Proc. Natl. Acad. Sci. USA Vol. 90, pp. 4636-4640, May 1993 Plant Biology Flower-bud formation in explants of photoperiodic and day-neutral Nicotiana biotypes and its bearing on the regulation of flower formation (photoperiodism in plants/plant tissue culture) MANGALATHU S. RAJEEVAN* AND ANTON LANGt MSU-DOE Plant Research Laboratory, Michigan State University, East Lansing, MI 48824-1312 Contributed by Anton Lang, December 31, 1992

ABSTRACT The capacity to form flower buds in thin-layer other photoperiodic biotypes of Nicotiana, with a qual. and a explants was studied in flowering plants of several species, quantitative (quant.) photoperiodic response, and also another cultivars, and lines of Nicohana differing in their response to DNP, Nicotiana rustica. The latter was used because it was photoperiod. This capacity was found in all biotypes examined reported (5) to be recalcitrant to flower formation in vitro, and could extend into sepals and corofla. It varied greatly, explants forming only few flower buds and only after forma- depending on genotype, source tissue and its developmental tion of two leaves. We examined the capacity of the different stage, and composition of the culture medium, particularly the plants to form flower buds in thin-layer explants and its levels ofglucose, auxin, and cytokinin. It was greatest in the two variations; we discuss some of the information research on in day-neutral plants examined, Samsun tobacco and Nicohana vitro flowering provides to our understanding ofthe regulation rustica, where it extended from the inflorescence region down of flower formation in general. the vegetative stem, in a basipetally decreasing gradient; it was least in the two qualitative photoperiodic plants studied, the MATERIALS long-day plantNicotiana silvestis and the short-day plant Mary- AND METHODS land Mammoth tobacco, the quantitative long-day plant Nico- Plants and Growing Conditions. The Nicotiana plants used tiana alata and the quantitative short-day plant Nicotiana oto- were (i) N. tabacum L. cv. Samsun, (ii) N. rustica L. cv. phora line 38-G-81, where it was limited to the pedicels (and, in Brasilia (DNPs), (iii) N. tabacum cv. Maryland Mammoth some cases, the sepals). Regardlessofthe photoperiodic response (qual. SDP), (iv) N. silvestris Speg. et Comes [qual. long-day of the source plants, the response was the same in explants plant (LDP)], (v) N. alata Link et Otto and (vi) N. otophora cultured under long and short days. The finding that capacity to Griseb. line 38C-G (quant. LDPs), and (vii) N. otophora line form flower buds in explants is present in all Nicotiana biotypes 38-G-81 and (viii) N. tomentosiformis Goodsp. (quant. SDPs). studied supports the idea that it is regulated by the same Maryland Mammoth and N. silvestris were from lines propa- mechanism(s), regardless ofthe plant's photoperiodic character. gated in A.L.'s laboratories for ca. 40 years. Seeds ofSamsun However, the source plants were all in the flowering stage, and were kindly provided by K. Tran Than Van (Le Phytotron, no flower-bud formation can be obtained in explants from Gif-sur-Yvette, France), those ofN. rustica cv. Brasilia were strictly vegetative Nicotana plants. Hence, flower formation in provided by H. H. Smith (Brookhaven National Laboratory, the explants is not identical with de novo flower formation in a Upton, NY), and those ofN. tomentosiformis and the two N. hitherto vegetative plant: it is rather the expression of a floral otophora lines were provided by the U.S. Department of state already established in the plant, although it can vary widely Agriculture Tobacco Research Laboratory (Oxford, NC). All in extent and spatial distribution. Culture conditions that permit plants have been checked for their photoperiodic response flower-bud formation in an explant are conditions that maintain under our standard growing conditions. the floral state and encourage its expression; conditions under The SDPs were started in a greenhouse on 20-hr LDs and which no flower buds are formed reduce this state and/or after 7-10 wk were transferred to 8-hr SDs, to induce flower prevent its expression. formation. The LDPs were grown for 10 wk in SDs and then in LDs. The greenhouse temperature was maintained at an A large part ofresearch on the physiology offlower formation average of 28°C. The plants were grown in a greenhouse soil in the last decades has been conducted using explants from mixture or, in later experiments, a commercial planting vegetative parts ofplants. The first majorwork ofthis kind was mixture; apart from water they received three times weekly that of Aghion-Prat (1), who used entire stem sections as a fertilizer solution. explants. In 1973, Tranh Than Van (2) introduced the use of Tissue Culture. Explants were prepared from pedicels of small segments consisting ofa few ofthe outermost cell layers fully opened and of unopened flowers, the latter being 1-1.5 of the stem, and these "thin-layer explants," having several cm in length without the pedicel, the basal region of inflo- advantages over stem sections, have been used in most rescence branches, various positions in the main stem, and in subsequent work on "in vitro" flower formation. Most ofthis some experiments from sepals and corollae of unopened work was done with day-neutral (DN) cultivars of tobacco flowers. Most explants (5-7 mm long, three to six cell layers) (Nicotiana tabacum L.); attempts at obtaining flower forma- were prepared when the first (terminal) flower of the source tion in explants from photoperiodic biotypes of Nicotiana plant was 3-4 days past anthesis, but sepal and corolla were much fewer and largely unsuccessful (for a briefreview, explants (3-4 mm long, excised under a dissecting micro- see ref. 3). This apparent difference between DN and photo- periodic plants has even led to some question as to whetherthe Abbreviations: BA, N6-benzylaminopurine; DN(P), day-neutral regulation offlower formation in these plants is fundamentally (plant); LD(P), long-day (plant); NAA, a-naphthaleneacetic acid; different. Some years ago we (4) showed that thin-layer qual., qualitative; quant., quantitative; SD(P), short-day (plant). explants of the tobacco cultivar Maryland *Present address: U.S. Department of Agriculture-Agricultural Re- Mammoth, a qual- search Service, Microbial Products Research Unit, P.O. Box 5677, itative short-day plant (qual. SDP), were capable of forming Athens, GA 30613. flower buds. We have expanded this work, including several tTo whom reprint requests should be addressed. 4636 Downloaded by guest on September 27, 2021 Plant Biology: Rajeevan and Lang Proc. Natl. Acad. Sci. USA 90 (1993) 4637

scope) were prepared using source plants before any flower Table 2. Distribution of flower-forming capacity in explants was open. Unopened flowers were sterilized in toto before of Nicotiana excising explants; of the other source organs, large sections Explant source were sterilized and explants were excised from these. Ster- Plant Cor Sep Ped IB SI SII SIII ilization was for 5 min in ca. 1.25% NaOCl, preceded by dipping in 95% ethanol and followed by rinsing in sterile N. tabacum cv. ++ +++ +++ ++ + distilled water. Explants were excised under aseptic condi- Samsun 87 100 100 87 45 0 tions, using scalpels with exchangeable blades. N. tabacum cv. - + ++ The culture media consisted of macro- and micronutrients Maryland Mammoth 55 100 0 0 0 0 according to Murashige and Skoog (6), 100 mg of myoinositol N. silvestris - + + per liter, 0.1 mg of thiamine hydrochloride per liter, and 55 75 0 0 0 various amounts of glucose, the auxin a-naphthaleneacetic N. otophora + + acid (NAA), and the cytokinin N6-benzylaminopurine (BA). 38C-G 0 75 25 0 0 0 They were solidified with 0.64% agar and sterilized by N. otophora + autoclaving (15 min at 121°C). All cultures were grown in 38-G-81 0 72 0 0 0 0 Petri plates in controlled-environment rooms at 25C and 16 N. tomentosiformis + ++ + hr of light daily from fluorescent and incandescent lamps for 0 75 85 25 0 0 the DNP and LDP and 8 hr for the SDP. (These light regimes N. rustica + +++ +++ + + + were chosen before it was recognized that flower-bud for- 62 100 95 100 47 32 mation in the explants was independent of photoperiod; see Cor, corolla; Sep, sepal; Ped, pedicel; IB, inflorescence branches; Table 3.) The culture period was 4 wk (after which time the SI, main stem above node 1 (nodes were counted basipetally, node flower buds undergo little, if any, further development); the 1 being the first node bearing a leaf >5 cm in length); SII, main stem cultures were examined under a dissecting microscope to between nodes 5 and 10 (in N. silvestris, nodes 4 and 8); SIII, main count the number of flower and vegetative buds formed and stem below node 15. + + + = Ten flower buds per explant and above; determine the developmental state reached by the former. + + = five to nine buds per explant; + = four and fewer buds per explant; denominator = percentage ofexplants with flower buds; not tested. Media as in Table 1. RESULTS The extent of flower-bud formation in the explants varied Flower-Bud Formation in Nicotiana Explants: Extent and considerably, regarding its extent in the plant, the fraction of Variation. Our main results are compiled in Tables 1 and 2. Flower-bud formation was obtained in explants from all explants forming flower buds, the number ofbuds per explant, species, cultivars, and lines of Nicotiana examined, in some and the development of the buds. It was greatest in the two cases not only from vegetative parts but also from sepals and DNPs studied, Samsun tobacco and N. rustica cv. Brasilia. All in the one case where it was tried from the corolla. In all cases explants from pedicels and sepals of Samsun formed flower the buds formed directly in the original tissue, with no buds whose number could reach 20 and above. Flower buds preceding callus formation, although sometimes some callus were also obtained in explants from inflorescence axes and was formed simultaneously. Except in those from sepals and stem, in a basipetally decreasing gradient. The buds in Samsun pedicels of Samsun tobacco, the explants also formed some explants were mostly fully developed, having sepals, petals, vegetative (shoot) buds; the largest numbers (ca. 10-20) were stamens, and a pistil; some of them developed into nearly found in N. silvestris explants. mature flowers (Fig. 1, which shows such a flower on a calyx Table 1. Flower-bud formation in thin-layer explants from pedicels of flowering Nicotiana plants Growth regulator in medium, Explants Maximum AM Source of with floral Flower buds per development Species, cv., line BA NAA explant* buds, % explantt of budst N. tabacum cv. 1 1 OF 100 21.0 ± 5.1 III Samsun (DNP) N. tabacum cv. Maryland Mammoth (qual. SDP) 5 0.5 OF 100 5.8 ± 1.9 III N. silvestris 1 0.1 OF 75 2.0 ± 1.5 III (qual. LDP) FB 55 2.5 ± 2.0 I N. alata 20 2 OF 33.3 1.1 ± 0.8 I (quant. LDP) FB 83 4.0 ± 2.6 I N. otophora 38C-G 10 1 OF 25 0.5 ± 0.3 II (quant. LDP) FB 75 2.5 ± 1.8 II N. otophora 38-G-81 10 1 OF 0 0 (quant. SDP) FB 72.5 2.6 ± 1.8 II N. tomentosiformis 10 1 OF 0 0 (quant. SDP) FB 75 2.4 ± 1.7 II N. rustica cv. 25 2.5 OF 100 9.5 ± 4.6 III Brasilia (DNP) FB 100 16.0 ± 2.6 III All media contained 3% glucose, except that for Maryland Mammoth, which contained 6% glucose. Twenty to 40 explants were used per treatment. *OF, pedicels of fully opened flowers; FB, pedicels of unopened flowers. tMean ± SD. 4I, sepal initials only; II, sepal, petal, and stamen initials; III, complete flower buds. Downloaded by guest on September 27, 2021 4638 Plant Biology: Rajeevan and Lang Proc. NatL Acad. Sci. USA 90 (1993) were 75% and 1-3 (average, 2). In either plant, at least under optimal culture conditions, the buds were fully developed (Figs. 3 and 4). In N. alata and N. otophora 38-G-81, flower buds were obtained in 83% and 72% of the explants and the average numbers of flower buds per explant were 4 and 2.5, respectively; in N. otophora (Fig. 5), however, the buds reached the stamen stage, but in N. alata only the sepal stage. The quant. LDP N. otophora 38C-G and the quant. SDP N. tomentosiformis occupy a position intermediary between the two DNPs studied and plants like N. silvestris and Maryland Mammoth. Flower buds were formed in explants from pedi- cels and inflorescence branches, and in N. tomentosiformis also from the uppermost part of the main stem, in about three-quarters ofthe pedicel explants and at an average ofca. 2.5 buds per explant. Factors Influencing Flower-Bud Formation in Nicoiana Ex- plants. In searches for factors that may regulate flower-bud formation in Nicotiana explants we found two that have a great influence on this process-namely, the age (i.e., the developmental state) ofthe source tissue and the composition of the medium (Table 1). the tissue age FIG. 1. Calyx explant from an unopened flower of DN tobacco culture Regarding Samsun. Several flower buds are evident; one has developed into an factor, in Samsun and Maryland Mammoth tobacco and in N. almost mature flower. For this and all other figures, the culture silvestris, flower-bud formation was greatest in explants taken medium is as in Table 1. Explants in all figures were photographed from pedicels of open flowers; in the other plants, flower-bud after 4 wk of culture. (x4.) formation was greatest in explants from unopened flowers- e.g., in N. silvestris the extent of flower-bud formation in explant). In Brasilia, numerous, mostly fully developed buds explants of either age was similar but buds in explants from were formed in the central portion of the explant (Fig. 2), unopened flowers reached only the sepal stage. Conversely, in sometimes with calli on the lateral parts. N. alata, N. otophora, N. tomentosiformis, and N. rustica cv. In explants from the two qualitative photoperiodic plants Brasilia, flower-bud formation in explants from the older examined, the LDP N. silvestris and the SDP Maryland pedicels was reduced; in two of these plants such pedicel Mammoth, and from the quant. LDP N. alata and the quant. explants produced neither flower nor vegetative buds-i.e., SDP N. otophora 38-G-81, flower buds were obtained only in seemed to have lost all proliferation capacity. The influence of explants from pedicels and in the two first-named plants also medium composition is also pronounced. As shown in our in sepal explants. (In N. alata, no other explants were tried.) previous report (4) optimal results with Maryland Mammoth In Maryland Mammoth 100% of the explants formed an explants were obtained with 6% glucose; explants from all average ofca. 6 buds per explant; in N. silvestris the numbers other biotypes did better with 3%. For explants of DN to-

FIG. 2. Calyx explant from an open flower of N. rustica cv. Brasilia (DNP). Flower buds are in the center part; callus is at the sides. (X6.) Downloaded by guest on September 27, 2021 Plant Biology: Rajeevan and Lang Proc. Natl. Acad. Sci. USA 90 (1993) 4639

FIG. 3. Complete flower bud in a pedicel explant from an open flower of Maryland Mammoth (qual. SDP). s, Sepal; p, petal; st, stamen; o, ovary. (x20.) bacco, 1 ,uM (each) NAA and BA was optimal but forthe other biotypes studied the optimal levels of the two growth regula- FIG. 5. Flower buds in a pedicel explantfrom an unopened flower tors can differ by up to an order of magnitude. In Maryland of N. otophora 38-G-81 (quant. SDP). (x 18.) Mammoth, raising the level of BA to 5 uM and reducing the NAA level to 0.5 ,uM greatly improved the response (see ref. different photoperiodic types: DN, LD, SD-the latter two 4); in N. silvestris best results were obtained with as little as with a quant. and a qual. day length requirement. Approaches 0.1 ,lM NAA in the medium. In explants from quant. LDP and important in obtaining these results were the use of pedicels quant. SDP greatest flower-bud formation was found with 10 as source organ for explants, variations in the age of this ,uM or 20 ,uM BA and 1 or 2 ,uM of NAA and in the DNP N. source organ, and variations in the composition ofthe culture rustica with as much as 25 and 2.5 ,uM, respectively. medium. That pedicels were the optimal source organ for Flower-Bud Formation in Explants Cultured in LD and SD. obtaining flower formation in thin-layer explants had been Since our experiments dealt with plants of different photo- shown for DN Samsun tobacco by van den Ende et al. (7); periodic behavior it was obviously important to know following their example we included pedicel explants in all whether the response of the explants themselves was influ- experiments, and in three of the seven plants examined for enced by photoperiod. This question was studied with ex- distribution of flower-bud-forming capacity, pedicels were plants from pedicels ofN. silvestris, Maryland Mammoth, N. the only vegetative organs whose explants formed flower otophora 38C-G, and N. otophora 38-G-81. The results buds (in two of them, Maryland Mammoth and N. silvestris, (Table 3) showed that flower-bud formation (and also forma- flower buds were also formed in explants from sepals). The tion of vegetative buds; not shown) was the same in 16- and age ofthe source organ also proved critical. In two plants (N. in 8-hr photoperiods. otophora 38-G-81, N. tomentosiformis) only pedicel explants from unopened flowers formed flower buds; in others the DISCUSSION differences between explants from open and unopened flower were quantitative but sometimes quite striking, in some cases Our results show that the potential for the formation offlower better results being obtained with the former and in others, buds is present in explants of Nicotiana plants belonging to with the latter. Lastly, although 3% glucose in the medium was best for explants of the other seven plants examined, explants of Maryland Mammoth did distinctly better on 6% glucose (see ref. 4). The higher glucose demand in Maryland Mammoth is not an osmotic phenomenon since replacing 3% of the glucose with mannitol, a compound of similar molec- Table 3. Flower-bud formation in pedicel explants of photoperiodic Nicotiana plants cultured in long and short days (16 and 8 hr of light daily, respectively) Explants with Flower buds Plant Photoperiod floral buds, % per explant* Maryland Mammoth SD 100 7.1 ± 2.1 (qual. SDP) LD 100 6.4 ± 1.6 N. silvestris LD 75 2.0 + 1.5 (qual. LDP) SD 78 1.3 ± 1.1 N. otophora 38C-G LD 75 2.5 + 2.4 (quant. LDP) SD 78 3.4 + 2.4 N. otophora 38-G-81 SD 75 2.6 ± 1.8 (quant. SDP) LD 72 1.9 ± 1.7 FIG. 4. Flower buds in a pedicel explant from an open flower ofN. Media as in Table 1. silvestris (qual. LDP). s, p, st, and o as in Fig. 3. (x40.) *Mean ± SD. Downloaded by guest on September 27, 2021 4640 Plant Biology: Rajeevan and Lang Proc. Natl. Acad. Sci. USA 90 (1993) ular weight but nonmetabolizable for tobacco explants (8), McDaniel and coworkers (11, 12) on positional pattern and did not duplicate the effect of 6% glucose (data not shown). stability of floral determination in Wisconsin 38 and Mary- Although medium with 1 uM (each) auxin and cytokinin is land Mammoth tobacco and in N. silvestris. Their findings widely used in work with explants of DN tobacco, better show that a florally determined state is reached in all these flower-bud formation in all other biotypes we studied was plants but is lesser in the two photoperiodic plants than the obtained with different levels of one or both regulators. Our DN one. Such determination states are of course well known results do not demonstrate clear-cut correlations between the in different phases of plant development-e.g., in shoot and effect of the experimental variations and the photoperiodic root apices, phase changes, and habituated cell cultures (e.g., these ref. 13). They are induced by specific internal or external nature ofthe plants, but it ought to be borne in mind that signals acting at appropriate developmental stages and can be variations-and the entire experiments with explants from very stable but can be reversed under the influence of other sepals and corolla-were relatively limited.t factors. The finding that, regardless of their responses to photope- Our approaches provide no information on the processes riod, all Nicotiana biotypes included in our studies can form that result in the floral state nor its variations. One conclusion flower buds in explants indicates that the "mechanism" or can, however, be made. The floral state does not seem to be mechanisms regulating this response are the same, at least correlated with the hormone-like "signals" known to partic- qualitatively. When assessing the bearing of these results on ipate in the regulation of flower formation in whole plants. our understanding of the regulation of flower formation in The existence of graft-transmissible flower-inducing materi- general it is essential to recognize that all research on in vitro als of this kind ("florigen") has been demonstrated in many flower formation in Nicotiana, including ours, was per- diverse plants, including DN, SD, and LD species and formed using explants from plants that were in the flowering cultivars of Nicotiana (3). More recently, quite similar but state or at least very close to it. McDaniel et al. (9) have flower-inhibiting materials ("antiflorigen") have been shown recently obtained, using large stem sections of young, dis- to exist in N. silvestris (and another LPD, Hyoscyamus niger tinctly vegetative plants of a DN tobacco (cv. Wisconsin 38), (3), whereas McDaniel et al. (11) have shown the existence of some, although quite slight (2%), flowering response. How- other transmissible flower inhibitors in the roots of DN ever, with typical explants from strictly vegetative plants, tobacco (Wisconsin 38) and N. silvestris. However, N. whether stem sections or thin layers, only vegetative buds silvestris is a potent "donor" for florigen and antiflorigen, have been obtained (e.g., refs. 1, 2, 10). Flower formation in whereas Maryland Mammoth produces less florigen and, if at explants from flowering Nicotiana plants is not the same as all, only very low levels of antiflorigen (3); yet the degrees of flower initiation de novo as it takes place in the whole plant the floral state in the two plants are quite similar. And, when it is progressing from the vegetative to the reproductive whereas Wisconsin 38 and N. silvestris produce the root phase of its life; rather, it is the manifestation of a "floral inhibitor, their floral states are quantitatively quite different. state" that is established in the plant after it has completed Results of Bridgen and Veilleux (14) with grafts between its vegetative phase. This interpretation is strongly supported tobacco explants capable and not capable of flower-bud by the finding that flower-bud formation in the explants is formation point in the same direction: the graft partners independent of day length (Table 3). Plants that have reached behaved strictly according to their inherent "program." The the floral state are past the stage when flower formation is floral state is evidently generated and remains localized in the regulated by photoperiod. Culture conditions under which cells. flower buds are formed are conditions that maintain the floral state in the explants and permit its manifestation when the Excellent technical assistance of Sharon DeBar and very useful correlations existing in the intact comments by Prof. Carl N. McDaniel (Rensselaer Polytechnic In- explant is removed from the stitute, Troy, NY) are gratefully acknowledged. This work was plant. Conditions under which no flower buds are formed in supported by the U.S. Department of Energy under Contract DE- explants inhibit this state and might lead to its disappearance. AC02-76ERO-1338 and the National Science Foundation through Studies of flower formation in explants from flowering Ni- Grant PCM-7910791. cotiana plants, and plants in the flowering state in general, may provide important insights into the nature of the floral 1. Aghion-Prat, D. (1965) Physiol. Veg. 3, 229-303. state and the development of flowers but not into the tran- 2. Tran Thanh Van, K. (1973) Planta 115, 87-92. sition ofthe hitherto vegetative plant to flower initiation-the 3. Lang, A. (1989) in Handbook ofFlowering, ed. Halevy, A. H. single most crucial event in the life of a seed plant. (CRC, Boca Raton, FL), Vol. 6, pp. 427-483. Though the floral state is reached in all Nicotiana plants 4. Rajeevan, M. S. & Lang, A. (1987) Planta 171, 560-564. it can be expressed to quite a different degree. 5. Tran Thanh Van, K. & Cousson, A. (1982) in Variability in investigated, Plants Regeneratedfrom Tissue Cultures, eds. Earle, D. D. & It appears to be strongest in the DNPs Samsun and N. rustica Demarly, Y. (Praeger, New York), pp. 121-139. where it extends into relatively old portions of the stem in a 6. Murashige, T. & Skoog, F. (1962) Physiol. Plant. 15, 473-497. basipetally decreasing gradient. For DN tobacco cultivars 7. van den Ende, G., Croes, A. E., Kemp, A. & Barendse, this is well known from earlier reports (e.g., refs. 1-3); in N. G. W. M. (1984) Physiol. Plant. 61, 114-118. rustica it may in fact be even somewhat more extended than 8. van den Ende, G., Barendse, G. W. M., Kemp, A. & Croes, in Samsun tobacco. In the photoperiodic biotypes it was A. F. (1984) J. Exp. Bot. 35, 1853-1859. more limited, to the pedicels (and in some, the sepals) or 9. McDaniel, C. N., Sangrey, K. A. & Jegla, D. E. (1989) Dev. extending not further than the youngest parts of the stem. A Biol. 134, 473-478. clear correlation with the photoperiodic type was not evident: 10. Konstantinova, T. N., Aksenova, N. P., Bavrina, T. V. & was found in LDPs and Chailakhyan, M. K. (1969) Dokl. Akad. Nauk SSSR 187, 466- similar extent of the floral state SDPs, 469. and in qual. and quant. types. Our results on the extent of the 11. McDaniel, C. N., Singer, S. R., Gebhardt, J. S. & Dennin, floral state in the plant are in good agreement with results of K; A. (1987) in Manipulation ofFlowering, ed. Atherton, J. G. (Butterworths, London), pp. 109-120. tWe have no explanation for the difference ofour results in N. rustica 12. Gebhardt, J. S. & McDaniel, C. N. (1987) Planta 172, 526-530. from those of previous authors (5), who obtained flower-bud for- 13. Wareing, P. F. (1987) in Manipulation ofFlowering, ed. Ath- mation in this material, if at all, only after formation of two leaves, erton, J. G. (Butterworths, London), pp. 83-92. except speculation about subtle and possibly unrecognized differ- 14. Bridgen, M. P. & Veilleux, R. E. (1985) J. Am. Hort. Soc. 110, ences in the culture of the source plants and the explants. 233-236. Downloaded by guest on September 27, 2021