Seed Germination of North American Orchids. I. Native California and Related Species of Calypso, Epipactis, Goodyera, Piperia, and Platanthera

Seeds of native California and related orchids were germinated in vitro on five basal media and 43 modifications. Germination periods extended from several weeks to almost 2 yr. Seeds from immature capsules germinated faster and in higher proportions than those from mature fruits. The most suitable media for orchid germination are those devised for the culture of barley embryos (Norstog) and a modified Curtis medium containing urea and calcium carbonate instead of ammonium nitrate and calcium nitrate.


Introduction
Attempts to germinate seeds of terrestrial orchids have been partially successful (DOWNIE 1940(DOWNIE , 1949KNUDSON 1941, STOUTAMIRE 1965HADLEY 1970;HARVAIS 1974;FAST 1976;CLEMENTS and ELLYARD 1979). Methods successful for one species are not always applicable to others, and procedures for orchids from one region may not be suitable for those from another. Because of this and since they are much more dependent on their mycorrhizal fungi for germination than epiphytic species (STOUT-AMIRE 1974;WARCUP 1975), seeds of terrestrial orchids are difficult to germinate in vitro. Since the available information is mostly empirical, very little is known about the germination and development of terrestrial orchids, especially those from North Temperate regions (reviews by WITHNER ized by immersion in filtered calcium hypochlorite solution (7 g/100 ml distilled water) for 10 min and opened under sterile conditions; their contents were scraped and placed on the agar surface. Mature seeds were sterilized by soaking in the sterilizing solution for 10 min. Short glass tubes (cut from disposable pipettes) stuffed with cotton at both ends and repipetting bulbs were used to sterilize the seeds, to wash them with sterile distilled water, and to dispense them into culture flasks (HARRISON 1970;HARRISON and ARDITTI 1970).
Modifications of the Curtis medium (CURTIS 1936(CURTIS , 1943 were used for the germination of all species (tables 1, 2). Additional media were employed for Epipactis gigantea and Calypso bulbosa (tables 1, 2). The fungus, Rhizoctonia repens (provided by Dr. GEOFFREY HADLEY, University of Aberdeen) was cultured on a shaker (60 oscillation/min) in a defined medium (POWELL and ARDITTI 1975) for 1 mo, and the liquid was freed of hyphae by filtration. The filtrate was sterilized by Millipore filtration (pore size 0.22 j.m) and incorporated into culture media at 100 ml/liter. Seeds were germinated and seedlings maintained under several combinations of illumination, pH, and concentration and composition of culture media (tables 1-7).

Results
Seeds in these experiments germinated by first forming protocorms, which is typical for orchids (ARDITTI 1967(ARDITTI , 1979. In each species ca. 90%o of the protocorms were initially white, even under illumination, but turned green with time (tables 2-4, 6, 7). Some protocorms of Calypso bulbosa and Platanthera saccata were green from the outset (tables 3, 7). In Epipactis, Goodyera, Piperia, and Platanthera (the last two were at one time part of the same genus, Habenaria), rhizome formation generally occurred after protocorms appeared (tables 4, 6, 7). No rhizomes formed in Calypso (table 3) developed shoots before the appearance of rhizomes (tables 4, 6, 7). Roots appeared following the shoots in all species (tables 3, 4, 6, 7), after which the seedlings developed into small plantlets.
Mature seeds of C. bulbosa germinated only on the Norstog medium and on a modification of the Curtis solution. Immature seeds germinated on both fulland half-strength Curtis medium (tables 2, 3).
Epipactis gigantea seeds from immature capsules germinated better than those from ripe fruits (table  4). Ethrel at 5-20 mg/liter improved germination only marginally, and the solvent, 20 ml/liter of 70%7 ethanol, had a similar effect (table 4). There was no germination on media containing 2.5 g/ liter of Ethrel or 2 g/liter of graphite. Germination does not seem to have been affected by changes in pH ranging from 5 to 7.5 (table 4). Seeds of E. gigantea germinated well on the Norstog medium; those of E. atrorubens did not.
In the dark, protocorms were always white, and most turned green on being moved to light. However, some protocorms produced by immature seeds of E. gigantea (tables 4, 5) developed white rhizomes, roots, and leaves and formed plantlets in small bushy clumps ca. 3 cm high. These clumps increased in diameter and number of shoots but not in height, and some have not turned green after 2 yr of culture  Curtis (1936Curtis ( , 1943, HC = half-strength Curtis (1936Curtis ( , 1943; MC = modified Curtis (1936Curtis ( , 1943, same medium as full-strength Curtis except that CaNOs and NH4NOs were replaced with 250 mg/liter urea and 148 mg/liter CaCOa; N = Norstog (1973); RE4 = Robert Ernst medium minus peptone plus vitamins and hormones; RE6 = Robert Ernst, personal communication; vit = vitamins; WS = Wolter-Skoog (1966). The listing of two or more media (separated by commas) is an indication of similar results. The temperature was 22 + 2 C for both light and dark cultures; light intensity was ca. 0.8 mw/cm2; photoperiods were 16 h. For details see tables 4-7.
c Seeds from immature capsules. d Seeds from mature capsules.
(tables 4, 5). The addition of naphthaleneacetic acid (0.1 mg/liter) and benzyladenine (1 mg/liter) separately and together or only kinetin (1 mg/liter) had no visible effects on color and growth of these seedlings (table 5). Irrespective of medium, some of them spontaneously produced green shoots and roots after ca. 18 mo of culture. Seed lots of G. oblongifolia differed in percentage of germination. The addition of cytokinins or vitamins improved the germination of some seeds. On fulland half-strength Curtis medium plus cytokinin, germination was better in the dark. Goodyera tesselata formed plantlets faster than G. oblongifolia (table 6).
Except in one instance, seeds of Platanthera saccata germinated poorly within 6 mo (tables 2, 7). Percentage of germination on fulland half-strength Curtis medium was not affected by illumination. On a modified Curtis medium, germination was better under illumination than in the dark. Addition of Rhizoctonia repens filtrate did not enhance germination in either the dark or light (table 7). Platanthera hyperborea and Piperia maritima germinated on the Norstog medium while P. saccata did not; P. hyperborea germinated rapidly but at a low percentage on the modified Curtis medium. Piperia elegans produced only white protocorms on a vitamin-enriched Curtis solution (table 7).
Several times, seeds of P. saccata on a Curtis medium under illumination formed protocorms which developed enlarged roots, 6-8 cm long, with numerous absorbing hairs. Green shoots remained 1 cm long or less or were entirely absent (table 7). Seeds in one culture (of two) from a single batch germinated underneath the agar surface and de-veloped into white protocorms. The protocorms enlarged but did not differentiate further or produce chlorophyll until they were transferred to the surface of fresh medium (table 7). Enlarged roots with very small shoots also developed when P. saccata was cultured on half-strength Curtis medium or full-strength Curtis solution with vitamins (table  7). Seeds on the vitamin-enriched solution germinated in the dark and were subsequently moved to the light. Piperia and Platanthera protocorms also developed in a manner similar to that of other orchids.

Discussion
Orchid seed germination differs from that of other seeds because of the absence of an endosperm, radicle, and leaf rudiments. Swelling of the embryo is followed by the formation of a round, top-shaped body called a protocorm. Other organs subsequently appear (see ARDITTI 1967ARDITTI , 1979. We have defined germination (tables 2-7) as the formation of green or white protocorms. Development is discussed in terms of the appearance of chlorophyll, absorbing hairs, rhizomes (in some species), shoots, and roots (tables 2-7).
Calypso bulbosa seeds from ripe capsules germinated very poorly or not at all (table 3). However, the germination of immature seed was as high as 80% (table 3). Differences were also evident in the germination of ripe and unripe seeds of Epipactis gigantea (table 4). This suggests that seeds of these species become dormant as they mature, which is in line with previous reports regarding orchid seed dormancy, suggesting it may exist only in some sp ecies (CURTIS 1936;VERMEULEN 1947;KNUDSON   Cd - ; Z4 1950;LIDDELL 1953;MCINTIRE, VEITCH, and WRIGLEY 1972;WILDHABER 1974;STOUTAMIRE 1974;WRIGLEY 1976;ARDITTI 1979). Another possibility is simply a decrease in viability with maturation. Illumination reduced germination in C. bulbosa and E. gigantea and had a varied effect on Goodyera oblongifolia, G. tesselata, and all species of Piperia and Platanthera (tables 2-4, 6, 7). Seeds of epiphytic orchids germinate in both the light and dark but require illumination for further development (AR-DITTI 1967, 1979, which is also true for some terrestrial species. Other terrestrials germinate best in the dark or are inhibited by light (ARDITTI 1967(ARDITTI , 1979STOUTAMIRE 1974). Thus, generalizations regarding the effects of light on orchid seed germination are not possible because of insufficient information. Our findings may require a reexamination of the view that the inhibitory effect of light on the germination of seeds of North Temperate terrestrial orchids is a protective mechanism (STOUTAMIRE 1974). If it is, two obvious questions are why (a) seeds of epiphytic and tropical terrestrial orchids, which can also be subject to drying, do not have a similar mechanism and (b) only some terrestrial species germinated as part of this project seem to be inhibited by light. An answer to both seems to be that, if a protective mechanism exists, it is not universal even among terrestrial species of North Temperate origin.
The inconsistent effect of vitamins on the germination of G. oblongifolia, Platanthera saccata, and Piperia elegans (table 6) is in line with previous reports regarding other orchids (WITHNER 1959(WITHNER , 1974ARDITTI 1967ARDITTI , 1977ARDITTI , 1979STOUTAMIRE 1974). Also, like other orchid seeds (ARDITTI 1967(ARDITTI , 1979,     (table 4). Plant hormone effects on the germination of orchid seeds (especially terrestrial species) differ (WITHNER 1959(WITHNER , 1974ARDITTI 1967ARDITTI , 1979HADLEY 1970;HARVAIS 1973;STOUTAMIRE 1974;TAMANAHA, SHIMIzu, and ARDITTI 1979). Our results are similar: Auxins, cytokinins, or Ethrel seems to hinder the germination of E. gigantea and enhance that of G. oblongifolia in the dark (tables 4, 6). In E. gigantea the effects of Ethrel are marginal and similar to those of the ethanol used as solvent (table 4). If the slightly enhanced germination is an ethanol effect, it confirms previous reports (THURSTON, SPENCER, and ARDITTI 1979). However, our findings should be treated with caution since the germination rates were very low and because 2.5 mg/liter of Ethrel had no effect at all.
Albino seedlings of nonsaprophytic orchids (VAJRABHAYA 1977) usually do not survive for long. Therefore, the development of white seedlings of E. gigantea is not unusual, but their survival is. If they are genetic albinos, it is also not surprising that the addition of hormones and complex additives to the culture media failed to induce chlorophyll formation (table 5). The spontaneous appearance of green shoots suggests that at least some of these seedlings are not genetic albinos. In nature a certain proportion of Epipactis are chlorophyll free and either perish or are parasitic on their endophytes. Furthermore, it is possible that some or all of these seedlings turn green as a result of a stimulus provided by the fungus.
Disproportionally large roots and very small shoots, like those of P. saccata seedlings, have occasionally been observed in cultures of epiphytic genera. Therefore, it is possible that this growth pattern is not typical for the species.
Seeds of some terrestrial orchids are difficult to germinate asymbiotically (ARDITTI 1967(ARDITTI , 1979STOUTAMIRE 1974) and others are not (HADLEY 1970;HARVAIS 1973;FAST 1976;CLEMENTS and ELLYARD 1979). Therefore, our findings suggest that procedures for the germination of different species can be developed.