Amer. J. Bot. 75(7): 1080-1085. 1988. ISOZYME NUMBER IN SUBTRIBE ONCIDIINAE (ORCHIDACEAE): AN EVALUATION OF POLYPLOIDY1 MARK W. CHASE2 AND RICHARD G. OLMSTEAD3 Department of Biology, University of Michigan, Ann Arbor, Michigan 48109; and Department of Botany KB-15, University of Washington, Seattle, Washington 98195 ABSTRACT The Oncidiinae has attracted attention because of the variation it exhibits in chromosome number, n = 5-30, which is greater than the range in the rest of the Orchidaceae. The genus Psygmorchis, with n = 5 and 7, has been a particular focus ofcontroversy, and many authors have suggested that 5 and 7 are the base numbers for the subtribe. The other taxa in the subtribe presumably evolved through hybridization and polyploidy. Other workers have found that the lowest counts correlate with derived morphological conditions and have hypothesized that these low numbers result from aneuploid reductions, while higher numbers are associated with an­ cestral morphologies and are not the result ofpolyploidy. These two hypotheses were evaluated by determining isozyme numbers for 13 enzymes in species that span the chromosomal range known for the Oncidiinae (n = 5-30). Isozyme number has been shown to be a reliable indicator ofpolyploidyin angiosperms because polyploidsdisplay isozyme multiplicity relative to diploids. This analysis revealed no differences among species in isozyme number for the enzymes ex­ amined. Therefore, ourdata reject the hypothesis that species with higher chromosome numbers are polyploid. ONCIDIUM Sw. and subtribe Oncidiinae Ben­ A great deal of controversy has centered tham have attracted attention from orchid cy­ around Psygmorchis Dodson & Dressler, a re­ tologists because oftheir range ofchromosome cent segregate of Oncidium (Dodson and numbers. The great majority of orchid genera Dressler, 1972). Psygmorchis, with four rap­ and higher taxa exhibit little variation in chro­ idly developing (i.e., maturing in less than six mosome number, and the most frequent hap­ months) species that occur exclusively on the loid counts are n = 19 or 20 (all counts, unless outer portions of the canopy, has counts of n otherwise specified, are from a review by Ta­ = 5 and 7, and these represent the lowest num­ naka and Kamemoto, 1984). Some putatively bers known in the Orchidaceae. The relation­ ancestral taxa, such as the Cypripedioideae, ship of Psygmorchis to the rest of the Onci­ have lower counts (n = 10 in Cypripedium L.; diinae has been the subject of considerable but see Atwood, 1984), but the general pattern speculation (Dodson, 1957, 1958; Sinoto, 1962; is that of consistency within genera and often Garay, 1963; Sanford, 1964; Dodson and within tribes and subtribes. Oncidium s.l. and Dressler, 1972; Charanasri, Kamemoto, and the Oncidiinae stand in stark contrast. This Takeshita, 1973; Charanasri and Kamemoto, large (70 genera; 1,200 species) and floristically 1975; Dressler, 1981; Chase, 1986a, 1987). important Neotropical subtribe has reliable re­ Two explanations have been advanced to ports of n = 5, 7, 12-22, 24-26, 28, and 30, account for the range ofchromosome numbers with n = 28 the most frequent number. This exhibited by the Oncidiinae. The earlier hy­ range exceeds that of the rest of the Orchida­ pothesis was that n = 5 or 7 represents the base ceae (ifclearly polyploid individuals and races number for the Oncidiinae (Dodson, 1957; are excluded, i.e., those with conspecifics of Sinoto, 1962; Sanford, 1964) and that higher lower numbers). numbers were produced by hybridization and polyploidy accompanied by small aneuploid reductions (Garay, 1963; Charanasri and Ka­ I Received for publication 31 July 1987; revision ac­ memoto, 1975). The second hypothesis is that cepted 7 October 1987. aneuploid reductions are responsible for the We wish to thank the Drs. Soltis for reading and criti­ cizing an earlier version ofthis report. A National Science low counts found in Psygmorchis. With the Foundation Fellowship in Environmental Biology (to benefit of a much larger set of chromosome MWC) and a Doctoral Dissertation Improvement Grant counts, Dodsonand Dressler(1972) and Dress­ (to RGO) supported this study. ler (1981) argued that both n = 5 and 7 are 2 Current address: Dept. of Biology, Univ. of North Carolina, Chapel Hill, NC 27514. much too low to be base numbers for the On­ 3 Current address: Dept. ofBiology, Univ. of Michigan, cidiinae or even the Orchidaceae. Dodson and Ann Arbor, MI 48109. Dressler (1.972) stated that Psygmorchis is 1080 July 1988] CHASE AND OLMSTEAD- ISOZYME NUMBER IN ONCIDIINAE 1081 TABLE I. Species sampled, number ofspecimens sampled, and chromosome numberfor each species Number of Species; voucher number- specimens sampled Chromosome number Ada chlorops (Endres & Reichenb. f.) N. H. Williams (84018,84041) 2 n = 30pb Brassia maculata R. Brown (82205) I n = 30e Leochilus carinatus Knowles & Westc. (82170, 82188, 83137) 3 n = 21d L. crocodiliceps (Reichenb. f.) Kranzlin (83218,83219) 2 n = 24d Maxillaria picta Hooker (87047) I n = 20e Miltonia spectabilis Lindley (85080, 81028) 2 n = 30e Notylia barkeri Lindley (82070,82149,83282) 3 n = 21e Oncidium bicallosum Lindley (83000) 2 n = 14e O. cebolleta Sw, (83132) 2 n = 17e O. splendidum A. Richard (84552) I n = 18e O. wydleri Reichenb. f. (84431, 86069, 86070) 3 n = 28e Psygmorchis pusilla (L.) Dodson & Dressler (87047) 2 n = SC a All collections by MWC; all vouchered by flowers in FAA in the author's collection. b Chase, unpublished. c Tanaka and Kamemoto, 1984. d Chase, 1986b. another example of an ephemeral with a low MATERIALS AND METHODs-Eleven species count derived by reduction from a higher base representing the full range of chromosome number (a fa Stebbins, 1958). Chase (1986a, number known for subtribe Oncidiinae (n = 1987) argued that not only was the low number 5, 14, 17, 18, 21, 24, 28, 30) were assayed for in Psygmorehis the result ofreduction, but that 13 enzyme systems. One species from the the most primitive members ofthe Oncidiinae, closely related subtribe, Maxillariinae, Max­ in terms of vegetative and floral morphology, illaria picta (n = 20), was also included for were those with a chromosome number ofn = comparison. The species sampled, the number 30. Thus, all numbers in the entire subtribe of individuals per species, and the chromo­ result from aneuploidy of the original ploidy. some number of each species are shown in In this study, these two contrasting hypoth­ Table 1. The small sample sizes, while insuf­ eses for the range ofchromosome numbers in ficient for estimates of within-species genetic the Oncidiinae were evaluated by employing variability, are sufficient to evaluate ploidy enzyme electrophoresis to determine iso­ (Gottlieb, 1981a). zyme numbers for 13 enzymes in species rep­ Enzymes were extracted from fresh leafma­ resenting the complete range, n = 5-30. Dip­ terial in a grinding buffer composed of 0.1 M loid plants must exhibit at least a minimal tris-HCl pH 7.5, 0.001 M EDTA, 0.01 M KCl, number of isozymes (Gottlieb, 1982, 1983), 0.01 M MgCl, and 8% w/v PVP 40,000 with and aneuploidy will not change these numbers 0.1 % v/v 2-mercaptoethanol added just before (although presumably their chromosomal ar­ beginning the extraction (Soltis et al., 1983) rangement maybe altered). Several studies have and run on 12.5% starch gels at 4 C. Three established that allopolyploids exhibit an in­ buffer systems were used to resolve 13 enzyme crease in number of isozymes due to the ad­ systems (Table 2). Gels were stained according dition ofdivergent genomes (Roose and Gott­ to procedures described in Soltis et al. (1983). lieb, 1976; Hart and Langston, 1977; Crawford, 1985; Werth, Guttman, and Eshbaugh, 1985; RESULTS-The number of isozymes ob­ Soltis and Reiseberg, 1986). Ifpolyploidy has served for each ofthe 13 enzyme systems was been involved in the production (evolution) of the same in all species tested. Sampling a small the variety of numbers present in the Onci­ number ofindividuals ofmost species enabled diinae, then at least some enzyme systems the distinction to be made between individuals should show evidence of this process by dif­ with multiple bands representing distinct iso­ ferences in the numbers of isozymes present. zymes and individuals exhibiting heterozy­ Conversely, if the species with lower number gosity at a single locus. Gel photographs offour represent lineages that originated by aneuploid of the enzymes are presented in Fig. 1. This reduction, then constancy of structural gene study included a phylogenetically diverse as­ number would be expected (Roose and Gott­ semblage oftaxa (within subtribe Oncidiinae) lieb, 1978; Gottlieb, 1981a, 1982; Crawford with a wide range of succulence, which made and Smith, 1982). uniform results difficult to obtain for all spec- 1082 AMERICAN JOURNAL OF BOTANY [Vol. 75 TABLE 2. Thirteen enzyme systems with the number ofloci observed and the buffer system used to resolveeach enzyme. All species examined exhibited the same numbers ofcoding loci Number of isozymes Enzyme observed Buffer system- Aspartate amino transferase (AAT) I 8b Aconitase (ACN) 2 9 Acid phosphatase (APH) I 8 Fructose-l,6-diphosphatase (F 1,6DP) 2 1 Glyceraldehyde-3-phosphate dehydrogenase (G3PDH) I 1 Isocitrate dehydrogenase (lDH) 3 9 Leucine aminopeptidase (LAP) I 8 Malate dehydrogenase (MDH) 3 9 6-phosphogluconate dehydrogenase (6PGD) 1 9 Phosphoglucose isomerase (PGI) I 8 Phosphoglucomutase (PGM) 1 9 Superoxide dismutase (SOD) 2 8,9 Triosephosphate isomerase (TPI) 2 8 a Soltis et aI., 1983. b Buffer system 8 is a modification of Soltis et al. (1983) as follows: gel 1 buffer: 0.003 M LiOH, 0.02 M boric acid, 0.33 M tris, and 0.005 M citric acid, pH 8.0; electrode buffer: 0.039 M LiOH and 0.263 M boric acid, pH 8.0. imens (see Fig. 1, 6-phosphog1uconate dehy­ for the 3-banded condition is unknown. Ifthis drogenase). Nevertheless, good results were is a case ofduplication, the simplerexplanation obtainedfor species with thelowest and highest is that the duplicate loci are the result ofchro­ chromosome numbers for all enzymes.