Somatic embryos as a synthetic seeding system using fluid drilling techniques for direct sowing of sweet potato (Ipomoea...


Material Information

Somatic embryos as a synthetic seeding system using fluid drilling techniques for direct sowing of sweet potato (Ipomoea batatas Poir.)
Physical Description:
vi, 194 leaves : ill., photos (some col.) ; 29 cm.
Schultheis, Jonathan Richard, 1958-
Publication Date:


Subjects / Keywords:
Sweet potatoes -- Propagation   ( lcsh )
Plant propagation   ( lcsh )
Dissertations, Academic -- Horticultural Science -- UF
Horticultural Science thesis Ph. D
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1989.
Includes bibliographical references (leaves 175-193).
Statement of Responsibility:
by Jonathan Richard Schultheis.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001563117
oclc - 22714814
notis - AHH6843
sobekcm - AA00006103_00001
System ID:

Full Text








I wish to thank Dr. Daniel J. Cantliffe for his full support during

my studies at the University of Florida. His help, encouragement, and

guidance during my graduate program were greatly appreciated. Beyond our

professional relationship, I value him as my good friend.

Special thanks is also extended to Dr. Indra K. Vasil for his

instruction and helpful suggestions during my program. I would also like

to thank Dr. Thomas E. Humphreys and Dr. Francis W. Zettler for their

encouragement and valued advice during this study. I gratefully

acknowledge Dr. Herbert H. Bryan for his help and support, particularly

in the Homestead field study.

Thanks are conveyed to faculty, staff, and students in the vegetable

crops department for their friendship and help throughout my graduate

studies. Specifically, I want to thank Marie Bieniek for her technical

assistance. Special thanks are conveyed to Dr. Raymond P. Chee for his

valuable input and suggestions during my research program.

My gratitude for the financial support during my program is extended

to the Gas Research Institute.

Finally, I want to thank my family for their constant love, support,

and prayers during my entire graduate program. Appreciation and a special

thanks is expressed to my wife, Nancy, and my daughter, Katharine Anne

"Katie," who endured many sacrifices yet provided me with their continuous

help, love, and encouragement during my graduate program.



ACKNOWLEDGEMENTS.................................................. ii

ABSTRACT .......................................................... v


1 INTRODUCTION................................................ 1

2 LITERATURE REVIEW............................................ 5

Botanical Classification..................................... 5
Seed Characteristics......................................... 5
Environmental Requirements for Optimum Sweet Potato Growth.. 7
Root Characterization and Development....................... 8
Cultural Management Practices and Their Effects on Yield..... 10
Tissue Culture............................................... 16
Somatic Embryogenesis ....................................... 17
Somaclonal Variation........................................ 39
Sweet Potato Tissue Culture.................................. 41
Delivery Methods for Somatic Embryos......................... 43
Fluid Drilling............................................... 49
Soil Covers.................................................. 54
Summary ..................................................... 55

AND GEL TYPE FOR FLUIDIZED SOWING........................... 57

Introduction.............................. .................. 57
Materials and Methods ....................................... 58
Results................................... .................. 62
Discussion................................ .................. 70
Summary ................. .................................... 74

FOR FLUIDIZED SOWING......................................... 77

Introduction.............................. .................. 77
Materials and Methods ....................................... 80
Results and Discussion....................................... 87
Summary ..................................................... 109


Materials and Methods ..............................
Results ...........................................
Summary.......................................... ..

.......... 112


A APPENDIX OF DATA TABLES ...................................... 157

B APPENDIX OF FIGURES.......................................... 164

LITERATURE CITED................................................... 175

BIOGRAPHICAL SKETCH................................................ 194

o . .
. .. .
. .. .
. . .
. . .

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




May 1989

Chairman: Daniel J. Cantliffe
Major Department: Horticultural Science

Lack of seed production is a major limiting factor for large-scale

production of many promising energy crops such as sweet potato. Somatic

embryos have great potential use as propagules in a synthetic seeding

system such as fluid drilling. The critical factors which affected the

rate, uniformity, and percentage embryo-to-plant formation in a

fluidized gel were examined.

Hydroxyethyl cellulose gel amended with Murashige and Skoog (MS)

basal medium was a compatible gel carrier for fluid drilling somatic

embryos and supported plant formation. Embryos at the elongated torpedo

developmental stage formed roots earlier and consistently produced a

greater percentage of plants than embryos at the cotyledonary and

torpedo stages. Embryo conversion to plants was doubled regardless of

embryo developmental stage when matured on agar-solidified basal medium

for 16 days rather than 25.

In gel amended with Hoagland, Gamborg, White, or MS media, the

most plant formation was obtained with Gamborg or MS medium. The

optimum concentrations of MS media for plant formation (40%) were one-

half and full strength. Micro and macro salts were essential for

optimum plant development, while vitamins were not.

No root or plant formation occurred in gel without carbohydrate or

with galactose. The best plant formation was obtained with fructose,

glucose or maltose. The optimum fructose and sucrose concentrations for

both root and plant development ranged from 23 to 93 mM. Plant dry

weight and nodes per plant increased with increasing carbohydrate


Vegetative growth and yield of storage roots of propagules derived

from somatic embryos, stock plants and/or zygotic seed were compared in

field experiments. More nodes per plant and greater vine length were

consistently obtained from stock plants and zygotic seeds versus somatic

embryos. Total yield and root size were greater from plants derived

from stock plants than somatic embryos. Under favorable environmental

conditions and extended growing seasons, plants from somatic embryos

produced normal large-sized roots.

The determination of a gel and those additives conducive for plant

growth is a key step towards commercial synthetic seed technology via

fluid drilling for vegetatively propagated species such as sweet potato.


Sweet potato (Ipomoea batatas (L.) Poir.) ranks seventh among

crops grown for food and tenth among all crops grown world-wide (FAO

Production Yearbook, 1986). Besides food production, sweet potato has

good potential as a biomass crop for ethanol and methane (Smith and

Frank, 1984.) The accumulated carbohydrate in the roots can be

efficiently digested and converted into ethanol or methane gas and used

for energy (Smith et al., 1987).

Methane yield conversion efficiencies from various groups of plant

species were compared; freshwater aquatics, forage and grasses, roots

and tubers, and saltwater species (Smith et al., 1987.) Roots and

tubers produced at least 36% greater yields than the other three groups

when equivalent plant quantities for each group were compared. Sweet

potato roots are not only efficiently converted to energy, but the crop

has added benefits. It can be grown with minimal energy inputs, has few

disease or insect pests (Bouwkamp, 1985), and can be produced in a short

130 day growing season (O'Hair et al., 1987). Two sweet potato crops

can be harvested annually in tropical growing regions, while in

temperate zones, cool season biomass crops such as turnip can follow a

sweet potato crop.

Lack of seed production is a major limiting factor for large-scale

production for many of the promising energy crop species such as sweet

potato. Thus, vegetative propagation is the plant establishment method


currently available for commercial plant production. Problems with this

planting method include the following: 1) length of time needed for

plants to survive stressful conditions, 2) maintenance and production of

disease-free plants, 3) spacial requirements for large-scale production,

and 4) labor costs to maintain and plant propagules (Cantliffe et al.,

1987). Direct seeding of asexually propagated biomass crops would make

mass plantings of these specie types much more cost-effective.

Unfortunately, in sweet potato, few or no seeds are ever produced since

they are self-incompatible, and those seeds which are produced result in

plants which genetically differ from the mother plant due to their

complex inheritance.

Micropropagation of high-value crops [i.e., ornamentals (Holdgate,

1977), citrus (Speigel-Roy and Vardi, 1984), pineapple (Rangan, 1984),

strawberry (Boxus et al., 1984), etc.] has resulted in successful

regeneration of plants. The use of a micropropagation system has the

advantage of producing plantlets which are fairly genetically uniform;

however, this system is limited to small-scale use since the

multiplication of propagules is relatively low and the plantlets must be

acclimated prior to planting. This leads to a relatively high cost per

plantlet which would be prohibitive for large-acreage crops.

In order for a tissue culture propagation method to be applicable

for large-acreage plantings, the following requirements would have to be

met. The propagation method must rapidly produce millions of propagules

and propagule cost must be competitive economically with conventional

propagation methods (Murashige, 1977).

Somatic embryogenesis is one such tissue culture method that could

be used to clone greater numbers of propagules in less space and shorter

times compared to normal seed or vegetative stock production. Some

problems must be overcome if somatic embryogenesis is to be implemented

as a synthetic seeding system. These problems include obtaining somatic

embryos which are singulated, which are uniformly mature, uniform in

size, and which grow rapidly into plants. Handling and planting of

these tissue-cultured propagules could then be patterned after

conventional seeding techniques.

Several delivery methods have been proposed for direct-field

synthetic seeding of somatic embryos. They are as follows: 1)

encapsulation of singulated embryos in an alginate gel capsule and

planting with a conventional drill (Redenbaugh et al., 1984, 1986; Jeon

et al., 1986), 2) embryo desiccation, then seeding with a conventional

drill (Gray et al., 1987), 3) simultaneous desiccation of embryos in a

water-soluble resin and planted as a wafer or on a seed tape (Kitto and

Janick, 1984ab), and 4) gel seeding embryos with fluid drilling

equipment (Drew, 1979, Baker, 1985).

Encapsulation research and development is in its infancy, cost per

propagule is high, and the emergence of the seedling is inhibited by the

capsule (Redenbaugh et al., 1987b). Desiccation has also been a problem

with this method since embryo injury and reduced plant formation has

occurred once the embryos are imbibed after being desiccated. Fluid

drilling, in comparison, is a developed planting technology that has

been used for planting zygotic seed for over 20 years (Gray, 1981.)

With fluid drilling, somatic embryos could be removed from tissue

culture without growth arrest or desiccation, and planted 'as is' or at

an advanced germinated stage. The gel carrier would not only protect

the somatic embryo but could act as an endosperm, since it can be


amended with nutrients, hormones, beneficial microbes, and other growth

factors. Various pesticides could also be incorporated into the gel for

added protection against disease or insects. Few studies have dealt

with fluid drilling of somatic embryos (Baker, 1985). Beyond this, few

field studies have compared yields of plants vegetatively propagated or

obtained from somatic embryos.

The objectives of this research were twofold. One part involved

determining the gel and additives that would facilitate the most rapid,

uniform, and complete embryo-to-plant formation when seeding with a

fluidized gel. Secondly, early plant growth and root yield comparisons

were made when plants were derived from original plant stock or derived

through somatic embryogenesis. Sweet potato was chosen for these

investigations because of its high potential for biomass production.


Botanical Classification

Sweet potato (Ipomoea batatas (L.) Poir.) is a member of the

Convolvulaceae (morning glory) family and is usually grown for its

edible roots. Sweet potato is often confused with yam since both

storage organs look and taste similar. Yams, however, belong to the

genus Dioscorea, (family Dioscoreaceae) are monocots, diploid (2n=20),

and storage organs are tubers whereas sweet potatoes are dicots,

hexaploid (2n-90), and the storage organs are botanically classified as

roots (Wilson and Collins, 1988). The literature often confuses the

issue by incorrectly calling the sweet potato's storage organ a tuber

(Kays, 1985).

Seed Characteristics

Sweet potato plants generally do not produce seeds since most

cultivars are self-incompatible (Hernandez and Miller, 1964; Wang, 1964;

Onwueme, 1978). Those plants which are produced from seeds are highly

heterozygous and differ genetically from the mother plant. Seeds mature

approximately 40-60 days after anthesis (Matsuo et al., 1984), and are

about 3 mm in length with one surface flattened and the other convex

(Hayward, 1938; Onwueme, 1978). They also have hard, nearly water-

impervious seed coats. Within the seed coat lies the embryo composed of

a hypocotyl connected to two large, tightly folded cotyledons. Seeds

are albuminous and the endosperm envelops the embryo. Scarification


either with sulfuric acid for about 45 minutes or by mechanically

clipping part of the seed coat is necessary for optimum germination,

which may reach approximately 90 percent (Jones and Dukes, 1982).

Germination usually occurs within one to two days, while plant emergence

takes an additional day or two under favorable growing conditions

(Hayward, 1938).

The analysis of metabolites in seeds can give clues as to the

importance of certain substances critical for the growth, development,

and germination of a seed. Several endogenous GAs were identified by

Matsuo et al. (1984) in both immature (15-30 days post-anthesis) and

mature (40-60 days post-anthesis) sweet potato seeds. In immature seeds

(15-30 days post-anthesis), GA19 and GA23 were the major GAs identified.

Also, abscisic acid (ABA) was found in immature seeds using gas-liquid

chromatography and mass spectrometry analysis. In mature seeds (40-60

days post-anthesis), GA3, 5, 8, 17, 19, 20, 23, and 44 were identified in

the seed tissue, with GA19 and GA23 the most concentrated at 200 and 160

pg/kg fresh weight, respectively. The authors determined that GA levels

increased in the seed up to 15 to 20 days after anthesis, then decreased

as the seed matured and dried. They suggested that the rapid cotyledon

growth 10 to 15 days after pollination might be attributed to the

increased GA levels found during this phase of seed development.

Gibberellin-like substances have been measured in mature seeds of

several families and species (Murakami, 1959). In particular, the

Convolvulaceae family had high levels of GA3, while specifically,

Ipomoea batatas had the highest concentrations (90 pg gibberellin

equivalents/100 g fresh weight in mature seed) of those species


evaluated in the family. Thus, GA appears to be of particular

importance in the development of sweet potato seed.

Environmental Requirements for Optimum Sweet Potato Growth


Sweet potatoes require a growing season that has at least four to

five frost-free months to produce ample-sized storage roots. Warm

conditions are required since temperatures at or below 15*C resulted in

little or no plant growth (Pierce, 1987). Above 15*C, Harter and

Whitney (1927) determined plant growth rates continued to increase up to

a temperature of 35*C, while Pierce (1987) reported optimum growth

occurred at air temperatures of 29*C.


The crop grows best in well-aerated, sandy-type soils, with a pH

ranging between 5 and 7.8 (O'Hair, 1984). However, good yields of 17.9

and 17.0 MT/ha have been obtained at soil pH values of 4.3 and 4.6

(Abruna et al., 1978). Soil aeration, when limiting, can inhibit

storage root formation and development. Wilson (1970) reported lack of

storage root production when sweet potatoes were grown under waterlogged



Sweet potatoes will grow and yield well under dry growing

conditions. Yields of 12.5 MT/ha, 15.0 MT/ha, and 19.7 MT/ha in seasons

which had only 10.1, 6.9, and 9.7 cm rainfall, respectively were

obtained in three separate investigations (Lamberth, 1957 cited in

Bouwkamp, 1985; Carpena et al., 1977; Villareal et al., 1979). In

several studies, supplemental irrigation improved yields and/or root

quality (root shape, soluble solids etc.) (Hernandez et al., 1965;


Jones, 1961; Constantin et al., 1974). Onwueme (1978) reported that

sweet potatoes grow best with an average rainfall of 50 cm over a

growing season. Moisture levels are particularly critical during plant

establishment (Bouwkamp, 1985; Godfrey-Sam-Aggrey, 1974; Anon., 1940).

Hernandez et. al. (1965) and Edmond and Amerman (1971) reported that

yields were most reduced when drought conditions occurred the first 40

days after transplanting.

Root Characterization and Development


Although sweet potato is grown extensively throughout the world,

the controlling mechanisms in storage root formation are poorly

understood. Most information dealing with this topic is based on

environmental and anatomical studies.

Root Origin. Type, and Size

Some extensive sweet potato root classification studies were

conducted by Wilson (1970,1982) and Kays (1985). Wilson has classified

roots into ten different groups based on morphology and origin. A

simpler classification scheme was proposed by Kays and will be briefly

described. Based on origin, Kays placed roots into two groups: 1)

adventitious roots, which emanate from underground stems, transplants,

or root pieces, and 2) lateral roots, which arise from preexisting

roots. Early in development young adventitious roots can be separated

into "thin" or "thick" roots, the former typically arising from the stem

internode, the latter from the stem node.

Anatomically, roots can be identified early in their development

by vascular arrangement. Pentarch and hexarch vascular arrangements are

destined to become storage roots while non-storage roots, which


typically are from "thin" roots, have a tetrarch arrangement. According

to Hahn and Hoyzo (1984), Togari reported that the number of storage

roots is determined the first 30 days after planting, which suggests

that environmental conditions during this period are critical for

obtaining maximum yields.

Young "thick" roots principally produce storage and pencil roots

(Kays, 1985). Storage roots are greater than 15 mm in diameter while

pencil roots, sometimes derived from "thin" roots, are 5 to 15 mm in

diameter. Primary fibrous roots typically arise from "thin" roots and

are less than 5 mm in diameter. The lateral roots of sweet potato arise

from preexisting roots and function as a nutrient and water-absorbing

system for the plant.

The vascular cambium and anomalous cambium are the two major

regions in the root which promote root enlargement through meristem

activity (Wilson and Lowe, 1973). Root enlargement from "thick" roots

occurs if environmental conditions are favorable; otherwise,

lignification of the growing region occurs, halting storage root

development with subsequent formation of non-storage root types.

Environmental Effects on Root Formation

Light. Several environmental factors influence storage root

formation and yields of sweet potato. Light can affect storage root

formation in two primary ways. Short day lengths promote storage root

formation (Kim, 1961) Roots exposed to light do not enlarge unless

covered with soil (Hahn and Hoyzo, 1984).

Oxeyen. Oxygen concentrations in the soil impact on the type of

roots which develop. Chua and Kays (1981) determined that a 2.5% oxygen

concentration in the root zone resulted in 12% dry weight in storage


roots, while a 21% oxygen concentration had 89% dry weight distributed

in storage roots. Dry, compacted soils also were reported to inhibit

storage root formation (Togari, 1950, cited in Kays, 1985). This can

likely be attributed to reduced oxygen concentrations in the soil as was

reported above by Chua and Kays (1981).

Fertilization. Fertilization, particularly with nitrogen (N) and

potassium (K), can influence storage root formation. High levels of N

increased vine growth at the expense of storage roots (Stino and Lashin,

1953). Large quantities of soil N reduced meristematic activity and

increased lignification which favored production of non-storage roots

(Togari, 1950, cited in Hahn and Hozyo, 1984). Potassium has

consistently increased yields and storage root formation (Bouwkamp,

1985) and will be discussed at length later.

Temperature. Sweet potato storage root yields are highly

temperature-dependent. The effect of temperature on root formation was

tested by varying the temperature in the root zone at 10*, 15*, 20,

25*, 30*, and 35C (Spence and Humphries, 1972). Storage root formation

was greatest at 25*C, while fibrous root formation was similar

regardless of temperature when tested above 15C. Thermoperiodism may

also be important in storage root formation. Daytime temperatures of

29C and nighttime temperatures of 21C increased storage root

development compared with a constant temperature of 29*C, which resulted

in greater vine than storage root growth (Kim, 1961).

Cultural Management Practices and Their Effects on Yield


Sweet potatoes will yield under conditions of low soil fertility;

however, root yields generally are improved by supplemental fertilizer


applicationss, particularly with K (Bouwkamp, 1985). Speights and

Paterson (1969) reported yields were increased 4.2 MT/ha with an

application of 70 kg/ha of K, while yields were further increased 2.2

MT/ha with an additional 70 kg/ha. Jones et al. (1979) determined that

yields increased linearly with K applications, and yield responses were

negatively correlated with extractable soil K. Yield increases were

also reported by Stino and Lashin (1953) with each 15 kg/ha addition of

K. Vine growth was determined to account for most of the K used during

the first three growing months, while K was accumulated and utilized

predominately by storage roots the last two months of the season (Scott,

1950). Potassium is of primary importance in sweet potato in starch

metabolism (Murata and Akazawa, 1968), translocation and photosynthesis

(Tsuno, 1970).

Several field experiments have been conducted to study the effect

of N on sweet potato yields. Results have been variable and less

consistent than with K. Addition of 100 versus 50 kg N/ha increased

sweet potato yields 17% (Speights and Paterson, 1969). Knavel (1971)

determined that root yields were increased after nine weeks with the

addition of N at 140 and 280 ppm. Conversely, Yong (1969) reported that

90 or 180 kg N/ha did not increase storage root yields compared with 45

kg N/ha, but a significant increase in vegetative weight was obtained.

Spence and Ahmad (1967) found that storage root formation occurred in N-

deficient plants in spite of severely restricted growth. Bouwkamp

(1985) suggested that variable responses to N may depend on several

factors such as N-supplying capacity of the soil or type of cultivar



Bouwkamp (1985) summarized nutrient studies of sweet potatoes

including manganese, phosphorus, magnesium, sulfur, and calcium and

which led to improved yields; however, as mentioned earlier, yield

improvement seemed most responsive to K applications. Storage root

formation was absent in magnesium, calcium, and K deficient plants

(Spence and Ahmad, 1967). Deficiency symptoms were also noted on

leaves, roots, and stems when dried tissue samples were below 0.08%

sulfur, 0.12% phosphorus, 0.75% K, 0.16% magnesium, and 0.2% calcium.

Length of Growing Season

Storage root development and yields were increased with longer

growing seasons. Thirty sweet potato lines and/or varieties were

harvested 108, 130 149 and 169 days after planting (Cordner and

Galeotti, 1961). An average yield of 151 bu/acre harvested 108 days

after planting represented 46% of the full season yield of 590 bu/acre

gathered 169 days after planting. The yields of some varieties and

cultivars were very low when harvested earlier. Yields of all varieties

and cultivars were improved with later harvests. Yong (1969) also

reported increased total yields with a longer growing season, and

corresponding decline in vine yields. Similarly, decreases of 50

bu/acre were recorded for every two weeks that planting was delayed in

South Carolina with the variety 'Porto Rico' (Beattie et al., 1938).

Spacing and Competition

Bouwkamp (1985) summarized the effects plant spacing had on yields

by evaluating several studies where spacings ranged from 15-110 cm

within the row for various cultivars. He concluded total yields were

improved at high plant densities (40,000 to 50,000 plants/ha) regardless

of cultivar.


Root size could be controlled by plant density. Roots were

smaller the closer the plant spacing (Bouwkamp and Scott, 1980). Also,

the percentage of smaller roots was cultivar-dependent. Walker and

Randle (1987) determined that cultivars with different growth habits

planted in adjacent rows competed with one another. They proposed

harvesting the middle row of three-row plots to reduce the competitive

effects of adjacent row cultivars. However, Bouwkamp (1986) asserted

the competitive effect of cultivars did not alter yields enough to

implement usage of multi-row plots. Recommendations for plant spacing

in Florida are 25-30 cm between plants or approximately 22,000 30,000

plants/ha. (Stall et al., 1984).

Plastic Mulch

Plastic mulch was used to improve yields of northern (Hochmuth and

Howell, 1983) and southern (Bryan, 1966) grown sweet potatoes. Black

plastic mulch increased leaf area, number of leaves, and total

vegetative weight compared with unmulched plots (Hochmuth and Howell,

1983). Marketable root yields were increased 7.1 MT/ha on mulched flat

beds and 8.4 MT/ha on mulched raised beds. Marketable sweet potato

yields were increased 53, 211, and 168% on black, clear, and white

plastic mulch, respectively, compared with unmulched plots (Bryan,



Cost of transplants is a major expense in the commercial

production of sweet potato (Kushman et al., 1970). As mentioned

earlier, commercial plantings are all vegetatively propagated. In

growing regions where sweet potatoes cannot be grown continuously, roots

must be saved, planted in beds, and sprouts or 'slips' which develop

from the roots used for propagation. In warmer climates where a crop

can be maintained perennially, or several weeks after slips have been

transplanted, vine cuttings may be used for crop establishment.

Effect of Establishment Practices on Storage Root Yields.

Several investigations have been conducted to determine the

effects of various propagule types and establishment practices on

storage root yields. Root plants (slips) and vine cuttings were

compared for their effect on storage root yields in four sweet potato

breeding lines or varieties (Cordner and Galeotti, 1961). Yields from

cuttings were similar to that obtained from slips. In another study

which compared vine cuttings to slips, yields were also similar when

total yields and the USDA size grades were evaluated (El-Kattan and

Stark, 1949).

Propagule length and its effect on yields has also been

investigated. Three studies were conducted in succeeding years in which

slips were removed from the mother root and separated into two size

categories, 23 to 25 cm (large) and 15 to 18 cm (small)(Beattie et al.,

1938). In one study, five weeks after transplanting, the large-plant

plots had greater vine growth than small-plant plots. However, the

better vegetative growth obtained from the large-plant plots did not

improve yields versus the small-plant plots. Mean yields were similar

throughout the three one-year studies and were not dependent on slip

length. A similar evaluation was made between vine cutting lengths of

23, 31, 46, and 61 cm in four studies for two cultivars (Godfrey-Sam-

Aggrey, 1974). Three months after planting, the 46 and 61 cm cuttings

of both cultivars,'Missus' and 'Madam,' increased root yields compared

to the smaller 23 and 31 cm cuttings. Vine production per hectare,


however, was similar regardless of propagule length. These results were

the opposite of that found in the previously mentioned study.

Hall (1986) examined vine cutting transplant lengths, number of

nodes planted underground, and orientation of transplants on sweet

potato yields. Total marketable root yields of 'Red Jewel' were 2.7

MT/ha greater with the 40-45 cm than the 20-25 cm vine cuttings. Yields

were equivalent when similar comparisons were made with 'Georgia Jet.'

Total marketable yield of either cultivar was unaffected by the number

of nodes planted underground or transplant orientation. Generally,

longer propagules improved yields, while variations in the number of

underground nodes and transplant orientation appeared to be more

cultivar dependent. Although longer propagules have been reported to

increase yields, they require more labor and expense.

Six methods of propagating sweet potatoes were investigated

(Templeton-Somers and Collins, 1986). Three used in vitro techniques:

1) shoots obtained from leaf explants, then rooted; 2) buds obtained

from roots of leaf explants, then rooted; and 3) lateral buds excised

and tissue cultured. The other three utilized in vivo techniques: 1)

nodes excised in 10 mm pieces and rooted in transplant trays; and 15 to

20 cm long 2) slips; and 3) vine cuttings. In the above study, the

plants propagated in vivo by either slips or vine cuttings had greater

total and grade one root yields per plant than those obtained via in

vitro techniques. A carryover study determined if differences due to

propagation method in the first planting continued the subsequent year.

To test this, roots were saved from each propagation treatment and

bedded, then slips were harvested and transplanted the following growing

season. No significant differences in yields were measured from slips


obtained from different propagules the previous year. The low yields

obtained from the in vitro propagules the previous year could not be

attributed to alteration in genetic composition.

To reduce propagation costs, the use of root pieces, as practiced

with potato, (Solanum tuberosum L.), was investigated as an alternative

for plant establishment (Bouwkamp, 1982). Twenty to 25 gm root pieces

in over 220 lines were screened throughout a nine-year period. The type

of root development was cultivar dependent. Some cultivars had several

good quality roots arise from the root piece, while in some others, the

root pieces enlarged with no additional root formation. Root pieces

were disease-susceptible in many cases, making plant establishment a

major problem. Use of root pieces led to inconsistent root yields, is

still in an experimental stage, and needs more work before its


Commercial plantings of sweet potatoes is primarily for food

production, but attention in recent years has focused on this crop as a

biomass crop for energy production. Southern Japan, in some instances,

has used sweet potato for energy production (Yen, 1974). Mass

production, however, is limited by the high cost of vegetative

propagation (Cantliffe et al., 1987). The use of tissue culture has

been suggested as a means to economically produce crops normally

propagated vegetatively.

Tissue Culture

The birth of plant tissue culture can be traced back to 1902 when

Haberlandt first attempted to culture isolated plant cells and

demonstrate totipotency. His efforts failed, but some of the theories

he put forth were later proven, among them, totipotency (Vasil and


Hildebrandt, 1965). Today, tissue culture continues to advance science

and new technologies.

Somatic Embryogenesis

Discovery. Definition, and Potential Advantages

A significant finding in tissue culture occurred when two

independent research groups discovered in 1958 that somatic embryos

would abundantly form from carrot callus tissue, then grow directly into

plants (Reinert, 1958; Steward et al., 1958). In the years that

followed, it was demonstrated that numerous plant species and families

were capable of somatic embryogenesis (Tisserat et al., 1979; Ammirato,


Somatic embryogenesis can be defined as the developmental process

which produces embryos from somatic cells. Sharp et al. (1980)

distinguished between two developmental patterns: 1) direct

embryogenesis, in which embryos originate directly from tissues without

callus proliferation; and 2) indirect embryogenesis, in which callus is

maintained and proliferated before embryo development. In either case,

the formation of somatic embryos closely follows the developmental

sequence in zygotic embryos which are globular, heart, and torpedo

stages, then plant formation.

Somatic embryogenesis differs from micropropagation through shoot

primordia and adventitious bud differentiation. With the latter two,

the shoot or bud is unipolar, while somatic embryos are bipolar

structures with closed vascular systems. Several advantages are gained

using somatic embryogenesis rather than micropropagation techniques: the

capacity to produce large numbers of embryos in limited space; no need


for physical separation of individual propagules; and direct embryo

conversion into plants (Lawrence, 1981).

Explant Selection for Somatic Embryogenesis

The first criterion to consider for induction of somatic

embryogenesis is explant source. Nearly all plant parts have been used

successfully to induce somatic embryogenesis: hypocotyls (Halperin,

1966), stems (Banks, 1979), embryos (Bayliss and Dunn, 1979), ovules

(Kochba et al., 1974), shoot tips (Lakshmi Sita et al., 1979), young

inflorescences (Chu et al., 1984), roots (Okamura et al., 1973), and

protoplasts (Vasil and Vasil, 1980). Somatic embryogenesis can be

induced from a variety of explants in certain species like carrot, i.e.,

hypocotyl (Halperin, 1966), petioles (Wetherall and Dougall, 1976),

taproots (Okamura et al., 1973) and young roots (Smith and Street,

1974); however, some species, particularly those in the Gramineae, may

respond in more limited regions in the plant, specifically the actively

growing regions (Ammirato, 1983.) For example, somatic embryo formation

was obtained from immature embryos of Triticum (Ozias-Akins and Vasil,

1982), immature inflorescences of Setaria itlica (Zhi-hong et al.,

1984), and apical meristems and leaf primordia of corn (Zea mays) (Fahey

et al., 1986). Environmental conditions (Ammirato, 1983) and

developmental which explant sources are obtained also influence

the capacity for embryogenesis (Vasil and Vasil, 1984).

Induction of Embryogenic Callus

Once an adequate explant source is determined, it must be

sterilized and placed on solid or in liquid medium to induce the

formation of embryogenic callus. Typically this medium, often termed

primary or induction medium, consists of inorganic salts, carbohydrates,


organic supplements such as vitamins, and hormones necessary for growth.

Specific requirements for optimum growth must be determined empirically

for a species, while requirements may even differ between cultivars of

of a species. Chen et al.(1987) tested 50 genotypes of three alfalfa

(Medicago spp.) species and determined that one particular combination

of medium and growth regulators resulted in formation of somatic embryos

in one genotype, while another combination was required for a different


Auxin is the critical growth regulator needed to induce

embryogenic callus formation (Tisserat et al., 1979). Auxins such as

2,4-dichlorophenoxyacetic acid (2,4-D), indoleacetic acid (IAA), and

napthlalene acetic acid (NAA) have been used successfully. The auxin

2,4-D was most commonly used among species (approximately 60% of the

time) for embryogenic callus induction (Ammirato, 1983; Evans et al.,

1981). Cytokinins may be useful for embryogenic callus production (Chee

and Cantliffe, 1988b), but these hormones may not be needed in some

families such as Umbelliferae (Ammirato, 1983). Schenck and Hildebrandt

(1972), however, reported low levels of cytokinins were essential for

most dicot cell cultures.

Embryo Induction and Maturation

Introduction. Conditions in the primary medium that favor callus

growth are not necessarily the same as those associated with embryo

formation and maturation. Embryo development is typically induced by

transfer of embryogenic callus to another medium containing with no

hormones or reduced hormone concentrations. Although embryo formation

may be easily induced, production of large quantities of embryos which

rapidly convert to plants may not. A number of studies are described


which have considered the effects of nutrient, carbohydrate, and growth

regulator formulations on the induction, development, and number of

somatic embryos produced.

The nutrient salt medium of Murashige and Skoog (MS) is most often

used for somatic embryogenesis. Evans et al. (1981) reported 70% usage

of this medium in somatic embryogenesis protocols. Other media, such as

Gamborg's (Gamborg et al., 1968), Schenk and Hildebrandt's (SH) (1972)

and White's (1943) have been successfully used or were superior for some

species (Evans et al., 1981). Sucrose is the carbon source most

frequently incorporated into tissue culture media (Ozias-Akins and

Vasil, 1984). Other sugar types such as galactose, glucose, maltose,

and fructose have been used with success (Verma and Dougall, 1977).

Lower levels of auxin, and various concentrations of cytokinin, abscisic

acid (ABA), and gibberellic acid (GA) have been reported in different

instances to improve either somatic embryo formation or maturation and

will be discussed later.

Inorganic media composition and comparisons. White (1943)

developed a medium for the culture of excised tomato roots. Murashige

and Skoog (1962) used White's medium as a reference to develop a medium

designed to increase production of tobacco callus. These researchers

determined that concentrations of N, K, phosphorus, magnesium, and

calcium were particularly important if optimal growth was to be

achieved. Gamborg's (1968) medium has been used for protoplast culture

and was specifically created for culture of soybean root cells. Schenck

and Hildebrandt (1972) developed a medium which promoted the growth of

callus in a variety of monocotyledonous and dicotyledonous plants.

Other nutrient media, usually less concentrated, such as Hoagland


solution (Hoagland and Arnon, 1950) have been used specifically for

hydroponic culture of plants.

Murashige and Skoog medium is used in many tissue culture

laboratories when the objective is plant formation; however, White's,

SH, Gamborg and other media will work as well. Compared with many

media, MS medium contains very high concentrations of inorganic salts.

Murashige and Skoog medium contains nearly two times more N and 20 times

more phosphorus than either Gamborg or SH media, with concentrations

being even greater than Whites medium (Table A-l). Another key

distinction between media is MS contains 20 mM ammonium compared to 1,

2.6 and 0 mM for Gamborg, SH, and White's media, respectively. The N

source, particularly ammonium or a reduced form of N has been reported

especially critical in somatic embryogenesis (Halperin and Wetherell,

1965; Stuart and Strickland, 1984ab).

Effect of nitrogen source. Halperin and Wetherall (1965), in one

of the earliest studies, reported ammonium was critical for

embryogenesis in wild carrot. The reduced N form, glutamine, could also

be substituted for ammonium and enhance embryogenesis (Halperin and

Wetherell, 1965; Halperin, 1966). Callus liquid cultures grown on

medium lacking reduced N formed adventitious roots while, in contrast,

cultures in medium with reduced N yielded large numbers of embryos.

Similarly, Wetherall and Dougall (1976) reported that reduced N

was critical for somatic embryogenesis in wild carrot. They determined

ammonium was the best reduced N source when cultures were grown in the

presence of 20 mM of nitrate, but also found that glutamate was the best

for embryogenesis when used as the sole nitrogen source at either the 15

or 30 mM levels. Kamada and Harada (1979) determined 10 mM alanine


stimulated embryo production in wild carrot nearly 40-fold more than the

optimum ammonium concentrations in the presence of 20 mM nitrate. To a

lesser extent, the amino acids glutamine, glutamate, aspartate,

arginine, proline and asparagine increased embryo production compared

with ammonium. Although greater numbers of embryos were produced with

the above amino acids, more mature embryos (heart and torpedo-shaped)

were produced only with alanine, glutamine, glutamate, and aspartate.

Besides the N form, Reinert (1967) reported that total N was a critical

factor for optimization of embryogenesis in wild carrot.

During the 1980s, the effect of N source on embryo development and

production was scrutinized more carefully on species other than wild

carrot. Walker and Sato (1981) determined that Medicago sativa L. cv

Regens (alfalfa) was similar to carrot in response to the ammonium ion.

Root formation was inhibited by ammonium levels 50 mM and above, while

somatic embryos formed at the minimal ammonium concentration tested of

12.5 mM up to a high concentration of 100 mM. Stuart and Strickland

(1984a) evaluated the effect of various amino acids on embryogenesis

when added to SH medium. Addition of 250 to 300 mM proline to SH medium

increased the number of somatic embryos nearly three times compared to

basal medium without proline. Glutamine, lysine, serine, asparagine,

alanine, and arginine, to a lesser extent, also increased embryo


Stuart and Strickland (1984a) correlated improved embryo quality

with increased embryo size, which led to higher percentages of embryos

converting to plants. Using these criteria, inclusion of glutamine,

arginine, alanine, and proline produced greater-sized embryos and led to

improved plant formation compared with embryos which developed on the


medium with the addition of only ammonium. Somatic embryogenesis was

dependent on the inclusion of ammonium in the medium, while the

improved somatic embryo number and quality obtained with proline was

dependent on ammonium (Stuart and Strickland, 1984b). The optimum

concentrations for embryogenesis was 100 mM proline and 25 mM ammonium.

The authors suggested that proline would improve somatic embryogenesis

in a number of species.

Optimum ammonium levels were 5 mM for embryo induction, while 20

mM was best for embryo development, increased somatic embryo yields, and

plant conversion of Medicago sativa cv falcata (Meijer and Brown, 1987).

Somatic embryos produced with 30 mM ammonium and nitrate often led to

formation of small, abnormal embryos. Contrary to reports by Stuart and

Strickland (1984ab), incorporation of amino acids such as proline did

not improve embryo production, but was often inhibitory.

The nitrate:ammonium ratio of 2:1 was determined best for embryo

production in Solanum melongena (eggplant) (Gleddie et al., 1983). When

the 2:1 ratio was maintained, significantly more embryos were produced

between 60 to 90 mM total nitrogen. No embryo formation occurred when

either nitrate or ammonium were absent from the medium. The authors

suggested the optimum 2:1 ratio for embryo production might be

attributed in part to pH.

Effect of potassium. Potassium stimulated embryo formation in

wild carrot (Reinert et al., 1967). Percentage of cultures with embryos

was increased from 3 to 45% when K concentrations were increased from 3

to 20 mM. Brown et al. (1976) determined 1 mM K was required for

optimum growth of callus, while 20 mM caused embryo induction. The


effects of other nutrients on embryogenesis have received little or no


Effect of carbohydrates. Sucrose was reported to be the most

commonly used and effective carbon source for somatic embryogenesis

(Ammirato, 1983). Maretzki et al. (1974) concluded either sucrose and

glucose were best as carbon sources for growth of tissue cultures;

however, they suggested that other carbohydrates could substitute for

glucose or sucrose depending on the species and growth phase. The

following are carbohydrate studies involved with somatic embryogenesis

and include some examples of carbohydrate substitution for the

traditionally used sucrose.

Inclusion of galactose in the medium stimulated somatic embryo

development in citrus culture (Spiegal-Roy and Vardi, 1984). Glucose

levels at 6 to 10% improved somatic embryo development in carrots

(Homes, 1967 cited in Tisseret et al., 1979). Embryogenic callus was

initiated in cotton (Gossypium hirsutum) on medium containing 3%

glucose; however, sucrose replaced glucose for callus maintenance,

embryo maturation, and plant regeneration (Shoemaker et al., 1986).

Ozias-Akins and Vasil (1982) reported growth responses of wheat

scutellar callus were similar when grown on 2% glucose or sucrose, while

7% galactose inhibited growth. More somatic embryos formed in soybean

when carbohydrate concentrations were reduced from 12 to 1.5% when using

either glucose and sucrose (Lazzeri et al., 1987). In eggplant, at 60

mM a sucrose source led to the greatest embryo formation, followed by

glucose, then fructose (Gleddie et al., 1983). No embryos formed when

galactose, xylose, or raffinose were added to the medium.


Various carbohydrate sources for somatic embryogenesis of alfalfa

were compared at the 1 and 3% levels (Strickland et al., 1987).

Maltose, malt extract, and fructose increased embryo yields compared

with 3% sucrose. Maltose was evaluated in the remaining experiments

since embryo yields and germination were improved most with this

treatment. Maltose improved embryo development (embryo length,

diameter, and percentage embryos with cotyledonary lateral appendages)

compared with sucrose. Embryo formation was influenced by ammonium

carbohydrate interactions. The greatest formation was with the

combination of 15 mM ammonium and 3% maltose.

Stuart et al. (1988) determined that protein accumulation in

alfalfa somatic embryos was similar to that found in seed (Stuart and

Nelson, 1988). The 11S protein was the best protein marker found in

both somatic and zygotic embryos. A comparison between 11S protein

measured in mature seeds versus somatic embryos might be a good

biochemical marker for determining the physiological stage of a somatic

embryo. The amount of 11S protein expressed in the somatic embryo was

approximately 10% of that found in seed, indicating that somatic embryo

development was much less than that found in a seed. The 11S protein

levels of embryos produced with maltose or sucrose were compared.

Embryos obtained with maltose rather than sucrose had greater 11S

protein levels. Maltose, when used in media for regeneration, yielded

larger embryos which germinated better than sucrose-treated cultures.

Various sucrose concentrations (0 to 400 mM) were tested to

determine their effect on embryo initiation and differentiation on a

diploid and tetraploid alfalfa genotype (Meijer and Brown, 1987).

Embryo induction was high (number of embryos, developmental stage not


delineated) over a sucrose concentration ranging from 25 to 100 mM in

both genotypes. Embryo differentiation was best at either 100 or 200 mM

sucrose. The 100 mM range corresponded closely with the 88 mM sucrose

(3%) used in most standard tissue culture media.

Several carbohydrates were tested at 60 mM for their effects on

embryo growth and development, and germination in wild carrot (Daucus

carota) suspension cultures (Verma and Dougall, 1977). Dry weights of

embryos 14 days after incubation were much greater with either

filter-sterilized or autoclaved sucrose compared with glucose, fructose,

mannose, glucose + fructose, or galactose. In addition, a greater

number of more fully developed embryos, and more germinated embryos were

obtained with sucrose compared with the monosaccharides. Fructose,

after autoclaving, was the only sugar which inhibited growth and

embryogenesis. In a study over time, sucrose produced more embryos than

the other sugars after 13 days; however, after 27 days, embryo

production was similar with sucrose, fructose, glucose, or maltose. The

authors suggested that an initial delay in growth and embryogenesis

which occurred with galactose, raffinose, and stachyose might be

explained by slow uptake rates and/or metabolism of these sugars. They

proposed the rate of embryo growth and development might be controlled

by the rate of uptake and/or metabolism of the carbohydrate to a common


Several in vitro studies with zygotic embryos illustrate the

importance of carbohydrate type and concentration on embryo growth and

development. For example, Cobb and Hannah (1986) determined that

embryos of the wild type and shrunken-2 corn (Zea mays) species matured

and germinated when placed in medium containing glucose, fructose, or


sucrose. However, greater germination, starch content, and embryo

weight was obtained with sucrose rather than fructose or glucose. The

dynamic changes necessary for a maturing embryo is illustrated with

jimson weed (Datura stramonium) (Rietsema et al., 1953). Greater

sucrose concentrations (8-12%) were needed for less mature embryos

(pre-heart stage), while little or no sucrose (0-0.2%) was needed for

torpedo-stage or mature embryos.

Effect of hormones. After embryogenic callus has been formed or

proembyos initiated, the hormone levels) and/or type(s) often must be

altered to permit the maturation of embryos. Elimination of all

hormones from the medium promotes embryo maturation in species such as

carrot (Smith and Street, 1974; McWilliam et al., 1974), corn (Lowe et

al., 1985), and chicory (Heirwegh et al., 1985). Also, a reduction in

hormone concentrations can promote embryo development. For example,

Echinochloa muricata somatic embryos matured when callus was transferred

from medium containing 10 to 15 mg/L to one with less than 5 mg/L 2,4-D

(Cobb et al., 1985). In celery, torpedo embryos formed when both 2,4-D

and kinetin levels were reduced (Al-Abta and Collins, 1978).

Other studies revealed cytokinins were specifically required for

embryo maturation and/or increased production. Addition of kinetin

(KIN) (100 pM) or benzylaminopurine (BAP)(100 or 10 pM) to basal medium

improved embryo maturation of winter oil rapeseed (Loh et al., 1983).

MacKinnon et al. (1987) determined that a change in auxin source, from

2,4-D to IAA, and the addition of BAP improved embryo formation,

maturation and plant formation in wheat and sorghum. Similarly, Walker

and Sato (1981) replaced 2,4-D with NAA and added KIN to promote alfalfa

embryo maturation. Fujimura and Komaine (1975) determined somatic


embryogenesis of citrus was improved when 2,4-D was removed and 0.1 /M

zeatin was placed in the basal medium. Use of a specific cytokinin was

necessary since KIN and BAP were also tested, but unlike zeatin, these

cytokinins inhibited embryo development.

In Norway spruce (Picea babies a series of transfers to media

containing different hormone concentrations and types were required for

somatic embryo initiation and development (Gupta and Durzan, 1986).

Callus was removed from basal medium containing 2 AM of both BAP and KIN

and 5 pM 2,4-D to a medium containing less 2,4-D (1 /M) in which

globular embryos formed. A subsequent transfer to a medium containing

no auxin was necessary for embryo maturation.

Generally, gibberellic acid is not included in the maturation

medium for somatic embryo development. However, Ammirato (1977)

determined a balance of GA, ABA, and zeatin was necessary for normal

development and maturation of caraway (Carum carvi) somatic embryos.

Addition of 1 mg/L GA stimulated differentiation of sandalwood (Santalum

album L.) somatic embryos (Lakshmi Sita et al., 1979).

Gibberellic acid plays an important role in zygotic seed

development and its active metabolism during embryo development has been

related to embryo growth. In Japanese morning glory (Pharbitis nil)

zygotic embryos, the maximum growth of the seed occurred just after GA

levels peaked (Murakami, 1961). Gibberellic acid concentration was more

concentrated in the endosperm than any other plant organ in Sechium

edule and its concentration remained nearly constant throughout seed

development (Lorenzi and Ceccarelli, 1983).

Abscisic acid has been reported to impose dormancy and inhibit

germination of zygotic embryos (Black, 1981). In addition, it has been


shown in several instances to have a role in preventing precocious

germination of zygotic seed (Crouch and Sussex, 1981; Hendrix and Radin,

1984; Choinski et al., 1981 etc.). The effects of ABA on somatic embryo

development have also been investigated.

Ammirato (1974) determined that optimum levels of ABA (0.1 yM)

resulted in normal somatic embryo development in caraway. ABA reduced

formation of abnormal structures, such as adventitious embryos, and

repressed precocious germination (Ammirato, 1974, 1977). Similar

findings were reported by Kamada and Harada (1981) and Baker (1985) with

carrot. Ranch et al. (1985) determined the addition of 1 or 10 pM ABA

to Gamborg's medium decreased precocious germination and increased the

frequency of embryo maturation in soybean, while Vasil and Vasil (1981)

found similar responses with pearl millet (Penisetum americanum). Low

concentrations (0.1 and 0.05 mg/L) ABA promoted maturation of barley

somatic embryos (Kott and Kasha, 1983). ABA increased embryo production

of orange (Citrus sinensis) at 2.0 MM ABA (Kochba et al., 1978).

Finklestein and Crouch (1986) compared the affects of ABA and

osmotic potential on somatic and zygotic embryos of Brassica napus as

they matured. Cruciferin was produced as long as the zygotic embryo

continued maturation. Thus, it was designated a marker protein and

could be used to determine embryo maturation in somatic embryos.

Precocious germination, common in somatic embryogenesis, was inhibited

with the addition of either ABA or high osmoticum. The rate of

cruciferin synthesis was slowed in a precociously germinating embryo;

however, under conditions which suppressed precocious germination (ABA

and high osmoticum), the accumulation of the marker protein was similar

to that obtained by embryos grown in situ. Comparison of the kinetics


of ABA and osmotic effects on gene expression revealed osmotic effects

were more rapid than ABA.

Organic additives. Other undefined organic additives have been

beneficial for somatic embryo maturation. Litz et al. (1982) reported

coconut water was an important additive for improved embryogenesis of

mango. Casein hydrolysate has also enhanced somatic embryogenesis

(Ammirato and Steward, 1971). Activated charcoal proved useful for

better development of carrot somatic embryos, and led to embryo

development in cultures which failed to produce embryos when auxin was

eliminated (Fridborg et al., 1978). Embryo maturation was also improved

in onion in the same study. The charcoal was suggested to absorb

residual hormones in the medium or inhibitory substances which prevented

embryo maturation. Similar results were reported for ivy as 0.1%

charcoal maintained embryo integrity and development, and prevented

callusing (Banks, 1979).

Light. Other factors influencing embryo maturation are light,

osmotic potential, physical environment, pH, and temperature. High

light intensities have been reported necessary for somatic embryogenesis

in tobacco (Haccius and Lakshmanan, 1965 cited in Ammirato, 1983).

Induction of embryogenesis in eggplant was also dependent on light

(Gleddie et al., 1983). Gupta and Durzan (1986) reported maturation of

globular embryos to cotyledonary embryos with light at specific

wavelengths (2.8, 2.0, and 0.5 uW cm-2 nm-1 in blue, red, and far red).

In contrast, Ammirato (1974) reported caraway somatic embryo maturation

was more normal when grown without light.

Osmoticum. Carrot somatic embryo development has been controlled

by subjecting developing embryos to high carbohydrate levels thereby

increasing osmotic potential. Reduced precocious germination and

smaller embryos with cotyledons were obtained in cultures containing 12%

sucrose (Ammirato and Steward, 1971). Gleddie et al. (1983)

supplemented MS medium with 0 to 0.6 M mannitol and determined that as

osmotic potential increased, embryo frequency decreased, with no

embryogenesis above 0.4 mannitol. Crouch and Sussex (1981) determined

high osmoticum prevented premature germination of somatic embryos and

permitted continued protein accumulation like an embryo maturing in


pH. Wetherell and Dougall (1976) reported carrot embryo formation

best at a pH of 5.4, and was two times greater than if the pH was lower

than 5 or greater than 6. In a later study, Dougall and Verma (1978)

examined the effect of N sources on pH and proposed media containing

ammonium as a sole source of N resulted in variation in pH that would be

deleterious to cell growth and embryo formation. A broader pH range was

reported to facilitate the growth of Ipomoea suspension cultures (Martin

and Rose, 1975).

Temperature. Temperature can also be critical for obtaining

better embryo quality and an increased number of somatic embryos.

Gleddie et al. (1983) determined more embryos formed earlier at 30C

versus 25C. Redenbaugh et al. (1986) reported somatic embryo production

in celery was extremely temperature dependent. At 240C, 10 high quality

embryos formed per Petri dish. At 270C, 50 embryos formed and if

temperatures fluctuated between the two extremes, no embryos formed.

Summary of Embryo Initiation. Development, and Maturation

Much of the literature reviewed has dealt with the various factors

affecting somatic embryo formation and maturation. The gain of


knowledge in this developmental phase can have a significant impact on

the subsequent success of plantlet formation. As evidence, in alfalfa,

treatments which improved embryo maturation and quality resulted in

increased embryo to plant conversion from 0.5% in 1982 to 60% in 1986

(Redenbaugh et al., 1987a). The environmental conditions during embryo

development and maturation affect the overall quality and vigor of the

seed (Delouche, 1980) and most probably affect that of somatic embryos.

Development of Embryos to Plants

Overall survey. Once mature embryos are obtained, the next

developmental phase is their conversion to plants. Most studies in

somatic embryogenesis have not focused on protocols that result in plant

formation. Studies have been more limited when planting and rooting of

somatic embryos are considered; however, scattered reports, particularly

in the last five years, have documented the percentages of embryos

forming plants under varying conditions (i.e., Redenbaugh et al., 1986,

1987,; Stamp and Henshaw, 1982; Lazzeri et al., 1987). Some of the most

extensive work on embryo conversion to plants has been by Plant Genetics

Inc. in Davis, California, where the primary objective was to

economically produce somatic embryos as propagules for commercial


Hormones. In some cases, various hormones have been included in

the regeneration medium to enhance plant formation. Typically, auxin

levels are lowered, and often some type of cytokinin is included in the

regeneration medium. MacKinnon et al. (1987) reduced and changed the

auxin used in the medium from 5 and 2 mg/L 2,4-D to 0.1 and 1.0 mg/L IAA

for wheat and sorghum, respectively. In addition, 0.5 mg/L BAP was

included in the plant regeneration media of both species. The best


regenerating lines of wheat and sorghum produced 193 and 87 plants/gm

callus, respectively. The addition of an auxin (0.5 mg/L NAA) and

cytokinin (2 mg/L BAP) was also used for plant regeneration of millet

(Setaria italica) (Zhi-hong et al., 1984). With zygotic seed, usually

auxins inhibit or are ineffective as germination promoters, but can help

in some instances (Nickell, 1982).

In some cases, plant regeneration has been improved by the

addition of cytokinin to the medium. Regeneration of Vigna aconitifolia

was best in basal medium containing 2.2 jM BAP an average of 12.4 plants

from 50 mg callus (Kumar et al., 1988). Only 2.8 plants formed when BAP

was not added to the medium. The percentage of calli with embryos

forming plants was increased from 7 to 43% in rice (Oryza sativa L.)

when 1 mg/L KIN was added to MS medium (Abe and Futsuhara, 1985).

However, the addition of 0.02 mg/L 2,4-D reduced plant formation, even

when KIN was included in the medium. Root and shoot formation from

filbert (Corylus avellana) somatic embryos was similar regardless of the

addition or deletion of BAP (Perez et al., 1983). Cytokinins generally

have been found less effective and more limited in their role in

affecting seed germination othe than overcoming dormancy (Nickell,

1982). Kahn et al. (1971), however, reported that cytokinin removed the

block to germination imposed by inhibitors of germination in several

species. They suggested cytokinin permitted the completion of

gibberellin-mediated germinative processes and thereby acted in the

permissive role for germination.

Root and shoot formation can be controlled by changing the auxin

cytokinin ratio. Various combinations of these hormones with different

species or cultivars within a species have promoted plant formation.


For example, in Norway spruce, 12% of the calli with embryos formed

plants when placed on a modified MS medium supplemented with 5 pM 2,4-D

and 5 pM KIN (Hakman and Arnold, 1985). Embryos of cassava (Manihot

esculenta), 4 mm in length, rapidly formed roots and leaves when placed

in MS basal medium with 0.1 mg/L BAP and 0.01 mg/L 2,4-D (Stamp and

Henshaw, 1982). However, with soybean, Lazzeri et al. (1987) determined

that no additional plant formation occurred when different hormone

concentrations and ratios were tested using NAA, BAP, KIN, and zeatin.

The addition of GA has been beneficial for somatic embryo to plant

conversion in some cases. Gibberellic acid at 0.3 pM plus 0.6 pM IBA

improved plant regeneration of soybean (Ranch et al., 1985). After

embryos were in culture two to three weeks, Ghazi et al. (1986)

determined that the addition of 0.104 mg/L GA3 to basal medium resulted

in 20% plant formation compared to 0% when no hormones were added. More

root formation was obtained in wheat when 1 mg/L GA3 was included in the

basal medium (Ozias-Akins and Vasil, 1982). However, no increase in

shoot formation was observed. Chuang and Chang (1987) placed Dysoma

pleiantha somatic embryos on either Gamborg or MS medium containing 1

mg/L BA and 1 mg/L GA. More than 90% of these embryos formed plants.

Rooting of horse chestnut somatic embryos was stimulated by auxin (IAA

or IBA) and GA3, the highest germination being 65%, while cytokinins

(KIN,BAP, and zeatin) inhibited rooting (Radojevic, 1988).

Concentrations of GA3 were varied at 0.5 and 1.0 mg/L, with and

without BAP in combination to improve germination of grape somatic

embryos (Stamp and Meredith, 1988). Plant formation was improved up to

42% with 1.0 mg/L GA; while 0 or 0.5 mg/L GA, or the addition of BAP led

to reduced plant formation (0 to 9%). Filbert shoot elongation was


increased at 30 uM GA; however, 150 and 300 pM levels induced abnormal

development resulting in weaker shoots and less root formation (Perez et

al., 1983).

Other growth substances can also enhance plant formation. Barley

(Hordeum vulgare) plant regeneration was better when MS medium was

supplemented with 3% 2,3,5-Triiodobenzoic acid (47%) compared either

with MS medium containing 1 pM IAA and 1 pM BAP (6%) or medium without

growth substances (29%) (Rengel and Jelaski, 1986).

Often the removal of all hormones in the medium facilitates plant

regeneration. For Norway spruce, the removal of all hormones was

necessary for plant formation (Dunstan et al., 1982). Similarly, about

30% of Digitalis obscura L. embryos developed into plants when placed on

basal medium containing no growth regulators (Arrillaga et al., 1986).

Carbohydrates. Carbohydrate type and level can affect somatic

embryo induction and maturation, but it can also influence subsequent

embryo germination and plant formation. Drew (1979) investigated the

possibility of stocking carrot somatic embryos with sucrose after three

subcultures in media with various sucrose levels. To accomplish this,

sucrose levels were increased two, five, and ten times or left

unchanged. The idea was for the embryos to accumulate an energy source

so that germination and plant formation could occur at sowing without

additional sucrose. However, under non-sterile conditions, only three

plants formed after eight weeks, while most stems that had elongated

became chlorotic and ceased growth. In concurrent studies, embryos with

cotyledons were placed on filter paper bridges supplemented with White's

nutrient medium at various sucrose levels. Without sucrose in the

medium, only 15% of the embryos produced plants under in vitro


conditions, while 1.5% strength sucrose led to 95% plant regeneration in

53 days.

Ahwoowalia and Maretzki (1983) determined that 6% sucrose

stimulated root formation of sugar cane somatic embryos, while 3%

sucrose promoted rooting and plant conversion from mature somatic

embryos. Rietsema et al. (1953) had similar results with excised jimson

weed (Datura stramonium) zygotic embryos in that as the embryo matured,

the optimum sucrose concentration needed for growth and development was


A sucrose hormone interaction on somatic embryo rooting was found

with orange (Citrus sinensis cv Shamouti) (Kochba et al., 1974).

Rooting was best at 5 and 6% sucrose concentrations when adenine sulfate

(ADS) and GA were added to the medium, while rooting without ADS or GA

resulted over a broader sucrose concentration range of 2 to 6%. Only

15% of the embryos rooted on 10% sucrose concentration. In soybean,

nearly 40% of somatic embryos placed on one-half strength MS medium

containing 1.5% sucrose produced plants (Shoemaker et al, 1986).

Chicory plant regeneration was similar when sucrose concentrations were

lowered from 2 to 1% (Heirwegh et al., 1985).

Glucose and sucrose concentrations were adjusted to 1.5, 3, 6, and

12% and their effects on percentage rooting of soybean somatic embryos

tested (Lazzeri et al., 1987). Roots were produced on 73% of the

embryos when placed in medium with 1.5% sucrose, while 30, 15, and 0%

roots formed at 3, 6, and 12% sucrose, respectively. Root formation

(45%) occurred only at the lowest glucose concentration (1.5%). Root

formation was similar when either fructose or sucrose were used at 2%.

Arnold and Hakman (1988) tested the effect of 30, 60, and 90 mM sucrose


on shoot formation of Norway spruce somatic embryos. They found

one-half strength basal medium containing 60 mM sucrose increased shoot

formation 10 to 20% compared with the other treatments.

Redenbaugh et al. (1987) found that in alfalfa, glucose improved

embryo to plant conversion most when compared with sucrose. To a lesser

extent, maltose, cellubiose, lactose, and fructose also improved embryo

conversion in alfalfa when compared to sucrose. No plants were obtained

with galactose. All carbohydrates were autoclaved in the above study.

The effect of carbohydrate breakdown due to autoclaving was also

investigated. While autoclaved maltose resulted in 37% conversion,

filter-sterilization of maltose reduced conversion to 23%. Autoclaved

maltose led to a similar percentage plant formation compared with

autoclaved or filter-sterilized glucose. The authors suggested the

improved embryo conversion was due to a glucose moiety obtained by

hydrolyzation of maltose when autoclaved.

Nutrient media. The role of various nutrient media on soybean

plant quality was investigated by Lazzeri et al. (1987). They reported

inorganic salts affected plant quality more than any other component in

the regeneration medium. In order of best to worst, media for

regeneration of red clover (Trifolium pratense) were: hydroponic

nutrient medium > 1/2 MS > White > Gamborg > 1/2 Gamborg > MS. Gamborg

and MS media were compared for their effect on apical development and

germination of soybean embryos at two, four, and six weeks (Ranch et

al., 1985). The Gamborg basal medium hastened and increased plant

formation when results were pooled from four genotypes. All embryos

which exhibited apical development formed plants within 30 days when

transferred to Gamborg or MS medium containing 0.5 pM IBA.

The percentage of somatic embryos with shoots, that rooted on

media containing either half-strength MS macrosalts or full strength MS

macrosalts and supplemented with 5.7 pM IAA was the same (Rengal and

Jelaska, 1986). Some embryos formed only shoots (10-20%). The authors

suggested this to be a "prolonged effect" response to BAP or auxins of

the induction media.

Stuart and Strickland (1984a) matured alfalfa somatic embryos on

SH basal medium containing various amino acids. They determined that

plant formation (production of trifoliate leaves) depended on the N

source. The percentage plant regeneration ranged between 33 to 67%

after 30 days. Although the addition of proline was mentioned earlier

as a stimulator of embryo production, superior plant formation was

obtained with the addition of 30 mM glutamine instead of proline. The

next best amino acid sources for plant formation were alanine and

arginine. The authors also reported that embryo size was positively

correlated with embryo conversion, that is, larger embryos consistently

resulted in a greater number of more vigorous plants.

Light. Arnold and Hakman (1988) investigated the effect of light

(20 hours) versus darkness on germinating Norway spruce embryos.

Germination (not defined) occurred both in light and dark incubation

conditions. In the dark, 35% of the embryos formed plants, while in the

light, 25% of the embryos formed plants after three weeks. However,

embryos exhibited hypertrophy when left in the dark for two weeks.

Therefore, the authors suggested the embryos be transferred to light as

soon as the roots began to elongate.


Yield Comparisons Between Plants Derived from Somatic Embryos or
Vegetatively Propagated

One field study compared yields from regenerated plants of napier

grass (Pennisetum purpureum Schum.) somatic embryos with yields from

vegetatively propagated plants (Rajasekaran et al., 1987). Plants

derived from tissue culture yielded 2,400 and 3,510 kg ha-1 more biomass

than vegetatively propagated plants in two separate studies. The

authors attributed the greater yields from tissue culture-derived plants

to earlier establishment and formation of significantly more tillers

than vegetatively propagated plants.

Somaclonal Variation

One of the major advantages gained by propagating via somatic

embryogenesis is clonal fidelity. Clonal fidelity must be maintained or

synthetic seeding would not be a viable option. Several investigators

have reported variability in plant tissue culture (Bayliss, 1980; Larkin

and Snowcraft, 1981). In fact, some researchers have suggested tissue

cultures would serve as an excellent source for creating new variability

resulting in new genotypes (Shepard et al., 1980; Larkin et al., 1984;

Zong-xiu et al., 1983; etc.). To date, however, no commercially grown

variety has been developed through somaclonal variation (Vasil, 1987).

Clearly, in some cases, plants obtained from tissue culture have

varied from the donor plant. Some of the most convincing evidence of

genetic variation from tissue culture was documented through cytological

examination. For example, McCoy et al. (1982) observed chromosome

breakages, partial chromosome loss, and various aneuploid types in oat

cell cultures. Similarly, McCoy and Phillips (1982) observed aneuploidy

in a number of corn cell cultures, while Sacristan (1971) found


chromosome rearrangements and polyploids in Crepis capillaris cell


Vasil (1987) points out that the type of cell culture

(differentiated vs undifferentiated) from which the cytological and

phenotypical variations of cells or plants were obtained often are not

well characterized, and Vasil concludes that genetic changes do not

occur in embryogenic cell cultures. He goes on to say that somatic

embryos do not develop beyond an early developmental stage akin to what

occurs in nature when an embryo aborts and loses its regenerative

capacity. This view is supported by Browers and Orton (1982) who found

celery plants were regenerated when cell cultures contained primarily

diploid cells whereas regeneration was not obtained with nondiploid

cells. Although the nutrient medium, transposition events, and the loss

of spacial and temporal control of cell division and differentiation

that operate in intact tissues and plants may lead to genetic

variability in culture, similar genetic changes can occur in vivo, but

often escape detection (Vasil, 1987). Thus, a natural variability is

already conferred into the differentiated tissues of a plant.

Observation of pearl millet (Pennesetum americanum) embryogenic cells

when cultured over a extended time period revealed that most cells

remained cytologically stable and that selection pressures favored the

formation of somatic embryos and plants from normal cells (Swedlund and

Vasil, 1985). From these reports, it seems apparent that plants

obtained from somatic embryos from embryogenic cells are likely to

maintain their genetic integrity.


Sweet Potato Tissue Culture


Although sweet potato ranks seventh among agricultural crops

world-wide, relatively few studies have dealt with cell and tissue

cultures of this species. Root and adventitious shoot formation was

obtained from callus derived from root plugs (Gunkel et al., 1972;

Yamaguchi and Nakajima, 1973). Responses to various hormone levels and

their effects on the rate of callus proliferation and shoot regeneration

were tested in two cultivars (Litz and Conover, 1978). Murashige and

Skoog medium supplemented with IAA stimulated callus production in

'White Star' explants. The addition of activated charcoal inhibited

callus growth, and 1 mg/L BAP promoted shoot proliferation averaging 8.5

shoots five weeks after culture establishment. Cultivar PI 315343

averaged 3.4 less shoots, while rooting and subsequent plant formation

was enhanced with 1 mg/L NAA.

Root and shoot formation was obtained from callus derived from

leaf explants on MS medium supplemented with 10-20 mg/L adenine or

0.1-0.5 mg/L KIN (Sehgel, 1975). Sehgal (1978) also reported nodulation

followed by plant formation from anther callus obtained on MS medium

containing adenine (10 or 20 ppm) plus 2,4-D (Ippm). Adventitious

shoots were generated from root disks in one of three species tested,

while shoots readily formed in all three species from petiole sections

(Hwang et al., 1983). Callus formed best from stem, leaf, and storage

root explants when placed on MS medium supplemented with 1.0 mg/L NAA

and 10 mg/L IBA. Roots and shoots also formed when roots were

subcultured on medium containing 1.0 mg/L NAA and 0.1 mg/L IBA. Plant

formation was improved by light.

Somatic Embryogenesis

Callus was derived from anthers (Tsay and Tseng, 1979). Optimum

callus production was obtained on MS medium supplemented with 2,4-D and

KIN. Somatic embryos formed in one cultivar when basal medium contained

0.1-1.0 mg/L ABA; however, no somatic embryos formed in the other two

cultivars. Plants were regenerated when embryos were placed on basal

medium supplemented with 1 mg/L IAA and 4 mg/L KIN.

Somatic embryos differentiated from callus derived from auxiliary

bud shoot tips in six of nine genotypes (Jarret et al., 1984).

Auxiliary shoot tips removed nearest the shoot apex resulted in the

greatest formation of embryogenic callus. Embryogenesis was facilitated

by 2,4-D concentrations between 0.1 to 3.0 mg/L for most genetypes.

Plant formation was obtained when embryos were removed and placed on

basal medium with half-strength MS salts and no hormones.

Embryogenic callus of sweet potato was generated when leaves,

shoot-tips, stems, and roots of the cultivars 'White Star' and 'GaTG 3'

were placed on MS medium with 1 mg/L 2,4-D (Liu and Cantliffe, 1984a).

However, embryogenic callus was most consistently obtained from leaves

(23%). Callus was readily produced between 2,4-D concentrations of 0.5

to 2.0 mg/L, while above 2.0 mg/L, callus formation was inhibited.

Embryos germinated and formed plants when removed from a medium with

hormones to one without. Plants flowered, five months after being grown

in growth chambers and some storage roots formed which were 15 cm or

larger in diameter.

Liu and Cantliffe (1984b) optimized 2,4-D and sucrose

concentrations for induction of embryogenic callus. Embryogenic callus

formed most frequently with 1 mg/L 2,4-D and 3% sucrose, while no


embryogenic callus formed without 2,4-D. Apical meristem domes alone or

with the youngest pair of leaf primordia most frequently formed

embryogenic callus (80-85%) than domes with two pairs or more leaf

primordia (30% or less).

Embryo formation was best when callus was fractionated and sieved,

such that the 500 to 2000 pM sizes were collected and plated on MS

medium containing one-half the concentration of ammonium nitrate (Chee

and Cantliffe, 1988a). Chee and Cantliffe (1988b) determined three

developmental stages in which somatic embryos become plants:l) late

torpedo; 2) cotyledonary; and 3) expanded torpedo stage. Larger embryo

sizes reportedly did not equate to more mature embryos and somatic

embryos generally had small cotyledons and large hypocotyls compared to

zygotic embryos. Conversion to plants from the above three embryo types

was approximately 50% when removed and subcultured on MS medium without


Delivery Methods for Somatic Embryos

Current Planting Scheme

Until recently, the ability to obtain at least a few plants from

somatic embryos was considered a major success. However, during the

1980s, more emphasis has been placed on improving embryo vigor and early

seedling growth, and increasing the percentage embryo to plant

conversion. With progress being made toward these goals, increased

attention has also been placed on development of delivery methods

suitable for economic propagation of mass quantities of somatic embryos.

To date, most plant conversion systems require several

intermediate steps prior to successful plant establishment in the field.

For example, Haydu and Vasil (1981) germinated napier grass (Pennisetum


purpureum) somatic embryos in tissue cultures until they were one to

five centimeters long. These small plants were then transferred to a

hormone-free medium in culture tubes to establish a more vigorous root

system. The next step was potting the plantlets into soil-vermiculite

in a growth chamber, then acclimating the plants to lower relative

humidities. Finally, plants were transplanted in the greenhouse and/or

the field.

For some species, plantlet formation is a very slow process.

Celery somatic embryos required five weeks before forming roots and

leaves (Dunstan et al., 1982). After adequate growth was obtained, the

plants were transferred to soil in propagation boxes in the greenhouse,

then to individual pots. Speigal-Roy and Vardi (1984) developed a

procedure for plant regeneration of citrus which required several

transfers from solid and liquid media. After plants reached a certain

size, they were cultured in tubes on paper bridges for further

development prior to transfer to soil. Sixteen to 18 weeks were

required before plants were obtained from somatic embryos and were

growing in the greenhouse.

Extensive transfer steps were also required to establish plants

from papaya Carica papaya L. somatic embryos (Litz and Conover, 1982).

Embryos were germinated and formed plants on White's medium supplemented

with 0.1-2.0 mg/L NAA and 0.05-0.20 mg/BAP. Plants were then moved to a

soilless potting mix and hardened off under intermittent mist for 2-2.5

weeks. These few examples illustrate the laborious and time consuming

task before plants from somatic embryos can be transplanted in the



Proposed Handling and Planting Methods of Somatic Embryos

Several delivery methods for propagating somatic embryos directly

from in vitro conditions to the field or greenhouse have been proposed.

They include: 1) simultaneous desiccation of somatic embryos using a

water soluble resin and planting in a wafer or seed tape (Kitto and

Janick, 1985ab); 2) desiccation, then planting dried somatic embryos

with a conventional drill (Gray et al., 1987); 3) encapsulation of

singulated somatic embryos in an alginate gel capsule, then planting

using a conventional drill (Redenbaugh et al., 1984,1986; Jeon et al.,

1986); 4) gel seeding somatic embryos with fluid drilling equipment

(Drew, 1979; Baker, 1985; Schultheis et al., 1986; Cantliffe et al.,

1987); and 5) an automated system for direct transfer of somatic embryos

to the greenhouse (Levin et al., 1988).

Desiccation in polyethyleneoxide. Desiccation is part of the

normal process during seed development and maturation (Bewley and Black,

1978). Desiccation of somatic embryos has been investigated with the

intent of arresting embryo growth, providing storage capability, and the

ability to plant as normal dry seeds (Kitto and Janick, 1985abc; Gray et

al., 1987; Gray, 1987).

Kitto and Janick (1985ab) were the first to successfully desiccate

somatic embryos of carrot and have them survive upon rehydration.

Addition of unspecified numbers of carrot embryos from suspension were

incorporated with polyethyleneoxide (Polyox), a water-soluble plastic

resin, and dried to form wafers under sterile conditions in a laminar

flow hood at ambient temperatures and humidities (Kitto and Janick,

1985b). A constant weight was achieved after 6.5 hours drying. Three

percent of the embryos survived encapsulation once rehydrated, although


it was unclear in the report as to whether plants formed. Embryo

survival decreased rapidly over drying time. No survival was obtained

when embryos were dried to a constant weight and not coated; however,

some embryos survived after 32 hours when they were coated with Polyox.

Hardening treatments, including 12% sucrose, chilling at 4C, high

inoculum density, and/or 1 pM ABA prior to desiccation reportedly

increased embryo survival compared with the nontreated control (Kitto

and Janick, 1985c). The use of a Polyox seed coat seemed to damage the

embryos during the coating and drying process as supported by low

percentage survival data. Desiccation of somatic embryos in Polyox does

not appear to be a practical seeding method, especially since coating

somatic embryos singly was difficult and that no plants formed after the

coating process.

Desiccation. Desiccation of somatic embryos, as a separate step,

prior to encapsulation with a coating material has been proposed (Gray,

1987; Gray et al., 1987). Somatic embryos of orchardgrass became

quiescent when dried from 75 to 13% moisture (Gray et al., 1987). After

21 days of desiccated storage at 230C, 12% of embryos formed roots

(germination), while 4% grew into plants. Developmental stage before

desiccation was critical for embryo germination and plant formation.

Only those embryos with well differentiated scutellar and coleoptilar

regions prior to desiccation were able to germinate and/or form plants

after imbibition.

Similar studies were conducted with grape somatic embryos (Gray,

1987). More germination (58%) and plant formation (20%) were obtained

with grape compared with orchardgrass (12% and 4% respectively) after 21

days of desiccated storage. Only 5% of grape somatic embryos


regenerated plants when embryos were not desiccated. Gray suggested

that desiccation may be an additional dormancy breaking mechanism in

grape. Breaking dormancy of grape somatic embryos by low temperatures

or with GA has already been documented (Rajasekaran and Mullins, 1979).

Parrott et al. (1988) reported that plant formation from soybean

somatic embryos typically was spread over a period of nine months.

However, after two weeks, when results from seven genotypes were pooled,

plant regeneration averaged 60% when embryos were desiccated compared

with 1% for embryos not desiccated. This study differed from the

proceeding desiccation studies in that it included a maturation period

of 28 days after which embryos were removed from their explant tissue

prior to desiccation.

Desiccation treatments have been useful for plant regeneration of

soybean and grape somatic embryos; however, the percentage plant

formation is still unacceptably low. Also, as storage times of the

desiccated embryos were increased, plant vigor and regeneration quickly

decreased (Gray et al., 1987). Relatively little is known about the

biological effects of desiccation. Thus, better control of the drying

process and more understanding of the desiccation process is necessary

before drying of somatic embryos can be incorporated in a synthetic seed


Encapsulation in soft gel capsule. A third delivery method

proposed was encapsulation of somatic embryos in a soft gel capsule

(Redenbaugh et al., 1984, 1986; Jeng et al., 1986). Several compounds

were tested for suitability for encapsulation, but calcium alginate, a

food thickener, derived from brown algae was selected (Redenbaugh et


al., 1986). This was selected due to its low embryo toxicity and ease

in forming capsules around the somatic embryo.

The gel capsules surrounding the embryo can be amended with

nutrients, hormones, and carbohydrates to facilitate the rapid growth

and survival of embryos. Most encapsulation research has dealt with

alfalfa somatic embryos which to date have consistently averaged 60%

plant formation under in vitro conditions and 20% when grown in the

greenhouse (Redenbaugh et al., cited in Fujii et al., 1987). The best

conversion frequency was 65% with celery somatic embryos when grown

under in vitro conditions with carefully selected embryos (Redenbaugh et

al., 1986). Other crops encapsulated and forming plants in vitro were

Brassica species, cotton, lettuce, and corn (Redenbaugh et al., 1987a).

Although encapsulated somatic embryos have been successfully converted

to plants in several species, there are problems with the alginate

capsules. Water-soluble nutrients have been reported to rapidly leach

out of the capsule (Redenbaugh et al., 1987b). Root and shoot emergence

have also been inhibited. In addition, Redenbaugh (1987b) stated there

is "lack of automation technology to allow medium to large-scale

production of single-embryo capsules." Finally, storage may be a

problem since embryo viability may decline over time due to inhibition

of embryo respiration in the capsule (Fujii et al., 1987). This seeding

technology is relatively new and will require much more intensive study

before it can be applied commercially.

Fluid drilling. Fluid drilling is the sowing of seeds in a

protective, flowable gel. The system was developed in England during

the 1970s (Gray, 1981). The concept of fluid drilling zygotic seed was


first conceived by Currah et al. (1974), while Drew (1979) was one of

the first to suggest the idea of fluid drilling somatic embryos.

Fluid Drilling


Fluid drilling has improved plant establishment, and led to

earlier, improved yields in a number of small seed species such as

carrot (Currah et al., 1974), tomato (Gray et al., 1979), parsnip (Gray

and Steckel, 1977), celery (Biddington et al., 1975), lettuce (Gray,

1978) and pepper (Ghate and Phatak, 1982). Actively growing or

pregerminated seeds can be sown in the gel without damage. The gel can

also be amended with nutrients, carbohydrates, beneficial microorganisms

(mycorhizae, bacteria), and pesticides to stimulate embryo growth and

improve plantlet survival. The microenvironment created around the

seeds when fluid drilled can be modified and optimized such that the

above advantages can be realized.

Gel Type

Due to their various chemical compositions, certain gels may be

better seed carriers for fluid drilling. For example, some gels may

facilitate better oxygen movement, some may contain nutrients which

enhance plant growth, and others may hold more moisture. Several gels

have been used for fluid drilling zygotic seed. Some of the more

commonly used gels are: Laponite (magnesium silicate clay,), Liqua-gel

(potassium starch acrylamide manufactured), Planta-gel (copolymer of

potassium acrylate and acrylamide), Terrasorb (starch or synthetic

copolymer), Hydrozorb 30 (potassium acrylate), and N-gel (various

cellulose based materials, formerly called Natrosol).


A laboratory study by Brocklehurst (1979) determined the

polyacrylate mineral colloid gel was superior to the natural gum gels

since the later caused senescence of lettuce seedlings four days after

planting. Sosa-Coronel and Motes (1979) evaluated seven commercial gels

for planting carrot, onion, and pepper seed. Pregerminated seeds were

suspended in gels for four and 24 hours prior to seeding in the

greenhouse. Superior seedling growth was obtained for all species sown

with Viterra 2, Natrosol 250, and Bacto-Agar gels, while Gum Blend and

CLD were toxic.

Oxygen movement is dependent on gel type and temperature. Frazier

et al. (1982) determined that oxygen movement in gels was slower as

temperature decreased. Slower oxygen movement in the gel could be

critical since low temperatures at planting might restrict the

metabolism of the pregerminated seed. Seed viability was related to

oxygen diffusion in the gel. Of all the gels tested, Natrosol had the

highest oxygen diffusion rate and led to the most radicle emergence and

growth of snapdragon seedlings when suspended in gel. Pill and

Fieldhouse (1982) ran a similar comparison and also found that Natrosol

was the only gel that tomato seed could be stored in for more than four

days without reducing the percentage and rate of emergence. Bryan et

al. (1982) observed slower plant emergence and reduced pepper stands

when germinated seeds were sown in Laponite gel than without gel during

wet growing conditions. Laponite has been reported to reduce 02 uptake

by the seed (Brocklehurst, 1979).

Protection from Damage and Desiccation

Gels will protect actively growing seeds with or without roots.

In addition, gels absorb water to many times their weight. This keeps

moisture around the seeds and reduces the possibility of desiccation.

Gels may retain moisture around the seeds several days after planting

(Bryan et al., 1978).

Gel Additives

Ghate (1982) studied the theological properties of gels when

amended with chemical additives. Little change occurred in the gel

without additives after one day when temperatures varied between 10 and

37*C. The addition of fertilizer caused gels to become thin while,

relative consistency was maintained after the addition of more gel.

Nutrient salts have been added to gels at planting to enhance

germination and early seedling growth from zygotic embryos. Nitrogen,

phosphorus, and K were added to guar gum gel with little effect on

lettuce or cabbage growth (Costigan and Locascio, 1982). Finch-Savage

and Cox (1982) reported improved early seedling growth of carrots with

the addition of monosodium phosphate at concentrations up to 30 g/L, but

nitrate (5-20 g/L) reduced seedling growth. In a similar study,

Finch-Savage and Cox (1983) obtained earlier maturity in lettuce and

onion when pregerminated seeds were fluid drilled in gel supplemented

with either sodium phosphate or ammonium nitrate compared with

nontreated, dry sown seed. No benefit was obtained when less than 10

g/L ammonium nitrate was incorporated in the gel.

Faster, more complete germination of pepper seeds was reported

when cytokinins, humic acids, or alpha-keto acids were added to the gel

and planted under wet environmental conditions (Bryan et al., 1982).

The addition of GA3 and GA4/7 to gel at 2 and 2.5 mg/L increased the

growth of tomato seedlings compared with gel without GA (Ohep and

Cantliffe, 1980). Tomato plants were taller when the gel was


supplemented with either Cytex or Cytozyme, while increased plant

height, dry weight, and leaf area under field conditions were obtained

when humic acids were incorporated. Gibberellic acid and diphenamid

hastened the emergence rates of pepper by one to three days (Ghate and

Phatak, 1983).

Various pesticides have improved plant establishment. Ohep et al.

(1984) reduced damping off of tomato seedlings by incorporating ethazol

plus thiophanate-methyl, fenaminosulf, and ethazol alone or combined

with benomyl, chloroneb, or captain into the gel. White (1979)

controlled white rot of onions more efficiently by placing 25 mg a.i./L

metalaxyl in the gel, while much higher concentrations of metalaxyl were

required to overcome the problem when incorporated in the soil.

Similarly, iprodione, when placed in gel, effectively controlled white

rot of salad onion at one-quarter the rate used for dry seeds

(Entwistle, 1979).

Hadar et al. (1984) added the biocontrol agents Trichoderma

harsianum and T. kininoii to gel to prevent seed rot in peas (Pisum

sativum L.) and navy beans (Phaseolus vulgaris L.) when sown in pythium-

infested soil. They determined that 95% of the plants emerged when

seeded in gel amended with T. harsianum and kininoii compared with 15%

when planted with dry seeds. Inclusion of Rhizobium in the gel resulted

in more than three times more root nodules on navy bean compared to

inoculation of dry seed (Hardwick and Hardaker, 1977).

The growth of Trichoderma hamatum, a biocontrol and antagonist

used to control Rhizoctonia solani, was compared in Viterra, Natrosol,

and Laponite gel carriers (Mihuta-Grimm and Rowe, 1986). Superior

growth of Trichoderma isolates was obtained in Natrosol versus Viterra


and Laponite gels. Due to better growth of the antagonist in Natrosol,

this gel was supplemented with Trichoderma to control damping off when

fluid sowing radish seeds. Four delivery systems for application of

Trichoderma isolates were compared; 1) fluid drilling in Natrosol, 2)

coated seed, 3) application of wheat-bran grown isolated applied in the

furrow before planting and 4) soil drenching with five day broth grown

cultures. The best disease control was consistently obtained with fluid


A peat-vermiculite mix was combined with Liqua-gel (gel mix) and

resulted in earlier, more uniform, and complete plant emergence of

pepper compared with gel seeding in either Laponite or Liqua-gel gels,

plug-mix, or dry sown seed (Schultheis et al., 1988). The increased

availability of moisture and nutrients were suggested as a reason for

improved plant stands.

Gel Studies with Somatic Embryos

Baker (1985) determined that a mixture of peat and vermiculite was

superior to vermiculite alone as a soil planting medium for carrot

somatic embryos since the former alleviated water stress. However,

vermiculite was the best soil cover since it did not crust and inhibit

emergence. Baker also compared Natrosol and Planta gel for suitability

as carriers for fluid drilling somatic embryos. No difference was

measured between gels with respect to embryo emergence, which was

approximately 5 to 30% under in vitro conditions. The polyacrylamide

material was used in subsequent greenhouse studies. In addition to

Baker, Schultheis et al.(1986) reported that somatic embryos of sweet

potato grew quite rapidly in N-gel when compared to three other gel



Naphthalene acetic acid and/or GA were placed in the gel to hasten

and improve survival of carrot somatic embryos; however, no advantage

was gained from the various hormone combinations used (Baker, 1985).

Survival ranged from 1 to 7% under greenhouse conditions after seven

days, but no plants developed into plants. Embryos survived in the gel

for several days, but stopped growth after a few days.

Soil Covers

Soil covers have been used to provide an environment favorable for

plant establishment. A large number of materials have been tested as

anticrustants, from synthetic organic polymers to plain sand. Few

materials, unfortunately, have been shown to have broad applications.

Their effectiveness apparently is closely tied to soil type, species,

and environmental conditions, particularly moisture and temperature.

Hemphill (1982) evaluated the effect of several anticrustants on plant

establishment for a number of small seeded

species. He found the most consistent establishment was obtained when

using vermiculite followed by phosphoric acid.

In more recent studies, several soil amendments were evaluated for

their effects on cabbage (Perkins-Veazie and Cantliffe, 1987), tomato

(Odell and Cantliffe, 1987), and lettuce (Seale and Cantliffe, 1986)

stand establishment under a range of growing conditions. A calcine clay

(GrowSorb, MidFlorida Mining), high temperature fired (LVM) 24/48,

consistently improved percentage plant emergence. Seale and Cantliffe

(1986) suggested the advantage obtained with GrowSorb might be

attributed to its ability to maintain favorable moisture conditions

during germination. Another promising soil amendment was gel mix

(mixture of peat and gel) (Schultheis et al., 1988). This hastened


emergence and usually increased seedling growth both in tomato (Odell

and Cantliffe, 1987) and lettuce (Seale and Cantliffe, 1986).


In summary, there are many factors such as explant source, light,

temperature, nutrition, hormones, carbohydrates which influence the

quantity and quality of somatic embryos produced. These same factors,

as mentioned previously, affect the formation of roots, shoots and

plants obtained from somatic embryos. Although somatic embryos are

similar to zygotic embryos, the former lack many of the stored reserves

zygotic embryos have available for their early seedling growth. Thus,

it is important to find which "growth factors" are necessary to achieve

rapid and complete plant formation from somatic embryos. It is also

necessary to use a synthetic seeding method which could be easily

adapted for field use. The incorporation of additives around the

embryos to sustain and improve their growth and development into plants

would be required for the successful implementation of synthetic


The research which follows has focused on the fluid drilling

method. Fluid drilling offers several advantages over the other

planting methods mentioned earlier. They include the ease of amending

the gel with additives around the embryo, the ability to plant the

embryo in an advanced growth stage with no need for desiccation, the

high moisture retention of the gel which could protect the embryo from

desiccation and injury at planting, and the available technology and

equipment available for large-acreage planting.

The primary goal of the following studies was to find a gel

carrier and amendments suited for fluid drilling somatic embryos. Once


the gel carrier was determined, a series of experiments were conducted

to determine the gel additives that would promote rapid, uniform, and

complete plant conversion from somatic embryos. Early plant growth and

root yield comparisons were also made between plants obtained from

somatic embryos and vegetatively propagated.



Patterns of somatic embryony in sweet potato (Ipomoea batatas

Poir.) were recently determined (Chee and Cantliffe, 1988). The authors

identified embryos at three developmental stages which had the potential

of becoming plants: 1) torpedo, 2) cotyledonary, and 3) elongated

torpedo. Embryos capable of plant development usually ranged from 1 to

4 mm long.

Selection of the proper embryo developmental stage and size could

improve germination and subsequent plant formation of sweet potato.

Gray et al. (1987) determined that orchardgrass (Dactylis glomerata)

somatic embryos must have well-developed scutellar and coleoptilar

regions for germination to occur. Morphological features of the embryos

were suggested by Stuart et al. (1985) to be an identifiable marker in

determining which embryos would more readily convert to plants. Embryo

length has been positively correlated with germination and/or plant

formation in zygotic and somatic embryos (Gray et al., 1987; Austin et

al., 1969; Soffer and Smith, 1974; Stuart et al., 1985).

Several delivery methods have been proposed for direct-field

synthetic seeding of somatic embryos: 1) encapsulation of singulated

embryos in an alginate gel capsule (Redenbaugh et al., 1984, 1986); 2)


embryo desiccation in a hardened material (Gray et al., 1987); 3) embryo

encapsulation in a soft gel capsule and their subsequent desiccation

(Jeon et al., 1986); 4) simultaneous desiccation of embryos in water

soluble resin (Kitto and Janick, 1985ab); 5) an automated system for

direct transfer of embryos to the greenhouse (Levin et al., 1988) and 6)

fluid drilling (Drew, 1979:Baker, 1985; Schultheis and Cantliffe, 1988).

Drew (1979) was one of the first to suggest fluid drilling as a delivery

method but few researchers have investigated synthetic seeding using

this method (Baker, 1985; Schultheis and Cantliffe, 1988).

The purposes of these studies were two-fold. One goal was to find

a gel carrier which was compatible for fluid drilling somatic embryos.

Upon identification of the gel, the second goal was to determine the

effect of embryo developmental stage and embryo size on the timing and

percentage of root and plant development to improve synthetic seeding

via fluid drilling.

Materials and Methods

Embryo Production

Sweet potato plants, cv. White Star, were vegetatively propagated

and maintained in a growth chamber at 27C and 16 hours light/12 hours

dark photoperiod. Vigorous vegetative growth was maintained by frequent

fertilizer applications, watering, and pruning of the stems. Shoot tips

2-cm long were excised and surface sterilized in 1.2% hypochlorite and

Tween-20 for five minutes. Apical domes with up to two leaf primordia

pairs were plated on agar-solidified medium containing 10 pM 2,4-D, and

basal medium containing inorganic salts of Murashige and Skoog (MS)

(1962), 500 pM myo-inositol, 5 pM thiamine.HC1, 10 pM nicotinic acid, 5

pM pyridoxine.HC1 (Table A-1), 3% wt/vol sucrose, and 0.7% wt/vol


'Phytagar' (GIBCO Lab., Grand Island, NY). The pH was adjusted to 5.8

using 1.0 N NaOH solution, then autoclaved for 20 minutes. Twenty ml of

medium was poured into presterilized 100 x 15 mm plastic Petri plates.

Incubation was in the dark at 27*C with unmonitored light interruptions

necessary for routine observation.

After eight weeks, embryogenic callus was transferred to Petri

plates containing basal medium plus 10 pM 2,4-D and 1 pM 6-

benzylaminopurine. Six to eight weeks later, embryogenic callus was

bulked from several plates and placed in beakers. Cell aggregates were

gently broken apart into natural units with a glass slide, then

separated into a 355 to 710 micron size fraction, and a 710 micron and

larger fraction with copper sieves. The largest-sized fraction was used

for embryo production and plated on basal medium without growth

regulators, 10 mM ammonium nitrate (half strength), and 0.8% wt/vol

Phytagar. The other fraction was used to maintain and proliferate

embryogenic callus. Incubation was one week in the dark with

unmonitored light interruptions, then in light in a 10/14 light/dark

cycle at 27*C until embryos were selected.

Gel Study

A solution containing MS inorganic salts, 500 pM myo-inositol, 5

pM thiamine.HCl and 3% wt/vol sucrose was adjusted to a pH of 6.5, then

mixed with each gel at a rate of 2.0% wt/vol. Four gel types were

compared: 1) copolymer of potassium acrylate and acrylamide (PC)

(Planta-gel, Nepera, Inc. Harriman, N.Y.); 2) potassium starch

polyacrylamide (PSA) (Liqua-gel, Miller Chemical Co., Hanover, Pa.); 3)

potassium acrylate (PA) (Hydrozorb 30, American Colloid, Arlington

Heights, Ill.); and 4) hydroxyethyl cellulose (HEC) (N-gel H, Aqualon


Inc., Wilmington, Del.). The gels were autoclaved for 20 min and about

8 ml of each gel dispensed into quadrants of 100 x 15 mm Petri dishes.

Although the pH was adjusted to 6.5 prior to autoclaving, the pH was 0.5

to 1.0 unit lower after HEC and PA gels were autoclaved, while the pH of

the PSA gel remained near 6.5 and the pH of the PC gel was increased to

about 7.0.

One embryo was placed into the separated quadrants from a group of

randomly selected embryos. Ten embryo samples were used in each of

three replications. Embryo survival was rated by color; green (healthy,

viable embryo), some browning (embryo unhealthy), or completely brown or

black (embryo death) three and six days after being placed in gel.

Embryos were observed after 21 days for root and plant development.

Data were analyzed using analysis of variance and means separated with

Duncan's multiple range test.

Influence of Embryo Size. Maturation Time. and Stage of Development on
the Percentage of Embryos which Grew into Roots and Plants

Embryos at the torpedo, cotyledonary (recognized by formation of

lobed cotyledonary-like lateral appendages), and elongated torpedo stage

(hypocotyl expanded) of development were selected (Fig. 3-1). One of

each embryo type was placed in a 25 x 100 mm Petri plate (Lab-Tek, Nunc,

Inc., Naperville, Il) containing 20 ml gel. The gel was amended with

the inorganic nutrients of MS, and the same organic constituents used

for callus production. Length and diameter of all embryos were measured

using a stereomicroscope with a calibrated ocular lens. Embryos were

observed for root and plant development six, 12, 21, and 31 days after

plating. A radicle at least 2 mm long was defined as a "root" whereas a

well-developed shoot apex (with leaf primordia) and root was defined as

a "plant." Two experiments were conducted. Before embryos were placed

Fig. 3-1. Embryo developmental stages capable of plant conversion.

Torpedo embryo stage with rudimentary cotyledons (left),
cotyledonary embryo stage with small lobed
cotyledonary-like appendages (middle), and elongated
torpedo stage with elongated hypocotyl (right).

in the gel, they were incubated (allowed to mature) on agar-solidified

basal medium for 25 days in experiment 1 and 16 days in experiment 2.

Each experimental treatment was replicated three times, each replicate

had 10 embryo samples. Significance between means were determined using

standard error analysis.

Disregarding embryo development stage, somatic embryos were

separated into three size categories by length: 1) 1.0 to 1.95, 2) 2.0

to 2.95, and 3) 3.0 to 3.95 mm. Differences in root and plant formation

were determined at 31 days using least significant differences analysis.


Gel Experiment

The suitability of a gel carrier for growth and development of

somatic embryos was easily observed by the embryo color change shortly

after placement in the gels. All embryos placed in the HEC gel were

green and healthy after three and six days (Table 3-1). However,

placement of embryos in the acrylate or acrylamide type gels resulted in

at least 23% of the embryos with partial necrosis after three days.

After six days, the PSA gel was the best of the acrylamide gel types

with 40% green embryos. Nearly all embryos were dead after six days

when placed in the other two acrylate or acrylamide gels.

Some embryos grew into plants after being cultured 14 days in HEC

gel (Fig. 3-2). No root, shoot, or plant formation was observed when

embryos were placed in the other gels. Thus, in subsequent experiments,

the HEC gel carrier was used. A rate of 2.5% wt/vol was used instead of

the 2.0% because the thicker 2.5% consistency was most suited for gel


Table 3-1. Embryo survival three and six days after placement in gel.

% Response

Gelz 3 days 6 days
Green Necrosis Dead Green Necrosis Dead

cellulose 100ay Ob Oc 100a Ob Oc

Copolymer 17c 57a 27b Oc 13b 87a

Potassium starch
polyacrylamide 70b 23ab 7bc 40b 36a 23b

acrylate 3c 40a 57a Oc 3b 97a

zMurashige and Skoog inorganic salts, 500 pM myoinositol, 5 pM
thiamine.HC1, and 3% wt/vol sucrose were incorporated in gels.

YMeans separated within columns using Duncan's multiple range test, 5%

Fig 3-2. Effect of gels on embryo survival and plantlet development.

Lack of embryo growth in the acrylamide and/or acrylate gels
(upper left and right quadrants, and lower left quadrant)
and plant formation obtained in hydroxyethyl cellulose gel
(lower right quadrant).

A American Colloid Company's Hydrozorb 30 gel.
L Liqua-gel by Miller Chemical Co.
V Viterra (Planta-gel) by Nepera Inc.
N N-gel by Aqualon, Inc.

Influence of Embryo Size. Maturation Time. and Stage of Development on
the Percentage of Embryos which Formed Roots or Developed into Plants

Experiment 1

Elongated torpedo embryos formed roots earlier than the other

embryo types (Fig. 3-3A). After six days, 37% of the elongated embryos

had roots compared with 7% for the torpedo and cotyledonary embryos.

After 31 days in culture, the total percentage root formation was 58%

for the torpedo and 52% for the cotyledonary embryos, while 88% of the

elongated embryos produced roots.

Plant formation was similar regardless of embryo developmental

stage at 6, 12 and 21 days (Fig. 3-3B). However, the percentage of

embryos which converted to plants from elongated torpedo embryos was

greater than torpedo or cotyledonary embryos at 31 days (38 vs. 21 and

17%, respectively).

Experiment 2

Using embryos at different stages of development had less effect

on the percentage of embryos which formed roots and plants in experiment

2 (Fig. 3-4) than in experiment 1 (Fig.3-3). However, root development

again occurred earlier in the elongated torpedo than either the

cotyledonary or torpedo embryos after 12 days (Fig. 3-4A). The final

percentage root formation was greater with embryos at the elongated

torpedo development stage than those at the cotyledonary development

stage. As in experiment 1, the percentage of embryos which grew into

plants were similar regardless of embryo development stage through day

21 (Fig 3-4B). However, more embryos at the elongated and torpedo

development stage grew into plants compared with cotyledonary stage

embryos (63 and 59 vs. 41%, respectively).

Embryos at three developmental stages removed from maturation
medium after 25 days and their effects on the percentage
root and plant formation after being placed in hydroxyethyl
cellulose gel with the inorganic salts of Murashige and Skoog
medium, 500 MM myo-inositol, 5 pM thiamine.HCl, and 3% wt/vol
sucrose, experiment 1.

A. Percentage root development of embryos at the torpedo,
cotyledonary, and elongated torpedo developmental stage
at 0, 6, 12, 21, and 31 days.

B. Percentage plant development of embryos at the torpedo,
cotyledonary, elongated torpedo developmental stage at
0, 6, 12, 21, and 31 days.

Fig. 3-3.




0 6 12 21 31


0 6 12 21 31


Embryos at three developmental stages removed from maturation
medium after 16 days and their effects on the percentage
root and plant formation after being placed in hydroxyethyl
cellulose gel with the inorganic salts of Murashige and Skoog
medium, 500 iM myo-inositol, 5 pM thiamine.HCl, and 3% wt/vol
sucrose, experiment 2.

A. Percentage root development of embryos at the torpedo,
cotyledonary, and elongated torpedo developmental stage
at 0, 6, 12, 21, and 31 days.

B. Percentage plant development of embryos at the torpedo,
cotyledonary, elongated torpedo developmental stage at
0, 6, 12, 21, and 31 days.

Fig. 3-4.




0 6 12 21 31
















/ .


.- I
______ da, ............~....~ii ----- I -------\ --- \ --


In both experiments, embryos at the elongated torpedo

developmental stage had the greatest average embryo length. The next

longest embryos were those at the cotyledonary stage and the shortest

embryos were at the torpedo stage (Table 3-2). Selection of the

elongated torpedo embryos, which were the largest-size, consistently led

to improved root and plant development compared with embryos at the

other two developmental stages (Fig. 3-3 and 3-4). However, at least as

many of the smaller torpedo embryos grew into plants compared with the

larger cotyledonary embryos. Thus, embryo growth into plants was more

closely associated with developmental stage than to increased embryo


Embryos at all developmental stages were separated according to

length into three categories from both experiments. The percentage of

embryos which formed roots and plants were similar regardless of embryo

size (Table 3-3).


The superior embryo growth and plant formation obtained in the HEC

gel can be attributed to a number of factors. Considering that the

capacity for most gels to absorb water is great and that oxygen is not

very soluble in water, the effects of total quantity and rate of oxygen

movement in the gel was likely to have been extremely influential in the

growth and survival of the embryos in HEC gels. Frazier et al. (1982)

determined that seed viability across gel types was related to the

oxygen diffusion rate (ODR). They reported HEC had better ODR than

either Laponite (magnesium silicate clay) and Permasorb gels. Eastin

(personal communication, Kamterter Inc., Lincoln, Neb.) confirmed these

Table 3-2. Relationship of

Embryo No.
Developmental stage obs.

embryo developmental stage to embryo size.z

Embryo size (mm)





Expanded torpedo

Torpedo 29

Cotyledonary 29

Expanded torpedo 25

ZAverage mean standard error.

----------- Experiment

2.10 0.07Y

2.57 0.15

3.21 0.09

----------- Experiment

1.87 0.06

2.47 0.11

3.55 0.15








2 -----------

0.92 0.04

1.11 0.04

0.93 0.04

Table 3-3. Effect of embryo length on percentage root and plant
formation after 31 days in culture. Embryos were sorted
by size regardless of developmental stage.z

Embryo No.
length embryo Embryo size (mm) % Response
(mm) samples Length Width Root Plant

1.00 1.95 63 1.71 .02Y 0.92 .02 67x 39"

2.00 2.95 76 2.38 .03 1.04 .02 74 38

3.00 3.95 41 3.39 .04 1.00 .03 79 43

ZData pooled from 2 experiments.

YAverage mean standard error.

XLSD(0.05) for comparing root formation means 16%.

WLSD(0.05) for comparing plant formation means 15%.

reports by placing seeds in HEC gel and subjecting them to anaerobic

conditions. He found more oxygen was present in the gel immediately

after than prior to treatment.

The pH of the gels was probably not a major factor affecting

embryo survival and growth. Although gel pHs ranged between 5.5 and

7.2, tissue cultures can grow within a wide range of pHs (Skirvin et

al., 1986).

The HEC gel has other positive features which make it attractive

for planting somatic embryos. In addition to the high ODR, the pH was

measured at 6.5 prior to autoclaving and 5.8 after autoclaving which

were at levels where nutrients were available for plantlet growth.

Also, the HEC gel maintained its viscosity when amended with salts,

while many gels break down. Ghate (1982) reported that low levels of

fertilizer salts reduced the viscosity of all gels tested and their

usefulness as carriers of seed; however, the HEC gel was not included in

his study. It has been reported that HEC gels maintain their viscosity

better than polyacrylamide gels (Ward, 1980). The maintenance of gel

viscosity will be critical for synthetic seeding since additives must be

incorporated into the gel to replace the missing "growth factors" needed

by the naked somatic embryo.

The occurrence of faster root formation from elongated embryos was

probably due to these embryos being at an more advanced developmental

stage and the possible initiation of germination as evidenced by their

elongated hypocotyls. The elongated torpedo embryos, in addition, may

have had better shoot meristem development which increased plant

production compared with embryos at the cotyledonary and torpedo

developmental stages.


For all embryo developmental stages, the percentage that formed

plants was much less than the percentage that formed roots. This

indicated that the shoot meristem was not as well organized or developed

as the root meristem. Chee et al. (1989) reported the lack of shoot

meristem often coincided with root formation in sweet potato somatic

embryo conversion studies. Precocious germination has been reported to

commonly occur in somatic embryos resulting in stunted and abnormal

embryos in species such as carrot and caraway (Ammirato, 1985). Better

apical meristem development is necessary during embryo maturation for

improved and more consistent plant formation.

More roots and plants were produced from embryos in experiment 2

than 1 as nearly two times more plants were produced from embryos at all

developmental stages in the former than latter experiment. The improved

root and plant production obtained in experiment 2 was probably related

to the relatively short incubation time that embryos were cultured on

maturation medium (16 days in experiment 2 vs. 25 days in experiment 1).

The time factor/variation may have affected the availability of either

nutrients and/or carbohydrates obtained from supporting tissues or

medium during embryo maturation. In these experiments, the shorter

incubation time period apparently resulted in better embryo quality.

This may have been due to more availability of nutrients or organic

reserves, while these reserves may have been depleted over time and led

to reduced embryo quality in experiment 1. Delouche (1980) reported

that environmental conditions during zygotic seed development played a

key role in the overall quality of the seeds produced. The same

conclusion most likely holds true for somatic embryos.


Embryo size has been correlated to improved seed vigor carrot

(Daucus carota)(Austin et al., 1969) and lettuce (Lactuca sativa)(Soffer

and Smith, 1974) or somatic embryo quality in alfalfa (Medicago

sativa)(Stuart et al., 1985), but this was not found to be the case in

these studies. The quality of somatic embryos was apparently more

influenced by length of maturation time than embryo size. Redenbaugh et

al.(1987) reported embryo quality as the most important factor for

somatic embryo to plant conversion. They defined quality as the ability

of the embryo to convert to a plant.

Improved seed or embryo quality is the key to the successful

establishment of any crop, whether by direct or synthetic seeding. Use

of protein markers (such as those that accumulate during seed

maturation) rather than morphological markers would appear to be more

effective in determining the overall embryo quality since the latter

does not reveal the physiological status of embryo development. Work in

this regard has been reported with somatic embryos in rapeseed (Sussex

and Crouch, 1982) and soybean (Stuart and Nelson, 1988) However, this

method is destructive and time consuming, while use of a morphological

marker is not. A computer sorting system has been developed for sorting

embryos by size and shape (Grand d'Esnon et al., 1988). Gray (1987)

effectively used morphology to discern which orchardgrass somatic

embryos would convert to plants.


Mass planting of many potentially important biomass crops is not

cost-effective because many of these crops cannot be propagated by seed.

Large quantities of somatic embryos can be cloned rapidly via

embryogenesis. However, the planting of these propagules has been

limited to small-scale use because a number of intermediate transfer

steps are required to insure embryo and plant survival. Direct planting

of somatic embryos produced from tissue culture could be accomplished on

a large-scale by using a sowing method such as fluid drilling. Studies

were conducted to find a gel carrier amenable for fluid sowing somatic

embryos and to identify an embryo maturation time, developmental stage,

and size which led to the fastest and/or most complete root and plant


To determine the compatibility of somatic embryos with gel,

embryos were placed in gel carriers (2% wt/vol) amended with basal

medium containing MS inorganic salts, vitamins and sucrose 3% (wt/vol).

Most embryos died after six days when placed in gels composed of

acrylamide and/or acrylate. All embryos placed in HEC gel were healthy

after six days, while some embryos developed into plants after two

weeks. Thus, HEC was determined to be suited for fluid drilling of

somatic embryos and was used in all subsequent experiments.

Embryos were incubated on agar-solidified basal medium for 25 days

in one experiment and 16 days in another. After incubation, embryos at

the torpedo, cotyledonary, and elongated torpedo stages of development

were selected and placed in the gel. More roots formed earlier from

embryos at the elongated torpedo stage than the cotyledonary and torpedo

stages of development, regardless of incubation time. When embryos were

incubated for 25 days before placement in gel, root development for the

elongated torpedo stage embryos was greater and plant conversion was two

times more at 31 days compared to the other stage embryos. However,

when embryos were incubated for only 16 days before placement in gel,

root and plant development were greater than when incubated for 25 days.


Also, when embryos were incubated for 16 days, the percentage root and

plant formation were similar regardless of embryo developmental stage.

The exception was the lower plant formation obtained with cotyledonary


Thus, the shorter 16 day incubation time is critical for more

plant conversion regardless of embryo developmental stage. The

selection of elongated torpedo embryos consistently resulted in the

greatest root and plant formation regardless of incubation time.

Embryos were also sorted according to size regardless of

developmental stage. In spite of size differences, the percentage of

embryos which formed roots and plants were similar. Embryos could be

sorted by size since elongated torpedo embryos are usually larger;

however, to insure the highest percentage plant conversion, elongated

torpedo embryos should first be selected before sorting.

The compatibility of HEC gel with somatic embryos, the use of

shorter incubation times during embryo development, and the selection of

embryos at the elongated torpedo stage of development are an important

steps towards facilitating the development of synthetic seeding via

fluid drilling.



Sweet potato ranks seventh among food crops grown world-wide (FAO

Production Yearbook, 1986). Although sweet potato is used primarily as

a food source, some cultivars store large quantities of starch and have

good biomass potential (Smith and Frank, 1984). The starches can be

readily converted in these cultivars to ethanol or methane (Smith et

al., 1987).

Commercial plantings of sweet potato are planted by vegetative

propagation. Mass plantings of sweet potatoes for biomass would be too

costly since planting costs would exceed the returns from methane

(Cantliffe et al., 1987). This cost limitation could be partially

overcome by direct field sowing of synthetic seed propagules produced

via somatic embryogenesis.

Use of somatic embryogenesis in a synthetic seeding system offers

several potential advantages. They are: large quantities of propagules

produced in limited space, maintenance of genetic uniformity, rapid

propagule multiplication, and direct planting of somatic embryos into

the field, thus eliminating costly transplanting (Fujii et al., 1987).

Somatic embryos have been routinely obtained in sweet potato (Liu

and Cantliffe, 1984; Chee and Cantliffe, 1988). Although somatic

embryos are nearly identical to zygotic embryos (Steward, 1963;


Wetherall and Halperin,, 1963), the former lack protective seed coats

and the storage reserves that are available to zygotic embryos during

germination. A synthetic seeding method must be developed so that the

somatic embryo will be protected and the necessary growth requirements

strategically placed so that rapid and complete plant establishment will

occur when embryos are sown in the field.

The suspension of somatic embryos in a viscous gel (fluid

drilling) has been suggested as a synthetic seeding method (Drew, 1979).

A hydroxyethyl cellulose gel (N-gel) was proposed, as reported in the

previous chapter, as an ideal gel carrier for fluid drilling somatic


Most reviews dealing with somatic embryogenesis have concentrated

on developmental embryogenesis and have given little attention to the

efficient rooting and shoot formation of somatic embryos. (Ammirato,

1983; Tisserat et al., 1979). The response of somatic embryos to

different nutrient types and concentrations has varied among species

examined. For example, the greatest number of vigorous soybean plants

were obtained with half-strength Murashige and Skoog (MS) medium

(Lazzeri et al., 1987), whereas in barley, the percentage of embryos

which grew into plants was similar regardless whether half or full-

strength MS was used (Rengel and Jelaski, 1986). Stuart and Strickland

(1984a) found that the production of alfalfa plants was dependent on the

source of nitrogen (N), glutamine (30mM) being the best reduced N


A few investigators have determined the effect of different

carbohydrates on root and/or plant production. Redenbaugh et al. (1987)

determined that glucose led to superior alfalfa plant production


compared with maltose, cellobiose, lactose, fructose, and sucrose. Root

production reportedly increased in soybean with 44 mM sucrose compared

with 87 mM glucose (Lazzeri et al., 1987).

The effect of carbohydrate concentration on rooting and plant

development has varied among species. Soybean somatic embryos produced

roots in 73% of the embryos when placed in medium with 44 mM sucrose,

while 30, 15, and OX roots were produced from embryos when basal medium

was supplemented with 87, 154, and 349 mM sucrose, respectively (Lazzeri

et al., 1987). In sugarcane, rooting was increased with 154 mM sucrose;

however, shoot production was reduced (Ahwoowalia and Maretzki, 1983).

When the sucrose concentration was lowered to 87 mM, both root and shoot

production in sugarcane were improved. Chicory plant production from

somatic embryos was not affected by reducing the sucrose concentration

from 58 to 29 mM (Heirwegh et al., 1985). In Norway spruce, 60 mM

sucrose promoted shoot production when compared with 30 or 90 mM (Arnold

and Hakman, 1988).

The addition of auxin to basal medium has been beneficial for

improved plantlet formation in some cases (MacKinnon et al., 1987; Zhi

hou et al., 1984), while the addition of GA reportedly improved root and

plant development from somatic embryos in species such as soybean (Ranch

et al., 1985), horsechestnut (Radojevic, 1988), citrus (Kochba et al.,

1974) and grape (Stamp and Meridith, 1988).

The percentage of somatic embryos which develop into plants and

the rate in which the embryo grows into a plant must be improved if

field plantings are to be successful, especially when sown under hostile

conditions of environmental stress. The following studies were

conducted to optimize nutrient, carbohydrate, and hormone formulations

for root and plant production in sweet potato in a fluidized gel. The

determination of additives that improve plant formation in a gel carrier

would be an important step in the development of the fluid drilling

method for synthetic seeding.

Materials and Methods

Embryo Production

Sweet potato plants, cv. White Star, were grown and maintained in

growth chambers at 270C. The photoperiod was 16 hours light and eight

hours dark and light intensity was 980 pEm-2s-1. Healthy shoot tips 2-cm

long were removed from the plants, and surface sterilized in 1.2%

hypochlorite and Tween-20 for five minutes. Apical domes with up to two

leaf primordia were plated on agar-solidified medium containing 10 AM

2,4-D, and basal medium containing the inorganic salts of MS (1962), 500

pM myoinositol, 5 pM thiamine.HCl, 5 pM pyridoxine.HCl, 10 pM nicotinic

acid (Table A-l), 87 mM (3% wt/vol) sucrose, and 0.7% wt/vol 'Phytagar'

(GIBCO Lab., Grand Island, NY). Prior to autoclaving, the pH was

adjusted to 5.8 using 1.0 N NaOH solution. Twenty ml of medium was

dispensed into presterilized 100 x 15 mm plastic Petri dishes.

Incubation was in the dark at 27*C with unmonitored light interruptions

that were necessary for routine observations.

Embryogenic callus was removed from plates after eight weeks and

placed in basal medium containing 10 pM 2,4-D and 1 pM 6-

benzylaminopurine. Six to eight weeks later, embryogenic callus was

removed from plates and collected in beakers. Cell aggregates were

broken apart at their natural cleavage lines using a glass slide, then

separated using copper sieves into a 355 to 710 micron fraction and a

710 micron and larger fraction. The largest-sized fraction was plated


for embryo production on basal medium containing no growth regulators,

10 mM ammonium nitrate (half-strength) and 0.8% Phytagar. Incubation

was one week in the dark, then for the remaining duration in a 10/14

light/dark cycle at 27C.

Gel Additive Studies

All gel additive studies were conducted in vitro. The treatment

media solutions were mixed and the pH adjusted to 6.5 prior to the

addition of 2.5 X wt/vol hydroxyethyl cellulose gel (N-gel, Aqualon,

Wilmington, Del.). Gel solutions were autoclaved for 20 minutes, then

25 ml of each gel treatment was poured into a presterilized 25 x 100 mm

Petri plate (Lab-Tek, Nunc, Inc., Naperville, Il). Embryos were used

when growth was arrested at either the torpedo, cotyledonary, or

elongated torpedo stages of development after 16 to 25 days (Fig. 3-1).

In each experiment, embryos at the same developmental stage were chosen

for each embryo sample and distributed equally among treatments.

Embryos were incubated in growth chambers at 27C in a 14 hour light and

10 hour dark photoperiod. Light intensity was 250 /iE m-2s-1.

Nutritional studies

Three types of nutritional studies were conducted to determine

their effects on the rate and/or percentage of root, shoot, and plant

production. One study compared the effects of various concentrations

and types of inorganic salts in previously published nutrient media,

including Hoagland's solution (1950), Gamborg's B5 medium (1968),

White's medium as reported by Singh (1981), and MS (1962) medium (Table

A-2). A gel with no inorganic salts served as a control treatment. The

second study tested the concentration of MS basal medium (Table A-1) at


0, 1/8, 1/4, 1/2, and 1 (normal) strength. A third nutrient study

examined the constituents contained in MS basal medium and their effects

on root and plant development. The experimental treatments included: 1)

macro and micro inorganic salts plus the organic components, 2) macro

and micro inorganic salts, 3) macro inorganic salts, 4) N, phosphorus

(P), and potassium (K), and 5) no nutrients (control).

Experiments were repeated twice and results were combined in each

of the three nutritional studies. All gel treatments used in the

nutrition studies were supplemented with 87 mM sucrose. The final root

and plant formation was determined at 28 days in the MS basal medium

concentration experiments and at 21 days in the other two nutrition

studies. A root of at least 2 mm was counted as "root formation", while

a root plus a well organized shoot tip, which later formed leaves (Fig.

4-1), was counted as a "differentiated plant" for all experiments which

follow. The rate and uniformity of embryo development was determined by

a modification of the method of Gerson and Honma (1978). Root or

plantlet formation Sum (Days to development)(No. developed)/Total

number developed. Root and plant formation were determined each day

through day 10, then at 12, 14, 16, 20, and 21 and/or 28 days after

placing the embryos in the gel.

The nutrient components and concentrations were examined by

comparing different media (media comparison study) on four plates, each

with ten embryos. In the studies which examined the concentrations of

MS basal medium, four replicates with 10 and 11 embryo samples each were

used in experiments 1 and 2, respectively. In the studies which

determined the effect of constituents in MS medium, every treatment in

the first experiment had four plates containing 12 embryos each, while

Fig. 4-1. Plant conversion of embryos.

Examples of embryos with roots and developing shoot apexes.
These were considered as able to form plants at this stage
of development and counted thusly.



in the second experiment, each treatment had two replicates with 12

embryos each.

Carbohydrate studies

Fructose, galactose, glucose, maltose, and sucrose sugars were

compared in two experiments and data were combined to determine which

carbohydrates) led to faster and/or greater root, shoot, and plant

production. Each carbohydrate source was tested at 47 and 93 mM

concentrations. The results from both carbohydrate concentrations were

combined and analyzed to distinguish differences between the five

carbohydrates. Murashige and Skoog basal medium was supplemented with

vitamins and incorporated in the gel for all the carbohydrate treatments

(Table A-l). Data for each experiment were an average of four blocks of

ten embryos.

Another study was conducted two times and data were combined to

define the fructose and sucrose concentrations most conducive for

optimum early seedling growth, and for rapid and complete embryo to

plant conversion. The concentrations in both experiments were tested at

0, 23, 47, 93, 186, and 374 mM sucrose or fructose. An additional

treatment concentration of 748 mM fructose was included in the first

experiment, while in the second experiment, a 12 mM sucrose

concentration was included. The first experiment had four replications,

the second five replications. Thirteen embryo samples were evaluated

per replication for the first experiment and ten in the second for each

treatment. The rate and uniformity of root and plant formation were

determined. In addition, seedling growth was measured by counting the

nodes on each plant, as indicated by an expanded leaf, and plant fresh

and dry weights.

Hormone studies

The basal medium of the gel used in these studies contained half

strength MS with no organic additives except for 47 mM (1.6% wt/vol)

sucrose. One study evaluated the effects of gibberellic acid (GA)

(GA4,/7 and GA3) at 0, 0.01, 0.1, 1.0, 10.0, and 100.0 pM on the

percentage and rate which roots and plants formed. The gel was

initially mixed with basal medium at 5% wt/vol. The GA compounds were

filter-sterilized with a 45 micron Nalgene syringe filter. Next, the GA

treatment solutions (GA plus sterilized distilled water) were added and

mixed with the gel basal medium to obtain the final GA concentrations

and 2.5% wt/vol gel concentration. The experiment was replicated four

times with ten embryos per replication.

The effect of naphthalene acetic acid (NAA) on the rate and

percentage root and plant conversion was tested at the 0. 0.1, 0.2, 0.5,

0.75, 1.0, 2.0 and 5.0 pM levels. The gel additives (basal medium plus

sucrose) used in this study were the same as those incorporated in the

gel in the GA study. Each treatment had four replications with ten

embryo samples.

Statistical Analysis

All studies were analyzed as a completely randomized design. When

experiments were combined, they were analyzed as separate blocks. All

data that are reported as a percentage were arc sine transformed. Least

significant differences and regression analyses were used to detect

differences between treatments which evaluated quantitative

measurements. Studies which compared means from qualitative treatments

were separated using Duncan's multiple range test.


Results and Discussion

Nutritional Studies

In the studies which compared various nutrient media (Table A-2),

the greatest root, shoot, and plant formation was obtained with both

Gamborg (B5) and MS media (Table 4-1), compared to either White's medium

or Hoagland's solution. A major difference in the composition of the

four media tested was the N content with White's, Gamborg's, and MS's

media, and Hoagland's solution containing 2, 30, 60 and 15 mM total N,

respectively (Table A-2).

Thus, the N quantity appeared to be of prime importance for sweet

potato plant production from somatic embryos. Between 30 and 60 mM N

was adequate for plant conversion since Gamborg medium contained

approximately 30 mM and MS medium contained 60 mM, and both facilitated

plant production (Fig. 4-2). In comparison, Hoagland solution contained

15 mM and White's medium 2 mM N which resulted in the production of only

a few weak plants. Although the final percentage root, shoot, and plant

production from embryos were affected by nutrient formulation, the speed

and uniformity were not (Table 4-1).

The following studies were conducted to evaluate concentration

levels and the nutrient components of MS medium to determine which were

needed for optimum plant development. The MS medium was used in the

following studies since this was one of the media which promoted the

most root, shoot, and plant formation in sweet potato (Table 4-1). In

addition, MS contains a substantial quantity of ammonium which may be

critical for embryo to plant conversion. The ammonium form of N has

been associated with improved production and quality of soybean somatic

Table 4-1.

Media comparison study. Effect of various media (nutrient
components and concentrations) on the percentage and mean
days to root, shoot, and plant formation when embryos were
placed in hydroxyethyl cellulose gel.z

Media? Root Shoot Plant

--------------- Final percentage response" ---------------
Gamborg (B5) 68av 26a 25a

Hoagland 43b 3b lb

Murashige &
Skoog 59ab 29a 24a

White 48b 3b 3b

----------------- Mean days to response ------------------

Gamborg (Bs) 8.7a 15.3a 13.8a

Hoagland 9.2a 16.0a 16.0a

Murashige &
Skoog 9.5a 13.6a 13.8a

White 7.4a 15.Oa 15.Oa

ZData combined from two experiments.

YMedia formulations contained only inorganic salts.

XEmbryos placed in a gel treatment containing 87 mM
yielded no roots, shoots, or plants.

sucrose (control)

"Final percentage roots, shoots, or plants obtained from embryos were
determined at 21 days. Significance determined after arc sine

VMeans separated within columns using Duncan's multiple range test, 5%

Fig. 4-2. Plant production and growth as influenced by nutrient media
and media concentration of Murashige and Skoog from embryos
placed in hydroxyethyl cellulose gel.

A. Plant production and growth were greatest when MS and
Gamborg nutrient formulations were incorporated in the
gel compared with White's medium and Hoagland's

B. Half and full-strength concentrations of MS resulted
in the greatest plant production.








embryos as measured by improved plantlet production (Stuart and

Strickland, 1984a).

The number of embryos which formed roots was similar regardless of

the MS basal medium concentration (Fig. 4-3). The exception was when

the MS medium was excluded (control treatment) from the gel which

resulted in no root production. It has been suggested that the gel

material supplies nutrients for improved germination and early seedling

growth of zygotic seeded species (Schultheis et al., 1988; Taylor,

1986). This did not occur with the somatic embryos in these studies

since no root formation (germination) occurred when the gel lacked

nutrient additives.

The final percentage of plants produced in gels containing half-

strength MS basal medium were comparable to those produced in the full-

strength concentration (Fig. 4-3). However, as the concentration of the

basal medium was reduced to quarter-strength, so was the percentage of

plants produced compared with basal medium at either half or full-

strength concentrations. No plants were produced in gels without MS


This study gives further support to the idea that N is of primary

importance for optimum embryo to plant conversion. As with MS and

Gamborg media, full-strength and half-strength MS medium contained 60

and 30 mM N respectively, and led to equivalent percentage plant

production, while quarter and eighth-strength (15 and 7.5 mM N,

respectively) resulted in a reduced number of plants (Fig. 4-2). The

mean days for root and plant production were similar regardless of MS

concentration (Table 4-2).

0-- Root




0 0.125 0.25 0.5 1.0


Effect of concentrations of Murashige and Skoog medium
standard error on the percentage root and plant production at
28 days. Data combined from two experiments.

Fig. 4-3.

S------- 0 Plant

Table 4-2. MS concentration study. Effect of Murashige and Skoog (MS)
concentrations on mean days to root and plant formation when
embryos were placed in hydroxyethyl cellulose gel.z

Mean days to Mean days to
MS concentration root formation plant formation

0 -y

0.125 9.2x

0.25 9.8 18.3w

0.5 10.3 19.5

1.0 9.7 15.2

zData combined from two experiments. Final percentage roots and plants
obtained from embryos were determined after 21 days. Significance
determined after arc sine transformation.

YNo roots or too few plants formed.

XLSD (0.05) for comparing mean days to root formation 3.2 days.

WLSD (0.05) for comparing mean days to plant formation 6.3 days.


A third nutrition experiment was conducted to determine which

constituents in the medium were needed for plant development. The

greatest percentage of embryos rooted when macro nutrients were added to

the gel, while root formation was less when N, P, and K were the only

elements incorporated in the gel (Table 4-3). Shoot production and thus

plant production was best with and without organic constituents, while

the elimination of micronutrients reduced the formation of plants. Most

protocols used for plant production from somatic embryos have used a

basal medium recipe which included vitamins (Lazzeri et al., 1987; Ranch

et al., 1986; Gupta and Durzan, 1986; etc). However, vitamins were not

needed to optimize plant production from somatic embryos of sweet


Gel treatments which had no nutrient amendments or contained only

N, P, and K had no shoot or plant production (Table 4-3). The inclusion

of both the macro and micro inorganic salts were needed for optimum

plant production in sweet potato. These essential elements were most

likely required since the nutritional reserves that are typically found

in seeds were probably much lower or lacking in somatic embryos as a

result of underdeveloped cotyledons and no endosperm. The lack of well-

developed cotyledons in somatic embryos compared with zygotic embryos

has been reported for sweet potato (Chee and Cantliffe, 1988). A

comparison of rapeseed somatic embryos with zygotic embryos revealed

that proportionally lower quantities of the same proteins were found in

somatic versus zygotic embryos (Crouch and Sussex, 1981).

No difference in the rate of root, shoot, or plant production

could be attributed to the treatments in the constituent study or in any

of the other nutrition studies. The rate of plant formation is probably