- Email: [email protected]
A phylogenetic analysis of subtribe Pleurothallidinae (Orchidaceae)
Botanical Journal of the Linnean Society (1995), 117: 13–28. With 3 figures
A phylogenetic analysis of subtribe Pleurothallidinae (Orchidaceae) RAY NEYLAND AND LOWELL E. URBATSCH Department of Plant Biology, Louisiana State University, Baton Rouge, Louisiana 70803-1705, U.S.A. AND ALEC M. PRIDGEON Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3DS Received May 1994, accepted for publication November 1994
A cladistic analysis of subtribe, Pleurothallidinae (Orchidaceae) is based on 45 anatomical: morphological characters. The ingroup members comprise 24 genera; the large genus Pleurothallis consists of two subgenera and ten species complexes. Three taxa representing subtribes Laeliinae and Arpophyllinae are designated as outgroup. Eight most parsimonious trees were discovered using computer assisted software (length 230; CI 0.27). The hypothesis that subtribe Pleurothallidinae has undergone a unilinear reduction in the number of pollinia is not supported by this study. Although the eight-pollinia state as represented by Octomeria apparently is plesiomorphic, the two-pollinia and four-pollinia states arose early in the phylogeny of the subtribe. Both twoand four-pollinia states subsequently reappeared as parallelisms. The six-pollinia state exhibited in Brachionidium is autapomorphic. This cladistic analysis suggests that Pleurothallis is not a natural genus and, perhaps may be divided into several discrete genera. ADDITIONAL KEY WORDS:—Pleurothallis – cladistics – phenetics – systematics.
CONTENTS Introduction . . . . . Methods . . . . . . Results . . . . . . Discussion . . . . . . Systematic overview . . Taxonomy of Pleurothallis . Evolution of pollinia number Acknowledgements . . . . References . . . . . . Appendix . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
13 14 16 16 17 23 25 26 26 26
INTRODUCTION
Pleurothallidinae Lindl. (Orchidaceae) is a Neotropical subtribe included in tribe Epidendreae (Dressler, 1993). Its members range from Brazil and Peru northward through Central America into Mexico and southern Florida; the 0024–4074/95/010013+16 $08.00/0
13
© 1995 The Linnean Society of London
14
NEYLAND ET AL.
greatest diversity of species occurs in cloud forest habitats (Pridgeon, 1982a,b). An articulation associated with an abscission zone between ovary and pedicel distinguishes the subtribe (Garay, 1956; Luer, 1986a; Dressler, 1993; Neyland & Urbatsch, 1993). In all other orchids the articulation occurs at the base of the pedicel (Dressler, 1981). Recent revisions by Luer (1986a, 1991) attributed 29 genera and as many as 4000 species to the subtribe. Luer acknowledged his treatment as artificial and was unable to bring phylogenetic order to the group. An earlier study by Pridgeon (1982a) used 45 anatomical:morphological characters to undertake a numerical analysis of the subtribe and suggested that the phenetic relationships identified in his study represented seven major phylogenetic lines. An important character in Pridgeon’s analysis is number of pollinia. He supported the hypothesis by Dressler (1981, 1993) that the Pleurothallidinae shows a reduction series in the number of pollinia from eight to six to four to two with eight representing the plesiomorphic condition. However, Luer (1986a) apparently believed that the direction may have proceeded in the opposite direction from two to eight. Most genera and an overwhelming number of species bear two pollinia; therefore, the relatively few members with more than two pollinia are considered pivotal in this study. For example, Octomeria bears eight pollinia, whereas Restrepia, Restrepiella, Restrepiopsis, Barbosella and Dresslerella each bear four. According to Garay (1956), Brachionidium has six; however, Luer (1986a) included the eight-pollinia species of the invalid genus, Yolanda, in Brachionidium. According to Luer (1986c), 29 subgenera comprise the very large genus Pleurothallis to which over 2000 epithets have been attributed. In Pridgeon’s (1982a) numerical analysis, two subgenera and 10 species complexes were recognized. Dressler predicted (personal communication 1992), that Pleurothallis eventually may be divided into several distinct genera corresponding with their three basic habits. Most species have the Specklinia habit with short stems and oblanceolate leaves that taper basally. Some have the Acianthera habit with long stems and relatively wide leaf bases. The Pleurothallis cardiothallis group has long stems, cordate leaves, and fascicled flowers. The primary objective of this study is to infer a hypothesis of phylogenetic relationship for Pleurothallidinae by combining the character-state data matrix constructed for taxa studied in Pridgeon’s (1982a) phenetic analysis (here designated as the ingroup) with a data matrix constructed for an appropriate outgroup and subjecting the combined data sets to cladistic analysis. The resulting cladograms are used to: (1) discuss the cladistic relationships among the genera and Pleurothallis species complexes of Pleurothallidinae; (2) evaluate whether the genus Pleurothallis represents a monophyletic group or is polyphyletic and thus can be divided into several genera, and (3) discuss the trend in character evolution with respect to pollinia number.
METHODS
The ingroup members in this study are those used as terminal taxa in Pridgeon’s (1982a) phenetic analysis. These taxa consist of 24 genera; Pleurothallis consists of two subgenera further divided into 10 species complexes. For a
PLEUROTHALLIDINAE
15
detailed list of ingroup material examined see Pridgeon (1982a). Pridgeon’s classification scheme is presented in Table 1. Three taxa were selected as outgroup members based on their status as representing possible sisters to subtribe Pleurothallidinae. These include: Arpophyllum giganteum Hartw., (Arpophyllinae); Brassavola nodosa (L.) Lindl., (Laeliinae) and Epidendrum ciliaris L., (Laeliinae). Voucher specimens are housed at Louisiana State University (LSU) (Table 2). According to Dressler (1960), Arpophyllum appears to represent a survivor of the ancestral stock from which the Pleurothallidinae were derived. Pridgeon (1982a) stated that it is plausible that Octomeria (an ingroup member) and Arpophyllum have a common ancestor and represent divergent evolutionary lines from that progenitor. According to Dressler (personal communication 1992) and Pridgeon (personal communication
TABLE 1. Classification scheme of Pleurothallidinae from Pridgeon (1982b). — –––––––––––––––––––––––––––––––––––––––––– Number of pollinia Taxon –––––––––––––––––––––––––––––––––––––––––––––––– — 8 Octomeria Pleurothallopsis 6 Brachionidium Chamelophyton 4 Barbosella Dresslerella Restrepia Restrepiella Restrepiopsis 2 Acostaea Andreettaea Condylago Crocodeilanthe Cryptophoranthus Dracula Dryadella Lepanthes Lepanthopsis Masdevallia Phloeophila Physothallis Platystele Pleurothallis subg. Pleurothallis P. alexandrae complex P. blaisdellii complex P. dura complex P. loranthophylla complex P. peduncularis complex P. ruscifolia complex P. scoparum complex P. tuerckheimii complex P. subg. Specklinia P. grobyi complex P. sclerophylla complex Porroglossum Salpistele Scaphosepalum Stelis Trisetella –––––––––––––––––––––––––––––––––––––––––––––––– —
16
NEYLAND ET AL.
TABLE 2. Outgroup members examined ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Accession Taxon Source Number ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Arpophyllum giganteum Hartw. University of Connecticut Neyland 90 Brassavola nodosa (L.) Lindl. Louisiana State University Neyland 72 Epidendrum ciliare L. Louisiana State University Neyland 81 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
1992), subtribe, Laeliinae may also be considered a likely outgroup to the Pleurothallidinae. Furthermore, a close relationship with Laeliinae is also suggested by the cladistic analysis of Burns-Balogh and Funk (1966), which suggested that the tribe Epidendreae is sister to the ‘Pleurothallis Group’. Forty-five multi-state and binary characters were used in this analysis (Appendix). Character states assigned to the ingroup members are those used in Pridgeon’s (1982a) numerical analysis. These characters were expected to yield phylogenetically significant information for the reasons detailed in Pridgeon’s (1982b) diagnosis. Character states were determined for the outgroup members using light microscopy, scanning electron microscopy (SEM), direct observation, and literature sources. Character states for all are summarized in Table 3. Multi-state transformation series were unordered so that character changes could occur without arbitrary constraints. Cladistic analyses performed using PAUP (version 3.0s, Swofford, 1991). Multi-state characters were unordered using Fitch criterion (Fitch, 1971). Most parsimonious trees were sought from a heuristic stepwise addition search employing 100 replications. MacClade (version 3.0, Maddison and Maddison, 1992) was used to explore hypothesized phylogenetic trees further and to visualize character evolution. RESULTS
Using the three designated taxa as outgroup, eight most parsimonious cladograms were discovered. The data matrix included 36 phylogenetically informative characters. All cladograms have 230 steps and a consistency index (CI) of 0.27. A strict consensus tree is presented in Figure 1. In six of the parsimony trees, the Pleurothallis blaisdellii, P. sclerophylla, P. tuerckheimii and P. loranthophylla complexes form a clade that is distinct from the clade composed of Cryptophoranthus, the Pleurothallis dura and P. ruscifolia complexes, Lepanthopsis and Restrepia. In the remaining two trees, the two clades are combined into a single clade. One most parasimonious tree representing the topology where the two clades are distinct (reflecting six of the eight resulting topologies) is presented in Figure 2. DISCUSSION
The consistency index of 0.27 represents a high level of homoplasy and may be a function of the rapid rates of evolution believed to have occurred in Orchidaceae (Dressler, 1993). However, the use of characters with continuous states (characters: 6, 7, 11, 12, 15) and polymorphic states (characters: 15, 17, 18, 19, 30, 36, 39) used in this study may also have contributed to this
PLEUROTHALLIDINAE
17
level of homoplasy (Appendix). Branch lengths representing unambiguous synapomorphies are generally considered a measure of clade stability. As indicated by one of the most parsimonious cladograms, nodal support as defined by the number of synapomorphies is often weak (Fig. 2). This level of support is due in part to the low ratio of terminal taxa (34) to characters (45). Despite the high level of homoplasy and generally weak branch support, the data set used in this study brings a substantial level of resolution to the ingroup and provides a cladistic framework for a phylogenetic discussion of Pleurothallidinae. Systematic overview Octomeria and the Pleurothallis peduncularis complex form the most basal clade in this study (Fig. 2). The position of Octomeria is consistent with the hypothesis TABLE 3. Character states for Pleurothallidinae and outgroup members. Characters and character states are identified in Appendix. Missing data are indicated by question marks — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Taxa 1 2 3 4 5 6 7 8 9 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Ingroup Acostaea 2 0 0 0 0 2 2 0 1 Barbosella 2 0 0 0 0 0 1 0 0 Brachionidium 1 0 0 0 0 2 2 1 0 Condylago 2 0 0 0 0 2 0 1 1 Cryptophoranthus 2 0 0 0 1 1 2 1 1 Dracula 2 0 0 0 0 2 2 1 1 Dresslerella 2 1 0 0 0 2 2 0 0 Dryadella 2 0 0 0 0 1 2 1 1 Lepanthes 2 0 0 0 0 2 2 0 1 Lepanthopis 2 0 0 0 1 2 0 1 1 Masdevallia 2 0 0 0 1 2 2 0 1 Physosiphon 2 0 0 0 1 2 2 0 1 Physothallis 2 0 0 0 0 1 1 0 1 Platystele 2 0 0 0 0 2 2 1 1 Pleurothallis ruscifolia complex 2 0 0 0 0 1 1 1 1 Pleurothallis alexandrae complex 0 0 1 0 0 0 1 0 0 Pleurothallis blaisdellii complex 2 0 0 0 1 1 0 0 1 Pleurothallis dura complex 2 0 0 0 1 1 2 1 1 Pleurothallis grobyi complex 2 0 0 0 0 1 0 1 1 Pleurothallis loranthophylla complex 2 0 0 0 1 1 1 1 1 Pleurothallis peduncularis complex 0 0 0 1 1 1 1 1 0 Pleurothallis sclerophylla complex 2 0 0 0 0 2 0 1 1 Pleurothallis scoparium complex 2 0 0 0 0 1 2 1 1 Pleurothallis subgen. Specklinia 2 0 0 0 0 2 1 1 1 Pleurothallis tuerckheimii complex 2 0 0 0 0 2 1 1 1 Porroglossum 2 0 0 0 0 1 2 0 1 Restrepia 2 0 0 0 0 1 1 0 1 Restrepiella 0 0 0 0 1 0 1 1 0 Restrepiopis 1 0 0 0 0 1 2 1 0 Salpistele 2 0 0 0 1 2 2 1 1 Scaphosepalum 2 0 0 0 0 2 2 1 1 Stelis 2 0 0 0 1 2 2 1 1 Tristella 2 0 0 0 1 2 2 1 1 Octomeria 0 0 0 0 0 0 0 0 0 Outgroup Arpophylum giganteum 0 0 0 0 0 0 1 0 0 Brassavola nodosa 0 0 0 0 1 1 2 0 1 Epidendrum ciliaris 0 0 0 0 1 1 0 1 1 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
18
NEYLAND ET AL.
TABLE 3 (continued). — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 1 1 1 1 1 1 1 1 1 Taxa 0 1 2 3 4 5 6 7 8 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Ingroup Acostaea 1 0 1 0 0 1 1 1 3 Barbosella 1 0 1 0 0 2 1 0 1 Brachionidium 1 1 3 0 0 0 0 0 3 Condylago 1 0 1 0 0 1 1 1 3 Cryptophoranthus 1 0 0 0 0 3 1 1 0 Dracula 1 0 1 0 0 0 0 0 3 Dresslerella 0 2 3 0 0 3 1 0 0 Dryadella 1 0 3 0 0 2 1 2 1 Lepanthes 1 0 0 0 0 2 1 0 0 Lepanthopis 1 0 1 0 0 2 1 1 0 Masdevallia 1 1 2 0 0 3 1 2 1 Physosiphon 1 1 2 0 0 2 1 1 2 Physothallis 1 2 0 0 0 2 1 1 3 Platystele 1 0 1 0 1 2 1 1 3 Pleurothallis ruscifolia complex 1 0 1 0 0 2 1 1 0 Pleurothallis alexandrae complex 0 2 3 0 0 2 1 0 0 Pleurothallis blaisdellii complex 1 1 1 0 0 1 1 1 0 Pleurothallis dura complex 1 1 0 0 0 2 1 1 0 Pleurothallis grobyi complex 1 0 0 0 0 2 1 2 2 Pleurothallis loranthophylla complex 1 1 0 0 0 3 1 1 1 Pleurothallis penduncularis complex 1 1 2 1 0 2 1 0 0 Pleurothallis sclerophylla complex 1 1 1 0 0 2 1 1 1 Pleurothallis scoparium complex 1 1 1 0 0 3 1 2 2 Pleurothallis subgen. Specklinia 1 0 0 0 0 2 1 0 1 Pleurothallis tuerckheimii complex 1 1 2 0 0 2 1 1 0 Porroglossum 1 1 2 0 0 3 1 2 0 Restrepia 1 0 1 0 0 1 1 1 0 Restrepiella 1 0 0 0 0 2 1 0 0 Restrepiopsis 1 0 0 0 0 2 1 1 0 Salpistele 1 0 1 0 0 1 1 1 3 Scaphosepalum 1 0 1 0 0 3 1 2 1 Stelis 1 0 2 0 0 2 1 1 0 Trisetella 1 1 2 0 0 1 1 2 2 Octomeria 1 1 2 0 0 2 1 0 2 Outgroup Arpophyllum giganteum 0 1 3 0 0 0 0 0 0 Brassavola nodosa 0 2 3 0 0 1 1 0 0 Epidendrum ciliaris 0 1 3 0 0 0 0 0 0 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
that this taxon represents the most basal element in Pleurothallidinae. The apparently close relationship between Octomeria and the P. peduncularis complex was suggested by Pridgeon (1982a). The P. alexandrae complex appears as sister to the Octomeria-P. peduncularis clade. Pridgeon (1982a) suggested a close relationship among these taxa and distinguished the P. alexandrea complex from the P. peduncularis complex by floral specializations that led to fly pollination. Ascending the cladogram, a grade of genera exhibiting four pollinia (Dresslerella, Barbosella, Restrepiella, Restrepiopsis) are next encountered (Fig. 2). A close association among all Pleurothallidinae exhibiting four pollinia was suggested by Pridgeon (1982a). However, Restrepia (also exhibiting four pollinia) appears in a separate clade and thus appears not closely related to taxa in this grade (Fig. 2). The Cryptophoranthus, Lepanthopsis, Pleurothallis dura complex, Pleurothallis ruscifolia complex and Restrepia clade is distinguished in this study by stem cuticles ³5 mm
PLEUROTHALLIDINAE
19
TABLE 3 (continued). — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 1 2 2 2 2 2 2 2 2 Taxa 9 0 1 2 3 4 5 6 7 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Ingroup Acostaea 0 1 0 1 0 0 1 1 1 Barbosella 0 1 1 0 1 0 1 0 1 Brachionidium 0 1 0 2 1 0 ? ? ? Condylago 0 1 0 2 1 0 1 1 1 Cryptophoranthus 2 1 0 2 0 0 1 0 1 Dracula 0 1 1 2 1 0 1 1 1 Dresslerella 0 1 0 2 1 0 1 0 1 Dryadella 1 1 1 2 0 0 1 1 1 Lepanthes 0 1 0 2 1 0 1 1 1 Lepanthopsis 2 1 0 2 1 0 ? ? ? Masdevallia 2 1 1 2 1 0 1 1 1 Physosiphon 1 1 0 2 1 0 1 1 1 Physothallis 0 1 0 2 0 0 0 1 1 Platystele 1 1 0 0 0 0 1 1 1 Pleurothallis ruscifolia complex 2 1 0 2 1 0 1 0 0 Pleurothallis alaxandrae complex 0 0 0 2 1 0 0 0 1 Pleurothallis blaisdellii complex 1 1 0 2 1 0 0 0 1 Pleurothallis dura complex 2 1 0 2 0 0 ? ? ? Pleurothallis grobyi complex 0 1 0 2 0 0 1 1 1 Pleurothallis loranthophylla complex 2 1 0 2 1 0 0 1 1 Pleurothallis peduncularis complex 0 0 0 2 1 0 1 0 0 Pleurothallis sclerophylla complex 1 1 0 2 1 0 0 1 0 Pleurothallis scoparium complex 1 1 0 2 1 1 1 1 1 Pleurothallis subgen Specklinia 0 1 0 2 0 0 1 1 1 Pleurothallis tuerckheimii complex 1 1 0 1 1 0 1 1 0 Porroglossum 1 1 1 2 0 0 1 1 1 Restrepia 2 1 0 2 1 0 1 0 1 Restrepiella 2 1 1 2 1 0 0 0 1 Restrepiopsis 2 1 0 1 1 0 ? ? ? Salpistele 0 1 0 1 0 0 ? ? ? Scaphosepalum 1 1 0 2 1 0 1 1 1 Stelis 2 1 1 2 0 0 1 1 1 Trisetella 1 1 1 2 1 0 1 1 1 Octomeria 0 0 0 2 0 0 0 0 0 Outgroup Arpophyllum giganteum 0 0 0 2 1 0 0 0 0 Brassavola nodosa 0 0 0 2 0 0 0 1 0 Epidendrum ciliaris 2 0 1 2 1 0 0 0 0 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
(character 25), and a variable number of root protoxylem poles (character 34). The position of Lepanthopsis in this clade suggests this taxon is not closely allied with Lepanthes as was suggested by Pridgeon (1982a), who distinguished Lepanthes from Lepanthopsis primarily by the loss of spirally thickened idioblasts. Both genera are vegetatively similar with respect to lepanthiform (tubular) sheaths of the ramicaul. The Pleurothallis blaisdelii, P. sclerophylla, P. tuerckheimii, and P. loranthophylla species complexes form a clade distinguished in this study by the absence of a column foot (character 42) and by a variable degree of fusion between the labellum and column (character 39). A single synapomorphy in six of the eight most parsimonious trees discovered separates these two clades. Therefore, this distinction between the two clades is considered weak. Consisting of two terrestrial species Physothallis appears as sister to the Pleurothallis grobyi complex. Claiming that the deeply connate dorsal sepal is
20
NEYLAND ET AL.
TABLE 3 (continued). — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 2 2 3 3 3 3 3 3 3 Taxa 8 9 0 1 2 3 4 5 6 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Ingroup Acostaea 1 1 1 0 1 0 2 1 0 Barbosella 1 1 0 0 1 0 2 1 2 Brachionidium ? ? ? ? ? ? ? ? 1 Condylago 0 1 ? ? ? ? ? ? 0 Cryptophoranthus 0 1 2 0 1 0 1 1 0 Dracula 1 1 0 0 1 0 0 1 0 Dresslerella 0 0 1 3 1 1 2 1 0 Dryadella 1 1 2 2 1 0 2 1 0 Lepanthes 1 1 ? ? ? ? ? ? 1 Lepanthopsis ? ? ? ? ? ? ? ? 0 Masdevallia 1 1 3 0 1 0 1 1 0 Physosiphon 0 1 2 0 1 0 2 1 0 Physothallis 0 1 ? ? ? ? ? ? 0 Phytystele 1 1 2 0 1 0 2 1 0 Pleurothallis ruscifolia complex 0 1 0 3 0 1 1 1 1 Pleurothallis alexandrae complex 0 1 0 1 1 0 0 0 0 Pleurothallis blaisdellii complex 0 1 ? ? ? ? ? ? 0 Pleurothallis dura complex ? ? ? ? ? ? ? ? 0 Pleurothallis grobyi complex 1 1 ? ? ? ? ? ? 0 Pleurothallis loranthophylla complex 0 1 ? ? ? ? ? ? 0 Pleurothallis peduncularis complex 0 1 3 0 1 0 0 0 0 Pleurothallis sclerophylla complex 0 1 ? ? ? ? ? ? 0 Pleurothalis scoparium complex 0 1 ? ? ? ? ? ? 0 Pleurothallis subgen. Specklinia 1 1 0 0 1 0 2 0 1 Pleurothallis tuerckheimii complex 0 1 0 1 1 0 2 1 0 Porroglossum 1 1 3 3 1 0 1 1 1 Restrepia 0 1 3 0 1 1 1 1 0 Restrepiella 0 1 0 2 1 0 0 1 0 Restrepiopsis ? ? ? ? ? ? ? ? 0 Salpistele ? ? ? ? ? ? ? ? 0 Scaphosepalum 1 1 3 3 1 1 2 1 2 Stelis 0 1 0 2 1 0 2 1 1 Trisetella 1 1 2 0 1 0 2 1 0 Octomeria 0 1 1 0 1 0 0 0 0 Outgroup Arpophyllum giganteum 0 0 0 0 1 0 0 0 2 Brassavola nodosa 0 1 0 0 1 0 0 0 0 Epidendrum ciliaris 1 1 0 0 1 0 2 0 0 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
insufficient for defining this taxon at the generic level, Luer (1986a) relegated Physothallis to subgeneric status in Pleurothallis. The clade composed of the Pleurothallis subgen. Specklinia and Stelis is distinguished in this study by differentiated leaf mesophyll (character 18), incumbent anther (character 43), and variability with respect to resupination (character 36). Although the large genus Stelis (×500 species) appears to be a natural genus, the great similarity in floral morphology has made taxon distinction at the species level difficult. Pridgeon (1982a) suggested a close relationship existed between Stelis and Brachionidium on the basis that both taxa share a footless and bibrachiate column. However, this cladistics analysis suggests the two taxa represent disparate elements in Pleurothallidinae (Fig. 2). The large clade composed of Dryadella, Masdevallia, Porroglossum, Trisetella, Physosiphon, Scaphosepalum Platystele and the Pleurothallis scoparium complex is
PLEUROTHALLIDINAE
21
TABLE 3 (continued). — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 3 3 3 4 4 4 4 4 4 Taxa 7 8 9 0 1 2 3 4 5 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Ingroup Acostaea 2 0 2 1 0 2 0 3 0 Barbosella 2 0 2 0 0 2 0 2 0 Brachionidium 0 0 2 0 0 0 0 1 1 Condylago 2 0 2 1 0 2 0 3 0 Cryptophoranthus 2 1 2 0 0 2 0 3 0 Dracula 2 0 1 0 1 2 0 3 0 Dresslerella 2 0 2 0 0 2 0 2 0 Dryadella 2 0 2 0 0 2 0 3 0 Lepanthes 2 0 0 0 0 0 2 3 0 Lepanthopsis 2 0 1 0 0 0 0 3 1 Masdevallia 2 0 2 0 0 2 0 3 0 Physosiphon 2 0 2 0 0 2 0 3 0 Physothallis 2 0 2 0 0 2 0 3 0 Platystele 2 0 1 0 0 1 0 3 1 Pleurothallis ruscifolia complex 1 0 1 0 0 1 0 3 0 Pleurothallis alexandrae complex 2 0 2 0 0 2 0 3 0 Pleurothallis blaisdellii complex 2 0 2 0 0 2 0 3 0 Pleurothallis dura complex 2 0 2 0 0 2 0 3 0 Pleurothallis grobyi complex 2 0 2 0 0 2 0 3 0 Pleurothallis loranthophylla complex 0 0 2 0 0 2 0 3 0 Pleurothallis peduncularis complex 2 0 2 0 0 2 0 3 0 Plerurothallis sclerophylla complex 2 0 2 0 0 2 0 3 0 Pleurothallis scoparium complex 2 0 2 0 0 2 0 3 0 Pleurothallis subgen. Specklinia 2 0 2 0 0 2 1 3 0 Pleurothallis tuerckheimii complex 2 0 2 0 0 2 0 3 0 Porroglossum 2 0 2 1 0 2 1 3 0 Restrepia 2 0 0 0 0 0 0 2 0 Restrepiella 2 0 2 0 0 2 0 2 0 Restrepiopsis 2 0 2 0 0 2 0 2 0 Salpistele 2 0 0 0 0 0 3 3 0 Scaphosepalum 2 0 2 0 0 2 0 3 0 Stelis 2 0 2 0 0 1 1 3 1 Trisetella 2 0 2 0 0 2 0 3 0 Octomeria 0 0 1 0 0 1 0 0 0 Outgroup Arpophyllum giganteum 0 0 0 0 0 2 0 0 0 Brassavola nodosa 2 0 0 0 0 0 0 0 0 Epidendrum ciliaris 2 0 0 0 0 0 0 2 0 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
distinguished in this study by single-celled velamen (character 30) and variability with respect to spirally thickened idioblasts (character 19). Platystele appears as the most basal taxon in this clade. Consisting of approximately 73 species, this genus is distinguished by Luer (1986a) as having a lateral racemose inflorescence and ramicauls shorter than the leaves. Luer (1986a) considered Platystele closely related to Pleurothallis. Dryadella consists of about 40 species that formally resided in Masdevallia based on superficial similarities (Luer, 1986a). As suggested by the cladogram, Dryadella does not appear closely allied with Masdevallia. Scaphosepalum (approximately 30 species), distinguished by Luer (1988) as having a pair of cushion-like osmophores positioned on the lateral sepals, appears as sister to the Pleurothallis scoparium complex (Fig. 2). According to Luer (1986a), Trisetella (approximately 15 species) is distinguished by a column with a hooded ventral anther and was previously treated as a section of Masdevallia. Results
22
NEYLAND ET AL.
Figure 1. Strict consensus tree of eight trees discovered from a parsimony search for the taxa included in this study.
from this study indicate that Trisetella and Masdevallia are closely allied, and the treatment of Trisetella as a genus is justified. Masdevallia (approximately 350 species) is distinguished by Luer (1986b) as having calliferous petals and a lip hinged to a free incurved extension from the apex of the column foot. Luer (1987) stated that Porroglossum resembles Masdevallia by having short ramicauls and sepals with long tails; however, its sensitive labellum sufficiently distinguishes it from Masdevallia. The results of this study indicate that Masdevallia and Porroglossum are sister taxa. The clade composed of Acostaea, Salpistele, Brachionidium, Dracula and Condylago is distinguished in this study by having 1–3 adaxial hypodermal layers in the leaf (character 15). Pridgeon (1982a) suggested that Brachionidium was closely allied with Octomeria and is distinguished from Octomeria by the loss of two pollinia. However, the relatively derived position of Brachionidium suggests the two taxa are disparate elements in Pleurothallidinae (Fig. 2). Pridgeon (1992a) considered Dracula to be closely allied with Brachionidium and distinguished Dracula from Brachionidium by the loss of four pollinia, perianth specialization, and
PLEUROTHALLIDINAE
23
Figure 2. One of the eight most parsimonious trees discovered from a parsimony search for the taxa included in this study. The value above each branch represents the number of unambiguous synapomorphies.
presence of glandular trichomes on leaf surfaces. This cladistic analysis indicates that Dracula and Brachionidium are sister taxa (Fig. 2). Acostaea, Condylago, and Porroglossum each exhibit sensitive labella. According to Luer (1987), because the labellum in each of these genera is structurally and functionally distinct, these taxa are not closely related. As suggested by the results of this study, the sensitive labellum arose independently within each of these three taxa (Fig. 2). Luer (1991) stated that Salpistele may be closely allied with Lepanthes; however, he considered the trumpet-shaped column and lack of lepanthiform sheaths sufficient to distinguish Salpistele as a genus. This cladistic analysis suggests that Salpistele is a sister to Acostaea and not closely related to Lepanthes (Fig. 2). Taxonomy of Pleurothallis The results of this cladistic study suggests that the large genus Pleurothallis is polyphyletic and, therefore, may be divided into several genera. One genus
24
NEYLAND ET AL.
would be represented by the P. peduncularis complex that was segregated as Myoxanthus by Luer (1992). The members of this genus are characterized by well-developed ramicauls with scruffy sheaths and short rhizomes. A second genus may be represented by the P. alexandrae complex which was elevated to Myoxanthus subgenus Silenia by Luer (1992). Both complexes occupy a relatively basal position within the subtribe and are far removed from the more derived Pleurothallis complexes. The basis for separating these two closely related complexes is that this analysis indicates that the P. peduncularis complex appears more closely related to Octomeria than to the Pleurothallis alexandrae complex (Fig. 2). Among the more derived Pleurothallis complexes, P. dura and P. ruscifolia share a close affinity and occupy the same clade. Luer (1986a) elevated the P. dura complex to Trichosalpinx subg. Tubella. Members of this subgenus produce short ramicauls, connate lateral sepals and ciliate petals. Luer (1986c) placed the P. ruscifolia complex in Pleurothallis subg. Pleurothallis sect. Pleurothallis subsect. Pleurothallis series Pleurothallis. Members of this series produce flowers borne in short-fascicled racemes. Because this complex includes the type species P. ruscifolia ( Jacq.) R. Br., the name Pleurothallis must be retained. This cladistic analysis indicates that both complexes are more closely related to other genera than to each other. Therefore, Luer’s placement of the P. dura complex within Trichosalpinx seems justified (Fig. 2). The Pleurothallis loranthophylla, P. tuerckheimii, P. sclerophylla and P. blaisdellii complexes constitute a single clade. Luer (1986c) placed the P. loranthophylla complex in Pleurothallis subg. Rhynchopera. Luer characterized this subgenus as an easily recognizable homogenous group with well-developed ramicauls that bear a petiolate leaf; the simple lip is firmly united to the short, pedestal-like foot of the column. The P. tuerckheimii complex is included in Pleurothallis subg. Dracontia in Luer’s treatment. Its well-developed ramicauls bear a tubular sheath near the middle and are subtended by two or three other sheaths at the base; the lip is thick with erect, thin, wing-like basal lobes. Luer placed the P. sclerophylla complex in Pleurothallis subg. Specklinia sect. Acuminatae. In that section, as in Pleurothallis subg. Dracontia, tubular sheaths also are inserted on the ramicaul. However, the lip is oblong-ligulate often with a pair of calli; the base is simply hinged to the stout column foot. The P. blaisdellii complex is placed in Trichosalpinx subg. Trichosalpinx by Luer (1986a). That subgenus is characterized by stout ramicauls, free lateral sepals and petals with entire margins. Because these four complexes are grouped in the same clade, justification for their taxonomic separation into discrete genera is uncertain. If considered a single taxonomic unit, the clade appears to warrant elevation to generic status (Fig. 1). Moving toward the most derived taxa, the Pleurothallis grobyi complex, P. subg. Specklinia, and the P. scoparium complex show little affinity and are thus logical candidates for generic status (Fig. 2). The P. grobyi complex was placed within Pleurothallis subg. Specklinia sect. Hymenodanthae by Luer (1986c). Luer described that section as having numerous, closely allied caespitose species with an abbreviated ramicaul and racemose inflorescence that is usually longer than the leaf. Pleurothallis subg. Specklinia includes numerous species characterized by the lip hinged to the column foot. The column is well developed; its apex covers the anther, rostellum, and stigma. Luer placed the P. scoparium complex within
PLEUROTHALLIDINAE
25
Pleurothallis subg. Scopula. He attributed four species to the subgenus; they are characterized by elliptical leaves with a tuft of single-flowered peduncles emerging from what appears to be the midrib near the leaf apex. Evolution of pollinia number The number of pollinia (character 44) are plotted onto one most parsimonious tree to illustrate character state changes (Fig. 3). Equivocal branches were resolved by the MacClade (Maddison & Maddison, 1992) equivocal cycling feature. This analysis suggests that although the eight-pollinia state may represent the primitive condition as espoused by Dressler and Pridgeon, the number of pollinia did not follow a unilinear reduction series. Instead it appears that the eight pollinia state represented by Octomeria, the two-pollinia state represented by the P. peduncularis and P. alexandrae complexes, and the four-pollinia state represented by the grade comprising Restrepiopsis, Restrepiella, Barbosella and Dresslerella all arose early in the phylogeny of the subtribe. The two-pollinia state reappears as a parallelism and is represented by most of the genera and a vast majority of species within the subtribe. Because Restrepia appears in a clade removed from the other four-pollinia taxa, its character state also is considered a parallelism. The six-pollinia condition of Brachionidium represents an autapomorphy. This study represents the first attempt to bring cladistic resolution to the
Figure 3. Character state tree showing the distribution of states for the number of pollinia (character 44) for the taxa included in this study.
26
NEYLAND ET AL.
subtribe Pleurothallidinae. Because the characters used in this study exhibit a high level of homoplasy and branch nodes on cladograms are often weakly supported, the results of this analysis are considered tentative. Other data to include molecular, cytological and palynological characters should be evaluated in order to clarify further the phylogeny of this vast subfamily. ACKNOWLEDGEMENTS
We thank Carlyle A. Luer for providing technical assistance; William Cullina, University of Connecticut, for supplying plant material for the project; and Meredith Blackwell, Robert Burckhalter, and Shirley Tucker for reviewing the manuscript. Support from the LSU Plant Biology Department is gratefully acknowledged. REFERENCES Dressler RL. 1960. The relationships of Meiracyllium (Orchidaceae). Brittonia 12: 222–225. Dressler RL. 1981. The Orchids: Natural History and Classification. Cambridge, Massachusetts: Harvard University Press. Dressler RL. 1993. Phylogeny and Classification of the Orchid Family. Portland, Oregon: Dioscorides Press. Fitch WM. 1971. Toward defining the course of evolution: minimal change for a specific tree topology. Systematic Zoology 20: 406–416. Garay LA. 1956. Studies in American orchids. II. The genus Brachionidium Lindl. Canadian Journal of Botany 34: 721–743. Luer CA. 1986a. Icones Pleurothallidinarum I. Systematics of the Pleurothallidinae. St. Louis: Missouri Botanical Garden. Luer CA. 1986b. Icones Pleurothallidinarum II. Systematics of Masdevallia. St. Louis: Missouri Botanical Garden. Luer CA. 1986c. Icones Pleurothallidinarum III. Systematics of Pleurothallis. St. Louis: Missouri Botanical Garden. Luer CA. 1987. Icones Pleurothallidinarum IV. Systematics of Acostaea Condylago and Porroglossum. St. Louis: Missouri Botanical Garden. Luer CA. 1988. Icones Pleurothallidinarum V. Systematics of Dresslerella and Scaphosepalum. St. Louis: Missouri Botanical Garden. Luer CA. 1991. Icones Pleurothallidinarum VIII. Systematics of Lepanthopsis, Octomeria subgenus Pleurothallopsis, Restrepiella, Restrepiopsis, Salpistele, and Teagueia. St. Louis: Missouri Botanical Garden. Luer CA. 1992. Icones Pleurothallidinarum IX. Systematics of Myoxanthus. St. Louis: Missouri Botanical Garden. Maddison WP, Maddison, DR. 1992. MacClade: Analysis of phylogeny and character evolution. Version 3.0. Sunderland, Massachusetts: Sinauer Associates. Neyland R, Urbatsch LE. 1993. Anatomy and morphology of the articulation between ovary and pedicel in Pleurothallidinae. Lindleyana. 8(4): 189–192. Pridgeon AM. 1982a. Numerical analyses in the classification of the Pleurothallidinae (Orchidaceae). Botanical Journal of the Linnean Society 85: 103–131. Pridgeon AM. 1982b. Diagnostic anatomical characters in the Pleurothallidinae (Orchidaceae). American Journal of Botany 69: 921–938. Swofford DL. 1991. Phylogenetic Analysis Using Parsimony. Version 3.0s. Champaign, Illinois: Illinois Natural History Survey.
APPENDIX ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Attribute Character State ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — 1. Glandular trichomes (leaf ) (0) Absent (1) Limited to one surface (2) Present on both surfaces 2. Epidermal papillae (leaf ) (0) Absent (1) Present 3. Multicellular muscilage trichomes (leaf ) (0) Absent (1) Present 4. Hispid leaf sheaths (leaf ) (0) Absent (1) Present
PLEUROTHALLIDINAE
27
APPENDIX (continued). ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Attribute Character State ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — 5. Cuticle surface (leaf ) (0) Smooth (1) Papillose 6. Cuticle thickness (leaf ) (0) 15–22 mm (1) 6–14 mm (2) ³3 mm 7. Epidermal radial wall thickness (leaf ) (0) ×5–7 mm (1) 3–5 mm (2) ³3 mm 8. Epidermal cell wall thickening (0) Differentially thickened (leaf in transection) (1) Uniformly thickened 9. Epidermal cell shape (0) Dome-shaped (1) Not dome-shaped 10. Level of stomatal apparatus (leaf ) (0) Raised above the epidermis (1) Flush with the epidermis 11. Mean guard cell length (leaf ) (0) 24.92–31.22 mm (1) 32.39–40.46 mm (2) ×40.46 mm 12. Mean stomatal length:width ratio (leaf ) (0) 0.96:1.0–1.04:1.0 (1) 1.06:1.0–1.15:1.0 (2) 1.17:1.0–1.25:1.0 (3) 1.27:1.0–1.36:1.0 13. Epidermal raphide clusters (leaf ) (0) Absent (1) Present 14. Epidermal oil droplets (leaf ) (0) Absent (1) Present 15. Number of adaxial hypodermal layers (leaf ) (0) 0 (1) 1–3 (2) ×3 (3) Variable number of layers 16. Abaxial hypodermis (leaf ) (0) Absent (1) Present 17. Hypodermal thickenings (leaf ) (0) Absent (1) Present (2) Variable presence of thickenings 18. Mesophyll cell differentiation into (0) Differentiated palisade and mesophyll cells (leaf ) (1) Scarcely differentiated (2) Variably differentiation (3) Not differentiated 19. Spirally thickened idioblasts (leaf ) (0) Absent (1) Presence of thickenings variable (2) Present 20. Number of vein series (leaf ) (0) 2 (1) 1 21. Parenchymatous bundle sheath (0) Absent (1) Present 22. Sclerotic bundle sheath (leaf ) (0) Absent (1) Presence of bundle sheath variable (2) Present 23. Marginal bundle rotation (leaf ) (0) Rotated (1) Not rotated 24. Embedded peduncle (0) Absent (1) Present 25. Cuticle thickness (stem) (0) 5–15 mm (1) ³5 mm 26. Epidermal cell shape (stem in transection) (0) Dome-shaped (1) Not dome-shaped 27. Sclerotic epidermis (secondary stem) (0) Present (1) Absent 28. Sclerotic subepidermal layers (0) Present (secondary stem) (1) Absent continued
28
NEYLAND ET AL.
APPENDIX (continued) ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Attribute Character State ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — 29. Vascular cylinder (stem) (0) Eccentric (1) Not eccentric 30. Number of velamen layers (root) (0) ×2 (1) 2 (2) 1 (3) Variable number of layers 31. Exodermal thickenings (root) (0) U-shaped (1) O-shaped (2) U- and O-shaped (3) Variably shaped 32. Cortex (root) (0) Central large layer (1) Uniform 33. Endodermal thickenings (root) (0) O-shaped (1) Variably shaped 34. Number of protoxylem strands (root) (0) −9 (1) Variable in number (2) ¾8 35. Sclerotic pith (0) Absent (1) Present 36. Resupination (flower) (0) Resupinate (1) Presence of resupination variable (2) Nonresupinate 37. Relative shape of perianth (flower) (0) Parts similar (1) Presence of similar variable (2) Parts dissimilar 38. Sepal fusion (flower) (0) Apices fused (1) Apices not fused 39. Fusion of labellum and column (flower) (0) Firmly fused (1) Presence of fused labellum variable (2) Articulate 40. Labellum sensitivity (flower) (0) Not sensitive (1) Sensitive 41. Labellum divided into epichile and (0) No hypochile (flower) (1) Yes 42. Column foot (flower) (0) Absent (1) Presence of column for variable (2) Present 43. Another position (flower) (0) Incumbent (1) Dorsal (2) Apical (3) Trumpet-shaped 44. Number of pollinia (flower) (0) 8 (1) 6 (2) 4 (3) 2 45. Number of stigma lobes (flower) (0) 1 (1) 2 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
A phylogenetic analysis of subtribe Pleurothallidinae (Orchidaceae) RAY NEYLAND AND LOWELL E. URBATSCH Department of Plant Biology, Louisiana State University, Baton Rouge, Louisiana 70803-1705, U.S.A. AND ALEC M. PRIDGEON Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3DS Received May 1994, accepted for publication November 1994
A cladistic analysis of subtribe, Pleurothallidinae (Orchidaceae) is based on 45 anatomical: morphological characters. The ingroup members comprise 24 genera; the large genus Pleurothallis consists of two subgenera and ten species complexes. Three taxa representing subtribes Laeliinae and Arpophyllinae are designated as outgroup. Eight most parsimonious trees were discovered using computer assisted software (length 230; CI 0.27). The hypothesis that subtribe Pleurothallidinae has undergone a unilinear reduction in the number of pollinia is not supported by this study. Although the eight-pollinia state as represented by Octomeria apparently is plesiomorphic, the two-pollinia and four-pollinia states arose early in the phylogeny of the subtribe. Both twoand four-pollinia states subsequently reappeared as parallelisms. The six-pollinia state exhibited in Brachionidium is autapomorphic. This cladistic analysis suggests that Pleurothallis is not a natural genus and, perhaps may be divided into several discrete genera. ADDITIONAL KEY WORDS:—Pleurothallis – cladistics – phenetics – systematics.
CONTENTS Introduction . . . . . Methods . . . . . . Results . . . . . . Discussion . . . . . . Systematic overview . . Taxonomy of Pleurothallis . Evolution of pollinia number Acknowledgements . . . . References . . . . . . Appendix . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
13 14 16 16 17 23 25 26 26 26
INTRODUCTION
Pleurothallidinae Lindl. (Orchidaceae) is a Neotropical subtribe included in tribe Epidendreae (Dressler, 1993). Its members range from Brazil and Peru northward through Central America into Mexico and southern Florida; the 0024–4074/95/010013+16 $08.00/0
13
© 1995 The Linnean Society of London
14
NEYLAND ET AL.
greatest diversity of species occurs in cloud forest habitats (Pridgeon, 1982a,b). An articulation associated with an abscission zone between ovary and pedicel distinguishes the subtribe (Garay, 1956; Luer, 1986a; Dressler, 1993; Neyland & Urbatsch, 1993). In all other orchids the articulation occurs at the base of the pedicel (Dressler, 1981). Recent revisions by Luer (1986a, 1991) attributed 29 genera and as many as 4000 species to the subtribe. Luer acknowledged his treatment as artificial and was unable to bring phylogenetic order to the group. An earlier study by Pridgeon (1982a) used 45 anatomical:morphological characters to undertake a numerical analysis of the subtribe and suggested that the phenetic relationships identified in his study represented seven major phylogenetic lines. An important character in Pridgeon’s analysis is number of pollinia. He supported the hypothesis by Dressler (1981, 1993) that the Pleurothallidinae shows a reduction series in the number of pollinia from eight to six to four to two with eight representing the plesiomorphic condition. However, Luer (1986a) apparently believed that the direction may have proceeded in the opposite direction from two to eight. Most genera and an overwhelming number of species bear two pollinia; therefore, the relatively few members with more than two pollinia are considered pivotal in this study. For example, Octomeria bears eight pollinia, whereas Restrepia, Restrepiella, Restrepiopsis, Barbosella and Dresslerella each bear four. According to Garay (1956), Brachionidium has six; however, Luer (1986a) included the eight-pollinia species of the invalid genus, Yolanda, in Brachionidium. According to Luer (1986c), 29 subgenera comprise the very large genus Pleurothallis to which over 2000 epithets have been attributed. In Pridgeon’s (1982a) numerical analysis, two subgenera and 10 species complexes were recognized. Dressler predicted (personal communication 1992), that Pleurothallis eventually may be divided into several distinct genera corresponding with their three basic habits. Most species have the Specklinia habit with short stems and oblanceolate leaves that taper basally. Some have the Acianthera habit with long stems and relatively wide leaf bases. The Pleurothallis cardiothallis group has long stems, cordate leaves, and fascicled flowers. The primary objective of this study is to infer a hypothesis of phylogenetic relationship for Pleurothallidinae by combining the character-state data matrix constructed for taxa studied in Pridgeon’s (1982a) phenetic analysis (here designated as the ingroup) with a data matrix constructed for an appropriate outgroup and subjecting the combined data sets to cladistic analysis. The resulting cladograms are used to: (1) discuss the cladistic relationships among the genera and Pleurothallis species complexes of Pleurothallidinae; (2) evaluate whether the genus Pleurothallis represents a monophyletic group or is polyphyletic and thus can be divided into several genera, and (3) discuss the trend in character evolution with respect to pollinia number.
METHODS
The ingroup members in this study are those used as terminal taxa in Pridgeon’s (1982a) phenetic analysis. These taxa consist of 24 genera; Pleurothallis consists of two subgenera further divided into 10 species complexes. For a
PLEUROTHALLIDINAE
15
detailed list of ingroup material examined see Pridgeon (1982a). Pridgeon’s classification scheme is presented in Table 1. Three taxa were selected as outgroup members based on their status as representing possible sisters to subtribe Pleurothallidinae. These include: Arpophyllum giganteum Hartw., (Arpophyllinae); Brassavola nodosa (L.) Lindl., (Laeliinae) and Epidendrum ciliaris L., (Laeliinae). Voucher specimens are housed at Louisiana State University (LSU) (Table 2). According to Dressler (1960), Arpophyllum appears to represent a survivor of the ancestral stock from which the Pleurothallidinae were derived. Pridgeon (1982a) stated that it is plausible that Octomeria (an ingroup member) and Arpophyllum have a common ancestor and represent divergent evolutionary lines from that progenitor. According to Dressler (personal communication 1992) and Pridgeon (personal communication
TABLE 1. Classification scheme of Pleurothallidinae from Pridgeon (1982b). — –––––––––––––––––––––––––––––––––––––––––– Number of pollinia Taxon –––––––––––––––––––––––––––––––––––––––––––––––– — 8 Octomeria Pleurothallopsis 6 Brachionidium Chamelophyton 4 Barbosella Dresslerella Restrepia Restrepiella Restrepiopsis 2 Acostaea Andreettaea Condylago Crocodeilanthe Cryptophoranthus Dracula Dryadella Lepanthes Lepanthopsis Masdevallia Phloeophila Physothallis Platystele Pleurothallis subg. Pleurothallis P. alexandrae complex P. blaisdellii complex P. dura complex P. loranthophylla complex P. peduncularis complex P. ruscifolia complex P. scoparum complex P. tuerckheimii complex P. subg. Specklinia P. grobyi complex P. sclerophylla complex Porroglossum Salpistele Scaphosepalum Stelis Trisetella –––––––––––––––––––––––––––––––––––––––––––––––– —
16
NEYLAND ET AL.
TABLE 2. Outgroup members examined ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Accession Taxon Source Number ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Arpophyllum giganteum Hartw. University of Connecticut Neyland 90 Brassavola nodosa (L.) Lindl. Louisiana State University Neyland 72 Epidendrum ciliare L. Louisiana State University Neyland 81 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
1992), subtribe, Laeliinae may also be considered a likely outgroup to the Pleurothallidinae. Furthermore, a close relationship with Laeliinae is also suggested by the cladistic analysis of Burns-Balogh and Funk (1966), which suggested that the tribe Epidendreae is sister to the ‘Pleurothallis Group’. Forty-five multi-state and binary characters were used in this analysis (Appendix). Character states assigned to the ingroup members are those used in Pridgeon’s (1982a) numerical analysis. These characters were expected to yield phylogenetically significant information for the reasons detailed in Pridgeon’s (1982b) diagnosis. Character states were determined for the outgroup members using light microscopy, scanning electron microscopy (SEM), direct observation, and literature sources. Character states for all are summarized in Table 3. Multi-state transformation series were unordered so that character changes could occur without arbitrary constraints. Cladistic analyses performed using PAUP (version 3.0s, Swofford, 1991). Multi-state characters were unordered using Fitch criterion (Fitch, 1971). Most parsimonious trees were sought from a heuristic stepwise addition search employing 100 replications. MacClade (version 3.0, Maddison and Maddison, 1992) was used to explore hypothesized phylogenetic trees further and to visualize character evolution. RESULTS
Using the three designated taxa as outgroup, eight most parsimonious cladograms were discovered. The data matrix included 36 phylogenetically informative characters. All cladograms have 230 steps and a consistency index (CI) of 0.27. A strict consensus tree is presented in Figure 1. In six of the parsimony trees, the Pleurothallis blaisdellii, P. sclerophylla, P. tuerckheimii and P. loranthophylla complexes form a clade that is distinct from the clade composed of Cryptophoranthus, the Pleurothallis dura and P. ruscifolia complexes, Lepanthopsis and Restrepia. In the remaining two trees, the two clades are combined into a single clade. One most parasimonious tree representing the topology where the two clades are distinct (reflecting six of the eight resulting topologies) is presented in Figure 2. DISCUSSION
The consistency index of 0.27 represents a high level of homoplasy and may be a function of the rapid rates of evolution believed to have occurred in Orchidaceae (Dressler, 1993). However, the use of characters with continuous states (characters: 6, 7, 11, 12, 15) and polymorphic states (characters: 15, 17, 18, 19, 30, 36, 39) used in this study may also have contributed to this
PLEUROTHALLIDINAE
17
level of homoplasy (Appendix). Branch lengths representing unambiguous synapomorphies are generally considered a measure of clade stability. As indicated by one of the most parsimonious cladograms, nodal support as defined by the number of synapomorphies is often weak (Fig. 2). This level of support is due in part to the low ratio of terminal taxa (34) to characters (45). Despite the high level of homoplasy and generally weak branch support, the data set used in this study brings a substantial level of resolution to the ingroup and provides a cladistic framework for a phylogenetic discussion of Pleurothallidinae. Systematic overview Octomeria and the Pleurothallis peduncularis complex form the most basal clade in this study (Fig. 2). The position of Octomeria is consistent with the hypothesis TABLE 3. Character states for Pleurothallidinae and outgroup members. Characters and character states are identified in Appendix. Missing data are indicated by question marks — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Taxa 1 2 3 4 5 6 7 8 9 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Ingroup Acostaea 2 0 0 0 0 2 2 0 1 Barbosella 2 0 0 0 0 0 1 0 0 Brachionidium 1 0 0 0 0 2 2 1 0 Condylago 2 0 0 0 0 2 0 1 1 Cryptophoranthus 2 0 0 0 1 1 2 1 1 Dracula 2 0 0 0 0 2 2 1 1 Dresslerella 2 1 0 0 0 2 2 0 0 Dryadella 2 0 0 0 0 1 2 1 1 Lepanthes 2 0 0 0 0 2 2 0 1 Lepanthopis 2 0 0 0 1 2 0 1 1 Masdevallia 2 0 0 0 1 2 2 0 1 Physosiphon 2 0 0 0 1 2 2 0 1 Physothallis 2 0 0 0 0 1 1 0 1 Platystele 2 0 0 0 0 2 2 1 1 Pleurothallis ruscifolia complex 2 0 0 0 0 1 1 1 1 Pleurothallis alexandrae complex 0 0 1 0 0 0 1 0 0 Pleurothallis blaisdellii complex 2 0 0 0 1 1 0 0 1 Pleurothallis dura complex 2 0 0 0 1 1 2 1 1 Pleurothallis grobyi complex 2 0 0 0 0 1 0 1 1 Pleurothallis loranthophylla complex 2 0 0 0 1 1 1 1 1 Pleurothallis peduncularis complex 0 0 0 1 1 1 1 1 0 Pleurothallis sclerophylla complex 2 0 0 0 0 2 0 1 1 Pleurothallis scoparium complex 2 0 0 0 0 1 2 1 1 Pleurothallis subgen. Specklinia 2 0 0 0 0 2 1 1 1 Pleurothallis tuerckheimii complex 2 0 0 0 0 2 1 1 1 Porroglossum 2 0 0 0 0 1 2 0 1 Restrepia 2 0 0 0 0 1 1 0 1 Restrepiella 0 0 0 0 1 0 1 1 0 Restrepiopis 1 0 0 0 0 1 2 1 0 Salpistele 2 0 0 0 1 2 2 1 1 Scaphosepalum 2 0 0 0 0 2 2 1 1 Stelis 2 0 0 0 1 2 2 1 1 Tristella 2 0 0 0 1 2 2 1 1 Octomeria 0 0 0 0 0 0 0 0 0 Outgroup Arpophylum giganteum 0 0 0 0 0 0 1 0 0 Brassavola nodosa 0 0 0 0 1 1 2 0 1 Epidendrum ciliaris 0 0 0 0 1 1 0 1 1 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
18
NEYLAND ET AL.
TABLE 3 (continued). — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 1 1 1 1 1 1 1 1 1 Taxa 0 1 2 3 4 5 6 7 8 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Ingroup Acostaea 1 0 1 0 0 1 1 1 3 Barbosella 1 0 1 0 0 2 1 0 1 Brachionidium 1 1 3 0 0 0 0 0 3 Condylago 1 0 1 0 0 1 1 1 3 Cryptophoranthus 1 0 0 0 0 3 1 1 0 Dracula 1 0 1 0 0 0 0 0 3 Dresslerella 0 2 3 0 0 3 1 0 0 Dryadella 1 0 3 0 0 2 1 2 1 Lepanthes 1 0 0 0 0 2 1 0 0 Lepanthopis 1 0 1 0 0 2 1 1 0 Masdevallia 1 1 2 0 0 3 1 2 1 Physosiphon 1 1 2 0 0 2 1 1 2 Physothallis 1 2 0 0 0 2 1 1 3 Platystele 1 0 1 0 1 2 1 1 3 Pleurothallis ruscifolia complex 1 0 1 0 0 2 1 1 0 Pleurothallis alexandrae complex 0 2 3 0 0 2 1 0 0 Pleurothallis blaisdellii complex 1 1 1 0 0 1 1 1 0 Pleurothallis dura complex 1 1 0 0 0 2 1 1 0 Pleurothallis grobyi complex 1 0 0 0 0 2 1 2 2 Pleurothallis loranthophylla complex 1 1 0 0 0 3 1 1 1 Pleurothallis penduncularis complex 1 1 2 1 0 2 1 0 0 Pleurothallis sclerophylla complex 1 1 1 0 0 2 1 1 1 Pleurothallis scoparium complex 1 1 1 0 0 3 1 2 2 Pleurothallis subgen. Specklinia 1 0 0 0 0 2 1 0 1 Pleurothallis tuerckheimii complex 1 1 2 0 0 2 1 1 0 Porroglossum 1 1 2 0 0 3 1 2 0 Restrepia 1 0 1 0 0 1 1 1 0 Restrepiella 1 0 0 0 0 2 1 0 0 Restrepiopsis 1 0 0 0 0 2 1 1 0 Salpistele 1 0 1 0 0 1 1 1 3 Scaphosepalum 1 0 1 0 0 3 1 2 1 Stelis 1 0 2 0 0 2 1 1 0 Trisetella 1 1 2 0 0 1 1 2 2 Octomeria 1 1 2 0 0 2 1 0 2 Outgroup Arpophyllum giganteum 0 1 3 0 0 0 0 0 0 Brassavola nodosa 0 2 3 0 0 1 1 0 0 Epidendrum ciliaris 0 1 3 0 0 0 0 0 0 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
that this taxon represents the most basal element in Pleurothallidinae. The apparently close relationship between Octomeria and the P. peduncularis complex was suggested by Pridgeon (1982a). The P. alexandrae complex appears as sister to the Octomeria-P. peduncularis clade. Pridgeon (1982a) suggested a close relationship among these taxa and distinguished the P. alexandrea complex from the P. peduncularis complex by floral specializations that led to fly pollination. Ascending the cladogram, a grade of genera exhibiting four pollinia (Dresslerella, Barbosella, Restrepiella, Restrepiopsis) are next encountered (Fig. 2). A close association among all Pleurothallidinae exhibiting four pollinia was suggested by Pridgeon (1982a). However, Restrepia (also exhibiting four pollinia) appears in a separate clade and thus appears not closely related to taxa in this grade (Fig. 2). The Cryptophoranthus, Lepanthopsis, Pleurothallis dura complex, Pleurothallis ruscifolia complex and Restrepia clade is distinguished in this study by stem cuticles ³5 mm
PLEUROTHALLIDINAE
19
TABLE 3 (continued). — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 1 2 2 2 2 2 2 2 2 Taxa 9 0 1 2 3 4 5 6 7 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Ingroup Acostaea 0 1 0 1 0 0 1 1 1 Barbosella 0 1 1 0 1 0 1 0 1 Brachionidium 0 1 0 2 1 0 ? ? ? Condylago 0 1 0 2 1 0 1 1 1 Cryptophoranthus 2 1 0 2 0 0 1 0 1 Dracula 0 1 1 2 1 0 1 1 1 Dresslerella 0 1 0 2 1 0 1 0 1 Dryadella 1 1 1 2 0 0 1 1 1 Lepanthes 0 1 0 2 1 0 1 1 1 Lepanthopsis 2 1 0 2 1 0 ? ? ? Masdevallia 2 1 1 2 1 0 1 1 1 Physosiphon 1 1 0 2 1 0 1 1 1 Physothallis 0 1 0 2 0 0 0 1 1 Platystele 1 1 0 0 0 0 1 1 1 Pleurothallis ruscifolia complex 2 1 0 2 1 0 1 0 0 Pleurothallis alaxandrae complex 0 0 0 2 1 0 0 0 1 Pleurothallis blaisdellii complex 1 1 0 2 1 0 0 0 1 Pleurothallis dura complex 2 1 0 2 0 0 ? ? ? Pleurothallis grobyi complex 0 1 0 2 0 0 1 1 1 Pleurothallis loranthophylla complex 2 1 0 2 1 0 0 1 1 Pleurothallis peduncularis complex 0 0 0 2 1 0 1 0 0 Pleurothallis sclerophylla complex 1 1 0 2 1 0 0 1 0 Pleurothallis scoparium complex 1 1 0 2 1 1 1 1 1 Pleurothallis subgen Specklinia 0 1 0 2 0 0 1 1 1 Pleurothallis tuerckheimii complex 1 1 0 1 1 0 1 1 0 Porroglossum 1 1 1 2 0 0 1 1 1 Restrepia 2 1 0 2 1 0 1 0 1 Restrepiella 2 1 1 2 1 0 0 0 1 Restrepiopsis 2 1 0 1 1 0 ? ? ? Salpistele 0 1 0 1 0 0 ? ? ? Scaphosepalum 1 1 0 2 1 0 1 1 1 Stelis 2 1 1 2 0 0 1 1 1 Trisetella 1 1 1 2 1 0 1 1 1 Octomeria 0 0 0 2 0 0 0 0 0 Outgroup Arpophyllum giganteum 0 0 0 2 1 0 0 0 0 Brassavola nodosa 0 0 0 2 0 0 0 1 0 Epidendrum ciliaris 2 0 1 2 1 0 0 0 0 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
(character 25), and a variable number of root protoxylem poles (character 34). The position of Lepanthopsis in this clade suggests this taxon is not closely allied with Lepanthes as was suggested by Pridgeon (1982a), who distinguished Lepanthes from Lepanthopsis primarily by the loss of spirally thickened idioblasts. Both genera are vegetatively similar with respect to lepanthiform (tubular) sheaths of the ramicaul. The Pleurothallis blaisdelii, P. sclerophylla, P. tuerckheimii, and P. loranthophylla species complexes form a clade distinguished in this study by the absence of a column foot (character 42) and by a variable degree of fusion between the labellum and column (character 39). A single synapomorphy in six of the eight most parsimonious trees discovered separates these two clades. Therefore, this distinction between the two clades is considered weak. Consisting of two terrestrial species Physothallis appears as sister to the Pleurothallis grobyi complex. Claiming that the deeply connate dorsal sepal is
20
NEYLAND ET AL.
TABLE 3 (continued). — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 2 2 3 3 3 3 3 3 3 Taxa 8 9 0 1 2 3 4 5 6 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Ingroup Acostaea 1 1 1 0 1 0 2 1 0 Barbosella 1 1 0 0 1 0 2 1 2 Brachionidium ? ? ? ? ? ? ? ? 1 Condylago 0 1 ? ? ? ? ? ? 0 Cryptophoranthus 0 1 2 0 1 0 1 1 0 Dracula 1 1 0 0 1 0 0 1 0 Dresslerella 0 0 1 3 1 1 2 1 0 Dryadella 1 1 2 2 1 0 2 1 0 Lepanthes 1 1 ? ? ? ? ? ? 1 Lepanthopsis ? ? ? ? ? ? ? ? 0 Masdevallia 1 1 3 0 1 0 1 1 0 Physosiphon 0 1 2 0 1 0 2 1 0 Physothallis 0 1 ? ? ? ? ? ? 0 Phytystele 1 1 2 0 1 0 2 1 0 Pleurothallis ruscifolia complex 0 1 0 3 0 1 1 1 1 Pleurothallis alexandrae complex 0 1 0 1 1 0 0 0 0 Pleurothallis blaisdellii complex 0 1 ? ? ? ? ? ? 0 Pleurothallis dura complex ? ? ? ? ? ? ? ? 0 Pleurothallis grobyi complex 1 1 ? ? ? ? ? ? 0 Pleurothallis loranthophylla complex 0 1 ? ? ? ? ? ? 0 Pleurothallis peduncularis complex 0 1 3 0 1 0 0 0 0 Pleurothallis sclerophylla complex 0 1 ? ? ? ? ? ? 0 Pleurothalis scoparium complex 0 1 ? ? ? ? ? ? 0 Pleurothallis subgen. Specklinia 1 1 0 0 1 0 2 0 1 Pleurothallis tuerckheimii complex 0 1 0 1 1 0 2 1 0 Porroglossum 1 1 3 3 1 0 1 1 1 Restrepia 0 1 3 0 1 1 1 1 0 Restrepiella 0 1 0 2 1 0 0 1 0 Restrepiopsis ? ? ? ? ? ? ? ? 0 Salpistele ? ? ? ? ? ? ? ? 0 Scaphosepalum 1 1 3 3 1 1 2 1 2 Stelis 0 1 0 2 1 0 2 1 1 Trisetella 1 1 2 0 1 0 2 1 0 Octomeria 0 1 1 0 1 0 0 0 0 Outgroup Arpophyllum giganteum 0 0 0 0 1 0 0 0 2 Brassavola nodosa 0 1 0 0 1 0 0 0 0 Epidendrum ciliaris 1 1 0 0 1 0 2 0 0 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
insufficient for defining this taxon at the generic level, Luer (1986a) relegated Physothallis to subgeneric status in Pleurothallis. The clade composed of the Pleurothallis subgen. Specklinia and Stelis is distinguished in this study by differentiated leaf mesophyll (character 18), incumbent anther (character 43), and variability with respect to resupination (character 36). Although the large genus Stelis (×500 species) appears to be a natural genus, the great similarity in floral morphology has made taxon distinction at the species level difficult. Pridgeon (1982a) suggested a close relationship existed between Stelis and Brachionidium on the basis that both taxa share a footless and bibrachiate column. However, this cladistics analysis suggests the two taxa represent disparate elements in Pleurothallidinae (Fig. 2). The large clade composed of Dryadella, Masdevallia, Porroglossum, Trisetella, Physosiphon, Scaphosepalum Platystele and the Pleurothallis scoparium complex is
PLEUROTHALLIDINAE
21
TABLE 3 (continued). — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 3 3 3 4 4 4 4 4 4 Taxa 7 8 9 0 1 2 3 4 5 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Ingroup Acostaea 2 0 2 1 0 2 0 3 0 Barbosella 2 0 2 0 0 2 0 2 0 Brachionidium 0 0 2 0 0 0 0 1 1 Condylago 2 0 2 1 0 2 0 3 0 Cryptophoranthus 2 1 2 0 0 2 0 3 0 Dracula 2 0 1 0 1 2 0 3 0 Dresslerella 2 0 2 0 0 2 0 2 0 Dryadella 2 0 2 0 0 2 0 3 0 Lepanthes 2 0 0 0 0 0 2 3 0 Lepanthopsis 2 0 1 0 0 0 0 3 1 Masdevallia 2 0 2 0 0 2 0 3 0 Physosiphon 2 0 2 0 0 2 0 3 0 Physothallis 2 0 2 0 0 2 0 3 0 Platystele 2 0 1 0 0 1 0 3 1 Pleurothallis ruscifolia complex 1 0 1 0 0 1 0 3 0 Pleurothallis alexandrae complex 2 0 2 0 0 2 0 3 0 Pleurothallis blaisdellii complex 2 0 2 0 0 2 0 3 0 Pleurothallis dura complex 2 0 2 0 0 2 0 3 0 Pleurothallis grobyi complex 2 0 2 0 0 2 0 3 0 Pleurothallis loranthophylla complex 0 0 2 0 0 2 0 3 0 Pleurothallis peduncularis complex 2 0 2 0 0 2 0 3 0 Plerurothallis sclerophylla complex 2 0 2 0 0 2 0 3 0 Pleurothallis scoparium complex 2 0 2 0 0 2 0 3 0 Pleurothallis subgen. Specklinia 2 0 2 0 0 2 1 3 0 Pleurothallis tuerckheimii complex 2 0 2 0 0 2 0 3 0 Porroglossum 2 0 2 1 0 2 1 3 0 Restrepia 2 0 0 0 0 0 0 2 0 Restrepiella 2 0 2 0 0 2 0 2 0 Restrepiopsis 2 0 2 0 0 2 0 2 0 Salpistele 2 0 0 0 0 0 3 3 0 Scaphosepalum 2 0 2 0 0 2 0 3 0 Stelis 2 0 2 0 0 1 1 3 1 Trisetella 2 0 2 0 0 2 0 3 0 Octomeria 0 0 1 0 0 1 0 0 0 Outgroup Arpophyllum giganteum 0 0 0 0 0 2 0 0 0 Brassavola nodosa 2 0 0 0 0 0 0 0 0 Epidendrum ciliaris 2 0 0 0 0 0 0 2 0 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —
distinguished in this study by single-celled velamen (character 30) and variability with respect to spirally thickened idioblasts (character 19). Platystele appears as the most basal taxon in this clade. Consisting of approximately 73 species, this genus is distinguished by Luer (1986a) as having a lateral racemose inflorescence and ramicauls shorter than the leaves. Luer (1986a) considered Platystele closely related to Pleurothallis. Dryadella consists of about 40 species that formally resided in Masdevallia based on superficial similarities (Luer, 1986a). As suggested by the cladogram, Dryadella does not appear closely allied with Masdevallia. Scaphosepalum (approximately 30 species), distinguished by Luer (1988) as having a pair of cushion-like osmophores positioned on the lateral sepals, appears as sister to the Pleurothallis scoparium complex (Fig. 2). According to Luer (1986a), Trisetella (approximately 15 species) is distinguished by a column with a hooded ventral anther and was previously treated as a section of Masdevallia. Results
22
NEYLAND ET AL.
Figure 1. Strict consensus tree of eight trees discovered from a parsimony search for the taxa included in this study.
from this study indicate that Trisetella and Masdevallia are closely allied, and the treatment of Trisetella as a genus is justified. Masdevallia (approximately 350 species) is distinguished by Luer (1986b) as having calliferous petals and a lip hinged to a free incurved extension from the apex of the column foot. Luer (1987) stated that Porroglossum resembles Masdevallia by having short ramicauls and sepals with long tails; however, its sensitive labellum sufficiently distinguishes it from Masdevallia. The results of this study indicate that Masdevallia and Porroglossum are sister taxa. The clade composed of Acostaea, Salpistele, Brachionidium, Dracula and Condylago is distinguished in this study by having 1–3 adaxial hypodermal layers in the leaf (character 15). Pridgeon (1982a) suggested that Brachionidium was closely allied with Octomeria and is distinguished from Octomeria by the loss of two pollinia. However, the relatively derived position of Brachionidium suggests the two taxa are disparate elements in Pleurothallidinae (Fig. 2). Pridgeon (1992a) considered Dracula to be closely allied with Brachionidium and distinguished Dracula from Brachionidium by the loss of four pollinia, perianth specialization, and
PLEUROTHALLIDINAE
23
Figure 2. One of the eight most parsimonious trees discovered from a parsimony search for the taxa included in this study. The value above each branch represents the number of unambiguous synapomorphies.
presence of glandular trichomes on leaf surfaces. This cladistic analysis indicates that Dracula and Brachionidium are sister taxa (Fig. 2). Acostaea, Condylago, and Porroglossum each exhibit sensitive labella. According to Luer (1987), because the labellum in each of these genera is structurally and functionally distinct, these taxa are not closely related. As suggested by the results of this study, the sensitive labellum arose independently within each of these three taxa (Fig. 2). Luer (1991) stated that Salpistele may be closely allied with Lepanthes; however, he considered the trumpet-shaped column and lack of lepanthiform sheaths sufficient to distinguish Salpistele as a genus. This cladistic analysis suggests that Salpistele is a sister to Acostaea and not closely related to Lepanthes (Fig. 2). Taxonomy of Pleurothallis The results of this cladistic study suggests that the large genus Pleurothallis is polyphyletic and, therefore, may be divided into several genera. One genus
24
NEYLAND ET AL.
would be represented by the P. peduncularis complex that was segregated as Myoxanthus by Luer (1992). The members of this genus are characterized by well-developed ramicauls with scruffy sheaths and short rhizomes. A second genus may be represented by the P. alexandrae complex which was elevated to Myoxanthus subgenus Silenia by Luer (1992). Both complexes occupy a relatively basal position within the subtribe and are far removed from the more derived Pleurothallis complexes. The basis for separating these two closely related complexes is that this analysis indicates that the P. peduncularis complex appears more closely related to Octomeria than to the Pleurothallis alexandrae complex (Fig. 2). Among the more derived Pleurothallis complexes, P. dura and P. ruscifolia share a close affinity and occupy the same clade. Luer (1986a) elevated the P. dura complex to Trichosalpinx subg. Tubella. Members of this subgenus produce short ramicauls, connate lateral sepals and ciliate petals. Luer (1986c) placed the P. ruscifolia complex in Pleurothallis subg. Pleurothallis sect. Pleurothallis subsect. Pleurothallis series Pleurothallis. Members of this series produce flowers borne in short-fascicled racemes. Because this complex includes the type species P. ruscifolia ( Jacq.) R. Br., the name Pleurothallis must be retained. This cladistic analysis indicates that both complexes are more closely related to other genera than to each other. Therefore, Luer’s placement of the P. dura complex within Trichosalpinx seems justified (Fig. 2). The Pleurothallis loranthophylla, P. tuerckheimii, P. sclerophylla and P. blaisdellii complexes constitute a single clade. Luer (1986c) placed the P. loranthophylla complex in Pleurothallis subg. Rhynchopera. Luer characterized this subgenus as an easily recognizable homogenous group with well-developed ramicauls that bear a petiolate leaf; the simple lip is firmly united to the short, pedestal-like foot of the column. The P. tuerckheimii complex is included in Pleurothallis subg. Dracontia in Luer’s treatment. Its well-developed ramicauls bear a tubular sheath near the middle and are subtended by two or three other sheaths at the base; the lip is thick with erect, thin, wing-like basal lobes. Luer placed the P. sclerophylla complex in Pleurothallis subg. Specklinia sect. Acuminatae. In that section, as in Pleurothallis subg. Dracontia, tubular sheaths also are inserted on the ramicaul. However, the lip is oblong-ligulate often with a pair of calli; the base is simply hinged to the stout column foot. The P. blaisdellii complex is placed in Trichosalpinx subg. Trichosalpinx by Luer (1986a). That subgenus is characterized by stout ramicauls, free lateral sepals and petals with entire margins. Because these four complexes are grouped in the same clade, justification for their taxonomic separation into discrete genera is uncertain. If considered a single taxonomic unit, the clade appears to warrant elevation to generic status (Fig. 1). Moving toward the most derived taxa, the Pleurothallis grobyi complex, P. subg. Specklinia, and the P. scoparium complex show little affinity and are thus logical candidates for generic status (Fig. 2). The P. grobyi complex was placed within Pleurothallis subg. Specklinia sect. Hymenodanthae by Luer (1986c). Luer described that section as having numerous, closely allied caespitose species with an abbreviated ramicaul and racemose inflorescence that is usually longer than the leaf. Pleurothallis subg. Specklinia includes numerous species characterized by the lip hinged to the column foot. The column is well developed; its apex covers the anther, rostellum, and stigma. Luer placed the P. scoparium complex within
PLEUROTHALLIDINAE
25
Pleurothallis subg. Scopula. He attributed four species to the subgenus; they are characterized by elliptical leaves with a tuft of single-flowered peduncles emerging from what appears to be the midrib near the leaf apex. Evolution of pollinia number The number of pollinia (character 44) are plotted onto one most parsimonious tree to illustrate character state changes (Fig. 3). Equivocal branches were resolved by the MacClade (Maddison & Maddison, 1992) equivocal cycling feature. This analysis suggests that although the eight-pollinia state may represent the primitive condition as espoused by Dressler and Pridgeon, the number of pollinia did not follow a unilinear reduction series. Instead it appears that the eight pollinia state represented by Octomeria, the two-pollinia state represented by the P. peduncularis and P. alexandrae complexes, and the four-pollinia state represented by the grade comprising Restrepiopsis, Restrepiella, Barbosella and Dresslerella all arose early in the phylogeny of the subtribe. The two-pollinia state reappears as a parallelism and is represented by most of the genera and a vast majority of species within the subtribe. Because Restrepia appears in a clade removed from the other four-pollinia taxa, its character state also is considered a parallelism. The six-pollinia condition of Brachionidium represents an autapomorphy. This study represents the first attempt to bring cladistic resolution to the
Figure 3. Character state tree showing the distribution of states for the number of pollinia (character 44) for the taxa included in this study.
26
NEYLAND ET AL.
subtribe Pleurothallidinae. Because the characters used in this study exhibit a high level of homoplasy and branch nodes on cladograms are often weakly supported, the results of this analysis are considered tentative. Other data to include molecular, cytological and palynological characters should be evaluated in order to clarify further the phylogeny of this vast subfamily. ACKNOWLEDGEMENTS
We thank Carlyle A. Luer for providing technical assistance; William Cullina, University of Connecticut, for supplying plant material for the project; and Meredith Blackwell, Robert Burckhalter, and Shirley Tucker for reviewing the manuscript. Support from the LSU Plant Biology Department is gratefully acknowledged. REFERENCES Dressler RL. 1960. The relationships of Meiracyllium (Orchidaceae). Brittonia 12: 222–225. Dressler RL. 1981. The Orchids: Natural History and Classification. Cambridge, Massachusetts: Harvard University Press. Dressler RL. 1993. Phylogeny and Classification of the Orchid Family. Portland, Oregon: Dioscorides Press. Fitch WM. 1971. Toward defining the course of evolution: minimal change for a specific tree topology. Systematic Zoology 20: 406–416. Garay LA. 1956. Studies in American orchids. II. The genus Brachionidium Lindl. Canadian Journal of Botany 34: 721–743. Luer CA. 1986a. Icones Pleurothallidinarum I. Systematics of the Pleurothallidinae. St. Louis: Missouri Botanical Garden. Luer CA. 1986b. Icones Pleurothallidinarum II. Systematics of Masdevallia. St. Louis: Missouri Botanical Garden. Luer CA. 1986c. Icones Pleurothallidinarum III. Systematics of Pleurothallis. St. Louis: Missouri Botanical Garden. Luer CA. 1987. Icones Pleurothallidinarum IV. Systematics of Acostaea Condylago and Porroglossum. St. Louis: Missouri Botanical Garden. Luer CA. 1988. Icones Pleurothallidinarum V. Systematics of Dresslerella and Scaphosepalum. St. Louis: Missouri Botanical Garden. Luer CA. 1991. Icones Pleurothallidinarum VIII. Systematics of Lepanthopsis, Octomeria subgenus Pleurothallopsis, Restrepiella, Restrepiopsis, Salpistele, and Teagueia. St. Louis: Missouri Botanical Garden. Luer CA. 1992. Icones Pleurothallidinarum IX. Systematics of Myoxanthus. St. Louis: Missouri Botanical Garden. Maddison WP, Maddison, DR. 1992. MacClade: Analysis of phylogeny and character evolution. Version 3.0. Sunderland, Massachusetts: Sinauer Associates. Neyland R, Urbatsch LE. 1993. Anatomy and morphology of the articulation between ovary and pedicel in Pleurothallidinae. Lindleyana. 8(4): 189–192. Pridgeon AM. 1982a. Numerical analyses in the classification of the Pleurothallidinae (Orchidaceae). Botanical Journal of the Linnean Society 85: 103–131. Pridgeon AM. 1982b. Diagnostic anatomical characters in the Pleurothallidinae (Orchidaceae). American Journal of Botany 69: 921–938. Swofford DL. 1991. Phylogenetic Analysis Using Parsimony. Version 3.0s. Champaign, Illinois: Illinois Natural History Survey.
APPENDIX ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Attribute Character State ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — 1. Glandular trichomes (leaf ) (0) Absent (1) Limited to one surface (2) Present on both surfaces 2. Epidermal papillae (leaf ) (0) Absent (1) Present 3. Multicellular muscilage trichomes (leaf ) (0) Absent (1) Present 4. Hispid leaf sheaths (leaf ) (0) Absent (1) Present
PLEUROTHALLIDINAE
27
APPENDIX (continued). ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Attribute Character State ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — 5. Cuticle surface (leaf ) (0) Smooth (1) Papillose 6. Cuticle thickness (leaf ) (0) 15–22 mm (1) 6–14 mm (2) ³3 mm 7. Epidermal radial wall thickness (leaf ) (0) ×5–7 mm (1) 3–5 mm (2) ³3 mm 8. Epidermal cell wall thickening (0) Differentially thickened (leaf in transection) (1) Uniformly thickened 9. Epidermal cell shape (0) Dome-shaped (1) Not dome-shaped 10. Level of stomatal apparatus (leaf ) (0) Raised above the epidermis (1) Flush with the epidermis 11. Mean guard cell length (leaf ) (0) 24.92–31.22 mm (1) 32.39–40.46 mm (2) ×40.46 mm 12. Mean stomatal length:width ratio (leaf ) (0) 0.96:1.0–1.04:1.0 (1) 1.06:1.0–1.15:1.0 (2) 1.17:1.0–1.25:1.0 (3) 1.27:1.0–1.36:1.0 13. Epidermal raphide clusters (leaf ) (0) Absent (1) Present 14. Epidermal oil droplets (leaf ) (0) Absent (1) Present 15. Number of adaxial hypodermal layers (leaf ) (0) 0 (1) 1–3 (2) ×3 (3) Variable number of layers 16. Abaxial hypodermis (leaf ) (0) Absent (1) Present 17. Hypodermal thickenings (leaf ) (0) Absent (1) Present (2) Variable presence of thickenings 18. Mesophyll cell differentiation into (0) Differentiated palisade and mesophyll cells (leaf ) (1) Scarcely differentiated (2) Variably differentiation (3) Not differentiated 19. Spirally thickened idioblasts (leaf ) (0) Absent (1) Presence of thickenings variable (2) Present 20. Number of vein series (leaf ) (0) 2 (1) 1 21. Parenchymatous bundle sheath (0) Absent (1) Present 22. Sclerotic bundle sheath (leaf ) (0) Absent (1) Presence of bundle sheath variable (2) Present 23. Marginal bundle rotation (leaf ) (0) Rotated (1) Not rotated 24. Embedded peduncle (0) Absent (1) Present 25. Cuticle thickness (stem) (0) 5–15 mm (1) ³5 mm 26. Epidermal cell shape (stem in transection) (0) Dome-shaped (1) Not dome-shaped 27. Sclerotic epidermis (secondary stem) (0) Present (1) Absent 28. Sclerotic subepidermal layers (0) Present (secondary stem) (1) Absent continued
28
NEYLAND ET AL.
APPENDIX (continued) ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Attribute Character State ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — 29. Vascular cylinder (stem) (0) Eccentric (1) Not eccentric 30. Number of velamen layers (root) (0) ×2 (1) 2 (2) 1 (3) Variable number of layers 31. Exodermal thickenings (root) (0) U-shaped (1) O-shaped (2) U- and O-shaped (3) Variably shaped 32. Cortex (root) (0) Central large layer (1) Uniform 33. Endodermal thickenings (root) (0) O-shaped (1) Variably shaped 34. Number of protoxylem strands (root) (0) −9 (1) Variable in number (2) ¾8 35. Sclerotic pith (0) Absent (1) Present 36. Resupination (flower) (0) Resupinate (1) Presence of resupination variable (2) Nonresupinate 37. Relative shape of perianth (flower) (0) Parts similar (1) Presence of similar variable (2) Parts dissimilar 38. Sepal fusion (flower) (0) Apices fused (1) Apices not fused 39. Fusion of labellum and column (flower) (0) Firmly fused (1) Presence of fused labellum variable (2) Articulate 40. Labellum sensitivity (flower) (0) Not sensitive (1) Sensitive 41. Labellum divided into epichile and (0) No hypochile (flower) (1) Yes 42. Column foot (flower) (0) Absent (1) Presence of column for variable (2) Present 43. Another position (flower) (0) Incumbent (1) Dorsal (2) Apical (3) Trumpet-shaped 44. Number of pollinia (flower) (0) 8 (1) 6 (2) 4 (3) 2 45. Number of stigma lobes (flower) (0) 1 (1) 2 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —