RHODORA, Vol. 99, No. 898, 152-160, 1997 pp. NATURAL HYBRIDIZATION THE GENUS IN CIRSIUM: ALTISSIMUM DISCOLOR—CYTOLOGICAL AND X C. C. MORPHOLOGICAL EVIDENCE Simon Dabydeen Biology Department, Frostburg State University, MD 21532 Frostburg, abstract. Cytological and morphological evidence indicates natural hy- bridization between Cirsium altissimum and C. discolor. postulated that It is the high degree of hybrid sterility correlated with meiotic abnormalities. is Key Words: Cirsium altissimum, Cirsium discolor, hybridization, hybrid Nebraska. sterility, In the genus Cirsium (family Asteraceae, Cardueae), tribe vari- and make ability intergradation of diagnostic characters morpho- many logical circumscription of species challenging and difficult. The difficulty partly explained by the occasional breakdown of is between sterility barriers closely related species, thus permitting (Ownbey interspecific hybridization Davidson 1951, 1964; 1963; Bloom Dabydeen 1977; 1987). Cirsium altissimum (L.) Spreng. and discolor (Muhl.) C. two Spreng. are biennial, sympatric species sharing similar mes- ophytic habitats within an extensive eastern North American range. Both species have a similar extended flowering ohenologv: structures immature simultaneously on same the plant. generally believed that It is members of the genus Cirsium employ both autogamous and al- logamous reproductive strategies. Petrak (1917) placed Cirsium altissimum and C. discolor in his A Altissima series to reflect close affinity. closer relationship is m C Oray seen s (1884) treatment of discolor as a variety of altissimum Cnicus (as altissimum). C Phenotypically, Cirsium altissimum and discolor are easily and distinguishable are separated primarily by the degree of lob- ing or indentation of the leaves. According Frankton and to C Moore (1963), in discolor the "cauline leaves deeply are pin- C narrow natifid into linear-lanceolate lobes." and altissimum in 152 — Dabydeen Cirsium 1997] Hybridization 153 the "cauline leaves are entire or subentire with shallowly sinuate lobes." Cytologically, in Cirsium altissimum and discolor chromo- C. some numbers have become extremely reliable in the determi- nation of the species, even though supernumerary chromosomes have been observed infrequently in C. discolor (Frankton and Moore Ownbey Bloom 1963; and Hsi 1963; 1977). Moore Frankton and measured chromosome (1963) somatic = = lengths of Cirsium and discolor (2n 20) C. altissimum (2n and found no between 18) significant differences the chro- total matin ) lengths of these taxa nor in the size ranges of the individual chromosomes. same Apparently, both genotypes contain the amount They of material. argued that the reduction in the chro- mosome number may of altissimum have been accomplished C. chromosomes chromosomes Bloom mentioned of (1977) that his results initial studies (un- C showed published) of Cirsium altissimum and discolor hybrids chromosomal two by that arrangement of these species differ a minimum of four reciprocal translocations and one paracentric inversion. Observation of a Cirsium plant with an unusual leaf lobing from mixed pattern (in Nebraska just across the Missouri River a population of C. altissimum and C. discolor plants in Iowa) prompted The mixed population was growing this investigation. in an area approximately by ten feet on top of the Missouri thirty River The was mostly with small bank. vegetated, area lightly shrubs and Beyond riverbank patches of shrubs and herbs. the tall few what appeared be abandoned a small dominated to trees in pasture lands. based on In Cirsium altissimum and discolor, reports anal- C. yses of morphological have revealed evidence of natural variation Ownbey To author's hybridization (Davidson 1963; 1964). this knowledge, no evidence supporting spontaneous hy- cytological bridization between altissimum and C. discolor exists in the C. chromosomal of ev- literature. Thus, paper the report this is first two idence of natural hybridization between these taxa. AND METHODS MATERIALS Cirsium In July 1990, a plant with features similar to altissi- and • 154 Rhodora 99 [Vol. riverbank This yards west of the Missouri River linking Iowa. This plant was one of a few plants left intact after a recent was brush-cut. observed that dried-up remnants of additional It Cirsium formed plants part of the debris, but no other live Cir- sium plants were observed in the vicinity of the putative hybrid riverbank mixed C C ulation of nine altissimum and seven discolor plants grew. No plants phenotypically from two were different the species C found at this The occurrence of a mixed altissimum and site. proximity brush-cul and/or their derivatives. The putative hybrid specimen was sampled in addition to spec- imens of the Iowa Cirsium altissimum and mixed discolor C. Two population. immature capitula and twenty- seven cypselae long aborted ovaries Seven immature ant. three and immature and five capitula thirtv-three mature mixture V/V). Cypselae were germinated chromosome Microsporocyt method Chromosome but firm itle r Chromosomes lip. microscope-earner X with 97 tern 5 ocular Data for morphological comparison were from mea- obtained surements of nineteen characters (Table The mean values of 1). measurements 10 for each character of each were plant used. ran and — Dabydeen 1997] Cirsium Hybridization 155 Cirsium discolor: iowa, Freemont Co., riverbank, 50 yards west of the Missouri River interstate bridge linking Nebraska 27 City, Jul Dabydeen 1990, 9001, 9002, 9003 (neb). Cirsium altissimum: iowa, Freemont Co., riverbank, 50 yards west of the Missouri River interstate bridge linking Nebraska City, 27 Jul Dabydeen 1990, 9004, 9005, 9006 (neb). X Cirsium altissimum Cirsium Nebraska, Otoe discolor: Co., roadside riverbank, 75 yards west of the Missouri River interstate bridge linking Iowa, 27 Jul 1990, Dabydeen 9007 (neb). RESULTS = Cirsium altissimum has a chromosome number of 2n 18 (Figures la and chromosome complement In C. discolor, the If). = is 2n 20 (Figures lb and The suspected hybrid individual lc). has a complement of 19 chromosomes (Figure All chromo- Id). some complements observed metaphase composed mitotic are at of metacentric submetacentric chromosomes. to In most florets of the two immature capitula of the suspected hybrid, microsporogenesis had been completed and only already, a few anthers with microsporocytes two stages of meiosis were at found. Meiotic analysis of the putative hybrid specimens revealed that some of the bivalents were closely associated on an end-to- end basis and chain formation. In Figure a chain illustrated 2g, of six chromosomes, six bivalents, and one univalent (arrow) are shown. Microsporogenesis was not normal; one of the at least four microspore appeared small (Figure nuclei relatively le). Only three of twenty-seven (11%) cypselae of suspected hybrid origin germinated; cypselae of both Cirsium altissimum and C. 100% discolor had germination. Measurements of nineteen characters (Table of the suspected 1) com- hybrid showed were intermediate individual that fourteen pared Three characters were approximately to the parental types. the same as either parental type, one character was larger, and one smaller than either of the parental types. DISCUSSION hybrid phenotypes often In interspecific hybridization events, many express morphological intermediacy of parental characters. In the suspected Cirsium hybrid specimen, fourteen characters, including four leaf features, were intermediate between those of Rhodora 156 99 [Vol. a Figure Photomicrographs of mitotic and meiotic chromosome comple- 1 . X ments of the genus Cirsium, 800 unless otherwise stated, C. altissimum: a. = = mitotic metaphase, 2n 18. b. C. discolor: mitotic metaphase, 2n 20. c. X C. discolor microsporocyte: metaphase 10 bivalents. d. C. altissimum C. I, X discolor, mitotic metaphase, 2n 19. C. altissimum C. discolor: ab- e. X normal microsporogenesis and micronucleus (arrow) 300. C. altissimum f. metaphase 9 microsporocyte: bivalents. I, — Dabydeen 1997] Cirsium Hybridization 157 <3 o Q> A Figure 2. photomicrograph (X800) and an interpretive drawing of the » meiotic chromosome complement of the genus Cirsium. C. altissimum X g. microsporocyte arrow formation arrow Table Comparison of character measurements of Cirsium altissimum, 1. C. discolor, and their hybrid derivative (all measurements in millimeters). altissimum C. X C. altissimum C. discolor C. discolor Mean Mean Mean Characters S.E. S.E. S.E. Involucre: length 29.7 0.48 28.2 0.85 26.8 1.01 width 21.6 0.85 22.3 0.85 23.7 1.03 Phyllaries: no. of rows 12.3 0.15 11.5 0.25 11.2 0.33 spine length 3.7 0.15 3.7 0.15 4.0 0.26 inner: length 26.1 0.53 24.3 0.47 22.4 0.70 inner: width 1.7 0.09 1.5 0.13 1.5 0.17 Pappus length 26.1 0.44 25.1 0.50 24.7 0.75 Corolla: lobe length 8.6 0.22 8.0 0.26 8.0 0.33 tube length 0.18 8.2 0.20 6.0 0.26 9.1 Stamen length 17.5 0.36 16.4 0.40 13.3 0.34 Stigma length 0.14 6.1 0.23 5.0 0.29 6.1 Style length 32.9 0.78 32.3 0.54 31.8 1.33 Leaf: length 100.8 3.78 103.4 1.96 112.5 3.57 0.62 width 18.4 1.30 16.4 0.62 15.4 0.90 22.0 0.77 lobe length 1.04 20.2 18.5 lobe base width 11.3 0.82 9.3 0.42 10.2 0.49 0.29 spine length 4.4 0.18 5.0 0.26 4.8 Intemode length 22.3 1.27 21.1 1.20 20.6 1.11 Flower length 39.0 0.84 39.0 1.62 36.0 1.33 Rhodora 158 99 [Vol. Ownbey C. altissimum and C. discolor (Table (1951) inves- 1). tigated C. discolor, C. altissimum, and their putative hybrid de- rivatives and determined that plants with intermediate leaf types were hybrids. = The diploid chromosome complement of 2n 18 in Cirsium altissimum has been reported previously and number the re- is Moore Ownbey garded as stable (Frankton and 1963; and Hsi Johnson chromosome num- 1963; 1974). In C. discolor, a diploid = ber of 20 has been determined (Frankton and Moore 1963; 2/i Ownbey Bloom and Hsi 1963; 1977). In this taxon, accessory chromosomes have been occasionally observed, but in a closely C related congener, muticum, frequent observations of extra chro- mosomes have been reported (Frankton and Moore 1963). Accessory chromosomes do usually not exert phenotypic influ- ence. Thus, based on morphological data (Table and also cy- 1) tological features (Figures Id and seems very likely that 2g), it chromosomes the nineteen observed in the putative hybrid spec- imen are products of interspecific hybridization between Cirsium altissimum and These chromosomes, C. discolor. nineteen there- fore, likely do not constitute a genotype of a C. discolor specimen with supernumerary chromosome. a probable 10 chro- that It is mosomes must have come from C. discolor and 9 chromosomes from C. altissimum. The observation of a chain of six chromosomes, six bivalents, and one univalent (Figure 2g) during meiosis of suspected hybrid microsporocytes suggested that 9 chromosomes of Cirsium altis- C simum and 9 chromosomes of discolor are homologous or homeologous, and that reciprocal translocation involved in the is formation of the meiotic The must have configuration. univalent Similar formation) Bloom other Cirsium taxa (Dabydeen 1987; 1977). nation of micronuclei during microsporogenesis wa: chromosomal to deficiency from meiotic ab Lie arising imilar micronuclei formation observed microsporo in X Cirsium flodmanii undulatum was C. hybrids attrib ging chromosomal (Dabvdeen 19871 materials The manv of failure ovaries may There external evidence of damaee would nrevent that fruit — — Dabydeen Cirsium 1997] Hybridization 159 Many formation. of the aborted ovaries were shriveled or ex- tremely The extremely low flattened. rate of viability of cypselae A may be due to hybridity. high degree of of Cirsium sterility hybrids has been (Ownbey Bloom reported 1951, 1964; 1977; Dabydeen Bloom 1987). (1977) attributed the high degree of ste- irregularities arising chromosomal rearran mixed Cirsium discolor and C. altissimum not expected. Hybridization is events have occurred many where two in instances these taxa proximity to unknown te, and/or It is suggested that the rather small size and possible recent mixed the absence of hybrids. At the Nebraska highly possible site, is it may that the brush-cut have destroyed a well-established mixed population, including additional hybrids. acknowledgments. The M. author wishes thank K. to McCarty, R. C. Lommasson, Roland, and K. Planalp for their J. advice and assistance. LITERATURE CITED W. Chromosomal between Cirsium discolor L. 1977. differentiation and Bot. 1-13. 2: Dabydeen, genus Cirsium: Cflodmanii 1987. Natural hybridization in the S. X Rhodora 369-373. undulatum. C. 89: Davidson, R. A. 1963. biometric survey of morphological variation In—itial 222-241. in Cirsium altissimum C. discolor complex. Brittonia 15: Frankton, and Moore. Cytotaxonomy of Cirsium muticum, C. R. 1963. J. Cirsium and Cirsium altissimum. Canad. Bot. 41: 73-84. discolor, J. — Feulgen, R. and H. Rossenbach. 1924. Mikroscopisch chemischer Na- vom Hoppe chweis Typus Thymonucleinsaure. Einer Nucleinsaure der Seyl 203-248. Z. 135: Amer. Acad. Arts Johnson, M. Cirsium, Carduus, 1974. Cynareae (Asteraceae) in Virginia. F. Onopordum. 152-160. Virginia Sci. 25: J. — C Ownbey, genus Cirsium G. B. 1951. Natural hybridization in the I. X muticum Michx. Torrey discolor (Muhl. ex Willd.) Spreng. C. Bull. Bot. Club 233-253. 78: Rhodora 99 160 [Vol. — — genus Cirsium altissimum 1964. Natural hybridization in the C. II. . C X Michigan 87-97. Bot. — discolor. 3: and Chromosome numbers some North American Y. Hsi. 1963. in Rhodora 339-354. species of the genus Cirsium. 65: Petrak, Die Nordamerikanischen Arten der Gattung Cirsium. Beih. 1917. F. 223-567. Bot. Centralbl. 35: Snow, R. 1963. Alcoholic hydrochloric acid carmine as a stain for chro- mosomes 9-13. in squash preparations. Stain Technol. 38: