by the low TF% rate shown by all species studied, varying from 14.5 in E. pubescens to 18.7 and 20.5 in S. rhombifolia and E. longipetalus, respectively (Table 1). This higher TF% rate shown by both species is the result of the presence of the second submetacentric/metacentric pair.

Secondary constrictions occur in at least one pair of acrocentric chromosomes, usually the smallest pair, but were not seen in E. tennellus. In E. longipetalus and S. rhombifolia the secondary constrictions are not in the smallest acrocentric pair (Figs 1, ¹⁰). E. macrophyllus and E. grandiflorus are the only species with two chromosome pairs having secondary constrictions (Table 1, Figs 1, 4, 6).

DISCUSSION

We have presented here karyotype documentations for six species (Table 1). With the exception of S. montevidensis, in which Taylor (1925) found 2n = 20 chromosomes, our findings also confirm previous investigations (Harada, 1956; Beal, 1960; Lepper, 1982; Uchiyama, 1989), with 2n = 22 chromosomes in all examined species of both genera and a basic number of x = 11. Another diploid number found in one species of the family, not available for the present study, is n = c. 14 (2n = c. 28) in E. subatalus (Coleman & Smith, 1969).

Together with the numerical data, other characteristics are shared between all species. All karyotypes present large chromosomes, strong bimodality (due to the presence of one pair of metacentric chromosomes that is larger than the other pairs) and low karyotype symmetry, as a consequence of the large numbers of acrocentric chromosomes. This pattern is also shown by E. cordifolius, S. calycina and S. eatoni (Beal, 1960), with only S. calycina presenting more than two metacentric/submetacentric pairs.

It is interesting to note that there are differences between karyotypes (Fig. 1), mainly in satellite position and morphology of the smallest chromosome pair. Some species present very similar karyotypes, as in E. aschersonianus and E. pubescens. They present the same satellite position and a small difference in chromosome symmetry index (TF%) (Table 1). Similarities were also observed between E. grandiflorus and E. macrophyllus, with different satellite position, but with chromosome length and TCL very similar. These species are easily confused by collectors (E.R. Pansarin, pers. comm.) and are separated mainly by leaf morphology, with the pellucid markings of E. grandiflorus absent in E. macrophyllus (Haynes & Holm-Nielsen, 1994).

Only E. longipetalus and S. rhombifolia have more than one pair of metacentric/submetacentric chromosomes. Similar karyotypes with the smallest chromosome pair also metacentric/submetacentric have been found in S. eatoni and S. calycina by Beal (1960) and in S. sagittifolia and S. guayanensis by Mehra & Pandita (1984).

Species of Alisma (Alismataceae), not found in the Neotropics, show large chromosomes, varying from 2.9 to 9.1 µm, and chromosome numbers from 2n = 14 to 26 (Mehra & Pandita, 1984). Contrasting with species of Echinodorus and Sagittaria that show mostly just one pair of metacentric chromosomes, Alisma plantago-aquatica and A. gramineum have five pairs of metacentric chromosomes and only two telocentric pairs, while A. lanceolatum (2n = 26) has 11 metacentric and two telocentric pairs (Mehra & Pandita, 1984). In this work, the authors concluded that the original basic number of Alismataceae was probably x = 6 metacentric chromosomes and a nombre fondamental (Matthey, 1945) of x = 12 arms in the ancestor. The number n = 7 would have originated from fission of one of the metacentric pairs, resulting in two pairs of telocentric chromosomes; n = 11 (genus Sagittaria) would be the result of fission at the centromeric region of all chromosomes but one. Alisma lanceolatum would have originated through polyploidy and fission of one pair of metacentrics (Mehra & Pandita, 1984).

Though the number of species studied by Mehra & Pandita (1984) was small, most karyotypes of Echinodorus and Sagittaria species studied here fit this hypothesis, with one pair of metacentric and 10 pairs of acrocentric chromosomes originating from six pairs of metacentric chromosomes through centric fission, keeping the nombre fondamental (NF) of x = 12 arms. Even in the previous documentation of 2n = 20 for S. montevidensis (Taylor, 1925), the NF is also x = 12 arms, with eight acrocentric pairs and the two largest chromosome pairs metacentric. On the other hand, the karyotypes of S. rhombifolia and E. longipetalus show a NF of x = 13 arms, which could be interpreted as a result of later chromosome repatternings, such as duplication or translocation events, giving rise to a second small pair of metacentric/submetacentric chromosomes.

In Limnocharitaceae, a family closely related to Alismataceae (Cronquist, 1988) and placed within it by Judd et al. (1999), the karyotypes also show large chromosomes but a gradual variation in length. Limnocharis flava and L. laforestii have 2n = 20 and Hydrocleys nymphoides 2n = 16 (Forni-Martins & Calligaris, 2002), with two metacentric pairs in both genera, plus a submetacentric pair in Hydrocleys. The same pattern of large chromosomes and high asymmetry was found, with sizes varying from 2.6 to 11.0 µm and their asymmetry index (TF%) from 17.8 to 19.6. Considering Limnocharis and Hydrocleys as members of Alismataceae (Judd et al., 1999) and following the evolutionary trends proposed by Mehra & Pandita (1984), the two genera could have originated from an ancestor with x = 6, through chromosome fission in four pairs, resulting in n = 10, with two metacentric pairs and eight acrocentric pairs in Limnocharis, and through chromosome fission and loss of an acrocentric pair in Hydrocleys, resulting in n = 8. The karyotypic evolution by chromosome fission in Limnocharitaceae was also suggested by Rao (1953).

In Butomaceae, chromosome numbers from 2n = 20 to 42 are documented, with length varying from 3.7 to 8.3 µm and karyotypes more bimodal than in Alismataceae (Rao, 1953; Harada, 1956; Sharma & Chatterjee, 1967). Members of the Hydrocharitaceae were found with a great variety of chromosome numbers, 2n = 16–72 (Harada, 1956), also with large chromosomes (1.6–10.0 µm), with a gradual range of size or with a bimodal distribution (Sharma & Chatterjee, 1967). Thus, Butomaceae and Hydrocharitaceae show karyotypic patterns similar to those of Alismataceae, especially regarding chromosome length, justifying the close relationship between these groups proposed by Judd et al. (1999) for Alismatales and by Les et al. (1997) for the Alismatidae group, from the analysis of rbcL sequences. The consensus tree obtained by Les et al. (1997) showed two major clades, the first containing a subclade comprising Alismataceae and Limnocharitaceae, and the other subclade containing Butomaceae, Hydrocharitaceae and Najadaceae.

In the second major clade, Les et al. (1997) placed ten families, with chromosome data available for two of them. In Potamogetonaceae, numbers are documented from 2n = 26 to c. 88 (Hollingsworth, Preston & Gornall, 1998), with small chromosomes (1.0–2.3 µm) varying gradually (Sharma & Chatterjee, 1967). Members of Aponogetonaceae have chromosome numbers of n = 12/24 (Harada, 1956), and 2n = 76, with small chromosomes (1.0–2.3 µm) (Sharma & Chatterjee, 1967). Thus, species belonging to this clade show smaller chromosomes and a higher ploidy level than individuals from the first clade.

All chromosome data obtained in the present work for Alismataceae, allied with data obtained from the Alismatidae group by different authors, establish a karyotypic pattern for the family, with large chromosomes and low symmetry, and the basic number x = 11 for both Echinodorus and Sagittaria. Following this pattern, chromosome data also support the incorporation of Limnocharitaceae within Alismataceae (Judd et al., 1999) and other arrangements proposed through molecular analysis (Les et al., 1997; Judd et al., 1999). Further morphological, anatomical, chemical, molecular and ecological studies are necessary for a better understanding of the group.

ACKNOWLEDGEMENTS

We thank Maria do Carmo Estanislau do Amaral and Emerson Ricardo Pansarin for species identification, Julie H. A. Dutilh for improvements to the manuscript, and Iara F. Bressan for helping with laboratory techniques. J.Y. Costa received a fellowship from the Fundação de Amparo à Pesquisa do estado de São Paulo (FAPESP ♯00/00767–0).